2009-03-15 22:06:04

by Adrian McMenamin

Subject: [PATCH] snd: aica - fix annoying compiler warning

Cast pointer to data member of struct firmware as a void to end an
annoying compiler warning.

fix annoying compiler warning

Signed-off-by: Adrian McMenamin <[email protected]>
---

diff --git a/sound/sh/aica.c b/sound/sh/aica.c
index 7c920f3..822b119 100644
--- a/sound/sh/aica.c
+++ b/sound/sh/aica.c
@@ -567,7 +567,7 @@ static int load_aica_firmware(void)
return err;
/* write firware into memory */
spu_disable();
- spu_memload(0, fw_entry->data, fw_entry->size);
+ spu_memload(0, (void *)fw_entry->data, fw_entry->size);
spu_enable();
release_firmware(fw_entry);
return err;

2009-03-16 06:32:15

by Takashi Iwai

Subject: Re: [PATCH] snd: aica - fix annoying compiler warning

At Sun, 15 Mar 2009 22:05:40 +0000,
Adrian McMenamin wrote:
>
> Cast pointer to data member of struct firmware as a void to end an
> annoying compiler warning.
>
> fix annoying compiler warning
>
> Signed-off-by: Adrian McMenamin <[email protected]>
> ---
>
> diff --git a/sound/sh/aica.c b/sound/sh/aica.c
> index 7c920f3..822b119 100644
> --- a/sound/sh/aica.c
> +++ b/sound/sh/aica.c
> @@ -567,7 +567,7 @@ static int load_aica_firmware(void)
> return err;
> /* write firware into memory */
> spu_disable();
> - spu_memload(0, fw_entry->data, fw_entry->size);
> + spu_memload(0, (void *)fw_entry->data, fw_entry->size);

IMO, it's better to fix spu_memload() to take const pointer instead of
cast.


thanks,

Takashi

2009-03-16 07:54:31

by Adrian McMenamin

Subject: Re: [PATCH] snd: aica - fix annoying compiler warning

On Mon, 2009-03-16 at 07:31 +0100, Takashi Iwai wrote:
> At Sun, 15 Mar 2009 22:05:40 +0000,
> Adrian McMenamin wrote:
> >
> > Cast pointer to data member of struct firmware as a void to end an
> > annoying compiler warning.
> >
> > fix annoying compiler warning
> >
> > Signed-off-by: Adrian McMenamin <[email protected]>
> > ---
> >
> > diff --git a/sound/sh/aica.c b/sound/sh/aica.c
> > index 7c920f3..822b119 100644
> > --- a/sound/sh/aica.c
> > +++ b/sound/sh/aica.c
> > @@ -567,7 +567,7 @@ static int load_aica_firmware(void)
> > return err;
> > /* write firware into memory */
> > spu_disable();
> > - spu_memload(0, fw_entry->data, fw_entry->size);
> > + spu_memload(0, (void *)fw_entry->data, fw_entry->size);
>
> IMO, it's better to fix spu_memload() to take const pointer instead of
> cast.
>

If that's what you want I can do that but it adds to kernel bloat by
having two functions essentially do the same thing.

2009-03-16 07:59:34

by Takashi Iwai

Subject: Re: [PATCH] snd: aica - fix annoying compiler warning

At Mon, 16 Mar 2009 07:54:09 +0000,
Adrian McMenamin wrote:
>
> On Mon, 2009-03-16 at 07:31 +0100, Takashi Iwai wrote:
> > At Sun, 15 Mar 2009 22:05:40 +0000,
> > Adrian McMenamin wrote:
> > >
> > > Cast pointer to data member of struct firmware as a void to end an
> > > annoying compiler warning.
> > >
> > > fix annoying compiler warning
> > >
> > > Signed-off-by: Adrian McMenamin <[email protected]>
> > > ---
> > >
> > > diff --git a/sound/sh/aica.c b/sound/sh/aica.c
> > > index 7c920f3..822b119 100644
> > > --- a/sound/sh/aica.c
> > > +++ b/sound/sh/aica.c
> > > @@ -567,7 +567,7 @@ static int load_aica_firmware(void)
> > > return err;
> > > /* write firware into memory */
> > > spu_disable();
> > > - spu_memload(0, fw_entry->data, fw_entry->size);
> > > + spu_memload(0, (void *)fw_entry->data, fw_entry->size);
> >
> > IMO, it's better to fix spu_memload() to take const pointer instead of
> > cast.
> >
>
> If that's what you want I can do that but it adds to kernel bloat by
> having two functions essentially do the same thing.

