Current refcounting for modules (done if CONFIG_MODULE_UNLOAD=y) is
using a lot of memory.
Each 'struct module' contains an [NR_CPUS] array of full cache lines.
This patch uses existing infrastructure (percpu_modalloc() &
percpu_modfree()) to allocate percpu space for the refcount storage.
Instead of wasting NR_CPUS*128 bytes (on i386), we now use
nr_cpu_ids*sizeof(local_t) bytes.
On a typical distro, where NR_CPUS=8, shiping 2000 modules, we reduce
size of module files by about 2 Mbytes. (1Kb per module)
Instead of having all refcounters in the same memory node - with TLB misses
because of vmalloc() - this new implementation permits to have better
NUMA properties, since each CPU will use storage on its preferred node,
thanks to percpu storage.
Signed-off-by: Eric Dumazet <dada1@cosmosbay.com>
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This essentially reverts commit 8edb08caf6.
It downgraded our mmap semaphore to a read-lock while mlocking pages, in
order to allow other threads (and external accesses like "ps" et al) to
walk the vma lists and take page faults etc. Which is a nice idea, but
the implementation does not work.
Because we cannot upgrade the lock back to a write lock without
releasing the mmap semaphore, the code had to release the lock entirely
and then re-take it as a writelock. However, that meant that the caller
possibly lost the vma chain that it was following, since now another
thread could come in and mmap/munmap the range.
The code tried to work around that by just looking up the vma again and
erroring out if that happened, but quite frankly, that was just a buggy
hack that doesn't actually protect against anything (the other thread
could just have replaced the vma with another one instead of totally
unmapping it).
The only way to downgrade to a read map _reliably_ is to do it at the
end, which is likely the right thing to do: do all the 'vma' operations
with the write-lock held, then downgrade to a read after completing them
all, and then do the "populate the newly mlocked regions" while holding
just the read lock. And then just drop the read-lock and return to user
space.
The (perhaps somewhat simpler) alternative is to just make all the
callers of mlock_vma_pages_range() know that the mmap lock got dropped,
and just re-grab the mmap semaphore if it needs to mlock more than one
vma region.
So we can do this "downgrade mmap sem while populating mlocked regions"
thing right, but the way it was done here was absolutely not correct.
Thus the revert, in the expectation that we will do it all correctly
some day.
Cc: Lee Schermerhorn <lee.schermerhorn@hp.com>
Cc: Rik van Riel <riel@redhat.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: stable@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* 'core-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
generic-ipi: use per cpu data for single cpu ipi calls
cpumask: convert lib/smp_processor_id to new cpumask ops
signals, debug: fix BUG: using smp_processor_id() in preemptible code in print_fatal_signal()
* 'timers-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
hrtimer: prevent negative expiry value after clock_was_set()
hrtimers: allow the hot-unplugging of all cpus
hrtimers: increase clock min delta threshold while interrupt hanging
* 'tracing-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
x86, ds, bts: cleanup/fix DS configuration
ring-buffer: reset timestamps when ring buffer is reset
trace: set max latency variable to zero on default
trace: stop all recording to ring buffer on ftrace_dump
trace: print ftrace_dump at KERN_EMERG log level
ring_buffer: reset write when reserve buffer fail
tracing/function-graph-tracer: fix a regression while suspend to disk
ring-buffer: fix alignment problem
* 'x86-fixes-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/linux-2.6-tip:
x86 setup: fix asm constraints in vesa_store_edid
xen: make sysfs files behave as their names suggest
x86: tone down mtrr_trim_uncached_memory() warning
x86: correct the CPUID pattern for MSR_IA32_MISC_ENABLE availability
* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/lrg/voltage-2.6:
leds: Fix bounds checking of wm8350->pmic.led
regulator: move bq24022 init back to module_init instead of subsys_initcall
Fix an off-by-two memory error in console selection.
The loop below goes from sel_start to sel_end (inclusive), so it writes
one more character. This one more character was added to the allocated
size (+1), but it was not multiplied by an UTF-8 multiplier.
This patch fixes a memory corruption when UTF-8 console is used and the
user selects a few characters, all of them 3-byte in UTF-8 (for example
a frame line).
