License cleanup: add SPDX GPL-2.0 license identifier to files with no license
Many source files in the tree are missing licensing information, which
makes it harder for compliance tools to determine the correct license.
By default all files without license information are under the default
license of the kernel, which is GPL version 2.
Update the files which contain no license information with the 'GPL-2.0'
SPDX license identifier. The SPDX identifier is a legally binding
shorthand, which can be used instead of the full boiler plate text.
This patch is based on work done by Thomas Gleixner and Kate Stewart and
Philippe Ombredanne.
How this work was done:
Patches were generated and checked against linux-4.14-rc6 for a subset of
the use cases:
- file had no licensing information it it.
- file was a */uapi/* one with no licensing information in it,
- file was a */uapi/* one with existing licensing information,
Further patches will be generated in subsequent months to fix up cases
where non-standard license headers were used, and references to license
had to be inferred by heuristics based on keywords.
The analysis to determine which SPDX License Identifier to be applied to
a file was done in a spreadsheet of side by side results from of the
output of two independent scanners (ScanCode & Windriver) producing SPDX
tag:value files created by Philippe Ombredanne. Philippe prepared the
base worksheet, and did an initial spot review of a few 1000 files.
The 4.13 kernel was the starting point of the analysis with 60,537 files
assessed. Kate Stewart did a file by file comparison of the scanner
results in the spreadsheet to determine which SPDX license identifier(s)
to be applied to the file. She confirmed any determination that was not
immediately clear with lawyers working with the Linux Foundation.
Criteria used to select files for SPDX license identifier tagging was:
- Files considered eligible had to be source code files.
- Make and config files were included as candidates if they contained >5
lines of source
- File already had some variant of a license header in it (even if <5
lines).
All documentation files were explicitly excluded.
The following heuristics were used to determine which SPDX license
identifiers to apply.
- when both scanners couldn't find any license traces, file was
considered to have no license information in it, and the top level
COPYING file license applied.
For non */uapi/* files that summary was:
SPDX license identifier # files
---------------------------------------------------|-------
GPL-2.0 11139
and resulted in the first patch in this series.
If that file was a */uapi/* path one, it was "GPL-2.0 WITH
Linux-syscall-note" otherwise it was "GPL-2.0". Results of that was:
SPDX license identifier # files
---------------------------------------------------|-------
GPL-2.0 WITH Linux-syscall-note 930
and resulted in the second patch in this series.
- if a file had some form of licensing information in it, and was one
of the */uapi/* ones, it was denoted with the Linux-syscall-note if
any GPL family license was found in the file or had no licensing in
it (per prior point). Results summary:
SPDX license identifier # files
---------------------------------------------------|------
GPL-2.0 WITH Linux-syscall-note 270
GPL-2.0+ WITH Linux-syscall-note 169
((GPL-2.0 WITH Linux-syscall-note) OR BSD-2-Clause) 21
((GPL-2.0 WITH Linux-syscall-note) OR BSD-3-Clause) 17
LGPL-2.1+ WITH Linux-syscall-note 15
GPL-1.0+ WITH Linux-syscall-note 14
((GPL-2.0+ WITH Linux-syscall-note) OR BSD-3-Clause) 5
LGPL-2.0+ WITH Linux-syscall-note 4
LGPL-2.1 WITH Linux-syscall-note 3
((GPL-2.0 WITH Linux-syscall-note) OR MIT) 3
((GPL-2.0 WITH Linux-syscall-note) AND MIT) 1
and that resulted in the third patch in this series.
- when the two scanners agreed on the detected license(s), that became
the concluded license(s).
- when there was disagreement between the two scanners (one detected a
license but the other didn't, or they both detected different
licenses) a manual inspection of the file occurred.
- In most cases a manual inspection of the information in the file
resulted in a clear resolution of the license that should apply (and
which scanner probably needed to revisit its heuristics).
- When it was not immediately clear, the license identifier was
confirmed with lawyers working with the Linux Foundation.
- If there was any question as to the appropriate license identifier,
the file was flagged for further research and to be revisited later
in time.
In total, over 70 hours of logged manual review was done on the
spreadsheet to determine the SPDX license identifiers to apply to the
source files by Kate, Philippe, Thomas and, in some cases, confirmation
by lawyers working with the Linux Foundation.
Kate also obtained a third independent scan of the 4.13 code base from
FOSSology, and compared selected files where the other two scanners
disagreed against that SPDX file, to see if there was new insights. The
Windriver scanner is based on an older version of FOSSology in part, so
they are related.
Thomas did random spot checks in about 500 files from the spreadsheets
for the uapi headers and agreed with SPDX license identifier in the
files he inspected. For the non-uapi files Thomas did random spot checks
in about 15000 files.
In initial set of patches against 4.14-rc6, 3 files were found to have
copy/paste license identifier errors, and have been fixed to reflect the
correct identifier.
Additionally Philippe spent 10 hours this week doing a detailed manual
inspection and review of the 12,461 patched files from the initial patch
version early this week with:
- a full scancode scan run, collecting the matched texts, detected
license ids and scores
- reviewing anything where there was a license detected (about 500+
files) to ensure that the applied SPDX license was correct
- reviewing anything where there was no detection but the patch license
was not GPL-2.0 WITH Linux-syscall-note to ensure that the applied
SPDX license was correct
This produced a worksheet with 20 files needing minor correction. This
worksheet was then exported into 3 different .csv files for the
different types of files to be modified.
These .csv files were then reviewed by Greg. Thomas wrote a script to
parse the csv files and add the proper SPDX tag to the file, in the
format that the file expected. This script was further refined by Greg
based on the output to detect more types of files automatically and to
distinguish between header and source .c files (which need different
comment types.) Finally Greg ran the script using the .csv files to
generate the patches.
Reviewed-by: Kate Stewart <kstewart@linuxfoundation.org>
Reviewed-by: Philippe Ombredanne <pombredanne@nexb.com>
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2017-11-01 14:07:57 +00:00
|
|
|
/* SPDX-License-Identifier: GPL-2.0 */
|
2005-04-16 22:20:36 +00:00
|
|
|
#ifndef _LINUX_SCHED_H
|
|
|
|
#define _LINUX_SCHED_H
|
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
/*
|
|
|
|
* Define 'struct task_struct' and provide the main scheduler
|
|
|
|
* APIs (schedule(), wakeup variants, etc.)
|
|
|
|
*/
|
2006-04-26 23:12:56 +00:00
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
#include <uapi/linux/sched.h>
|
2014-01-27 22:15:37 +00:00
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
#include <asm/current.h>
|
2005-04-16 22:20:36 +00:00
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
#include <linux/pid.h>
|
2005-04-16 22:20:36 +00:00
|
|
|
#include <linux/sem.h>
|
shm: make exit_shm work proportional to task activity
This is small set of patches our team has had kicking around for a few
versions internally that fixes tasks getting hung on shm_exit when there
are many threads hammering it at once.
Anton wrote a simple test to cause the issue:
http://ozlabs.org/~anton/junkcode/bust_shm_exit.c
Before applying this patchset, this test code will cause either hanging
tracebacks or pthread out of memory errors.
After this patchset, it will still produce output like:
root@somehost:~# ./bust_shm_exit 1024 160
...
INFO: rcu_sched detected stalls on CPUs/tasks: {} (detected by 116, t=2111 jiffies, g=241, c=240, q=7113)
INFO: Stall ended before state dump start
...
But the task will continue to run along happily, so we consider this an
improvement over hanging, even if it's a bit noisy.
This patch (of 3):
exit_shm obtains the ipc_ns shm rwsem for write and holds it while it
walks every shared memory segment in the namespace. Thus the amount of
work is related to the number of shm segments in the namespace not the
number of segments that might need to be cleaned.
In addition, this occurs after the task has been notified the thread has
exited, so the number of tasks waiting for the ns shm rwsem can grow
without bound until memory is exausted.
Add a list to the task struct of all shmids allocated by this task. Init
the list head in copy_process. Use the ns->rwsem for locking. Add
segments after id is added, remove before removing from id.
On unshare of NEW_IPCNS orphan any ids as if the task had exited, similar
to handling of semaphore undo.
I chose a define for the init sequence since its a simple list init,
otherwise it would require a function call to avoid include loops between
the semaphore code and the task struct. Converting the list_del to
list_del_init for the unshare cases would remove the exit followed by
init, but I left it blow up if not inited.
Signed-off-by: Milton Miller <miltonm@bga.com>
Signed-off-by: Jack Miller <millerjo@us.ibm.com>
Cc: Davidlohr Bueso <davidlohr@hp.com>
Cc: Manfred Spraul <manfred@colorfullife.com>
Cc: Anton Blanchard <anton@samba.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2014-08-08 21:23:19 +00:00
|
|
|
#include <linux/shm.h>
|
2017-02-06 21:06:35 +00:00
|
|
|
#include <linux/kcov.h>
|
|
|
|
#include <linux/mutex.h>
|
|
|
|
#include <linux/plist.h>
|
|
|
|
#include <linux/hrtimer.h>
|
2005-04-16 22:20:36 +00:00
|
|
|
#include <linux/seccomp.h>
|
2017-02-06 21:06:35 +00:00
|
|
|
#include <linux/nodemask.h>
|
2017-02-04 00:27:20 +00:00
|
|
|
#include <linux/rcupdate.h>
|
2019-01-18 12:27:29 +00:00
|
|
|
#include <linux/refcount.h>
|
2006-04-25 13:54:40 +00:00
|
|
|
#include <linux/resource.h>
|
2008-01-25 20:08:34 +00:00
|
|
|
#include <linux/latencytop.h>
|
2017-02-06 21:06:35 +00:00
|
|
|
#include <linux/sched/prio.h>
|
|
|
|
#include <linux/signal_types.h>
|
|
|
|
#include <linux/mm_types_task.h>
|
|
|
|
#include <linux/task_io_accounting.h>
|
rseq: Introduce restartable sequences system call
Expose a new system call allowing each thread to register one userspace
memory area to be used as an ABI between kernel and user-space for two
purposes: user-space restartable sequences and quick access to read the
current CPU number value from user-space.
* Restartable sequences (per-cpu atomics)
Restartables sequences allow user-space to perform update operations on
per-cpu data without requiring heavy-weight atomic operations.
The restartable critical sections (percpu atomics) work has been started
by Paul Turner and Andrew Hunter. It lets the kernel handle restart of
critical sections. [1] [2] The re-implementation proposed here brings a
few simplifications to the ABI which facilitates porting to other
architectures and speeds up the user-space fast path.
Here are benchmarks of various rseq use-cases.
Test hardware:
arm32: ARMv7 Processor rev 4 (v7l) "Cubietruck", 2-core
x86-64: Intel E5-2630 v3@2.40GHz, 16-core, hyperthreading
The following benchmarks were all performed on a single thread.
* Per-CPU statistic counter increment
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 344.0 31.4 11.0
x86-64: 15.3 2.0 7.7
* LTTng-UST: write event 32-bit header, 32-bit payload into tracer
per-cpu buffer
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 2502.0 2250.0 1.1
x86-64: 117.4 98.0 1.2
* liburcu percpu: lock-unlock pair, dereference, read/compare word
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 751.0 128.5 5.8
x86-64: 53.4 28.6 1.9
* jemalloc memory allocator adapted to use rseq
Using rseq with per-cpu memory pools in jemalloc at Facebook (based on
rseq 2016 implementation):
The production workload response-time has 1-2% gain avg. latency, and
the P99 overall latency drops by 2-3%.
* Reading the current CPU number
Speeding up reading the current CPU number on which the caller thread is
running is done by keeping the current CPU number up do date within the
cpu_id field of the memory area registered by the thread. This is done
by making scheduler preemption set the TIF_NOTIFY_RESUME flag on the
current thread. Upon return to user-space, a notify-resume handler
updates the current CPU value within the registered user-space memory
area. User-space can then read the current CPU number directly from
memory.
Keeping the current cpu id in a memory area shared between kernel and
user-space is an improvement over current mechanisms available to read
the current CPU number, which has the following benefits over
alternative approaches:
- 35x speedup on ARM vs system call through glibc
- 20x speedup on x86 compared to calling glibc, which calls vdso
executing a "lsl" instruction,
- 14x speedup on x86 compared to inlined "lsl" instruction,
- Unlike vdso approaches, this cpu_id value can be read from an inline
assembly, which makes it a useful building block for restartable
sequences.
- The approach of reading the cpu id through memory mapping shared
between kernel and user-space is portable (e.g. ARM), which is not the
case for the lsl-based x86 vdso.
On x86, yet another possible approach would be to use the gs segment
selector to point to user-space per-cpu data. This approach performs
similarly to the cpu id cache, but it has two disadvantages: it is
not portable, and it is incompatible with existing applications already
using the gs segment selector for other purposes.
Benchmarking various approaches for reading the current CPU number:
ARMv7 Processor rev 4 (v7l)
Machine model: Cubietruck
- Baseline (empty loop): 8.4 ns
- Read CPU from rseq cpu_id: 16.7 ns
- Read CPU from rseq cpu_id (lazy register): 19.8 ns
- glibc 2.19-0ubuntu6.6 getcpu: 301.8 ns
- getcpu system call: 234.9 ns
x86-64 Intel(R) Xeon(R) CPU E5-2630 v3 @ 2.40GHz:
- Baseline (empty loop): 0.8 ns
- Read CPU from rseq cpu_id: 0.8 ns
- Read CPU from rseq cpu_id (lazy register): 0.8 ns
- Read using gs segment selector: 0.8 ns
- "lsl" inline assembly: 13.0 ns
- glibc 2.19-0ubuntu6 getcpu: 16.6 ns
- getcpu system call: 53.9 ns
- Speed (benchmark taken on v8 of patchset)
Running 10 runs of hackbench -l 100000 seems to indicate, contrary to
expectations, that enabling CONFIG_RSEQ slightly accelerates the
scheduler:
Configuration: 2 sockets * 8-core Intel(R) Xeon(R) CPU E5-2630 v3 @
2.40GHz (directly on hardware, hyperthreading disabled in BIOS, energy
saving disabled in BIOS, turboboost disabled in BIOS, cpuidle.off=1
kernel parameter), with a Linux v4.6 defconfig+localyesconfig,
restartable sequences series applied.
* CONFIG_RSEQ=n
avg.: 41.37 s
std.dev.: 0.36 s
* CONFIG_RSEQ=y
avg.: 40.46 s
std.dev.: 0.33 s
- Size
On x86-64, between CONFIG_RSEQ=n/y, the text size increase of vmlinux is
567 bytes, and the data size increase of vmlinux is 5696 bytes.
[1] https://lwn.net/Articles/650333/
[2] http://www.linuxplumbersconf.org/2013/ocw/system/presentations/1695/original/LPC%20-%20PerCpu%20Atomics.pdf
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Joel Fernandes <joelaf@google.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Dave Watson <davejwatson@fb.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Chris Lameter <cl@linux.com>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Andrew Hunter <ahh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: "Paul E . McKenney" <paulmck@linux.vnet.ibm.com>
Cc: Paul Turner <pjt@google.com>
Cc: Boqun Feng <boqun.feng@gmail.com>
Cc: Josh Triplett <josh@joshtriplett.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Ben Maurer <bmaurer@fb.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: linux-api@vger.kernel.org
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/20151027235635.16059.11630.stgit@pjt-glaptop.roam.corp.google.com
Link: http://lkml.kernel.org/r/20150624222609.6116.86035.stgit@kitami.mtv.corp.google.com
Link: https://lkml.kernel.org/r/20180602124408.8430-3-mathieu.desnoyers@efficios.com
2018-06-02 12:43:54 +00:00
|
|
|
#include <linux/rseq.h>
|
2006-04-25 13:54:40 +00:00
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
/* task_struct member predeclarations (sorted alphabetically): */
|
2017-02-03 21:01:58 +00:00
|
|
|
struct audit_context;
|
|
|
|
struct backing_dev_info;
|
2010-02-23 07:55:42 +00:00
|
|
|
struct bio_list;
|
2011-03-08 12:19:51 +00:00
|
|
|
struct blk_plug;
|
2019-06-04 11:14:55 +00:00
|
|
|
struct capture_control;
|
2017-02-03 21:01:58 +00:00
|
|
|
struct cfs_rq;
|
|
|
|
struct fs_struct;
|
|
|
|
struct futex_pi_state;
|
|
|
|
struct io_context;
|
|
|
|
struct mempolicy;
|
2015-05-12 12:29:38 +00:00
|
|
|
struct nameidata;
|
2017-02-03 21:01:58 +00:00
|
|
|
struct nsproxy;
|
|
|
|
struct perf_event_context;
|
|
|
|
struct pid_namespace;
|
|
|
|
struct pipe_inode_info;
|
|
|
|
struct rcu_node;
|
|
|
|
struct reclaim_state;
|
|
|
|
struct robust_list_head;
|
2019-06-04 11:14:55 +00:00
|
|
|
struct root_domain;
|
|
|
|
struct rq;
|
2017-02-03 21:01:58 +00:00
|
|
|
struct sched_attr;
|
|
|
|
struct sched_param;
|
2007-07-09 16:52:00 +00:00
|
|
|
struct seq_file;
|
2017-02-03 21:01:58 +00:00
|
|
|
struct sighand_struct;
|
|
|
|
struct signal_struct;
|
|
|
|
struct task_delay_info;
|
2007-10-15 15:00:14 +00:00
|
|
|
struct task_group;
|
2005-04-16 22:20:36 +00:00
|
|
|
|
2005-09-29 22:18:21 +00:00
|
|
|
/*
|
|
|
|
* Task state bitmask. NOTE! These bits are also
|
|
|
|
* encoded in fs/proc/array.c: get_task_state().
|
|
|
|
*
|
|
|
|
* We have two separate sets of flags: task->state
|
|
|
|
* is about runnability, while task->exit_state are
|
|
|
|
* about the task exiting. Confusing, but this way
|
|
|
|
* modifying one set can't modify the other one by
|
|
|
|
* mistake.
|
|
|
|
*/
|
2017-02-06 21:06:35 +00:00
|
|
|
|
|
|
|
/* Used in tsk->state: */
|
2017-09-22 16:13:36 +00:00
|
|
|
#define TASK_RUNNING 0x0000
|
|
|
|
#define TASK_INTERRUPTIBLE 0x0001
|
|
|
|
#define TASK_UNINTERRUPTIBLE 0x0002
|
|
|
|
#define __TASK_STOPPED 0x0004
|
|
|
|
#define __TASK_TRACED 0x0008
|
2017-02-06 21:06:35 +00:00
|
|
|
/* Used in tsk->exit_state: */
|
2017-09-22 16:13:36 +00:00
|
|
|
#define EXIT_DEAD 0x0010
|
|
|
|
#define EXIT_ZOMBIE 0x0020
|
2017-02-06 21:06:35 +00:00
|
|
|
#define EXIT_TRACE (EXIT_ZOMBIE | EXIT_DEAD)
|
|
|
|
/* Used in tsk->state again: */
|
2017-09-22 16:37:28 +00:00
|
|
|
#define TASK_PARKED 0x0040
|
|
|
|
#define TASK_DEAD 0x0080
|
|
|
|
#define TASK_WAKEKILL 0x0100
|
|
|
|
#define TASK_WAKING 0x0200
|
2017-09-22 16:13:36 +00:00
|
|
|
#define TASK_NOLOAD 0x0400
|
|
|
|
#define TASK_NEW 0x0800
|
|
|
|
#define TASK_STATE_MAX 0x1000
|
2017-02-06 21:06:35 +00:00
|
|
|
|
|
|
|
/* Convenience macros for the sake of set_current_state: */
|
|
|
|
#define TASK_KILLABLE (TASK_WAKEKILL | TASK_UNINTERRUPTIBLE)
|
|
|
|
#define TASK_STOPPED (TASK_WAKEKILL | __TASK_STOPPED)
|
|
|
|
#define TASK_TRACED (TASK_WAKEKILL | __TASK_TRACED)
|
|
|
|
|
|
|
|
#define TASK_IDLE (TASK_UNINTERRUPTIBLE | TASK_NOLOAD)
|
|
|
|
|
|
|
|
/* Convenience macros for the sake of wake_up(): */
|
|
|
|
#define TASK_NORMAL (TASK_INTERRUPTIBLE | TASK_UNINTERRUPTIBLE)
|
|
|
|
|
|
|
|
/* get_task_state(): */
|
|
|
|
#define TASK_REPORT (TASK_RUNNING | TASK_INTERRUPTIBLE | \
|
|
|
|
TASK_UNINTERRUPTIBLE | __TASK_STOPPED | \
|
2017-09-22 16:37:28 +00:00
|
|
|
__TASK_TRACED | EXIT_DEAD | EXIT_ZOMBIE | \
|
|
|
|
TASK_PARKED)
|
2017-02-06 21:06:35 +00:00
|
|
|
|
|
|
|
#define task_is_traced(task) ((task->state & __TASK_TRACED) != 0)
|
|
|
|
|
|
|
|
#define task_is_stopped(task) ((task->state & __TASK_STOPPED) != 0)
|
|
|
|
|
|
|
|
#define task_is_stopped_or_traced(task) ((task->state & (__TASK_STOPPED | __TASK_TRACED)) != 0)
|
|
|
|
|
|
|
|
#define task_contributes_to_load(task) ((task->state & TASK_UNINTERRUPTIBLE) != 0 && \
|
|
|
|
(task->flags & PF_FROZEN) == 0 && \
|
|
|
|
(task->state & TASK_NOLOAD) == 0)
|
2005-04-16 22:20:36 +00:00
|
|
|
|
2014-09-24 08:18:55 +00:00
|
|
|
#ifdef CONFIG_DEBUG_ATOMIC_SLEEP
|
|
|
|
|
2018-04-30 12:51:01 +00:00
|
|
|
/*
|
|
|
|
* Special states are those that do not use the normal wait-loop pattern. See
|
|
|
|
* the comment with set_special_state().
|
|
|
|
*/
|
|
|
|
#define is_special_task_state(state) \
|
2018-06-07 09:45:49 +00:00
|
|
|
((state) & (__TASK_STOPPED | __TASK_TRACED | TASK_PARKED | TASK_DEAD))
|
2018-04-30 12:51:01 +00:00
|
|
|
|
2014-09-24 08:18:55 +00:00
|
|
|
#define __set_current_state(state_value) \
|
|
|
|
do { \
|
2018-04-30 12:51:01 +00:00
|
|
|
WARN_ON_ONCE(is_special_task_state(state_value));\
|
2014-09-24 08:18:55 +00:00
|
|
|
current->task_state_change = _THIS_IP_; \
|
|
|
|
current->state = (state_value); \
|
|
|
|
} while (0)
|
2018-04-30 12:51:01 +00:00
|
|
|
|
2014-09-24 08:18:55 +00:00
|
|
|
#define set_current_state(state_value) \
|
|
|
|
do { \
|
2018-04-30 12:51:01 +00:00
|
|
|
WARN_ON_ONCE(is_special_task_state(state_value));\
|
2014-09-24 08:18:55 +00:00
|
|
|
current->task_state_change = _THIS_IP_; \
|
2016-10-19 13:45:27 +00:00
|
|
|
smp_store_mb(current->state, (state_value)); \
|
2014-09-24 08:18:55 +00:00
|
|
|
} while (0)
|
|
|
|
|
2018-04-30 12:51:01 +00:00
|
|
|
#define set_special_state(state_value) \
|
|
|
|
do { \
|
|
|
|
unsigned long flags; /* may shadow */ \
|
|
|
|
WARN_ON_ONCE(!is_special_task_state(state_value)); \
|
|
|
|
raw_spin_lock_irqsave(¤t->pi_lock, flags); \
|
|
|
|
current->task_state_change = _THIS_IP_; \
|
|
|
|
current->state = (state_value); \
|
|
|
|
raw_spin_unlock_irqrestore(¤t->pi_lock, flags); \
|
|
|
|
} while (0)
|
2014-09-24 08:18:55 +00:00
|
|
|
#else
|
2005-09-13 08:25:14 +00:00
|
|
|
/*
|
|
|
|
* set_current_state() includes a barrier so that the write of current->state
|
|
|
|
* is correctly serialised wrt the caller's subsequent test of whether to
|
|
|
|
* actually sleep:
|
|
|
|
*
|
2016-10-19 13:45:27 +00:00
|
|
|
* for (;;) {
|
2005-09-13 08:25:14 +00:00
|
|
|
* set_current_state(TASK_UNINTERRUPTIBLE);
|
2016-10-19 13:45:27 +00:00
|
|
|
* if (!need_sleep)
|
|
|
|
* break;
|
|
|
|
*
|
|
|
|
* schedule();
|
|
|
|
* }
|
|
|
|
* __set_current_state(TASK_RUNNING);
|
|
|
|
*
|
|
|
|
* If the caller does not need such serialisation (because, for instance, the
|
|
|
|
* condition test and condition change and wakeup are under the same lock) then
|
|
|
|
* use __set_current_state().
|
|
|
|
*
|
|
|
|
* The above is typically ordered against the wakeup, which does:
|
|
|
|
*
|
2018-04-30 12:51:01 +00:00
|
|
|
* need_sleep = false;
|
|
|
|
* wake_up_state(p, TASK_UNINTERRUPTIBLE);
|
2016-10-19 13:45:27 +00:00
|
|
|
*
|
2018-07-16 18:06:03 +00:00
|
|
|
* where wake_up_state() executes a full memory barrier before accessing the
|
|
|
|
* task state.
|
2016-10-19 13:45:27 +00:00
|
|
|
*
|
|
|
|
* Wakeup will do: if (@state & p->state) p->state = TASK_RUNNING, that is,
|
|
|
|
* once it observes the TASK_UNINTERRUPTIBLE store the waking CPU can issue a
|
|
|
|
* TASK_RUNNING store which can collide with __set_current_state(TASK_RUNNING).
|
2005-09-13 08:25:14 +00:00
|
|
|
*
|
2018-04-30 12:51:01 +00:00
|
|
|
* However, with slightly different timing the wakeup TASK_RUNNING store can
|
2018-12-03 09:05:56 +00:00
|
|
|
* also collide with the TASK_UNINTERRUPTIBLE store. Losing that store is not
|
2018-04-30 12:51:01 +00:00
|
|
|
* a problem either because that will result in one extra go around the loop
|
|
|
|
* and our @cond test will save the day.
|
2005-09-13 08:25:14 +00:00
|
|
|
*
|
2016-10-19 13:45:27 +00:00
|
|
|
* Also see the comments of try_to_wake_up().
|
2005-09-13 08:25:14 +00:00
|
|
|
*/
|
2018-04-30 12:51:01 +00:00
|
|
|
#define __set_current_state(state_value) \
|
|
|
|
current->state = (state_value)
|
|
|
|
|
|
|
|
#define set_current_state(state_value) \
|
|
|
|
smp_store_mb(current->state, (state_value))
|
|
|
|
|
|
|
|
/*
|
|
|
|
* set_special_state() should be used for those states when the blocking task
|
|
|
|
* can not use the regular condition based wait-loop. In that case we must
|
|
|
|
* serialize against wakeups such that any possible in-flight TASK_RUNNING stores
|
|
|
|
* will not collide with our state change.
|
|
|
|
*/
|
|
|
|
#define set_special_state(state_value) \
|
|
|
|
do { \
|
|
|
|
unsigned long flags; /* may shadow */ \
|
|
|
|
raw_spin_lock_irqsave(¤t->pi_lock, flags); \
|
|
|
|
current->state = (state_value); \
|
|
|
|
raw_spin_unlock_irqrestore(¤t->pi_lock, flags); \
|
|
|
|
} while (0)
|
|
|
|
|
2014-09-24 08:18:55 +00:00
|
|
|
#endif
|
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
/* Task command name length: */
|
|
|
|
#define TASK_COMM_LEN 16
|
2005-04-16 22:20:36 +00:00
|
|
|
|
|
|
|
extern void scheduler_tick(void);
|
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
#define MAX_SCHEDULE_TIMEOUT LONG_MAX
|
|
|
|
|
|
|
|
extern long schedule_timeout(long timeout);
|
|
|
|
extern long schedule_timeout_interruptible(long timeout);
|
|
|
|
extern long schedule_timeout_killable(long timeout);
|
|
|
|
extern long schedule_timeout_uninterruptible(long timeout);
|
|
|
|
extern long schedule_timeout_idle(long timeout);
|
2005-04-16 22:20:36 +00:00
|
|
|
asmlinkage void schedule(void);
|
2011-03-21 11:09:35 +00:00
|
|
|
extern void schedule_preempt_disabled(void);
|
2005-04-16 22:20:36 +00:00
|
|
|
|
2016-10-28 16:58:10 +00:00
|
|
|
extern int __must_check io_schedule_prepare(void);
|
|
|
|
extern void io_schedule_finish(int token);
|
sched: Prevent recursion in io_schedule()
io_schedule() calls blk_flush_plug() which, depending on the
contents of current->plug, can initiate arbitrary blk-io requests.
Note that this contrasts with blk_schedule_flush_plug() which requires
all non-trivial work to be handed off to a separate thread.
This makes it possible for io_schedule() to recurse, and initiating
block requests could possibly call mempool_alloc() which, in times of
memory pressure, uses io_schedule().
Apart from any stack usage issues, io_schedule() will not behave
correctly when called recursively as delayacct_blkio_start() does
not allow for repeated calls.
So:
- use ->in_iowait to detect recursion. Set it earlier, and restore
it to the old value.
- move the call to "raw_rq" after the call to blk_flush_plug().
As this is some sort of per-cpu thing, we want some chance that
we are on the right CPU
- When io_schedule() is called recurively, use blk_schedule_flush_plug()
which cannot further recurse.
- as this makes io_schedule() a lot more complex and as io_schedule()
must match io_schedule_timeout(), but all the changes in io_schedule_timeout()
and make io_schedule a simple wrapper for that.
Signed-off-by: NeilBrown <neilb@suse.de>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
[ Moved the now rudimentary io_schedule() into sched.h. ]
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Tony Battersby <tonyb@cybernetics.com>
Link: http://lkml.kernel.org/r/20150213162600.059fffb2@notabene.brown
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-02-13 04:49:17 +00:00
|
|
|
extern long io_schedule_timeout(long timeout);
|
2016-10-28 16:58:10 +00:00
|
|
|
extern void io_schedule(void);
|
sched: Prevent recursion in io_schedule()
io_schedule() calls blk_flush_plug() which, depending on the
contents of current->plug, can initiate arbitrary blk-io requests.
Note that this contrasts with blk_schedule_flush_plug() which requires
all non-trivial work to be handed off to a separate thread.
This makes it possible for io_schedule() to recurse, and initiating
block requests could possibly call mempool_alloc() which, in times of
memory pressure, uses io_schedule().
Apart from any stack usage issues, io_schedule() will not behave
correctly when called recursively as delayacct_blkio_start() does
not allow for repeated calls.
So:
- use ->in_iowait to detect recursion. Set it earlier, and restore
it to the old value.
- move the call to "raw_rq" after the call to blk_flush_plug().
As this is some sort of per-cpu thing, we want some chance that
we are on the right CPU
- When io_schedule() is called recurively, use blk_schedule_flush_plug()
which cannot further recurse.
- as this makes io_schedule() a lot more complex and as io_schedule()
must match io_schedule_timeout(), but all the changes in io_schedule_timeout()
and make io_schedule a simple wrapper for that.
Signed-off-by: NeilBrown <neilb@suse.de>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
[ Moved the now rudimentary io_schedule() into sched.h. ]
Cc: Jens Axboe <axboe@kernel.dk>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Tony Battersby <tonyb@cybernetics.com>
Link: http://lkml.kernel.org/r/20150213162600.059fffb2@notabene.brown
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-02-13 04:49:17 +00:00
|
|
|
|
2012-11-21 23:58:35 +00:00
|
|
|
/**
|
2017-03-07 11:48:02 +00:00
|
|
|
* struct prev_cputime - snapshot of system and user cputime
|
2012-11-21 23:58:35 +00:00
|
|
|
* @utime: time spent in user mode
|
|
|
|
* @stime: time spent in system mode
|
2015-06-30 09:30:54 +00:00
|
|
|
* @lock: protects the above two fields
|
2012-11-21 23:58:35 +00:00
|
|
|
*
|
2015-06-30 09:30:54 +00:00
|
|
|
* Stores previous user/system time values such that we can guarantee
|
|
|
|
* monotonicity.
|
2012-11-21 23:58:35 +00:00
|
|
|
*/
|
2015-06-30 09:30:54 +00:00
|
|
|
struct prev_cputime {
|
|
|
|
#ifndef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE
|
2017-02-06 21:06:35 +00:00
|
|
|
u64 utime;
|
|
|
|
u64 stime;
|
|
|
|
raw_spinlock_t lock;
|
2015-06-30 09:30:54 +00:00
|
|
|
#endif
|
2012-11-21 23:58:35 +00:00
|
|
|
};
|
|
|
|
|
timers: fix itimer/many thread hang
Overview
This patch reworks the handling of POSIX CPU timers, including the
ITIMER_PROF, ITIMER_VIRT timers and rlimit handling. It was put together
with the help of Roland McGrath, the owner and original writer of this code.
The problem we ran into, and the reason for this rework, has to do with using
a profiling timer in a process with a large number of threads. It appears
that the performance of the old implementation of run_posix_cpu_timers() was
at least O(n*3) (where "n" is the number of threads in a process) or worse.
Everything is fine with an increasing number of threads until the time taken
for that routine to run becomes the same as or greater than the tick time, at
which point things degrade rather quickly.
This patch fixes bug 9906, "Weird hang with NPTL and SIGPROF."
Code Changes
This rework corrects the implementation of run_posix_cpu_timers() to make it
run in constant time for a particular machine. (Performance may vary between
one machine and another depending upon whether the kernel is built as single-
or multiprocessor and, in the latter case, depending upon the number of
running processors.) To do this, at each tick we now update fields in
signal_struct as well as task_struct. The run_posix_cpu_timers() function
uses those fields to make its decisions.
We define a new structure, "task_cputime," to contain user, system and
scheduler times and use these in appropriate places:
struct task_cputime {
cputime_t utime;
cputime_t stime;
unsigned long long sum_exec_runtime;
};
This is included in the structure "thread_group_cputime," which is a new
substructure of signal_struct and which varies for uniprocessor versus
multiprocessor kernels. For uniprocessor kernels, it uses "task_cputime" as
a simple substructure, while for multiprocessor kernels it is a pointer:
struct thread_group_cputime {
struct task_cputime totals;
};
struct thread_group_cputime {
struct task_cputime *totals;
};
We also add a new task_cputime substructure directly to signal_struct, to
cache the earliest expiration of process-wide timers, and task_cputime also
replaces the it_*_expires fields of task_struct (used for earliest expiration
of thread timers). The "thread_group_cputime" structure contains process-wide
timers that are updated via account_user_time() and friends. In the non-SMP
case the structure is a simple aggregator; unfortunately in the SMP case that
simplicity was not achievable due to cache-line contention between CPUs (in
one measured case performance was actually _worse_ on a 16-cpu system than
the same test on a 4-cpu system, due to this contention). For SMP, the
thread_group_cputime counters are maintained as a per-cpu structure allocated
using alloc_percpu(). The timer functions update only the timer field in
the structure corresponding to the running CPU, obtained using per_cpu_ptr().
We define a set of inline functions in sched.h that we use to maintain the
thread_group_cputime structure and hide the differences between UP and SMP
implementations from the rest of the kernel. The thread_group_cputime_init()
function initializes the thread_group_cputime structure for the given task.
The thread_group_cputime_alloc() is a no-op for UP; for SMP it calls the
out-of-line function thread_group_cputime_alloc_smp() to allocate and fill
in the per-cpu structures and fields. The thread_group_cputime_free()
function, also a no-op for UP, in SMP frees the per-cpu structures. The
thread_group_cputime_clone_thread() function (also a UP no-op) for SMP calls
thread_group_cputime_alloc() if the per-cpu structures haven't yet been
allocated. The thread_group_cputime() function fills the task_cputime
structure it is passed with the contents of the thread_group_cputime fields;
in UP it's that simple but in SMP it must also safely check that tsk->signal
is non-NULL (if it is it just uses the appropriate fields of task_struct) and,
if so, sums the per-cpu values for each online CPU. Finally, the three
functions account_group_user_time(), account_group_system_time() and
account_group_exec_runtime() are used by timer functions to update the
respective fields of the thread_group_cputime structure.
Non-SMP operation is trivial and will not be mentioned further.
The per-cpu structure is always allocated when a task creates its first new
thread, via a call to thread_group_cputime_clone_thread() from copy_signal().
It is freed at process exit via a call to thread_group_cputime_free() from
cleanup_signal().
All functions that formerly summed utime/stime/sum_sched_runtime values from
from all threads in the thread group now use thread_group_cputime() to
snapshot the values in the thread_group_cputime structure or the values in
the task structure itself if the per-cpu structure hasn't been allocated.
Finally, the code in kernel/posix-cpu-timers.c has changed quite a bit.
The run_posix_cpu_timers() function has been split into a fast path and a
slow path; the former safely checks whether there are any expired thread
timers and, if not, just returns, while the slow path does the heavy lifting.
With the dedicated thread group fields, timers are no longer "rebalanced" and
the process_timer_rebalance() function and related code has gone away. All
summing loops are gone and all code that used them now uses the
thread_group_cputime() inline. When process-wide timers are set, the new
task_cputime structure in signal_struct is used to cache the earliest
expiration; this is checked in the fast path.
Performance
The fix appears not to add significant overhead to existing operations. It
generally performs the same as the current code except in two cases, one in
which it performs slightly worse (Case 5 below) and one in which it performs
very significantly better (Case 2 below). Overall it's a wash except in those
two cases.
I've since done somewhat more involved testing on a dual-core Opteron system.
Case 1: With no itimer running, for a test with 100,000 threads, the fixed
kernel took 1428.5 seconds, 513 seconds more than the unfixed system,
all of which was spent in the system. There were twice as many
voluntary context switches with the fix as without it.
Case 2: With an itimer running at .01 second ticks and 4000 threads (the most
an unmodified kernel can handle), the fixed kernel ran the test in
eight percent of the time (5.8 seconds as opposed to 70 seconds) and
had better tick accuracy (.012 seconds per tick as opposed to .023
seconds per tick).
Case 3: A 4000-thread test with an initial timer tick of .01 second and an
interval of 10,000 seconds (i.e. a timer that ticks only once) had
very nearly the same performance in both cases: 6.3 seconds elapsed
for the fixed kernel versus 5.5 seconds for the unfixed kernel.
With fewer threads (eight in these tests), the Case 1 test ran in essentially
the same time on both the modified and unmodified kernels (5.2 seconds versus
5.8 seconds). The Case 2 test ran in about the same time as well, 5.9 seconds
versus 5.4 seconds but again with much better tick accuracy, .013 seconds per
tick versus .025 seconds per tick for the unmodified kernel.
Since the fix affected the rlimit code, I also tested soft and hard CPU limits.
Case 4: With a hard CPU limit of 20 seconds and eight threads (and an itimer
running), the modified kernel was very slightly favored in that while
it killed the process in 19.997 seconds of CPU time (5.002 seconds of
wall time), only .003 seconds of that was system time, the rest was
user time. The unmodified kernel killed the process in 20.001 seconds
of CPU (5.014 seconds of wall time) of which .016 seconds was system
time. Really, though, the results were too close to call. The results
were essentially the same with no itimer running.
Case 5: With a soft limit of 20 seconds and a hard limit of 2000 seconds
(where the hard limit would never be reached) and an itimer running,
the modified kernel exhibited worse tick accuracy than the unmodified
kernel: .050 seconds/tick versus .028 seconds/tick. Otherwise,
performance was almost indistinguishable. With no itimer running this
test exhibited virtually identical behavior and times in both cases.
In times past I did some limited performance testing. those results are below.
On a four-cpu Opteron system without this fix, a sixteen-thread test executed
in 3569.991 seconds, of which user was 3568.435s and system was 1.556s. On
the same system with the fix, user and elapsed time were about the same, but
system time dropped to 0.007 seconds. Performance with eight, four and one
thread were comparable. Interestingly, the timer ticks with the fix seemed
more accurate: The sixteen-thread test with the fix received 149543 ticks
for 0.024 seconds per tick, while the same test without the fix received 58720
for 0.061 seconds per tick. Both cases were configured for an interval of
0.01 seconds. Again, the other tests were comparable. Each thread in this
test computed the primes up to 25,000,000.
I also did a test with a large number of threads, 100,000 threads, which is
impossible without the fix. In this case each thread computed the primes only
up to 10,000 (to make the runtime manageable). System time dominated, at
1546.968 seconds out of a total 2176.906 seconds (giving a user time of
629.938s). It received 147651 ticks for 0.015 seconds per tick, still quite
accurate. There is obviously no comparable test without the fix.
Signed-off-by: Frank Mayhar <fmayhar@google.com>
Cc: Roland McGrath <roland@redhat.com>
Cc: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-09-12 16:54:39 +00:00
|
|
|
/**
|
|
|
|
* struct task_cputime - collected CPU time counts
|
2017-01-31 03:09:23 +00:00
|
|
|
* @utime: time spent in user mode, in nanoseconds
|
|
|
|
* @stime: time spent in kernel mode, in nanoseconds
|
timers: fix itimer/many thread hang
Overview
This patch reworks the handling of POSIX CPU timers, including the
ITIMER_PROF, ITIMER_VIRT timers and rlimit handling. It was put together
with the help of Roland McGrath, the owner and original writer of this code.
The problem we ran into, and the reason for this rework, has to do with using
a profiling timer in a process with a large number of threads. It appears
that the performance of the old implementation of run_posix_cpu_timers() was
at least O(n*3) (where "n" is the number of threads in a process) or worse.
Everything is fine with an increasing number of threads until the time taken
for that routine to run becomes the same as or greater than the tick time, at
which point things degrade rather quickly.
This patch fixes bug 9906, "Weird hang with NPTL and SIGPROF."
Code Changes
This rework corrects the implementation of run_posix_cpu_timers() to make it
run in constant time for a particular machine. (Performance may vary between
one machine and another depending upon whether the kernel is built as single-
or multiprocessor and, in the latter case, depending upon the number of
running processors.) To do this, at each tick we now update fields in
signal_struct as well as task_struct. The run_posix_cpu_timers() function
uses those fields to make its decisions.
We define a new structure, "task_cputime," to contain user, system and
scheduler times and use these in appropriate places:
struct task_cputime {
cputime_t utime;
cputime_t stime;
unsigned long long sum_exec_runtime;
};
This is included in the structure "thread_group_cputime," which is a new
substructure of signal_struct and which varies for uniprocessor versus
multiprocessor kernels. For uniprocessor kernels, it uses "task_cputime" as
a simple substructure, while for multiprocessor kernels it is a pointer:
struct thread_group_cputime {
struct task_cputime totals;
};
struct thread_group_cputime {
struct task_cputime *totals;
};
We also add a new task_cputime substructure directly to signal_struct, to
cache the earliest expiration of process-wide timers, and task_cputime also
replaces the it_*_expires fields of task_struct (used for earliest expiration
of thread timers). The "thread_group_cputime" structure contains process-wide
timers that are updated via account_user_time() and friends. In the non-SMP
case the structure is a simple aggregator; unfortunately in the SMP case that
simplicity was not achievable due to cache-line contention between CPUs (in
one measured case performance was actually _worse_ on a 16-cpu system than
the same test on a 4-cpu system, due to this contention). For SMP, the
thread_group_cputime counters are maintained as a per-cpu structure allocated
using alloc_percpu(). The timer functions update only the timer field in
the structure corresponding to the running CPU, obtained using per_cpu_ptr().
We define a set of inline functions in sched.h that we use to maintain the
thread_group_cputime structure and hide the differences between UP and SMP
implementations from the rest of the kernel. The thread_group_cputime_init()
function initializes the thread_group_cputime structure for the given task.
The thread_group_cputime_alloc() is a no-op for UP; for SMP it calls the
out-of-line function thread_group_cputime_alloc_smp() to allocate and fill
in the per-cpu structures and fields. The thread_group_cputime_free()
function, also a no-op for UP, in SMP frees the per-cpu structures. The
thread_group_cputime_clone_thread() function (also a UP no-op) for SMP calls
thread_group_cputime_alloc() if the per-cpu structures haven't yet been
allocated. The thread_group_cputime() function fills the task_cputime
structure it is passed with the contents of the thread_group_cputime fields;
in UP it's that simple but in SMP it must also safely check that tsk->signal
is non-NULL (if it is it just uses the appropriate fields of task_struct) and,
if so, sums the per-cpu values for each online CPU. Finally, the three
functions account_group_user_time(), account_group_system_time() and
account_group_exec_runtime() are used by timer functions to update the
respective fields of the thread_group_cputime structure.
Non-SMP operation is trivial and will not be mentioned further.
The per-cpu structure is always allocated when a task creates its first new
thread, via a call to thread_group_cputime_clone_thread() from copy_signal().
It is freed at process exit via a call to thread_group_cputime_free() from
cleanup_signal().
All functions that formerly summed utime/stime/sum_sched_runtime values from
from all threads in the thread group now use thread_group_cputime() to
snapshot the values in the thread_group_cputime structure or the values in
the task structure itself if the per-cpu structure hasn't been allocated.
Finally, the code in kernel/posix-cpu-timers.c has changed quite a bit.
The run_posix_cpu_timers() function has been split into a fast path and a
slow path; the former safely checks whether there are any expired thread
timers and, if not, just returns, while the slow path does the heavy lifting.
With the dedicated thread group fields, timers are no longer "rebalanced" and
the process_timer_rebalance() function and related code has gone away. All
summing loops are gone and all code that used them now uses the
thread_group_cputime() inline. When process-wide timers are set, the new
task_cputime structure in signal_struct is used to cache the earliest
expiration; this is checked in the fast path.
Performance
The fix appears not to add significant overhead to existing operations. It
generally performs the same as the current code except in two cases, one in
which it performs slightly worse (Case 5 below) and one in which it performs
very significantly better (Case 2 below). Overall it's a wash except in those
two cases.
I've since done somewhat more involved testing on a dual-core Opteron system.
Case 1: With no itimer running, for a test with 100,000 threads, the fixed
kernel took 1428.5 seconds, 513 seconds more than the unfixed system,
all of which was spent in the system. There were twice as many
voluntary context switches with the fix as without it.
Case 2: With an itimer running at .01 second ticks and 4000 threads (the most
an unmodified kernel can handle), the fixed kernel ran the test in
eight percent of the time (5.8 seconds as opposed to 70 seconds) and
had better tick accuracy (.012 seconds per tick as opposed to .023
seconds per tick).
Case 3: A 4000-thread test with an initial timer tick of .01 second and an
interval of 10,000 seconds (i.e. a timer that ticks only once) had
very nearly the same performance in both cases: 6.3 seconds elapsed
for the fixed kernel versus 5.5 seconds for the unfixed kernel.
With fewer threads (eight in these tests), the Case 1 test ran in essentially
the same time on both the modified and unmodified kernels (5.2 seconds versus
5.8 seconds). The Case 2 test ran in about the same time as well, 5.9 seconds
versus 5.4 seconds but again with much better tick accuracy, .013 seconds per
tick versus .025 seconds per tick for the unmodified kernel.
Since the fix affected the rlimit code, I also tested soft and hard CPU limits.
Case 4: With a hard CPU limit of 20 seconds and eight threads (and an itimer
running), the modified kernel was very slightly favored in that while
it killed the process in 19.997 seconds of CPU time (5.002 seconds of
wall time), only .003 seconds of that was system time, the rest was
user time. The unmodified kernel killed the process in 20.001 seconds
of CPU (5.014 seconds of wall time) of which .016 seconds was system
time. Really, though, the results were too close to call. The results
were essentially the same with no itimer running.
Case 5: With a soft limit of 20 seconds and a hard limit of 2000 seconds
(where the hard limit would never be reached) and an itimer running,
the modified kernel exhibited worse tick accuracy than the unmodified
kernel: .050 seconds/tick versus .028 seconds/tick. Otherwise,
performance was almost indistinguishable. With no itimer running this
test exhibited virtually identical behavior and times in both cases.
In times past I did some limited performance testing. those results are below.
On a four-cpu Opteron system without this fix, a sixteen-thread test executed
in 3569.991 seconds, of which user was 3568.435s and system was 1.556s. On
the same system with the fix, user and elapsed time were about the same, but
system time dropped to 0.007 seconds. Performance with eight, four and one
thread were comparable. Interestingly, the timer ticks with the fix seemed
more accurate: The sixteen-thread test with the fix received 149543 ticks
for 0.024 seconds per tick, while the same test without the fix received 58720
for 0.061 seconds per tick. Both cases were configured for an interval of
0.01 seconds. Again, the other tests were comparable. Each thread in this
test computed the primes up to 25,000,000.
I also did a test with a large number of threads, 100,000 threads, which is
impossible without the fix. In this case each thread computed the primes only
up to 10,000 (to make the runtime manageable). System time dominated, at
1546.968 seconds out of a total 2176.906 seconds (giving a user time of
629.938s). It received 147651 ticks for 0.015 seconds per tick, still quite
accurate. There is obviously no comparable test without the fix.
Signed-off-by: Frank Mayhar <fmayhar@google.com>
Cc: Roland McGrath <roland@redhat.com>
Cc: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-09-12 16:54:39 +00:00
|
|
|
* @sum_exec_runtime: total time spent on the CPU, in nanoseconds
|
2008-09-14 15:11:46 +00:00
|
|
|
*
|
2015-06-30 09:30:54 +00:00
|
|
|
* This structure groups together three kinds of CPU time that are tracked for
|
|
|
|
* threads and thread groups. Most things considering CPU time want to group
|
|
|
|
* these counts together and treat all three of them in parallel.
|
timers: fix itimer/many thread hang
Overview
This patch reworks the handling of POSIX CPU timers, including the
ITIMER_PROF, ITIMER_VIRT timers and rlimit handling. It was put together
with the help of Roland McGrath, the owner and original writer of this code.
The problem we ran into, and the reason for this rework, has to do with using
a profiling timer in a process with a large number of threads. It appears
that the performance of the old implementation of run_posix_cpu_timers() was
at least O(n*3) (where "n" is the number of threads in a process) or worse.
Everything is fine with an increasing number of threads until the time taken
for that routine to run becomes the same as or greater than the tick time, at
which point things degrade rather quickly.
This patch fixes bug 9906, "Weird hang with NPTL and SIGPROF."
Code Changes
This rework corrects the implementation of run_posix_cpu_timers() to make it
run in constant time for a particular machine. (Performance may vary between
one machine and another depending upon whether the kernel is built as single-
or multiprocessor and, in the latter case, depending upon the number of
running processors.) To do this, at each tick we now update fields in
signal_struct as well as task_struct. The run_posix_cpu_timers() function
uses those fields to make its decisions.
We define a new structure, "task_cputime," to contain user, system and
scheduler times and use these in appropriate places:
struct task_cputime {
cputime_t utime;
cputime_t stime;
unsigned long long sum_exec_runtime;
};
This is included in the structure "thread_group_cputime," which is a new
substructure of signal_struct and which varies for uniprocessor versus
multiprocessor kernels. For uniprocessor kernels, it uses "task_cputime" as
a simple substructure, while for multiprocessor kernels it is a pointer:
struct thread_group_cputime {
struct task_cputime totals;
};
struct thread_group_cputime {
struct task_cputime *totals;
};
We also add a new task_cputime substructure directly to signal_struct, to
cache the earliest expiration of process-wide timers, and task_cputime also
replaces the it_*_expires fields of task_struct (used for earliest expiration
of thread timers). The "thread_group_cputime" structure contains process-wide
timers that are updated via account_user_time() and friends. In the non-SMP
case the structure is a simple aggregator; unfortunately in the SMP case that
simplicity was not achievable due to cache-line contention between CPUs (in
one measured case performance was actually _worse_ on a 16-cpu system than
the same test on a 4-cpu system, due to this contention). For SMP, the
thread_group_cputime counters are maintained as a per-cpu structure allocated
using alloc_percpu(). The timer functions update only the timer field in
the structure corresponding to the running CPU, obtained using per_cpu_ptr().
We define a set of inline functions in sched.h that we use to maintain the
thread_group_cputime structure and hide the differences between UP and SMP
implementations from the rest of the kernel. The thread_group_cputime_init()
function initializes the thread_group_cputime structure for the given task.
