Allocate space for storing ZA on first access to SME and use that to save
and restore ZA state when context switching. We do this by using the vector
form of the LDR and STR ZA instructions, these do not require streaming
mode and have implementation recommendations that they avoid contention
issues in shared SMCU implementations.
Since ZA is architecturally guaranteed to be zeroed when enabled we do not
need to explicitly zero ZA, either we will be restoring from a saved copy
or trapping on first use of SME so we know that ZA must be disabled.
Signed-off-by: Mark Brown <broonie@kernel.org>
Reviewed-by: Catalin Marinas <catalin.marinas@arm.com>
Link: https://lore.kernel.org/r/20220419112247.711548-16-broonie@kernel.org
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
When in streaming mode we need to save and restore the streaming mode
SVE register state rather than the regular SVE register state. This uses
the streaming mode vector length and omits FFR but is otherwise identical,
if TIF_SVE is enabled when we are in streaming mode then streaming mode
takes precedence.
This does not handle use of streaming SVE state with KVM, ptrace or
signals. This will be updated in further patches.
Signed-off-by: Mark Brown <broonie@kernel.org>
Reviewed-by: Catalin Marinas <catalin.marinas@arm.com>
Link: https://lore.kernel.org/r/20220419112247.711548-15-broonie@kernel.org
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
The vector lengths used for SME are controlled through a similar set of
registers to those for SVE and enumerated using a similar algorithm with
some slight differences due to the fact that unlike SVE there are no
restrictions on which combinations of vector lengths can be supported
nor any mandatory vector lengths which must be implemented. Add a new
vector type and implement support for enumerating it.
One slightly awkward feature is that we need to read the current vector
length using a different instruction (or enter streaming mode which
would have the same issue and be higher cost). Rather than add an ops
structure we add special cases directly in the otherwise generic
vec_probe_vqs() function, this is a bit inelegant but it's the only
place where this is an issue.
Signed-off-by: Mark Brown <broonie@kernel.org>
Reviewed-by: Catalin Marinas <catalin.marinas@arm.com>
Link: https://lore.kernel.org/r/20220419112247.711548-10-broonie@kernel.org
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Currently when restoring the SVE state we supply the SVE vector length
as an argument to sve_load_state() and the underlying macros. This becomes
inconvenient with the addition of SME since we may need to restore any
combination of SVE and SME vector lengths, and we already separately
restore the vector length in the KVM code. We don't need to know the vector
length during the actual register load since the SME load instructions can
index into the data array for us.
Refactor the interface so we explicitly set the vector length separately
to restoring the SVE registers in preparation for adding SME support, no
functional change should be involved.
Signed-off-by: Mark Brown <broonie@kernel.org>
Link: https://lore.kernel.org/r/20211019172247.3045838-9-broonie@kernel.org
Signed-off-by: Will Deacon <will@kernel.org>
SME introduces streaming SVE mode in which FFR is not present and the
instructions for accessing it UNDEF. In preparation for handling this
update the low level SVE state access functions to take a flag specifying
if FFR should be handled. When saving the register state we store a zero
for FFR to guard against uninitialized data being read. No behaviour change
should be introduced by this patch.
Signed-off-by: Mark Brown <broonie@kernel.org>
Link: https://lore.kernel.org/r/20211019172247.3045838-5-broonie@kernel.org
Signed-off-by: Will Deacon <will@kernel.org>
Following optimisations of the SVE register handling we no longer load the
SVE state from a saved copy of the FPSIMD registers, we convert directly
in registers or from one saved state to another. Remove the function so we
don't need to update it during further refactoring.
Signed-off-by: Mark Brown <broonie@kernel.org>
Link: https://lore.kernel.org/r/20211019172247.3045838-3-broonie@kernel.org
Signed-off-by: Will Deacon <will@kernel.org>
The use of macros for the actual function bodies means legibility is always
going to be a bit of a challenge, especially while we can't rely on SVE
support in the toolchain, but this helps a little.