When memory redzones are enabled, a redzone corruption is reported.
When they are not enabled, trashing of random memory occurs.
Signed-off-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
* git://git.kernel.org/pub/scm/linux/kernel/git/davem/net-2.6:
igb: fix link reporting when using sgmii
igb: prevent skb_over panic w/ mtu smaller than 1K
igb: Fix DCA errors and do not use context index for 82576
ipv6: compile fix for ip6mr.c
packet: Avoid lock_sock in mmap handler
sfc: Replace stats_enabled flag with a disable count
sfc: SFX7101/SFT9001: Fix AN advertisements
sfc: SFT9001: Always enable XNP exchange on SFT9001 rev B
sfc: Update board info for hardware monitor on SFN4111T-R5 and later
sfc: Test for PHYXS faults whenever we cannot test link state bits
sfc: Reinitialise the PHY completely in case of a PHY or NIC reset
sfc: Fix post-reset MAC selection
sfc: SFN4111T: Fix GPIO sharing between I2C and FLASH_CFG_1
sfc: SFT9001: Fix speed reporting in 1G PHY loopback
sfc: SFX7101: Remove workaround for bad link training
sfc: SFT9001: Enable robust link training
sky2: fix hard hang with netconsoling and iface going up
The mmap_region() code would temporarily set the VM_ACCOUNT flag for
anonymous shared mappings just to inform shmem_zero_setup() that it
should enable accounting for the resulting shm object. It would then
clear the flag after calling ->mmap (for the /dev/zero case) or doing
shmem_zero_setup() (for the MAP_ANON case).
This just resulted in vma merge issues, but also made for just
unnecessary confusion. Use the already-existing VM_NORESERVE flag for
this instead, and let shmem_{zero|file}_setup() just figure it out from
that.
This also happens to make it obvious that the new DRI2 GEM layer uses a
non-reserving backing store for its object allocation - which is quite
possibly not intentional. But since I didn't want to change semantics
in this patch, I left it alone, and just updated the caller to use the
new flag semantics.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
When using sgmii the link was not being properly passed up to the driver
from the underlying link management functions. This change corrects it so
that get_link_status is cleared when a link has been found.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
A panic has been observed with frame sizes smaller than 1K. This has been
root caused to the hardware spanning larger frames across multiple buffers
and then reporting the original frame size in the first descriptor. To
prevent this we can enable set the LPE bit which in turn will restrict
packet sizes to those set in the RLPML register.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
82576 was being incorrectly flagged as needing a context index. It does not as
each ring has it's own table of 2 contexts.
Driver was registering after registering the driver instead of the other way around.
Signed-off-by: Alexander Duyck <alexander.h.duyck@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
net/ipv6/ip6mr.c: In function 'pim6_rcv':
net/ipv6/ip6mr.c:368: error: implicit declaration of function 'csum_ipv6_magic'
Signed-off-by: Dave Jones <davej@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Impact: fix potential miscompile (currently believed non-manifest)
As the comment explains, the VBE DDC call can clobber any register.
Tell the compiler about that fact.
Signed-off-by: Andreas Schwab <schwab@suse.de>
Signed-off-by: H. Peter Anvin <hpa@linux.intel.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
As the mmap handler gets called under mmap_sem, and we may grab
mmap_sem elsewhere under the socket lock to access user data, we
should avoid grabbing the socket lock in the mmap handler.
Since the only thing we care about in the mmap handler is for
pg_vec* to be invariant, i.e., to exclude packet_set_ring, we
can achieve this by simply using a new mutex.
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Tested-by: Martin MOKREJŠ <mmokrejs@ribosome.natur.cuni.cz>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently we use a spin-lock to serialise statistics fetches and also
to inhibit them for short periods of time, plus a flag to
enable/disable statistics fetches for longer periods of time, during
online reset. This was apparently insufficient to deal with the several
reasons for stats being disabled.
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
All 10Xpress PHYs require autonegotiation all the time; enforce this
in the set_settings() method and do not treat it as a workaround.
Remove claimed support for 100M HD mode since it is not supported by
current firmware.
Do not set speed override bits when AN is enabled, and do not use
register 1.49192 for AN configuration as it can override what we set
elsewhere.
