[PATCH] drm/amd/powerplay: use ARRAY_SIZE() for size of array

[PATCH] drm/amd/powerplay: use ARRAY_SIZE() for size of array

[Muhammad Falak R Wani]

Use ARRAY_SIZE() for the size calculation of the array. Also move the
condition evaulation function out of the for loop.
Although, any respectable c-compiler would optimize this and evaluate
the function only once outside the loop, but the optimzation engine
of gcc is bit brain-dead, and at times needs some hand holding.

Signed-off-by: Muhammad Falak R Wani <falakreyaz@gmail.com>
---
 drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c
index da18f44..718a551 100644
--- a/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c
+++ b/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c
@@ -636,10 +636,11 @@ static int cz_smu_populate_firmware_entries(struct pp_smumgr *smumgr)
 	int ret;
 	enum cgs_ucode_id ucode_id;
 	struct cgs_firmware_info info = {0};
+	int n = ARRAY_SIZE(firmware_list);
 
 	cz_smu->driver_buffer_length = 0;
 
-	for (i = 0; i < sizeof(firmware_list)/sizeof(*firmware_list); i++) {
+	for (i = 0; i < n; i++) {
 
 		firmware_type = cz_translate_firmware_enum_to_arg(smumgr,
 					firmware_list[i]);
-- 
1.9.1

Re: [PATCH] drm/amd/powerplay: use ARRAY_SIZE() for size of array

[Eric Engestrom]

On Wed, May 11, 2016 at 11:18:43PM +0530, Muhammad Falak R Wani wrote:
> Use ARRAY_SIZE() for the size calculation of the array. Also move the
> condition evaulation function out of the for loop.
> Although, any respectable c-compiler would optimize this and evaluate
> the function only once outside the loop, but the optimzation engine
> of gcc is bit brain-dead, and at times needs some hand holding.

`sizeof` is actually a compile-time thing, so at worst, if no
optimisation is made, the runtime result is a division of two
literals, eg. `for (i = 0; i < 64/8; i++)` (which I doubt any
compiler would leave as is anyway)

So, +1 on using ARRAY_SIZE, -1 on creating a new variable (which
is not even `const` btw)

> 
> Signed-off-by: Muhammad Falak R Wani <falakreyaz@gmail.com>
> ---
>  drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c
> index da18f44..718a551 100644
> --- a/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c
> +++ b/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c
> @@ -636,10 +636,11 @@ static int cz_smu_populate_firmware_entries(struct pp_smumgr *smumgr)
>  	int ret;
>  	enum cgs_ucode_id ucode_id;
>  	struct cgs_firmware_info info = {0};
> +	int n = ARRAY_SIZE(firmware_list);
>  
>  	cz_smu->driver_buffer_length = 0;
>  
> -	for (i = 0; i < sizeof(firmware_list)/sizeof(*firmware_list); i++) {
> +	for (i = 0; i < n; i++) {
>  
>  		firmware_type = cz_translate_firmware_enum_to_arg(smumgr,
>  					firmware_list[i]);
> -- 
> 1.9.1
> 
> _______________________________________________
> dri-devel mailing list
> dri-devel@lists.freedesktop.org
> https://lists.freedesktop.org/mailman/listinfo/dri-devel

Re: [PATCH] drm/amd/powerplay: use ARRAY_SIZE() for size of array

[Jani Nikula]

On Wed, 11 May 2016, Muhammad Falak R Wani <falakreyaz@gmail.com> wrote:
> Use ARRAY_SIZE() for the size calculation of the array. Also move the
> condition evaulation function out of the for loop.
> Although, any respectable c-compiler would optimize this and evaluate
> the function only once outside the loop, but the optimzation engine
> of gcc is bit brain-dead, and at times needs some hand holding.

This just caught my eye. ARRAY_SIZE is a macro that expands to a compile
time constant. Arguably adding the the local variable here gives GCC
more chances to go wrong than keeping the ARRAY_SIZE in the loop
condition.

BR,
Jani.

>
> Signed-off-by: Muhammad Falak R Wani <falakreyaz@gmail.com>
> ---
>  drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c
> index da18f44..718a551 100644
> --- a/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c
> +++ b/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c
> @@ -636,10 +636,11 @@ static int cz_smu_populate_firmware_entries(struct pp_smumgr *smumgr)
>  	int ret;
>  	enum cgs_ucode_id ucode_id;
>  	struct cgs_firmware_info info = {0};
> +	int n = ARRAY_SIZE(firmware_list);
>  
>  	cz_smu->driver_buffer_length = 0;
>  
> -	for (i = 0; i < sizeof(firmware_list)/sizeof(*firmware_list); i++) {
> +	for (i = 0; i < n; i++) {
>  
>  		firmware_type = cz_translate_firmware_enum_to_arg(smumgr,
>  					firmware_list[i]);

-- 
Jani Nikula, Intel Open Source Technology Center

Re: [PATCH] drm/amd/powerplay: use ARRAY_SIZE() for size of array

[Muhammad Falak R Wani]