Signed-off-by: Mark Brown <broonie@kernel.org>
Link: https://lore.kernel.org/r/20210812201143.35578-1-broonie@kernel.org
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
When the SVE vector length is 128 bits then there are no bits in the Z
registers which are not shared with the V registers so we can skip them
when zeroing state not shared with FPSIMD, this results in a minor
performance improvement.
Signed-off-by: Mark Brown <broonie@kernel.org>
Reviewed-by: Dave Martin <Dave.Martin@arm.com>
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Link: https://lore.kernel.org/r/20210512151131.27877-4-broonie@kernel.org
Signed-off-by: Will Deacon <will@kernel.org>
This makes the code a bit clearer and as a result we can also make the
indentation more normal, there is no change to the generated code.
Signed-off-by: Mark Brown <broonie@kernel.org>
Reviewed-by: Dave Martin <Dave.Martin@arm.com>
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Link: https://lore.kernel.org/r/20210512151131.27877-3-broonie@kernel.org
Signed-off-by: Will Deacon <will@kernel.org>
Trivial refactoring to support further work, no change to generated code.
Signed-off-by: Mark Brown <broonie@kernel.org>
Reviewed-by: Dave Martin <Dave.Martin@arm.com>
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Link: https://lore.kernel.org/r/20210512151131.27877-2-broonie@kernel.org
Signed-off-by: Will Deacon <will@kernel.org>
When we enable SVE usage in userspace after taking a SVE access trap we
need to ensure that the portions of the register state that are not
shared with the FPSIMD registers are zeroed. Currently we do this by
forcing the FPSIMD registers to be saved to the task struct and converting
them there. This is wasteful in the common case where the task state is
loaded into the registers and we will immediately return to userspace
since we can initialise the SVE state directly in registers instead of
accessing multiple copies of the register state in memory.
Instead in that common case do the conversion in the registers and
update the task metadata so that we can return to userspace without
spilling the register state to memory unless there is some other reason
to do so.
Signed-off-by: Mark Brown <broonie@kernel.org>
Link: https://lore.kernel.org/r/20210312190313.24598-1-broonie@kernel.org
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
In a follow-up patch, we may save the FPSIMD rather than the full SVE
state when the state has to be zeroed on return to userspace (e.g
during a syscall).
Introduce an helper to load SVE vectors from FPSIMD state and zero the rest
of SVE registers.
Signed-off-by: Julien Grall <julien.grall@arm.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Reviewed-by: Dave Martin <Dave.Martin@arm.com>
Link: https://lore.kernel.org/r/20200828181155.17745-7-broonie@kernel.org
Signed-off-by: Will Deacon <will@kernel.org>
Introduce a new helper that will zero all SVE registers but the first
128-bits of each vector. This will be used by subsequent patches to
avoid costly store/maipulate/reload sequences in places like do_sve_acc().
Signed-off-by: Julien Grall <julien.grall@arm.com>
Signed-off-by: Mark Brown <broonie@kernel.org>
Reviewed-by: Dave Martin <Dave.Martin@arm.com>
Link: https://lore.kernel.org/r/20200828181155.17745-6-broonie@kernel.org
Signed-off-by: Will Deacon <will@kernel.org>
In an effort to clarify and simplify the annotation of assembly functions
in the kernel new macros have been introduced. These replace ENTRY and
ENDPROC and also add a new annotation for static functions which previously
had no ENTRY equivalent. Update the annotations in the core kernel code to
the new macros.
Signed-off-by: Mark Brown <broonie@kernel.org>
Acked-by: Mark Rutland <mark.rutland@arm.com>
Link: https://lore.kernel.org/r/20200501115430.37315-3-broonie@kernel.org
Signed-off-by: Will Deacon <will@kernel.org>
Based on 1 normalized pattern(s):
this program is free software you can redistribute it and or modify
it under the terms of the gnu general public license version 2 as
published by the free software foundation this program is
distributed in the hope that it will be useful but without any
warranty without even the implied warranty of merchantability or
fitness for a particular purpose see the gnu general public license
for more details you should have received a copy of the gnu general
public license along with this program if not see http www gnu org
licenses
extracted by the scancode license scanner the SPDX license identifier
GPL-2.0-only
has been chosen to replace the boilerplate/reference in 503 file(s).