The thread_group_cputime_alloc() is a no-op for UP; for SMP it calls the
out-of-line function thread_group_cputime_alloc_smp() to allocate and fill
in the per-cpu structures and fields. The thread_group_cputime_free()
function, also a no-op for UP, in SMP frees the per-cpu structures. The
thread_group_cputime_clone_thread() function (also a UP no-op) for SMP calls
thread_group_cputime_alloc() if the per-cpu structures haven't yet been
allocated. The thread_group_cputime() function fills the task_cputime
structure it is passed with the contents of the thread_group_cputime fields;
in UP it's that simple but in SMP it must also safely check that tsk->signal
is non-NULL (if it is it just uses the appropriate fields of task_struct) and,
if so, sums the per-cpu values for each online CPU. Finally, the three
functions account_group_user_time(), account_group_system_time() and
account_group_exec_runtime() are used by timer functions to update the
respective fields of the thread_group_cputime structure.
Non-SMP operation is trivial and will not be mentioned further.
The per-cpu structure is always allocated when a task creates its first new
thread, via a call to thread_group_cputime_clone_thread() from copy_signal().
It is freed at process exit via a call to thread_group_cputime_free() from
cleanup_signal().
All functions that formerly summed utime/stime/sum_sched_runtime values from
from all threads in the thread group now use thread_group_cputime() to
snapshot the values in the thread_group_cputime structure or the values in
the task structure itself if the per-cpu structure hasn't been allocated.
Finally, the code in kernel/posix-cpu-timers.c has changed quite a bit.
The run_posix_cpu_timers() function has been split into a fast path and a
slow path; the former safely checks whether there are any expired thread
timers and, if not, just returns, while the slow path does the heavy lifting.
With the dedicated thread group fields, timers are no longer "rebalanced" and
the process_timer_rebalance() function and related code has gone away. All
summing loops are gone and all code that used them now uses the
thread_group_cputime() inline. When process-wide timers are set, the new
task_cputime structure in signal_struct is used to cache the earliest
expiration; this is checked in the fast path.
Performance
The fix appears not to add significant overhead to existing operations. It
generally performs the same as the current code except in two cases, one in
which it performs slightly worse (Case 5 below) and one in which it performs
very significantly better (Case 2 below). Overall it's a wash except in those
two cases.
I've since done somewhat more involved testing on a dual-core Opteron system.
Case 1: With no itimer running, for a test with 100,000 threads, the fixed
kernel took 1428.5 seconds, 513 seconds more than the unfixed system,
all of which was spent in the system. There were twice as many
voluntary context switches with the fix as without it.
Case 2: With an itimer running at .01 second ticks and 4000 threads (the most
an unmodified kernel can handle), the fixed kernel ran the test in
eight percent of the time (5.8 seconds as opposed to 70 seconds) and
had better tick accuracy (.012 seconds per tick as opposed to .023
seconds per tick).
Case 3: A 4000-thread test with an initial timer tick of .01 second and an
interval of 10,000 seconds (i.e. a timer that ticks only once) had
very nearly the same performance in both cases: 6.3 seconds elapsed
for the fixed kernel versus 5.5 seconds for the unfixed kernel.
With fewer threads (eight in these tests), the Case 1 test ran in essentially
the same time on both the modified and unmodified kernels (5.2 seconds versus
5.8 seconds). The Case 2 test ran in about the same time as well, 5.9 seconds
versus 5.4 seconds but again with much better tick accuracy, .013 seconds per
tick versus .025 seconds per tick for the unmodified kernel.
Since the fix affected the rlimit code, I also tested soft and hard CPU limits.
Case 4: With a hard CPU limit of 20 seconds and eight threads (and an itimer
running), the modified kernel was very slightly favored in that while
it killed the process in 19.997 seconds of CPU time (5.002 seconds of
wall time), only .003 seconds of that was system time, the rest was
user time. The unmodified kernel killed the process in 20.001 seconds
of CPU (5.014 seconds of wall time) of which .016 seconds was system
time. Really, though, the results were too close to call. The results
were essentially the same with no itimer running.
Case 5: With a soft limit of 20 seconds and a hard limit of 2000 seconds
(where the hard limit would never be reached) and an itimer running,
the modified kernel exhibited worse tick accuracy than the unmodified
kernel: .050 seconds/tick versus .028 seconds/tick. Otherwise,
performance was almost indistinguishable. With no itimer running this
test exhibited virtually identical behavior and times in both cases.
In times past I did some limited performance testing. those results are below.
On a four-cpu Opteron system without this fix, a sixteen-thread test executed
in 3569.991 seconds, of which user was 3568.435s and system was 1.556s. On
the same system with the fix, user and elapsed time were about the same, but
system time dropped to 0.007 seconds. Performance with eight, four and one
thread were comparable. Interestingly, the timer ticks with the fix seemed
more accurate: The sixteen-thread test with the fix received 149543 ticks
for 0.024 seconds per tick, while the same test without the fix received 58720
for 0.061 seconds per tick. Both cases were configured for an interval of
0.01 seconds. Again, the other tests were comparable. Each thread in this
test computed the primes up to 25,000,000.
I also did a test with a large number of threads, 100,000 threads, which is
impossible without the fix. In this case each thread computed the primes only
up to 10,000 (to make the runtime manageable). System time dominated, at
1546.968 seconds out of a total 2176.906 seconds (giving a user time of
629.938s). It received 147651 ticks for 0.015 seconds per tick, still quite
accurate. There is obviously no comparable test without the fix.
Signed-off-by: Frank Mayhar <fmayhar@google.com>
Cc: Roland McGrath <roland@redhat.com>
Cc: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-09-12 16:54:39 +00:00
|
|
|
*/
|
|
|
|
struct task_cputime {
|
2017-02-06 21:06:35 +00:00
|
|
|
u64 utime;
|
|
|
|
u64 stime;
|
|
|
|
unsigned long long sum_exec_runtime;
|
timers: fix itimer/many thread hang
Overview
This patch reworks the handling of POSIX CPU timers, including the
ITIMER_PROF, ITIMER_VIRT timers and rlimit handling. It was put together
with the help of Roland McGrath, the owner and original writer of this code.
The problem we ran into, and the reason for this rework, has to do with using
a profiling timer in a process with a large number of threads. It appears
that the performance of the old implementation of run_posix_cpu_timers() was
at least O(n*3) (where "n" is the number of threads in a process) or worse.
Everything is fine with an increasing number of threads until the time taken
for that routine to run becomes the same as or greater than the tick time, at
which point things degrade rather quickly.
This patch fixes bug 9906, "Weird hang with NPTL and SIGPROF."
Code Changes
This rework corrects the implementation of run_posix_cpu_timers() to make it
run in constant time for a particular machine. (Performance may vary between
one machine and another depending upon whether the kernel is built as single-
or multiprocessor and, in the latter case, depending upon the number of
running processors.) To do this, at each tick we now update fields in
signal_struct as well as task_struct. The run_posix_cpu_timers() function
uses those fields to make its decisions.
We define a new structure, "task_cputime," to contain user, system and
scheduler times and use these in appropriate places:
struct task_cputime {
cputime_t utime;
cputime_t stime;
unsigned long long sum_exec_runtime;
};
This is included in the structure "thread_group_cputime," which is a new
substructure of signal_struct and which varies for uniprocessor versus
multiprocessor kernels. For uniprocessor kernels, it uses "task_cputime" as
a simple substructure, while for multiprocessor kernels it is a pointer:
struct thread_group_cputime {
struct task_cputime totals;
};
struct thread_group_cputime {
struct task_cputime *totals;
};
We also add a new task_cputime substructure directly to signal_struct, to
cache the earliest expiration of process-wide timers, and task_cputime also
replaces the it_*_expires fields of task_struct (used for earliest expiration
of thread timers). The "thread_group_cputime" structure contains process-wide
timers that are updated via account_user_time() and friends. In the non-SMP
case the structure is a simple aggregator; unfortunately in the SMP case that
simplicity was not achievable due to cache-line contention between CPUs (in
one measured case performance was actually _worse_ on a 16-cpu system than
the same test on a 4-cpu system, due to this contention). For SMP, the
thread_group_cputime counters are maintained as a per-cpu structure allocated
using alloc_percpu(). The timer functions update only the timer field in
the structure corresponding to the running CPU, obtained using per_cpu_ptr().
We define a set of inline functions in sched.h that we use to maintain the
thread_group_cputime structure and hide the differences between UP and SMP
implementations from the rest of the kernel. The thread_group_cputime_init()
function initializes the thread_group_cputime structure for the given task.
The thread_group_cputime_alloc() is a no-op for UP; for SMP it calls the
out-of-line function thread_group_cputime_alloc_smp() to allocate and fill
in the per-cpu structures and fields. The thread_group_cputime_free()
function, also a no-op for UP, in SMP frees the per-cpu structures. The
thread_group_cputime_clone_thread() function (also a UP no-op) for SMP calls
thread_group_cputime_alloc() if the per-cpu structures haven't yet been
allocated. The thread_group_cputime() function fills the task_cputime
structure it is passed with the contents of the thread_group_cputime fields;
in UP it's that simple but in SMP it must also safely check that tsk->signal
is non-NULL (if it is it just uses the appropriate fields of task_struct) and,
if so, sums the per-cpu values for each online CPU. Finally, the three
functions account_group_user_time(), account_group_system_time() and
account_group_exec_runtime() are used by timer functions to update the
respective fields of the thread_group_cputime structure.
Non-SMP operation is trivial and will not be mentioned further.
The per-cpu structure is always allocated when a task creates its first new
thread, via a call to thread_group_cputime_clone_thread() from copy_signal().
It is freed at process exit via a call to thread_group_cputime_free() from
cleanup_signal().
All functions that formerly summed utime/stime/sum_sched_runtime values from
from all threads in the thread group now use thread_group_cputime() to
snapshot the values in the thread_group_cputime structure or the values in
the task structure itself if the per-cpu structure hasn't been allocated.
Finally, the code in kernel/posix-cpu-timers.c has changed quite a bit.
The run_posix_cpu_timers() function has been split into a fast path and a
slow path; the former safely checks whether there are any expired thread
timers and, if not, just returns, while the slow path does the heavy lifting.
With the dedicated thread group fields, timers are no longer "rebalanced" and
the process_timer_rebalance() function and related code has gone away. All
summing loops are gone and all code that used them now uses the
thread_group_cputime() inline. When process-wide timers are set, the new
task_cputime structure in signal_struct is used to cache the earliest
expiration; this is checked in the fast path.
Performance
The fix appears not to add significant overhead to existing operations. It
generally performs the same as the current code except in two cases, one in
which it performs slightly worse (Case 5 below) and one in which it performs
very significantly better (Case 2 below). Overall it's a wash except in those
two cases.
I've since done somewhat more involved testing on a dual-core Opteron system.
Case 1: With no itimer running, for a test with 100,000 threads, the fixed
kernel took 1428.5 seconds, 513 seconds more than the unfixed system,
all of which was spent in the system. There were twice as many
voluntary context switches with the fix as without it.
Case 2: With an itimer running at .01 second ticks and 4000 threads (the most
an unmodified kernel can handle), the fixed kernel ran the test in
eight percent of the time (5.8 seconds as opposed to 70 seconds) and
had better tick accuracy (.012 seconds per tick as opposed to .023
seconds per tick).
Case 3: A 4000-thread test with an initial timer tick of .01 second and an
interval of 10,000 seconds (i.e. a timer that ticks only once) had
very nearly the same performance in both cases: 6.3 seconds elapsed
for the fixed kernel versus 5.5 seconds for the unfixed kernel.
With fewer threads (eight in these tests), the Case 1 test ran in essentially
the same time on both the modified and unmodified kernels (5.2 seconds versus
5.8 seconds). The Case 2 test ran in about the same time as well, 5.9 seconds
versus 5.4 seconds but again with much better tick accuracy, .013 seconds per
tick versus .025 seconds per tick for the unmodified kernel.
Since the fix affected the rlimit code, I also tested soft and hard CPU limits.
Case 4: With a hard CPU limit of 20 seconds and eight threads (and an itimer
running), the modified kernel was very slightly favored in that while
it killed the process in 19.997 seconds of CPU time (5.002 seconds of
wall time), only .003 seconds of that was system time, the rest was
user time. The unmodified kernel killed the process in 20.001 seconds
of CPU (5.014 seconds of wall time) of which .016 seconds was system
time. Really, though, the results were too close to call. The results
were essentially the same with no itimer running.
Case 5: With a soft limit of 20 seconds and a hard limit of 2000 seconds
(where the hard limit would never be reached) and an itimer running,
the modified kernel exhibited worse tick accuracy than the unmodified
kernel: .050 seconds/tick versus .028 seconds/tick. Otherwise,
performance was almost indistinguishable. With no itimer running this
test exhibited virtually identical behavior and times in both cases.
In times past I did some limited performance testing. those results are below.
On a four-cpu Opteron system without this fix, a sixteen-thread test executed
in 3569.991 seconds, of which user was 3568.435s and system was 1.556s. On
the same system with the fix, user and elapsed time were about the same, but
system time dropped to 0.007 seconds. Performance with eight, four and one
thread were comparable. Interestingly, the timer ticks with the fix seemed
more accurate: The sixteen-thread test with the fix received 149543 ticks
for 0.024 seconds per tick, while the same test without the fix received 58720
for 0.061 seconds per tick. Both cases were configured for an interval of
0.01 seconds. Again, the other tests were comparable. Each thread in this
test computed the primes up to 25,000,000.
I also did a test with a large number of threads, 100,000 threads, which is
impossible without the fix. In this case each thread computed the primes only
up to 10,000 (to make the runtime manageable). System time dominated, at
1546.968 seconds out of a total 2176.906 seconds (giving a user time of
629.938s). It received 147651 ticks for 0.015 seconds per tick, still quite
accurate. There is obviously no comparable test without the fix.
Signed-off-by: Frank Mayhar <fmayhar@google.com>
Cc: Roland McGrath <roland@redhat.com>
Cc: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-09-12 16:54:39 +00:00
|
|
|
};
|
2015-06-30 09:30:54 +00:00
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
/* Alternate field names when used on cache expirations: */
|
|
|
|
#define virt_exp utime
|
|
|
|
#define prof_exp stime
|
|
|
|
#define sched_exp sum_exec_runtime
|
timers: fix itimer/many thread hang
Overview
This patch reworks the handling of POSIX CPU timers, including the
ITIMER_PROF, ITIMER_VIRT timers and rlimit handling. It was put together
with the help of Roland McGrath, the owner and original writer of this code.
The problem we ran into, and the reason for this rework, has to do with using
a profiling timer in a process with a large number of threads. It appears
that the performance of the old implementation of run_posix_cpu_timers() was
at least O(n*3) (where "n" is the number of threads in a process) or worse.
Everything is fine with an increasing number of threads until the time taken
for that routine to run becomes the same as or greater than the tick time, at
which point things degrade rather quickly.
This patch fixes bug 9906, "Weird hang with NPTL and SIGPROF."
Code Changes
This rework corrects the implementation of run_posix_cpu_timers() to make it
run in constant time for a particular machine. (Performance may vary between
one machine and another depending upon whether the kernel is built as single-
or multiprocessor and, in the latter case, depending upon the number of
running processors.) To do this, at each tick we now update fields in
signal_struct as well as task_struct. The run_posix_cpu_timers() function
uses those fields to make its decisions.
We define a new structure, "task_cputime," to contain user, system and
scheduler times and use these in appropriate places:
struct task_cputime {
cputime_t utime;
cputime_t stime;
unsigned long long sum_exec_runtime;
};
This is included in the structure "thread_group_cputime," which is a new
substructure of signal_struct and which varies for uniprocessor versus
multiprocessor kernels. For uniprocessor kernels, it uses "task_cputime" as
a simple substructure, while for multiprocessor kernels it is a pointer:
struct thread_group_cputime {
struct task_cputime totals;
};
struct thread_group_cputime {
struct task_cputime *totals;
};
We also add a new task_cputime substructure directly to signal_struct, to
cache the earliest expiration of process-wide timers, and task_cputime also
replaces the it_*_expires fields of task_struct (used for earliest expiration
of thread timers). The "thread_group_cputime" structure contains process-wide
timers that are updated via account_user_time() and friends. In the non-SMP
case the structure is a simple aggregator; unfortunately in the SMP case that
simplicity was not achievable due to cache-line contention between CPUs (in
one measured case performance was actually _worse_ on a 16-cpu system than
the same test on a 4-cpu system, due to this contention). For SMP, the
thread_group_cputime counters are maintained as a per-cpu structure allocated
using alloc_percpu(). The timer functions update only the timer field in
the structure corresponding to the running CPU, obtained using per_cpu_ptr().
We define a set of inline functions in sched.h that we use to maintain the
thread_group_cputime structure and hide the differences between UP and SMP
implementations from the rest of the kernel. The thread_group_cputime_init()
function initializes the thread_group_cputime structure for the given task.
The thread_group_cputime_alloc() is a no-op for UP; for SMP it calls the
out-of-line function thread_group_cputime_alloc_smp() to allocate and fill
in the per-cpu structures and fields. The thread_group_cputime_free()
function, also a no-op for UP, in SMP frees the per-cpu structures. The
thread_group_cputime_clone_thread() function (also a UP no-op) for SMP calls
thread_group_cputime_alloc() if the per-cpu structures haven't yet been
allocated. The thread_group_cputime() function fills the task_cputime
structure it is passed with the contents of the thread_group_cputime fields;
in UP it's that simple but in SMP it must also safely check that tsk->signal
is non-NULL (if it is it just uses the appropriate fields of task_struct) and,
if so, sums the per-cpu values for each online CPU. Finally, the three
functions account_group_user_time(), account_group_system_time() and
account_group_exec_runtime() are used by timer functions to update the
respective fields of the thread_group_cputime structure.
Non-SMP operation is trivial and will not be mentioned further.
The per-cpu structure is always allocated when a task creates its first new
thread, via a call to thread_group_cputime_clone_thread() from copy_signal().
It is freed at process exit via a call to thread_group_cputime_free() from
cleanup_signal().
All functions that formerly summed utime/stime/sum_sched_runtime values from
from all threads in the thread group now use thread_group_cputime() to
snapshot the values in the thread_group_cputime structure or the values in
the task structure itself if the per-cpu structure hasn't been allocated.
Finally, the code in kernel/posix-cpu-timers.c has changed quite a bit.
The run_posix_cpu_timers() function has been split into a fast path and a
slow path; the former safely checks whether there are any expired thread
timers and, if not, just returns, while the slow path does the heavy lifting.
With the dedicated thread group fields, timers are no longer "rebalanced" and
the process_timer_rebalance() function and related code has gone away. All
summing loops are gone and all code that used them now uses the
thread_group_cputime() inline. When process-wide timers are set, the new
task_cputime structure in signal_struct is used to cache the earliest
expiration; this is checked in the fast path.
Performance
The fix appears not to add significant overhead to existing operations. It
generally performs the same as the current code except in two cases, one in
which it performs slightly worse (Case 5 below) and one in which it performs
very significantly better (Case 2 below). Overall it's a wash except in those
two cases.
I've since done somewhat more involved testing on a dual-core Opteron system.
Case 1: With no itimer running, for a test with 100,000 threads, the fixed
kernel took 1428.5 seconds, 513 seconds more than the unfixed system,
all of which was spent in the system. There were twice as many
voluntary context switches with the fix as without it.
Case 2: With an itimer running at .01 second ticks and 4000 threads (the most
an unmodified kernel can handle), the fixed kernel ran the test in
eight percent of the time (5.8 seconds as opposed to 70 seconds) and
had better tick accuracy (.012 seconds per tick as opposed to .023
seconds per tick).
Case 3: A 4000-thread test with an initial timer tick of .01 second and an
interval of 10,000 seconds (i.e. a timer that ticks only once) had
very nearly the same performance in both cases: 6.3 seconds elapsed
for the fixed kernel versus 5.5 seconds for the unfixed kernel.
With fewer threads (eight in these tests), the Case 1 test ran in essentially
the same time on both the modified and unmodified kernels (5.2 seconds versus
5.8 seconds). The Case 2 test ran in about the same time as well, 5.9 seconds
versus 5.4 seconds but again with much better tick accuracy, .013 seconds per
tick versus .025 seconds per tick for the unmodified kernel.
Since the fix affected the rlimit code, I also tested soft and hard CPU limits.
Case 4: With a hard CPU limit of 20 seconds and eight threads (and an itimer
running), the modified kernel was very slightly favored in that while
it killed the process in 19.997 seconds of CPU time (5.002 seconds of
wall time), only .003 seconds of that was system time, the rest was
user time. The unmodified kernel killed the process in 20.001 seconds
of CPU (5.014 seconds of wall time) of which .016 seconds was system
time. Really, though, the results were too close to call. The results
were essentially the same with no itimer running.
Case 5: With a soft limit of 20 seconds and a hard limit of 2000 seconds
(where the hard limit would never be reached) and an itimer running,
the modified kernel exhibited worse tick accuracy than the unmodified
kernel: .050 seconds/tick versus .028 seconds/tick. Otherwise,
performance was almost indistinguishable. With no itimer running this
test exhibited virtually identical behavior and times in both cases.
In times past I did some limited performance testing. those results are below.
On a four-cpu Opteron system without this fix, a sixteen-thread test executed
in 3569.991 seconds, of which user was 3568.435s and system was 1.556s. On
the same system with the fix, user and elapsed time were about the same, but
system time dropped to 0.007 seconds. Performance with eight, four and one
thread were comparable. Interestingly, the timer ticks with the fix seemed
more accurate: The sixteen-thread test with the fix received 149543 ticks
for 0.024 seconds per tick, while the same test without the fix received 58720
for 0.061 seconds per tick. Both cases were configured for an interval of
0.01 seconds. Again, the other tests were comparable. Each thread in this
test computed the primes up to 25,000,000.
I also did a test with a large number of threads, 100,000 threads, which is
impossible without the fix. In this case each thread computed the primes only
up to 10,000 (to make the runtime manageable). System time dominated, at
1546.968 seconds out of a total 2176.906 seconds (giving a user time of
629.938s). It received 147651 ticks for 0.015 seconds per tick, still quite
accurate. There is obviously no comparable test without the fix.
Signed-off-by: Frank Mayhar <fmayhar@google.com>
Cc: Roland McGrath <roland@redhat.com>
Cc: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2008-09-12 16:54:39 +00:00
|
|
|
|
2017-06-29 17:15:10 +00:00
|
|
|
enum vtime_state {
|
|
|
|
/* Task is sleeping or running in a CPU with VTIME inactive: */
|
|
|
|
VTIME_INACTIVE = 0,
|
|
|
|
/* Task runs in userspace in a CPU with VTIME active: */
|
|
|
|
VTIME_USER,
|
|
|
|
/* Task runs in kernelspace in a CPU with VTIME active: */
|
|
|
|
VTIME_SYS,
|
|
|
|
};
|
|
|
|
|
|
|
|
struct vtime {
|
|
|
|
seqcount_t seqcount;
|
|
|
|
unsigned long long starttime;
|
|
|
|
enum vtime_state state;
|
2017-06-29 17:15:11 +00:00
|
|
|
u64 utime;
|
|
|
|
u64 stime;
|
|
|
|
u64 gtime;
|
2017-06-29 17:15:10 +00:00
|
|
|
};
|
|
|
|
|
2019-06-21 08:42:02 +00:00
|
|
|
/*
|
|
|
|
* Utilization clamp constraints.
|
|
|
|
* @UCLAMP_MIN: Minimum utilization
|
|
|
|
* @UCLAMP_MAX: Maximum utilization
|
|
|
|
* @UCLAMP_CNT: Utilization clamp constraints count
|
|
|
|
*/
|
|
|
|
enum uclamp_id {
|
|
|
|
UCLAMP_MIN = 0,
|
|
|
|
UCLAMP_MAX,
|
|
|
|
UCLAMP_CNT
|
|
|
|
};
|
|
|
|
|
2005-04-16 22:20:36 +00:00
|
|
|
struct sched_info {
|
2017-02-06 10:44:12 +00:00
|
|
|
#ifdef CONFIG_SCHED_INFO
|
2017-02-06 21:06:35 +00:00
|
|
|
/* Cumulative counters: */
|
|
|
|
|
|
|
|
/* # of times we have run on this CPU: */
|
|
|
|
unsigned long pcount;
|
|
|
|
|
|
|
|
/* Time spent waiting on a runqueue: */
|
|
|
|
unsigned long long run_delay;
|
|
|
|
|
|
|
|
/* Timestamps: */
|
|
|
|
|
|
|
|
/* When did we last run on a CPU? */
|
|
|
|
unsigned long long last_arrival;
|
|
|
|
|
|
|
|
/* When were we last queued to run? */
|
|
|
|
unsigned long long last_queued;
|
2005-04-16 22:20:36 +00:00
|
|
|
|
2015-06-25 18:23:37 +00:00
|
|
|
#endif /* CONFIG_SCHED_INFO */
|
2017-02-06 10:44:12 +00:00
|
|
|
};
|
2005-04-16 22:20:36 +00:00
|
|
|
|
sched/fair: Generalize the load/util averages resolution definition
Integer metric needs fixed point arithmetic. In sched/fair, a few
metrics, e.g., weight, load, load_avg, util_avg, freq, and capacity,
may have different fixed point ranges, which makes their update and
usage error-prone.
In order to avoid the errors relating to the fixed point range, we
definie a basic fixed point range, and then formalize all metrics to
base on the basic range.
The basic range is 1024 or (1 << 10). Further, one can recursively
apply the basic range to have larger range.
Pointed out by Ben Segall, weight (visible to user, e.g., NICE-0 has
1024) and load (e.g., NICE_0_LOAD) have independent ranges, but they
must be well calibrated.
Signed-off-by: Yuyang Du <yuyang.du@intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: bsegall@google.com
Cc: dietmar.eggemann@arm.com
Cc: lizefan@huawei.com
Cc: morten.rasmussen@arm.com
Cc: pjt@google.com
Cc: umgwanakikbuti@gmail.com
Cc: vincent.guittot@linaro.org
Link: http://lkml.kernel.org/r/1459829551-21625-2-git-send-email-yuyang.du@intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2016-04-05 04:12:26 +00:00
|
|
|
/*
|
|
|
|
* Integer metrics need fixed point arithmetic, e.g., sched/fair
|
|
|
|
* has a few: load, load_avg, util_avg, freq, and capacity.
|
|
|
|
*
|
|
|
|
* We define a basic fixed point arithmetic range, and then formalize
|
|
|
|
* all these metrics based on that basic range.
|
|
|
|
*/
|
2017-02-06 21:06:35 +00:00
|
|
|
# define SCHED_FIXEDPOINT_SHIFT 10
|
|
|
|
# define SCHED_FIXEDPOINT_SCALE (1L << SCHED_FIXEDPOINT_SHIFT)
|
sched/fair: Generalize the load/util averages resolution definition
Integer metric needs fixed point arithmetic. In sched/fair, a few
metrics, e.g., weight, load, load_avg, util_avg, freq, and capacity,
may have different fixed point ranges, which makes their update and
usage error-prone.
In order to avoid the errors relating to the fixed point range, we
definie a basic fixed point range, and then formalize all metrics to
base on the basic range.
The basic range is 1024 or (1 << 10). Further, one can recursively
apply the basic range to have larger range.
Pointed out by Ben Segall, weight (visible to user, e.g., NICE-0 has
1024) and load (e.g., NICE_0_LOAD) have independent ranges, but they
must be well calibrated.
Signed-off-by: Yuyang Du <yuyang.du@intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: bsegall@google.com
Cc: dietmar.eggemann@arm.com
Cc: lizefan@huawei.com
Cc: morten.rasmussen@arm.com
Cc: pjt@google.com
Cc: umgwanakikbuti@gmail.com
Cc: vincent.guittot@linaro.org
Link: http://lkml.kernel.org/r/1459829551-21625-2-git-send-email-yuyang.du@intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2016-04-05 04:12:26 +00:00
|
|
|
|
2019-06-21 08:42:02 +00:00
|
|
|
/* Increase resolution of cpu_capacity calculations */
|
|
|
|
# define SCHED_CAPACITY_SHIFT SCHED_FIXEDPOINT_SHIFT
|
|
|
|
# define SCHED_CAPACITY_SCALE (1L << SCHED_CAPACITY_SHIFT)
|
|
|
|
|
2007-07-09 16:51:58 +00:00
|
|
|
struct load_weight {
|
2017-02-06 21:06:35 +00:00
|
|
|
unsigned long weight;
|
|
|
|
u32 inv_weight;
|
2007-07-09 16:51:58 +00:00
|
|
|
};
|
|
|
|
|
sched/fair: Add util_est on top of PELT
The util_avg signal computed by PELT is too variable for some use-cases.
For example, a big task waking up after a long sleep period will have its
utilization almost completely decayed. This introduces some latency before
schedutil will be able to pick the best frequency to run a task.
The same issue can affect task placement. Indeed, since the task
utilization is already decayed at wakeup, when the task is enqueued in a
CPU, this can result in a CPU running a big task as being temporarily
represented as being almost empty. This leads to a race condition where
other tasks can be potentially allocated on a CPU which just started to run
a big task which slept for a relatively long period.
Moreover, the PELT utilization of a task can be updated every [ms], thus
making it a continuously changing value for certain longer running
tasks. This means that the instantaneous PELT utilization of a RUNNING
task is not really meaningful to properly support scheduler decisions.
For all these reasons, a more stable signal can do a better job of
representing the expected/estimated utilization of a task/cfs_rq.
Such a signal can be easily created on top of PELT by still using it as
an estimator which produces values to be aggregated on meaningful
events.
This patch adds a simple implementation of util_est, a new signal built on
top of PELT's util_avg where:
util_est(task) = max(task::util_avg, f(task::util_avg@dequeue))
This allows to remember how big a task has been reported by PELT in its
previous activations via f(task::util_avg@dequeue), which is the new
_task_util_est(struct task_struct*) function added by this patch.
If a task should change its behavior and it runs longer in a new
activation, after a certain time its util_est will just track the
original PELT signal (i.e. task::util_avg).
The estimated utilization of cfs_rq is defined only for root ones.
That's because the only sensible consumer of this signal are the
scheduler and schedutil when looking for the overall CPU utilization
due to FAIR tasks.
For this reason, the estimated utilization of a root cfs_rq is simply
defined as:
util_est(cfs_rq) = max(cfs_rq::util_avg, cfs_rq::util_est::enqueued)
where:
cfs_rq::util_est::enqueued = sum(_task_util_est(task))
for each RUNNABLE task on that root cfs_rq
It's worth noting that the estimated utilization is tracked only for
objects of interests, specifically:
- Tasks: to better support tasks placement decisions
- root cfs_rqs: to better support both tasks placement decisions as
well as frequencies selection
Signed-off-by: Patrick Bellasi <patrick.bellasi@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Joel Fernandes <joelaf@google.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Morten Rasmussen <morten.rasmussen@arm.com>
Cc: Paul Turner <pjt@google.com>
Cc: Rafael J . Wysocki <rafael.j.wysocki@intel.com>
Cc: Steve Muckle <smuckle@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Todd Kjos <tkjos@android.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Cc: Viresh Kumar <viresh.kumar@linaro.org>
Link: http://lkml.kernel.org/r/20180309095245.11071-2-patrick.bellasi@arm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2018-03-09 09:52:42 +00:00
|
|
|
/**
|
|
|
|
* struct util_est - Estimation utilization of FAIR tasks
|
|
|
|
* @enqueued: instantaneous estimated utilization of a task/cpu
|
|
|
|
* @ewma: the Exponential Weighted Moving Average (EWMA)
|
|
|
|
* utilization of a task
|
|
|
|
*
|
|
|
|
* Support data structure to track an Exponential Weighted Moving Average
|
|
|
|
* (EWMA) of a FAIR task's utilization. New samples are added to the moving
|
|
|
|
* average each time a task completes an activation. Sample's weight is chosen
|
|
|
|
* so that the EWMA will be relatively insensitive to transient changes to the
|
|
|
|
* task's workload.
|
|
|
|
*
|
|
|
|
* The enqueued attribute has a slightly different meaning for tasks and cpus:
|
|
|
|
* - task: the task's util_avg at last task dequeue time
|
|
|
|
* - cfs_rq: the sum of util_est.enqueued for each RUNNABLE task on that CPU
|
|
|
|
* Thus, the util_est.enqueued of a task represents the contribution on the
|
|
|
|
* estimated utilization of the CPU where that task is currently enqueued.
|
|
|
|
*
|
|
|
|
* Only for tasks we track a moving average of the past instantaneous
|
|
|
|
* estimated utilization. This allows to absorb sporadic drops in utilization
|
|
|
|
* of an otherwise almost periodic task.
|
|
|
|
*/
|
|
|
|
struct util_est {
|
|
|
|
unsigned int enqueued;
|
|
|
|
unsigned int ewma;
|
|
|
|
#define UTIL_EST_WEIGHT_SHIFT 2
|
2018-04-05 08:05:21 +00:00
|
|
|
} __attribute__((__aligned__(sizeof(u64))));
|
sched/fair: Add util_est on top of PELT
The util_avg signal computed by PELT is too variable for some use-cases.
For example, a big task waking up after a long sleep period will have its
utilization almost completely decayed. This introduces some latency before
schedutil will be able to pick the best frequency to run a task.
The same issue can affect task placement. Indeed, since the task
utilization is already decayed at wakeup, when the task is enqueued in a
CPU, this can result in a CPU running a big task as being temporarily
represented as being almost empty. This leads to a race condition where
other tasks can be potentially allocated on a CPU which just started to run
a big task which slept for a relatively long period.
Moreover, the PELT utilization of a task can be updated every [ms], thus
making it a continuously changing value for certain longer running
tasks. This means that the instantaneous PELT utilization of a RUNNING
task is not really meaningful to properly support scheduler decisions.
For all these reasons, a more stable signal can do a better job of
representing the expected/estimated utilization of a task/cfs_rq.
Such a signal can be easily created on top of PELT by still using it as
an estimator which produces values to be aggregated on meaningful
events.
This patch adds a simple implementation of util_est, a new signal built on
top of PELT's util_avg where:
util_est(task) = max(task::util_avg, f(task::util_avg@dequeue))
This allows to remember how big a task has been reported by PELT in its
previous activations via f(task::util_avg@dequeue), which is the new
_task_util_est(struct task_struct*) function added by this patch.
If a task should change its behavior and it runs longer in a new
activation, after a certain time its util_est will just track the
original PELT signal (i.e. task::util_avg).
The estimated utilization of cfs_rq is defined only for root ones.
That's because the only sensible consumer of this signal are the
scheduler and schedutil when looking for the overall CPU utilization
due to FAIR tasks.
For this reason, the estimated utilization of a root cfs_rq is simply
defined as:
util_est(cfs_rq) = max(cfs_rq::util_avg, cfs_rq::util_est::enqueued)
where:
cfs_rq::util_est::enqueued = sum(_task_util_est(task))
for each RUNNABLE task on that root cfs_rq
It's worth noting that the estimated utilization is tracked only for
objects of interests, specifically:
- Tasks: to better support tasks placement decisions
- root cfs_rqs: to better support both tasks placement decisions as
well as frequencies selection
Signed-off-by: Patrick Bellasi <patrick.bellasi@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Joel Fernandes <joelaf@google.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Morten Rasmussen <morten.rasmussen@arm.com>
Cc: Paul Turner <pjt@google.com>
Cc: Rafael J . Wysocki <rafael.j.wysocki@intel.com>
Cc: Steve Muckle <smuckle@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Todd Kjos <tkjos@android.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Cc: Viresh Kumar <viresh.kumar@linaro.org>
Link: http://lkml.kernel.org/r/20180309095245.11071-2-patrick.bellasi@arm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2018-03-09 09:52:42 +00:00
|
|
|
|
sched/fair: Rewrite runnable load and utilization average tracking
The idea of runnable load average (let runnable time contribute to weight)
was proposed by Paul Turner and Ben Segall, and it is still followed by
this rewrite. This rewrite aims to solve the following issues:
1. cfs_rq's load average (namely runnable_load_avg and blocked_load_avg) is
updated at the granularity of an entity at a time, which results in the
cfs_rq's load average is stale or partially updated: at any time, only
one entity is up to date, all other entities are effectively lagging
behind. This is undesirable.
To illustrate, if we have n runnable entities in the cfs_rq, as time
elapses, they certainly become outdated:
t0: cfs_rq { e1_old, e2_old, ..., en_old }
and when we update:
t1: update e1, then we have cfs_rq { e1_new, e2_old, ..., en_old }
t2: update e2, then we have cfs_rq { e1_old, e2_new, ..., en_old }
...
We solve this by combining all runnable entities' load averages together
in cfs_rq's avg, and update the cfs_rq's avg as a whole. This is based
on the fact that if we regard the update as a function, then:
w * update(e) = update(w * e) and
update(e1) + update(e2) = update(e1 + e2), then
w1 * update(e1) + w2 * update(e2) = update(w1 * e1 + w2 * e2)
therefore, by this rewrite, we have an entirely updated cfs_rq at the
time we update it:
t1: update cfs_rq { e1_new, e2_new, ..., en_new }
t2: update cfs_rq { e1_new, e2_new, ..., en_new }
...
2. cfs_rq's load average is different between top rq->cfs_rq and other
task_group's per CPU cfs_rqs in whether or not blocked_load_average
contributes to the load.
The basic idea behind runnable load average (the same for utilization)
is that the blocked state is taken into account as opposed to only
accounting for the currently runnable state. Therefore, the average
should include both the runnable/running and blocked load averages.
This rewrite does that.
In addition, we also combine runnable/running and blocked averages
of all entities into the cfs_rq's average, and update it together at
once. This is based on the fact that:
update(runnable) + update(blocked) = update(runnable + blocked)
This significantly reduces the code as we don't need to separately
maintain/update runnable/running load and blocked load.
3. How task_group entities' share is calculated is complex and imprecise.
We reduce the complexity in this rewrite to allow a very simple rule:
the task_group's load_avg is aggregated from its per CPU cfs_rqs's
load_avgs. Then group entity's weight is simply proportional to its
own cfs_rq's load_avg / task_group's load_avg. To illustrate,
if a task_group has { cfs_rq1, cfs_rq2, ..., cfs_rqn }, then,
task_group_avg = cfs_rq1_avg + cfs_rq2_avg + ... + cfs_rqn_avg, then
cfs_rqx's entity's share = cfs_rqx_avg / task_group_avg * task_group's share
To sum up, this rewrite in principle is equivalent to the current one, but
fixes the issues described above. Turns out, it significantly reduces the
code complexity and hence increases clarity and efficiency. In addition,
the new averages are more smooth/continuous (no spurious spikes and valleys)
and updated more consistently and quickly to reflect the load dynamics.
As a result, we have less load tracking overhead, better performance,
and especially better power efficiency due to more balanced load.
Signed-off-by: Yuyang Du <yuyang.du@intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: arjan@linux.intel.com
Cc: bsegall@google.com
Cc: dietmar.eggemann@arm.com
Cc: fengguang.wu@intel.com
Cc: len.brown@intel.com
Cc: morten.rasmussen@arm.com
Cc: pjt@google.com
Cc: rafael.j.wysocki@intel.com
Cc: umgwanakikbuti@gmail.com
Cc: vincent.guittot@linaro.org
Link: http://lkml.kernel.org/r/1436918682-4971-3-git-send-email-yuyang.du@intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-07-15 00:04:37 +00:00
|
|
|
/*
|
2016-04-05 04:12:28 +00:00
|
|
|
* The load_avg/util_avg accumulates an infinite geometric series
|
|
|
|
* (see __update_load_avg() in kernel/sched/fair.c).
|
|
|
|
*
|
|
|
|
* [load_avg definition]
|
|
|
|
*
|
|
|
|
* load_avg = runnable% * scale_load_down(load)
|
|
|
|
*
|
|
|
|
* where runnable% is the time ratio that a sched_entity is runnable.
|
|
|
|
* For cfs_rq, it is the aggregated load_avg of all runnable and
|
sched/fair: Rewrite runnable load and utilization average tracking
The idea of runnable load average (let runnable time contribute to weight)
was proposed by Paul Turner and Ben Segall, and it is still followed by
this rewrite. This rewrite aims to solve the following issues:
1. cfs_rq's load average (namely runnable_load_avg and blocked_load_avg) is
updated at the granularity of an entity at a time, which results in the
cfs_rq's load average is stale or partially updated: at any time, only
one entity is up to date, all other entities are effectively lagging
behind. This is undesirable.
To illustrate, if we have n runnable entities in the cfs_rq, as time
elapses, they certainly become outdated:
t0: cfs_rq { e1_old, e2_old, ..., en_old }
and when we update:
t1: update e1, then we have cfs_rq { e1_new, e2_old, ..., en_old }
t2: update e2, then we have cfs_rq { e1_old, e2_new, ..., en_old }
...
We solve this by combining all runnable entities' load averages together
in cfs_rq's avg, and update the cfs_rq's avg as a whole. This is based
on the fact that if we regard the update as a function, then:
w * update(e) = update(w * e) and
update(e1) + update(e2) = update(e1 + e2), then
w1 * update(e1) + w2 * update(e2) = update(w1 * e1 + w2 * e2)
therefore, by this rewrite, we have an entirely updated cfs_rq at the
time we update it:
t1: update cfs_rq { e1_new, e2_new, ..., en_new }
t2: update cfs_rq { e1_new, e2_new, ..., en_new }
...
2. cfs_rq's load average is different between top rq->cfs_rq and other
task_group's per CPU cfs_rqs in whether or not blocked_load_average
contributes to the load.
The basic idea behind runnable load average (the same for utilization)
is that the blocked state is taken into account as opposed to only
accounting for the currently runnable state. Therefore, the average
should include both the runnable/running and blocked load averages.
This rewrite does that.
In addition, we also combine runnable/running and blocked averages
of all entities into the cfs_rq's average, and update it together at
once. This is based on the fact that:
update(runnable) + update(blocked) = update(runnable + blocked)
This significantly reduces the code as we don't need to separately
maintain/update runnable/running load and blocked load.
3. How task_group entities' share is calculated is complex and imprecise.
We reduce the complexity in this rewrite to allow a very simple rule:
the task_group's load_avg is aggregated from its per CPU cfs_rqs's
load_avgs. Then group entity's weight is simply proportional to its
own cfs_rq's load_avg / task_group's load_avg. To illustrate,
if a task_group has { cfs_rq1, cfs_rq2, ..., cfs_rqn }, then,
task_group_avg = cfs_rq1_avg + cfs_rq2_avg + ... + cfs_rqn_avg, then
cfs_rqx's entity's share = cfs_rqx_avg / task_group_avg * task_group's share
To sum up, this rewrite in principle is equivalent to the current one, but
fixes the issues described above. Turns out, it significantly reduces the
code complexity and hence increases clarity and efficiency. In addition,
the new averages are more smooth/continuous (no spurious spikes and valleys)
and updated more consistently and quickly to reflect the load dynamics.
As a result, we have less load tracking overhead, better performance,
and especially better power efficiency due to more balanced load.
Signed-off-by: Yuyang Du <yuyang.du@intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: arjan@linux.intel.com
Cc: bsegall@google.com
Cc: dietmar.eggemann@arm.com
Cc: fengguang.wu@intel.com
Cc: len.brown@intel.com
Cc: morten.rasmussen@arm.com
Cc: pjt@google.com
Cc: rafael.j.wysocki@intel.com
Cc: umgwanakikbuti@gmail.com
Cc: vincent.guittot@linaro.org
Link: http://lkml.kernel.org/r/1436918682-4971-3-git-send-email-yuyang.du@intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-07-15 00:04:37 +00:00
|
|
|
* blocked sched_entities.
|
2016-04-05 04:12:28 +00:00
|
|
|
*
|
|
|
|
* [util_avg definition]
|
|
|
|
*
|
|
|
|
* util_avg = running% * SCHED_CAPACITY_SCALE
|
|
|
|
*
|
|
|
|
* where running% is the time ratio that a sched_entity is running on
|
|
|
|
* a CPU. For cfs_rq, it is the aggregated util_avg of all runnable
|
|
|
|
* and blocked sched_entities.
|
|
|
|
*
|
sched/fair: Update scale invariance of PELT
The current implementation of load tracking invariance scales the
contribution with current frequency and uarch performance (only for
utilization) of the CPU. One main result of this formula is that the
figures are capped by current capacity of CPU. Another one is that the
load_avg is not invariant because not scaled with uarch.
The util_avg of a periodic task that runs r time slots every p time slots
varies in the range :
U * (1-y^r)/(1-y^p) * y^i < Utilization < U * (1-y^r)/(1-y^p)
with U is the max util_avg value = SCHED_CAPACITY_SCALE
At a lower capacity, the range becomes:
U * C * (1-y^r')/(1-y^p) * y^i' < Utilization < U * C * (1-y^r')/(1-y^p)
with C reflecting the compute capacity ratio between current capacity and
max capacity.
so C tries to compensate changes in (1-y^r') but it can't be accurate.
Instead of scaling the contribution value of PELT algo, we should scale the
running time. The PELT signal aims to track the amount of computation of
tasks and/or rq so it seems more correct to scale the running time to
reflect the effective amount of computation done since the last update.
In order to be fully invariant, we need to apply the same amount of
running time and idle time whatever the current capacity. Because running
at lower capacity implies that the task will run longer, we have to ensure
that the same amount of idle time will be applied when system becomes idle
and no idle time has been "stolen". But reaching the maximum utilization
value (SCHED_CAPACITY_SCALE) means that the task is seen as an
always-running task whatever the capacity of the CPU (even at max compute
capacity). In this case, we can discard this "stolen" idle times which
becomes meaningless.
In order to achieve this time scaling, a new clock_pelt is created per rq.
The increase of this clock scales with current capacity when something
is running on rq and synchronizes with clock_task when rq is idle. With
this mechanism, we ensure the same running and idle time whatever the
current capacity. This also enables to simplify the pelt algorithm by
removing all references of uarch and frequency and applying the same
contribution to utilization and loads. Furthermore, the scaling is done
only once per update of clock (update_rq_clock_task()) instead of during
each update of sched_entities and cfs/rt/dl_rq of the rq like the current
implementation. This is interesting when cgroup are involved as shown in
the results below:
On a hikey (octo Arm64 platform).
Performance cpufreq governor and only shallowest c-state to remove variance
generated by those power features so we only track the impact of pelt algo.
each test runs 16 times:
./perf bench sched pipe
(higher is better)
kernel tip/sched/core + patch
ops/seconds ops/seconds diff
cgroup
root 59652(+/- 0.18%) 59876(+/- 0.24%) +0.38%
level1 55608(+/- 0.27%) 55923(+/- 0.24%) +0.57%
level2 52115(+/- 0.29%) 52564(+/- 0.22%) +0.86%
hackbench -l 1000
(lower is better)
kernel tip/sched/core + patch
duration(sec) duration(sec) diff
cgroup
root 4.453(+/- 2.37%) 4.383(+/- 2.88%) -1.57%
level1 4.859(+/- 8.50%) 4.830(+/- 7.07%) -0.60%
level2 5.063(+/- 9.83%) 4.928(+/- 9.66%) -2.66%
Then, the responsiveness of PELT is improved when CPU is not running at max
capacity with this new algorithm. I have put below some examples of
duration to reach some typical load values according to the capacity of the
CPU with current implementation and with this patch. These values has been
computed based on the geometric series and the half period value:
Util (%) max capacity half capacity(mainline) half capacity(w/ patch)
972 (95%) 138ms not reachable 276ms
486 (47.5%) 30ms 138ms 60ms
256 (25%) 13ms 32ms 26ms
On my hikey (octo Arm64 platform) with schedutil governor, the time to
reach max OPP when starting from a null utilization, decreases from 223ms
with current scale invariance down to 121ms with the new algorithm.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Morten.Rasmussen@arm.com
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: bsegall@google.com
Cc: dietmar.eggemann@arm.com
Cc: patrick.bellasi@arm.com
Cc: pjt@google.com
Cc: pkondeti@codeaurora.org
Cc: quentin.perret@arm.com
Cc: rjw@rjwysocki.net
Cc: srinivas.pandruvada@linux.intel.com
Cc: thara.gopinath@linaro.org
Link: https://lkml.kernel.org/r/1548257214-13745-3-git-send-email-vincent.guittot@linaro.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2019-01-23 15:26:53 +00:00
|
|
|
* load_avg and util_avg don't direcly factor frequency scaling and CPU
|
|
|
|
* capacity scaling. The scaling is done through the rq_clock_pelt that
|
|
|
|
* is used for computing those signals (see update_rq_clock_pelt())
|
2016-04-05 04:12:28 +00:00
|
|
|
*
|
sched/fair: Update scale invariance of PELT
The current implementation of load tracking invariance scales the
contribution with current frequency and uarch performance (only for
utilization) of the CPU. One main result of this formula is that the
figures are capped by current capacity of CPU. Another one is that the
load_avg is not invariant because not scaled with uarch.
The util_avg of a periodic task that runs r time slots every p time slots
varies in the range :
U * (1-y^r)/(1-y^p) * y^i < Utilization < U * (1-y^r)/(1-y^p)
with U is the max util_avg value = SCHED_CAPACITY_SCALE
At a lower capacity, the range becomes:
U * C * (1-y^r')/(1-y^p) * y^i' < Utilization < U * C * (1-y^r')/(1-y^p)
with C reflecting the compute capacity ratio between current capacity and
max capacity.
so C tries to compensate changes in (1-y^r') but it can't be accurate.
Instead of scaling the contribution value of PELT algo, we should scale the
running time. The PELT signal aims to track the amount of computation of
tasks and/or rq so it seems more correct to scale the running time to
reflect the effective amount of computation done since the last update.
In order to be fully invariant, we need to apply the same amount of
running time and idle time whatever the current capacity. Because running
at lower capacity implies that the task will run longer, we have to ensure
that the same amount of idle time will be applied when system becomes idle
and no idle time has been "stolen". But reaching the maximum utilization
value (SCHED_CAPACITY_SCALE) means that the task is seen as an
always-running task whatever the capacity of the CPU (even at max compute
capacity). In this case, we can discard this "stolen" idle times which
becomes meaningless.
In order to achieve this time scaling, a new clock_pelt is created per rq.
The increase of this clock scales with current capacity when something
is running on rq and synchronizes with clock_task when rq is idle. With
this mechanism, we ensure the same running and idle time whatever the
current capacity. This also enables to simplify the pelt algorithm by
removing all references of uarch and frequency and applying the same
contribution to utilization and loads. Furthermore, the scaling is done
only once per update of clock (update_rq_clock_task()) instead of during
each update of sched_entities and cfs/rt/dl_rq of the rq like the current
implementation. This is interesting when cgroup are involved as shown in
the results below:
On a hikey (octo Arm64 platform).
Performance cpufreq governor and only shallowest c-state to remove variance
generated by those power features so we only track the impact of pelt algo.
each test runs 16 times:
./perf bench sched pipe
(higher is better)
kernel tip/sched/core + patch
ops/seconds ops/seconds diff
cgroup
root 59652(+/- 0.18%) 59876(+/- 0.24%) +0.38%
level1 55608(+/- 0.27%) 55923(+/- 0.24%) +0.57%
level2 52115(+/- 0.29%) 52564(+/- 0.22%) +0.86%
hackbench -l 1000
(lower is better)
kernel tip/sched/core + patch
duration(sec) duration(sec) diff
cgroup
root 4.453(+/- 2.37%) 4.383(+/- 2.88%) -1.57%
level1 4.859(+/- 8.50%) 4.830(+/- 7.07%) -0.60%
level2 5.063(+/- 9.83%) 4.928(+/- 9.66%) -2.66%
Then, the responsiveness of PELT is improved when CPU is not running at max
capacity with this new algorithm. I have put below some examples of
duration to reach some typical load values according to the capacity of the
CPU with current implementation and with this patch. These values has been
computed based on the geometric series and the half period value:
Util (%) max capacity half capacity(mainline) half capacity(w/ patch)
972 (95%) 138ms not reachable 276ms
486 (47.5%) 30ms 138ms 60ms
256 (25%) 13ms 32ms 26ms
On my hikey (octo Arm64 platform) with schedutil governor, the time to
reach max OPP when starting from a null utilization, decreases from 223ms
with current scale invariance down to 121ms with the new algorithm.