Always set the AN selector bits to 1 (802.3).
Fix confusion between Next Page and Extended Next Page.
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This workaround is not specific to rev A.
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Depending on the loopback mode, there may be no pertinent link state
bits. In this case we test the PHYXS RX fault bit instead. Make
sure to do this in all cases where there are no link state bits.
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In particular, set pause advertising bits properly.
A PHY reset is not necessary to recover from the register self-test,
so use a "invisible" reset there instead.
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Modify falcon_switch_mac() to always set NIC_STAT_REG, even if the the
MAC is the same as it was before. This ensures that the value is
correct after an online reset.
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Change sfn4111t_reset() to change only GPIO output enables so that it
doesn't break subsequent I2C operations.
Update comments to explain exactly what we're doing.
Add a short sleep to make sure the FLASH_CFG_1 value is latched before
any subsequent I2C operations.
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Instead of disabling AN in loopback, just prevent restarting AN and
override the speed in sft9001_get_settings().
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Early versions of the SFX7101 firmware could complete link training in
a state where it would not adequately cancel noise (Solarflare bug
10750). We previously worked around this by resetting the PHY after
seeing many Ethernet CRC errors. This workaround is unsafe since it
takes no account of the interval between errors; it also appears to
be unnecessary with production firmware. Therefore remove it.
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Enable a firmware option that appears to be necessary for reliable
operation.
Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Printing anything over netconsole before hw is up and running is,
of course, not going to work.
Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com>
Acked-by: Stephen Hemminger <shemminger@vyatta.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This workaround was needed when regulator/ was not linked before both
power/ and usb/otg/ in drivers/Makefile. Now that it is even linked
before mfd/, this patch makes sure that bq24022 isn't probed before the
GPIO expander is set up.
Signed-off-by: Philipp Zabel <philipp.zabel@gmail.com>
Signed-off-by: Liam Girdwood <lrg@slimlogic.co.uk>
Impact: prevent false positive WARN_ON() in clockevents_program_event()
clock_was_set() changes the base->offset of CLOCK_REALTIME and
enforces the reprogramming of the clockevent device to expire timers
which are based on CLOCK_REALTIME. If the clock change is large enough
then the subtraction of the timer expiry value and base->offset can
become negative which triggers the warning in
clockevents_program_event().
Check the subtraction result and set a negative value to 0.
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Impact: fix CPU hotplug hang on Power6 testbox
On architectures that support offlining all cpus (at least powerpc/pseries),
hot-unpluging the tick_do_timer_cpu can result in a system hang.
This comes from the fact that if the cpu going down happens to be the
cpu doing the tick, then as the tick_do_timer_cpu handover happens after the
cpu is dead (via the CPU_DEAD notification), we're left without ticks,
jiffies are frozen and any task relying on timers (msleep, ...) is stuck.
That's particularly the case for the cpu looping in __cpu_die() waiting
for the dying cpu to be dead.
This patch addresses this by having the tick_do_timer_cpu handover happen
earlier during the CPU_DYING notification. For this, a new clockevent
notification type is introduced (CLOCK_EVT_NOTIFY_CPU_DYING) which is triggered
in hrtimer_cpu_notify().
Signed-off-by: Sebastien Dugue <sebastien.dugue@bull.net>
Cc: <stable@kernel.org>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
Impact: avoid timer IRQ hanging slow systems
While using the function graph tracer on a virtualized system, the
hrtimer_interrupt can hang the system on an infinite loop.
This can be caused in several situations:
- the hardware is very slow and HZ is set too high
- something intrusive is slowing the system down (tracing under emulation)
... and the next clock events to program are always before the current time.
This patch implements a reasonable compromise: if such a situation is
detected, we share the CPUs time in 1/4 to process the hrtimer interrupts.
This is enough to let the system running without serious starvation.
It has been successfully tested under VirtualBox with 1000 HZ and 100 HZ
with function graph tracer launched. On both cases, the clock events were
increased until about 25 ms periodic ticks, which means 40 HZ.
So we change a hard to debug hang into a warning message and a system that
still manages to limp along.
