2023-09-12 16:07:50

by Jisheng Zhang

[permalink] [raw]
Subject: [PATCH] riscv: don't probe unaligned access speed if already done

If misaligned_access_speed percpu var isn't so called "HWPROBE
MISALIGNED UNKNOWN", it means the probe has happened(this is possible
for example, hotplug off then hotplug on one cpu), and the percpu var
has been set, don't probe again in this case.

Signed-off-by: Jisheng Zhang <[email protected]>
---
arch/riscv/kernel/cpufeature.c | 4 ++++
1 file changed, 4 insertions(+)

diff --git a/arch/riscv/kernel/cpufeature.c b/arch/riscv/kernel/cpufeature.c
index 1cfbba65d11a..e12cd22755c7 100644
--- a/arch/riscv/kernel/cpufeature.c
+++ b/arch/riscv/kernel/cpufeature.c
@@ -568,6 +568,10 @@ void check_unaligned_access(int cpu)
void *src;
long speed = RISCV_HWPROBE_MISALIGNED_SLOW;

+ /* We are already set since the last check */
+ if (per_cpu(misaligned_access_speed, cpu) != RISCV_HWPROBE_MISALIGNED_UNKNOWN)
+ return;
+
page = alloc_pages(GFP_NOWAIT, get_order(MISALIGNED_BUFFER_SIZE));
if (!page) {
pr_warn("Can't alloc pages to measure memcpy performance");
--
2.40.1


2023-09-13 07:59:02

by Conor Dooley

[permalink] [raw]
Subject: Re: [PATCH] riscv: don't probe unaligned access speed if already done

On Tue, Sep 12, 2023 at 11:40:40PM +0800, Jisheng Zhang wrote:
> If misaligned_access_speed percpu var isn't so called "HWPROBE
> MISALIGNED UNKNOWN", it means the probe has happened(this is possible
> for example, hotplug off then hotplug on one cpu), and the percpu var
> has been set, don't probe again in this case.
>
> Signed-off-by: Jisheng Zhang <[email protected]>

Presumably this should grow a
Fixes: 584ea6564bca ("RISC-V: Probe for unaligned access speed")
on application.
Patch seems fine to me,
Reviewed-by: Conor Dooley <[email protected]>

Thanks,
Conor.

> ---
> arch/riscv/kernel/cpufeature.c | 4 ++++
> 1 file changed, 4 insertions(+)
>
> diff --git a/arch/riscv/kernel/cpufeature.c b/arch/riscv/kernel/cpufeature.c
> index 1cfbba65d11a..e12cd22755c7 100644
> --- a/arch/riscv/kernel/cpufeature.c
> +++ b/arch/riscv/kernel/cpufeature.c
> @@ -568,6 +568,10 @@ void check_unaligned_access(int cpu)
> void *src;
> long speed = RISCV_HWPROBE_MISALIGNED_SLOW;
>
> + /* We are already set since the last check */
> + if (per_cpu(misaligned_access_speed, cpu) != RISCV_HWPROBE_MISALIGNED_UNKNOWN)
> + return;
> +
> page = alloc_pages(GFP_NOWAIT, get_order(MISALIGNED_BUFFER_SIZE));
> if (!page) {
> pr_warn("Can't alloc pages to measure memcpy performance");
> --
> 2.40.1
>

Attachments:
(No filename) (1.32 kB)
 signature.asc (235.00 B)
 Download all attachments
Subject: Re: [PATCH] riscv: don't probe unaligned access speed if already done

Hello:

This patch was applied to riscv/linux.git (for-next)
by Palmer Dabbelt <[email protected]>:

On Tue, 12 Sep 2023 23:40:40 +0800 you wrote:
> If misaligned_access_speed percpu var isn't so called "HWPROBE
> MISALIGNED UNKNOWN", it means the probe has happened(this is possible
> for example, hotplug off then hotplug on one cpu), and the percpu var
> has been set, don't probe again in this case.
>
> Signed-off-by: Jisheng Zhang <[email protected]>
>
> [...]

Here is the summary with links:
- riscv: don't probe unaligned access speed if already done
https://git.kernel.org/riscv/c/c20d36cc2a20

You are awesome, thank you!
--
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html