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Morten.Rasmussen@arm.com
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: bsegall@google.com
Cc: dietmar.eggemann@arm.com
Cc: patrick.bellasi@arm.com
Cc: pjt@google.com
Cc: pkondeti@codeaurora.org
Cc: quentin.perret@arm.com
Cc: rjw@rjwysocki.net
Cc: srinivas.pandruvada@linux.intel.com
Cc: thara.gopinath@linaro.org
Link: https://lkml.kernel.org/r/1548257214-13745-3-git-send-email-vincent.guittot@linaro.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2019-01-23 15:26:53 +00:00
|
|
|
* N.B., the above ratios (runnable% and running%) themselves are in the
|
|
|
|
* range of [0, 1]. To do fixed point arithmetics, we therefore scale them
|
|
|
|
* to as large a range as necessary. This is for example reflected by
|
|
|
|
* util_avg's SCHED_CAPACITY_SCALE.
|
2016-04-05 04:12:28 +00:00
|
|
|
*
|
|
|
|
* [Overflow issue]
|
|
|
|
*
|
|
|
|
* The 64-bit load_sum can have 4353082796 (=2^64/47742/88761) entities
|
|
|
|
* with the highest load (=88761), always runnable on a single cfs_rq,
|
|
|
|
* and should not overflow as the number already hits PID_MAX_LIMIT.
|
|
|
|
*
|
|
|
|
* For all other cases (including 32-bit kernels), struct load_weight's
|
|
|
|
* weight will overflow first before we do, because:
|
|
|
|
*
|
|
|
|
* Max(load_avg) <= Max(load.weight)
|
|
|
|
*
|
|
|
|
* Then it is the load_weight's responsibility to consider overflow
|
|
|
|
* issues.
|
sched/fair: Rewrite runnable load and utilization average tracking
The idea of runnable load average (let runnable time contribute to weight)
was proposed by Paul Turner and Ben Segall, and it is still followed by
this rewrite. This rewrite aims to solve the following issues:
1. cfs_rq's load average (namely runnable_load_avg and blocked_load_avg) is
updated at the granularity of an entity at a time, which results in the
cfs_rq's load average is stale or partially updated: at any time, only
one entity is up to date, all other entities are effectively lagging
behind. This is undesirable.
To illustrate, if we have n runnable entities in the cfs_rq, as time
elapses, they certainly become outdated:
t0: cfs_rq { e1_old, e2_old, ..., en_old }
and when we update:
t1: update e1, then we have cfs_rq { e1_new, e2_old, ..., en_old }
t2: update e2, then we have cfs_rq { e1_old, e2_new, ..., en_old }
...
We solve this by combining all runnable entities' load averages together
in cfs_rq's avg, and update the cfs_rq's avg as a whole. This is based
on the fact that if we regard the update as a function, then:
w * update(e) = update(w * e) and
update(e1) + update(e2) = update(e1 + e2), then
w1 * update(e1) + w2 * update(e2) = update(w1 * e1 + w2 * e2)
therefore, by this rewrite, we have an entirely updated cfs_rq at the
time we update it:
t1: update cfs_rq { e1_new, e2_new, ..., en_new }
t2: update cfs_rq { e1_new, e2_new, ..., en_new }
...
2. cfs_rq's load average is different between top rq->cfs_rq and other
task_group's per CPU cfs_rqs in whether or not blocked_load_average
contributes to the load.
The basic idea behind runnable load average (the same for utilization)
is that the blocked state is taken into account as opposed to only
accounting for the currently runnable state. Therefore, the average
should include both the runnable/running and blocked load averages.
This rewrite does that.
In addition, we also combine runnable/running and blocked averages
of all entities into the cfs_rq's average, and update it together at
once. This is based on the fact that:
update(runnable) + update(blocked) = update(runnable + blocked)
This significantly reduces the code as we don't need to separately
maintain/update runnable/running load and blocked load.
3. How task_group entities' share is calculated is complex and imprecise.
We reduce the complexity in this rewrite to allow a very simple rule:
the task_group's load_avg is aggregated from its per CPU cfs_rqs's
load_avgs. Then group entity's weight is simply proportional to its
own cfs_rq's load_avg / task_group's load_avg. To illustrate,
if a task_group has { cfs_rq1, cfs_rq2, ..., cfs_rqn }, then,
task_group_avg = cfs_rq1_avg + cfs_rq2_avg + ... + cfs_rqn_avg, then
cfs_rqx's entity's share = cfs_rqx_avg / task_group_avg * task_group's share
To sum up, this rewrite in principle is equivalent to the current one, but
fixes the issues described above. Turns out, it significantly reduces the
code complexity and hence increases clarity and efficiency. In addition,
the new averages are more smooth/continuous (no spurious spikes and valleys)
and updated more consistently and quickly to reflect the load dynamics.
As a result, we have less load tracking overhead, better performance,
and especially better power efficiency due to more balanced load.
Signed-off-by: Yuyang Du <yuyang.du@intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: arjan@linux.intel.com
Cc: bsegall@google.com
Cc: dietmar.eggemann@arm.com
Cc: fengguang.wu@intel.com
Cc: len.brown@intel.com
Cc: morten.rasmussen@arm.com
Cc: pjt@google.com
Cc: rafael.j.wysocki@intel.com
Cc: umgwanakikbuti@gmail.com
Cc: vincent.guittot@linaro.org
Link: http://lkml.kernel.org/r/1436918682-4971-3-git-send-email-yuyang.du@intel.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-07-15 00:04:37 +00:00
|
|
|
*/
|
2012-10-04 11:18:29 +00:00
|
|
|
struct sched_avg {
|
2017-02-06 21:06:35 +00:00
|
|
|
u64 last_update_time;
|
|
|
|
u64 load_sum;
|
2017-05-06 13:59:54 +00:00
|
|
|
u64 runnable_load_sum;
|
2017-02-06 21:06:35 +00:00
|
|
|
u32 util_sum;
|
|
|
|
u32 period_contrib;
|
|
|
|
unsigned long load_avg;
|
2017-05-06 13:59:54 +00:00
|
|
|
unsigned long runnable_load_avg;
|
2017-02-06 21:06:35 +00:00
|
|
|
unsigned long util_avg;
|
sched/fair: Add util_est on top of PELT
The util_avg signal computed by PELT is too variable for some use-cases.
For example, a big task waking up after a long sleep period will have its
utilization almost completely decayed. This introduces some latency before
schedutil will be able to pick the best frequency to run a task.
The same issue can affect task placement. Indeed, since the task
utilization is already decayed at wakeup, when the task is enqueued in a
CPU, this can result in a CPU running a big task as being temporarily
represented as being almost empty. This leads to a race condition where
other tasks can be potentially allocated on a CPU which just started to run
a big task which slept for a relatively long period.
Moreover, the PELT utilization of a task can be updated every [ms], thus
making it a continuously changing value for certain longer running
tasks. This means that the instantaneous PELT utilization of a RUNNING
task is not really meaningful to properly support scheduler decisions.
For all these reasons, a more stable signal can do a better job of
representing the expected/estimated utilization of a task/cfs_rq.
Such a signal can be easily created on top of PELT by still using it as
an estimator which produces values to be aggregated on meaningful
events.
This patch adds a simple implementation of util_est, a new signal built on
top of PELT's util_avg where:
util_est(task) = max(task::util_avg, f(task::util_avg@dequeue))
This allows to remember how big a task has been reported by PELT in its
previous activations via f(task::util_avg@dequeue), which is the new
_task_util_est(struct task_struct*) function added by this patch.
If a task should change its behavior and it runs longer in a new
activation, after a certain time its util_est will just track the
original PELT signal (i.e. task::util_avg).
The estimated utilization of cfs_rq is defined only for root ones.
That's because the only sensible consumer of this signal are the
scheduler and schedutil when looking for the overall CPU utilization
due to FAIR tasks.
For this reason, the estimated utilization of a root cfs_rq is simply
defined as:
util_est(cfs_rq) = max(cfs_rq::util_avg, cfs_rq::util_est::enqueued)
where:
cfs_rq::util_est::enqueued = sum(_task_util_est(task))
for each RUNNABLE task on that root cfs_rq
It's worth noting that the estimated utilization is tracked only for
objects of interests, specifically:
- Tasks: to better support tasks placement decisions
- root cfs_rqs: to better support both tasks placement decisions as
well as frequencies selection
Signed-off-by: Patrick Bellasi <patrick.bellasi@arm.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Dietmar Eggemann <dietmar.eggemann@arm.com>
Cc: Joel Fernandes <joelaf@google.com>
Cc: Juri Lelli <juri.lelli@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Morten Rasmussen <morten.rasmussen@arm.com>
Cc: Paul Turner <pjt@google.com>
Cc: Rafael J . Wysocki <rafael.j.wysocki@intel.com>
Cc: Steve Muckle <smuckle@google.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Todd Kjos <tkjos@android.com>
Cc: Vincent Guittot <vincent.guittot@linaro.org>
Cc: Viresh Kumar <viresh.kumar@linaro.org>
Link: http://lkml.kernel.org/r/20180309095245.11071-2-patrick.bellasi@arm.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2018-03-09 09:52:42 +00:00
|
|
|
struct util_est util_est;
|
2018-04-05 08:05:21 +00:00
|
|
|
} ____cacheline_aligned;
|
2012-10-04 11:18:29 +00:00
|
|
|
|
2010-03-11 02:37:45 +00:00
|
|
|
struct sched_statistics {
|
2017-02-06 10:44:12 +00:00
|
|
|
#ifdef CONFIG_SCHEDSTATS
|
2017-02-06 21:06:35 +00:00
|
|
|
u64 wait_start;
|
|
|
|
u64 wait_max;
|
|
|
|
u64 wait_count;
|
|
|
|
u64 wait_sum;
|
|
|
|
u64 iowait_count;
|
|
|
|
u64 iowait_sum;
|
|
|
|
|
|
|
|
u64 sleep_start;
|
|
|
|
u64 sleep_max;
|
|
|
|
s64 sum_sleep_runtime;
|
|
|
|
|
|
|
|
u64 block_start;
|
|
|
|
u64 block_max;
|
|
|
|
u64 exec_max;
|
|
|
|
u64 slice_max;
|
|
|
|
|
|
|
|
u64 nr_migrations_cold;
|
|
|
|
u64 nr_failed_migrations_affine;
|
|
|
|
u64 nr_failed_migrations_running;
|
|
|
|
u64 nr_failed_migrations_hot;
|
|
|
|
u64 nr_forced_migrations;
|
|
|
|
|
|
|
|
u64 nr_wakeups;
|
|
|
|
u64 nr_wakeups_sync;
|
|
|
|
u64 nr_wakeups_migrate;
|
|
|
|
u64 nr_wakeups_local;
|
|
|
|
u64 nr_wakeups_remote;
|
|
|
|
u64 nr_wakeups_affine;
|
|
|
|
u64 nr_wakeups_affine_attempts;
|
|
|
|
u64 nr_wakeups_passive;
|
|
|
|
u64 nr_wakeups_idle;
|
2010-03-11 02:37:45 +00:00
|
|
|
#endif
|
2017-02-06 10:44:12 +00:00
|
|
|
};
|
2010-03-11 02:37:45 +00:00
|
|
|
|
|
|
|
struct sched_entity {
|
2017-02-06 21:06:35 +00:00
|
|
|
/* For load-balancing: */
|
|
|
|
struct load_weight load;
|
2017-05-06 13:59:54 +00:00
|
|
|
unsigned long runnable_weight;
|
2017-02-06 21:06:35 +00:00
|
|
|
struct rb_node run_node;
|
|
|
|
struct list_head group_node;
|
|
|
|
unsigned int on_rq;
|
2010-03-11 02:37:45 +00:00
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
u64 exec_start;
|
|
|
|
u64 sum_exec_runtime;
|
|
|
|
u64 vruntime;
|
|
|
|
u64 prev_sum_exec_runtime;
|
2010-03-11 02:37:45 +00:00
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
u64 nr_migrations;
|
2010-03-11 02:37:45 +00:00
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
struct sched_statistics statistics;
|
2007-08-02 15:41:40 +00:00
|
|
|
|
2007-07-09 16:51:58 +00:00
|
|
|
#ifdef CONFIG_FAIR_GROUP_SCHED
|
2017-02-06 21:06:35 +00:00
|
|
|
int depth;
|
|
|
|
struct sched_entity *parent;
|
2007-07-09 16:51:58 +00:00
|
|
|
/* rq on which this entity is (to be) queued: */
|
2017-02-06 21:06:35 +00:00
|
|
|
struct cfs_rq *cfs_rq;
|
2007-07-09 16:51:58 +00:00
|
|
|
/* rq "owned" by this entity/group: */
|
2017-02-06 21:06:35 +00:00
|
|
|
struct cfs_rq *my_q;
|
2007-07-09 16:51:58 +00:00
|
|
|
#endif
|
2013-02-07 15:47:07 +00:00
|
|
|
|
2013-06-26 05:05:39 +00:00
|
|
|
#ifdef CONFIG_SMP
|
sched/core: Move sched_entity::avg into separate cache line
The sched_entity::avg collides with read-mostly sched_entity data.
The perf c2c tool showed many read HITM accesses across
many CPUs for sched_entity's cfs_rq and my_q, while having
at the same time tons of stores for avg.
After placing sched_entity::avg into separate cache line,
the perf bench sched pipe showed around 20 seconds speedup.
NOTE I cut out all perf events except for cycles and
instructions from following output.
Before:
$ perf stat -r 5 perf bench sched pipe -l 10000000
# Running 'sched/pipe' benchmark:
# Executed 10000000 pipe operations between two processes
Total time: 270.348 [sec]
27.034805 usecs/op
36989 ops/sec
...
245,537,074,035 cycles # 1.433 GHz
187,264,548,519 instructions # 0.77 insns per cycle
272.653840535 seconds time elapsed ( +- 1.31% )
After:
$ perf stat -r 5 perf bench sched pipe -l 10000000
# Running 'sched/pipe' benchmark:
# Executed 10000000 pipe operations between two processes
Total time: 251.076 [sec]
25.107678 usecs/op
39828 ops/sec
...
244,573,513,928 cycles # 1.572 GHz
187,409,641,157 instructions # 0.76 insns per cycle
251.679315188 seconds time elapsed ( +- 0.31% )
Signed-off-by: Jiri Olsa <jolsa@kernel.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@kernel.org>
Cc: Don Zickus <dzickus@redhat.com>
Cc: Joe Mario <jmario@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/1449606239-28602-1-git-send-email-jolsa@kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2015-12-08 20:23:59 +00:00
|
|
|
/*
|
|
|
|
* Per entity load average tracking.
|
|
|
|
*
|
|
|
|
* Put into separate cache line so it does not
|
|
|
|
* collide with read-mostly values above.
|
|
|
|
*/
|
2018-04-05 08:05:21 +00:00
|
|
|
struct sched_avg avg;
|
2012-10-04 11:18:29 +00:00
|
|
|
#endif
|
2007-07-09 16:51:58 +00:00
|
|
|
};
|
2006-07-03 07:25:42 +00:00
|
|
|
|
2008-01-25 20:08:27 +00:00
|
|
|
struct sched_rt_entity {
|
2017-02-06 21:06:35 +00:00
|
|
|
struct list_head run_list;
|
|
|
|
unsigned long timeout;
|
|
|
|
unsigned long watchdog_stamp;
|
|
|
|
unsigned int time_slice;
|
|
|
|
unsigned short on_rq;
|
|
|
|
unsigned short on_list;
|
|
|
|
|
|
|
|
struct sched_rt_entity *back;
|
2008-02-13 14:45:40 +00:00
|
|
|
#ifdef CONFIG_RT_GROUP_SCHED
|
2017-02-06 21:06:35 +00:00
|
|
|
struct sched_rt_entity *parent;
|
2008-01-25 20:08:30 +00:00
|
|
|
/* rq on which this entity is (to be) queued: */
|
2017-02-06 21:06:35 +00:00
|
|
|
struct rt_rq *rt_rq;
|
2008-01-25 20:08:30 +00:00
|
|
|
/* rq "owned" by this entity/group: */
|
2017-02-06 21:06:35 +00:00
|
|
|
struct rt_rq *my_q;
|
2008-01-25 20:08:30 +00:00
|
|
|
#endif
|
2016-10-28 08:22:25 +00:00
|
|
|
} __randomize_layout;
|
2008-01-25 20:08:27 +00:00
|
|
|
|
sched/deadline: Add SCHED_DEADLINE structures & implementation
Introduces the data structures, constants and symbols needed for
SCHED_DEADLINE implementation.
Core data structure of SCHED_DEADLINE are defined, along with their
initializers. Hooks for checking if a task belong to the new policy
are also added where they are needed.
Adds a scheduling class, in sched/dl.c and a new policy called
SCHED_DEADLINE. It is an implementation of the Earliest Deadline
First (EDF) scheduling algorithm, augmented with a mechanism (called
Constant Bandwidth Server, CBS) that makes it possible to isolate
the behaviour of tasks between each other.
The typical -deadline task will be made up of a computation phase
(instance) which is activated on a periodic or sporadic fashion. The
expected (maximum) duration of such computation is called the task's
runtime; the time interval by which each instance need to be completed
is called the task's relative deadline. The task's absolute deadline
is dynamically calculated as the time instant a task (better, an
instance) activates plus the relative deadline.
The EDF algorithms selects the task with the smallest absolute
deadline as the one to be executed first, while the CBS ensures each
task to run for at most its runtime every (relative) deadline
length time interval, avoiding any interference between different
tasks (bandwidth isolation).
Thanks to this feature, also tasks that do not strictly comply with
the computational model sketched above can effectively use the new
policy.
To summarize, this patch:
- introduces the data structures, constants and symbols needed;
- implements the core logic of the scheduling algorithm in the new
scheduling class file;
- provides all the glue code between the new scheduling class and
the core scheduler and refines the interactions between sched/dl
and the other existing scheduling classes.
Signed-off-by: Dario Faggioli <raistlin@linux.it>
Signed-off-by: Michael Trimarchi <michael@amarulasolutions.com>
Signed-off-by: Fabio Checconi <fchecconi@gmail.com>
Signed-off-by: Juri Lelli <juri.lelli@gmail.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1383831828-15501-4-git-send-email-juri.lelli@gmail.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2013-11-28 10:14:43 +00:00
|
|
|
struct sched_dl_entity {
|
2017-02-06 21:06:35 +00:00
|
|
|
struct rb_node rb_node;
|
sched/deadline: Add SCHED_DEADLINE structures & implementation
Introduces the data structures, constants and symbols needed for
SCHED_DEADLINE implementation.
Core data structure of SCHED_DEADLINE are defined, along with their
initializers. Hooks for checking if a task belong to the new policy
are also added where they are needed.
Adds a scheduling class, in sched/dl.c and a new policy called
SCHED_DEADLINE. It is an implementation of the Earliest Deadline
First (EDF) scheduling algorithm, augmented with a mechanism (called
Constant Bandwidth Server, CBS) that makes it possible to isolate
the behaviour of tasks between each other.
The typical -deadline task will be made up of a computation phase
(instance) which is activated on a periodic or sporadic fashion. The
expected (maximum) duration of such computation is called the task's
runtime; the time interval by which each instance need to be completed
is called the task's relative deadline. The task's absolute deadline
is dynamically calculated as the time instant a task (better, an
instance) activates plus the relative deadline.
The EDF algorithms selects the task with the smallest absolute
deadline as the one to be executed first, while the CBS ensures each
task to run for at most its runtime every (relative) deadline
length time interval, avoiding any interference between different
tasks (bandwidth isolation).
Thanks to this feature, also tasks that do not strictly comply with
the computational model sketched above can effectively use the new
policy.
To summarize, this patch:
- introduces the data structures, constants and symbols needed;
- implements the core logic of the scheduling algorithm in the new
scheduling class file;
- provides all the glue code between the new scheduling class and
the core scheduler and refines the interactions between sched/dl
and the other existing scheduling classes.
Signed-off-by: Dario Faggioli <raistlin@linux.it>
Signed-off-by: Michael Trimarchi <michael@amarulasolutions.com>
Signed-off-by: Fabio Checconi <fchecconi@gmail.com>
Signed-off-by: Juri Lelli <juri.lelli@gmail.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1383831828-15501-4-git-send-email-juri.lelli@gmail.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2013-11-28 10:14:43 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Original scheduling parameters. Copied here from sched_attr
|
2014-05-09 03:21:27 +00:00
|
|
|
* during sched_setattr(), they will remain the same until
|
|
|
|
* the next sched_setattr().
|
sched/deadline: Add SCHED_DEADLINE structures & implementation
Introduces the data structures, constants and symbols needed for
SCHED_DEADLINE implementation.
Core data structure of SCHED_DEADLINE are defined, along with their
initializers. Hooks for checking if a task belong to the new policy
are also added where they are needed.
Adds a scheduling class, in sched/dl.c and a new policy called
SCHED_DEADLINE. It is an implementation of the Earliest Deadline
First (EDF) scheduling algorithm, augmented with a mechanism (called
Constant Bandwidth Server, CBS) that makes it possible to isolate
the behaviour of tasks between each other.
The typical -deadline task will be made up of a computation phase
(instance) which is activated on a periodic or sporadic fashion. The
expected (maximum) duration of such computation is called the task's
runtime; the time interval by which each instance need to be completed
is called the task's relative deadline. The task's absolute deadline
is dynamically calculated as the time instant a task (better, an
instance) activates plus the relative deadline.
The EDF algorithms selects the task with the smallest absolute
deadline as the one to be executed first, while the CBS ensures each
task to run for at most its runtime every (relative) deadline
length time interval, avoiding any interference between different
tasks (bandwidth isolation).
Thanks to this feature, also tasks that do not strictly comply with
the computational model sketched above can effectively use the new
policy.
To summarize, this patch:
- introduces the data structures, constants and symbols needed;
- implements the core logic of the scheduling algorithm in the new
scheduling class file;
- provides all the glue code between the new scheduling class and
the core scheduler and refines the interactions between sched/dl
and the other existing scheduling classes.
Signed-off-by: Dario Faggioli <raistlin@linux.it>
Signed-off-by: Michael Trimarchi <michael@amarulasolutions.com>
Signed-off-by: Fabio Checconi <fchecconi@gmail.com>
Signed-off-by: Juri Lelli <juri.lelli@gmail.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1383831828-15501-4-git-send-email-juri.lelli@gmail.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2013-11-28 10:14:43 +00:00
|
|
|
*/
|
2017-02-06 21:06:35 +00:00
|
|
|
u64 dl_runtime; /* Maximum runtime for each instance */
|
|
|
|
u64 dl_deadline; /* Relative deadline of each instance */
|
|
|
|
u64 dl_period; /* Separation of two instances (period) */
|
2017-05-29 14:24:02 +00:00
|
|
|
u64 dl_bw; /* dl_runtime / dl_period */
|
sched/deadline: Use the revised wakeup rule for suspending constrained dl tasks
We have been facing some problems with self-suspending constrained
deadline tasks. The main reason is that the original CBS was not
designed for such sort of tasks.
One problem reported by Xunlei Pang takes place when a task
suspends, and then is awakened before the deadline, but so close
to the deadline that its remaining runtime can cause the task
to have an absolute density higher than allowed. In such situation,
the original CBS assumes that the task is facing an early activation,
and so it replenishes the task and set another deadline, one deadline
in the future. This rule works fine for implicit deadline tasks.
Moreover, it allows the system to adapt the period of a task in which
the external event source suffered from a clock drift.
However, this opens the window for bandwidth leakage for constrained
deadline tasks. For instance, a task with the following parameters:
runtime = 5 ms
deadline = 7 ms
[density] = 5 / 7 = 0.71
period = 1000 ms
If the task runs for 1 ms, and then suspends for another 1ms,
it will be awakened with the following parameters:
remaining runtime = 4
laxity = 5
presenting a absolute density of 4 / 5 = 0.80.
In this case, the original CBS would assume the task had an early
wakeup. Then, CBS will reset the runtime, and the absolute deadline will
be postponed by one relative deadline, allowing the task to run.
The problem is that, if the task runs this pattern forever, it will keep
receiving bandwidth, being able to run 1ms every 2ms. Following this
behavior, the task would be able to run 500 ms in 1 sec. Thus running
more than the 5 ms / 1 sec the admission control allowed it to run.
Trying to address the self-suspending case, Luca Abeni, Giuseppe
Lipari, and Juri Lelli [1] revisited the CBS in order to deal with
self-suspending tasks. In the new approach, rather than
replenishing/postponing the absolute deadline, the revised wakeup rule
adjusts the remaining runtime, reducing it to fit into the allowed
density.
A revised version of the idea is:
At a given time t, the maximum absolute density of a task cannot be
higher than its relative density, that is:
runtime / (deadline - t) <= dl_runtime / dl_deadline
Knowing the laxity of a task (deadline - t), it is possible to move
it to the other side of the equality, thus enabling to define max
remaining runtime a task can use within the absolute deadline, without
over-running the allowed density:
runtime = (dl_runtime / dl_deadline) * (deadline - t)
For instance, in our previous example, the task could still run:
runtime = ( 5 / 7 ) * 5
runtime = 3.57 ms
Without causing damage for other deadline tasks. It is note worthy
that the laxity cannot be negative because that would cause a negative
runtime. Thus, this patch depends on the patch:
df8eac8cafce ("sched/deadline: Throttle a constrained deadline task activated after the deadline")
Which throttles a constrained deadline task activated after the
deadline.
Finally, it is also possible to use the revised wakeup rule for
all other tasks, but that would require some more discussions
about pros and cons.
Reported-by: Xunlei Pang <xpang@redhat.com>
Signed-off-by: Daniel Bristot de Oliveira <bristot@redhat.com>
[peterz: replaced dl_is_constrained with dl_is_implicit]
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Juri Lelli <juri.lelli@arm.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Luca Abeni <luca.abeni@santannapisa.it>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Romulo Silva de Oliveira <romulo.deoliveira@ufsc.br>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Tommaso Cucinotta <tommaso.cucinotta@sssup.it>
Link: http://lkml.kernel.org/r/5c800ab3a74a168a84ee5f3f84d12a02e11383be.1495803804.git.bristot@redhat.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2017-05-29 14:24:03 +00:00
|
|
|
u64 dl_density; /* dl_runtime / dl_deadline */
|
sched/deadline: Add SCHED_DEADLINE structures & implementation
Introduces the data structures, constants and symbols needed for
SCHED_DEADLINE implementation.
Core data structure of SCHED_DEADLINE are defined, along with their
initializers. Hooks for checking if a task belong to the new policy
are also added where they are needed.
Adds a scheduling class, in sched/dl.c and a new policy called
SCHED_DEADLINE. It is an implementation of the Earliest Deadline
First (EDF) scheduling algorithm, augmented with a mechanism (called
Constant Bandwidth Server, CBS) that makes it possible to isolate
the behaviour of tasks between each other.
The typical -deadline task will be made up of a computation phase
(instance) which is activated on a periodic or sporadic fashion. The
expected (maximum) duration of such computation is called the task's
runtime; the time interval by which each instance need to be completed
is called the task's relative deadline. The task's absolute deadline
is dynamically calculated as the time instant a task (better, an
instance) activates plus the relative deadline.
The EDF algorithms selects the task with the smallest absolute
deadline as the one to be executed first, while the CBS ensures each
task to run for at most its runtime every (relative) deadline
length time interval, avoiding any interference between different
tasks (bandwidth isolation).
Thanks to this feature, also tasks that do not strictly comply with
the computational model sketched above can effectively use the new
policy.
To summarize, this patch:
- introduces the data structures, constants and symbols needed;
- implements the core logic of the scheduling algorithm in the new
scheduling class file;
- provides all the glue code between the new scheduling class and
the core scheduler and refines the interactions between sched/dl
and the other existing scheduling classes.
Signed-off-by: Dario Faggioli <raistlin@linux.it>
Signed-off-by: Michael Trimarchi <michael@amarulasolutions.com>
Signed-off-by: Fabio Checconi <fchecconi@gmail.com>
Signed-off-by: Juri Lelli <juri.lelli@gmail.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1383831828-15501-4-git-send-email-juri.lelli@gmail.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2013-11-28 10:14:43 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Actual scheduling parameters. Initialized with the values above,
|
2018-12-03 09:05:56 +00:00
|
|
|
* they are continuously updated during task execution. Note that
|
sched/deadline: Add SCHED_DEADLINE structures & implementation
Introduces the data structures, constants and symbols needed for
SCHED_DEADLINE implementation.
Core data structure of SCHED_DEADLINE are defined, along with their
initializers. Hooks for checking if a task belong to the new policy
are also added where they are needed.
Adds a scheduling class, in sched/dl.c and a new policy called
SCHED_DEADLINE. It is an implementation of the Earliest Deadline
First (EDF) scheduling algorithm, augmented with a mechanism (called
Constant Bandwidth Server, CBS) that makes it possible to isolate
the behaviour of tasks between each other.
The typical -deadline task will be made up of a computation phase
(instance) which is activated on a periodic or sporadic fashion. The
expected (maximum) duration of such computation is called the task's
runtime; the time interval by which each instance need to be completed
is called the task's relative deadline. The task's absolute deadline
is dynamically calculated as the time instant a task (better, an
instance) activates plus the relative deadline.
The EDF algorithms selects the task with the smallest absolute
deadline as the one to be executed first, while the CBS ensures each
task to run for at most its runtime every (relative) deadline
length time interval, avoiding any interference between different
tasks (bandwidth isolation).
Thanks to this feature, also tasks that do not strictly comply with
the computational model sketched above can effectively use the new
policy.
To summarize, this patch:
- introduces the data structures, constants and symbols needed;
- implements the core logic of the scheduling algorithm in the new
scheduling class file;
- provides all the glue code between the new scheduling class and
the core scheduler and refines the interactions between sched/dl
and the other existing scheduling classes.
Signed-off-by: Dario Faggioli <raistlin@linux.it>
Signed-off-by: Michael Trimarchi <michael@amarulasolutions.com>
Signed-off-by: Fabio Checconi <fchecconi@gmail.com>
Signed-off-by: Juri Lelli <juri.lelli@gmail.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1383831828-15501-4-git-send-email-juri.lelli@gmail.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2013-11-28 10:14:43 +00:00
|
|
|
* the remaining runtime could be < 0 in case we are in overrun.
|
|
|
|
*/
|
2017-02-06 21:06:35 +00:00
|
|
|
s64 runtime; /* Remaining runtime for this instance */
|
|
|
|
u64 deadline; /* Absolute deadline for this instance */
|
|
|
|
unsigned int flags; /* Specifying the scheduler behaviour */
|
sched/deadline: Add SCHED_DEADLINE structures & implementation
Introduces the data structures, constants and symbols needed for
SCHED_DEADLINE implementation.
Core data structure of SCHED_DEADLINE are defined, along with their
initializers. Hooks for checking if a task belong to the new policy
are also added where they are needed.
Adds a scheduling class, in sched/dl.c and a new policy called
SCHED_DEADLINE. It is an implementation of the Earliest Deadline
First (EDF) scheduling algorithm, augmented with a mechanism (called
Constant Bandwidth Server, CBS) that makes it possible to isolate
the behaviour of tasks between each other.
The typical -deadline task will be made up of a computation phase
(instance) which is activated on a periodic or sporadic fashion. The
expected (maximum) duration of such computation is called the task's
runtime; the time interval by which each instance need to be completed
is called the task's relative deadline. The task's absolute deadline
is dynamically calculated as the time instant a task (better, an
instance) activates plus the relative deadline.
The EDF algorithms selects the task with the smallest absolute
deadline as the one to be executed first, while the CBS ensures each
task to run for at most its runtime every (relative) deadline
length time interval, avoiding any interference between different
tasks (bandwidth isolation).
Thanks to this feature, also tasks that do not strictly comply with
the computational model sketched above can effectively use the new
policy.
To summarize, this patch:
- introduces the data structures, constants and symbols needed;
- implements the core logic of the scheduling algorithm in the new
scheduling class file;
- provides all the glue code between the new scheduling class and
the core scheduler and refines the interactions between sched/dl
and the other existing scheduling classes.
Signed-off-by: Dario Faggioli <raistlin@linux.it>
Signed-off-by: Michael Trimarchi <michael@amarulasolutions.com>
Signed-off-by: Fabio Checconi <fchecconi@gmail.com>
Signed-off-by: Juri Lelli <juri.lelli@gmail.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1383831828-15501-4-git-send-email-juri.lelli@gmail.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2013-11-28 10:14:43 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Some bool flags:
|
|
|
|
*
|
|
|
|
* @dl_throttled tells if we exhausted the runtime. If so, the
|
|
|
|
* task has to wait for a replenishment to be performed at the
|
|
|
|
* next firing of dl_timer.
|
|
|
|
*
|
sched/deadline: Add SCHED_DEADLINE inheritance logic
Some method to deal with rt-mutexes and make sched_dl interact with
the current PI-coded is needed, raising all but trivial issues, that
needs (according to us) to be solved with some restructuring of
the pi-code (i.e., going toward a proxy execution-ish implementation).
This is under development, in the meanwhile, as a temporary solution,
what this commits does is:
- ensure a pi-lock owner with waiters is never throttled down. Instead,
when it runs out of runtime, it immediately gets replenished and it's
deadline is postponed;
- the scheduling parameters (relative deadline and default runtime)
used for that replenishments --during the whole period it holds the
pi-lock-- are the ones of the waiting task with earliest deadline.
Acting this way, we provide some kind of boosting to the lock-owner,
still by using the existing (actually, slightly modified by the previous
commit) pi-architecture.
We would stress the fact that this is only a surely needed, all but
clean solution to the problem. In the end it's only a way to re-start
discussion within the community. So, as always, comments, ideas, rants,
etc.. are welcome! :-)
Signed-off-by: Dario Faggioli <raistlin@linux.it>
Signed-off-by: Juri Lelli <juri.lelli@gmail.com>
[ Added !RT_MUTEXES build fix. ]
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1383831828-15501-11-git-send-email-juri.lelli@gmail.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2013-11-07 13:43:44 +00:00
|
|
|
* @dl_boosted tells if we are boosted due to DI. If so we are
|
|
|
|
* outside bandwidth enforcement mechanism (but only until we
|
2014-04-15 11:49:04 +00:00
|
|
|
* exit the critical section);
|
|
|
|
*
|
2017-02-06 21:06:35 +00:00
|
|
|
* @dl_yielded tells if task gave up the CPU before consuming
|
2014-04-15 11:49:04 +00:00
|
|
|
* all its available runtime during the last job.
|
2017-05-18 20:13:29 +00:00
|
|
|
*
|
|
|
|
* @dl_non_contending tells if the task is inactive while still
|
|
|
|
* contributing to the active utilization. In other words, it
|
|
|
|
* indicates if the inactive timer has been armed and its handler
|
|
|
|
* has not been executed yet. This flag is useful to avoid race
|
|
|
|
* conditions between the inactive timer handler and the wakeup
|
|
|
|
* code.
|
2017-12-12 11:10:24 +00:00
|
|
|
*
|
|
|
|
* @dl_overrun tells if the task asked to be informed about runtime
|
|
|
|
* overruns.
|
sched/deadline: Add SCHED_DEADLINE structures & implementation
Introduces the data structures, constants and symbols needed for
SCHED_DEADLINE implementation.
Core data structure of SCHED_DEADLINE are defined, along with their
initializers. Hooks for checking if a task belong to the new policy
are also added where they are needed.
Adds a scheduling class, in sched/dl.c and a new policy called
SCHED_DEADLINE. It is an implementation of the Earliest Deadline
First (EDF) scheduling algorithm, augmented with a mechanism (called
Constant Bandwidth Server, CBS) that makes it possible to isolate
the behaviour of tasks between each other.
The typical -deadline task will be made up of a computation phase
(instance) which is activated on a periodic or sporadic fashion. The
expected (maximum) duration of such computation is called the task's
runtime; the time interval by which each instance need to be completed
is called the task's relative deadline. The task's absolute deadline
is dynamically calculated as the time instant a task (better, an
instance) activates plus the relative deadline.
The EDF algorithms selects the task with the smallest absolute
deadline as the one to be executed first, while the CBS ensures each
task to run for at most its runtime every (relative) deadline
length time interval, avoiding any interference between different
tasks (bandwidth isolation).
Thanks to this feature, also tasks that do not strictly comply with
the computational model sketched above can effectively use the new
policy.
To summarize, this patch:
- introduces the data structures, constants and symbols needed;
- implements the core logic of the scheduling algorithm in the new
scheduling class file;
- provides all the glue code between the new scheduling class and
the core scheduler and refines the interactions between sched/dl
and the other existing scheduling classes.
Signed-off-by: Dario Faggioli <raistlin@linux.it>
Signed-off-by: Michael Trimarchi <michael@amarulasolutions.com>
Signed-off-by: Fabio Checconi <fchecconi@gmail.com>
Signed-off-by: Juri Lelli <juri.lelli@gmail.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1383831828-15501-4-git-send-email-juri.lelli@gmail.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2013-11-28 10:14:43 +00:00
|
|
|
*/
|
2017-10-13 07:01:22 +00:00
|
|
|
unsigned int dl_throttled : 1;
|
|
|
|
unsigned int dl_boosted : 1;
|
|
|
|
unsigned int dl_yielded : 1;
|
|
|
|
unsigned int dl_non_contending : 1;
|
2017-12-12 11:10:24 +00:00
|
|
|
unsigned int dl_overrun : 1;
|
sched/deadline: Add SCHED_DEADLINE structures & implementation
Introduces the data structures, constants and symbols needed for
SCHED_DEADLINE implementation.
Core data structure of SCHED_DEADLINE are defined, along with their
initializers. Hooks for checking if a task belong to the new policy
are also added where they are needed.
Adds a scheduling class, in sched/dl.c and a new policy called
SCHED_DEADLINE. It is an implementation of the Earliest Deadline
First (EDF) scheduling algorithm, augmented with a mechanism (called
Constant Bandwidth Server, CBS) that makes it possible to isolate
the behaviour of tasks between each other.
The typical -deadline task will be made up of a computation phase
(instance) which is activated on a periodic or sporadic fashion. The
expected (maximum) duration of such computation is called the task's
runtime; the time interval by which each instance need to be completed
is called the task's relative deadline. The task's absolute deadline
is dynamically calculated as the time instant a task (better, an
instance) activates plus the relative deadline.
The EDF algorithms selects the task with the smallest absolute
deadline as the one to be executed first, while the CBS ensures each
task to run for at most its runtime every (relative) deadline
length time interval, avoiding any interference between different
tasks (bandwidth isolation).
Thanks to this feature, also tasks that do not strictly comply with
the computational model sketched above can effectively use the new
policy.
To summarize, this patch:
- introduces the data structures, constants and symbols needed;
- implements the core logic of the scheduling algorithm in the new
scheduling class file;
- provides all the glue code between the new scheduling class and
the core scheduler and refines the interactions between sched/dl
and the other existing scheduling classes.
Signed-off-by: Dario Faggioli <raistlin@linux.it>
Signed-off-by: Michael Trimarchi <michael@amarulasolutions.com>
Signed-off-by: Fabio Checconi <fchecconi@gmail.com>
Signed-off-by: Juri Lelli <juri.lelli@gmail.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1383831828-15501-4-git-send-email-juri.lelli@gmail.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2013-11-28 10:14:43 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Bandwidth enforcement timer. Each -deadline task has its
|
|
|
|
* own bandwidth to be enforced, thus we need one timer per task.
|
|
|
|
*/
|
2017-02-06 21:06:35 +00:00
|
|
|
struct hrtimer dl_timer;
|
2017-05-18 20:13:29 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Inactive timer, responsible for decreasing the active utilization
|
|
|
|
* at the "0-lag time". When a -deadline task blocks, it contributes
|
|
|
|
* to GRUB's active utilization until the "0-lag time", hence a
|
|
|
|
* timer is needed to decrease the active utilization at the correct
|
|
|
|
* time.
|
|
|
|
*/
|
|
|
|
struct hrtimer inactive_timer;
|
sched/deadline: Add SCHED_DEADLINE structures & implementation
Introduces the data structures, constants and symbols needed for
SCHED_DEADLINE implementation.
Core data structure of SCHED_DEADLINE are defined, along with their
initializers. Hooks for checking if a task belong to the new policy
are also added where they are needed.
Adds a scheduling class, in sched/dl.c and a new policy called
SCHED_DEADLINE. It is an implementation of the Earliest Deadline
First (EDF) scheduling algorithm, augmented with a mechanism (called
Constant Bandwidth Server, CBS) that makes it possible to isolate
the behaviour of tasks between each other.
The typical -deadline task will be made up of a computation phase
(instance) which is activated on a periodic or sporadic fashion. The
expected (maximum) duration of such computation is called the task's
runtime; the time interval by which each instance need to be completed
is called the task's relative deadline. The task's absolute deadline
is dynamically calculated as the time instant a task (better, an
instance) activates plus the relative deadline.
The EDF algorithms selects the task with the smallest absolute
deadline as the one to be executed first, while the CBS ensures each
task to run for at most its runtime every (relative) deadline
length time interval, avoiding any interference between different
tasks (bandwidth isolation).
Thanks to this feature, also tasks that do not strictly comply with
the computational model sketched above can effectively use the new
policy.
To summarize, this patch:
- introduces the data structures, constants and symbols needed;
- implements the core logic of the scheduling algorithm in the new
scheduling class file;
- provides all the glue code between the new scheduling class and
the core scheduler and refines the interactions between sched/dl
and the other existing scheduling classes.
Signed-off-by: Dario Faggioli <raistlin@linux.it>
Signed-off-by: Michael Trimarchi <michael@amarulasolutions.com>
Signed-off-by: Fabio Checconi <fchecconi@gmail.com>
Signed-off-by: Juri Lelli <juri.lelli@gmail.com>
Signed-off-by: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/1383831828-15501-4-git-send-email-juri.lelli@gmail.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
2013-11-28 10:14:43 +00:00
|
|
|
};
|
2013-02-07 15:47:07 +00:00
|
|
|
|
2019-06-21 08:42:02 +00:00
|
|
|
#ifdef CONFIG_UCLAMP_TASK
|
|
|
|
/* Number of utilization clamp buckets (shorter alias) */
|
|
|
|
#define UCLAMP_BUCKETS CONFIG_UCLAMP_BUCKETS_COUNT
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Utilization clamp for a scheduling entity
|
|
|
|
* @value: clamp value "assigned" to a se
|
|
|
|
* @bucket_id: bucket index corresponding to the "assigned" value
|
2019-06-21 08:42:05 +00:00
|
|
|
* @active: the se is currently refcounted in a rq's bucket
|
2019-06-21 08:42:07 +00:00
|
|
|
* @user_defined: the requested clamp value comes from user-space
|
2019-06-21 08:42:02 +00:00
|
|
|
*
|
|
|
|
* The bucket_id is the index of the clamp bucket matching the clamp value
|
|
|
|
* which is pre-computed and stored to avoid expensive integer divisions from
|
|
|
|
* the fast path.
|
2019-06-21 08:42:05 +00:00
|
|
|
*
|
|
|
|
* The active bit is set whenever a task has got an "effective" value assigned,
|
|
|
|
* which can be different from the clamp value "requested" from user-space.
|
|
|
|
* This allows to know a task is refcounted in the rq's bucket corresponding
|
|
|
|
* to the "effective" bucket_id.
|
2019-06-21 08:42:07 +00:00
|
|
|
*
|
|
|
|
* The user_defined bit is set whenever a task has got a task-specific clamp
|
|
|
|
* value requested from userspace, i.e. the system defaults apply to this task
|
|
|
|
* just as a restriction. This allows to relax default clamps when a less
|
|
|
|
* restrictive task-specific value has been requested, thus allowing to
|
|
|
|
* implement a "nice" semantic. For example, a task running with a 20%
|
|
|
|
* default boost can still drop its own boosting to 0%.
|
2019-06-21 08:42:02 +00:00
|
|
|
*/
|
|
|
|
struct uclamp_se {
|
|
|
|
unsigned int value : bits_per(SCHED_CAPACITY_SCALE);
|
|
|
|
unsigned int bucket_id : bits_per(UCLAMP_BUCKETS);
|
2019-06-21 08:42:05 +00:00
|
|
|
unsigned int active : 1;
|
2019-06-21 08:42:07 +00:00
|
|
|
unsigned int user_defined : 1;
|
2019-06-21 08:42:02 +00:00
|
|
|
};
|
|
|
|
#endif /* CONFIG_UCLAMP_TASK */
|
|
|
|
|
2014-08-14 23:01:53 +00:00
|
|
|
union rcu_special {
|
|
|
|
struct {
|
2017-02-06 21:06:35 +00:00
|
|
|
u8 blocked;
|
|
|
|
u8 need_qs;
|
rcu: Speed up expedited GPs when interrupting RCU reader
In PREEMPT kernels, an expedited grace period might send an IPI to a
CPU that is executing an RCU read-side critical section. In that case,
it would be nice if the rcu_read_unlock() directly interacted with the
RCU core code to immediately report the quiescent state. And this does
happen in the case where the reader has been preempted. But it would
also be a nice performance optimization if immediate reporting also
happened in the preemption-free case.
This commit therefore adds an ->exp_hint field to the task_struct structure's
->rcu_read_unlock_special field. The IPI handler sets this hint when
it has interrupted an RCU read-side critical section, and this causes
the outermost rcu_read_unlock() call to invoke rcu_read_unlock_special(),
which, if preemption is enabled, reports the quiescent state immediately.
If preemption is disabled, then the report is required to be deferred
until preemption (or bottom halves or interrupts or whatever) is re-enabled.
Because this is a hint, it does nothing for more complicated cases. For
example, if the IPI interrupts an RCU reader, but interrupts are disabled
across the rcu_read_unlock(), but another rcu_read_lock() is executed
before interrupts are re-enabled, the hint will already have been cleared.
If you do crazy things like this, reporting will be deferred until some
later RCU_SOFTIRQ handler, context switch, cond_resched(), or similar.
Reported-by: Joel Fernandes <joel@joelfernandes.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.ibm.com>
Acked-by: Joel Fernandes (Google) <joel@joelfernandes.org>
2018-10-16 11:12:58 +00:00
|
|
|
u8 exp_hint; /* Hint for performance. */
|
rcu: Check for wakeup-safe conditions in rcu_read_unlock_special()
When RCU core processing is offloaded from RCU_SOFTIRQ to the rcuc
kthreads, a full and unconditional wakeup is required to initiate RCU
core processing. In contrast, when RCU core processing is carried
out by RCU_SOFTIRQ, a raise_softirq() suffices. Of course, there are
situations where raise_softirq() does a full wakeup, but these do not
occur with normal usage of rcu_read_unlock().
The reason that full wakeups can be problematic is that the scheduler
sometimes invokes rcu_read_unlock() with its pi or rq locks held,
which can of course result in deadlock in CONFIG_PREEMPT=y kernels when
rcu_read_unlock() invokes the scheduler. Scheduler invocations can happen
in the following situations: (1) The just-ended reader has been subjected
to RCU priority boosting, in which case rcu_read_unlock() must deboost,
(2) Interrupts were disabled across the call to rcu_read_unlock(), so
the quiescent state must be deferred, requiring a wakeup of the rcuc
kthread corresponding to the current CPU.
Now, the scheduler may hold one of its locks across rcu_read_unlock()
only if preemption has been disabled across the entire RCU read-side
critical section, which in the days prior to RCU flavor consolidation
meant that rcu_read_unlock() never needed to do wakeups. However, this
is no longer the case for any but the first rcu_read_unlock() following a
condition (e.g., preempted RCU reader) requiring special rcu_read_unlock()
attention. For example, an RCU read-side critical section might be
preempted, but preemption might be disabled across the rcu_read_unlock().
The rcu_read_unlock() must defer the quiescent state, and therefore
leaves the task queued on its leaf rcu_node structure. If a scheduler
interrupt occurs, the scheduler might well invoke rcu_read_unlock() with
one of its locks held. However, the preempted task is still queued, so
rcu_read_unlock() will attempt to defer the quiescent state once more.
When RCU core processing is carried out by RCU_SOFTIRQ, this works just
fine: The raise_softirq() function simply sets a bit in a per-CPU mask
and the RCU core processing will be undertaken upon return from interrupt.
Not so when RCU core processing is carried out by the rcuc kthread: In this
case, the required wakeup can result in deadlock.
The initial solution to this problem was to use set_tsk_need_resched() and
set_preempt_need_resched() to force a future context switch, which allows
rcu_preempt_note_context_switch() to report the deferred quiescent state
to RCU's core processing. Unfortunately for expedited grace periods,
there can be a significant delay between the call for a context switch
and the actual context switch.
This commit therefore introduces a ->deferred_qs flag to the task_struct
structure's rcu_special structure. This flag is initially false, and
is set to true by the first call to rcu_read_unlock() requiring special
attention, then finally reset back to false when the quiescent state is
finally reported. Then rcu_read_unlock() attempts full wakeups only when
->deferred_qs is false, that is, on the first rcu_read_unlock() requiring
special attention. Note that a chain of RCU readers linked by some other
sort of reader may find that a later rcu_read_unlock() is once again able
to do a full wakeup, courtesy of an intervening preemption:
rcu_read_lock();
/* preempted */
local_irq_disable();
rcu_read_unlock(); /* Can do full wakeup, sets ->deferred_qs. */
rcu_read_lock();
local_irq_enable();
preempt_disable()
rcu_read_unlock(); /* Cannot do full wakeup, ->deferred_qs set. */
rcu_read_lock();
preempt_enable();
/* preempted, >deferred_qs reset. */
local_irq_disable();
rcu_read_unlock(); /* Can again do full wakeup, sets ->deferred_qs. */
Such linked RCU readers do not yet seem to appear in the Linux kernel, and
it is probably best if they don't. However, RCU needs to handle them, and
some variations on this theme could make even raise_softirq() unsafe due to
the possibility of its doing a full wakeup. This commit therefore also
avoids invoking raise_softirq() when the ->deferred_qs set flag is set.
Signed-off-by: Paul E. McKenney <paulmck@linux.ibm.com>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
2019-03-24 22:25:51 +00:00
|
|
|
u8 deferred_qs;
|
2015-08-02 20:53:17 +00:00
|
|
|
} b; /* Bits. */
|
rcu: Speed up expedited GPs when interrupting RCU reader
In PREEMPT kernels, an expedited grace period might send an IPI to a
CPU that is executing an RCU read-side critical section. In that case,
it would be nice if the rcu_read_unlock() directly interacted with the
RCU core code to immediately report the quiescent state. And this does
happen in the case where the reader has been preempted. But it would
also be a nice performance optimization if immediate reporting also
happened in the preemption-free case.
This commit therefore adds an ->exp_hint field to the task_struct structure's
->rcu_read_unlock_special field. The IPI handler sets this hint when
it has interrupted an RCU read-side critical section, and this causes
the outermost rcu_read_unlock() call to invoke rcu_read_unlock_special(),
which, if preemption is enabled, reports the quiescent state immediately.
If preemption is disabled, then the report is required to be deferred
until preemption (or bottom halves or interrupts or whatever) is re-enabled.
Because this is a hint, it does nothing for more complicated cases. For
example, if the IPI interrupts an RCU reader, but interrupts are disabled
across the rcu_read_unlock(), but another rcu_read_lock() is executed
before interrupts are re-enabled, the hint will already have been cleared.
If you do crazy things like this, reporting will be deferred until some
later RCU_SOFTIRQ handler, context switch, cond_resched(), or similar.
Reported-by: Joel Fernandes <joel@joelfernandes.org>
Signed-off-by: Paul E. McKenney <paulmck@linux.ibm.com>
Acked-by: Joel Fernandes (Google) <joel@joelfernandes.org>
2018-10-16 11:12:58 +00:00
|
|
|
u32 s; /* Set of bits. */
|
2014-08-14 23:01:53 +00:00
|
|
|
};
|
2009-08-27 22:00:12 +00:00
|
|
|
|
2010-09-02 14:50:03 +00:00
|
|
|
enum perf_event_task_context {
|
|
|
|
perf_invalid_context = -1,
|
|
|
|
perf_hw_context = 0,
|
2010-09-07 15:34:50 +00:00
|
|
|
perf_sw_context,
|
2010-09-02 14:50:03 +00:00
|
|
|
perf_nr_task_contexts,
|
|
|
|
};
|
|
|
|
|
2017-02-01 16:09:06 +00:00
|
|
|
struct wake_q_node {
|
|
|
|
struct wake_q_node *next;
|
|
|
|
};
|
|
|
|
|
2005-04-16 22:20:36 +00:00
|
|
|
struct task_struct {
|
2016-09-13 21:29:24 +00:00
|
|
|
#ifdef CONFIG_THREAD_INFO_IN_TASK
|
|
|
|
/*
|
|
|
|
* For reasons of header soup (see current_thread_info()), this
|
|
|
|
* must be the first element of task_struct.
|
|
|
|
*/
|
2017-02-06 21:06:35 +00:00
|
|
|
struct thread_info thread_info;
|
2016-09-13 21:29:24 +00:00
|
|
|
#endif
|
2017-02-06 21:06:35 +00:00
|
|
|
/* -1 unrunnable, 0 runnable, >0 stopped: */
|
|
|
|
volatile long state;
|
2017-04-06 05:43:33 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* This begins the randomizable portion of task_struct. Only
|
|
|
|
* scheduling-critical items should be added above here.
|
|
|
|
*/
|
|
|
|
randomized_struct_fields_start
|
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
void *stack;
|
2019-01-18 12:27:29 +00:00
|
|
|
refcount_t usage;
|
2017-02-06 21:06:35 +00:00
|
|
|
/* Per task flags (PF_*), defined further below: */
|
|
|
|
unsigned int flags;
|
|
|
|
unsigned int ptrace;
|
2005-04-16 22:20:36 +00:00
|
|
|
|
[PATCH] sched: implement smpnice
Problem:
The introduction of separate run queues per CPU has brought with it "nice"
enforcement problems that are best described by a simple example.
