All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Alex Bennée" <alex.bennee@linaro.org>
To: Dave Martin <Dave.Martin@arm.com>
Cc: Christoffer Dall <cdall@kernel.org>,
	Ard Biesheuvel <ard.biesheuvel@linaro.org>,
	Marc Zyngier <marc.zyngier@arm.com>,
	Catalin Marinas <catalin.marinas@arm.com>,
	Will Deacon <will.deacon@arm.com>,
	kvmarm@lists.cs.columbia.edu,
	linux-arm-kernel@lists.infradead.org
Subject: Re: [PATCH v10 07/18] arm64: fpsimd: Eliminate task->mm checks
Date: Thu, 24 May 2018 10:19:26 +0100	[thread overview]
Message-ID: <87bmd5tnq9.fsf@linaro.org> (raw)
In-Reply-To: <1527005119-6842-8-git-send-email-Dave.Martin@arm.com>


Dave Martin <Dave.Martin@arm.com> writes:

> Currently the FPSIMD handling code uses the condition task->mm ==
> NULL as a hint that task has no FPSIMD register context.
>
> The ->mm check is only there to filter out tasks that cannot
> possibly have FPSIMD context loaded, for optimisation purposes.
> Also, TIF_FOREIGN_FPSTATE must always be checked anyway before
> saving FPSIMD context back to memory.  For these reasons, the ->mm
> checks are not useful, providing that that TIF_FOREIGN_FPSTATE is
> maintained in a consistent way for kernel threads.
>
> This is true by construction however: TIF_FOREIGN_FPSTATE is never
> cleared except when returning to userspace or returning from a
> signal: thus, for a true kernel thread no FPSIMD context is ever
> loaded, TIF_FOREIGN_FPSTATE will remain set and no context will
> ever be saved.
>
> This patch removes the redundant checks and special-case code.
>
> Signed-off-by: Dave Martin <Dave.Martin@arm.com>
> Cc: Catalin Marinas <catalin.marinas@arm.com>
> Cc: Will Deacon <will.deacon@arm.com>
> Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>

With Christoffer's commit text:

Reviewed-by: Alex Bennée <alex.bennee@linaro.org>

>
> ---
>
> Changes since v9:
>
>  * New patch.  Introduced during debugging, since the ->mm checks
>    appear bogus and/or redundant, so are likely to be hiding or
>    causing bugs.
> ---
>  arch/arm64/include/asm/thread_info.h |  1 +
>  arch/arm64/kernel/fpsimd.c           | 38 ++++++++++++------------------------
>  2 files changed, 14 insertions(+), 25 deletions(-)
>
> diff --git a/arch/arm64/include/asm/thread_info.h b/arch/arm64/include/asm/thread_info.h
> index 740aa03c..a2ac914 100644
> --- a/arch/arm64/include/asm/thread_info.h
> +++ b/arch/arm64/include/asm/thread_info.h
> @@ -47,6 +47,7 @@ struct thread_info {
>
>  #define INIT_THREAD_INFO(tsk)						\
>  {									\
> +	.flags		= _TIF_FOREIGN_FPSTATE,				\
>  	.preempt_count	= INIT_PREEMPT_COUNT,				\
>  	.addr_limit	= KERNEL_DS,					\
>  }
> diff --git a/arch/arm64/kernel/fpsimd.c b/arch/arm64/kernel/fpsimd.c
> index 3aa100a..1222491 100644
> --- a/arch/arm64/kernel/fpsimd.c
> +++ b/arch/arm64/kernel/fpsimd.c
> @@ -891,31 +891,21 @@ asmlinkage void do_fpsimd_exc(unsigned int esr, struct pt_regs *regs)
>
>  void fpsimd_thread_switch(struct task_struct *next)
>  {
> +	bool wrong_task, wrong_cpu;
> +
>  	if (!system_supports_fpsimd())
>  		return;
> -	/*
> -	 * Save the current FPSIMD state to memory, but only if whatever is in
> -	 * the registers is in fact the most recent userland FPSIMD state of
> -	 * 'current'.
> -	 */
> -	if (current->mm)
> -		fpsimd_save();
>
> -	if (next->mm) {
> -		/*
> -		 * If we are switching to a task whose most recent userland
> -		 * FPSIMD state is already in the registers of *this* cpu,
> -		 * we can skip loading the state from memory. Otherwise, set
> -		 * the TIF_FOREIGN_FPSTATE flag so the state will be loaded
> -		 * upon the next return to userland.
> -		 */
> -		bool wrong_task = __this_cpu_read(fpsimd_last_state.st) !=
> +	/* Save unsaved fpsimd state, if any: */
> +	fpsimd_save();
> +
> +	/* Fix up TIF_FOREIGN_FPSTATE to correctly describe next's state: */
> +	wrong_task = __this_cpu_read(fpsimd_last_state.st) !=
>  					&next->thread.uw.fpsimd_state;
> -		bool wrong_cpu = next->thread.fpsimd_cpu != smp_processor_id();
> +	wrong_cpu = next->thread.fpsimd_cpu != smp_processor_id();
>
> -		update_tsk_thread_flag(next, TIF_FOREIGN_FPSTATE,
> -				       wrong_task || wrong_cpu);
> -	}
> +	update_tsk_thread_flag(next, TIF_FOREIGN_FPSTATE,
> +			       wrong_task || wrong_cpu);
>  }
>
>  void fpsimd_flush_thread(void)
> @@ -1120,9 +1110,8 @@ void kernel_neon_begin(void)
>
>  	__this_cpu_write(kernel_neon_busy, true);
>
> -	/* Save unsaved task fpsimd state, if any: */
> -	if (current->mm)
> -		fpsimd_save();
> +	/* Save unsaved fpsimd state, if any: */
> +	fpsimd_save();
>
>  	/* Invalidate any task state remaining in the fpsimd regs: */
>  	fpsimd_flush_cpu_state();
> @@ -1244,8 +1233,7 @@ static int fpsimd_cpu_pm_notifier(struct notifier_block *self,
>  {
>  	switch (cmd) {
>  	case CPU_PM_ENTER:
> -		if (current->mm)
> -			fpsimd_save();
> +		fpsimd_save();
>  		fpsimd_flush_cpu_state();
>  		break;
>  	case CPU_PM_EXIT:


--
Alex Bennée

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

WARNING: multiple messages have this Message-ID (diff)
From: alex.bennee@linaro.org (Alex Bennée)
To: linux-arm-kernel@lists.infradead.org
Subject: [PATCH v10 07/18] arm64: fpsimd: Eliminate task->mm checks
Date: Thu, 24 May 2018 10:19:26 +0100	[thread overview]
Message-ID: <87bmd5tnq9.fsf@linaro.org> (raw)
In-Reply-To: <1527005119-6842-8-git-send-email-Dave.Martin@arm.com>


Dave Martin <Dave.Martin@arm.com> writes:

> Currently the FPSIMD handling code uses the condition task->mm ==
> NULL as a hint that task has no FPSIMD register context.
>
> The ->mm check is only there to filter out tasks that cannot
> possibly have FPSIMD context loaded, for optimisation purposes.
> Also, TIF_FOREIGN_FPSTATE must always be checked anyway before
> saving FPSIMD context back to memory.  For these reasons, the ->mm
> checks are not useful, providing that that TIF_FOREIGN_FPSTATE is
> maintained in a consistent way for kernel threads.
>
> This is true by construction however: TIF_FOREIGN_FPSTATE is never
> cleared except when returning to userspace or returning from a
> signal: thus, for a true kernel thread no FPSIMD context is ever
> loaded, TIF_FOREIGN_FPSTATE will remain set and no context will
> ever be saved.
>
> This patch removes the redundant checks and special-case code.
>
> Signed-off-by: Dave Martin <Dave.Martin@arm.com>
> Cc: Catalin Marinas <catalin.marinas@arm.com>
> Cc: Will Deacon <will.deacon@arm.com>
> Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>

With Christoffer's commit text:

Reviewed-by: Alex Benn?e <alex.bennee@linaro.org>

>
> ---
>
> Changes since v9:
>
>  * New patch.  Introduced during debugging, since the ->mm checks
>    appear bogus and/or redundant, so are likely to be hiding or
>    causing bugs.
> ---
>  arch/arm64/include/asm/thread_info.h |  1 +
>  arch/arm64/kernel/fpsimd.c           | 38 ++++++++++++------------------------
>  2 files changed, 14 insertions(+), 25 deletions(-)
>
> diff --git a/arch/arm64/include/asm/thread_info.h b/arch/arm64/include/asm/thread_info.h
> index 740aa03c..a2ac914 100644
> --- a/arch/arm64/include/asm/thread_info.h
> +++ b/arch/arm64/include/asm/thread_info.h
> @@ -47,6 +47,7 @@ struct thread_info {
>
>  #define INIT_THREAD_INFO(tsk)						\
>  {									\
> +	.flags		= _TIF_FOREIGN_FPSTATE,				\
>  	.preempt_count	= INIT_PREEMPT_COUNT,				\
>  	.addr_limit	= KERNEL_DS,					\
>  }
> diff --git a/arch/arm64/kernel/fpsimd.c b/arch/arm64/kernel/fpsimd.c
> index 3aa100a..1222491 100644
> --- a/arch/arm64/kernel/fpsimd.c
> +++ b/arch/arm64/kernel/fpsimd.c
> @@ -891,31 +891,21 @@ asmlinkage void do_fpsimd_exc(unsigned int esr, struct pt_regs *regs)
>
>  void fpsimd_thread_switch(struct task_struct *next)
>  {
> +	bool wrong_task, wrong_cpu;
> +
>  	if (!system_supports_fpsimd())
>  		return;
> -	/*
> -	 * Save the current FPSIMD state to memory, but only if whatever is in
> -	 * the registers is in fact the most recent userland FPSIMD state of
> -	 * 'current'.
> -	 */
> -	if (current->mm)
> -		fpsimd_save();
>
> -	if (next->mm) {
> -		/*
> -		 * If we are switching to a task whose most recent userland
> -		 * FPSIMD state is already in the registers of *this* cpu,
> -		 * we can skip loading the state from memory. Otherwise, set
> -		 * the TIF_FOREIGN_FPSTATE flag so the state will be loaded
> -		 * upon the next return to userland.
> -		 */
> -		bool wrong_task = __this_cpu_read(fpsimd_last_state.st) !=
> +	/* Save unsaved fpsimd state, if any: */
> +	fpsimd_save();
> +
> +	/* Fix up TIF_FOREIGN_FPSTATE to correctly describe next's state: */
> +	wrong_task = __this_cpu_read(fpsimd_last_state.st) !=
>  					&next->thread.uw.fpsimd_state;
> -		bool wrong_cpu = next->thread.fpsimd_cpu != smp_processor_id();
> +	wrong_cpu = next->thread.fpsimd_cpu != smp_processor_id();
>
> -		update_tsk_thread_flag(next, TIF_FOREIGN_FPSTATE,
> -				       wrong_task || wrong_cpu);
> -	}
> +	update_tsk_thread_flag(next, TIF_FOREIGN_FPSTATE,
> +			       wrong_task || wrong_cpu);
>  }
>
>  void fpsimd_flush_thread(void)
> @@ -1120,9 +1110,8 @@ void kernel_neon_begin(void)
>
>  	__this_cpu_write(kernel_neon_busy, true);
>
> -	/* Save unsaved task fpsimd state, if any: */
> -	if (current->mm)
> -		fpsimd_save();
> +	/* Save unsaved fpsimd state, if any: */
> +	fpsimd_save();
>
>  	/* Invalidate any task state remaining in the fpsimd regs: */
>  	fpsimd_flush_cpu_state();
> @@ -1244,8 +1233,7 @@ static int fpsimd_cpu_pm_notifier(struct notifier_block *self,
>  {
>  	switch (cmd) {
>  	case CPU_PM_ENTER:
> -		if (current->mm)
> -			fpsimd_save();
> +		fpsimd_save();
>  		fpsimd_flush_cpu_state();
>  		break;
>  	case CPU_PM_EXIT:


--
Alex Benn?e

  parent reply	other threads:[~2018-05-24  9:19 UTC|newest]