On Thu, May 12, 2016 at 12:53:48PM +0300, Jani Nikula wrote:
> On Wed, 11 May 2016, Muhammad Falak R Wani <falakreyaz@gmail.com> wrote:
> > Use ARRAY_SIZE() for the size calculation of the array. Also move the
> > condition evaulation function out of the for loop.
> > Although, any respectable c-compiler would optimize this and evaluate
> > the function only once outside the loop, but the optimzation engine
> > of gcc is bit brain-dead, and at times needs some hand holding.
> 
> This just caught my eye. ARRAY_SIZE is a macro that expands to a compile
> time constant. Arguably adding the the local variable here gives GCC
> more chances to go wrong than keeping the ARRAY_SIZE in the loop
> condition.
> 
> BR,
> Jani.
> 
> >
> > Signed-off-by: Muhammad Falak R Wani <falakreyaz@gmail.com>
> > ---
> >  drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c | 3 ++-
> >  1 file changed, 2 insertions(+), 1 deletion(-)
> >
> > diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c b/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c
> > index da18f44..718a551 100644
> > --- a/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c
> > +++ b/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c
> > @@ -636,10 +636,11 @@ static int cz_smu_populate_firmware_entries(struct pp_smumgr *smumgr)
> >  	int ret;
> >  	enum cgs_ucode_id ucode_id;
> >  	struct cgs_firmware_info info = {0};
> > +	int n = ARRAY_SIZE(firmware_list);
> >  
> >  	cz_smu->driver_buffer_length = 0;
> >  
> > -	for (i = 0; i < sizeof(firmware_list)/sizeof(*firmware_list); i++) {
> > +	for (i = 0; i < n; i++) {
> >  
> >  		firmware_type = cz_translate_firmware_enum_to_arg(smumgr,
> >  					firmware_list[i]);
> 
> -- 
> Jani Nikula, Intel Open Source Technology Center
Should i send a new patch, with these modifications. Sorry for my
childish mistake, i got a little too carried away.

RE: [PATCH] drm/amd/powerplay: use ARRAY_SIZE() for size of array

[Deucher, Alexander]

> -----Original Message-----
> From: Muhammad Falak R Wani [mailto:falakreyaz@gmail.com]
> Sent: Friday, May 13, 2016 1:17 PM
> To: Jani Nikula
> Cc: Koenig, Christian; Nils Wallménius; Zhou, Jammy; linux-
> kernel@vger.kernel.org; dri-devel@lists.freedesktop.org; Deucher,
> Alexander; Zhu, Rex; Dan Carpenter
> Subject: Re: [PATCH] drm/amd/powerplay: use ARRAY_SIZE() for size of
> array
> 
> On Thu, May 12, 2016 at 12:53:48PM +0300, Jani Nikula wrote:
> > On Wed, 11 May 2016, Muhammad Falak R Wani <falakreyaz@gmail.com>
> wrote:
> > > Use ARRAY_SIZE() for the size calculation of the array. Also move the
> > > condition evaulation function out of the for loop.
> > > Although, any respectable c-compiler would optimize this and evaluate
> > > the function only once outside the loop, but the optimzation engine
> > > of gcc is bit brain-dead, and at times needs some hand holding.
> >
> > This just caught my eye. ARRAY_SIZE is a macro that expands to a compile
> > time constant. Arguably adding the the local variable here gives GCC
> > more chances to go wrong than keeping the ARRAY_SIZE in the loop
> > condition.
> >
> > BR,
> > Jani.
> >
> > >
> > > Signed-off-by: Muhammad Falak R Wani <falakreyaz@gmail.com>
> > > ---
> > >  drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c | 3 ++-
> > >  1 file changed, 2 insertions(+), 1 deletion(-)
> > >
> > > diff --git a/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c
> b/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c
> > > index da18f44..718a551 100644
> > > --- a/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c
> > > +++ b/drivers/gpu/drm/amd/powerplay/smumgr/cz_smumgr.c
> > > @@ -636,10 +636,11 @@ static int
> cz_smu_populate_firmware_entries(struct pp_smumgr *smumgr)
> > >  	int ret;
> > >  	enum cgs_ucode_id ucode_id;
> > >  	struct cgs_firmware_info info = {0};
> > > +	int n = ARRAY_SIZE(firmware_list);
> > >
> > >  	cz_smu->driver_buffer_length = 0;
> > >
> > > -	for (i = 0; i < sizeof(firmware_list)/sizeof(*firmware_list); i++) {
> > > +	for (i = 0; i < n; i++) {
> > >
> > >  		firmware_type =
> cz_translate_firmware_enum_to_arg(smumgr,
> > >  					firmware_list[i]);
> >
> > --
> > Jani Nikula, Intel Open Source Technology Center
> Should i send a new patch, with these modifications. Sorry for my
> childish mistake, i got a little too carried away.

Yes, please send an updated patch.

Thanks,

Alex

Re: [PATCH] drm/amd/powerplay: use ARRAY_SIZE() for size of array

[Dan Carpenter]

On Fri, May 13, 2016 at 10:47:05PM +0530, Muhammad Falak R Wani wrote:
> Sorry for my childish mistake, i got a little too carried away.

No need to appologize, it's just a disagreement about style issues.

regards,
dan carpenter