I meant a fix like below...


Takashi

---
diff --git a/sound/sh/aica.c b/sound/sh/aica.c
index f551233..fad0c47 100644
--- a/sound/sh/aica.c
+++ b/sound/sh/aica.c
@@ -119,10 +119,10 @@ static void spu_memset(u32 toi, u32 what, int length)
}

/* spu_memload - write to SPU address space */
-static void spu_memload(u32 toi, void *from, int length)
+static void spu_memload(u32 toi, const void *from, int length)
{
unsigned long flags;
- u32 *froml = from;
+ const u32 *froml = from;
u32 __iomem *to = (u32 __iomem *) (SPU_MEMORY_BASE + toi);
int i;
u32 val;

2009-03-30 23:15:30

by Paul Mundt

Subject: Re: [PATCH] snd: aica - fix annoying compiler warning

On Mon, Mar 16, 2009 at 08:59:20AM +0100, Takashi Iwai wrote:
> At Mon, 16 Mar 2009 07:54:09 +0000,
> Adrian McMenamin wrote:
> >
> > On Mon, 2009-03-16 at 07:31 +0100, Takashi Iwai wrote:
> > > At Sun, 15 Mar 2009 22:05:40 +0000,
> > > Adrian McMenamin wrote:
> > > >
> > > > Cast pointer to data member of struct firmware as a void to end an
> > > > annoying compiler warning.
> > > >
> > > > fix annoying compiler warning
> > > >
> > > > Signed-off-by: Adrian McMenamin <[email protected]>
> > > > ---
> > > >
> > > > diff --git a/sound/sh/aica.c b/sound/sh/aica.c
> > > > index 7c920f3..822b119 100644
> > > > --- a/sound/sh/aica.c
> > > > +++ b/sound/sh/aica.c
> > > > @@ -567,7 +567,7 @@ static int load_aica_firmware(void)
> > > > return err;
> > > > /* write firware into memory */
> > > > spu_disable();
> > > > - spu_memload(0, fw_entry->data, fw_entry->size);
> > > > + spu_memload(0, (void *)fw_entry->data, fw_entry->size);
> > >
> > > IMO, it's better to fix spu_memload() to take const pointer instead of
> > > cast.
> > >
> >
> > If that's what you want I can do that but it adds to kernel bloat by
> > having two functions essentially do the same thing.
>
> I meant a fix like below...
>
Is anything happening with this, or should I just ignore it?

> ---
> diff --git a/sound/sh/aica.c b/sound/sh/aica.c
> index f551233..fad0c47 100644
> --- a/sound/sh/aica.c
> +++ b/sound/sh/aica.c
> @@ -119,10 +119,10 @@ static void spu_memset(u32 toi, u32 what, int length)
> }
>
> /* spu_memload - write to SPU address space */
> -static void spu_memload(u32 toi, void *from, int length)
> +static void spu_memload(u32 toi, const void *from, int length)
> {
> unsigned long flags;
> - u32 *froml = from;
> + const u32 *froml = from;
> u32 __iomem *to = (u32 __iomem *) (SPU_MEMORY_BASE + toi);
> int i;
> u32 val;
> --
> To unsubscribe from this list: send the line "unsubscribe linux-sh" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html

