[U-Boot] [PATCH 7/7] GCC4.6: Squash warning in cmd_mem.c

[Simon Glass]

Hi Merek,

On Mon, Sep 26, 2011 at 11:24 AM, Marek Vasut <marek.vasut at gmail.com> wrote:
> On Monday, September 26, 2011 08:05:43 PM Simon Glass wrote:
>> Hi Marek,
>>
>> On Sun, Sep 25, 2011 at 5:26 PM, Marek Vasut <marek.vasut at gmail.com> wrote:
>> > cmd_mem.c: In function ‘do_mem_loop’:
>> > cmd_mem.c:474:25: warning: variable ‘junk’ set but not used
>> > [-Wunused-but-set-variable]
>> >
>> > The assigned variable can be removed because the pointers are volatile so
>> > accesses to their addresses are always generated.
>> >
>> > Signed-off-by: Marek Vasut <marek.vasut at gmail.com>
>> > ---
>> >  common/cmd_mem.c |    8 ++++----
>> >  1 files changed, 4 insertions(+), 4 deletions(-)
>> >
>> > diff --git a/common/cmd_mem.c b/common/cmd_mem.c
>> > index 4daa1b3..e84cc4e 100644
>> > --- a/common/cmd_mem.c
>> > +++ b/common/cmd_mem.c
>> > @@ -471,7 +471,7 @@ int do_mem_base (cmd_tbl_t *cmdtp, int flag, int
>> > argc, char * const argv[])
>> >
>> >  int do_mem_loop (cmd_tbl_t *cmdtp, int flag, int argc, char * const
>> > argv[]) {
>> > -       ulong   addr, length, i, junk;
>> > +       ulong   addr, length, i;
>> >        int     size;
>> >        volatile uint   *longp;
>> >        volatile ushort *shortp;
>> > @@ -518,7 +518,7 @@ int do_mem_loop (cmd_tbl_t *cmdtp, int flag, int
>> > argc, char * const argv[]) longp = (uint *)addr;
>> >                        i = length;
>> >                        while (i-- > 0)
>> > -                               junk = *longp++;
>> > +                               *longp++;
>> >                }
>> >        }
>> >        if (size == 2) {
>> > @@ -526,14 +526,14 @@ int do_mem_loop (cmd_tbl_t *cmdtp, int flag, int
>> > argc, char * const argv[]) shortp = (ushort *)addr;
>> >                        i = length;
>> >                        while (i-- > 0)
>> > -                               junk = *shortp++;
>> > +                               *shortp++;
>> >                }
>> >        }
>> >        for (;;) {
>> >                cp = (u_char *)addr;
>> >                i = length;
>> >                while (i-- > 0)
>> > -                       junk = *cp++;
>> > +                       *cp++;
>> >        }
>> >  }
>>
>> I checked the ARM assembler output and it looks fine.
>>
>> Acked-by: Simon Glass <sjg at chromium.org>
>>
>> Re the 'length == 1' case (above your patch) site, it is assigning to
>> 'i' here. Could/should we remove that assignment also?
>
> The loop below depends on that ... ?

Which loop? It doesn't look like the value of i is used?

Regards,
Simon

>
>>
>> Regards,
>> Simon
>>
>> > --
>> > 1.7.5.4
>> >
>> > _______________________________________________
>> > U-Boot mailing list
>> > U-Boot at lists.denx.de
>> > http://lists.denx.de/mailman/listinfo/u-boot
>