For the sake of argument suppose that on a single CPU machine with a
nice==19 hard spinner and a nice==0 hard spinner running that the nice==0
task gets 95% of the CPU and the nice==19 task gets 5% of the CPU. Now
suppose that there is a system with 2 CPUs and 2 nice==19 hard spinners and
2 nice==0 hard spinners running. The user of this system would be entitled
to expect that the nice==0 tasks each get 95% of a CPU and the nice==19
tasks only get 5% each. However, whether this expectation is met is pretty
much down to luck as there are four equally likely distributions of the
tasks to the CPUs that the load balancing code will consider to be balanced
with loads of 2.0 for each CPU. Two of these distributions involve one
nice==0 and one nice==19 task per CPU and in these circumstances the users
expectations will be met. The other two distributions both involve both
nice==0 tasks being on one CPU and both nice==19 being on the other CPU and
each task will get 50% of a CPU and the user's expectations will not be
met.
Solution:
The solution to this problem that is implemented in the attached patch is
to use weighted loads when determining if the system is balanced and, when
an imbalance is detected, to move an amount of weighted load between run
queues (as opposed to a number of tasks) to restore the balance. Once
again, the easiest way to explain why both of these measures are necessary
is to use a simple example. Suppose that (in a slight variation of the
above example) that we have a two CPU system with 4 nice==0 and 4 nice=19
hard spinning tasks running and that the 4 nice==0 tasks are on one CPU and
the 4 nice==19 tasks are on the other CPU. The weighted loads for the two
CPUs would be 4.0 and 0.2 respectively and the load balancing code would
move 2 tasks resulting in one CPU with a load of 2.0 and the other with
load of 2.2. If this was considered to be a big enough imbalance to
justify moving a task and that task was moved using the current
move_tasks() then it would move the highest priority task that it found and
this would result in one CPU with a load of 3.0 and the other with a load
of 1.2 which would result in the movement of a task in the opposite
direction and so on -- infinite loop. If, on the other hand, an amount of
load to be moved is calculated from the imbalance (in this case 0.1) and
move_tasks() skips tasks until it find ones whose contributions to the
weighted load are less than this amount it would move two of the nice==19
tasks resulting in a system with 2 nice==0 and 2 nice=19 on each CPU with
loads of 2.1 for each CPU.
One of the advantages of this mechanism is that on a system where all tasks
have nice==0 the load balancing calculations would be mathematically
identical to the current load balancing code.
Notes:
struct task_struct:
has a new field load_weight which (in a trade off of space for speed)
stores the contribution that this task makes to a CPU's weighted load when
it is runnable.
struct runqueue:
has a new field raw_weighted_load which is the sum of the load_weight
values for the currently runnable tasks on this run queue. This field
always needs to be updated when nr_running is updated so two new inline
functions inc_nr_running() and dec_nr_running() have been created to make
sure that this happens. This also offers a convenient way to optimize away
this part of the smpnice mechanism when CONFIG_SMP is not defined.
int try_to_wake_up():
in this function the value SCHED_LOAD_BALANCE is used to represent the load
contribution of a single task in various calculations in the code that
decides which CPU to put the waking task on. While this would be a valid
on a system where the nice values for the runnable tasks were distributed
evenly around zero it will lead to anomalous load balancing if the
distribution is skewed in either direction. To overcome this problem
SCHED_LOAD_SCALE has been replaced by the load_weight for the relevant task
or by the average load_weight per task for the queue in question (as
appropriate).
int move_tasks():
The modifications to this function were complicated by the fact that
active_load_balance() uses it to move exactly one task without checking
whether an imbalance actually exists. This precluded the simple
overloading of max_nr_move with max_load_move and necessitated the addition
of the latter as an extra argument to the function. The internal
implementation is then modified to move up to max_nr_move tasks and
max_load_move of weighted load. This slightly complicates the code where
move_tasks() is called and if ever active_load_balance() is changed to not
use move_tasks() the implementation of move_tasks() should be simplified
accordingly.
struct sched_group *find_busiest_group():
Similar to try_to_wake_up(), there are places in this function where
SCHED_LOAD_SCALE is used to represent the load contribution of a single
task and the same issues are created. A similar solution is adopted except
that it is now the average per task contribution to a group's load (as
opposed to a run queue) that is required. As this value is not directly
available from the group it is calculated on the fly as the queues in the
groups are visited when determining the busiest group.
A key change to this function is that it is no longer to scale down
*imbalance on exit as move_tasks() uses the load in its scaled form.
void set_user_nice():
has been modified to update the task's load_weight field when it's nice
value and also to ensure that its run queue's raw_weighted_load field is
updated if it was runnable.
From: "Siddha, Suresh B" <suresh.b.siddha@intel.com>
With smpnice, sched groups with highest priority tasks can mask the imbalance
between the other sched groups with in the same domain. This patch fixes some
of the listed down scenarios by not considering the sched groups which are
lightly loaded.
a) on a simple 4-way MP system, if we have one high priority and 4 normal
priority tasks, with smpnice we would like to see the high priority task
scheduled on one cpu, two other cpus getting one normal task each and the
fourth cpu getting the remaining two normal tasks. but with current
smpnice extra normal priority task keeps jumping from one cpu to another
cpu having the normal priority task. This is because of the
busiest_has_loaded_cpus, nr_loaded_cpus logic.. We are not including the
cpu with high priority task in max_load calculations but including that in
total and avg_load calcuations.. leading to max_load < avg_load and load
balance between cpus running normal priority tasks(2 Vs 1) will always show
imbalanace as one normal priority and the extra normal priority task will
keep moving from one cpu to another cpu having normal priority task..
b) 4-way system with HT (8 logical processors). Package-P0 T0 has a
highest priority task, T1 is idle. Package-P1 Both T0 and T1 have 1 normal
priority task each.. P2 and P3 are idle. With this patch, one of the
normal priority tasks on P1 will be moved to P2 or P3..
c) With the current weighted smp nice calculations, it doesn't always make
sense to look at the highest weighted runqueue in the busy group..
Consider a load balance scenario on a DP with HT system, with Package-0
containing one high priority and one low priority, Package-1 containing one
low priority(with other thread being idle).. Package-1 thinks that it need
to take the low priority thread from Package-0. And find_busiest_queue()
returns the cpu thread with highest priority task.. And ultimately(with
help of active load balance) we move high priority task to Package-1. And
same continues with Package-0 now, moving high priority task from package-1
to package-0.. Even without the presence of active load balance, load
balance will fail to balance the above scenario.. Fix find_busiest_queue
to use "imbalance" when it is lightly loaded.
[kernel@kolivas.org: sched: store weighted load on up]
[kernel@kolivas.org: sched: add discrete weighted cpu load function]
[suresh.b.siddha@intel.com: sched: remove dead code]
Signed-off-by: Peter Williams <pwil3058@bigpond.com.au>
Cc: "Siddha, Suresh B" <suresh.b.siddha@intel.com>
Cc: "Chen, Kenneth W" <kenneth.w.chen@intel.com>
Acked-by: Ingo Molnar <mingo@elte.hu>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Signed-off-by: Con Kolivas <kernel@kolivas.org>
Cc: John Hawkes <hawkes@sgi.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-06-27 09:54:34 +00:00
|
|
|
#ifdef CONFIG_SMP
|
2017-02-06 21:06:35 +00:00
|
|
|
struct llist_node wake_entry;
|
|
|
|
int on_cpu;
|
2016-09-13 21:29:24 +00:00
|
|
|
#ifdef CONFIG_THREAD_INFO_IN_TASK
|
2017-02-06 21:06:35 +00:00
|
|
|
/* Current CPU: */
|
|
|
|
unsigned int cpu;
|
2016-09-13 21:29:24 +00:00
|
|
|
#endif
|
2017-02-06 21:06:35 +00:00
|
|
|
unsigned int wakee_flips;
|
|
|
|
unsigned long wakee_flip_decay_ts;
|
|
|
|
struct task_struct *last_wakee;
|
2013-10-07 10:29:16 +00:00
|
|
|
|
2018-01-30 10:45:55 +00:00
|
|
|
/*
|
|
|
|
* recent_used_cpu is initially set as the last CPU used by a task
|
|
|
|
* that wakes affine another task. Waker/wakee relationships can
|
|
|
|
* push tasks around a CPU where each wakeup moves to the next one.
|
|
|
|
* Tracking a recently used CPU allows a quick search for a recently
|
|
|
|
* used CPU that may be idle.
|
|
|
|
*/
|
|
|
|
int recent_used_cpu;
|
2017-02-06 21:06:35 +00:00
|
|
|
int wake_cpu;
|
[PATCH] sched: implement smpnice
Problem:
The introduction of separate run queues per CPU has brought with it "nice"
enforcement problems that are best described by a simple example.
For the sake of argument suppose that on a single CPU machine with a
nice==19 hard spinner and a nice==0 hard spinner running that the nice==0
task gets 95% of the CPU and the nice==19 task gets 5% of the CPU. Now
suppose that there is a system with 2 CPUs and 2 nice==19 hard spinners and
2 nice==0 hard spinners running. The user of this system would be entitled
to expect that the nice==0 tasks each get 95% of a CPU and the nice==19
tasks only get 5% each. However, whether this expectation is met is pretty
much down to luck as there are four equally likely distributions of the
tasks to the CPUs that the load balancing code will consider to be balanced
with loads of 2.0 for each CPU. Two of these distributions involve one
nice==0 and one nice==19 task per CPU and in these circumstances the users
expectations will be met. The other two distributions both involve both
nice==0 tasks being on one CPU and both nice==19 being on the other CPU and
each task will get 50% of a CPU and the user's expectations will not be
met.
Solution:
The solution to this problem that is implemented in the attached patch is
to use weighted loads when determining if the system is balanced and, when
an imbalance is detected, to move an amount of weighted load between run
queues (as opposed to a number of tasks) to restore the balance. Once
again, the easiest way to explain why both of these measures are necessary
is to use a simple example. Suppose that (in a slight variation of the
above example) that we have a two CPU system with 4 nice==0 and 4 nice=19
hard spinning tasks running and that the 4 nice==0 tasks are on one CPU and
the 4 nice==19 tasks are on the other CPU. The weighted loads for the two
CPUs would be 4.0 and 0.2 respectively and the load balancing code would
move 2 tasks resulting in one CPU with a load of 2.0 and the other with
load of 2.2. If this was considered to be a big enough imbalance to
justify moving a task and that task was moved using the current
move_tasks() then it would move the highest priority task that it found and
this would result in one CPU with a load of 3.0 and the other with a load
of 1.2 which would result in the movement of a task in the opposite
direction and so on -- infinite loop. If, on the other hand, an amount of
load to be moved is calculated from the imbalance (in this case 0.1) and
move_tasks() skips tasks until it find ones whose contributions to the
weighted load are less than this amount it would move two of the nice==19
tasks resulting in a system with 2 nice==0 and 2 nice=19 on each CPU with
loads of 2.1 for each CPU.
One of the advantages of this mechanism is that on a system where all tasks
have nice==0 the load balancing calculations would be mathematically
identical to the current load balancing code.
Notes:
struct task_struct:
has a new field load_weight which (in a trade off of space for speed)
stores the contribution that this task makes to a CPU's weighted load when
it is runnable.
struct runqueue:
has a new field raw_weighted_load which is the sum of the load_weight
values for the currently runnable tasks on this run queue. This field
always needs to be updated when nr_running is updated so two new inline
functions inc_nr_running() and dec_nr_running() have been created to make
sure that this happens. This also offers a convenient way to optimize away
this part of the smpnice mechanism when CONFIG_SMP is not defined.
int try_to_wake_up():
in this function the value SCHED_LOAD_BALANCE is used to represent the load
contribution of a single task in various calculations in the code that
decides which CPU to put the waking task on. While this would be a valid
on a system where the nice values for the runnable tasks were distributed
evenly around zero it will lead to anomalous load balancing if the
distribution is skewed in either direction. To overcome this problem
SCHED_LOAD_SCALE has been replaced by the load_weight for the relevant task
or by the average load_weight per task for the queue in question (as
appropriate).
int move_tasks():
The modifications to this function were complicated by the fact that
active_load_balance() uses it to move exactly one task without checking
whether an imbalance actually exists. This precluded the simple
overloading of max_nr_move with max_load_move and necessitated the addition
of the latter as an extra argument to the function. The internal
implementation is then modified to move up to max_nr_move tasks and
max_load_move of weighted load. This slightly complicates the code where
move_tasks() is called and if ever active_load_balance() is changed to not
use move_tasks() the implementation of move_tasks() should be simplified
accordingly.
struct sched_group *find_busiest_group():
Similar to try_to_wake_up(), there are places in this function where
SCHED_LOAD_SCALE is used to represent the load contribution of a single
task and the same issues are created. A similar solution is adopted except
that it is now the average per task contribution to a group's load (as
opposed to a run queue) that is required. As this value is not directly
available from the group it is calculated on the fly as the queues in the
groups are visited when determining the busiest group.
A key change to this function is that it is no longer to scale down
*imbalance on exit as move_tasks() uses the load in its scaled form.
void set_user_nice():
has been modified to update the task's load_weight field when it's nice
value and also to ensure that its run queue's raw_weighted_load field is
updated if it was runnable.
From: "Siddha, Suresh B" <suresh.b.siddha@intel.com>
With smpnice, sched groups with highest priority tasks can mask the imbalance
between the other sched groups with in the same domain. This patch fixes some
of the listed down scenarios by not considering the sched groups which are
lightly loaded.
a) on a simple 4-way MP system, if we have one high priority and 4 normal
priority tasks, with smpnice we would like to see the high priority task
scheduled on one cpu, two other cpus getting one normal task each and the
fourth cpu getting the remaining two normal tasks. but with current
smpnice extra normal priority task keeps jumping from one cpu to another
cpu having the normal priority task. This is because of the
busiest_has_loaded_cpus, nr_loaded_cpus logic.. We are not including the
cpu with high priority task in max_load calculations but including that in
total and avg_load calcuations.. leading to max_load < avg_load and load
balance between cpus running normal priority tasks(2 Vs 1) will always show
imbalanace as one normal priority and the extra normal priority task will
keep moving from one cpu to another cpu having normal priority task..
b) 4-way system with HT (8 logical processors). Package-P0 T0 has a
highest priority task, T1 is idle. Package-P1 Both T0 and T1 have 1 normal
priority task each.. P2 and P3 are idle. With this patch, one of the
normal priority tasks on P1 will be moved to P2 or P3..
c) With the current weighted smp nice calculations, it doesn't always make
sense to look at the highest weighted runqueue in the busy group..
Consider a load balance scenario on a DP with HT system, with Package-0
containing one high priority and one low priority, Package-1 containing one
low priority(with other thread being idle).. Package-1 thinks that it need
to take the low priority thread from Package-0. And find_busiest_queue()
returns the cpu thread with highest priority task.. And ultimately(with
help of active load balance) we move high priority task to Package-1. And
same continues with Package-0 now, moving high priority task from package-1
to package-0.. Even without the presence of active load balance, load
balance will fail to balance the above scenario.. Fix find_busiest_queue
to use "imbalance" when it is lightly loaded.
[kernel@kolivas.org: sched: store weighted load on up]
[kernel@kolivas.org: sched: add discrete weighted cpu load function]
[suresh.b.siddha@intel.com: sched: remove dead code]
Signed-off-by: Peter Williams <pwil3058@bigpond.com.au>
Cc: "Siddha, Suresh B" <suresh.b.siddha@intel.com>
Cc: "Chen, Kenneth W" <kenneth.w.chen@intel.com>
Acked-by: Ingo Molnar <mingo@elte.hu>
Cc: Nick Piggin <nickpiggin@yahoo.com.au>
Signed-off-by: Con Kolivas <kernel@kolivas.org>
Cc: John Hawkes <hawkes@sgi.com>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-06-27 09:54:34 +00:00
|
|
|
#endif
|
2017-02-06 21:06:35 +00:00
|
|
|
int on_rq;
|
|
|
|
|
|
|
|
int prio;
|
|
|
|
int static_prio;
|
|
|
|
int normal_prio;
|
|
|
|
unsigned int rt_priority;
|
2007-07-09 16:52:00 +00:00
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
const struct sched_class *sched_class;
|
|
|
|
struct sched_entity se;
|
|
|
|
struct sched_rt_entity rt;
|
2012-06-22 11:36:05 +00:00
|
|
|
#ifdef CONFIG_CGROUP_SCHED
|
2017-02-06 21:06:35 +00:00
|
|
|
struct task_group *sched_task_group;
|
2012-06-22 11:36:05 +00:00
|
|
|
#endif
|
2017-02-06 21:06:35 +00:00
|
|
|
struct sched_dl_entity dl;
|
2005-04-16 22:20:36 +00:00
|
|
|
|
2019-06-21 08:42:02 +00:00
|
|
|
#ifdef CONFIG_UCLAMP_TASK
|
2019-06-21 08:42:05 +00:00
|
|
|
/* Clamp values requested for a scheduling entity */
|
|
|
|
struct uclamp_se uclamp_req[UCLAMP_CNT];
|
|
|
|
/* Effective clamp values used for a scheduling entity */
|
2019-06-21 08:42:02 +00:00
|
|
|
struct uclamp_se uclamp[UCLAMP_CNT];
|
|
|
|
#endif
|
|
|
|
|
2007-07-26 11:40:43 +00:00
|
|
|
#ifdef CONFIG_PREEMPT_NOTIFIERS
|
2017-02-06 21:06:35 +00:00
|
|
|
/* List of struct preempt_notifier: */
|
|
|
|
struct hlist_head preempt_notifiers;
|
2007-07-26 11:40:43 +00:00
|
|
|
#endif
|
|
|
|
|
2006-09-29 08:59:40 +00:00
|
|
|
#ifdef CONFIG_BLK_DEV_IO_TRACE
|
2017-02-06 21:06:35 +00:00
|
|
|
unsigned int btrace_seq;
|
2006-09-29 08:59:40 +00:00
|
|
|
#endif
|
2005-04-16 22:20:36 +00:00
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
unsigned int policy;
|
|
|
|
int nr_cpus_allowed;
|
2019-04-23 14:26:36 +00:00
|
|
|
const cpumask_t *cpus_ptr;
|
|
|
|
cpumask_t cpus_mask;
|
2005-04-16 22:20:36 +00:00
|
|
|
|
2010-06-29 23:49:16 +00:00
|
|
|
#ifdef CONFIG_PREEMPT_RCU
|
2017-02-06 21:06:35 +00:00
|
|
|
int rcu_read_lock_nesting;
|
|
|
|
union rcu_special rcu_read_unlock_special;
|
|
|
|
struct list_head rcu_node_entry;
|
|
|
|
struct rcu_node *rcu_blocked_node;
|
2014-09-22 18:00:48 +00:00
|
|
|
#endif /* #ifdef CONFIG_PREEMPT_RCU */
|
2017-02-06 21:06:35 +00:00
|
|
|
|
2014-06-27 20:42:20 +00:00
|
|
|
#ifdef CONFIG_TASKS_RCU
|
2017-02-06 21:06:35 +00:00
|
|
|
unsigned long rcu_tasks_nvcsw;
|
2017-05-25 15:51:48 +00:00
|
|
|
u8 rcu_tasks_holdout;
|
|
|
|
u8 rcu_tasks_idx;
|
2017-02-06 21:06:35 +00:00
|
|
|
int rcu_tasks_idle_cpu;
|
2017-05-25 15:51:48 +00:00
|
|
|
struct list_head rcu_tasks_holdout_list;
|
2014-06-27 20:42:20 +00:00
|
|
|
#endif /* #ifdef CONFIG_TASKS_RCU */
|
2008-01-25 20:08:24 +00:00
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
struct sched_info sched_info;
|
2005-04-16 22:20:36 +00:00
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
struct list_head tasks;
|
2010-11-30 18:51:33 +00:00
|
|
|
#ifdef CONFIG_SMP
|
2017-02-06 21:06:35 +00:00
|
|
|
struct plist_node pushable_tasks;
|
|
|
|
struct rb_node pushable_dl_tasks;
|
2010-11-30 18:51:33 +00:00
|
|
|
#endif
|
2005-04-16 22:20:36 +00:00
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
struct mm_struct *mm;
|
|
|
|
struct mm_struct *active_mm;
|
2017-02-03 10:03:31 +00:00
|
|
|
|
|
|
|
/* Per-thread vma caching: */
|
2017-02-06 21:06:35 +00:00
|
|
|
struct vmacache vmacache;
|
2017-02-03 10:03:31 +00:00
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
#ifdef SPLIT_RSS_COUNTING
|
|
|
|
struct task_rss_stat rss_stat;
|
2010-03-05 21:41:40 +00:00
|
|
|
#endif
|
2017-02-06 21:06:35 +00:00
|
|
|
int exit_state;
|
|
|
|
int exit_code;
|
|
|
|
int exit_signal;
|
|
|
|
/* The signal sent when the parent dies: */
|
|
|
|
int pdeath_signal;
|
|
|
|
/* JOBCTL_*, siglock protected: */
|
|
|
|
unsigned long jobctl;
|
|
|
|
|
|
|
|
/* Used for emulating ABI behavior of previous Linux versions: */
|
|
|
|
unsigned int personality;
|
|
|
|
|
|
|
|
/* Scheduler bits, serialized by scheduler locks: */
|
|
|
|
unsigned sched_reset_on_fork:1;
|
|
|
|
unsigned sched_contributes_to_load:1;
|
|
|
|
unsigned sched_migrated:1;
|
|
|
|
unsigned sched_remote_wakeup:1;
|
psi: pressure stall information for CPU, memory, and IO
When systems are overcommitted and resources become contended, it's hard
to tell exactly the impact this has on workload productivity, or how close
the system is to lockups and OOM kills. In particular, when machines work
multiple jobs concurrently, the impact of overcommit in terms of latency
and throughput on the individual job can be enormous.
In order to maximize hardware utilization without sacrificing individual
job health or risk complete machine lockups, this patch implements a way
to quantify resource pressure in the system.
A kernel built with CONFIG_PSI=y creates files in /proc/pressure/ that
expose the percentage of time the system is stalled on CPU, memory, or IO,
respectively. Stall states are aggregate versions of the per-task delay
accounting delays:
cpu: some tasks are runnable but not executing on a CPU
memory: tasks are reclaiming, or waiting for swapin or thrashing cache
io: tasks are waiting for io completions
These percentages of walltime can be thought of as pressure percentages,
and they give a general sense of system health and productivity loss
incurred by resource overcommit. They can also indicate when the system
is approaching lockup scenarios and OOMs.
To do this, psi keeps track of the task states associated with each CPU
and samples the time they spend in stall states. Every 2 seconds, the
samples are averaged across CPUs - weighted by the CPUs' non-idle time to
eliminate artifacts from unused CPUs - and translated into percentages of
walltime. A running average of those percentages is maintained over 10s,
1m, and 5m periods (similar to the loadaverage).
[hannes@cmpxchg.org: doc fixlet, per Randy]
Link: http://lkml.kernel.org/r/20180828205625.GA14030@cmpxchg.org
[hannes@cmpxchg.org: code optimization]
Link: http://lkml.kernel.org/r/20180907175015.GA8479@cmpxchg.org
[hannes@cmpxchg.org: rename psi_clock() to psi_update_work(), per Peter]
Link: http://lkml.kernel.org/r/20180907145404.GB11088@cmpxchg.org
[hannes@cmpxchg.org: fix build]
Link: http://lkml.kernel.org/r/20180913014222.GA2370@cmpxchg.org
Link: http://lkml.kernel.org/r/20180828172258.3185-9-hannes@cmpxchg.org
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Daniel Drake <drake@endlessm.com>
Tested-by: Suren Baghdasaryan <surenb@google.com>
Cc: Christopher Lameter <cl@linux.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Johannes Weiner <jweiner@fb.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Enderborg <peter.enderborg@sony.com>
Cc: Randy Dunlap <rdunlap@infradead.org>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Vinayak Menon <vinmenon@codeaurora.org>
Cc: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2018-10-26 22:06:27 +00:00
|
|
|
#ifdef CONFIG_PSI
|
|
|
|
unsigned sched_psi_wake_requeue:1;
|
|
|
|
#endif
|
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
/* Force alignment to the next boundary: */
|
|
|
|
unsigned :0;
|
|
|
|
|
|
|
|
/* Unserialized, strictly 'current' */
|
|
|
|
|
|
|
|
/* Bit to tell LSMs we're in execve(): */
|
|
|
|
unsigned in_execve:1;
|
|
|
|
unsigned in_iowait:1;
|
|
|
|
#ifndef TIF_RESTORE_SIGMASK
|
|
|
|
unsigned restore_sigmask:1;
|
signal: consolidate {TS,TLF}_RESTORE_SIGMASK code
In general, there's no need for the "restore sigmask" flag to live in
ti->flags. alpha, ia64, microblaze, powerpc, sh, sparc (64-bit only),
tile, and x86 use essentially identical alternative implementations,
placing the flag in ti->status.
Replace those optimized implementations with an equally good common
implementation that stores it in a bitfield in struct task_struct and
drop the custom implementations.
Additional architectures can opt in by removing their
TIF_RESTORE_SIGMASK defines.
Link: http://lkml.kernel.org/r/8a14321d64a28e40adfddc90e18a96c086a6d6f9.1468522723.git.luto@kernel.org
Signed-off-by: Andy Lutomirski <luto@kernel.org>
Tested-by: Michael Ellerman <mpe@ellerman.id.au> [powerpc]
Cc: Richard Henderson <rth@twiddle.net>
Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: Matt Turner <mattst88@gmail.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Michal Simek <monstr@monstr.eu>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Cc: Rich Felker <dalias@libc.org>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Chris Metcalf <cmetcalf@mellanox.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Borislav Petkov <bp@suse.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Dmitry Safonov <dsafonov@virtuozzo.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-08-02 21:05:36 +00:00
|
|
|
#endif
|
2015-11-06 02:46:09 +00:00
|
|
|
#ifdef CONFIG_MEMCG
|
2018-08-17 22:47:11 +00:00
|
|
|
unsigned in_user_fault:1;
|
2016-01-20 23:02:32 +00:00
|
|
|
#endif
|
2015-04-17 18:05:30 +00:00
|
|
|
#ifdef CONFIG_COMPAT_BRK
|
2017-02-06 21:06:35 +00:00
|
|
|
unsigned brk_randomized:1;
|
2015-04-17 18:05:30 +00:00
|
|
|
#endif
|
cgroup, kthread: close race window where new kthreads can be migrated to non-root cgroups
Creation of a kthread goes through a couple interlocked stages between
the kthread itself and its creator. Once the new kthread starts
running, it initializes itself and wakes up the creator. The creator
then can further configure the kthread and then let it start doing its
job by waking it up.
In this configuration-by-creator stage, the creator is the only one
that can wake it up but the kthread is visible to userland. When
altering the kthread's attributes from userland is allowed, this is
fine; however, for cases where CPU affinity is critical,
kthread_bind() is used to first disable affinity changes from userland
and then set the affinity. This also prevents the kthread from being
migrated into non-root cgroups as that can affect the CPU affinity and
many other things.
Unfortunately, the cgroup side of protection is racy. While the
PF_NO_SETAFFINITY flag prevents further migrations, userland can win
the race before the creator sets the flag with kthread_bind() and put
the kthread in a non-root cgroup, which can lead to all sorts of
problems including incorrect CPU affinity and starvation.
This bug got triggered by userland which periodically tries to migrate
all processes in the root cpuset cgroup to a non-root one. Per-cpu
workqueue workers got caught while being created and ended up with
incorrected CPU affinity breaking concurrency management and sometimes
stalling workqueue execution.
This patch adds task->no_cgroup_migration which disallows the task to
be migrated by userland. kthreadd starts with the flag set making
every child kthread start in the root cgroup with migration
disallowed. The flag is cleared after the kthread finishes
initialization by which time PF_NO_SETAFFINITY is set if the kthread
should stay in the root cgroup.
It'd be better to wait for the initialization instead of failing but I
couldn't think of a way of implementing that without adding either a
new PF flag, or sleeping and retrying from waiting side. Even if
userland depends on changing cgroup membership of a kthread, it either
has to be synchronized with kthread_create() or periodically repeat,
so it's unlikely that this would break anything.
v2: Switch to a simpler implementation using a new task_struct bit
field suggested by Oleg.
Signed-off-by: Tejun Heo <tj@kernel.org>
Suggested-by: Oleg Nesterov <oleg@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Reported-and-debugged-by: Chris Mason <clm@fb.com>
Cc: stable@vger.kernel.org # v4.3+ (we can't close the race on < v4.3)
Signed-off-by: Tejun Heo <tj@kernel.org>
2017-03-16 20:54:24 +00:00
|
|
|
#ifdef CONFIG_CGROUPS
|
|
|
|
/* disallow userland-initiated cgroup migration */
|
|
|
|
unsigned no_cgroup_migration:1;
|
cgroup: cgroup v2 freezer
Cgroup v1 implements the freezer controller, which provides an ability
to stop the workload in a cgroup and temporarily free up some
resources (cpu, io, network bandwidth and, potentially, memory)
for some other tasks. Cgroup v2 lacks this functionality.
This patch implements freezer for cgroup v2.
Cgroup v2 freezer tries to put tasks into a state similar to jobctl
stop. This means that tasks can be killed, ptraced (using
PTRACE_SEIZE*), and interrupted. It is possible to attach to
a frozen task, get some information (e.g. read registers) and detach.
It's also possible to migrate a frozen tasks to another cgroup.
This differs cgroup v2 freezer from cgroup v1 freezer, which mostly
tried to imitate the system-wide freezer. However uninterruptible
sleep is fine when all tasks are going to be frozen (hibernation case),
it's not the acceptable state for some subset of the system.
Cgroup v2 freezer is not supporting freezing kthreads.
If a non-root cgroup contains kthread, the cgroup still can be frozen,
but the kthread will remain running, the cgroup will be shown
as non-frozen, and the notification will not be delivered.
* PTRACE_ATTACH is not working because non-fatal signal delivery
is blocked in frozen state.
There are some interface differences between cgroup v1 and cgroup v2
freezer too, which are required to conform the cgroup v2 interface
design principles:
1) There is no separate controller, which has to be turned on:
the functionality is always available and is represented by
cgroup.freeze and cgroup.events cgroup control files.
2) The desired state is defined by the cgroup.freeze control file.
Any hierarchical configuration is allowed.
3) The interface is asynchronous. The actual state is available
using cgroup.events control file ("frozen" field). There are no
dedicated transitional states.
4) It's allowed to make any changes with the cgroup hierarchy
(create new cgroups, remove old cgroups, move tasks between cgroups)
no matter if some cgroups are frozen.
Signed-off-by: Roman Gushchin <guro@fb.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
No-objection-from-me-by: Oleg Nesterov <oleg@redhat.com>
Cc: kernel-team@fb.com
2019-04-19 17:03:04 +00:00
|
|
|
/* task is frozen/stopped (used by the cgroup freezer) */
|
|
|
|
unsigned frozen:1;
|
cgroup, kthread: close race window where new kthreads can be migrated to non-root cgroups
Creation of a kthread goes through a couple interlocked stages between
the kthread itself and its creator. Once the new kthread starts
running, it initializes itself and wakes up the creator. The creator
then can further configure the kthread and then let it start doing its
job by waking it up.
In this configuration-by-creator stage, the creator is the only one
that can wake it up but the kthread is visible to userland. When
altering the kthread's attributes from userland is allowed, this is
fine; however, for cases where CPU affinity is critical,
kthread_bind() is used to first disable affinity changes from userland
and then set the affinity. This also prevents the kthread from being
migrated into non-root cgroups as that can affect the CPU affinity and
many other things.
Unfortunately, the cgroup side of protection is racy. While the
PF_NO_SETAFFINITY flag prevents further migrations, userland can win
the race before the creator sets the flag with kthread_bind() and put
the kthread in a non-root cgroup, which can lead to all sorts of
problems including incorrect CPU affinity and starvation.
This bug got triggered by userland which periodically tries to migrate
all processes in the root cpuset cgroup to a non-root one. Per-cpu
workqueue workers got caught while being created and ended up with
incorrected CPU affinity breaking concurrency management and sometimes
stalling workqueue execution.
This patch adds task->no_cgroup_migration which disallows the task to
be migrated by userland. kthreadd starts with the flag set making
every child kthread start in the root cgroup with migration
disallowed. The flag is cleared after the kthread finishes
initialization by which time PF_NO_SETAFFINITY is set if the kthread
should stay in the root cgroup.
It'd be better to wait for the initialization instead of failing but I
couldn't think of a way of implementing that without adding either a
new PF flag, or sleeping and retrying from waiting side. Even if
userland depends on changing cgroup membership of a kthread, it either
has to be synchronized with kthread_create() or periodically repeat,
so it's unlikely that this would break anything.
v2: Switch to a simpler implementation using a new task_struct bit
field suggested by Oleg.
Signed-off-by: Tejun Heo <tj@kernel.org>
Suggested-by: Oleg Nesterov <oleg@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Reported-and-debugged-by: Chris Mason <clm@fb.com>
Cc: stable@vger.kernel.org # v4.3+ (we can't close the race on < v4.3)
Signed-off-by: Tejun Heo <tj@kernel.org>
2017-03-16 20:54:24 +00:00
|
|
|
#endif
|
2018-07-03 15:14:55 +00:00
|
|
|
#ifdef CONFIG_BLK_CGROUP
|
|
|
|
/* to be used once the psi infrastructure lands upstream. */
|
|
|
|
unsigned use_memdelay:1;
|
|
|
|
#endif
|
2014-12-13 00:55:15 +00:00
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
unsigned long atomic_flags; /* Flags requiring atomic access. */
|
2014-05-21 22:23:46 +00:00
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
struct restart_block restart_block;
|
2015-02-12 23:01:14 +00:00
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
pid_t pid;
|
|
|
|
pid_t tgid;
|
2006-09-26 08:52:38 +00:00
|
|
|
|
Kbuild: rename CC_STACKPROTECTOR[_STRONG] config variables
The changes to automatically test for working stack protector compiler
support in the Kconfig files removed the special STACKPROTECTOR_AUTO
option that picked the strongest stack protector that the compiler
supported.
That was all a nice cleanup - it makes no sense to have the AUTO case
now that the Kconfig phase can just determine the compiler support
directly.
HOWEVER.
It also meant that doing "make oldconfig" would now _disable_ the strong
stackprotector if you had AUTO enabled, because in a legacy config file,
the sane stack protector configuration would look like
CONFIG_HAVE_CC_STACKPROTECTOR=y
# CONFIG_CC_STACKPROTECTOR_NONE is not set
# CONFIG_CC_STACKPROTECTOR_REGULAR is not set
# CONFIG_CC_STACKPROTECTOR_STRONG is not set
CONFIG_CC_STACKPROTECTOR_AUTO=y
and when you ran this through "make oldconfig" with the Kbuild changes,
it would ask you about the regular CONFIG_CC_STACKPROTECTOR (that had
been renamed from CONFIG_CC_STACKPROTECTOR_REGULAR to just
CONFIG_CC_STACKPROTECTOR), but it would think that the STRONG version
used to be disabled (because it was really enabled by AUTO), and would
disable it in the new config, resulting in:
CONFIG_HAVE_CC_STACKPROTECTOR=y
CONFIG_CC_HAS_STACKPROTECTOR_NONE=y
CONFIG_CC_STACKPROTECTOR=y
# CONFIG_CC_STACKPROTECTOR_STRONG is not set
CONFIG_CC_HAS_SANE_STACKPROTECTOR=y
That's dangerously subtle - people could suddenly find themselves with
the weaker stack protector setup without even realizing.
The solution here is to just rename not just the old RECULAR stack
protector option, but also the strong one. This does that by just
removing the CC_ prefix entirely for the user choices, because it really
is not about the compiler support (the compiler support now instead
automatially impacts _visibility_ of the options to users).
This results in "make oldconfig" actually asking the user for their
choice, so that we don't have any silent subtle security model changes.
The end result would generally look like this:
CONFIG_HAVE_CC_STACKPROTECTOR=y
CONFIG_CC_HAS_STACKPROTECTOR_NONE=y
CONFIG_STACKPROTECTOR=y
CONFIG_STACKPROTECTOR_STRONG=y
CONFIG_CC_HAS_SANE_STACKPROTECTOR=y
where the "CC_" versions really are about internal compiler
infrastructure, not the user selections.
Acked-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2018-06-14 03:21:18 +00:00
|
|
|
#ifdef CONFIG_STACKPROTECTOR
|
2017-02-06 21:06:35 +00:00
|
|
|
/* Canary value for the -fstack-protector GCC feature: */
|
|
|
|
unsigned long stack_canary;
|
2009-08-18 06:06:02 +00:00
|
|
|
#endif
|
2012-05-11 00:59:08 +00:00
|
|
|
/*
|
2017-02-06 21:06:35 +00:00
|
|
|
* Pointers to the (original) parent process, youngest child, younger sibling,
|
2012-05-11 00:59:08 +00:00
|
|
|
* older sibling, respectively. (p->father can be replaced with
|
2008-03-25 01:36:23 +00:00
|
|
|
* p->real_parent->pid)
|
2005-04-16 22:20:36 +00:00
|
|
|
*/
|
2017-02-06 21:06:35 +00:00
|
|
|
|
|
|
|
/* Real parent process: */
|
|
|
|
struct task_struct __rcu *real_parent;
|
|
|
|
|
|
|
|
/* Recipient of SIGCHLD, wait4() reports: */
|
|
|
|
struct task_struct __rcu *parent;
|
|
|
|
|
2005-04-16 22:20:36 +00:00
|
|
|
/*
|
2017-02-06 21:06:35 +00:00
|
|
|
* Children/sibling form the list of natural children:
|
2005-04-16 22:20:36 +00:00
|
|
|
*/
|
2017-02-06 21:06:35 +00:00
|
|
|
struct list_head children;
|
|
|
|
struct list_head sibling;
|
|
|
|
struct task_struct *group_leader;
|
2005-04-16 22:20:36 +00:00
|
|
|
|
2008-03-25 01:36:23 +00:00
|
|
|
/*
|
2017-02-06 21:06:35 +00:00
|
|
|
* 'ptraced' is the list of tasks this task is using ptrace() on.
|
|
|
|
*
|
2008-03-25 01:36:23 +00:00
|
|
|
* This includes both natural children and PTRACE_ATTACH targets.
|
2017-02-06 21:06:35 +00:00
|
|
|
* 'ptrace_entry' is this task's link on the p->parent->ptraced list.
|
2008-03-25 01:36:23 +00:00
|
|
|
*/
|
2017-02-06 21:06:35 +00:00
|
|
|
struct list_head ptraced;
|
|
|
|
struct list_head ptrace_entry;
|
2008-03-25 01:36:23 +00:00
|
|
|
|
2005-04-16 22:20:36 +00:00
|
|
|
/* PID/PID hash table linkage. */
|
2017-09-26 18:06:43 +00:00
|
|
|
struct pid *thread_pid;
|
|
|
|
struct hlist_node pid_links[PIDTYPE_MAX];
|
2017-02-06 21:06:35 +00:00
|
|
|
struct list_head thread_group;
|
|
|
|
struct list_head thread_node;
|
|
|
|
|
|
|
|
struct completion *vfork_done;
|
2005-04-16 22:20:36 +00:00
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
/* CLONE_CHILD_SETTID: */
|
|
|
|
int __user *set_child_tid;
|
2005-04-16 22:20:36 +00:00
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
/* CLONE_CHILD_CLEARTID: */
|
|
|
|
int __user *clear_child_tid;
|
|
|
|
|
|
|
|
u64 utime;
|
|
|
|
u64 stime;
|
2016-11-15 02:06:51 +00:00
|
|
|
#ifdef CONFIG_ARCH_HAS_SCALED_CPUTIME
|
2017-02-06 21:06:35 +00:00
|
|
|
u64 utimescaled;
|
|
|
|
u64 stimescaled;
|
2016-11-15 02:06:51 +00:00
|
|
|
#endif
|
2017-02-06 21:06:35 +00:00
|
|
|
u64 gtime;
|
|
|
|
struct prev_cputime prev_cputime;
|
2012-12-16 19:00:34 +00:00
|
|
|
#ifdef CONFIG_VIRT_CPU_ACCOUNTING_GEN
|
2017-06-29 17:15:10 +00:00
|
|
|
struct vtime vtime;
|
2009-12-02 08:26:47 +00:00
|
|
|
#endif
|
2015-06-07 13:54:30 +00:00
|
|
|
|
|
|
|
#ifdef CONFIG_NO_HZ_FULL
|
2017-02-06 21:06:35 +00:00
|
|
|
atomic_t tick_dep_mask;
|
2015-06-07 13:54:30 +00:00
|
|
|
#endif
|
2017-02-06 21:06:35 +00:00
|
|
|
/* Context switch counts: */
|
|
|
|
unsigned long nvcsw;
|
|
|
|
unsigned long nivcsw;
|
|
|
|
|
|
|
|
/* Monotonic time in nsecs: */
|
|
|
|
u64 start_time;
|
|
|
|
|
|
|
|
/* Boot based time in nsecs: */
|
|
|
|
u64 real_start_time;
|
|
|
|
|
|
|
|
/* MM fault and swap info: this can arguably be seen as either mm-specific or thread-specific: */
|
|
|
|
unsigned long min_flt;
|
|
|
|
unsigned long maj_flt;
|
2005-04-16 22:20:36 +00:00
|
|
|
|
2017-01-21 05:09:08 +00:00
|
|
|
#ifdef CONFIG_POSIX_TIMERS
|
2017-02-06 21:06:35 +00:00
|
|
|
struct task_cputime cputime_expires;
|
|
|
|
struct list_head cpu_timers[3];
|
2017-01-21 05:09:08 +00:00
|
|
|
#endif
|
2005-04-16 22:20:36 +00:00
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
/* Process credentials: */
|
|
|
|
|
|
|
|
/* Tracer's credentials at attach: */
|
|
|
|
const struct cred __rcu *ptracer_cred;
|
|
|
|
|
|
|
|
/* Objective and real subjective task credentials (COW): */
|
|
|
|
const struct cred __rcu *real_cred;
|
|
|
|
|
|
|
|
/* Effective (overridable) subjective task credentials (COW): */
|
|
|
|
const struct cred __rcu *cred;
|
|
|
|
|
keys: Cache result of request_key*() temporarily in task_struct
If a filesystem uses keys to hold authentication tokens, then it needs a
token for each VFS operation that might perform an authentication check -
either by passing it to the server, or using to perform a check based on
authentication data cached locally.
For open files this isn't a problem, since the key should be cached in the
file struct since it represents the subject performing operations on that
file descriptor.
During pathwalk, however, there isn't anywhere to cache the key, except
perhaps in the nameidata struct - but that isn't exposed to the
filesystems. Further, a pathwalk can incur a lot of operations, calling
one or more of the following, for instance:
->lookup()
->permission()
->d_revalidate()
->d_automount()
->get_acl()
->getxattr()
on each dentry/inode it encounters - and each one may need to call
request_key(). And then, at the end of pathwalk, it will call the actual
operation:
->mkdir()
->mknod()
->getattr()
->open()
...
which may need to go and get the token again.
However, it is very likely that all of the operations on a single
dentry/inode - and quite possibly a sequence of them - will all want to use
the same authentication token, which suggests that caching it would be a
good idea.
To this end:
(1) Make it so that a positive result of request_key() and co. that didn't
require upcalling to userspace is cached temporarily in task_struct.
(2) The cache is 1 deep, so a new result displaces the old one.
(3) The key is released by exit and by notify-resume.
(4) The cache is cleared in a newly forked process.
Signed-off-by: David Howells <dhowells@redhat.com>
2019-06-19 15:10:15 +00:00
|
|
|
#ifdef CONFIG_KEYS
|
|
|
|
/* Cached requested key. */
|
|
|
|
struct key *cached_requested_key;
|
|
|
|
#endif
|
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
/*
|
|
|
|
* executable name, excluding path.
|
|
|
|
*
|
|
|
|
* - normally initialized setup_new_exec()
|
|
|
|
* - access it with [gs]et_task_comm()
|
|
|
|
* - lock it with task_lock()
|
|
|
|
*/
|
|
|
|
char comm[TASK_COMM_LEN];
|
|
|
|
|
|
|
|
struct nameidata *nameidata;
|
|
|
|
|
2006-09-29 08:59:40 +00:00
|
|
|
#ifdef CONFIG_SYSVIPC
|
2017-02-06 21:06:35 +00:00
|
|
|
struct sysv_sem sysvsem;
|
|
|
|
struct sysv_shm sysvshm;
|
2006-09-29 08:59:40 +00:00
|
|
|
#endif
|
2009-01-15 19:08:40 +00:00
|
|
|
#ifdef CONFIG_DETECT_HUNG_TASK
|
2017-02-06 21:06:35 +00:00
|
|
|
unsigned long last_switch_count;
|
2018-08-22 04:55:52 +00:00
|
|
|
unsigned long last_switch_time;
|
2008-01-25 20:08:02 +00:00
|
|
|
#endif
|
2017-02-06 21:06:35 +00:00
|
|
|
/* Filesystem information: */
|
|
|
|
struct fs_struct *fs;
|
|
|
|
|
|
|
|
/* Open file information: */
|
|
|
|
struct files_struct *files;
|
|
|
|
|
|
|
|
/* Namespaces: */
|
|
|
|
struct nsproxy *nsproxy;
|
|
|
|
|
|
|
|
/* Signal handlers: */
|
|
|
|
struct signal_struct *signal;
|
|
|
|
struct sighand_struct *sighand;
|
|
|
|
sigset_t blocked;
|
|
|
|
sigset_t real_blocked;
|
|
|
|
/* Restored if set_restore_sigmask() was used: */
|
|
|
|
sigset_t saved_sigmask;
|
|
|
|
struct sigpending pending;
|
|
|
|
unsigned long sas_ss_sp;
|
|
|
|
size_t sas_ss_size;
|
|
|
|
unsigned int sas_ss_flags;
|
|
|
|
|
|
|
|
struct callback_head *task_works;
|
|
|
|
|
2019-01-22 22:06:39 +00:00
|
|
|
#ifdef CONFIG_AUDIT
|
2008-01-10 09:53:18 +00:00
|
|
|
#ifdef CONFIG_AUDITSYSCALL
|
2019-02-02 03:45:17 +00:00
|
|
|
struct audit_context *audit_context;
|
|
|
|
#endif
|
2017-02-06 21:06:35 +00:00
|
|
|
kuid_t loginuid;
|
|
|
|
unsigned int sessionid;
|
2008-01-10 09:53:18 +00:00
|
|
|
#endif
|
2017-02-06 21:06:35 +00:00
|
|
|
struct seccomp seccomp;
|
|
|
|
|
|
|
|
/* Thread group tracking: */
|
|
|
|
u32 parent_exec_id;
|
|
|
|
u32 self_exec_id;
|
2005-04-16 22:20:36 +00:00
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
/* Protection against (de-)allocation: mm, files, fs, tty, keyrings, mems_allowed, mempolicy: */
|
|
|
|
spinlock_t alloc_lock;
|
2005-04-16 22:20:36 +00:00
|
|
|
|
2006-06-27 09:54:51 +00:00
|
|
|
/* Protection of the PI data structures: */
|
2017-02-06 21:06:35 +00:00
|
|
|
raw_spinlock_t pi_lock;
|
2006-06-27 09:54:51 +00:00
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
struct wake_q_node wake_q;
|
2015-05-01 15:27:50 +00:00
|
|
|
|
2006-06-27 09:54:53 +00:00
|
|
|
#ifdef CONFIG_RT_MUTEXES
|
2017-02-06 21:06:35 +00:00
|
|
|
/* PI waiters blocked on a rt_mutex held by this task: */
|
2017-09-08 23:15:01 +00:00
|
|
|
struct rb_root_cached pi_waiters;
|
2017-03-23 14:56:08 +00:00
|
|
|
/* Updated under owner's pi_lock and rq lock */
|
|
|
|
struct task_struct *pi_top_task;
|
2017-02-06 21:06:35 +00:00
|
|
|
/* Deadlock detection and priority inheritance handling: */
|
|
|
|
struct rt_mutex_waiter *pi_blocked_on;
|
2006-06-27 09:54:53 +00:00
|
|
|
#endif
|
|
|
|
|
2006-01-09 23:59:20 +00:00
|
|
|
#ifdef CONFIG_DEBUG_MUTEXES
|
2017-02-06 21:06:35 +00:00
|
|
|
/* Mutex deadlock detection: */
|
|
|
|
struct mutex_waiter *blocked_on;
|
2006-01-09 23:59:20 +00:00
|
|
|
#endif
|
2017-02-06 21:06:35 +00:00
|
|
|
|
2006-07-03 07:24:42 +00:00
|
|
|
#ifdef CONFIG_TRACE_IRQFLAGS
|
2017-02-06 21:06:35 +00:00
|
|
|
unsigned int irq_events;
|
|
|
|
unsigned long hardirq_enable_ip;
|
|
|
|
unsigned long hardirq_disable_ip;
|
|
|
|
unsigned int hardirq_enable_event;
|
|
|
|
unsigned int hardirq_disable_event;
|
|
|
|
int hardirqs_enabled;
|
|
|
|
int hardirq_context;
|
|
|
|
unsigned long softirq_disable_ip;
|
|
|
|
unsigned long softirq_enable_ip;
|
|
|
|
unsigned int softirq_disable_event;
|
|
|
|
unsigned int softirq_enable_event;
|
|
|
|
int softirqs_enabled;
|
|
|
|
int softirq_context;
|
2006-07-03 07:24:42 +00:00
|
|
|
#endif
|
2017-02-06 21:06:35 +00:00
|
|
|
|
[PATCH] lockdep: core
Do 'make oldconfig' and accept all the defaults for new config options -
reboot into the kernel and if everything goes well it should boot up fine and
you should have /proc/lockdep and /proc/lockdep_stats files.
Typically if the lock validator finds some problem it will print out
voluminous debug output that begins with "BUG: ..." and which syslog output
can be used by kernel developers to figure out the precise locking scenario.
What does the lock validator do? It "observes" and maps all locking rules as
they occur dynamically (as triggered by the kernel's natural use of spinlocks,
rwlocks, mutexes and rwsems). Whenever the lock validator subsystem detects a
new locking scenario, it validates this new rule against the existing set of
rules. If this new rule is consistent with the existing set of rules then the
new rule is added transparently and the kernel continues as normal. If the
new rule could create a deadlock scenario then this condition is printed out.
When determining validity of locking, all possible "deadlock scenarios" are
considered: assuming arbitrary number of CPUs, arbitrary irq context and task
context constellations, running arbitrary combinations of all the existing
locking scenarios. In a typical system this means millions of separate
scenarios. This is why we call it a "locking correctness" validator - for all
rules that are observed the lock validator proves it with mathematical
certainty that a deadlock could not occur (assuming that the lock validator
implementation itself is correct and its internal data structures are not
corrupted by some other kernel subsystem). [see more details and conditionals
of this statement in include/linux/lockdep.h and
Documentation/lockdep-design.txt]
Furthermore, this "all possible scenarios" property of the validator also
enables the finding of complex, highly unlikely multi-CPU multi-context races
via single single-context rules, increasing the likelyhood of finding bugs
drastically. In practical terms: the lock validator already found a bug in
the upstream kernel that could only occur on systems with 3 or more CPUs, and
which needed 3 very unlikely code sequences to occur at once on the 3 CPUs.
