Re: [PATCH] drm/amd/amdgpu: Remove static from variable in RLCG Reg RW.

[Paul Menzel <pmenzel@molgen.mpg.de>]

Dear Gavin,


Thank you for your patch.

Am 13.04.22 um 17:26 schrieb Gavin Wan:

Should you re-roll your patch (v2), please remove the dot/period from 
the end of the git commit message summary (subject).

> [why] These static variables saves the RLC Scratch registers address.

s/saves/save/

>        When we installed multiple GPUs (for example: XGMI setting) and

s/installed/install/

>        multiple GPUs call the function at same time. The RLC Scratch

… same time, the RLC …

>        registers address are changed each other. Then it caused

s/caused/causes/

>        reading/writing to wrong GPU.

I see from other patches posted here, that [why] is put on a separate 
line, so you do not need to indent the text.

[why]

These static …

> 
> [fix] Removed the static from the variables. The variables are
>        in stack.

Same here, though *how* instead of *fix* seems more common.

s/Removed/Remove/
s/in stack/on the stack/

> 
> Signed-off-by: Gavin Wan <Gavin.Wan@amd.com>
> Change-Id: Iee78849291d4f7a9688ecc5165bec70ee85cdfbe

Without the Gerrit URL that line is useless.


Kind regards.

Paul


> ---
>   drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c | 10 +++++-----
>   1 file changed, 5 insertions(+), 5 deletions(-)
> 
> diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c
> index d5eea031c3e3..d18a05a20566 100644
> --- a/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c
> +++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_virt.c
> @@ -868,11 +868,11 @@ static u32 amdgpu_virt_rlcg_reg_rw(struct amdgpu_device *adev, u32 offset, u32 v
>   	uint32_t timeout = 50000;
>   	uint32_t i, tmp;
>   	uint32_t ret = 0;
> -	static void *scratch_reg0;
> -	static void *scratch_reg1;
> -	static void *scratch_reg2;
> -	static void *scratch_reg3;
> -	static void *spare_int;
> +	void *scratch_reg0;
> +	void *scratch_reg1;
> +	void *scratch_reg2;
> +	void *scratch_reg3;
> +	void *spare_int;
>   
>   	if (!adev->gfx.rlc.rlcg_reg_access_supported) {
>   		dev_err(adev->dev,