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Signed-off-by: Ingo Molnar <mingo@elte.hu>
In Linus' current -git the cpumask member is now a pointer.
Signed-off-by: Manuel Lauss <mano@roarinelk.homelinux.net>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
This kernel symbol provides a way for drivers to switch on alternate
function for a certain GPIO pin. Turning it off is done implicitly when
changing the GPIO direction, as that would be fixed when using the given
pin als alternate function.
Signed-off-by: Phil Sutter <n0-1@freewrt.org>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Remove the {set,get}_434_reg() prototypes, as the functions have been
removed. Also move the prototypes for {get,set}_latch_u5() to the correct
place.
Signed-off-by: Phil Sutter <n0-1@freewrt.org>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
As rb532_dev3_ctl_res is not used by any platform device, it can be dropped
when not used for holding the physical address of the device 3 controller.
Also a size of one byte should suffice when ioremapping the physical
address mentioned above, as only a single byte is being read from and
written to it.
Signed-off-by: Phil Sutter <n0-1@freewrt.org>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
These kernel symbols are unused. Also, since dev3 init has been moved to
devices.c, set_434_reg() breaks compiling as it uses dev3.
Signed-off-by: Phil Sutter <n0-1@freewrt.org>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
This code doesn't belong to gpio.c, as it's completely unrelated to
GPIO. As dev1 and dev2 init code is in devices.c, it seems to be a more
adequate place.
Signed-off-by: Phil Sutter <n0-1@freewrt.org>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
The data to be written is just a byte, so use writeb instead of writel.
Also, dev3.base contains the address, not the data so referencing here
is wrong.
Signed-off-by: Phil Sutter <n0-1@freewrt.org>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
This register just contains the address of the actual resource, so
initialisation has to be the same as cf_slot0_res and nand_slot0_res.
Signed-off-by: Phil Sutter <n0-1@freewrt.org>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
As the korina ethernet driver uses platform_get_drvdata() to extract the
driver specific data from the platform device, driver_data has to be
used here.
Signed-off-by: Phil Sutter <n0-1@freewrt.org>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Auto-detection works just fine, so use it instead of specifying the type
manually. Also define a platform device for the uart, as suggested by
David Daney.
Signed-off-by: Phil Sutter <n0-1@freewrt.org>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
As the pata-rb532-cf driver calls gpio_direction_input(), the calls to
rb532_gpio_set_func() and rb532_gpio_direction_input() are not needed since
the alternate function is automatically being disabled when changing the
GPIO pin direction.
The later two calls to rb532_gpio_set_{ilevel,istat}() are implicitly being
done by the IRQ initialisation of pata-rb532-cf.
Signed-off-by: Phil Sutter <n0-1@freewrt.org>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
When a driver calls gpio_set_direction_{input,output}(), it obviously
doesn't want the alternate function for that pin to be active (as the
direction would not matter in that case). This patch ensures alternate
function is disabled when the direction is being changed.
Signed-off-by: Phil Sutter <n0-1@freewrt.org>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
Interrupt Group 4 mapps the GPIO pins enabled as interrupt sources;
add defines to make this clear when addressing them later in code.
The mapped GPIOs support triggering on either level high or low. To
achieve this, the set_type() function calls rb532_gpio_set_ilevel() for
interrupts of the above mentioned group.
As there is no way to alter the triggering characteristics of the other
interrupts, accept level triggering on status high only. (This is just a
guess; but as the system boots fine and interrupt-driven devices (e.g.
serial console) work with no implications, it seems to be right.)
To clear a GPIO mapped IRQ, the source has to be cleared (i.e., the
interrupt status bit of the corresponding GPIO pin). This is done inside
rb532_disable_irq().
After applying these changes I could undo most of my former "fixes" to
pata-rb532-cf. Particularly all interrupt handling can be done
generically via set_irq_type() as it was before.
Signed-off-by: Phil Sutter <n0-1@freewrt.org>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
The code is rather based on trial-and-error than knowledge. Verified Via
Rhine functionality in PIO as well as MMIO mode.
[Looks sane -- Ralf]
Signed-off-by: Phil Sutter <n0-1@freewrt.org>
Tested-by: Florian Fainelli <florian@openwrt.org>
Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