That bug was found and reported on a single-CPU system (!). So in essence a
race will be found "piecemail-wise", triggering all the necessary components
for the race, without having to reproduce the race scenario itself! In its
short existence the lock validator found and reported many bugs before they
actually caused a real deadlock.
To further increase the efficiency of the validator, the mapping is not per
"lock instance", but per "lock-class". For example, all struct inode objects
in the kernel have inode->inotify_mutex. If there are 10,000 inodes cached,
then there are 10,000 lock objects. But ->inotify_mutex is a single "lock
type", and all locking activities that occur against ->inotify_mutex are
"unified" into this single lock-class. The advantage of the lock-class
approach is that all historical ->inotify_mutex uses are mapped into a single
(and as narrow as possible) set of locking rules - regardless of how many
different tasks or inode structures it took to build this set of rules. The
set of rules persist during the lifetime of the kernel.
To see the rough magnitude of checking that the lock validator does, here's a
portion of /proc/lockdep_stats, fresh after bootup:
lock-classes: 694 [max: 2048]
direct dependencies: 1598 [max: 8192]
indirect dependencies: 17896
all direct dependencies: 16206
dependency chains: 1910 [max: 8192]
in-hardirq chains: 17
in-softirq chains: 105
in-process chains: 1065
stack-trace entries: 38761 [max: 131072]
combined max dependencies: 2033928
hardirq-safe locks: 24
hardirq-unsafe locks: 176
softirq-safe locks: 53
softirq-unsafe locks: 137
irq-safe locks: 59
irq-unsafe locks: 176
The lock validator has observed 1598 actual single-thread locking patterns,
and has validated all possible 2033928 distinct locking scenarios.
More details about the design of the lock validator can be found in
Documentation/lockdep-design.txt, which can also found at:
http://redhat.com/~mingo/lockdep-patches/lockdep-design.txt
[bunk@stusta.de: cleanups]
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-07-03 07:24:50 +00:00
|
|
|
#ifdef CONFIG_LOCKDEP
|
2017-02-06 21:06:35 +00:00
|
|
|
# define MAX_LOCK_DEPTH 48UL
|
|
|
|
u64 curr_chain_key;
|
|
|
|
int lockdep_depth;
|
|
|
|
unsigned int lockdep_recursion;
|
|
|
|
struct held_lock held_locks[MAX_LOCK_DEPTH];
|
[PATCH] lockdep: core
Do 'make oldconfig' and accept all the defaults for new config options -
reboot into the kernel and if everything goes well it should boot up fine and
you should have /proc/lockdep and /proc/lockdep_stats files.
Typically if the lock validator finds some problem it will print out
voluminous debug output that begins with "BUG: ..." and which syslog output
can be used by kernel developers to figure out the precise locking scenario.
What does the lock validator do? It "observes" and maps all locking rules as
they occur dynamically (as triggered by the kernel's natural use of spinlocks,
rwlocks, mutexes and rwsems). Whenever the lock validator subsystem detects a
new locking scenario, it validates this new rule against the existing set of
rules. If this new rule is consistent with the existing set of rules then the
new rule is added transparently and the kernel continues as normal. If the
new rule could create a deadlock scenario then this condition is printed out.
When determining validity of locking, all possible "deadlock scenarios" are
considered: assuming arbitrary number of CPUs, arbitrary irq context and task
context constellations, running arbitrary combinations of all the existing
locking scenarios. In a typical system this means millions of separate
scenarios. This is why we call it a "locking correctness" validator - for all
rules that are observed the lock validator proves it with mathematical
certainty that a deadlock could not occur (assuming that the lock validator
implementation itself is correct and its internal data structures are not
corrupted by some other kernel subsystem). [see more details and conditionals
of this statement in include/linux/lockdep.h and
Documentation/lockdep-design.txt]
Furthermore, this "all possible scenarios" property of the validator also
enables the finding of complex, highly unlikely multi-CPU multi-context races
via single single-context rules, increasing the likelyhood of finding bugs
drastically. In practical terms: the lock validator already found a bug in
the upstream kernel that could only occur on systems with 3 or more CPUs, and
which needed 3 very unlikely code sequences to occur at once on the 3 CPUs.
That bug was found and reported on a single-CPU system (!). So in essence a
race will be found "piecemail-wise", triggering all the necessary components
for the race, without having to reproduce the race scenario itself! In its
short existence the lock validator found and reported many bugs before they
actually caused a real deadlock.
To further increase the efficiency of the validator, the mapping is not per
"lock instance", but per "lock-class". For example, all struct inode objects
in the kernel have inode->inotify_mutex. If there are 10,000 inodes cached,
then there are 10,000 lock objects. But ->inotify_mutex is a single "lock
type", and all locking activities that occur against ->inotify_mutex are
"unified" into this single lock-class. The advantage of the lock-class
approach is that all historical ->inotify_mutex uses are mapped into a single
(and as narrow as possible) set of locking rules - regardless of how many
different tasks or inode structures it took to build this set of rules. The
set of rules persist during the lifetime of the kernel.
To see the rough magnitude of checking that the lock validator does, here's a
portion of /proc/lockdep_stats, fresh after bootup:
lock-classes: 694 [max: 2048]
direct dependencies: 1598 [max: 8192]
indirect dependencies: 17896
all direct dependencies: 16206
dependency chains: 1910 [max: 8192]
in-hardirq chains: 17
in-softirq chains: 105
in-process chains: 1065
stack-trace entries: 38761 [max: 131072]
combined max dependencies: 2033928
hardirq-safe locks: 24
hardirq-unsafe locks: 176
softirq-safe locks: 53
softirq-unsafe locks: 137
irq-safe locks: 59
irq-unsafe locks: 176
The lock validator has observed 1598 actual single-thread locking patterns,
and has validated all possible 2033928 distinct locking scenarios.
More details about the design of the lock validator can be found in
Documentation/lockdep-design.txt, which can also found at:
http://redhat.com/~mingo/lockdep-patches/lockdep-design.txt
[bunk@stusta.de: cleanups]
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Arjan van de Ven <arjan@linux.intel.com>
Signed-off-by: Adrian Bunk <bunk@stusta.de>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-07-03 07:24:50 +00:00
|
|
|
#endif
|
2017-02-06 21:06:35 +00:00
|
|
|
|
2016-01-20 23:00:55 +00:00
|
|
|
#ifdef CONFIG_UBSAN
|
2017-02-06 21:06:35 +00:00
|
|
|
unsigned int in_ubsan;
|
2016-01-20 23:00:55 +00:00
|
|
|
#endif
|
2006-01-09 23:59:20 +00:00
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
/* Journalling filesystem info: */
|
|
|
|
void *journal_info;
|
2005-04-16 22:20:36 +00:00
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
/* Stacked block device info: */
|
|
|
|
struct bio_list *bio_list;
|
When stacked block devices are in-use (e.g. md or dm), the recursive calls
to generic_make_request can use up a lot of space, and we would rather they
didn't.
As generic_make_request is a void function, and as it is generally not
expected that it will have any effect immediately, it is safe to delay any
call to generic_make_request until there is sufficient stack space
available.
As ->bi_next is reserved for the driver to use, it can have no valid value
when generic_make_request is called, and as __make_request implicitly
assumes it will be NULL (ELEVATOR_BACK_MERGE fork of switch) we can be
certain that all callers set it to NULL. We can therefore safely use
bi_next to link pending requests together, providing we clear it before
making the real call.
So, we choose to allow each thread to only be active in one
generic_make_request at a time. If a subsequent (recursive) call is made,
the bio is linked into a per-thread list, and is handled when the active
call completes.
As the list of pending bios is per-thread, there are no locking issues to
worry about.
I say above that it is "safe to delay any call...". There are, however,
some behaviours of a make_request_fn which would make it unsafe. These
include any behaviour that assumes anything will have changed after a
recursive call to generic_make_request.
These could include:
- waiting for that call to finish and call it's bi_end_io function.
md use to sometimes do this (marking the superblock dirty before
completing a write) but doesn't any more
- inspecting the bio for fields that generic_make_request might
change, such as bi_sector or bi_bdev. It is hard to see a good
reason for this, and I don't think anyone actually does it.
- inspecing the queue to see if, e.g. it is 'full' yet. Again, I
think this is very unlikely to be useful, or to be done.
Signed-off-by: Neil Brown <neilb@suse.de>
Cc: Jens Axboe <axboe@kernel.dk>
Cc: <dm-devel@redhat.com>
Alasdair G Kergon <agk@redhat.com> said:
I can see nothing wrong with this in principle.
For device-mapper at the moment though it's essential that, while the bio
mappings may now get delayed, they still get processed in exactly
the same order as they were passed to generic_make_request().
My main concern is whether the timing changes implicit in this patch
will make the rare data-corrupting races in the existing snapshot code
more likely. (I'm working on a fix for these races, but the unfinished
patch is already several hundred lines long.)
It would be helpful if some people on this mailing list would test
this patch in various scenarios and report back.
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
2007-05-01 07:53:42 +00:00
|
|
|
|
2011-03-08 12:19:51 +00:00
|
|
|
#ifdef CONFIG_BLOCK
|
2017-02-06 21:06:35 +00:00
|
|
|
/* Stack plugging: */
|
|
|
|
struct blk_plug *plug;
|
2011-03-08 12:19:51 +00:00
|
|
|
#endif
|
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
/* VM state: */
|
|
|
|
struct reclaim_state *reclaim_state;
|
|
|
|
|
|
|
|
struct backing_dev_info *backing_dev_info;
|
2005-04-16 22:20:36 +00:00
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
struct io_context *io_context;
|
2005-04-16 22:20:36 +00:00
|
|
|
|
2019-03-05 23:45:41 +00:00
|
|
|
#ifdef CONFIG_COMPACTION
|
|
|
|
struct capture_control *capture_control;
|
|
|
|
#endif
|
2017-02-06 21:06:35 +00:00
|
|
|
/* Ptrace state: */
|
|
|
|
unsigned long ptrace_message;
|
2018-09-25 09:27:20 +00:00
|
|
|
kernel_siginfo_t *last_siginfo;
|
2005-04-16 22:20:36 +00:00
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
struct task_io_accounting ioac;
|
psi: pressure stall information for CPU, memory, and IO
When systems are overcommitted and resources become contended, it's hard
to tell exactly the impact this has on workload productivity, or how close
the system is to lockups and OOM kills. In particular, when machines work
multiple jobs concurrently, the impact of overcommit in terms of latency
and throughput on the individual job can be enormous.
In order to maximize hardware utilization without sacrificing individual
job health or risk complete machine lockups, this patch implements a way
to quantify resource pressure in the system.
A kernel built with CONFIG_PSI=y creates files in /proc/pressure/ that
expose the percentage of time the system is stalled on CPU, memory, or IO,
respectively. Stall states are aggregate versions of the per-task delay
accounting delays:
cpu: some tasks are runnable but not executing on a CPU
memory: tasks are reclaiming, or waiting for swapin or thrashing cache
io: tasks are waiting for io completions
These percentages of walltime can be thought of as pressure percentages,
and they give a general sense of system health and productivity loss
incurred by resource overcommit. They can also indicate when the system
is approaching lockup scenarios and OOMs.
To do this, psi keeps track of the task states associated with each CPU
and samples the time they spend in stall states. Every 2 seconds, the
samples are averaged across CPUs - weighted by the CPUs' non-idle time to
eliminate artifacts from unused CPUs - and translated into percentages of
walltime. A running average of those percentages is maintained over 10s,
1m, and 5m periods (similar to the loadaverage).
[hannes@cmpxchg.org: doc fixlet, per Randy]
Link: http://lkml.kernel.org/r/20180828205625.GA14030@cmpxchg.org
[hannes@cmpxchg.org: code optimization]
Link: http://lkml.kernel.org/r/20180907175015.GA8479@cmpxchg.org
[hannes@cmpxchg.org: rename psi_clock() to psi_update_work(), per Peter]
Link: http://lkml.kernel.org/r/20180907145404.GB11088@cmpxchg.org
[hannes@cmpxchg.org: fix build]
Link: http://lkml.kernel.org/r/20180913014222.GA2370@cmpxchg.org
Link: http://lkml.kernel.org/r/20180828172258.3185-9-hannes@cmpxchg.org
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Daniel Drake <drake@endlessm.com>
Tested-by: Suren Baghdasaryan <surenb@google.com>
Cc: Christopher Lameter <cl@linux.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Johannes Weiner <jweiner@fb.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Enderborg <peter.enderborg@sony.com>
Cc: Randy Dunlap <rdunlap@infradead.org>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Vinayak Menon <vinmenon@codeaurora.org>
Cc: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2018-10-26 22:06:27 +00:00
|
|
|
#ifdef CONFIG_PSI
|
|
|
|
/* Pressure stall state */
|
|
|
|
unsigned int psi_flags;
|
|
|
|
#endif
|
2017-02-06 21:06:35 +00:00
|
|
|
#ifdef CONFIG_TASK_XACCT
|
|
|
|
/* Accumulated RSS usage: */
|
|
|
|
u64 acct_rss_mem1;
|
|
|
|
/* Accumulated virtual memory usage: */
|
|
|
|
u64 acct_vm_mem1;
|
|
|
|
/* stime + utime since last update: */
|
|
|
|
u64 acct_timexpd;
|
2005-04-16 22:20:36 +00:00
|
|
|
#endif
|
|
|
|
#ifdef CONFIG_CPUSETS
|
2017-02-06 21:06:35 +00:00
|
|
|
/* Protected by ->alloc_lock: */
|
|
|
|
nodemask_t mems_allowed;
|
|
|
|
/* Seqence number to catch updates: */
|
|
|
|
seqcount_t mems_allowed_seq;
|
|
|
|
int cpuset_mem_spread_rotor;
|
|
|
|
int cpuset_slab_spread_rotor;
|
2005-04-16 22:20:36 +00:00
|
|
|
#endif
|
Task Control Groups: basic task cgroup framework
Generic Process Control Groups
--------------------------
There have recently been various proposals floating around for
resource management/accounting and other task grouping subsystems in
the kernel, including ResGroups, User BeanCounters, NSProxy
cgroups, and others. These all need the basic abstraction of being
able to group together multiple processes in an aggregate, in order to
track/limit the resources permitted to those processes, or control
other behaviour of the processes, and all implement this grouping in
different ways.
This patchset provides a framework for tracking and grouping processes
into arbitrary "cgroups" and assigning arbitrary state to those
groupings, in order to control the behaviour of the cgroup as an
aggregate.
The intention is that the various resource management and
virtualization/cgroup efforts can also become task cgroup
clients, with the result that:
- the userspace APIs are (somewhat) normalised
- it's easier to test e.g. the ResGroups CPU controller in
conjunction with the BeanCounters memory controller, or use either of
them as the resource-control portion of a virtual server system.
- the additional kernel footprint of any of the competing resource
management systems is substantially reduced, since it doesn't need
to provide process grouping/containment, hence improving their
chances of getting into the kernel
This patch:
Add the main task cgroups framework - the cgroup filesystem, and the
basic structures for tracking membership and associating subsystem state
objects to tasks.
Signed-off-by: Paul Menage <menage@google.com>
Cc: Serge E. Hallyn <serue@us.ibm.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Dave Hansen <haveblue@us.ibm.com>
Cc: Balbir Singh <balbir@in.ibm.com>
Cc: Paul Jackson <pj@sgi.com>
Cc: Kirill Korotaev <dev@openvz.org>
Cc: Herbert Poetzl <herbert@13thfloor.at>
Cc: Srivatsa Vaddagiri <vatsa@in.ibm.com>
Cc: Cedric Le Goater <clg@fr.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-19 06:39:30 +00:00
|
|
|
#ifdef CONFIG_CGROUPS
|
2017-02-06 21:06:35 +00:00
|
|
|
/* Control Group info protected by css_set_lock: */
|
|
|
|
struct css_set __rcu *cgroups;
|
|
|
|
/* cg_list protected by css_set_lock and tsk->alloc_lock: */
|
|
|
|
struct list_head cg_list;
|
Task Control Groups: basic task cgroup framework
Generic Process Control Groups
--------------------------
There have recently been various proposals floating around for
resource management/accounting and other task grouping subsystems in
the kernel, including ResGroups, User BeanCounters, NSProxy
cgroups, and others. These all need the basic abstraction of being
able to group together multiple processes in an aggregate, in order to
track/limit the resources permitted to those processes, or control
other behaviour of the processes, and all implement this grouping in
different ways.
This patchset provides a framework for tracking and grouping processes
into arbitrary "cgroups" and assigning arbitrary state to those
groupings, in order to control the behaviour of the cgroup as an
aggregate.
The intention is that the various resource management and
virtualization/cgroup efforts can also become task cgroup
clients, with the result that:
- the userspace APIs are (somewhat) normalised
- it's easier to test e.g. the ResGroups CPU controller in
conjunction with the BeanCounters memory controller, or use either of
them as the resource-control portion of a virtual server system.
- the additional kernel footprint of any of the competing resource
management systems is substantially reduced, since it doesn't need
to provide process grouping/containment, hence improving their
chances of getting into the kernel
This patch:
Add the main task cgroups framework - the cgroup filesystem, and the
basic structures for tracking membership and associating subsystem state
objects to tasks.
Signed-off-by: Paul Menage <menage@google.com>
Cc: Serge E. Hallyn <serue@us.ibm.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Dave Hansen <haveblue@us.ibm.com>
Cc: Balbir Singh <balbir@in.ibm.com>
Cc: Paul Jackson <pj@sgi.com>
Cc: Kirill Korotaev <dev@openvz.org>
Cc: Herbert Poetzl <herbert@13thfloor.at>
Cc: Srivatsa Vaddagiri <vatsa@in.ibm.com>
Cc: Cedric Le Goater <clg@fr.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-10-19 06:39:30 +00:00
|
|
|
#endif
|
2019-01-29 22:44:36 +00:00
|
|
|
#ifdef CONFIG_X86_CPU_RESCTRL
|
2017-07-25 21:14:33 +00:00
|
|
|
u32 closid;
|
2017-07-25 21:14:34 +00:00
|
|
|
u32 rmid;
|
2016-10-28 22:04:46 +00:00
|
|
|
#endif
|
2007-10-17 06:27:30 +00:00
|
|
|
#ifdef CONFIG_FUTEX
|
2017-02-06 21:06:35 +00:00
|
|
|
struct robust_list_head __user *robust_list;
|
2006-03-27 09:16:24 +00:00
|
|
|
#ifdef CONFIG_COMPAT
|
|
|
|
struct compat_robust_list_head __user *compat_robust_list;
|
|
|
|
#endif
|
2017-02-06 21:06:35 +00:00
|
|
|
struct list_head pi_state_list;
|
|
|
|
struct futex_pi_state *pi_state_cache;
|
2008-05-15 11:09:15 +00:00
|
|
|
#endif
|
perf: Do the big rename: Performance Counters -> Performance Events
Bye-bye Performance Counters, welcome Performance Events!
In the past few months the perfcounters subsystem has grown out its
initial role of counting hardware events, and has become (and is
becoming) a much broader generic event enumeration, reporting, logging,
monitoring, analysis facility.
Naming its core object 'perf_counter' and naming the subsystem
'perfcounters' has become more and more of a misnomer. With pending
code like hw-breakpoints support the 'counter' name is less and
less appropriate.
All in one, we've decided to rename the subsystem to 'performance
events' and to propagate this rename through all fields, variables
and API names. (in an ABI compatible fashion)
The word 'event' is also a bit shorter than 'counter' - which makes
it slightly more convenient to write/handle as well.
Thanks goes to Stephane Eranian who first observed this misnomer and
suggested a rename.
User-space tooling and ABI compatibility is not affected - this patch
should be function-invariant. (Also, defconfigs were not touched to
keep the size down.)
This patch has been generated via the following script:
FILES=$(find * -type f | grep -vE 'oprofile|[^K]config')
sed -i \
-e 's/PERF_EVENT_/PERF_RECORD_/g' \
-e 's/PERF_COUNTER/PERF_EVENT/g' \
-e 's/perf_counter/perf_event/g' \
-e 's/nb_counters/nb_events/g' \
-e 's/swcounter/swevent/g' \
-e 's/tpcounter_event/tp_event/g' \
$FILES
for N in $(find . -name perf_counter.[ch]); do
M=$(echo $N | sed 's/perf_counter/perf_event/g')
mv $N $M
done
FILES=$(find . -name perf_event.*)
sed -i \
-e 's/COUNTER_MASK/REG_MASK/g' \
-e 's/COUNTER/EVENT/g' \
-e 's/\<event\>/event_id/g' \
-e 's/counter/event/g' \
-e 's/Counter/Event/g' \
$FILES
... to keep it as correct as possible. This script can also be
used by anyone who has pending perfcounters patches - it converts
a Linux kernel tree over to the new naming. We tried to time this
change to the point in time where the amount of pending patches
is the smallest: the end of the merge window.
Namespace clashes were fixed up in a preparatory patch - and some
stylistic fallout will be fixed up in a subsequent patch.
( NOTE: 'counters' are still the proper terminology when we deal
with hardware registers - and these sed scripts are a bit
over-eager in renaming them. I've undone some of that, but
in case there's something left where 'counter' would be
better than 'event' we can undo that on an individual basis
instead of touching an otherwise nicely automated patch. )
Suggested-by: Stephane Eranian <eranian@google.com>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Acked-by: Paul Mackerras <paulus@samba.org>
Reviewed-by: Arjan van de Ven <arjan@linux.intel.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: David Howells <dhowells@redhat.com>
Cc: Kyle McMartin <kyle@mcmartin.ca>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: <linux-arch@vger.kernel.org>
LKML-Reference: <new-submission>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-09-21 10:02:48 +00:00
|
|
|
#ifdef CONFIG_PERF_EVENTS
|
2017-02-06 21:06:35 +00:00
|
|
|
struct perf_event_context *perf_event_ctxp[perf_nr_task_contexts];
|
|
|
|
struct mutex perf_event_mutex;
|
|
|
|
struct list_head perf_event_list;
|
perf_counter: Dynamically allocate tasks' perf_counter_context struct
This replaces the struct perf_counter_context in the task_struct with
a pointer to a dynamically allocated perf_counter_context struct. The
main reason for doing is this is to allow us to transfer a
perf_counter_context from one task to another when we do lazy PMU
switching in a later patch.
This has a few side-benefits: the task_struct becomes a little smaller,
we save some memory because only tasks that have perf_counters attached
get a perf_counter_context allocated for them, and we can remove the
inclusion of <linux/perf_counter.h> in sched.h, meaning that we don't
end up recompiling nearly everything whenever perf_counter.h changes.
The perf_counter_context structures are reference-counted and freed
when the last reference is dropped. A context can have references
from its task and the counters on its task. Counters can outlive the
task so it is possible that a context will be freed well after its
task has exited.
Contexts are allocated on fork if the parent had a context, or
otherwise the first time that a per-task counter is created on a task.
In the latter case, we set the context pointer in the task struct
locklessly using an atomic compare-and-exchange operation in case we
raced with some other task in creating a context for the subject task.
This also removes the task pointer from the perf_counter struct. The
task pointer was not used anywhere and would make it harder to move a
context from one task to another. Anything that needed to know which
task a counter was attached to was already using counter->ctx->task.
The __perf_counter_init_context function moves up in perf_counter.c
so that it can be called from find_get_context, and now initializes
the refcount, but is otherwise unchanged.
We were potentially calling list_del_counter twice: once from
__perf_counter_exit_task when the task exits and once from
__perf_counter_remove_from_context when the counter's fd gets closed.
This adds a check in list_del_counter so it doesn't do anything if
the counter has already been removed from the lists.
Since perf_counter_task_sched_in doesn't do anything if the task doesn't
have a context, and leaves cpuctx->task_ctx = NULL, this adds code to
__perf_install_in_context to set cpuctx->task_ctx if necessary, i.e. in
the case where the current task adds the first counter to itself and
thus creates a context for itself.
This also adds similar code to __perf_counter_enable to handle a
similar situation which can arise when the counters have been disabled
using prctl; that also leaves cpuctx->task_ctx = NULL.
[ Impact: refactor counter context management to prepare for new feature ]
Signed-off-by: Paul Mackerras <paulus@samba.org>
Acked-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
Cc: Corey Ashford <cjashfor@linux.vnet.ibm.com>
Cc: Marcelo Tosatti <mtosatti@redhat.com>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
LKML-Reference: <18966.10075.781053.231153@cargo.ozlabs.ibm.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2009-05-22 04:17:31 +00:00
|
|
|
#endif
|
2014-02-07 19:58:39 +00:00
|
|
|
#ifdef CONFIG_DEBUG_PREEMPT
|
2017-02-06 21:06:35 +00:00
|
|
|
unsigned long preempt_disable_ip;
|
2014-02-07 19:58:39 +00:00
|
|
|
#endif
|
2008-05-15 11:09:15 +00:00
|
|
|
#ifdef CONFIG_NUMA
|
2017-02-06 21:06:35 +00:00
|
|
|
/* Protected by alloc_lock: */
|
|
|
|
struct mempolicy *mempolicy;
|
2017-07-06 22:39:59 +00:00
|
|
|
short il_prev;
|
2017-02-06 21:06:35 +00:00
|
|
|
short pref_node_fork;
|
2007-10-17 06:27:30 +00:00
|
|
|
#endif
|
2012-10-25 12:16:43 +00:00
|
|
|
#ifdef CONFIG_NUMA_BALANCING
|
2017-02-06 21:06:35 +00:00
|
|
|
int numa_scan_seq;
|
|
|
|
unsigned int numa_scan_period;
|
|
|
|
unsigned int numa_scan_period_max;
|
|
|
|
int numa_preferred_nid;
|
|
|
|
unsigned long numa_migrate_retry;
|
|
|
|
/* Migration stamp: */
|
|
|
|
u64 node_stamp;
|
|
|
|
u64 last_task_numa_placement;
|
|
|
|
u64 last_sum_exec_runtime;
|
|
|
|
struct callback_head numa_work;
|
|
|
|
|
2019-07-16 15:20:47 +00:00
|
|
|
/*
|
|
|
|
* This pointer is only modified for current in syscall and
|
|
|
|
* pagefault context (and for tasks being destroyed), so it can be read
|
|
|
|
* from any of the following contexts:
|
|
|
|
* - RCU read-side critical section
|
|
|
|
* - current->numa_group from everywhere
|
|
|
|
* - task's runqueue locked, task not running
|
|
|
|
*/
|
|
|
|
struct numa_group __rcu *numa_group;
|
2013-10-07 10:29:21 +00:00
|
|
|
|
2013-10-07 10:28:59 +00:00
|
|
|
/*
|
2014-10-31 00:13:31 +00:00
|
|
|
* numa_faults is an array split into four regions:
|
|
|
|
* faults_memory, faults_cpu, faults_memory_buffer, faults_cpu_buffer
|
|
|
|
* in this precise order.
|
|
|
|
*
|
|
|
|
* faults_memory: Exponential decaying average of faults on a per-node
|
|
|
|
* basis. Scheduling placement decisions are made based on these
|
|
|
|
* counts. The values remain static for the duration of a PTE scan.
|
|
|
|
* faults_cpu: Track the nodes the process was running on when a NUMA
|
|
|
|
* hinting fault was incurred.
|
|
|
|
* faults_memory_buffer and faults_cpu_buffer: Record faults per node
|
|
|
|
* during the current scan window. When the scan completes, the counts
|
|
|
|
* in faults_memory and faults_cpu decay and these values are copied.
|
2013-10-07 10:28:59 +00:00
|
|
|
*/
|
2017-02-06 21:06:35 +00:00
|
|
|
unsigned long *numa_faults;
|
|
|
|
unsigned long total_numa_faults;
|
2013-10-07 10:28:59 +00:00
|
|
|
|
2013-10-07 10:29:36 +00:00
|
|
|
/*
|
|
|
|
* numa_faults_locality tracks if faults recorded during the last
|
2015-03-25 22:55:42 +00:00
|
|
|
* scan window were remote/local or failed to migrate. The task scan
|
|
|
|
* period is adapted based on the locality of the faults with different
|
|
|
|
* weights depending on whether they were shared or private faults
|
2013-10-07 10:29:36 +00:00
|
|
|
*/
|
2017-02-06 21:06:35 +00:00
|
|
|
unsigned long numa_faults_locality[3];
|
2013-10-07 10:29:36 +00:00
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
unsigned long numa_pages_migrated;
|
2012-10-25 12:16:43 +00:00
|
|
|
#endif /* CONFIG_NUMA_BALANCING */
|
|
|
|
|
rseq: Introduce restartable sequences system call
Expose a new system call allowing each thread to register one userspace
memory area to be used as an ABI between kernel and user-space for two
purposes: user-space restartable sequences and quick access to read the
current CPU number value from user-space.
* Restartable sequences (per-cpu atomics)
Restartables sequences allow user-space to perform update operations on
per-cpu data without requiring heavy-weight atomic operations.
The restartable critical sections (percpu atomics) work has been started
by Paul Turner and Andrew Hunter. It lets the kernel handle restart of
critical sections. [1] [2] The re-implementation proposed here brings a
few simplifications to the ABI which facilitates porting to other
architectures and speeds up the user-space fast path.
Here are benchmarks of various rseq use-cases.
Test hardware:
arm32: ARMv7 Processor rev 4 (v7l) "Cubietruck", 2-core
x86-64: Intel E5-2630 v3@2.40GHz, 16-core, hyperthreading
The following benchmarks were all performed on a single thread.
* Per-CPU statistic counter increment
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 344.0 31.4 11.0
x86-64: 15.3 2.0 7.7
* LTTng-UST: write event 32-bit header, 32-bit payload into tracer
per-cpu buffer
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 2502.0 2250.0 1.1
x86-64: 117.4 98.0 1.2
* liburcu percpu: lock-unlock pair, dereference, read/compare word
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 751.0 128.5 5.8
x86-64: 53.4 28.6 1.9
* jemalloc memory allocator adapted to use rseq
Using rseq with per-cpu memory pools in jemalloc at Facebook (based on
rseq 2016 implementation):
The production workload response-time has 1-2% gain avg. latency, and
the P99 overall latency drops by 2-3%.
* Reading the current CPU number
Speeding up reading the current CPU number on which the caller thread is
running is done by keeping the current CPU number up do date within the
cpu_id field of the memory area registered by the thread. This is done
by making scheduler preemption set the TIF_NOTIFY_RESUME flag on the
current thread. Upon return to user-space, a notify-resume handler
updates the current CPU value within the registered user-space memory
area. User-space can then read the current CPU number directly from
memory.
Keeping the current cpu id in a memory area shared between kernel and
user-space is an improvement over current mechanisms available to read
the current CPU number, which has the following benefits over
alternative approaches:
- 35x speedup on ARM vs system call through glibc
- 20x speedup on x86 compared to calling glibc, which calls vdso
executing a "lsl" instruction,
- 14x speedup on x86 compared to inlined "lsl" instruction,
- Unlike vdso approaches, this cpu_id value can be read from an inline
assembly, which makes it a useful building block for restartable
sequences.
- The approach of reading the cpu id through memory mapping shared
between kernel and user-space is portable (e.g. ARM), which is not the
case for the lsl-based x86 vdso.
On x86, yet another possible approach would be to use the gs segment
selector to point to user-space per-cpu data. This approach performs
similarly to the cpu id cache, but it has two disadvantages: it is
not portable, and it is incompatible with existing applications already
using the gs segment selector for other purposes.
Benchmarking various approaches for reading the current CPU number:
ARMv7 Processor rev 4 (v7l)
Machine model: Cubietruck
- Baseline (empty loop): 8.4 ns
- Read CPU from rseq cpu_id: 16.7 ns
- Read CPU from rseq cpu_id (lazy register): 19.8 ns
- glibc 2.19-0ubuntu6.6 getcpu: 301.8 ns
- getcpu system call: 234.9 ns
x86-64 Intel(R) Xeon(R) CPU E5-2630 v3 @ 2.40GHz:
- Baseline (empty loop): 0.8 ns
- Read CPU from rseq cpu_id: 0.8 ns
- Read CPU from rseq cpu_id (lazy register): 0.8 ns
- Read using gs segment selector: 0.8 ns
- "lsl" inline assembly: 13.0 ns
- glibc 2.19-0ubuntu6 getcpu: 16.6 ns
- getcpu system call: 53.9 ns
- Speed (benchmark taken on v8 of patchset)
Running 10 runs of hackbench -l 100000 seems to indicate, contrary to
expectations, that enabling CONFIG_RSEQ slightly accelerates the
scheduler:
Configuration: 2 sockets * 8-core Intel(R) Xeon(R) CPU E5-2630 v3 @
2.40GHz (directly on hardware, hyperthreading disabled in BIOS, energy
saving disabled in BIOS, turboboost disabled in BIOS, cpuidle.off=1
kernel parameter), with a Linux v4.6 defconfig+localyesconfig,
restartable sequences series applied.
* CONFIG_RSEQ=n
avg.: 41.37 s
std.dev.: 0.36 s
* CONFIG_RSEQ=y
avg.: 40.46 s
std.dev.: 0.33 s
- Size
On x86-64, between CONFIG_RSEQ=n/y, the text size increase of vmlinux is
567 bytes, and the data size increase of vmlinux is 5696 bytes.
[1] https://lwn.net/Articles/650333/
[2] http://www.linuxplumbersconf.org/2013/ocw/system/presentations/1695/original/LPC%20-%20PerCpu%20Atomics.pdf
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Joel Fernandes <joelaf@google.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Dave Watson <davejwatson@fb.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Chris Lameter <cl@linux.com>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Andrew Hunter <ahh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: "Paul E . McKenney" <paulmck@linux.vnet.ibm.com>
Cc: Paul Turner <pjt@google.com>
Cc: Boqun Feng <boqun.feng@gmail.com>
Cc: Josh Triplett <josh@joshtriplett.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Ben Maurer <bmaurer@fb.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: linux-api@vger.kernel.org
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/20151027235635.16059.11630.stgit@pjt-glaptop.roam.corp.google.com
Link: http://lkml.kernel.org/r/20150624222609.6116.86035.stgit@kitami.mtv.corp.google.com
Link: https://lkml.kernel.org/r/20180602124408.8430-3-mathieu.desnoyers@efficios.com
2018-06-02 12:43:54 +00:00
|
|
|
#ifdef CONFIG_RSEQ
|
|
|
|
struct rseq __user *rseq;
|
|
|
|
u32 rseq_sig;
|
|
|
|
/*
|
|
|
|
* RmW on rseq_event_mask must be performed atomically
|
|
|
|
* with respect to preemption.
|
|
|
|
*/
|
|
|
|
unsigned long rseq_event_mask;
|
|
|
|
#endif
|
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
struct tlbflush_unmap_batch tlb_ubc;
|
2015-09-04 22:47:32 +00:00
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
struct rcu_head rcu;
|
2006-04-11 11:52:07 +00:00
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
/* Cache last used pipe for splice(): */
|
|
|
|
struct pipe_inode_info *splice_pipe;
|
net: use a per task frag allocator
We currently use a per socket order-0 page cache for tcp_sendmsg()
operations.
This page is used to build fragments for skbs.
Its done to increase probability of coalescing small write() into
single segments in skbs still in write queue (not yet sent)
But it wastes a lot of memory for applications handling many mostly
idle sockets, since each socket holds one page in sk->sk_sndmsg_page
Its also quite inefficient to build TSO 64KB packets, because we need
about 16 pages per skb on arches where PAGE_SIZE = 4096, so we hit
page allocator more than wanted.
This patch adds a per task frag allocator and uses bigger pages,
if available. An automatic fallback is done in case of memory pressure.
(up to 32768 bytes per frag, thats order-3 pages on x86)
This increases TCP stream performance by 20% on loopback device,
but also benefits on other network devices, since 8x less frags are
mapped on transmit and unmapped on tx completion. Alexander Duyck
mentioned a probable performance win on systems with IOMMU enabled.
Its possible some SG enabled hardware cant cope with bigger fragments,
but their ndo_start_xmit() should already handle this, splitting a
fragment in sub fragments, since some arches have PAGE_SIZE=65536
Successfully tested on various ethernet devices.
(ixgbe, igb, bnx2x, tg3, mellanox mlx4)
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Ben Hutchings <bhutchings@solarflare.com>
Cc: Vijay Subramanian <subramanian.vijay@gmail.com>
Cc: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Vijay Subramanian <subramanian.vijay@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-09-23 23:04:42 +00:00
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
struct page_frag task_frag;
|
net: use a per task frag allocator
We currently use a per socket order-0 page cache for tcp_sendmsg()
operations.
This page is used to build fragments for skbs.
Its done to increase probability of coalescing small write() into
single segments in skbs still in write queue (not yet sent)
But it wastes a lot of memory for applications handling many mostly
idle sockets, since each socket holds one page in sk->sk_sndmsg_page
Its also quite inefficient to build TSO 64KB packets, because we need
about 16 pages per skb on arches where PAGE_SIZE = 4096, so we hit
page allocator more than wanted.
This patch adds a per task frag allocator and uses bigger pages,
if available. An automatic fallback is done in case of memory pressure.
(up to 32768 bytes per frag, thats order-3 pages on x86)
This increases TCP stream performance by 20% on loopback device,
but also benefits on other network devices, since 8x less frags are
mapped on transmit and unmapped on tx completion. Alexander Duyck
mentioned a probable performance win on systems with IOMMU enabled.
Its possible some SG enabled hardware cant cope with bigger fragments,
but their ndo_start_xmit() should already handle this, splitting a
fragment in sub fragments, since some arches have PAGE_SIZE=65536
Successfully tested on various ethernet devices.
(ixgbe, igb, bnx2x, tg3, mellanox mlx4)
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Ben Hutchings <bhutchings@solarflare.com>
Cc: Vijay Subramanian <subramanian.vijay@gmail.com>
Cc: Alexander Duyck <alexander.h.duyck@intel.com>
Tested-by: Vijay Subramanian <subramanian.vijay@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2012-09-23 23:04:42 +00:00
|
|
|
|
2017-02-01 17:00:26 +00:00
|
|
|
#ifdef CONFIG_TASK_DELAY_ACCT
|
|
|
|
struct task_delay_info *delays;
|
2006-12-08 10:39:47 +00:00
|
|
|
#endif
|
2017-02-01 17:00:26 +00:00
|
|
|
|
2006-12-08 10:39:47 +00:00
|
|
|
#ifdef CONFIG_FAULT_INJECTION
|
2017-02-06 21:06:35 +00:00
|
|
|
int make_it_fail;
|
2017-07-14 21:49:52 +00:00
|
|
|
unsigned int fail_nth;
|
2006-07-14 07:24:36 +00:00
|
|
|
#endif
|
writeback: per task dirty rate limit
Add two fields to task_struct.
1) account dirtied pages in the individual tasks, for accuracy
2) per-task balance_dirty_pages() call intervals, for flexibility
The balance_dirty_pages() call interval (ie. nr_dirtied_pause) will
scale near-sqrt to the safety gap between dirty pages and threshold.
The main problem of per-task nr_dirtied is, if 1k+ tasks start dirtying
pages at exactly the same time, each task will be assigned a large
initial nr_dirtied_pause, so that the dirty threshold will be exceeded
long before each task reached its nr_dirtied_pause and hence call
balance_dirty_pages().
The solution is to watch for the number of pages dirtied on each CPU in
between the calls into balance_dirty_pages(). If it exceeds ratelimit_pages
(3% dirty threshold), force call balance_dirty_pages() for a chance to
set bdi->dirty_exceeded. In normal situations, this safeguarding
condition is not expected to trigger at all.
On the sqrt in dirty_poll_interval():
It will serve as an initial guess when dirty pages are still in the
freerun area.
When dirty pages are floating inside the dirty control scope [freerun,
limit], a followup patch will use some refined dirty poll interval to
get the desired pause time.
thresh-dirty (MB) sqrt
1 16
2 22
4 32
8 45
16 64
32 90
64 128
128 181
256 256
512 362
1024 512
The above table means, given 1MB (or 1GB) gap and the dd tasks polling
balance_dirty_pages() on every 16 (or 512) pages, the dirty limit won't
be exceeded as long as there are less than 16 (or 512) concurrent dd's.
So sqrt naturally leads to less overheads and more safe concurrent tasks
for large memory servers, which have large (thresh-freerun) gaps.
peter: keep the per-CPU ratelimit for safeguarding the 1k+ tasks case
CC: Peter Zijlstra <a.p.zijlstra@chello.nl>
Reviewed-by: Andrea Righi <andrea@betterlinux.com>
Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
2011-06-12 00:10:12 +00:00
|
|
|
/*
|
2017-02-06 21:06:35 +00:00
|
|
|
* When (nr_dirtied >= nr_dirtied_pause), it's time to call
|
|
|
|
* balance_dirty_pages() for a dirty throttling pause:
|
writeback: per task dirty rate limit
Add two fields to task_struct.
1) account dirtied pages in the individual tasks, for accuracy
2) per-task balance_dirty_pages() call intervals, for flexibility
The balance_dirty_pages() call interval (ie. nr_dirtied_pause) will
scale near-sqrt to the safety gap between dirty pages and threshold.
The main problem of per-task nr_dirtied is, if 1k+ tasks start dirtying
pages at exactly the same time, each task will be assigned a large
initial nr_dirtied_pause, so that the dirty threshold will be exceeded
long before each task reached its nr_dirtied_pause and hence call
balance_dirty_pages().
The solution is to watch for the number of pages dirtied on each CPU in
between the calls into balance_dirty_pages(). If it exceeds ratelimit_pages
(3% dirty threshold), force call balance_dirty_pages() for a chance to
set bdi->dirty_exceeded. In normal situations, this safeguarding
condition is not expected to trigger at all.
On the sqrt in dirty_poll_interval():
It will serve as an initial guess when dirty pages are still in the
freerun area.
When dirty pages are floating inside the dirty control scope [freerun,
limit], a followup patch will use some refined dirty poll interval to
get the desired pause time.
thresh-dirty (MB) sqrt
1 16
2 22
4 32
8 45
16 64
32 90
64 128
128 181
256 256
512 362
1024 512
The above table means, given 1MB (or 1GB) gap and the dd tasks polling
balance_dirty_pages() on every 16 (or 512) pages, the dirty limit won't
be exceeded as long as there are less than 16 (or 512) concurrent dd's.
So sqrt naturally leads to less overheads and more safe concurrent tasks
for large memory servers, which have large (thresh-freerun) gaps.
peter: keep the per-CPU ratelimit for safeguarding the 1k+ tasks case
CC: Peter Zijlstra <a.p.zijlstra@chello.nl>
Reviewed-by: Andrea Righi <andrea@betterlinux.com>
Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
2011-06-12 00:10:12 +00:00
|
|
|
*/
|
2017-02-06 21:06:35 +00:00
|
|
|
int nr_dirtied;
|
|
|
|
int nr_dirtied_pause;
|
|
|
|
/* Start of a write-and-pause period: */
|
|
|
|
unsigned long dirty_paused_when;
|
writeback: per task dirty rate limit
Add two fields to task_struct.
1) account dirtied pages in the individual tasks, for accuracy
2) per-task balance_dirty_pages() call intervals, for flexibility
The balance_dirty_pages() call interval (ie. nr_dirtied_pause) will
scale near-sqrt to the safety gap between dirty pages and threshold.
The main problem of per-task nr_dirtied is, if 1k+ tasks start dirtying
pages at exactly the same time, each task will be assigned a large
initial nr_dirtied_pause, so that the dirty threshold will be exceeded
long before each task reached its nr_dirtied_pause and hence call
balance_dirty_pages().
The solution is to watch for the number of pages dirtied on each CPU in
between the calls into balance_dirty_pages(). If it exceeds ratelimit_pages
(3% dirty threshold), force call balance_dirty_pages() for a chance to
set bdi->dirty_exceeded. In normal situations, this safeguarding
condition is not expected to trigger at all.
On the sqrt in dirty_poll_interval():
It will serve as an initial guess when dirty pages are still in the
freerun area.
When dirty pages are floating inside the dirty control scope [freerun,
limit], a followup patch will use some refined dirty poll interval to
get the desired pause time.
thresh-dirty (MB) sqrt
1 16
2 22
4 32
8 45
16 64
32 90
64 128
128 181
256 256
512 362
1024 512
The above table means, given 1MB (or 1GB) gap and the dd tasks polling
balance_dirty_pages() on every 16 (or 512) pages, the dirty limit won't
be exceeded as long as there are less than 16 (or 512) concurrent dd's.
So sqrt naturally leads to less overheads and more safe concurrent tasks
for large memory servers, which have large (thresh-freerun) gaps.
peter: keep the per-CPU ratelimit for safeguarding the 1k+ tasks case
CC: Peter Zijlstra <a.p.zijlstra@chello.nl>
Reviewed-by: Andrea Righi <andrea@betterlinux.com>
Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
2011-06-12 00:10:12 +00:00
|
|
|
|
2008-01-25 20:08:34 +00:00
|
|
|
#ifdef CONFIG_LATENCYTOP
|
2017-02-06 21:06:35 +00:00
|
|
|
int latency_record_count;
|
|
|
|
struct latency_record latency_record[LT_SAVECOUNT];
|
2008-01-25 20:08:34 +00:00
|
|
|
#endif
|
2008-09-01 22:52:40 +00:00
|
|
|
/*
|
2017-02-06 21:06:35 +00:00
|
|
|
* Time slack values; these are used to round up poll() and
|
2008-09-01 22:52:40 +00:00
|
|
|
* select() etc timeout values. These are in nanoseconds.
|
|
|
|
*/
|
2017-02-06 21:06:35 +00:00
|
|
|
u64 timer_slack_ns;
|
|
|
|
u64 default_timer_slack_ns;
|
2008-11-06 08:37:40 +00:00
|
|
|
|
kasan: add kernel address sanitizer infrastructure
Kernel Address sanitizer (KASan) is a dynamic memory error detector. It
provides fast and comprehensive solution for finding use-after-free and
out-of-bounds bugs.
KASAN uses compile-time instrumentation for checking every memory access,
therefore GCC > v4.9.2 required. v4.9.2 almost works, but has issues with
putting symbol aliases into the wrong section, which breaks kasan
instrumentation of globals.
This patch only adds infrastructure for kernel address sanitizer. It's
not available for use yet. The idea and some code was borrowed from [1].
Basic idea:
The main idea of KASAN is to use shadow memory to record whether each byte
of memory is safe to access or not, and use compiler's instrumentation to
check the shadow memory on each memory access.
Address sanitizer uses 1/8 of the memory addressable in kernel for shadow
memory and uses direct mapping with a scale and offset to translate a
memory address to its corresponding shadow address.
Here is function to translate address to corresponding shadow address:
unsigned long kasan_mem_to_shadow(unsigned long addr)
{
return (addr >> KASAN_SHADOW_SCALE_SHIFT) + KASAN_SHADOW_OFFSET;
}
where KASAN_SHADOW_SCALE_SHIFT = 3.
So for every 8 bytes there is one corresponding byte of shadow memory.
The following encoding used for each shadow byte: 0 means that all 8 bytes
of the corresponding memory region are valid for access; k (1 <= k <= 7)
means that the first k bytes are valid for access, and other (8 - k) bytes
are not; Any negative value indicates that the entire 8-bytes are
inaccessible. Different negative values used to distinguish between
different kinds of inaccessible memory (redzones, freed memory) (see
mm/kasan/kasan.h).
To be able to detect accesses to bad memory we need a special compiler.
Such compiler inserts a specific function calls (__asan_load*(addr),
__asan_store*(addr)) before each memory access of size 1, 2, 4, 8 or 16.
These functions check whether memory region is valid to access or not by
checking corresponding shadow memory. If access is not valid an error
printed.
Historical background of the address sanitizer from Dmitry Vyukov:
"We've developed the set of tools, AddressSanitizer (Asan),
ThreadSanitizer and MemorySanitizer, for user space. We actively use
them for testing inside of Google (continuous testing, fuzzing,
running prod services). To date the tools have found more than 10'000
scary bugs in Chromium, Google internal codebase and various
open-source projects (Firefox, OpenSSL, gcc, clang, ffmpeg, MySQL and
lots of others): [2] [3] [4].
The tools are part of both gcc and clang compilers.
We have not yet done massive testing under the Kernel AddressSanitizer
(it's kind of chicken and egg problem, you need it to be upstream to
start applying it extensively). To date it has found about 50 bugs.
Bugs that we've found in upstream kernel are listed in [5].
We've also found ~20 bugs in out internal version of the kernel. Also
people from Samsung and Oracle have found some.
[...]
As others noted, the main feature of AddressSanitizer is its
performance due to inline compiler instrumentation and simple linear
shadow memory. User-space Asan has ~2x slowdown on computational
programs and ~2x memory consumption increase. Taking into account that
kernel usually consumes only small fraction of CPU and memory when
running real user-space programs, I would expect that kernel Asan will
have ~10-30% slowdown and similar memory consumption increase (when we
finish all tuning).
I agree that Asan can well replace kmemcheck. We have plans to start
working on Kernel MemorySanitizer that finds uses of unitialized
memory. Asan+Msan will provide feature-parity with kmemcheck. As
others noted, Asan will unlikely replace debug slab and pagealloc that
can be enabled at runtime. Asan uses compiler instrumentation, so even
if it is disabled, it still incurs visible overheads.
Asan technology is easily portable to other architectures. Compiler
instrumentation is fully portable. Runtime has some arch-dependent
parts like shadow mapping and atomic operation interception. They are
relatively easy to port."
Comparison with other debugging features:
========================================
KMEMCHECK:
- KASan can do almost everything that kmemcheck can. KASan uses
compile-time instrumentation, which makes it significantly faster than
kmemcheck. The only advantage of kmemcheck over KASan is detection of
uninitialized memory reads.
Some brief performance testing showed that kasan could be
x500-x600 times faster than kmemcheck:
$ netperf -l 30
MIGRATED TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to localhost (127.0.0.1) port 0 AF_INET
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/sec
no debug: 87380 16384 16384 30.00 41624.72
kasan inline: 87380 16384 16384 30.00 12870.54
kasan outline: 87380 16384 16384 30.00 10586.39
kmemcheck: 87380 16384 16384 30.03 20.23
- Also kmemcheck couldn't work on several CPUs. It always sets
number of CPUs to 1. KASan doesn't have such limitation.
DEBUG_PAGEALLOC:
- KASan is slower than DEBUG_PAGEALLOC, but KASan works on sub-page
granularity level, so it able to find more bugs.
SLUB_DEBUG (poisoning, redzones):
- SLUB_DEBUG has lower overhead than KASan.
- SLUB_DEBUG in most cases are not able to detect bad reads,
KASan able to detect both reads and writes.
- In some cases (e.g. redzone overwritten) SLUB_DEBUG detect
bugs only on allocation/freeing of object. KASan catch
bugs right before it will happen, so we always know exact
place of first bad read/write.
[1] https://code.google.com/p/address-sanitizer/wiki/AddressSanitizerForKernel
[2] https://code.google.com/p/address-sanitizer/wiki/FoundBugs
[3] https://code.google.com/p/thread-sanitizer/wiki/FoundBugs
[4] https://code.google.com/p/memory-sanitizer/wiki/FoundBugs
[5] https://code.google.com/p/address-sanitizer/wiki/AddressSanitizerForKernel#Trophies
Based on work by Andrey Konovalov.
Signed-off-by: Andrey Ryabinin <a.ryabinin@samsung.com>
Acked-by: Michal Marek <mmarek@suse.cz>
Signed-off-by: Andrey Konovalov <adech.fo@gmail.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Konstantin Serebryany <kcc@google.com>
Cc: Dmitry Chernenkov <dmitryc@google.com>
Cc: Yuri Gribov <tetra2005@gmail.com>
Cc: Konstantin Khlebnikov <koct9i@gmail.com>
Cc: Sasha Levin <sasha.levin@oracle.com>
Cc: Christoph Lameter <cl@linux.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Christoph Lameter <cl@linux.com>
Cc: Pekka Enberg <penberg@kernel.org>
Cc: David Rientjes <rientjes@google.com>
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-02-13 22:39:17 +00:00
|
|
|
#ifdef CONFIG_KASAN
|
2017-02-06 21:06:35 +00:00
|
|
|
unsigned int kasan_depth;
|
kasan: add kernel address sanitizer infrastructure
Kernel Address sanitizer (KASan) is a dynamic memory error detector. It
provides fast and comprehensive solution for finding use-after-free and
out-of-bounds bugs.