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Reviewed-by: Alexios Zavras <alexios.zavras@intel.com>
Reviewed-by: Allison Randal <allison@lohutok.net>
Reviewed-by: Enrico Weigelt <info@metux.net>
Cc: linux-spdx@vger.kernel.org
Link: https://lkml.kernel.org/r/20190602204653.811534538@linutronix.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Writes to ZCR_EL1 are self-synchronising, and so may be expensive
in typical implementations.
This patch adopts the approach used for costly system register
writes elsewhere in the kernel: the system register write is
suppressed if it would not change the stored value.
Since the common case will be that of switching between tasks that
use the same vector length as one another, prediction hit rates on
the conditional branch should be reasonably good, with lower
expected amortised cost than the unconditional execution of a
heavyweight self-synchronising instruction.
Signed-off-by: Dave Martin <Dave.Martin@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Manipulating the SVE architectural state, including the vector and
predicate registers, first-fault register and the vector length,
requires the use of dedicated instructions added by SVE.
This patch adds suitable assembly functions for saving and
restoring the SVE registers and querying the vector length.
Setting of the vector length is done as part of register restore.
Since people building kernels may not all get an SVE-enabled
toolchain for a while, this patch uses macros that generate
explicit opcodes in place of assembler mnemonics.
Signed-off-by: Dave Martin <Dave.Martin@arm.com>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
Support for kernel-mode NEON to be nested and/or used in hardirq
context adds significant complexity, and the benefits may be
marginal. In practice, kernel-mode NEON is not used in hardirq
context, and is rarely used in softirq context (by certain mac80211
drivers).
This patch implements an arm64 may_use_simd() function to allow
clients to check whether kernel-mode NEON is usable in the current
context, and simplifies kernel_neon_{begin,end}() to handle only
saving of the task FPSIMD state (if any). Without nesting, there
is no other state to save.
The partial fpsimd save/restore functions become redundant as a
result of these changes, so they are removed too.
The save/restore model is changed to operate directly on
task_struct without additional percpu storage. This simplifies the
code and saves a bit of memory, but means that softirqs must now be
disabled when manipulating the task fpsimd state from task context:
correspondingly, preempt_{en,dis}sable() calls are upgraded to
local_bh_{en,dis}able() as appropriate. fpsimd_thread_switch()
already runs with hardirqs disabled and so is already protected
from softirqs.
These changes should make it easier to support kernel-mode NEON in
the presence of the Scalable Vector extension in the future.
Signed-off-by: Dave Martin <Dave.Martin@arm.com>
Reviewed-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Commit 190f1ca85d ("arm64: add support for kernel mode NEON in interrupt
context") introduced a typing error in fpsimd_save_partial_state ENDPROC.
This patch fixes the typing error.
Acked-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: byungchul.park <byungchul.park@lge.com>
Signed-off-by: Will Deacon <will.deacon@arm.com>
This patch modifies kernel_neon_begin() and kernel_neon_end(), so
they may be called from any context. To address the case where only
a couple of registers are needed, kernel_neon_begin_partial(u32) is
introduced which takes as a parameter the number of bottom 'n' NEON
q-registers required. To mark the end of such a partial section, the
regular kernel_neon_end() should be used.
Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org>
In order to be able to reuse the save-restore code in KVM, move
it to a pair of macros, similar to what the 32bit code does.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
This patch adds support for FP/ASIMD register bank saving and restoring
during context switch and FP exception handling to generate SIGFPE.
There are 32 128-bit registers and the context switching is currently
done non-lazily. Benchmarks on real hardware are required before
implementing lazy FP state saving/restoring.
Signed-off-by: Will Deacon <will.deacon@arm.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Acked-by: Tony Lindgren <tony@atomide.com>
Acked-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Nicolas Pitre <nico@linaro.org>
Acked-by: Olof Johansson <olof@lixom.net>
Acked-by: Santosh Shilimkar <santosh.shilimkar@ti.com>
Mark Brown