Thread overview: 138+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-05-22 16:05 [PATCH v10 00/18] KVM: arm64: Optimise FPSIMD context switching Dave Martin
2018-05-22 16:05 ` Dave Martin
2018-05-22 16:05 ` [PATCH v10 01/18] arm64: fpsimd: Fix TIF_FOREIGN_FPSTATE after invalidating cpu regs Dave Martin
2018-05-22 16:05   ` Dave Martin
2018-05-23 11:33   ` Christoffer Dall
2018-05-23 11:33     ` Christoffer Dall
2018-05-23 13:44   ` Alex Bennée
2018-05-23 13:44     ` Alex Bennée
2018-05-23 13:46   ` Catalin Marinas
2018-05-23 13:46     ` Catalin Marinas
2018-05-22 16:05 ` [PATCH v10 02/18] thread_info: Add update_thread_flag() helpers Dave Martin
2018-05-22 16:05   ` Dave Martin
2018-05-23 13:46   ` Alex Bennée
2018-05-23 13:46     ` Alex Bennée
2018-05-23 13:57     ` Dave Martin
2018-05-23 13:57       ` Dave Martin
2018-05-23 14:35       ` Alex Bennée
2018-05-23 14:35         ` Alex Bennée
2018-05-22 16:05 ` [PATCH v10 03/18] arm64: Use update{,_tsk}_thread_flag() Dave Martin
2018-05-22 16:05   ` Dave Martin
2018-05-23 13:48   ` Alex Bennée
2018-05-23 13:48     ` Alex Bennée
2018-05-22 16:05 ` [PATCH v10 04/18] KVM: arm/arm64: Introduce kvm_arch_vcpu_run_pid_change Dave Martin
2018-05-22 16:05   ` Dave Martin
2018-05-23 14:34   ` Alex Bennée
2018-05-23 14:34     ` Alex Bennée
2018-05-23 14:40     ` Dave Martin
2018-05-23 14:40       ` Dave Martin
2018-05-24  8:11       ` Christoffer Dall
2018-05-24  8:11         ` Christoffer Dall
2018-05-24  9:18         ` Alex Bennée
2018-05-24  9:18           ` Alex Bennée
2018-05-24 10:04           ` Dave Martin
2018-05-24 10:04             ` Dave Martin
2018-05-22 16:05 ` [PATCH v10 05/18] KVM: arm64: Convert lazy FPSIMD context switch trap to C Dave Martin
2018-05-22 16:05   ` Dave Martin
2018-05-23 19:35   ` Alex Bennée
2018-05-23 19:35     ` Alex Bennée
2018-05-24  8:12     ` Christoffer Dall
2018-05-24  8:12       ` Christoffer Dall
2018-05-24  8:54       ` Dave Martin
2018-05-24  8:54         ` Dave Martin
2018-05-24  9:14         ` Alex Bennée
2018-05-24  9:14           ` Alex Bennée
2018-05-22 16:05 ` [PATCH v10 06/18] arm64: fpsimd: Generalise context saving for non-task contexts Dave Martin
2018-05-22 16:05   ` Dave Martin
2018-05-23 20:15   ` Alex Bennée
2018-05-23 20:15     ` Alex Bennée
2018-05-24  9:03     ` Dave Martin
2018-05-24  9:03       ` Dave Martin
2018-05-24  9:41       ` Alex Bennée
2018-05-24  9:41         ` Alex Bennée
2018-05-22 16:05 ` [PATCH v10 07/18] arm64: fpsimd: Eliminate task->mm checks Dave Martin
2018-05-22 16:05   ` Dave Martin
2018-05-23 11:48   ` Christoffer Dall
2018-05-23 11:48     ` Christoffer Dall
2018-05-23 13:31     ` Dave Martin
2018-05-23 13:31       ` Dave Martin
2018-05-23 14:56       ` Catalin Marinas
2018-05-23 14:56         ` Catalin Marinas
2018-05-23 15:03         ` Dave Martin
2018-05-23 15:03           ` Dave Martin
2018-05-23 16:42           ` Catalin Marinas
2018-05-23 16:42             ` Catalin Marinas
2018-05-24  8:33           ` Christoffer Dall
2018-05-24  8:33             ` Christoffer Dall
2018-05-24  9:16             ` Alex Bennée
2018-05-24  9:16               ` Alex Bennée
2018-05-24  9:50             ` Dave Martin
2018-05-24  9:50               ` Dave Martin
2018-05-24 10:06               ` Christoffer Dall
2018-05-24 10:06                 ` Christoffer Dall
2018-05-24 14:37                 ` Dave Martin
2018-05-24 14:37                   ` Dave Martin
2018-05-25  9:00                   ` Christoffer Dall
2018-05-25  9:00                     ` Christoffer Dall