KASAN uses compile-time instrumentation for checking every memory access,
therefore GCC > v4.9.2 required. v4.9.2 almost works, but has issues with
putting symbol aliases into the wrong section, which breaks kasan
instrumentation of globals.
This patch only adds infrastructure for kernel address sanitizer. It's
not available for use yet. The idea and some code was borrowed from [1].
Basic idea:
The main idea of KASAN is to use shadow memory to record whether each byte
of memory is safe to access or not, and use compiler's instrumentation to
check the shadow memory on each memory access.
Address sanitizer uses 1/8 of the memory addressable in kernel for shadow
memory and uses direct mapping with a scale and offset to translate a
memory address to its corresponding shadow address.
Here is function to translate address to corresponding shadow address:
unsigned long kasan_mem_to_shadow(unsigned long addr)
{
return (addr >> KASAN_SHADOW_SCALE_SHIFT) + KASAN_SHADOW_OFFSET;
}
where KASAN_SHADOW_SCALE_SHIFT = 3.
So for every 8 bytes there is one corresponding byte of shadow memory.
The following encoding used for each shadow byte: 0 means that all 8 bytes
of the corresponding memory region are valid for access; k (1 <= k <= 7)
means that the first k bytes are valid for access, and other (8 - k) bytes
are not; Any negative value indicates that the entire 8-bytes are
inaccessible. Different negative values used to distinguish between
different kinds of inaccessible memory (redzones, freed memory) (see
mm/kasan/kasan.h).
To be able to detect accesses to bad memory we need a special compiler.
Such compiler inserts a specific function calls (__asan_load*(addr),
__asan_store*(addr)) before each memory access of size 1, 2, 4, 8 or 16.
These functions check whether memory region is valid to access or not by
checking corresponding shadow memory. If access is not valid an error
printed.
Historical background of the address sanitizer from Dmitry Vyukov:
"We've developed the set of tools, AddressSanitizer (Asan),
ThreadSanitizer and MemorySanitizer, for user space. We actively use
them for testing inside of Google (continuous testing, fuzzing,
running prod services). To date the tools have found more than 10'000
scary bugs in Chromium, Google internal codebase and various
open-source projects (Firefox, OpenSSL, gcc, clang, ffmpeg, MySQL and
lots of others): [2] [3] [4].
The tools are part of both gcc and clang compilers.
We have not yet done massive testing under the Kernel AddressSanitizer
(it's kind of chicken and egg problem, you need it to be upstream to
start applying it extensively). To date it has found about 50 bugs.
Bugs that we've found in upstream kernel are listed in [5].
We've also found ~20 bugs in out internal version of the kernel. Also
people from Samsung and Oracle have found some.
[...]
As others noted, the main feature of AddressSanitizer is its
performance due to inline compiler instrumentation and simple linear
shadow memory. User-space Asan has ~2x slowdown on computational
programs and ~2x memory consumption increase. Taking into account that
kernel usually consumes only small fraction of CPU and memory when
running real user-space programs, I would expect that kernel Asan will
have ~10-30% slowdown and similar memory consumption increase (when we
finish all tuning).
I agree that Asan can well replace kmemcheck. We have plans to start
working on Kernel MemorySanitizer that finds uses of unitialized
memory. Asan+Msan will provide feature-parity with kmemcheck. As
others noted, Asan will unlikely replace debug slab and pagealloc that
can be enabled at runtime. Asan uses compiler instrumentation, so even
if it is disabled, it still incurs visible overheads.
Asan technology is easily portable to other architectures. Compiler
instrumentation is fully portable. Runtime has some arch-dependent
parts like shadow mapping and atomic operation interception. They are
relatively easy to port."
Comparison with other debugging features:
========================================
KMEMCHECK:
- KASan can do almost everything that kmemcheck can. KASan uses
compile-time instrumentation, which makes it significantly faster than
kmemcheck. The only advantage of kmemcheck over KASan is detection of
uninitialized memory reads.
Some brief performance testing showed that kasan could be
x500-x600 times faster than kmemcheck:
$ netperf -l 30
MIGRATED TCP STREAM TEST from 0.0.0.0 (0.0.0.0) port 0 AF_INET to localhost (127.0.0.1) port 0 AF_INET
Recv Send Send
Socket Socket Message Elapsed
Size Size Size Time Throughput
bytes bytes bytes secs. 10^6bits/sec
no debug: 87380 16384 16384 30.00 41624.72
kasan inline: 87380 16384 16384 30.00 12870.54
kasan outline: 87380 16384 16384 30.00 10586.39
kmemcheck: 87380 16384 16384 30.03 20.23
- Also kmemcheck couldn't work on several CPUs. It always sets
number of CPUs to 1. KASan doesn't have such limitation.
DEBUG_PAGEALLOC:
- KASan is slower than DEBUG_PAGEALLOC, but KASan works on sub-page
granularity level, so it able to find more bugs.
SLUB_DEBUG (poisoning, redzones):
- SLUB_DEBUG has lower overhead than KASan.
- SLUB_DEBUG in most cases are not able to detect bad reads,
KASan able to detect both reads and writes.
- In some cases (e.g. redzone overwritten) SLUB_DEBUG detect
bugs only on allocation/freeing of object. KASan catch
bugs right before it will happen, so we always know exact
place of first bad read/write.
[1] https://code.google.com/p/address-sanitizer/wiki/AddressSanitizerForKernel
[2] https://code.google.com/p/address-sanitizer/wiki/FoundBugs
[3] https://code.google.com/p/thread-sanitizer/wiki/FoundBugs
[4] https://code.google.com/p/memory-sanitizer/wiki/FoundBugs
[5] https://code.google.com/p/address-sanitizer/wiki/AddressSanitizerForKernel#Trophies
Based on work by Andrey Konovalov.
Signed-off-by: Andrey Ryabinin <a.ryabinin@samsung.com>
Acked-by: Michal Marek <mmarek@suse.cz>
Signed-off-by: Andrey Konovalov <adech.fo@gmail.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Konstantin Serebryany <kcc@google.com>
Cc: Dmitry Chernenkov <dmitryc@google.com>
Cc: Yuri Gribov <tetra2005@gmail.com>
Cc: Konstantin Khlebnikov <koct9i@gmail.com>
Cc: Sasha Levin <sasha.levin@oracle.com>
Cc: Christoph Lameter <cl@linux.com>
Cc: Joonsoo Kim <iamjoonsoo.kim@lge.com>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Christoph Lameter <cl@linux.com>
Cc: Pekka Enberg <penberg@kernel.org>
Cc: David Rientjes <rientjes@google.com>
Cc: Stephen Rothwell <sfr@canb.auug.org.au>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-02-13 22:39:17 +00:00
|
|
|
#endif
|
2017-02-06 21:06:35 +00:00
|
|
|
|
2008-11-25 20:07:04 +00:00
|
|
|
#ifdef CONFIG_FUNCTION_GRAPH_TRACER
|
2017-02-06 21:06:35 +00:00
|
|
|
/* Index of current stored address in ret_stack: */
|
|
|
|
int curr_ret_stack;
|
function_graph: Use new curr_ret_depth to manage depth instead of curr_ret_stack
Currently, the depth of the ret_stack is determined by curr_ret_stack index.
The issue is that there's a race between setting of the curr_ret_stack and
calling of the callback attached to the return of the function.
Commit 03274a3ffb44 ("tracing/fgraph: Adjust fgraph depth before calling
trace return callback") moved the calling of the callback to after the
setting of the curr_ret_stack, even stating that it was safe to do so, when
in fact, it was the reason there was a barrier() there (yes, I should have
commented that barrier()).
Not only does the curr_ret_stack keep track of the current call graph depth,
it also keeps the ret_stack content from being overwritten by new data.
The function profiler, uses the "subtime" variable of ret_stack structure
and by moving the curr_ret_stack, it allows for interrupts to use the same
structure it was using, corrupting the data, and breaking the profiler.
To fix this, there needs to be two variables to handle the call stack depth
and the pointer to where the ret_stack is being used, as they need to change
at two different locations.
Cc: stable@kernel.org
Fixes: 03274a3ffb449 ("tracing/fgraph: Adjust fgraph depth before calling trace return callback")
Reviewed-by: Masami Hiramatsu <mhiramat@kernel.org>
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
2018-11-19 13:07:12 +00:00
|
|
|
int curr_ret_depth;
|
2017-02-06 21:06:35 +00:00
|
|
|
|
|
|
|
/* Stack of return addresses for return function tracing: */
|
|
|
|
struct ftrace_ret_stack *ret_stack;
|
|
|
|
|
|
|
|
/* Timestamp for last schedule: */
|
|
|
|
unsigned long long ftrace_timestamp;
|
|
|
|
|
2008-11-23 05:22:56 +00:00
|
|
|
/*
|
|
|
|
* Number of functions that haven't been traced
|
2017-02-06 21:06:35 +00:00
|
|
|
* because of depth overrun:
|
2008-11-23 05:22:56 +00:00
|
|
|
*/
|
2017-02-06 21:06:35 +00:00
|
|
|
atomic_t trace_overrun;
|
|
|
|
|
|
|
|
/* Pause tracing: */
|
|
|
|
atomic_t tracing_graph_pause;
|
2008-11-23 05:22:56 +00:00
|
|
|
#endif
|
2017-02-06 21:06:35 +00:00
|
|
|
|
2008-12-03 20:36:57 +00:00
|
|
|
#ifdef CONFIG_TRACING
|
2017-02-06 21:06:35 +00:00
|
|
|
/* State flags for use by tracers: */
|
|
|
|
unsigned long trace;
|
|
|
|
|
|
|
|
/* Bitmask and counter of trace recursion: */
|
|
|
|
unsigned long trace_recursion;
|
tracing: add same level recursion detection
The tracing infrastructure allows for recursion. That is, an interrupt
may interrupt the act of tracing an event, and that interrupt may very well
perform its own trace. This is a recursive trace, and is fine to do.
The problem arises when there is a bug, and the utility doing the trace
calls something that recurses back into the tracer. This recursion is not
caused by an external event like an interrupt, but by code that is not
expected to recurse. The result could be a lockup.
This patch adds a bitmask to the task structure that keeps track
of the trace recursion. To find the interrupt depth, the following
algorithm is used:
level = hardirq_count() + softirq_count() + in_nmi;
Here, level will be the depth of interrutps and softirqs, and even handles
the nmi. Then the corresponding bit is set in the recursion bitmask.
If the bit was already set, we know we had a recursion at the same level
and we warn about it and fail the writing to the buffer.
After the data has been committed to the buffer, we clear the bit.
No atomics are needed. The only races are with interrupts and they reset
the bitmask before returning anywy.
[ Impact: detect same irq level trace recursion ]
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
2009-04-17 01:41:52 +00:00
|
|
|
#endif /* CONFIG_TRACING */
|
2017-02-06 21:06:35 +00:00
|
|
|
|
kernel: add kcov code coverage
kcov provides code coverage collection for coverage-guided fuzzing
(randomized testing). Coverage-guided fuzzing is a testing technique
that uses coverage feedback to determine new interesting inputs to a
system. A notable user-space example is AFL
(http://lcamtuf.coredump.cx/afl/). However, this technique is not
widely used for kernel testing due to missing compiler and kernel
support.
kcov does not aim to collect as much coverage as possible. It aims to
collect more or less stable coverage that is function of syscall inputs.
To achieve this goal it does not collect coverage in soft/hard
interrupts and instrumentation of some inherently non-deterministic or
non-interesting parts of kernel is disbled (e.g. scheduler, locking).
Currently there is a single coverage collection mode (tracing), but the
API anticipates additional collection modes. Initially I also
implemented a second mode which exposes coverage in a fixed-size hash
table of counters (what Quentin used in his original patch). I've
dropped the second mode for simplicity.
This patch adds the necessary support on kernel side. The complimentary
compiler support was added in gcc revision 231296.
We've used this support to build syzkaller system call fuzzer, which has
found 90 kernel bugs in just 2 months:
https://github.com/google/syzkaller/wiki/Found-Bugs
We've also found 30+ bugs in our internal systems with syzkaller.
Another (yet unexplored) direction where kcov coverage would greatly
help is more traditional "blob mutation". For example, mounting a
random blob as a filesystem, or receiving a random blob over wire.
Why not gcov. Typical fuzzing loop looks as follows: (1) reset
coverage, (2) execute a bit of code, (3) collect coverage, repeat. A
typical coverage can be just a dozen of basic blocks (e.g. an invalid
input). In such context gcov becomes prohibitively expensive as
reset/collect coverage steps depend on total number of basic
blocks/edges in program (in case of kernel it is about 2M). Cost of
kcov depends only on number of executed basic blocks/edges. On top of
that, kernel requires per-thread coverage because there are always
background threads and unrelated processes that also produce coverage.
With inlined gcov instrumentation per-thread coverage is not possible.
kcov exposes kernel PCs and control flow to user-space which is
insecure. But debugfs should not be mapped as user accessible.
Based on a patch by Quentin Casasnovas.
[akpm@linux-foundation.org: make task_struct.kcov_mode have type `enum kcov_mode']
[akpm@linux-foundation.org: unbreak allmodconfig]
[akpm@linux-foundation.org: follow x86 Makefile layout standards]
Signed-off-by: Dmitry Vyukov <dvyukov@google.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Cc: syzkaller <syzkaller@googlegroups.com>
Cc: Vegard Nossum <vegard.nossum@oracle.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Tavis Ormandy <taviso@google.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Quentin Casasnovas <quentin.casasnovas@oracle.com>
Cc: Kostya Serebryany <kcc@google.com>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Alexander Potapenko <glider@google.com>
Cc: Kees Cook <keescook@google.com>
Cc: Bjorn Helgaas <bhelgaas@google.com>
Cc: Sasha Levin <sasha.levin@oracle.com>
Cc: David Drysdale <drysdale@google.com>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Andrey Ryabinin <ryabinin.a.a@gmail.com>
Cc: Kirill A. Shutemov <kirill@shutemov.name>
Cc: Jiri Slaby <jslaby@suse.cz>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-03-22 21:27:30 +00:00
|
|
|
#ifdef CONFIG_KCOV
|
2017-02-06 21:06:35 +00:00
|
|
|
/* Coverage collection mode enabled for this task (0 if disabled): */
|
sched/core / kcov: avoid kcov_area during task switch
During a context switch, we first switch_mm() to the next task's mm,
then switch_to() that new task. This means that vmalloc'd regions which
had previously been faulted in can transiently disappear in the context
of the prev task.
Functions instrumented by KCOV may try to access a vmalloc'd kcov_area
during this window, and as the fault handling code is instrumented, this
results in a recursive fault.
We must avoid accessing any kcov_area during this window. We can do so
with a new flag in kcov_mode, set prior to switching the mm, and cleared
once the new task is live. Since task_struct::kcov_mode isn't always a
specific enum kcov_mode value, this is made an unsigned int.
The manipulation is hidden behind kcov_{prepare,finish}_switch() helpers,
which are empty for !CONFIG_KCOV kernels.
The code uses macros because I can't use static inline functions without a
circular include dependency between <linux/sched.h> and <linux/kcov.h>,
since the definition of task_struct uses things defined in <linux/kcov.h>
Link: http://lkml.kernel.org/r/20180504135535.53744-4-mark.rutland@arm.com
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Acked-by: Andrey Ryabinin <aryabinin@virtuozzo.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2018-06-14 22:27:41 +00:00
|
|
|
unsigned int kcov_mode;
|
2017-02-06 21:06:35 +00:00
|
|
|
|
|
|
|
/* Size of the kcov_area: */
|
|
|
|
unsigned int kcov_size;
|
|
|
|
|
|
|
|
/* Buffer for coverage collection: */
|
|
|
|
void *kcov_area;
|
|
|
|
|
|
|
|
/* KCOV descriptor wired with this task or NULL: */
|
|
|
|
struct kcov *kcov;
|
kernel: add kcov code coverage
kcov provides code coverage collection for coverage-guided fuzzing
(randomized testing). Coverage-guided fuzzing is a testing technique
that uses coverage feedback to determine new interesting inputs to a
system. A notable user-space example is AFL
(http://lcamtuf.coredump.cx/afl/). However, this technique is not
widely used for kernel testing due to missing compiler and kernel
support.
kcov does not aim to collect as much coverage as possible. It aims to
collect more or less stable coverage that is function of syscall inputs.
To achieve this goal it does not collect coverage in soft/hard
interrupts and instrumentation of some inherently non-deterministic or
non-interesting parts of kernel is disbled (e.g. scheduler, locking).
Currently there is a single coverage collection mode (tracing), but the
API anticipates additional collection modes. Initially I also
implemented a second mode which exposes coverage in a fixed-size hash
table of counters (what Quentin used in his original patch). I've
dropped the second mode for simplicity.
This patch adds the necessary support on kernel side. The complimentary
compiler support was added in gcc revision 231296.
We've used this support to build syzkaller system call fuzzer, which has
found 90 kernel bugs in just 2 months:
https://github.com/google/syzkaller/wiki/Found-Bugs
We've also found 30+ bugs in our internal systems with syzkaller.
Another (yet unexplored) direction where kcov coverage would greatly
help is more traditional "blob mutation". For example, mounting a
random blob as a filesystem, or receiving a random blob over wire.
Why not gcov. Typical fuzzing loop looks as follows: (1) reset
coverage, (2) execute a bit of code, (3) collect coverage, repeat. A
typical coverage can be just a dozen of basic blocks (e.g. an invalid
input). In such context gcov becomes prohibitively expensive as
reset/collect coverage steps depend on total number of basic
blocks/edges in program (in case of kernel it is about 2M). Cost of
kcov depends only on number of executed basic blocks/edges. On top of
that, kernel requires per-thread coverage because there are always
background threads and unrelated processes that also produce coverage.
With inlined gcov instrumentation per-thread coverage is not possible.
kcov exposes kernel PCs and control flow to user-space which is
insecure. But debugfs should not be mapped as user accessible.
Based on a patch by Quentin Casasnovas.
[akpm@linux-foundation.org: make task_struct.kcov_mode have type `enum kcov_mode']
[akpm@linux-foundation.org: unbreak allmodconfig]
[akpm@linux-foundation.org: follow x86 Makefile layout standards]
Signed-off-by: Dmitry Vyukov <dvyukov@google.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Cc: syzkaller <syzkaller@googlegroups.com>
Cc: Vegard Nossum <vegard.nossum@oracle.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Tavis Ormandy <taviso@google.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Quentin Casasnovas <quentin.casasnovas@oracle.com>
Cc: Kostya Serebryany <kcc@google.com>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Alexander Potapenko <glider@google.com>
Cc: Kees Cook <keescook@google.com>
Cc: Bjorn Helgaas <bhelgaas@google.com>
Cc: Sasha Levin <sasha.levin@oracle.com>
Cc: David Drysdale <drysdale@google.com>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Cc: Andrey Ryabinin <ryabinin.a.a@gmail.com>
Cc: Kirill A. Shutemov <kirill@shutemov.name>
Cc: Jiri Slaby <jslaby@suse.cz>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-03-22 21:27:30 +00:00
|
|
|
#endif
|
2017-02-06 21:06:35 +00:00
|
|
|
|
2014-12-13 00:55:15 +00:00
|
|
|
#ifdef CONFIG_MEMCG
|
2017-02-06 21:06:35 +00:00
|
|
|
struct mem_cgroup *memcg_in_oom;
|
|
|
|
gfp_t memcg_oom_gfp_mask;
|
|
|
|
int memcg_oom_order;
|
memcg: punt high overage reclaim to return-to-userland path
Currently, try_charge() tries to reclaim memory synchronously when the
high limit is breached; however, if the allocation doesn't have
__GFP_WAIT, synchronous reclaim is skipped. If a process performs only
speculative allocations, it can blow way past the high limit. This is
actually easily reproducible by simply doing "find /". slab/slub
allocator tries speculative allocations first, so as long as there's
memory which can be consumed without blocking, it can keep allocating
memory regardless of the high limit.
This patch makes try_charge() always punt the over-high reclaim to the
return-to-userland path. If try_charge() detects that high limit is
breached, it adds the overage to current->memcg_nr_pages_over_high and
schedules execution of mem_cgroup_handle_over_high() which performs
synchronous reclaim from the return-to-userland path.
As long as kernel doesn't have a run-away allocation spree, this should
provide enough protection while making kmemcg behave more consistently.
It also has the following benefits.
- All over-high reclaims can use GFP_KERNEL regardless of the specific
gfp mask in use, e.g. GFP_NOFS, when the limit was breached.
- It copes with prio inversion. Previously, a low-prio task with
small memory.high might perform over-high reclaim with a bunch of
locks held. If a higher prio task needed any of these locks, it
would have to wait until the low prio task finished reclaim and
released the locks. By handing over-high reclaim to the task exit
path this issue can be avoided.
Signed-off-by: Tejun Heo <tj@kernel.org>
Acked-by: Michal Hocko <mhocko@kernel.org>
Reviewed-by: Vladimir Davydov <vdavydov@parallels.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2015-11-06 02:46:11 +00:00
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
/* Number of pages to reclaim on returning to userland: */
|
|
|
|
unsigned int memcg_nr_pages_over_high;
|
fs: fsnotify: account fsnotify metadata to kmemcg
Patch series "Directed kmem charging", v8.
The Linux kernel's memory cgroup allows limiting the memory usage of the
jobs running on the system to provide isolation between the jobs. All
the kernel memory allocated in the context of the job and marked with
__GFP_ACCOUNT will also be included in the memory usage and be limited
by the job's limit.
The kernel memory can only be charged to the memcg of the process in
whose context kernel memory was allocated. However there are cases
where the allocated kernel memory should be charged to the memcg
different from the current processes's memcg. This patch series
contains two such concrete use-cases i.e. fsnotify and buffer_head.
The fsnotify event objects can consume a lot of system memory for large
or unlimited queues if there is either no or slow listener. The events
are allocated in the context of the event producer. However they should
be charged to the event consumer. Similarly the buffer_head objects can
be allocated in a memcg different from the memcg of the page for which
buffer_head objects are being allocated.
To solve this issue, this patch series introduces mechanism to charge
kernel memory to a given memcg. In case of fsnotify events, the memcg
of the consumer can be used for charging and for buffer_head, the memcg
of the page can be charged. For directed charging, the caller can use
the scope API memalloc_[un]use_memcg() to specify the memcg to charge
for all the __GFP_ACCOUNT allocations within the scope.
This patch (of 2):
A lot of memory can be consumed by the events generated for the huge or
unlimited queues if there is either no or slow listener. This can cause
system level memory pressure or OOMs. So, it's better to account the
fsnotify kmem caches to the memcg of the listener.
However the listener can be in a different memcg than the memcg of the
producer and these allocations happen in the context of the event
producer. This patch introduces remote memcg charging API which the
producer can use to charge the allocations to the memcg of the listener.
There are seven fsnotify kmem caches and among them allocations from
dnotify_struct_cache, dnotify_mark_cache, fanotify_mark_cache and
inotify_inode_mark_cachep happens in the context of syscall from the
listener. So, SLAB_ACCOUNT is enough for these caches.
The objects from fsnotify_mark_connector_cachep are not accounted as
they are small compared to the notification mark or events and it is
unclear whom to account connector to since it is shared by all events
attached to the inode.
The allocations from the event caches happen in the context of the event
producer. For such caches we will need to remote charge the allocations
to the listener's memcg. Thus we save the memcg reference in the
fsnotify_group structure of the listener.
This patch has also moved the members of fsnotify_group to keep the size
same, at least for 64 bit build, even with additional member by filling
the holes.
[shakeelb@google.com: use GFP_KERNEL_ACCOUNT rather than open-coding it]
Link: http://lkml.kernel.org/r/20180702215439.211597-1-shakeelb@google.com
Link: http://lkml.kernel.org/r/20180627191250.209150-2-shakeelb@google.com
Signed-off-by: Shakeel Butt <shakeelb@google.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Cc: Michal Hocko <mhocko@kernel.org>
Cc: Jan Kara <jack@suse.cz>
Cc: Amir Goldstein <amir73il@gmail.com>
Cc: Greg Thelen <gthelen@google.com>
Cc: Vladimir Davydov <vdavydov.dev@gmail.com>
Cc: Roman Gushchin <guro@fb.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2018-08-17 22:46:39 +00:00
|
|
|
|
|
|
|
/* Used by memcontrol for targeted memcg charge: */
|
|
|
|
struct mem_cgroup *active_memcg;
|
2009-12-16 00:47:03 +00:00
|
|
|
#endif
|
2017-02-06 21:06:35 +00:00
|
|
|
|
2018-07-03 15:14:55 +00:00
|
|
|
#ifdef CONFIG_BLK_CGROUP
|
|
|
|
struct request_queue *throttle_queue;
|
|
|
|
#endif
|
|
|
|
|
uprobes/core: Handle breakpoint and singlestep exceptions
Uprobes uses exception notifiers to get to know if a thread hit
a breakpoint or a singlestep exception.
When a thread hits a uprobe or is singlestepping post a uprobe
hit, the uprobe exception notifier sets its TIF_UPROBE bit,
which will then be checked on its return to userspace path
(do_notify_resume() ->uprobe_notify_resume()), where the
consumers handlers are run (in task context) based on the
defined filters.
Uprobe hits are thread specific and hence we need to maintain
information about if a task hit a uprobe, what uprobe was hit,
the slot where the original instruction was copied for xol so
that it can be singlestepped with appropriate fixups.
In some cases, special care is needed for instructions that are
executed out of line (xol). These are architecture specific
artefacts, such as handling RIP relative instructions on x86_64.
Since the instruction at which the uprobe was inserted is
executed out of line, architecture specific fixups are added so
that the thread continues normal execution in the presence of a
uprobe.
Postpone the signals until we execute the probed insn.
post_xol() path does a recalc_sigpending() before return to
user-mode, this ensures the signal can't be lost.
Uprobes relies on DIE_DEBUG notification to notify if a
singlestep is complete.
Adds x86 specific uprobe exception notifiers and appropriate
hooks needed to determine a uprobe hit and subsequent post
processing.
Add requisite x86 fixups for xol for uprobes. Specific cases
needing fixups include relative jumps (x86_64), calls, etc.
Where possible, we check and skip singlestepping the
breakpointed instructions. For now we skip single byte as well
as few multibyte nop instructions. However this can be extended
to other instructions too.
Credits to Oleg Nesterov for suggestions/patches related to
signal, breakpoint, singlestep handling code.
Signed-off-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com>
Cc: Jim Keniston <jkenisto@linux.vnet.ibm.com>
Cc: Linux-mm <linux-mm@kvack.org>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Arnaldo Carvalho de Melo <acme@infradead.org>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/20120313180011.29771.89027.sendpatchset@srdronam.in.ibm.com
[ Performed various cleanliness edits ]
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2012-03-13 18:00:11 +00:00
|
|
|
#ifdef CONFIG_UPROBES
|
2017-02-06 21:06:35 +00:00
|
|
|
struct uprobe_task *utask;
|
uprobes/core: Handle breakpoint and singlestep exceptions
Uprobes uses exception notifiers to get to know if a thread hit
a breakpoint or a singlestep exception.
When a thread hits a uprobe or is singlestepping post a uprobe
hit, the uprobe exception notifier sets its TIF_UPROBE bit,
which will then be checked on its return to userspace path
(do_notify_resume() ->uprobe_notify_resume()), where the
consumers handlers are run (in task context) based on the
defined filters.
Uprobe hits are thread specific and hence we need to maintain
information about if a task hit a uprobe, what uprobe was hit,
the slot where the original instruction was copied for xol so
that it can be singlestepped with appropriate fixups.
In some cases, special care is needed for instructions that are
executed out of line (xol). These are architecture specific
artefacts, such as handling RIP relative instructions on x86_64.
Since the instruction at which the uprobe was inserted is
executed out of line, architecture specific fixups are added so
that the thread continues normal execution in the presence of a
uprobe.
Postpone the signals until we execute the probed insn.
post_xol() path does a recalc_sigpending() before return to
user-mode, this ensures the signal can't be lost.
Uprobes relies on DIE_DEBUG notification to notify if a
singlestep is complete.
Adds x86 specific uprobe exception notifiers and appropriate
hooks needed to determine a uprobe hit and subsequent post
processing.
Add requisite x86 fixups for xol for uprobes. Specific cases
needing fixups include relative jumps (x86_64), calls, etc.
Where possible, we check and skip singlestepping the
breakpointed instructions. For now we skip single byte as well
as few multibyte nop instructions. However this can be extended
to other instructions too.
Credits to Oleg Nesterov for suggestions/patches related to
signal, breakpoint, singlestep handling code.
Signed-off-by: Srikar Dronamraju <srikar@linux.vnet.ibm.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Ananth N Mavinakayanahalli <ananth@in.ibm.com>
Cc: Jim Keniston <jkenisto@linux.vnet.ibm.com>
Cc: Linux-mm <linux-mm@kvack.org>
Cc: Oleg Nesterov <oleg@redhat.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: Christoph Hellwig <hch@infradead.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Arnaldo Carvalho de Melo <acme@infradead.org>
Cc: Masami Hiramatsu <masami.hiramatsu.pt@hitachi.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Link: http://lkml.kernel.org/r/20120313180011.29771.89027.sendpatchset@srdronam.in.ibm.com
[ Performed various cleanliness edits ]
Signed-off-by: Ingo Molnar <mingo@elte.hu>
2012-03-13 18:00:11 +00:00
|
|
|
#endif
|
2013-03-23 23:11:31 +00:00
|
|
|
#if defined(CONFIG_BCACHE) || defined(CONFIG_BCACHE_MODULE)
|
2017-02-06 21:06:35 +00:00
|
|
|
unsigned int sequential_io;
|
|
|
|
unsigned int sequential_io_avg;
|
2013-03-23 23:11:31 +00:00
|
|
|
#endif
|
2014-09-24 08:18:55 +00:00
|
|
|
#ifdef CONFIG_DEBUG_ATOMIC_SLEEP
|
2017-02-06 21:06:35 +00:00
|
|
|
unsigned long task_state_change;
|
2014-09-24 08:18:55 +00:00
|
|
|
#endif
|
2017-02-06 21:06:35 +00:00
|
|
|
int pagefault_disabled;
|
2016-03-25 21:20:33 +00:00
|
|
|
#ifdef CONFIG_MMU
|
2017-02-06 21:06:35 +00:00
|
|
|
struct task_struct *oom_reaper_list;
|
2016-03-25 21:20:33 +00:00
|
|
|
#endif
|
2016-08-11 09:35:21 +00:00
|
|
|
#ifdef CONFIG_VMAP_STACK
|
2017-02-06 21:06:35 +00:00
|
|
|
struct vm_struct *stack_vm_area;
|
2016-08-11 09:35:21 +00:00
|
|
|
#endif
|
2016-09-16 05:45:48 +00:00
|
|
|
#ifdef CONFIG_THREAD_INFO_IN_TASK
|
2017-02-06 21:06:35 +00:00
|
|
|
/* A live task holds one reference: */
|
2019-01-18 12:27:30 +00:00
|
|
|
refcount_t stack_refcount;
|
livepatch: change to a per-task consistency model
Change livepatch to use a basic per-task consistency model. This is the
foundation which will eventually enable us to patch those ~10% of
security patches which change function or data semantics. This is the
biggest remaining piece needed to make livepatch more generally useful.
This code stems from the design proposal made by Vojtech [1] in November
2014. It's a hybrid of kGraft and kpatch: it uses kGraft's per-task
consistency and syscall barrier switching combined with kpatch's stack
trace switching. There are also a number of fallback options which make
it quite flexible.
Patches are applied on a per-task basis, when the task is deemed safe to
switch over. When a patch is enabled, livepatch enters into a
transition state where tasks are converging to the patched state.
Usually this transition state can complete in a few seconds. The same
sequence occurs when a patch is disabled, except the tasks converge from
the patched state to the unpatched state.
An interrupt handler inherits the patched state of the task it
interrupts. The same is true for forked tasks: the child inherits the
patched state of the parent.
Livepatch uses several complementary approaches to determine when it's
safe to patch tasks:
1. The first and most effective approach is stack checking of sleeping
tasks. If no affected functions are on the stack of a given task,
the task is patched. In most cases this will patch most or all of
the tasks on the first try. Otherwise it'll keep trying
periodically. This option is only available if the architecture has
reliable stacks (HAVE_RELIABLE_STACKTRACE).
2. The second approach, if needed, is kernel exit switching. A
task is switched when it returns to user space from a system call, a
user space IRQ, or a signal. It's useful in the following cases:
a) Patching I/O-bound user tasks which are sleeping on an affected
function. In this case you have to send SIGSTOP and SIGCONT to
force it to exit the kernel and be patched.
b) Patching CPU-bound user tasks. If the task is highly CPU-bound
then it will get patched the next time it gets interrupted by an
IRQ.
c) In the future it could be useful for applying patches for
architectures which don't yet have HAVE_RELIABLE_STACKTRACE. In
this case you would have to signal most of the tasks on the
system. However this isn't supported yet because there's
currently no way to patch kthreads without
HAVE_RELIABLE_STACKTRACE.
3. For idle "swapper" tasks, since they don't ever exit the kernel, they
instead have a klp_update_patch_state() call in the idle loop which
allows them to be patched before the CPU enters the idle state.
(Note there's not yet such an approach for kthreads.)
All the above approaches may be skipped by setting the 'immediate' flag
in the 'klp_patch' struct, which will disable per-task consistency and
patch all tasks immediately. This can be useful if the patch doesn't
change any function or data semantics. Note that, even with this flag
set, it's possible that some tasks may still be running with an old
version of the function, until that function returns.
There's also an 'immediate' flag in the 'klp_func' struct which allows
you to specify that certain functions in the patch can be applied
without per-task consistency. This might be useful if you want to patch
a common function like schedule(), and the function change doesn't need
consistency but the rest of the patch does.
For architectures which don't have HAVE_RELIABLE_STACKTRACE, the user
must set patch->immediate which causes all tasks to be patched
immediately. This option should be used with care, only when the patch
doesn't change any function or data semantics.
In the future, architectures which don't have HAVE_RELIABLE_STACKTRACE
may be allowed to use per-task consistency if we can come up with
another way to patch kthreads.
The /sys/kernel/livepatch/<patch>/transition file shows whether a patch
is in transition. Only a single patch (the topmost patch on the stack)
can be in transition at a given time. A patch can remain in transition
indefinitely, if any of the tasks are stuck in the initial patch state.
A transition can be reversed and effectively canceled by writing the
opposite value to the /sys/kernel/livepatch/<patch>/enabled file while
the transition is in progress. Then all the tasks will attempt to
converge back to the original patch state.
[1] https://lkml.kernel.org/r/20141107140458.GA21774@suse.cz
Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com>
Acked-by: Miroslav Benes <mbenes@suse.cz>
Acked-by: Ingo Molnar <mingo@kernel.org> # for the scheduler changes
Signed-off-by: Jiri Kosina <jkosina@suse.cz>
2017-02-14 01:42:40 +00:00
|
|
|
#endif
|
|
|
|
#ifdef CONFIG_LIVEPATCH
|
|
|
|
int patch_state;
|
2017-05-03 15:50:52 +00:00
|
|
|
#endif
|
2017-03-24 11:46:33 +00:00
|
|
|
#ifdef CONFIG_SECURITY
|
|
|
|
/* Used by LSM modules for access restriction: */
|
|
|
|
void *security;
|
2016-09-16 05:45:48 +00:00
|
|
|
#endif
|
2017-04-06 05:43:33 +00:00
|
|
|
|
2018-08-16 22:16:58 +00:00
|
|
|
#ifdef CONFIG_GCC_PLUGIN_STACKLEAK
|
|
|
|
unsigned long lowest_stack;
|
2018-08-16 22:17:01 +00:00
|
|
|
unsigned long prev_lowest_stack;
|
2018-08-16 22:16:58 +00:00
|
|
|
#endif
|
|
|
|
|
2017-04-06 05:43:33 +00:00
|
|
|
/*
|
|
|
|
* New fields for task_struct should be added above here, so that
|
|
|
|
* they are included in the randomized portion of task_struct.
|
|
|
|
*/
|
|
|
|
randomized_struct_fields_end
|
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
/* CPU-specific state of this task: */
|
|
|
|
struct thread_struct thread;
|
|
|
|
|
|
|
|
/*
|
|
|
|
* WARNING: on x86, 'thread_struct' contains a variable-sized
|
|
|
|
* structure. It *MUST* be at the end of 'task_struct'.
|
|
|
|
*
|
|
|
|
* Do not put anything below here!
|
|
|
|
*/
|
2005-04-16 22:20:36 +00:00
|
|
|
};
|
|
|
|
|
2007-10-26 08:17:22 +00:00
|
|
|
static inline struct pid *task_pid(struct task_struct *task)
|
2006-10-02 09:17:09 +00:00
|
|
|
{
|
2017-09-26 18:06:43 +00:00
|
|
|
return task->thread_pid;
|
2006-10-02 09:17:09 +00:00
|
|
|
}
|
|
|
|
|
2007-10-19 06:40:06 +00:00
|
|
|
/*
|
|
|
|
* the helpers to get the task's different pids as they are seen
|
|
|
|
* from various namespaces
|
|
|
|
*
|
|
|
|
* task_xid_nr() : global id, i.e. the id seen from the init namespace;
|
2008-02-08 12:19:15 +00:00
|
|
|
* task_xid_vnr() : virtual id, i.e. the id seen from the pid namespace of
|
|
|
|
* current.
|
2007-10-19 06:40:06 +00:00
|
|
|
* task_xid_nr_ns() : id seen from the ns specified;
|
|
|
|
*
|
|
|
|
* see also pid_nr() etc in include/linux/pid.h
|
|
|
|
*/
|
2017-02-06 21:06:35 +00:00
|
|
|
pid_t __task_pid_nr_ns(struct task_struct *task, enum pid_type type, struct pid_namespace *ns);
|
2007-10-19 06:40:06 +00:00
|
|
|
|
2007-10-26 08:17:22 +00:00
|
|
|
static inline pid_t task_pid_nr(struct task_struct *tsk)
|
2007-10-19 06:40:06 +00:00
|
|
|
{
|
|
|
|
return tsk->pid;
|
|
|
|
}
|
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
static inline pid_t task_pid_nr_ns(struct task_struct *tsk, struct pid_namespace *ns)
|
pids: refactor vnr/nr_ns helpers to make them safe
Inho, the safety rules for vnr/nr_ns helpers are horrible and buggy.
task_pid_nr_ns(task) needs rcu/tasklist depending on task == current.
As for "special" pids, vnr/nr_ns helpers always need rcu. However, if
task != current, they are unsafe even under rcu lock, we can't trust
task->group_leader without the special checks.
And almost every helper has a callsite which needs a fix.
Also, it is a bit annoying that the implementations of, say,
task_pgrp_vnr() and task_pgrp_nr_ns() are not "symmetrical".
This patch introduces the new helper, __task_pid_nr_ns(), which is always
safe to use, and turns all other helpers into the trivial wrappers.
After this I'll send another patch which converts task_tgid_xxx() as well,
they're are a bit special.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Cc: Louis Rilling <Louis.Rilling@kerlabs.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Pavel Emelyanov <xemul@openvz.org>
Cc: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-02 23:58:38 +00:00
|
|
|
{
|
|
|
|
return __task_pid_nr_ns(tsk, PIDTYPE_PID, ns);
|
|
|
|
}
|
2007-10-19 06:40:06 +00:00
|
|
|
|
|
|
|
static inline pid_t task_pid_vnr(struct task_struct *tsk)
|
|
|
|
{
|
pids: refactor vnr/nr_ns helpers to make them safe
Inho, the safety rules for vnr/nr_ns helpers are horrible and buggy.
task_pid_nr_ns(task) needs rcu/tasklist depending on task == current.
As for "special" pids, vnr/nr_ns helpers always need rcu. However, if
task != current, they are unsafe even under rcu lock, we can't trust
task->group_leader without the special checks.
And almost every helper has a callsite which needs a fix.
Also, it is a bit annoying that the implementations of, say,
task_pgrp_vnr() and task_pgrp_nr_ns() are not "symmetrical".
This patch introduces the new helper, __task_pid_nr_ns(), which is always
safe to use, and turns all other helpers into the trivial wrappers.
After this I'll send another patch which converts task_tgid_xxx() as well,
they're are a bit special.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Cc: Louis Rilling <Louis.Rilling@kerlabs.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Pavel Emelyanov <xemul@openvz.org>
Cc: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-02 23:58:38 +00:00
|
|
|
return __task_pid_nr_ns(tsk, PIDTYPE_PID, NULL);
|
2007-10-19 06:40:06 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
2007-10-26 08:17:22 +00:00
|
|
|
static inline pid_t task_tgid_nr(struct task_struct *tsk)
|
2007-10-19 06:40:06 +00:00
|
|
|
{
|
|
|
|
return tsk->tgid;
|
|
|
|
}
|
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
/**
|
|
|
|
* pid_alive - check that a task structure is not stale
|
|
|
|
* @p: Task structure to be checked.
|
|
|
|
*
|
|
|
|
* Test if a process is not yet dead (at most zombie state)
|
|
|
|
* If pid_alive fails, then pointers within the task structure
|
|
|
|
* can be stale and must not be dereferenced.
|
|
|
|
*
|
|
|
|
* Return: 1 if the process is alive. 0 otherwise.
|
|
|
|
*/
|
|
|
|
static inline int pid_alive(const struct task_struct *p)
|
|
|
|
{
|
2017-09-26 18:06:43 +00:00
|
|
|
return p->thread_pid != NULL;
|
2017-02-06 21:06:35 +00:00
|
|
|
}
|
2007-10-19 06:40:06 +00:00
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
static inline pid_t task_pgrp_nr_ns(struct task_struct *tsk, struct pid_namespace *ns)
|
2007-10-19 06:40:06 +00:00
|
|
|
{
|
pids: refactor vnr/nr_ns helpers to make them safe
Inho, the safety rules for vnr/nr_ns helpers are horrible and buggy.
task_pid_nr_ns(task) needs rcu/tasklist depending on task == current.
As for "special" pids, vnr/nr_ns helpers always need rcu. However, if
task != current, they are unsafe even under rcu lock, we can't trust
task->group_leader without the special checks.
And almost every helper has a callsite which needs a fix.
Also, it is a bit annoying that the implementations of, say,
task_pgrp_vnr() and task_pgrp_nr_ns() are not "symmetrical".
This patch introduces the new helper, __task_pid_nr_ns(), which is always
safe to use, and turns all other helpers into the trivial wrappers.
After this I'll send another patch which converts task_tgid_xxx() as well,
they're are a bit special.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Cc: Louis Rilling <Louis.Rilling@kerlabs.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Pavel Emelyanov <xemul@openvz.org>
Cc: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-02 23:58:38 +00:00
|
|
|
return __task_pid_nr_ns(tsk, PIDTYPE_PGID, ns);
|
2007-10-19 06:40:06 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static inline pid_t task_pgrp_vnr(struct task_struct *tsk)
|
|
|
|
{
|
pids: refactor vnr/nr_ns helpers to make them safe
Inho, the safety rules for vnr/nr_ns helpers are horrible and buggy.
task_pid_nr_ns(task) needs rcu/tasklist depending on task == current.
As for "special" pids, vnr/nr_ns helpers always need rcu. However, if
task != current, they are unsafe even under rcu lock, we can't trust
task->group_leader without the special checks.
And almost every helper has a callsite which needs a fix.
Also, it is a bit annoying that the implementations of, say,
task_pgrp_vnr() and task_pgrp_nr_ns() are not "symmetrical".
This patch introduces the new helper, __task_pid_nr_ns(), which is always
safe to use, and turns all other helpers into the trivial wrappers.
After this I'll send another patch which converts task_tgid_xxx() as well,
they're are a bit special.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Cc: Louis Rilling <Louis.Rilling@kerlabs.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Pavel Emelyanov <xemul@openvz.org>
Cc: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-02 23:58:38 +00:00
|
|
|
return __task_pid_nr_ns(tsk, PIDTYPE_PGID, NULL);
|
2007-10-19 06:40:06 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
static inline pid_t task_session_nr_ns(struct task_struct *tsk, struct pid_namespace *ns)
|
2007-10-19 06:40:06 +00:00
|
|
|
{
|
pids: refactor vnr/nr_ns helpers to make them safe
Inho, the safety rules for vnr/nr_ns helpers are horrible and buggy.
task_pid_nr_ns(task) needs rcu/tasklist depending on task == current.
As for "special" pids, vnr/nr_ns helpers always need rcu. However, if
task != current, they are unsafe even under rcu lock, we can't trust
task->group_leader without the special checks.
And almost every helper has a callsite which needs a fix.
Also, it is a bit annoying that the implementations of, say,
task_pgrp_vnr() and task_pgrp_nr_ns() are not "symmetrical".
This patch introduces the new helper, __task_pid_nr_ns(), which is always
safe to use, and turns all other helpers into the trivial wrappers.
After this I'll send another patch which converts task_tgid_xxx() as well,
they're are a bit special.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Cc: Louis Rilling <Louis.Rilling@kerlabs.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Pavel Emelyanov <xemul@openvz.org>
Cc: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-02 23:58:38 +00:00
|
|
|
return __task_pid_nr_ns(tsk, PIDTYPE_SID, ns);
|
2007-10-19 06:40:06 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static inline pid_t task_session_vnr(struct task_struct *tsk)
|
|
|
|
{
|
pids: refactor vnr/nr_ns helpers to make them safe
Inho, the safety rules for vnr/nr_ns helpers are horrible and buggy.
task_pid_nr_ns(task) needs rcu/tasklist depending on task == current.
As for "special" pids, vnr/nr_ns helpers always need rcu. However, if
task != current, they are unsafe even under rcu lock, we can't trust
task->group_leader without the special checks.
And almost every helper has a callsite which needs a fix.
Also, it is a bit annoying that the implementations of, say,
task_pgrp_vnr() and task_pgrp_nr_ns() are not "symmetrical".
This patch introduces the new helper, __task_pid_nr_ns(), which is always
safe to use, and turns all other helpers into the trivial wrappers.
After this I'll send another patch which converts task_tgid_xxx() as well,
they're are a bit special.
Signed-off-by: Oleg Nesterov <oleg@redhat.com>
Cc: Louis Rilling <Louis.Rilling@kerlabs.com>
Cc: "Eric W. Biederman" <ebiederm@xmission.com>
Cc: Pavel Emelyanov <xemul@openvz.org>
Cc: Sukadev Bhattiprolu <sukadev@linux.vnet.ibm.com>
Cc: Roland McGrath <roland@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2009-04-02 23:58:38 +00:00
|
|
|
return __task_pid_nr_ns(tsk, PIDTYPE_SID, NULL);
|
2007-10-19 06:40:06 +00:00
|
|
|
}
|
|
|
|
|
2017-08-21 15:35:02 +00:00
|
|
|
static inline pid_t task_tgid_nr_ns(struct task_struct *tsk, struct pid_namespace *ns)
|
|
|
|
{
|
2017-06-04 09:32:13 +00:00
|
|
|
return __task_pid_nr_ns(tsk, PIDTYPE_TGID, ns);
|
2017-08-21 15:35:02 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static inline pid_t task_tgid_vnr(struct task_struct *tsk)
|
|
|
|
{
|
2017-06-04 09:32:13 +00:00
|
|
|
return __task_pid_nr_ns(tsk, PIDTYPE_TGID, NULL);
|
2017-08-21 15:35:02 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static inline pid_t task_ppid_nr_ns(const struct task_struct *tsk, struct pid_namespace *ns)
|
|
|
|
{
|
|
|
|
pid_t pid = 0;
|
|
|
|
|
|
|
|
rcu_read_lock();
|
|
|
|
if (pid_alive(tsk))
|
|
|
|
pid = task_tgid_nr_ns(rcu_dereference(tsk->real_parent), ns);
|
|
|
|
rcu_read_unlock();
|
|
|
|
|
|
|
|
return pid;
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline pid_t task_ppid_nr(const struct task_struct *tsk)
|
|
|
|
{
|
|
|
|
return task_ppid_nr_ns(tsk, &init_pid_ns);
|
|
|
|
}
|
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
/* Obsolete, do not use: */
|
2009-04-02 23:58:39 +00:00
|
|
|
static inline pid_t task_pgrp_nr(struct task_struct *tsk)
|
|
|
|
{
|
|
|
|
return task_pgrp_nr_ns(tsk, &init_pid_ns);
|
|
|
|
}
|
2007-10-19 06:40:06 +00:00
|
|
|
|
2017-09-22 16:30:40 +00:00
|
|
|
#define TASK_REPORT_IDLE (TASK_REPORT + 1)
|
|
|
|
#define TASK_REPORT_MAX (TASK_REPORT_IDLE << 1)
|
|
|
|
|
2017-09-29 11:50:16 +00:00
|
|
|
static inline unsigned int task_state_index(struct task_struct *tsk)
|
2017-08-07 08:44:23 +00:00
|
|
|
{
|
2017-09-22 16:09:26 +00:00
|
|
|
unsigned int tsk_state = READ_ONCE(tsk->state);
|
|
|
|
unsigned int state = (tsk_state | tsk->exit_state) & TASK_REPORT;
|
2017-08-07 08:44:23 +00:00
|
|
|
|
2017-09-22 16:30:40 +00:00
|
|
|
BUILD_BUG_ON_NOT_POWER_OF_2(TASK_REPORT_MAX);
|
|
|
|
|
|
|
|
if (tsk_state == TASK_IDLE)
|
|
|
|
state = TASK_REPORT_IDLE;
|
|
|
|
|
2017-09-22 16:09:26 +00:00
|
|
|
return fls(state);
|
|
|
|
}
|
|
|
|
|
2017-09-29 11:50:16 +00:00
|
|
|
static inline char task_index_to_char(unsigned int state)
|
2017-09-22 16:09:26 +00:00
|
|
|
{
|
2017-09-22 16:37:28 +00:00
|
|
|
static const char state_char[] = "RSDTtXZPI";
|
2017-09-22 16:09:26 +00:00
|
|
|
|
2017-09-22 16:30:40 +00:00
|
|
|
BUILD_BUG_ON(1 + ilog2(TASK_REPORT_MAX) != sizeof(state_char) - 1);
|
2017-08-07 08:44:23 +00:00
|
|
|
|
2017-09-22 16:09:26 +00:00
|
|
|
return state_char[state];
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline char task_state_to_char(struct task_struct *tsk)
|
|
|
|
{
|
2017-09-29 11:50:16 +00:00
|
|
|
return task_index_to_char(task_state_index(tsk));
|
2017-08-07 08:44:23 +00:00
|
|
|
}
|
|
|
|
|
2006-09-29 09:00:07 +00:00
|
|
|
/**
|
2016-01-01 14:03:01 +00:00
|
|
|
* is_global_init - check if a task structure is init. Since init
|
|
|
|
* is free to have sub-threads we need to check tgid.
|
2006-10-06 07:44:01 +00:00
|
|
|
* @tsk: Task structure to be checked.
|
|
|
|
*
|
|
|
|
* Check if a task structure is the first user space task the kernel created.
|
2013-07-12 18:45:47 +00:00
|
|
|
*
|
|
|
|
* Return: 1 if the task structure is init. 0 otherwise.
|
2007-10-19 06:39:52 +00:00
|
|
|
*/
|
2007-10-26 08:17:22 +00:00
|
|
|
static inline int is_global_init(struct task_struct *tsk)
|
2007-10-19 06:40:09 +00:00
|
|
|
{
|
2016-01-01 14:03:01 +00:00
|
|
|
return task_tgid_nr(tsk) == 1;
|
2007-10-19 06:40:09 +00:00
|
|
|
}
|
2007-10-19 06:39:52 +00:00
|
|
|
|
2006-10-02 09:19:00 +00:00
|
|
|
extern struct pid *cad_pid;
|
|
|
|
|
2005-04-16 22:20:36 +00:00
|
|
|
/*
|
|
|
|
* Per process flags
|
|
|
|
*/
|
2017-02-06 21:06:35 +00:00
|
|
|
#define PF_IDLE 0x00000002 /* I am an IDLE thread */
|
|
|
|
#define PF_EXITING 0x00000004 /* Getting shut down */
|
|
|
|
#define PF_EXITPIDONE 0x00000008 /* PI exit done on shut down */
|
|
|
|
#define PF_VCPU 0x00000010 /* I'm a virtual CPU */
|
|
|
|
#define PF_WQ_WORKER 0x00000020 /* I'm a workqueue worker */
|
|
|
|
#define PF_FORKNOEXEC 0x00000040 /* Forked but didn't exec */
|
|
|
|
#define PF_MCE_PROCESS 0x00000080 /* Process policy on mce errors */
|
|
|
|
#define PF_SUPERPRIV 0x00000100 /* Used super-user privileges */
|
|
|
|
#define PF_DUMPCORE 0x00000200 /* Dumped core */
|
|
|
|
#define PF_SIGNALED 0x00000400 /* Killed by a signal */
|
|
|
|
#define PF_MEMALLOC 0x00000800 /* Allocating memory */
|
|
|
|
#define PF_NPROC_EXCEEDED 0x00001000 /* set_user() noticed that RLIMIT_NPROC was exceeded */
|
|
|
|
#define PF_USED_MATH 0x00002000 /* If unset the fpu must be initialized before use */
|
|
|
|
#define PF_USED_ASYNC 0x00004000 /* Used async_schedule*(), used by module init */
|
|
|
|
#define PF_NOFREEZE 0x00008000 /* This thread should not be frozen */
|
|
|
|
#define PF_FROZEN 0x00010000 /* Frozen for system suspend */
|
mm: introduce memalloc_nofs_{save,restore} API
GFP_NOFS context is used for the following 5 reasons currently:
- to prevent from deadlocks when the lock held by the allocation
context would be needed during the memory reclaim
- to prevent from stack overflows during the reclaim because the
allocation is performed from a deep context already
- to prevent lockups when the allocation context depends on other
reclaimers to make a forward progress indirectly
- just in case because this would be safe from the fs POV
- silence lockdep false positives
Unfortunately overuse of this allocation context brings some problems to
the MM. Memory reclaim is much weaker (especially during heavy FS
metadata workloads), OOM killer cannot be invoked because the MM layer
doesn't have enough information about how much memory is freeable by the
FS layer.