2009-04-01 23:42:20

by Adrian McMenamin

Subject: Re: [PATCH] snd: aica - fix annoying compiler warning

On Tue, 2009-03-31 at 08:10 +0900, Paul Mundt wrote:
> On Mon, Mar 16, 2009 at 08:59:20AM +0100, Takashi Iwai wrote:
> > At Mon, 16 Mar 2009 07:54:09 +0000,
> > Adrian McMenamin wrote:
> > >
> > > On Mon, 2009-03-16 at 07:31 +0100, Takashi Iwai wrote:
> > > > At Sun, 15 Mar 2009 22:05:40 +0000,
> > > > Adrian McMenamin wrote:
> > > > >
> > > > > Cast pointer to data member of struct firmware as a void to end an
> > > > > annoying compiler warning.
> > > > >
> > > > > fix annoying compiler warning
> > > > >
> > > > > Signed-off-by: Adrian McMenamin <[email protected]>
> > > > > ---
> > > > >
> > > > > diff --git a/sound/sh/aica.c b/sound/sh/aica.c
> > > > > index 7c920f3..822b119 100644
> > > > > --- a/sound/sh/aica.c
> > > > > +++ b/sound/sh/aica.c
> > > > > @@ -567,7 +567,7 @@ static int load_aica_firmware(void)
> > > > > return err;
> > > > > /* write firware into memory */
> > > > > spu_disable();
> > > > > - spu_memload(0, fw_entry->data, fw_entry->size);
> > > > > + spu_memload(0, (void *)fw_entry->data, fw_entry->size);
> > > >
> > > > IMO, it's better to fix spu_memload() to take const pointer instead of
> > > > cast.
> > > >
> > >
> > > If that's what you want I can do that but it adds to kernel bloat by
> > > having two functions essentially do the same thing.
> >
> > I meant a fix like below...
> >
> Is anything happening with this, or should I just ignore it?
>

I though Takashi's fix was fine/obvious but it doesn't appear to have
been pushed upstream (yet?)

2009-04-06 01:28:29

by Takashi Iwai

Subject: Re: [PATCH] snd: aica - fix annoying compiler warning

At Thu, 02 Apr 2009 00:41:50 +0100,
Adrian McMenamin wrote:
>
> On Tue, 2009-03-31 at 08:10 +0900, Paul Mundt wrote:
> > On Mon, Mar 16, 2009 at 08:59:20AM +0100, Takashi Iwai wrote:
> > > At Mon, 16 Mar 2009 07:54:09 +0000,
> > > Adrian McMenamin wrote:
> > > >
> > > > On Mon, 2009-03-16 at 07:31 +0100, Takashi Iwai wrote:
> > > > > At Sun, 15 Mar 2009 22:05:40 +0000,
> > > > > Adrian McMenamin wrote:
> > > > > >
> > > > > > Cast pointer to data member of struct firmware as a void to end an
> > > > > > annoying compiler warning.
> > > > > >
> > > > > > fix annoying compiler warning
> > > > > >
> > > > > > Signed-off-by: Adrian McMenamin <[email protected]>
> > > > > > ---
> > > > > >
> > > > > > diff --git a/sound/sh/aica.c b/sound/sh/aica.c
> > > > > > index 7c920f3..822b119 100644
> > > > > > --- a/sound/sh/aica.c
> > > > > > +++ b/sound/sh/aica.c
> > > > > > @@ -567,7 +567,7 @@ static int load_aica_firmware(void)
> > > > > > return err;
> > > > > > /* write firware into memory */
> > > > > > spu_disable();
> > > > > > - spu_memload(0, fw_entry->data, fw_entry->size);
> > > > > > + spu_memload(0, (void *)fw_entry->data, fw_entry->size);
> > > > >
> > > > > IMO, it's better to fix spu_memload() to take const pointer instead of
> > > > > cast.
> > > > >
> > > >
> > > > If that's what you want I can do that but it adds to kernel bloat by
> > > > having two functions essentially do the same thing.
> > >
> > > I meant a fix like below...
> > >
> > Is anything happening with this, or should I just ignore it?
> >
>
> I though Takashi's fix was fine/obvious but it doesn't appear to have
> been pushed upstream (yet?)

Well, then it'd be appreciated if you repost your patch with my fix.

thanks,

Takashi