2018-05-25  9:45                     ` Dave Martin
2018-05-25  9:45                       ` Dave Martin
2018-05-25 11:28                       ` Christoffer Dall
2018-05-25 11:28                         ` Christoffer Dall
2018-05-24  9:19   ` Alex Bennée [this message]
2018-05-24  9:19     ` Alex Bennée
2018-05-22 16:05 ` [PATCH v10 08/18] arm64/sve: Refactor user SVE trap maintenance for external use Dave Martin
2018-05-22 16:05   ` Dave Martin
2018-05-23 20:16   ` Alex Bennée
2018-05-23 20:16     ` Alex Bennée
2018-05-22 16:05 ` [PATCH v10 09/18] KVM: arm64: Repurpose vcpu_arch.debug_flags for general-purpose flags Dave Martin
2018-05-22 16:05   ` Dave Martin
2018-05-24  9:21   ` Alex Bennée
2018-05-24  9:21     ` Alex Bennée
2018-05-22 16:05 ` [PATCH v10 10/18] KVM: arm64: Optimise FPSIMD handling to reduce guest/host thrashing Dave Martin
2018-05-22 16:05   ` Dave Martin
2018-05-24 10:09   ` Alex Bennée
2018-05-24 10:09     ` Alex Bennée
2018-05-24 10:18     ` Dave Martin
2018-05-24 10:18       ` Dave Martin
2018-05-22 16:05 ` [PATCH v10 11/18] arm64/sve: Move read_zcr_features() out of cpufeature.h Dave Martin
2018-05-22 16:05   ` Dave Martin
2018-05-24 10:12   ` Alex Bennée
2018-05-24 10:12     ` Alex Bennée
2018-05-22 16:05 ` [PATCH v10 12/18] arm64/sve: Switch sve_pffr() argument from task to thread Dave Martin
2018-05-22 16:05   ` Dave Martin
2018-05-24 10:12   ` Alex Bennée
2018-05-24 10:12     ` Alex Bennée
2018-05-22 16:05 ` [PATCH v10 13/18] arm64/sve: Move sve_pffr() to fpsimd.h and make inline Dave Martin
2018-05-22 16:05   ` Dave Martin
2018-05-24 10:20   ` Alex Bennée
2018-05-24 10:20     ` Alex Bennée
2018-05-24 11:22     ` Dave Martin
2018-05-24 11:22       ` Dave Martin
2018-05-22 16:05 ` [PATCH v10 14/18] KVM: arm64: Save host SVE context as appropriate Dave Martin
2018-05-22 16:05   ` Dave Martin
2018-05-23 14:59   ` Catalin Marinas
2018-05-23 14:59     ` Catalin Marinas
2018-05-24  9:11   ` Christoffer Dall
2018-05-24  9:11     ` Christoffer Dall
2018-05-24 14:49   ` Alex Bennée
2018-05-24 14:49     ` Alex Bennée
2018-05-22 16:05 ` [PATCH v10 15/18] KVM: arm64: Remove eager host SVE state saving Dave Martin
2018-05-22 16:05   ` Dave Martin
2018-05-24 14:54   ` Alex Bennée
2018-05-24 14:54     ` Alex Bennée
2018-05-22 16:05 ` [PATCH v10 16/18] KVM: arm64: Remove redundant *exit_code changes in fpsimd_guest_exit() Dave Martin
2018-05-22 16:05   ` Dave Martin
2018-05-24  9:11   ` Christoffer Dall
2018-05-24  9:11     ` Christoffer Dall
2018-05-24 15:02   ` Alex Bennée
2018-05-24 15:02     ` Alex Bennée
2018-05-22 16:05 ` [PATCH v10 17/18] KVM: arm64: Fold redundant exit code checks out of fixup_guest_exit() Dave Martin
2018-05-22 16:05   ` Dave Martin
2018-05-24  9:12   ` Christoffer Dall
2018-05-24  9:12     ` Christoffer Dall
2018-05-24 15:06   ` Alex Bennée
2018-05-24 15:06     ` Alex Bennée
2018-05-22 16:05 ` [PATCH v10 18/18] KVM: arm64: Invoke FPSIMD context switch trap from C Dave Martin
2018-05-22 16:05   ` Dave Martin
2018-05-24 15:09   ` Alex Bennée
2018-05-24 15:09     ` Alex Bennée

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=87bmd5tnq9.fsf@linaro.org \
    --to=alex.bennee@linaro.org \
    --cc=Dave.Martin@arm.com \
    --cc=ard.biesheuvel@linaro.org \
    --cc=catalin.marinas@arm.com \
    --cc=cdall@kernel.org \
    --cc=kvmarm@lists.cs.columbia.edu \
    --cc=linux-arm-kernel@lists.infradead.org \
    --cc=marc.zyngier@arm.com \
    --cc=will.deacon@arm.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.