In many cases it is far from clear why the weaker context is even used
and so it might be used unnecessarily. We would like to get rid of
those as much as possible. One way to do that is to use the flag in
scopes rather than isolated cases. Such a scope is declared when really
necessary, tracked per task and all the allocation requests from within
the context will simply inherit the GFP_NOFS semantic.
Not only this is easier to understand and maintain because there are
much less problematic contexts than specific allocation requests, this
also helps code paths where FS layer interacts with other layers (e.g.
crypto, security modules, MM etc...) and there is no easy way to convey
the allocation context between the layers.
Introduce memalloc_nofs_{save,restore} API to control the scope of
GFP_NOFS allocation context. This is basically copying
memalloc_noio_{save,restore} API we have for other restricted allocation
context GFP_NOIO. The PF_MEMALLOC_NOFS flag already exists and it is
just an alias for PF_FSTRANS which has been xfs specific until recently.
There are no more PF_FSTRANS users anymore so let's just drop it.
PF_MEMALLOC_NOFS is now checked in the MM layer and drops __GFP_FS
implicitly same as PF_MEMALLOC_NOIO drops __GFP_IO. memalloc_noio_flags
is renamed to current_gfp_context because it now cares about both
PF_MEMALLOC_NOFS and PF_MEMALLOC_NOIO contexts. Xfs code paths preserve
their semantic. kmem_flags_convert() doesn't need to evaluate the flag
anymore.
This patch shouldn't introduce any functional changes.
Let's hope that filesystems will drop direct GFP_NOFS (resp. ~__GFP_FS)
usage as much as possible and only use a properly documented
memalloc_nofs_{save,restore} checkpoints where they are appropriate.
[akpm@linux-foundation.org: fix comment typo, reflow comment]
Link: http://lkml.kernel.org/r/20170306131408.9828-5-mhocko@kernel.org
Signed-off-by: Michal Hocko <mhocko@suse.com>
Acked-by: Vlastimil Babka <vbabka@suse.cz>
Cc: Dave Chinner <david@fromorbit.com>
Cc: Theodore Ts'o <tytso@mit.edu>
Cc: Chris Mason <clm@fb.com>
Cc: David Sterba <dsterba@suse.cz>
Cc: Jan Kara <jack@suse.cz>
Cc: Brian Foster <bfoster@redhat.com>
Cc: Darrick J. Wong <darrick.wong@oracle.com>
Cc: Nikolay Borisov <nborisov@suse.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2017-05-03 21:53:15 +00:00
|
|
|
#define PF_KSWAPD 0x00020000 /* I am kswapd */
|
|
|
|
#define PF_MEMALLOC_NOFS 0x00040000 /* All allocation requests will inherit GFP_NOFS */
|
|
|
|
#define PF_MEMALLOC_NOIO 0x00080000 /* All allocation requests will inherit GFP_NOIO */
|
2017-02-06 21:06:35 +00:00
|
|
|
#define PF_LESS_THROTTLE 0x00100000 /* Throttle me less: I clean memory */
|
|
|
|
#define PF_KTHREAD 0x00200000 /* I am a kernel thread */
|
|
|
|
#define PF_RANDOMIZE 0x00400000 /* Randomize virtual address space */
|
|
|
|
#define PF_SWAPWRITE 0x00800000 /* Allowed to write to swap */
|
psi: pressure stall information for CPU, memory, and IO
When systems are overcommitted and resources become contended, it's hard
to tell exactly the impact this has on workload productivity, or how close
the system is to lockups and OOM kills. In particular, when machines work
multiple jobs concurrently, the impact of overcommit in terms of latency
and throughput on the individual job can be enormous.
In order to maximize hardware utilization without sacrificing individual
job health or risk complete machine lockups, this patch implements a way
to quantify resource pressure in the system.
A kernel built with CONFIG_PSI=y creates files in /proc/pressure/ that
expose the percentage of time the system is stalled on CPU, memory, or IO,
respectively. Stall states are aggregate versions of the per-task delay
accounting delays:
cpu: some tasks are runnable but not executing on a CPU
memory: tasks are reclaiming, or waiting for swapin or thrashing cache
io: tasks are waiting for io completions
These percentages of walltime can be thought of as pressure percentages,
and they give a general sense of system health and productivity loss
incurred by resource overcommit. They can also indicate when the system
is approaching lockup scenarios and OOMs.
To do this, psi keeps track of the task states associated with each CPU
and samples the time they spend in stall states. Every 2 seconds, the
samples are averaged across CPUs - weighted by the CPUs' non-idle time to
eliminate artifacts from unused CPUs - and translated into percentages of
walltime. A running average of those percentages is maintained over 10s,
1m, and 5m periods (similar to the loadaverage).
[hannes@cmpxchg.org: doc fixlet, per Randy]
Link: http://lkml.kernel.org/r/20180828205625.GA14030@cmpxchg.org
[hannes@cmpxchg.org: code optimization]
Link: http://lkml.kernel.org/r/20180907175015.GA8479@cmpxchg.org
[hannes@cmpxchg.org: rename psi_clock() to psi_update_work(), per Peter]
Link: http://lkml.kernel.org/r/20180907145404.GB11088@cmpxchg.org
[hannes@cmpxchg.org: fix build]
Link: http://lkml.kernel.org/r/20180913014222.GA2370@cmpxchg.org
Link: http://lkml.kernel.org/r/20180828172258.3185-9-hannes@cmpxchg.org
Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Tested-by: Daniel Drake <drake@endlessm.com>
Tested-by: Suren Baghdasaryan <surenb@google.com>
Cc: Christopher Lameter <cl@linux.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Johannes Weiner <jweiner@fb.com>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Peter Enderborg <peter.enderborg@sony.com>
Cc: Randy Dunlap <rdunlap@infradead.org>
Cc: Shakeel Butt <shakeelb@google.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: Vinayak Menon <vinmenon@codeaurora.org>
Cc: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2018-10-26 22:06:27 +00:00
|
|
|
#define PF_MEMSTALL 0x01000000 /* Stalled due to lack of memory */
|
2019-01-08 17:23:56 +00:00
|
|
|
#define PF_UMH 0x02000000 /* I'm an Usermodehelper process */
|
2019-04-23 14:26:36 +00:00
|
|
|
#define PF_NO_SETAFFINITY 0x04000000 /* Userland is not allowed to meddle with cpus_mask */
|
2017-02-06 21:06:35 +00:00
|
|
|
#define PF_MCE_EARLY 0x08000000 /* Early kill for mce process policy */
|
2019-03-05 23:47:40 +00:00
|
|
|
#define PF_MEMALLOC_NOCMA 0x10000000 /* All allocation request will have _GFP_MOVABLE cleared */
|
2017-02-06 21:06:35 +00:00
|
|
|
#define PF_FREEZER_SKIP 0x40000000 /* Freezer should not count it as freezable */
|
|
|
|
#define PF_SUSPEND_TASK 0x80000000 /* This thread called freeze_processes() and should not be frozen */
|
2005-04-16 22:20:36 +00:00
|
|
|
|
|
|
|
/*
|
|
|
|
* Only the _current_ task can read/write to tsk->flags, but other
|
|
|
|
* tasks can access tsk->flags in readonly mode for example
|
|
|
|
* with tsk_used_math (like during threaded core dumping).
|
|
|
|
* There is however an exception to this rule during ptrace
|
|
|
|
* or during fork: the ptracer task is allowed to write to the
|
|
|
|
* child->flags of its traced child (same goes for fork, the parent
|
|
|
|
* can write to the child->flags), because we're guaranteed the
|
|
|
|
* child is not running and in turn not changing child->flags
|
|
|
|
* at the same time the parent does it.
|
|
|
|
*/
|
2017-02-06 21:06:35 +00:00
|
|
|
#define clear_stopped_child_used_math(child) do { (child)->flags &= ~PF_USED_MATH; } while (0)
|
|
|
|
#define set_stopped_child_used_math(child) do { (child)->flags |= PF_USED_MATH; } while (0)
|
|
|
|
#define clear_used_math() clear_stopped_child_used_math(current)
|
|
|
|
#define set_used_math() set_stopped_child_used_math(current)
|
|
|
|
|
2005-04-16 22:20:36 +00:00
|
|
|
#define conditional_stopped_child_used_math(condition, child) \
|
|
|
|
do { (child)->flags &= ~PF_USED_MATH, (child)->flags |= (condition) ? PF_USED_MATH : 0; } while (0)
|
2017-02-06 21:06:35 +00:00
|
|
|
|
|
|
|
#define conditional_used_math(condition) conditional_stopped_child_used_math(condition, current)
|
|
|
|
|
2005-04-16 22:20:36 +00:00
|
|
|
#define copy_to_stopped_child_used_math(child) \
|
|
|
|
do { (child)->flags &= ~PF_USED_MATH, (child)->flags |= current->flags & PF_USED_MATH; } while (0)
|
2017-02-06 21:06:35 +00:00
|
|
|
|
2005-04-16 22:20:36 +00:00
|
|
|
/* NOTE: this will return 0 or PF_USED_MATH, it will never return 1 */
|
2017-02-06 21:06:35 +00:00
|
|
|
#define tsk_used_math(p) ((p)->flags & PF_USED_MATH)
|
|
|
|
#define used_math() tsk_used_math(current)
|
2005-04-16 22:20:36 +00:00
|
|
|
|
2017-05-24 08:15:41 +00:00
|
|
|
static inline bool is_percpu_thread(void)
|
|
|
|
{
|
|
|
|
#ifdef CONFIG_SMP
|
|
|
|
return (current->flags & PF_NO_SETAFFINITY) &&
|
|
|
|
(current->nr_cpus_allowed == 1);
|
|
|
|
#else
|
|
|
|
return true;
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
2014-05-21 22:23:46 +00:00
|
|
|
/* Per-process atomic flags. */
|
2017-02-06 21:06:35 +00:00
|
|
|
#define PFA_NO_NEW_PRIVS 0 /* May not gain new privileges. */
|
|
|
|
#define PFA_SPREAD_PAGE 1 /* Spread page cache over cpuset */
|
|
|
|
#define PFA_SPREAD_SLAB 2 /* Spread some slab caches over cpuset */
|
2018-05-03 20:09:15 +00:00
|
|
|
#define PFA_SPEC_SSB_DISABLE 3 /* Speculative Store Bypass disabled */
|
|
|
|
#define PFA_SPEC_SSB_FORCE_DISABLE 4 /* Speculative Store Bypass force disabled*/
|
x86/speculation: Add prctl() control for indirect branch speculation
Add the PR_SPEC_INDIRECT_BRANCH option for the PR_GET_SPECULATION_CTRL and
PR_SET_SPECULATION_CTRL prctls to allow fine grained per task control of
indirect branch speculation via STIBP and IBPB.
Invocations:
Check indirect branch speculation status with
- prctl(PR_GET_SPECULATION_CTRL, PR_SPEC_INDIRECT_BRANCH, 0, 0, 0);
Enable indirect branch speculation with
- prctl(PR_SET_SPECULATION_CTRL, PR_SPEC_INDIRECT_BRANCH, PR_SPEC_ENABLE, 0, 0);
Disable indirect branch speculation with
- prctl(PR_SET_SPECULATION_CTRL, PR_SPEC_INDIRECT_BRANCH, PR_SPEC_DISABLE, 0, 0);
Force disable indirect branch speculation with
- prctl(PR_SET_SPECULATION_CTRL, PR_SPEC_INDIRECT_BRANCH, PR_SPEC_FORCE_DISABLE, 0, 0);
See Documentation/userspace-api/spec_ctrl.rst.
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Ingo Molnar <mingo@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Jiri Kosina <jkosina@suse.cz>
Cc: Tom Lendacky <thomas.lendacky@amd.com>
Cc: Josh Poimboeuf <jpoimboe@redhat.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: David Woodhouse <dwmw@amazon.co.uk>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Casey Schaufler <casey.schaufler@intel.com>
Cc: Asit Mallick <asit.k.mallick@intel.com>
Cc: Arjan van de Ven <arjan@linux.intel.com>
Cc: Jon Masters <jcm@redhat.com>
Cc: Waiman Long <longman9394@gmail.com>
Cc: Greg KH <gregkh@linuxfoundation.org>
Cc: Dave Stewart <david.c.stewart@intel.com>
Cc: Kees Cook <keescook@chromium.org>
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/20181125185005.866780996@linutronix.de
2018-11-25 18:33:53 +00:00
|
|
|
#define PFA_SPEC_IB_DISABLE 5 /* Indirect branch speculation restricted */
|
|
|
|
#define PFA_SPEC_IB_FORCE_DISABLE 6 /* Indirect branch speculation permanently restricted */
|
2019-01-16 22:01:36 +00:00
|
|
|
#define PFA_SPEC_SSB_NOEXEC 7 /* Speculative Store Bypass clear on execve() */
|
2014-05-21 22:23:46 +00:00
|
|
|
|
2014-09-25 01:40:40 +00:00
|
|
|
#define TASK_PFA_TEST(name, func) \
|
|
|
|
static inline bool task_##func(struct task_struct *p) \
|
|
|
|
{ return test_bit(PFA_##name, &p->atomic_flags); }
|
2017-02-06 21:06:35 +00:00
|
|
|
|
2014-09-25 01:40:40 +00:00
|
|
|
#define TASK_PFA_SET(name, func) \
|
|
|
|
static inline void task_set_##func(struct task_struct *p) \
|
|
|
|
{ set_bit(PFA_##name, &p->atomic_flags); }
|
2017-02-06 21:06:35 +00:00
|
|
|
|
2014-09-25 01:40:40 +00:00
|
|
|
#define TASK_PFA_CLEAR(name, func) \
|
|
|
|
static inline void task_clear_##func(struct task_struct *p) \
|
|
|
|
{ clear_bit(PFA_##name, &p->atomic_flags); }
|
|
|
|
|
|
|
|
TASK_PFA_TEST(NO_NEW_PRIVS, no_new_privs)
|
|
|
|
TASK_PFA_SET(NO_NEW_PRIVS, no_new_privs)
|
2014-05-21 22:23:46 +00:00
|
|
|
|
2014-09-25 01:41:02 +00:00
|
|
|
TASK_PFA_TEST(SPREAD_PAGE, spread_page)
|
|
|
|
TASK_PFA_SET(SPREAD_PAGE, spread_page)
|
|
|
|
TASK_PFA_CLEAR(SPREAD_PAGE, spread_page)
|
|
|
|
|
|
|
|
TASK_PFA_TEST(SPREAD_SLAB, spread_slab)
|
|
|
|
TASK_PFA_SET(SPREAD_SLAB, spread_slab)
|
|
|
|
TASK_PFA_CLEAR(SPREAD_SLAB, spread_slab)
|
2014-05-21 22:23:46 +00:00
|
|
|
|
2018-05-03 20:09:15 +00:00
|
|
|
TASK_PFA_TEST(SPEC_SSB_DISABLE, spec_ssb_disable)
|
|
|
|
TASK_PFA_SET(SPEC_SSB_DISABLE, spec_ssb_disable)
|
|
|
|
TASK_PFA_CLEAR(SPEC_SSB_DISABLE, spec_ssb_disable)
|
|
|
|
|
2019-01-16 22:01:36 +00:00
|
|
|
TASK_PFA_TEST(SPEC_SSB_NOEXEC, spec_ssb_noexec)
|
|
|
|
TASK_PFA_SET(SPEC_SSB_NOEXEC, spec_ssb_noexec)
|
|
|
|
TASK_PFA_CLEAR(SPEC_SSB_NOEXEC, spec_ssb_noexec)
|
|
|
|
|
2018-05-03 20:09:15 +00:00
|
|
|
TASK_PFA_TEST(SPEC_SSB_FORCE_DISABLE, spec_ssb_force_disable)
|
|
|
|
TASK_PFA_SET(SPEC_SSB_FORCE_DISABLE, spec_ssb_force_disable)
|
|
|
|
|
x86/speculation: Add prctl() control for indirect branch speculation
Add the PR_SPEC_INDIRECT_BRANCH option for the PR_GET_SPECULATION_CTRL and
PR_SET_SPECULATION_CTRL prctls to allow fine grained per task control of
indirect branch speculation via STIBP and IBPB.
Invocations:
Check indirect branch speculation status with
- prctl(PR_GET_SPECULATION_CTRL, PR_SPEC_INDIRECT_BRANCH, 0, 0, 0);
Enable indirect branch speculation with
- prctl(PR_SET_SPECULATION_CTRL, PR_SPEC_INDIRECT_BRANCH, PR_SPEC_ENABLE, 0, 0);
Disable indirect branch speculation with
- prctl(PR_SET_SPECULATION_CTRL, PR_SPEC_INDIRECT_BRANCH, PR_SPEC_DISABLE, 0, 0);
Force disable indirect branch speculation with
- prctl(PR_SET_SPECULATION_CTRL, PR_SPEC_INDIRECT_BRANCH, PR_SPEC_FORCE_DISABLE, 0, 0);
See Documentation/userspace-api/spec_ctrl.rst.
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Ingo Molnar <mingo@kernel.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Andy Lutomirski <luto@kernel.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Jiri Kosina <jkosina@suse.cz>
Cc: Tom Lendacky <thomas.lendacky@amd.com>
Cc: Josh Poimboeuf <jpoimboe@redhat.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: David Woodhouse <dwmw@amazon.co.uk>
Cc: Andi Kleen <ak@linux.intel.com>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Casey Schaufler <casey.schaufler@intel.com>
Cc: Asit Mallick <asit.k.mallick@intel.com>
Cc: Arjan van de Ven <arjan@linux.intel.com>
Cc: Jon Masters <jcm@redhat.com>
Cc: Waiman Long <longman9394@gmail.com>
Cc: Greg KH <gregkh@linuxfoundation.org>
Cc: Dave Stewart <david.c.stewart@intel.com>
Cc: Kees Cook <keescook@chromium.org>
Cc: stable@vger.kernel.org
Link: https://lkml.kernel.org/r/20181125185005.866780996@linutronix.de
2018-11-25 18:33:53 +00:00
|
|
|
TASK_PFA_TEST(SPEC_IB_DISABLE, spec_ib_disable)
|
|
|
|
TASK_PFA_SET(SPEC_IB_DISABLE, spec_ib_disable)
|
|
|
|
TASK_PFA_CLEAR(SPEC_IB_DISABLE, spec_ib_disable)
|
|
|
|
|
|
|
|
TASK_PFA_TEST(SPEC_IB_FORCE_DISABLE, spec_ib_force_disable)
|
|
|
|
TASK_PFA_SET(SPEC_IB_FORCE_DISABLE, spec_ib_force_disable)
|
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
static inline void
|
2017-04-07 00:03:26 +00:00
|
|
|
current_restore_flags(unsigned long orig_flags, unsigned long flags)
|
2012-07-31 23:44:07 +00:00
|
|
|
{
|
2017-04-07 00:03:26 +00:00
|
|
|
current->flags &= ~flags;
|
|
|
|
current->flags |= orig_flags & flags;
|
2012-07-31 23:44:07 +00:00
|
|
|
}
|
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
extern int cpuset_cpumask_can_shrink(const struct cpumask *cur, const struct cpumask *trial);
|
|
|
|
extern int task_can_attach(struct task_struct *p, const struct cpumask *cs_cpus_allowed);
|
2005-04-16 22:20:36 +00:00
|
|
|
#ifdef CONFIG_SMP
|
2017-02-06 21:06:35 +00:00
|
|
|
extern void do_set_cpus_allowed(struct task_struct *p, const struct cpumask *new_mask);
|
|
|
|
extern int set_cpus_allowed_ptr(struct task_struct *p, const struct cpumask *new_mask);
|
2005-04-16 22:20:36 +00:00
|
|
|
#else
|
2017-02-06 21:06:35 +00:00
|
|
|
static inline void do_set_cpus_allowed(struct task_struct *p, const struct cpumask *new_mask)
|
2011-05-19 06:08:58 +00:00
|
|
|
{
|
|
|
|
}
|
2017-02-06 21:06:35 +00:00
|
|
|
static inline int set_cpus_allowed_ptr(struct task_struct *p, const struct cpumask *new_mask)
|
2005-04-16 22:20:36 +00:00
|
|
|
{
|
2008-11-24 16:05:14 +00:00
|
|
|
if (!cpumask_test_cpu(0, new_mask))
|
2005-04-16 22:20:36 +00:00
|
|
|
return -EINVAL;
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
#endif
|
2009-09-24 15:34:38 +00:00
|
|
|
|
2014-05-23 10:20:42 +00:00
|
|
|
extern int yield_to(struct task_struct *p, bool preempt);
|
2006-07-03 07:25:41 +00:00
|
|
|
extern void set_user_nice(struct task_struct *p, long nice);
|
|
|
|
extern int task_prio(const struct task_struct *p);
|
2017-02-06 21:06:35 +00:00
|
|
|
|
2014-01-28 03:00:45 +00:00
|
|
|
/**
|
|
|
|
* task_nice - return the nice value of a given task.
|
|
|
|
* @p: the task in question.
|
|
|
|
*
|
|
|
|
* Return: The nice value [ -20 ... 0 ... 19 ].
|
|
|
|
*/
|
|
|
|
static inline int task_nice(const struct task_struct *p)
|
|
|
|
{
|
|
|
|
return PRIO_TO_NICE((p)->static_prio);
|
|
|
|
}
|
2017-02-06 21:06:35 +00:00
|
|
|
|
2006-07-03 07:25:41 +00:00
|
|
|
extern int can_nice(const struct task_struct *p, const int nice);
|
|
|
|
extern int task_curr(const struct task_struct *p);
|
2005-04-16 22:20:36 +00:00
|
|
|
extern int idle_cpu(int cpu);
|
2018-05-09 16:39:48 +00:00
|
|
|
extern int available_idle_cpu(int cpu);
|
2017-02-06 21:06:35 +00:00
|
|
|
extern int sched_setscheduler(struct task_struct *, int, const struct sched_param *);
|
|
|
|
extern int sched_setscheduler_nocheck(struct task_struct *, int, const struct sched_param *);
|
|
|
|
extern int sched_setattr(struct task_struct *, const struct sched_attr *);
|
2017-12-04 10:23:20 +00:00
|
|
|
extern int sched_setattr_nocheck(struct task_struct *, const struct sched_attr *);
|
2006-07-03 07:25:41 +00:00
|
|
|
extern struct task_struct *idle_task(int cpu);
|
2017-02-06 21:06:35 +00:00
|
|
|
|
2011-11-10 20:41:56 +00:00
|
|
|
/**
|
|
|
|
* is_idle_task - is the specified task an idle task?
|
2012-01-21 19:03:13 +00:00
|
|
|
* @p: the task in question.
|
2013-07-12 18:45:47 +00:00
|
|
|
*
|
|
|
|
* Return: 1 if @p is an idle task. 0 otherwise.
|
2011-11-10 20:41:56 +00:00
|
|
|
*/
|
2011-12-20 16:20:46 +00:00
|
|
|
static inline bool is_idle_task(const struct task_struct *p)
|
2011-11-10 20:41:56 +00:00
|
|
|
{
|
2016-11-29 07:03:05 +00:00
|
|
|
return !!(p->flags & PF_IDLE);
|
2011-11-10 20:41:56 +00:00
|
|
|
}
|
2017-02-06 21:06:35 +00:00
|
|
|
|
2006-07-03 07:25:41 +00:00
|
|
|
extern struct task_struct *curr_task(int cpu);
|
2016-09-20 18:29:40 +00:00
|
|
|
extern void ia64_set_curr_task(int cpu, struct task_struct *p);
|
2005-04-16 22:20:36 +00:00
|
|
|
|
|
|
|
void yield(void);
|
|
|
|
|
|
|
|
union thread_union {
|
2018-01-02 15:12:01 +00:00
|
|
|
#ifndef CONFIG_ARCH_TASK_STRUCT_ON_STACK
|
|
|
|
struct task_struct task;
|
|
|
|
#endif
|
2016-09-13 21:29:24 +00:00
|
|
|
#ifndef CONFIG_THREAD_INFO_IN_TASK
|
2005-04-16 22:20:36 +00:00
|
|
|
struct thread_info thread_info;
|
2016-09-13 21:29:24 +00:00
|
|
|
#endif
|
2005-04-16 22:20:36 +00:00
|
|
|
unsigned long stack[THREAD_SIZE/sizeof(long)];
|
|
|
|
};
|
|
|
|
|
2018-01-02 15:12:01 +00:00
|
|
|
#ifndef CONFIG_THREAD_INFO_IN_TASK
|
|
|
|
extern struct thread_info init_thread_info;
|
|
|
|
#endif
|
|
|
|
|
|
|
|
extern unsigned long init_stack[THREAD_SIZE / sizeof(unsigned long)];
|
|
|
|
|
2017-02-03 21:59:33 +00:00
|
|
|
#ifdef CONFIG_THREAD_INFO_IN_TASK
|
|
|
|
static inline struct thread_info *task_thread_info(struct task_struct *task)
|
|
|
|
{
|
|
|
|
return &task->thread_info;
|
|
|
|
}
|
|
|
|
#elif !defined(__HAVE_THREAD_FUNCTIONS)
|
|
|
|
# define task_thread_info(task) ((struct thread_info *)(task)->stack)
|
|
|
|
#endif
|
|
|
|
|
2007-10-19 06:40:06 +00:00
|
|
|
/*
|
|
|
|
* find a task by one of its numerical ids
|
|
|
|
*
|
|
|
|
* find_task_by_pid_ns():
|
|
|
|
* finds a task by its pid in the specified namespace
|
2007-10-19 06:40:16 +00:00
|
|
|
* find_task_by_vpid():
|
|
|
|
* finds a task by its virtual pid
|
2007-10-19 06:40:06 +00:00
|
|
|
*
|
2008-07-25 08:48:36 +00:00
|
|
|
* see also find_vpid() etc in include/linux/pid.h
|
2007-10-19 06:40:06 +00:00
|
|
|
*/
|
|
|
|
|
2007-10-19 06:40:16 +00:00
|
|
|
extern struct task_struct *find_task_by_vpid(pid_t nr);
|
2017-02-06 21:06:35 +00:00
|
|
|
extern struct task_struct *find_task_by_pid_ns(pid_t nr, struct pid_namespace *ns);
|
2007-10-19 06:40:06 +00:00
|
|
|
|
2018-02-06 23:40:17 +00:00
|
|
|
/*
|
|
|
|
* find a task by its virtual pid and get the task struct
|
|
|
|
*/
|
|
|
|
extern struct task_struct *find_get_task_by_vpid(pid_t nr);
|
|
|
|
|
2008-02-13 23:03:15 +00:00
|
|
|
extern int wake_up_state(struct task_struct *tsk, unsigned int state);
|
|
|
|
extern int wake_up_process(struct task_struct *tsk);
|
2011-05-11 16:18:05 +00:00
|
|
|
extern void wake_up_new_task(struct task_struct *tsk);
|
2017-02-06 21:06:35 +00:00
|
|
|
|
2005-04-16 22:20:36 +00:00
|
|
|
#ifdef CONFIG_SMP
|
2017-02-06 21:06:35 +00:00
|
|
|
extern void kick_process(struct task_struct *tsk);
|
2005-04-16 22:20:36 +00:00
|
|
|
#else
|
2017-02-06 21:06:35 +00:00
|
|
|
static inline void kick_process(struct task_struct *tsk) { }
|
2005-04-16 22:20:36 +00:00
|
|
|
#endif
|
|
|
|
|
2014-05-28 08:45:04 +00:00
|
|
|
extern void __set_task_comm(struct task_struct *tsk, const char *from, bool exec);
|
2017-02-06 21:06:35 +00:00
|
|
|
|
2014-05-28 08:45:04 +00:00
|
|
|
static inline void set_task_comm(struct task_struct *tsk, const char *from)
|
|
|
|
{
|
|
|
|
__set_task_comm(tsk, from, false);
|
|
|
|
}
|
2017-02-06 21:06:35 +00:00
|
|
|
|
2017-12-14 23:32:41 +00:00
|
|
|
extern char *__get_task_comm(char *to, size_t len, struct task_struct *tsk);
|
|
|
|
#define get_task_comm(buf, tsk) ({ \
|
|
|
|
BUILD_BUG_ON(sizeof(buf) != TASK_COMM_LEN); \
|
|
|
|
__get_task_comm(buf, sizeof(buf), tsk); \
|
|
|
|
})
|
2005-04-16 22:20:36 +00:00
|
|
|
|
|
|
|
#ifdef CONFIG_SMP
|
2011-04-05 15:23:58 +00:00
|
|
|
void scheduler_ipi(void);
|
2008-07-26 02:45:58 +00:00
|
|
|
extern unsigned long wait_task_inactive(struct task_struct *, long match_state);
|
2005-04-16 22:20:36 +00:00
|
|
|
#else
|
2011-04-05 15:23:39 +00:00
|
|
|
static inline void scheduler_ipi(void) { }
|
2017-02-06 21:06:35 +00:00
|
|
|
static inline unsigned long wait_task_inactive(struct task_struct *p, long match_state)
|
2008-07-26 02:45:58 +00:00
|
|
|
{
|
|
|
|
return 1;
|
|
|
|
}
|
2005-04-16 22:20:36 +00:00
|
|
|
#endif
|
|
|
|
|
2017-02-06 21:06:35 +00:00
|
|
|
/*
|
|
|
|
* Set thread flags in other task's structures.
|
|
|
|
* See asm/thread_info.h for TIF_xxxx flags available:
|
2005-04-16 22:20:36 +00:00
|
|
|
*/
|
|
|
|
static inline void set_tsk_thread_flag(struct task_struct *tsk, int flag)
|
|
|
|
{
|
2005-11-14 00:06:55 +00:00
|
|
|
set_ti_thread_flag(task_thread_info(tsk), flag);
|
2005-04-16 22:20:36 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static inline void clear_tsk_thread_flag(struct task_struct *tsk, int flag)
|
|
|
|
{
|
2005-11-14 00:06:55 +00:00
|
|
|
clear_ti_thread_flag(task_thread_info(tsk), flag);
|
2005-04-16 22:20:36 +00:00
|
|
|
}
|
|
|
|
|
2018-04-11 16:54:20 +00:00
|
|
|
static inline void update_tsk_thread_flag(struct task_struct *tsk, int flag,
|
|
|
|
bool value)
|
|
|
|
{
|
|
|
|
update_ti_thread_flag(task_thread_info(tsk), flag, value);
|
|
|
|
}
|
|
|
|
|
2005-04-16 22:20:36 +00:00
|
|
|
static inline int test_and_set_tsk_thread_flag(struct task_struct *tsk, int flag)
|
|
|
|
{
|
2005-11-14 00:06:55 +00:00
|
|
|
return test_and_set_ti_thread_flag(task_thread_info(tsk), flag);
|
2005-04-16 22:20:36 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static inline int test_and_clear_tsk_thread_flag(struct task_struct *tsk, int flag)
|
|
|
|
{
|
2005-11-14 00:06:55 +00:00
|
|
|
return test_and_clear_ti_thread_flag(task_thread_info(tsk), flag);
|
2005-04-16 22:20:36 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static inline int test_tsk_thread_flag(struct task_struct *tsk, int flag)
|
|
|
|
{
|
2005-11-14 00:06:55 +00:00
|
|
|
return test_ti_thread_flag(task_thread_info(tsk), flag);
|
2005-04-16 22:20:36 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
static inline void set_tsk_need_resched(struct task_struct *tsk)
|
|
|
|
{
|
|
|
|
set_tsk_thread_flag(tsk,TIF_NEED_RESCHED);
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void clear_tsk_need_resched(struct task_struct *tsk)
|
|
|
|
{
|
|
|
|
clear_tsk_thread_flag(tsk,TIF_NEED_RESCHED);
|
|
|
|
}
|
|
|
|
|
2008-04-23 11:13:29 +00:00
|
|
|
static inline int test_tsk_need_resched(struct task_struct *tsk)
|
|
|
|
{
|
|
|
|
return unlikely(test_tsk_thread_flag(tsk,TIF_NEED_RESCHED));
|
|
|
|
}
|
|
|
|
|
2005-04-16 22:20:36 +00:00
|
|
|
/*
|
|
|
|
* cond_resched() and cond_resched_lock(): latency reduction via
|
|
|
|
* explicit rescheduling in places that are safe. The return
|
|
|
|
* value indicates whether a reschedule was done in fact.
|
|
|
|
* cond_resched_lock() will drop the spinlock before scheduling,
|
|
|
|
*/
|
2016-09-19 10:57:53 +00:00
|
|
|
#ifndef CONFIG_PREEMPT
|
2008-05-11 23:04:48 +00:00
|
|
|
extern int _cond_resched(void);
|
2016-09-19 10:57:53 +00:00
|
|
|
#else
|
|
|
|
static inline int _cond_resched(void) { return 0; }
|
|
|
|
#endif
|
2009-07-16 13:44:29 +00:00
|
|
|
|
2009-07-16 13:44:29 +00:00
|
|
|
#define cond_resched() ({ \
|
2014-09-24 08:18:56 +00:00
|
|
|
___might_sleep(__FILE__, __LINE__, 0); \
|
2009-07-16 13:44:29 +00:00
|
|
|
_cond_resched(); \
|
|
|
|
})
|
2009-07-16 13:44:29 +00:00
|
|
|
|
2009-07-16 13:44:29 +00:00
|
|
|
extern int __cond_resched_lock(spinlock_t *lock);
|
|
|
|
|
|
|
|
#define cond_resched_lock(lock) ({ \
|
2014-09-24 08:18:56 +00:00
|
|
|
___might_sleep(__FILE__, __LINE__, PREEMPT_LOCK_OFFSET);\
|
2009-07-16 13:44:29 +00:00
|
|
|
__cond_resched_lock(lock); \
|
|
|
|
})
|
|
|
|
|
2013-05-22 05:50:31 +00:00
|
|
|
static inline void cond_resched_rcu(void)
|
|
|
|
{
|
|
|
|
#if defined(CONFIG_DEBUG_ATOMIC_SLEEP) || !defined(CONFIG_PREEMPT_RCU)
|
|
|
|
rcu_read_unlock();
|
|
|
|
cond_resched();
|
|
|
|
rcu_read_lock();
|
|
|
|
#endif
|
|
|
|
}
|
|
|
|
|
2005-04-16 22:20:36 +00:00
|
|
|
/*
|
|
|
|
* Does a critical section need to be broken due to another
|
2008-01-30 12:31:20 +00:00
|
|
|
* task waiting?: (technically does not depend on CONFIG_PREEMPT,
|
|
|
|
* but a general need for low latency)
|
2005-04-16 22:20:36 +00:00
|
|
|
*/
|
2008-01-30 12:31:20 +00:00
|
|
|
static inline int spin_needbreak(spinlock_t *lock)
|
2005-04-16 22:20:36 +00:00
|
|
|
{
|
2008-01-30 12:31:20 +00:00
|
|
|
#ifdef CONFIG_PREEMPT
|
|
|
|
return spin_is_contended(lock);
|
|
|
|
#else
|
2005-04-16 22:20:36 +00:00
|
|
|
return 0;
|
2008-01-30 12:31:20 +00:00
|
|
|
#endif
|
2005-04-16 22:20:36 +00:00
|
|
|
}
|
|
|
|
|
2013-09-27 15:30:03 +00:00
|
|
|
static __always_inline bool need_resched(void)
|
|
|
|
{
|
|
|
|
return unlikely(tif_need_resched());
|
|
|
|
}
|
|
|
|
|
2005-04-16 22:20:36 +00:00
|
|
|
/*
|
|
|
|
* Wrappers for p->thread_info->cpu access. No-op on UP.
|
|
|
|
*/
|
|
|
|
#ifdef CONFIG_SMP
|
|
|
|
|
|
|
|
static inline unsigned int task_cpu(const struct task_struct *p)
|
|
|
|
{
|
2016-09-13 21:29:24 +00:00
|
|
|
#ifdef CONFIG_THREAD_INFO_IN_TASK
|
2019-01-21 15:52:40 +00:00
|
|
|
return READ_ONCE(p->cpu);
|
2016-09-13 21:29:24 +00:00
|
|
|
#else
|
2019-01-21 15:52:40 +00:00
|
|
|
return READ_ONCE(task_thread_info(p)->cpu);
|
2016-09-13 21:29:24 +00:00
|
|
|
#endif
|
2005-04-16 22:20:36 +00:00
|
|
|
}
|
|
|
|
|
2007-07-09 16:51:58 +00:00
|
|
|
extern void set_task_cpu(struct task_struct *p, unsigned int cpu);
|
2005-04-16 22:20:36 +00:00
|
|
|
|
|
|
|
#else
|
|
|
|
|
|
|
|
static inline unsigned int task_cpu(const struct task_struct *p)
|
|
|
|
{
|
|
|
|
return 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void set_task_cpu(struct task_struct *p, unsigned int cpu)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
|
|
|
|
#endif /* CONFIG_SMP */
|
|
|
|
|
2016-11-02 09:08:28 +00:00
|
|
|
/*
|
|
|
|
* In order to reduce various lock holder preemption latencies provide an
|
|
|
|
* interface to see if a vCPU is currently running or not.
|
|
|
|
*
|
|
|
|
* This allows us to terminate optimistic spin loops and block, analogous to
|
|
|
|
* the native optimistic spin heuristic of testing if the lock owner task is
|
|
|
|
* running or not.
|
|
|
|
*/
|
|
|
|
#ifndef vcpu_is_preempted
|
|
|
|
# define vcpu_is_preempted(cpu) false
|
|
|
|
#endif
|
|
|
|
|
2008-11-24 16:05:14 +00:00
|
|
|
extern long sched_setaffinity(pid_t pid, const struct cpumask *new_mask);
|
|
|
|
extern long sched_getaffinity(pid_t pid, struct cpumask *mask);
|
2006-06-27 09:54:42 +00:00
|
|
|
|
2008-02-05 06:28:59 +00:00
|
|
|
#ifndef TASK_SIZE_OF
|
|
|
|
#define TASK_SIZE_OF(tsk) TASK_SIZE
|
|
|
|
#endif
|
|
|
|
|
rseq: Introduce restartable sequences system call
Expose a new system call allowing each thread to register one userspace
memory area to be used as an ABI between kernel and user-space for two
purposes: user-space restartable sequences and quick access to read the
current CPU number value from user-space.
* Restartable sequences (per-cpu atomics)
Restartables sequences allow user-space to perform update operations on
per-cpu data without requiring heavy-weight atomic operations.
The restartable critical sections (percpu atomics) work has been started
by Paul Turner and Andrew Hunter. It lets the kernel handle restart of
critical sections. [1] [2] The re-implementation proposed here brings a
few simplifications to the ABI which facilitates porting to other
architectures and speeds up the user-space fast path.
Here are benchmarks of various rseq use-cases.
Test hardware:
arm32: ARMv7 Processor rev 4 (v7l) "Cubietruck", 2-core
x86-64: Intel E5-2630 v3@2.40GHz, 16-core, hyperthreading
The following benchmarks were all performed on a single thread.
* Per-CPU statistic counter increment
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 344.0 31.4 11.0
x86-64: 15.3 2.0 7.7
* LTTng-UST: write event 32-bit header, 32-bit payload into tracer
per-cpu buffer
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 2502.0 2250.0 1.1
x86-64: 117.4 98.0 1.2
* liburcu percpu: lock-unlock pair, dereference, read/compare word
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 751.0 128.5 5.8
x86-64: 53.4 28.6 1.9
* jemalloc memory allocator adapted to use rseq
Using rseq with per-cpu memory pools in jemalloc at Facebook (based on
rseq 2016 implementation):
The production workload response-time has 1-2% gain avg. latency, and
the P99 overall latency drops by 2-3%.
* Reading the current CPU number
Speeding up reading the current CPU number on which the caller thread is
running is done by keeping the current CPU number up do date within the
cpu_id field of the memory area registered by the thread. This is done
by making scheduler preemption set the TIF_NOTIFY_RESUME flag on the
current thread. Upon return to user-space, a notify-resume handler
updates the current CPU value within the registered user-space memory
area. User-space can then read the current CPU number directly from
memory.
Keeping the current cpu id in a memory area shared between kernel and
user-space is an improvement over current mechanisms available to read
the current CPU number, which has the following benefits over
alternative approaches:
- 35x speedup on ARM vs system call through glibc
- 20x speedup on x86 compared to calling glibc, which calls vdso
executing a "lsl" instruction,
- 14x speedup on x86 compared to inlined "lsl" instruction,
- Unlike vdso approaches, this cpu_id value can be read from an inline
assembly, which makes it a useful building block for restartable
sequences.
- The approach of reading the cpu id through memory mapping shared
between kernel and user-space is portable (e.g. ARM), which is not the
case for the lsl-based x86 vdso.
On x86, yet another possible approach would be to use the gs segment
selector to point to user-space per-cpu data. This approach performs
similarly to the cpu id cache, but it has two disadvantages: it is
not portable, and it is incompatible with existing applications already
using the gs segment selector for other purposes.
Benchmarking various approaches for reading the current CPU number:
ARMv7 Processor rev 4 (v7l)
Machine model: Cubietruck
- Baseline (empty loop): 8.4 ns
- Read CPU from rseq cpu_id: 16.7 ns
- Read CPU from rseq cpu_id (lazy register): 19.8 ns
- glibc 2.19-0ubuntu6.6 getcpu: 301.8 ns
- getcpu system call: 234.9 ns
x86-64 Intel(R) Xeon(R) CPU E5-2630 v3 @ 2.40GHz:
- Baseline (empty loop): 0.8 ns
- Read CPU from rseq cpu_id: 0.8 ns
- Read CPU from rseq cpu_id (lazy register): 0.8 ns
- Read using gs segment selector: 0.8 ns
- "lsl" inline assembly: 13.0 ns
- glibc 2.19-0ubuntu6 getcpu: 16.6 ns
- getcpu system call: 53.9 ns
- Speed (benchmark taken on v8 of patchset)
Running 10 runs of hackbench -l 100000 seems to indicate, contrary to
expectations, that enabling CONFIG_RSEQ slightly accelerates the
scheduler:
Configuration: 2 sockets * 8-core Intel(R) Xeon(R) CPU E5-2630 v3 @
2.40GHz (directly on hardware, hyperthreading disabled in BIOS, energy
saving disabled in BIOS, turboboost disabled in BIOS, cpuidle.off=1
kernel parameter), with a Linux v4.6 defconfig+localyesconfig,
restartable sequences series applied.
* CONFIG_RSEQ=n
avg.: 41.37 s
std.dev.: 0.36 s
* CONFIG_RSEQ=y
avg.: 40.46 s
std.dev.: 0.33 s
- Size
On x86-64, between CONFIG_RSEQ=n/y, the text size increase of vmlinux is
567 bytes, and the data size increase of vmlinux is 5696 bytes.
[1] https://lwn.net/Articles/650333/
[2] http://www.linuxplumbersconf.org/2013/ocw/system/presentations/1695/original/LPC%20-%20PerCpu%20Atomics.pdf
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Joel Fernandes <joelaf@google.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Dave Watson <davejwatson@fb.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Chris Lameter <cl@linux.com>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Andrew Hunter <ahh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: "Paul E . McKenney" <paulmck@linux.vnet.ibm.com>
Cc: Paul Turner <pjt@google.com>
Cc: Boqun Feng <boqun.feng@gmail.com>
Cc: Josh Triplett <josh@joshtriplett.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Ben Maurer <bmaurer@fb.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: linux-api@vger.kernel.org
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/20151027235635.16059.11630.stgit@pjt-glaptop.roam.corp.google.com
Link: http://lkml.kernel.org/r/20150624222609.6116.86035.stgit@kitami.mtv.corp.google.com
Link: https://lkml.kernel.org/r/20180602124408.8430-3-mathieu.desnoyers@efficios.com
2018-06-02 12:43:54 +00:00
|
|
|
#ifdef CONFIG_RSEQ
|
|
|
|
|
|
|
|
/*
|
|
|
|
* Map the event mask on the user-space ABI enum rseq_cs_flags
|
|
|
|
* for direct mask checks.
|
|
|
|
*/
|
|
|
|
enum rseq_event_mask_bits {
|
|
|
|
RSEQ_EVENT_PREEMPT_BIT = RSEQ_CS_FLAG_NO_RESTART_ON_PREEMPT_BIT,
|
|
|
|
RSEQ_EVENT_SIGNAL_BIT = RSEQ_CS_FLAG_NO_RESTART_ON_SIGNAL_BIT,
|
|
|
|
RSEQ_EVENT_MIGRATE_BIT = RSEQ_CS_FLAG_NO_RESTART_ON_MIGRATE_BIT,
|
|
|
|
};
|
|
|
|
|
|
|
|
enum rseq_event_mask {
|
|
|
|
RSEQ_EVENT_PREEMPT = (1U << RSEQ_EVENT_PREEMPT_BIT),
|
|
|
|
RSEQ_EVENT_SIGNAL = (1U << RSEQ_EVENT_SIGNAL_BIT),
|
|
|
|
RSEQ_EVENT_MIGRATE = (1U << RSEQ_EVENT_MIGRATE_BIT),
|
|
|
|
};
|
|
|
|
|
|
|
|
static inline void rseq_set_notify_resume(struct task_struct *t)
|
|
|
|
{
|
|
|
|
if (t->rseq)
|
|
|
|
set_tsk_thread_flag(t, TIF_NOTIFY_RESUME);
|
|
|
|
}
|
|
|
|
|
2018-06-22 10:45:07 +00:00
|
|
|
void __rseq_handle_notify_resume(struct ksignal *sig, struct pt_regs *regs);
|
rseq: Introduce restartable sequences system call
Expose a new system call allowing each thread to register one userspace
memory area to be used as an ABI between kernel and user-space for two
purposes: user-space restartable sequences and quick access to read the
current CPU number value from user-space.
* Restartable sequences (per-cpu atomics)
Restartables sequences allow user-space to perform update operations on
per-cpu data without requiring heavy-weight atomic operations.
The restartable critical sections (percpu atomics) work has been started
by Paul Turner and Andrew Hunter. It lets the kernel handle restart of
critical sections. [1] [2] The re-implementation proposed here brings a
few simplifications to the ABI which facilitates porting to other
architectures and speeds up the user-space fast path.
Here are benchmarks of various rseq use-cases.
Test hardware:
arm32: ARMv7 Processor rev 4 (v7l) "Cubietruck", 2-core
x86-64: Intel E5-2630 v3@2.40GHz, 16-core, hyperthreading
The following benchmarks were all performed on a single thread.
* Per-CPU statistic counter increment
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 344.0 31.4 11.0
x86-64: 15.3 2.0 7.7
* LTTng-UST: write event 32-bit header, 32-bit payload into tracer
per-cpu buffer
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 2502.0 2250.0 1.1
x86-64: 117.4 98.0 1.2
* liburcu percpu: lock-unlock pair, dereference, read/compare word
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 751.0 128.5 5.8
x86-64: 53.4 28.6 1.9
* jemalloc memory allocator adapted to use rseq
Using rseq with per-cpu memory pools in jemalloc at Facebook (based on
rseq 2016 implementation):
The production workload response-time has 1-2% gain avg. latency, and
the P99 overall latency drops by 2-3%.
* Reading the current CPU number
Speeding up reading the current CPU number on which the caller thread is
running is done by keeping the current CPU number up do date within the
cpu_id field of the memory area registered by the thread. This is done
by making scheduler preemption set the TIF_NOTIFY_RESUME flag on the
current thread. Upon return to user-space, a notify-resume handler
updates the current CPU value within the registered user-space memory
area. User-space can then read the current CPU number directly from
memory.
Keeping the current cpu id in a memory area shared between kernel and
user-space is an improvement over current mechanisms available to read
the current CPU number, which has the following benefits over
alternative approaches:
- 35x speedup on ARM vs system call through glibc
- 20x speedup on x86 compared to calling glibc, which calls vdso
executing a "lsl" instruction,
- 14x speedup on x86 compared to inlined "lsl" instruction,
- Unlike vdso approaches, this cpu_id value can be read from an inline
assembly, which makes it a useful building block for restartable
sequences.
- The approach of reading the cpu id through memory mapping shared
between kernel and user-space is portable (e.g. ARM), which is not the
case for the lsl-based x86 vdso.
On x86, yet another possible approach would be to use the gs segment
selector to point to user-space per-cpu data. This approach performs
similarly to the cpu id cache, but it has two disadvantages: it is
not portable, and it is incompatible with existing applications already
using the gs segment selector for other purposes.
Benchmarking various approaches for reading the current CPU number:
ARMv7 Processor rev 4 (v7l)
Machine model: Cubietruck
- Baseline (empty loop): 8.4 ns
- Read CPU from rseq cpu_id: 16.7 ns
- Read CPU from rseq cpu_id (lazy register): 19.8 ns
- glibc 2.19-0ubuntu6.6 getcpu: 301.8 ns
- getcpu system call: 234.9 ns
x86-64 Intel(R) Xeon(R) CPU E5-2630 v3 @ 2.40GHz:
- Baseline (empty loop): 0.8 ns
- Read CPU from rseq cpu_id: 0.8 ns
- Read CPU from rseq cpu_id (lazy register): 0.8 ns
- Read using gs segment selector: 0.8 ns
- "lsl" inline assembly: 13.0 ns
- glibc 2.19-0ubuntu6 getcpu: 16.6 ns
- getcpu system call: 53.9 ns
- Speed (benchmark taken on v8 of patchset)
Running 10 runs of hackbench -l 100000 seems to indicate, contrary to
expectations, that enabling CONFIG_RSEQ slightly accelerates the
scheduler:
Configuration: 2 sockets * 8-core Intel(R) Xeon(R) CPU E5-2630 v3 @
2.40GHz (directly on hardware, hyperthreading disabled in BIOS, energy
saving disabled in BIOS, turboboost disabled in BIOS, cpuidle.off=1
kernel parameter), with a Linux v4.6 defconfig+localyesconfig,
restartable sequences series applied.
* CONFIG_RSEQ=n
avg.: 41.37 s
std.dev.: 0.36 s
* CONFIG_RSEQ=y
avg.: 40.46 s
std.dev.: 0.33 s
- Size
On x86-64, between CONFIG_RSEQ=n/y, the text size increase of vmlinux is
567 bytes, and the data size increase of vmlinux is 5696 bytes.
[1] https://lwn.net/Articles/650333/
[2] http://www.linuxplumbersconf.org/2013/ocw/system/presentations/1695/original/LPC%20-%20PerCpu%20Atomics.pdf
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Joel Fernandes <joelaf@google.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Dave Watson <davejwatson@fb.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Chris Lameter <cl@linux.com>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Andrew Hunter <ahh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: "Paul E . McKenney" <paulmck@linux.vnet.ibm.com>
Cc: Paul Turner <pjt@google.com>
Cc: Boqun Feng <boqun.feng@gmail.com>
Cc: Josh Triplett <josh@joshtriplett.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Ben Maurer <bmaurer@fb.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: linux-api@vger.kernel.org
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/20151027235635.16059.11630.stgit@pjt-glaptop.roam.corp.google.com
Link: http://lkml.kernel.org/r/20150624222609.6116.86035.stgit@kitami.mtv.corp.google.com
Link: https://lkml.kernel.org/r/20180602124408.8430-3-mathieu.desnoyers@efficios.com
2018-06-02 12:43:54 +00:00
|
|
|
|
2018-06-22 10:45:07 +00:00
|
|
|
static inline void rseq_handle_notify_resume(struct ksignal *ksig,
|
|
|
|
struct pt_regs *regs)
|
rseq: Introduce restartable sequences system call
Expose a new system call allowing each thread to register one userspace
memory area to be used as an ABI between kernel and user-space for two
purposes: user-space restartable sequences and quick access to read the
current CPU number value from user-space.
* Restartable sequences (per-cpu atomics)
Restartables sequences allow user-space to perform update operations on
per-cpu data without requiring heavy-weight atomic operations.
The restartable critical sections (percpu atomics) work has been started
by Paul Turner and Andrew Hunter. It lets the kernel handle restart of
critical sections. [1] [2] The re-implementation proposed here brings a
few simplifications to the ABI which facilitates porting to other
architectures and speeds up the user-space fast path.
Here are benchmarks of various rseq use-cases.
Test hardware:
arm32: ARMv7 Processor rev 4 (v7l) "Cubietruck", 2-core
x86-64: Intel E5-2630 v3@2.40GHz, 16-core, hyperthreading
The following benchmarks were all performed on a single thread.
* Per-CPU statistic counter increment
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 344.0 31.4 11.0
x86-64: 15.3 2.0 7.7
* LTTng-UST: write event 32-bit header, 32-bit payload into tracer
per-cpu buffer
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 2502.0 2250.0 1.1
x86-64: 117.4 98.0 1.2
* liburcu percpu: lock-unlock pair, dereference, read/compare word
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 751.0 128.5 5.8
x86-64: 53.4 28.6 1.9
* jemalloc memory allocator adapted to use rseq
Using rseq with per-cpu memory pools in jemalloc at Facebook (based on
rseq 2016 implementation):
The production workload response-time has 1-2% gain avg. latency, and
the P99 overall latency drops by 2-3%.
* Reading the current CPU number
Speeding up reading the current CPU number on which the caller thread is
running is done by keeping the current CPU number up do date within the
cpu_id field of the memory area registered by the thread. This is done
by making scheduler preemption set the TIF_NOTIFY_RESUME flag on the
current thread. Upon return to user-space, a notify-resume handler
updates the current CPU value within the registered user-space memory
area. User-space can then read the current CPU number directly from
memory.
Keeping the current cpu id in a memory area shared between kernel and
user-space is an improvement over current mechanisms available to read
the current CPU number, which has the following benefits over
alternative approaches:
- 35x speedup on ARM vs system call through glibc
- 20x speedup on x86 compared to calling glibc, which calls vdso
executing a "lsl" instruction,
- 14x speedup on x86 compared to inlined "lsl" instruction,
- Unlike vdso approaches, this cpu_id value can be read from an inline
assembly, which makes it a useful building block for restartable
sequences.
- The approach of reading the cpu id through memory mapping shared
between kernel and user-space is portable (e.g. ARM), which is not the
case for the lsl-based x86 vdso.
On x86, yet another possible approach would be to use the gs segment
selector to point to user-space per-cpu data. This approach performs
similarly to the cpu id cache, but it has two disadvantages: it is
not portable, and it is incompatible with existing applications already
using the gs segment selector for other purposes.
Benchmarking various approaches for reading the current CPU number:
ARMv7 Processor rev 4 (v7l)
Machine model: Cubietruck
- Baseline (empty loop): 8.4 ns
- Read CPU from rseq cpu_id: 16.7 ns
- Read CPU from rseq cpu_id (lazy register): 19.8 ns
- glibc 2.19-0ubuntu6.6 getcpu: 301.8 ns
- getcpu system call: 234.9 ns
x86-64 Intel(R) Xeon(R) CPU E5-2630 v3 @ 2.40GHz:
- Baseline (empty loop): 0.8 ns
- Read CPU from rseq cpu_id: 0.8 ns
- Read CPU from rseq cpu_id (lazy register): 0.8 ns
- Read using gs segment selector: 0.8 ns
- "lsl" inline assembly: 13.0 ns
- glibc 2.19-0ubuntu6 getcpu: 16.6 ns
- getcpu system call: 53.9 ns
- Speed (benchmark taken on v8 of patchset)
Running 10 runs of hackbench -l 100000 seems to indicate, contrary to
expectations, that enabling CONFIG_RSEQ slightly accelerates the
scheduler:
Configuration: 2 sockets * 8-core Intel(R) Xeon(R) CPU E5-2630 v3 @
2.40GHz (directly on hardware, hyperthreading disabled in BIOS, energy
saving disabled in BIOS, turboboost disabled in BIOS, cpuidle.off=1
kernel parameter), with a Linux v4.6 defconfig+localyesconfig,
restartable sequences series applied.
* CONFIG_RSEQ=n
avg.: 41.37 s
std.dev.: 0.36 s
* CONFIG_RSEQ=y
avg.: 40.46 s
std.dev.: 0.33 s
- Size
On x86-64, between CONFIG_RSEQ=n/y, the text size increase of vmlinux is
567 bytes, and the data size increase of vmlinux is 5696 bytes.
[1] https://lwn.net/Articles/650333/
[2] http://www.linuxplumbersconf.org/2013/ocw/system/presentations/1695/original/LPC%20-%20PerCpu%20Atomics.pdf
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Joel Fernandes <joelaf@google.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Dave Watson <davejwatson@fb.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Chris Lameter <cl@linux.com>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Andrew Hunter <ahh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: "Paul E . McKenney" <paulmck@linux.vnet.ibm.com>
Cc: Paul Turner <pjt@google.com>
Cc: Boqun Feng <boqun.feng@gmail.com>
Cc: Josh Triplett <josh@joshtriplett.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Ben Maurer <bmaurer@fb.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: linux-api@vger.kernel.org
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/20151027235635.16059.11630.stgit@pjt-glaptop.roam.corp.google.com
Link: http://lkml.kernel.org/r/20150624222609.6116.86035.stgit@kitami.mtv.corp.google.com
Link: https://lkml.kernel.org/r/20180602124408.8430-3-mathieu.desnoyers@efficios.com
2018-06-02 12:43:54 +00:00
|
|
|
{
|
|
|
|
if (current->rseq)
|
2018-06-22 10:45:07 +00:00
|
|
|
__rseq_handle_notify_resume(ksig, regs);
|
rseq: Introduce restartable sequences system call
Expose a new system call allowing each thread to register one userspace
memory area to be used as an ABI between kernel and user-space for two
purposes: user-space restartable sequences and quick access to read the
current CPU number value from user-space.
* Restartable sequences (per-cpu atomics)
Restartables sequences allow user-space to perform update operations on
per-cpu data without requiring heavy-weight atomic operations.
The restartable critical sections (percpu atomics) work has been started
by Paul Turner and Andrew Hunter. It lets the kernel handle restart of
critical sections. [1] [2] The re-implementation proposed here brings a
few simplifications to the ABI which facilitates porting to other
architectures and speeds up the user-space fast path.
Here are benchmarks of various rseq use-cases.
Test hardware:
arm32: ARMv7 Processor rev 4 (v7l) "Cubietruck", 2-core
x86-64: Intel E5-2630 v3@2.40GHz, 16-core, hyperthreading
The following benchmarks were all performed on a single thread.
* Per-CPU statistic counter increment
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 344.0 31.4 11.0
x86-64: 15.3 2.0 7.7
* LTTng-UST: write event 32-bit header, 32-bit payload into tracer
per-cpu buffer
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 2502.0 2250.0 1.1
x86-64: 117.4 98.0 1.2
* liburcu percpu: lock-unlock pair, dereference, read/compare word
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 751.0 128.5 5.8
x86-64: 53.4 28.6 1.9
* jemalloc memory allocator adapted to use rseq
Using rseq with per-cpu memory pools in jemalloc at Facebook (based on
rseq 2016 implementation):
The production workload response-time has 1-2% gain avg. latency, and
the P99 overall latency drops by 2-3%.
* Reading the current CPU number
Speeding up reading the current CPU number on which the caller thread is
running is done by keeping the current CPU number up do date within the
cpu_id field of the memory area registered by the thread. This is done
by making scheduler preemption set the TIF_NOTIFY_RESUME flag on the
current thread. Upon return to user-space, a notify-resume handler
updates the current CPU value within the registered user-space memory
area. User-space can then read the current CPU number directly from
memory.
Keeping the current cpu id in a memory area shared between kernel and
user-space is an improvement over current mechanisms available to read
the current CPU number, which has the following benefits over
alternative approaches:
- 35x speedup on ARM vs system call through glibc
- 20x speedup on x86 compared to calling glibc, which calls vdso
executing a "lsl" instruction,
- 14x speedup on x86 compared to inlined "lsl" instruction,
- Unlike vdso approaches, this cpu_id value can be read from an inline
assembly, which makes it a useful building block for restartable
sequences.
- The approach of reading the cpu id through memory mapping shared
between kernel and user-space is portable (e.g. ARM), which is not the
case for the lsl-based x86 vdso.
On x86, yet another possible approach would be to use the gs segment
selector to point to user-space per-cpu data. This approach performs
similarly to the cpu id cache, but it has two disadvantages: it is
not portable, and it is incompatible with existing applications already
using the gs segment selector for other purposes.
Benchmarking various approaches for reading the current CPU number:
ARMv7 Processor rev 4 (v7l)
Machine model: Cubietruck
- Baseline (empty loop): 8.4 ns
- Read CPU from rseq cpu_id: 16.7 ns
- Read CPU from rseq cpu_id (lazy register): 19.8 ns
- glibc 2.19-0ubuntu6.6 getcpu: 301.8 ns
- getcpu system call: 234.9 ns
x86-64 Intel(R) Xeon(R) CPU E5-2630 v3 @ 2.40GHz:
- Baseline (empty loop): 0.8 ns
- Read CPU from rseq cpu_id: 0.8 ns
- Read CPU from rseq cpu_id (lazy register): 0.8 ns
- Read using gs segment selector: 0.8 ns
- "lsl" inline assembly: 13.0 ns
- glibc 2.19-0ubuntu6 getcpu: 16.6 ns
- getcpu system call: 53.9 ns
- Speed (benchmark taken on v8 of patchset)
Running 10 runs of hackbench -l 100000 seems to indicate, contrary to
expectations, that enabling CONFIG_RSEQ slightly accelerates the
scheduler:
Configuration: 2 sockets * 8-core Intel(R) Xeon(R) CPU E5-2630 v3 @
2.40GHz (directly on hardware, hyperthreading disabled in BIOS, energy
saving disabled in BIOS, turboboost disabled in BIOS, cpuidle.off=1
kernel parameter), with a Linux v4.6 defconfig+localyesconfig,
restartable sequences series applied.
* CONFIG_RSEQ=n
avg.: 41.37 s
std.dev.: 0.36 s
* CONFIG_RSEQ=y
avg.: 40.46 s
std.dev.: 0.33 s
- Size
On x86-64, between CONFIG_RSEQ=n/y, the text size increase of vmlinux is
567 bytes, and the data size increase of vmlinux is 5696 bytes.
[1] https://lwn.net/Articles/650333/
[2] http://www.linuxplumbersconf.org/2013/ocw/system/presentations/1695/original/LPC%20-%20PerCpu%20Atomics.pdf
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Joel Fernandes <joelaf@google.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Dave Watson <davejwatson@fb.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Chris Lameter <cl@linux.com>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Andrew Hunter <ahh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: "Paul E . McKenney" <paulmck@linux.vnet.ibm.com>
Cc: Paul Turner <pjt@google.com>
Cc: Boqun Feng <boqun.feng@gmail.com>
Cc: Josh Triplett <josh@joshtriplett.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Ben Maurer <bmaurer@fb.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: linux-api@vger.kernel.org
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/20151027235635.16059.11630.stgit@pjt-glaptop.roam.corp.google.com
Link: http://lkml.kernel.org/r/20150624222609.6116.86035.stgit@kitami.mtv.corp.google.com
Link: https://lkml.kernel.org/r/20180602124408.8430-3-mathieu.desnoyers@efficios.com
2018-06-02 12:43:54 +00:00
|
|
|
}
|
|
|
|
|
2018-06-22 10:45:07 +00:00
|
|
|
static inline void rseq_signal_deliver(struct ksignal *ksig,
|
|
|
|
struct pt_regs *regs)
|
rseq: Introduce restartable sequences system call
Expose a new system call allowing each thread to register one userspace
memory area to be used as an ABI between kernel and user-space for two
purposes: user-space restartable sequences and quick access to read the
current CPU number value from user-space.
* Restartable sequences (per-cpu atomics)
Restartables sequences allow user-space to perform update operations on
per-cpu data without requiring heavy-weight atomic operations.
The restartable critical sections (percpu atomics) work has been started
by Paul Turner and Andrew Hunter. It lets the kernel handle restart of
critical sections. [1] [2] The re-implementation proposed here brings a
few simplifications to the ABI which facilitates porting to other
architectures and speeds up the user-space fast path.
Here are benchmarks of various rseq use-cases.
Test hardware:
arm32: ARMv7 Processor rev 4 (v7l) "Cubietruck", 2-core
x86-64: Intel E5-2630 v3@2.40GHz, 16-core, hyperthreading
The following benchmarks were all performed on a single thread.
* Per-CPU statistic counter increment
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 344.0 31.4 11.0
x86-64: 15.3 2.0 7.7
* LTTng-UST: write event 32-bit header, 32-bit payload into tracer
per-cpu buffer
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 2502.0 2250.0 1.1
x86-64: 117.4 98.0 1.2
* liburcu percpu: lock-unlock pair, dereference, read/compare word
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 751.0 128.5 5.8
x86-64: 53.4 28.6 1.9
* jemalloc memory allocator adapted to use rseq
Using rseq with per-cpu memory pools in jemalloc at Facebook (based on
rseq 2016 implementation):
The production workload response-time has 1-2% gain avg. latency, and
the P99 overall latency drops by 2-3%.
* Reading the current CPU number
Speeding up reading the current CPU number on which the caller thread is
running is done by keeping the current CPU number up do date within the
cpu_id field of the memory area registered by the thread. This is done
by making scheduler preemption set the TIF_NOTIFY_RESUME flag on the
current thread. Upon return to user-space, a notify-resume handler
updates the current CPU value within the registered user-space memory
area. User-space can then read the current CPU number directly from
memory.
Keeping the current cpu id in a memory area shared between kernel and
user-space is an improvement over current mechanisms available to read
the current CPU number, which has the following benefits over
alternative approaches:
- 35x speedup on ARM vs system call through glibc
- 20x speedup on x86 compared to calling glibc, which calls vdso
executing a "lsl" instruction,
- 14x speedup on x86 compared to inlined "lsl" instruction,
- Unlike vdso approaches, this cpu_id value can be read from an inline
assembly, which makes it a useful building block for restartable
sequences.
- The approach of reading the cpu id through memory mapping shared
between kernel and user-space is portable (e.g. ARM), which is not the
case for the lsl-based x86 vdso.
On x86, yet another possible approach would be to use the gs segment
selector to point to user-space per-cpu data. This approach performs
similarly to the cpu id cache, but it has two disadvantages: it is
not portable, and it is incompatible with existing applications already
using the gs segment selector for other purposes.
Benchmarking various approaches for reading the current CPU number:
ARMv7 Processor rev 4 (v7l)
Machine model: Cubietruck
- Baseline (empty loop): 8.4 ns
- Read CPU from rseq cpu_id: 16.7 ns
- Read CPU from rseq cpu_id (lazy register): 19.8 ns
- glibc 2.19-0ubuntu6.6 getcpu: 301.8 ns
- getcpu system call: 234.9 ns
x86-64 Intel(R) Xeon(R) CPU E5-2630 v3 @ 2.40GHz:
- Baseline (empty loop): 0.8 ns
- Read CPU from rseq cpu_id: 0.8 ns
- Read CPU from rseq cpu_id (lazy register): 0.8 ns
- Read using gs segment selector: 0.8 ns
- "lsl" inline assembly: 13.0 ns
- glibc 2.19-0ubuntu6 getcpu: 16.6 ns
- getcpu system call: 53.9 ns
- Speed (benchmark taken on v8 of patchset)
Running 10 runs of hackbench -l 100000 seems to indicate, contrary to
expectations, that enabling CONFIG_RSEQ slightly accelerates the
scheduler:
Configuration: 2 sockets * 8-core Intel(R) Xeon(R) CPU E5-2630 v3 @
2.40GHz (directly on hardware, hyperthreading disabled in BIOS, energy
saving disabled in BIOS, turboboost disabled in BIOS, cpuidle.off=1
kernel parameter), with a Linux v4.6 defconfig+localyesconfig,
restartable sequences series applied.
* CONFIG_RSEQ=n
avg.: 41.37 s
std.dev.: 0.36 s
* CONFIG_RSEQ=y
avg.: 40.46 s
std.dev.: 0.33 s
- Size
On x86-64, between CONFIG_RSEQ=n/y, the text size increase of vmlinux is
567 bytes, and the data size increase of vmlinux is 5696 bytes.
[1] https://lwn.net/Articles/650333/
[2] http://www.linuxplumbersconf.org/2013/ocw/system/presentations/1695/original/LPC%20-%20PerCpu%20Atomics.pdf
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Joel Fernandes <joelaf@google.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Dave Watson <davejwatson@fb.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Chris Lameter <cl@linux.com>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Andrew Hunter <ahh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: "Paul E . McKenney" <paulmck@linux.vnet.ibm.com>
Cc: Paul Turner <pjt@google.com>
Cc: Boqun Feng <boqun.feng@gmail.com>
Cc: Josh Triplett <josh@joshtriplett.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Ben Maurer <bmaurer@fb.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: linux-api@vger.kernel.org
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/20151027235635.16059.11630.stgit@pjt-glaptop.roam.corp.google.com
Link: http://lkml.kernel.org/r/20150624222609.6116.86035.stgit@kitami.mtv.corp.google.com
Link: https://lkml.kernel.org/r/20180602124408.8430-3-mathieu.desnoyers@efficios.com
2018-06-02 12:43:54 +00:00
|
|
|
{
|
|
|
|
preempt_disable();
|
|
|
|
__set_bit(RSEQ_EVENT_SIGNAL_BIT, ¤t->rseq_event_mask);
|
|
|
|
preempt_enable();
|
2018-06-22 10:45:07 +00:00
|
|
|
rseq_handle_notify_resume(ksig, regs);
|
rseq: Introduce restartable sequences system call
Expose a new system call allowing each thread to register one userspace
memory area to be used as an ABI between kernel and user-space for two
purposes: user-space restartable sequences and quick access to read the
current CPU number value from user-space.
* Restartable sequences (per-cpu atomics)
Restartables sequences allow user-space to perform update operations on
per-cpu data without requiring heavy-weight atomic operations.
The restartable critical sections (percpu atomics) work has been started
by Paul Turner and Andrew Hunter. It lets the kernel handle restart of
critical sections. [1] [2] The re-implementation proposed here brings a
few simplifications to the ABI which facilitates porting to other
architectures and speeds up the user-space fast path.
Here are benchmarks of various rseq use-cases.
Test hardware:
arm32: ARMv7 Processor rev 4 (v7l) "Cubietruck", 2-core
x86-64: Intel E5-2630 v3@2.40GHz, 16-core, hyperthreading
The following benchmarks were all performed on a single thread.
* Per-CPU statistic counter increment
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 344.0 31.4 11.0
x86-64: 15.3 2.0 7.7
* LTTng-UST: write event 32-bit header, 32-bit payload into tracer
per-cpu buffer
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 2502.0 2250.0 1.1
x86-64: 117.4 98.0 1.2
* liburcu percpu: lock-unlock pair, dereference, read/compare word
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 751.0 128.5 5.8
x86-64: 53.4 28.6 1.9
* jemalloc memory allocator adapted to use rseq
Using rseq with per-cpu memory pools in jemalloc at Facebook (based on
rseq 2016 implementation):
The production workload response-time has 1-2% gain avg. latency, and
the P99 overall latency drops by 2-3%.
* Reading the current CPU number
Speeding up reading the current CPU number on which the caller thread is
running is done by keeping the current CPU number up do date within the
cpu_id field of the memory area registered by the thread. This is done
by making scheduler preemption set the TIF_NOTIFY_RESUME flag on the
current thread. Upon return to user-space, a notify-resume handler
updates the current CPU value within the registered user-space memory
area. User-space can then read the current CPU number directly from
memory.
Keeping the current cpu id in a memory area shared between kernel and
user-space is an improvement over current mechanisms available to read
the current CPU number, which has the following benefits over
alternative approaches:
- 35x speedup on ARM vs system call through glibc
- 20x speedup on x86 compared to calling glibc, which calls vdso
executing a "lsl" instruction,
- 14x speedup on x86 compared to inlined "lsl" instruction,
- Unlike vdso approaches, this cpu_id value can be read from an inline
assembly, which makes it a useful building block for restartable
sequences.
- The approach of reading the cpu id through memory mapping shared
between kernel and user-space is portable (e.g. ARM), which is not the
case for the lsl-based x86 vdso.
On x86, yet another possible approach would be to use the gs segment
selector to point to user-space per-cpu data. This approach performs
similarly to the cpu id cache, but it has two disadvantages: it is
not portable, and it is incompatible with existing applications already
using the gs segment selector for other purposes.
Benchmarking various approaches for reading the current CPU number:
ARMv7 Processor rev 4 (v7l)
Machine model: Cubietruck
- Baseline (empty loop): 8.4 ns
- Read CPU from rseq cpu_id: 16.7 ns
- Read CPU from rseq cpu_id (lazy register): 19.8 ns
- glibc 2.19-0ubuntu6.6 getcpu: 301.8 ns
- getcpu system call: 234.9 ns
x86-64 Intel(R) Xeon(R) CPU E5-2630 v3 @ 2.40GHz:
- Baseline (empty loop): 0.8 ns
- Read CPU from rseq cpu_id: 0.8 ns
- Read CPU from rseq cpu_id (lazy register): 0.8 ns
- Read using gs segment selector: 0.8 ns
- "lsl" inline assembly: 13.0 ns
- glibc 2.19-0ubuntu6 getcpu: 16.6 ns
- getcpu system call: 53.9 ns
- Speed (benchmark taken on v8 of patchset)
Running 10 runs of hackbench -l 100000 seems to indicate, contrary to
expectations, that enabling CONFIG_RSEQ slightly accelerates the
scheduler:
Configuration: 2 sockets * 8-core Intel(R) Xeon(R) CPU E5-2630 v3 @
2.40GHz (directly on hardware, hyperthreading disabled in BIOS, energy
saving disabled in BIOS, turboboost disabled in BIOS, cpuidle.off=1
kernel parameter), with a Linux v4.6 defconfig+localyesconfig,
restartable sequences series applied.
* CONFIG_RSEQ=n
avg.: 41.37 s
std.dev.: 0.36 s
* CONFIG_RSEQ=y
avg.: 40.46 s
std.dev.: 0.33 s
- Size
On x86-64, between CONFIG_RSEQ=n/y, the text size increase of vmlinux is
567 bytes, and the data size increase of vmlinux is 5696 bytes.
[1] https://lwn.net/Articles/650333/
[2] http://www.linuxplumbersconf.org/2013/ocw/system/presentations/1695/original/LPC%20-%20PerCpu%20Atomics.pdf
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Joel Fernandes <joelaf@google.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Dave Watson <davejwatson@fb.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Chris Lameter <cl@linux.com>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Andrew Hunter <ahh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: "Paul E . McKenney" <paulmck@linux.vnet.ibm.com>
Cc: Paul Turner <pjt@google.com>
Cc: Boqun Feng <boqun.feng@gmail.com>
Cc: Josh Triplett <josh@joshtriplett.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Ben Maurer <bmaurer@fb.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: linux-api@vger.kernel.org
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/20151027235635.16059.11630.stgit@pjt-glaptop.roam.corp.google.com
Link: http://lkml.kernel.org/r/20150624222609.6116.86035.stgit@kitami.mtv.corp.google.com
Link: https://lkml.kernel.org/r/20180602124408.8430-3-mathieu.desnoyers@efficios.com
2018-06-02 12:43:54 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
/* rseq_preempt() requires preemption to be disabled. */
|
|
|
|
static inline void rseq_preempt(struct task_struct *t)
|
|
|
|
{
|
|
|
|
__set_bit(RSEQ_EVENT_PREEMPT_BIT, &t->rseq_event_mask);
|
|
|
|
rseq_set_notify_resume(t);
|
|
|
|
}
|
|
|
|
|
|
|
|
/* rseq_migrate() requires preemption to be disabled. */
|
|
|
|
static inline void rseq_migrate(struct task_struct *t)
|
|
|
|
{
|
|
|
|
__set_bit(RSEQ_EVENT_MIGRATE_BIT, &t->rseq_event_mask);
|
|
|
|
rseq_set_notify_resume(t);
|
|
|
|
}
|
|
|
|
|
|
|
|
/*
|
|
|
|
* If parent process has a registered restartable sequences area, the
|
2018-06-19 13:32:30 +00:00
|
|
|
* child inherits. Only applies when forking a process, not a thread.
|
rseq: Introduce restartable sequences system call
Expose a new system call allowing each thread to register one userspace
memory area to be used as an ABI between kernel and user-space for two
purposes: user-space restartable sequences and quick access to read the
current CPU number value from user-space.
* Restartable sequences (per-cpu atomics)
Restartables sequences allow user-space to perform update operations on
per-cpu data without requiring heavy-weight atomic operations.
The restartable critical sections (percpu atomics) work has been started
by Paul Turner and Andrew Hunter. It lets the kernel handle restart of
critical sections. [1] [2] The re-implementation proposed here brings a
few simplifications to the ABI which facilitates porting to other
architectures and speeds up the user-space fast path.
Here are benchmarks of various rseq use-cases.
Test hardware:
arm32: ARMv7 Processor rev 4 (v7l) "Cubietruck", 2-core
x86-64: Intel E5-2630 v3@2.40GHz, 16-core, hyperthreading
The following benchmarks were all performed on a single thread.
* Per-CPU statistic counter increment
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 344.0 31.4 11.0
x86-64: 15.3 2.0 7.7
* LTTng-UST: write event 32-bit header, 32-bit payload into tracer
per-cpu buffer
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 2502.0 2250.0 1.1
x86-64: 117.4 98.0 1.2
* liburcu percpu: lock-unlock pair, dereference, read/compare word
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 751.0 128.5 5.8
x86-64: 53.4 28.6 1.9
* jemalloc memory allocator adapted to use rseq
Using rseq with per-cpu memory pools in jemalloc at Facebook (based on
rseq 2016 implementation):
The production workload response-time has 1-2% gain avg. latency, and
the P99 overall latency drops by 2-3%.
* Reading the current CPU number
Speeding up reading the current CPU number on which the caller thread is
running is done by keeping the current CPU number up do date within the
cpu_id field of the memory area registered by the thread. This is done
by making scheduler preemption set the TIF_NOTIFY_RESUME flag on the
current thread. Upon return to user-space, a notify-resume handler
updates the current CPU value within the registered user-space memory
area. User-space can then read the current CPU number directly from
memory.
Keeping the current cpu id in a memory area shared between kernel and
user-space is an improvement over current mechanisms available to read
the current CPU number, which has the following benefits over
alternative approaches:
- 35x speedup on ARM vs system call through glibc
- 20x speedup on x86 compared to calling glibc, which calls vdso
executing a "lsl" instruction,
- 14x speedup on x86 compared to inlined "lsl" instruction,
- Unlike vdso approaches, this cpu_id value can be read from an inline
assembly, which makes it a useful building block for restartable
sequences.
- The approach of reading the cpu id through memory mapping shared
between kernel and user-space is portable (e.g. ARM), which is not the
case for the lsl-based x86 vdso.
On x86, yet another possible approach would be to use the gs segment
selector to point to user-space per-cpu data. This approach performs
similarly to the cpu id cache, but it has two disadvantages: it is
not portable, and it is incompatible with existing applications already
using the gs segment selector for other purposes.
Benchmarking various approaches for reading the current CPU number:
ARMv7 Processor rev 4 (v7l)
Machine model: Cubietruck
- Baseline (empty loop): 8.4 ns
- Read CPU from rseq cpu_id: 16.7 ns
- Read CPU from rseq cpu_id (lazy register): 19.8 ns
- glibc 2.19-0ubuntu6.6 getcpu: 301.8 ns
- getcpu system call: 234.9 ns
x86-64 Intel(R) Xeon(R) CPU E5-2630 v3 @ 2.40GHz:
- Baseline (empty loop): 0.8 ns
- Read CPU from rseq cpu_id: 0.8 ns
- Read CPU from rseq cpu_id (lazy register): 0.8 ns
- Read using gs segment selector: 0.8 ns
- "lsl" inline assembly: 13.0 ns
- glibc 2.19-0ubuntu6 getcpu: 16.6 ns
- getcpu system call: 53.9 ns
- Speed (benchmark taken on v8 of patchset)
Running 10 runs of hackbench -l 100000 seems to indicate, contrary to
expectations, that enabling CONFIG_RSEQ slightly accelerates the
scheduler:
Configuration: 2 sockets * 8-core Intel(R) Xeon(R) CPU E5-2630 v3 @
2.40GHz (directly on hardware, hyperthreading disabled in BIOS, energy
saving disabled in BIOS, turboboost disabled in BIOS, cpuidle.off=1
kernel parameter), with a Linux v4.6 defconfig+localyesconfig,
restartable sequences series applied.
* CONFIG_RSEQ=n
avg.: 41.37 s
std.dev.: 0.36 s
* CONFIG_RSEQ=y
avg.: 40.46 s
std.dev.: 0.33 s
- Size
On x86-64, between CONFIG_RSEQ=n/y, the text size increase of vmlinux is
567 bytes, and the data size increase of vmlinux is 5696 bytes.
[1] https://lwn.net/Articles/650333/
[2] http://www.linuxplumbersconf.org/2013/ocw/system/presentations/1695/original/LPC%20-%20PerCpu%20Atomics.pdf
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Joel Fernandes <joelaf@google.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Dave Watson <davejwatson@fb.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Chris Lameter <cl@linux.com>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Andrew Hunter <ahh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: "Paul E . McKenney" <paulmck@linux.vnet.ibm.com>
Cc: Paul Turner <pjt@google.com>
Cc: Boqun Feng <boqun.feng@gmail.com>
Cc: Josh Triplett <josh@joshtriplett.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Ben Maurer <bmaurer@fb.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: linux-api@vger.kernel.org
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/20151027235635.16059.11630.stgit@pjt-glaptop.roam.corp.google.com
Link: http://lkml.kernel.org/r/20150624222609.6116.86035.stgit@kitami.mtv.corp.google.com
Link: https://lkml.kernel.org/r/20180602124408.8430-3-mathieu.desnoyers@efficios.com
2018-06-02 12:43:54 +00:00
|
|
|
*/
|
|
|
|
static inline void rseq_fork(struct task_struct *t, unsigned long clone_flags)
|
|
|
|
{
|
|
|
|
if (clone_flags & CLONE_THREAD) {
|
|
|
|
t->rseq = NULL;
|
|
|
|
t->rseq_sig = 0;
|
|
|
|
t->rseq_event_mask = 0;
|
|
|
|
} else {
|
|
|
|
t->rseq = current->rseq;
|
|
|
|
t->rseq_sig = current->rseq_sig;
|
|
|
|
t->rseq_event_mask = current->rseq_event_mask;
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
static inline void rseq_execve(struct task_struct *t)
|
|
|
|
{
|
|
|
|
t->rseq = NULL;
|
|
|
|
t->rseq_sig = 0;
|
|
|
|
t->rseq_event_mask = 0;
|
|
|
|
}
|
|
|
|
|
|
|
|
#else
|
|
|
|
|
|
|
|
static inline void rseq_set_notify_resume(struct task_struct *t)
|
|
|
|
{
|
|
|
|
}
|
2018-06-22 10:45:07 +00:00
|
|
|
static inline void rseq_handle_notify_resume(struct ksignal *ksig,
|
|
|
|
struct pt_regs *regs)
|
rseq: Introduce restartable sequences system call
Expose a new system call allowing each thread to register one userspace
memory area to be used as an ABI between kernel and user-space for two
purposes: user-space restartable sequences and quick access to read the
current CPU number value from user-space.
* Restartable sequences (per-cpu atomics)
Restartables sequences allow user-space to perform update operations on
per-cpu data without requiring heavy-weight atomic operations.
The restartable critical sections (percpu atomics) work has been started
by Paul Turner and Andrew Hunter. It lets the kernel handle restart of
critical sections. [1] [2] The re-implementation proposed here brings a
few simplifications to the ABI which facilitates porting to other
architectures and speeds up the user-space fast path.
Here are benchmarks of various rseq use-cases.
Test hardware:
arm32: ARMv7 Processor rev 4 (v7l) "Cubietruck", 2-core
x86-64: Intel E5-2630 v3@2.40GHz, 16-core, hyperthreading
The following benchmarks were all performed on a single thread.
* Per-CPU statistic counter increment
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 344.0 31.4 11.0
x86-64: 15.3 2.0 7.7
* LTTng-UST: write event 32-bit header, 32-bit payload into tracer
per-cpu buffer
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 2502.0 2250.0 1.1
x86-64: 117.4 98.0 1.2
* liburcu percpu: lock-unlock pair, dereference, read/compare word
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 751.0 128.5 5.8
x86-64: 53.4 28.6 1.9
* jemalloc memory allocator adapted to use rseq
Using rseq with per-cpu memory pools in jemalloc at Facebook (based on
rseq 2016 implementation):
The production workload response-time has 1-2% gain avg. latency, and
the P99 overall latency drops by 2-3%.
* Reading the current CPU number
Speeding up reading the current CPU number on which the caller thread is
running is done by keeping the current CPU number up do date within the
cpu_id field of the memory area registered by the thread. This is done
by making scheduler preemption set the TIF_NOTIFY_RESUME flag on the
current thread. Upon return to user-space, a notify-resume handler
updates the current CPU value within the registered user-space memory
area. User-space can then read the current CPU number directly from
memory.
Keeping the current cpu id in a memory area shared between kernel and
user-space is an improvement over current mechanisms available to read
the current CPU number, which has the following benefits over
alternative approaches:
- 35x speedup on ARM vs system call through glibc
- 20x speedup on x86 compared to calling glibc, which calls vdso
executing a "lsl" instruction,
- 14x speedup on x86 compared to inlined "lsl" instruction,
- Unlike vdso approaches, this cpu_id value can be read from an inline
assembly, which makes it a useful building block for restartable
sequences.
- The approach of reading the cpu id through memory mapping shared
between kernel and user-space is portable (e.g. ARM), which is not the
case for the lsl-based x86 vdso.
On x86, yet another possible approach would be to use the gs segment
selector to point to user-space per-cpu data. This approach performs
similarly to the cpu id cache, but it has two disadvantages: it is
not portable, and it is incompatible with existing applications already
using the gs segment selector for other purposes.
Benchmarking various approaches for reading the current CPU number:
ARMv7 Processor rev 4 (v7l)
Machine model: Cubietruck
- Baseline (empty loop): 8.4 ns
- Read CPU from rseq cpu_id: 16.7 ns
- Read CPU from rseq cpu_id (lazy register): 19.8 ns
- glibc 2.19-0ubuntu6.6 getcpu: 301.8 ns
- getcpu system call: 234.9 ns
x86-64 Intel(R) Xeon(R) CPU E5-2630 v3 @ 2.40GHz:
- Baseline (empty loop): 0.8 ns
- Read CPU from rseq cpu_id: 0.8 ns
- Read CPU from rseq cpu_id (lazy register): 0.8 ns
- Read using gs segment selector: 0.8 ns
- "lsl" inline assembly: 13.0 ns
- glibc 2.19-0ubuntu6 getcpu: 16.6 ns
- getcpu system call: 53.9 ns
- Speed (benchmark taken on v8 of patchset)
Running 10 runs of hackbench -l 100000 seems to indicate, contrary to
expectations, that enabling CONFIG_RSEQ slightly accelerates the
scheduler:
Configuration: 2 sockets * 8-core Intel(R) Xeon(R) CPU E5-2630 v3 @
2.40GHz (directly on hardware, hyperthreading disabled in BIOS, energy
saving disabled in BIOS, turboboost disabled in BIOS, cpuidle.off=1
kernel parameter), with a Linux v4.6 defconfig+localyesconfig,
restartable sequences series applied.
* CONFIG_RSEQ=n
avg.: 41.37 s
std.dev.: 0.36 s
* CONFIG_RSEQ=y
avg.: 40.46 s
std.dev.: 0.33 s
- Size
On x86-64, between CONFIG_RSEQ=n/y, the text size increase of vmlinux is
567 bytes, and the data size increase of vmlinux is 5696 bytes.
[1] https://lwn.net/Articles/650333/
[2] http://www.linuxplumbersconf.org/2013/ocw/system/presentations/1695/original/LPC%20-%20PerCpu%20Atomics.pdf
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Joel Fernandes <joelaf@google.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Dave Watson <davejwatson@fb.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Chris Lameter <cl@linux.com>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Andrew Hunter <ahh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: "Paul E . McKenney" <paulmck@linux.vnet.ibm.com>
Cc: Paul Turner <pjt@google.com>
Cc: Boqun Feng <boqun.feng@gmail.com>
Cc: Josh Triplett <josh@joshtriplett.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Ben Maurer <bmaurer@fb.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: linux-api@vger.kernel.org
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/20151027235635.16059.11630.stgit@pjt-glaptop.roam.corp.google.com
Link: http://lkml.kernel.org/r/20150624222609.6116.86035.stgit@kitami.mtv.corp.google.com
Link: https://lkml.kernel.org/r/20180602124408.8430-3-mathieu.desnoyers@efficios.com
2018-06-02 12:43:54 +00:00
|
|
|
{
|
|
|
|
}
|
2018-06-22 10:45:07 +00:00
|
|
|
static inline void rseq_signal_deliver(struct ksignal *ksig,
|
|
|
|
struct pt_regs *regs)
|
rseq: Introduce restartable sequences system call
Expose a new system call allowing each thread to register one userspace
memory area to be used as an ABI between kernel and user-space for two
purposes: user-space restartable sequences and quick access to read the
current CPU number value from user-space.
* Restartable sequences (per-cpu atomics)
Restartables sequences allow user-space to perform update operations on
per-cpu data without requiring heavy-weight atomic operations.
The restartable critical sections (percpu atomics) work has been started
by Paul Turner and Andrew Hunter. It lets the kernel handle restart of
critical sections. [1] [2] The re-implementation proposed here brings a
few simplifications to the ABI which facilitates porting to other
architectures and speeds up the user-space fast path.
Here are benchmarks of various rseq use-cases.
Test hardware:
arm32: ARMv7 Processor rev 4 (v7l) "Cubietruck", 2-core
x86-64: Intel E5-2630 v3@2.40GHz, 16-core, hyperthreading
The following benchmarks were all performed on a single thread.
* Per-CPU statistic counter increment
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 344.0 31.4 11.0
x86-64: 15.3 2.0 7.7
* LTTng-UST: write event 32-bit header, 32-bit payload into tracer
per-cpu buffer
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 2502.0 2250.0 1.1
x86-64: 117.4 98.0 1.2
* liburcu percpu: lock-unlock pair, dereference, read/compare word
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 751.0 128.5 5.8
x86-64: 53.4 28.6 1.9
* jemalloc memory allocator adapted to use rseq
Using rseq with per-cpu memory pools in jemalloc at Facebook (based on
rseq 2016 implementation):
The production workload response-time has 1-2% gain avg. latency, and
the P99 overall latency drops by 2-3%.
* Reading the current CPU number
Speeding up reading the current CPU number on which the caller thread is
running is done by keeping the current CPU number up do date within the
cpu_id field of the memory area registered by the thread. This is done
by making scheduler preemption set the TIF_NOTIFY_RESUME flag on the
current thread. Upon return to user-space, a notify-resume handler
updates the current CPU value within the registered user-space memory
area. User-space can then read the current CPU number directly from
memory.
Keeping the current cpu id in a memory area shared between kernel and
user-space is an improvement over current mechanisms available to read
the current CPU number, which has the following benefits over
alternative approaches:
- 35x speedup on ARM vs system call through glibc
- 20x speedup on x86 compared to calling glibc, which calls vdso
executing a "lsl" instruction,
- 14x speedup on x86 compared to inlined "lsl" instruction,
- Unlike vdso approaches, this cpu_id value can be read from an inline
assembly, which makes it a useful building block for restartable
sequences.
- The approach of reading the cpu id through memory mapping shared
between kernel and user-space is portable (e.g. ARM), which is not the
case for the lsl-based x86 vdso.
On x86, yet another possible approach would be to use the gs segment
selector to point to user-space per-cpu data. This approach performs
similarly to the cpu id cache, but it has two disadvantages: it is
not portable, and it is incompatible with existing applications already
using the gs segment selector for other purposes.
Benchmarking various approaches for reading the current CPU number:
ARMv7 Processor rev 4 (v7l)
Machine model: Cubietruck
- Baseline (empty loop): 8.4 ns
- Read CPU from rseq cpu_id: 16.7 ns
- Read CPU from rseq cpu_id (lazy register): 19.8 ns
- glibc 2.19-0ubuntu6.6 getcpu: 301.8 ns
- getcpu system call: 234.9 ns
x86-64 Intel(R) Xeon(R) CPU E5-2630 v3 @ 2.40GHz:
- Baseline (empty loop): 0.8 ns
- Read CPU from rseq cpu_id: 0.8 ns
- Read CPU from rseq cpu_id (lazy register): 0.8 ns
- Read using gs segment selector: 0.8 ns
- "lsl" inline assembly: 13.0 ns
- glibc 2.19-0ubuntu6 getcpu: 16.6 ns
- getcpu system call: 53.9 ns
- Speed (benchmark taken on v8 of patchset)
Running 10 runs of hackbench -l 100000 seems to indicate, contrary to
expectations, that enabling CONFIG_RSEQ slightly accelerates the
scheduler:
Configuration: 2 sockets * 8-core Intel(R) Xeon(R) CPU E5-2630 v3 @
2.40GHz (directly on hardware, hyperthreading disabled in BIOS, energy
saving disabled in BIOS, turboboost disabled in BIOS, cpuidle.off=1
kernel parameter), with a Linux v4.6 defconfig+localyesconfig,
restartable sequences series applied.
* CONFIG_RSEQ=n
avg.: 41.37 s
std.dev.: 0.36 s
* CONFIG_RSEQ=y
avg.: 40.46 s
std.dev.: 0.33 s
- Size
On x86-64, between CONFIG_RSEQ=n/y, the text size increase of vmlinux is
567 bytes, and the data size increase of vmlinux is 5696 bytes.
[1] https://lwn.net/Articles/650333/
[2] http://www.linuxplumbersconf.org/2013/ocw/system/presentations/1695/original/LPC%20-%20PerCpu%20Atomics.pdf
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Joel Fernandes <joelaf@google.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Dave Watson <davejwatson@fb.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Chris Lameter <cl@linux.com>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Andrew Hunter <ahh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: "Paul E . McKenney" <paulmck@linux.vnet.ibm.com>
Cc: Paul Turner <pjt@google.com>
Cc: Boqun Feng <boqun.feng@gmail.com>
Cc: Josh Triplett <josh@joshtriplett.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Ben Maurer <bmaurer@fb.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: linux-api@vger.kernel.org
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/20151027235635.16059.11630.stgit@pjt-glaptop.roam.corp.google.com
Link: http://lkml.kernel.org/r/20150624222609.6116.86035.stgit@kitami.mtv.corp.google.com
Link: https://lkml.kernel.org/r/20180602124408.8430-3-mathieu.desnoyers@efficios.com
2018-06-02 12:43:54 +00:00
|
|
|
{
|
|
|
|
}
|
|
|
|
static inline void rseq_preempt(struct task_struct *t)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
static inline void rseq_migrate(struct task_struct *t)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
static inline void rseq_fork(struct task_struct *t, unsigned long clone_flags)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
static inline void rseq_execve(struct task_struct *t)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
|
|
|
|
#endif
|
|
|
|
|
2019-01-08 17:23:56 +00:00
|
|
|
void __exit_umh(struct task_struct *tsk);
|
|
|
|
|
|
|
|
static inline void exit_umh(struct task_struct *tsk)
|
|
|
|
{
|
|
|
|
if (unlikely(tsk->flags & PF_UMH))
|
|
|
|
__exit_umh(tsk);
|
|
|
|
}
|
|
|
|
|
rseq: Introduce restartable sequences system call
Expose a new system call allowing each thread to register one userspace
memory area to be used as an ABI between kernel and user-space for two
purposes: user-space restartable sequences and quick access to read the
current CPU number value from user-space.
* Restartable sequences (per-cpu atomics)
Restartables sequences allow user-space to perform update operations on
per-cpu data without requiring heavy-weight atomic operations.
The restartable critical sections (percpu atomics) work has been started
by Paul Turner and Andrew Hunter. It lets the kernel handle restart of
critical sections. [1] [2] The re-implementation proposed here brings a
few simplifications to the ABI which facilitates porting to other
architectures and speeds up the user-space fast path.
Here are benchmarks of various rseq use-cases.
Test hardware:
arm32: ARMv7 Processor rev 4 (v7l) "Cubietruck", 2-core
x86-64: Intel E5-2630 v3@2.40GHz, 16-core, hyperthreading
The following benchmarks were all performed on a single thread.
* Per-CPU statistic counter increment
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 344.0 31.4 11.0
x86-64: 15.3 2.0 7.7
* LTTng-UST: write event 32-bit header, 32-bit payload into tracer
per-cpu buffer
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 2502.0 2250.0 1.1
x86-64: 117.4 98.0 1.2
* liburcu percpu: lock-unlock pair, dereference, read/compare word
getcpu+atomic (ns/op) rseq (ns/op) speedup
arm32: 751.0 128.5 5.8
x86-64: 53.4 28.6 1.9
* jemalloc memory allocator adapted to use rseq
Using rseq with per-cpu memory pools in jemalloc at Facebook (based on
rseq 2016 implementation):
The production workload response-time has 1-2% gain avg. latency, and
the P99 overall latency drops by 2-3%.
* Reading the current CPU number
Speeding up reading the current CPU number on which the caller thread is
running is done by keeping the current CPU number up do date within the
cpu_id field of the memory area registered by the thread. This is done
by making scheduler preemption set the TIF_NOTIFY_RESUME flag on the
current thread. Upon return to user-space, a notify-resume handler
updates the current CPU value within the registered user-space memory
area. User-space can then read the current CPU number directly from
memory.
Keeping the current cpu id in a memory area shared between kernel and
user-space is an improvement over current mechanisms available to read
the current CPU number, which has the following benefits over
alternative approaches:
- 35x speedup on ARM vs system call through glibc
- 20x speedup on x86 compared to calling glibc, which calls vdso
executing a "lsl" instruction,
- 14x speedup on x86 compared to inlined "lsl" instruction,
- Unlike vdso approaches, this cpu_id value can be read from an inline
assembly, which makes it a useful building block for restartable
sequences.
- The approach of reading the cpu id through memory mapping shared
between kernel and user-space is portable (e.g. ARM), which is not the
case for the lsl-based x86 vdso.
On x86, yet another possible approach would be to use the gs segment
selector to point to user-space per-cpu data. This approach performs
similarly to the cpu id cache, but it has two disadvantages: it is
not portable, and it is incompatible with existing applications already
using the gs segment selector for other purposes.
Benchmarking various approaches for reading the current CPU number:
ARMv7 Processor rev 4 (v7l)
Machine model: Cubietruck
- Baseline (empty loop): 8.4 ns
- Read CPU from rseq cpu_id: 16.7 ns
- Read CPU from rseq cpu_id (lazy register): 19.8 ns
- glibc 2.19-0ubuntu6.6 getcpu: 301.8 ns
- getcpu system call: 234.9 ns
x86-64 Intel(R) Xeon(R) CPU E5-2630 v3 @ 2.40GHz:
- Baseline (empty loop): 0.8 ns
- Read CPU from rseq cpu_id: 0.8 ns
- Read CPU from rseq cpu_id (lazy register): 0.8 ns
- Read using gs segment selector: 0.8 ns
- "lsl" inline assembly: 13.0 ns
- glibc 2.19-0ubuntu6 getcpu: 16.6 ns
- getcpu system call: 53.9 ns
- Speed (benchmark taken on v8 of patchset)
Running 10 runs of hackbench -l 100000 seems to indicate, contrary to
expectations, that enabling CONFIG_RSEQ slightly accelerates the
scheduler:
Configuration: 2 sockets * 8-core Intel(R) Xeon(R) CPU E5-2630 v3 @
2.40GHz (directly on hardware, hyperthreading disabled in BIOS, energy
saving disabled in BIOS, turboboost disabled in BIOS, cpuidle.off=1
kernel parameter), with a Linux v4.6 defconfig+localyesconfig,
restartable sequences series applied.
* CONFIG_RSEQ=n
avg.: 41.37 s
std.dev.: 0.36 s
* CONFIG_RSEQ=y
avg.: 40.46 s
std.dev.: 0.33 s
- Size
On x86-64, between CONFIG_RSEQ=n/y, the text size increase of vmlinux is
567 bytes, and the data size increase of vmlinux is 5696 bytes.
[1] https://lwn.net/Articles/650333/
[2] http://www.linuxplumbersconf.org/2013/ocw/system/presentations/1695/original/LPC%20-%20PerCpu%20Atomics.pdf
Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Joel Fernandes <joelaf@google.com>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Dave Watson <davejwatson@fb.com>
Cc: Will Deacon <will.deacon@arm.com>
Cc: Andi Kleen <andi@firstfloor.org>
Cc: "H . Peter Anvin" <hpa@zytor.com>
Cc: Chris Lameter <cl@linux.com>
Cc: Russell King <linux@arm.linux.org.uk>
Cc: Andrew Hunter <ahh@google.com>
Cc: Michael Kerrisk <mtk.manpages@gmail.com>
Cc: "Paul E . McKenney" <paulmck@linux.vnet.ibm.com>
Cc: Paul Turner <pjt@google.com>
Cc: Boqun Feng <boqun.feng@gmail.com>
Cc: Josh Triplett <josh@joshtriplett.org>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Ben Maurer <bmaurer@fb.com>
Cc: Alexander Viro <viro@zeniv.linux.org.uk>
Cc: linux-api@vger.kernel.org
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Link: http://lkml.kernel.org/r/20151027235635.16059.11630.stgit@pjt-glaptop.roam.corp.google.com
Link: http://lkml.kernel.org/r/20150624222609.6116.86035.stgit@kitami.mtv.corp.google.com
Link: https://lkml.kernel.org/r/20180602124408.8430-3-mathieu.desnoyers@efficios.com
2018-06-02 12:43:54 +00:00
|
|
|
#ifdef CONFIG_DEBUG_RSEQ
|
|
|
|
|
|
|
|
void rseq_syscall(struct pt_regs *regs);
|
|
|
|
|
|
|
|
#else
|
|
|
|
|
|
|
|
static inline void rseq_syscall(struct pt_regs *regs)
|
|
|
|
{
|
|
|
|
}
|
|
|
|
|
|
|
|
#endif
|
|
|
|
|
2019-06-04 11:14:55 +00:00
|
|
|
const struct sched_avg *sched_trace_cfs_rq_avg(struct cfs_rq *cfs_rq);
|
|
|
|
char *sched_trace_cfs_rq_path(struct cfs_rq *cfs_rq, char *str, int len);
|
|
|
|
int sched_trace_cfs_rq_cpu(struct cfs_rq *cfs_rq);
|
|
|
|
|
|
|
|
const struct sched_avg *sched_trace_rq_avg_rt(struct rq *rq);
|
|
|
|
const struct sched_avg *sched_trace_rq_avg_dl(struct rq *rq);
|
|
|
|
const struct sched_avg *sched_trace_rq_avg_irq(struct rq *rq);
|
|
|
|
|
|
|
|
int sched_trace_rq_cpu(struct rq *rq);
|
|
|
|
|
|
|
|
const struct cpumask *sched_trace_rd_span(struct root_domain *rd);
|
|
|
|
|
2005-04-16 22:20:36 +00:00
|
|
|
#endif
|