On Thu, Jan 29, 2015 at 12:27 PM, Vinícius Tinti <viniciustinti@xxxxxxxxx> wrote:
I can simply turn down this kind of warning.On Thu, Jan 29, 2015 at 3:07 AM, Greg KH <greg@xxxxxxxxx> wrote:
> On Thu, Jan 29, 2015 at 02:16:51AM -0200, Vinícius Tinti wrote:
>> On Thu, Jan 29, 2015 at 2:08 AM, Greg KH <greg@xxxxxxxxx> wrote:
>> > On Thu, Jan 29, 2015 at 01:48:43AM -0200, Vinícius Tinti wrote:
>> >> This is a simple patch that initializes a function with NULL to avoid some
>> >> compiler warnings. In such cases should I proceed as a normal patch or it is
>> >> better to send to another ML like to one for trivial patches?
>> >>
>> >> Thanks,
>> >>
>> >> Tinti
>> >
>> >> >From a391789bf44afbdbe2a7b3c76301b5ece9f72475 Mon Sep 17 00:00:00 2001
>> >> From: =?UTF-8?q?Vin=C3=ADcius=20Tinti?= <viniciustinti@xxxxxxxxx>
>> >> Date: Thu, 29 Jan 2015 01:35:34 -0200
>> >> Subject: [PATCH] x86: LLVMLinux: Fix uninitialized function do_reloc
>> >> MIME-Version: 1.0
>> >> Content-Type: text/plain; charset=UTF-8
>> >> Content-Transfer-Encoding: 8bit
>> >>
>> >> Explicit initializes do_reloc function with NULL. Later the function is
>> >> either proper initialized of an error issued.
>> >>
>> >> Signed-off-by: Vinícius Tinti <viniciustinti@xxxxxxxxx>
>> >> ---
>> >> arch/x86/tools/relocs.c | 2 +-
>> >> 1 file changed, 1 insertion(+), 1 deletion(-)
>> >>
>> >> diff --git a/arch/x86/tools/relocs.c b/arch/x86/tools/relocs.c
>> >> index 0c2fae8..1d533f1 100644
>> >> --- a/arch/x86/tools/relocs.c
>> >> +++ b/arch/x86/tools/relocs.c
>> >> @@ -971,7 +971,7 @@ static void emit_relocs(int as_text, int use_real_mode)
>> >> int i;
>> >> int (*write_reloc)(uint32_t, FILE *) = write32;
>> >> int (*do_reloc)(struct section *sec, Elf_Rel *rel, Elf_Sym *sym,
>> >> - const char *symname);
>> >> + const char *symname) = NULL;
>> >
>> > I think you need to get an updated version of the compiler as this patch
>> > should not be needed at all. It doesn't cause a warning here for me
>> > without it.
>>
>> In fact it causes a warning on Clang which complains that:
>>
>> arch/x86/tools/relocs.c:977:6: warning: variable 'do_reloc' is used
>> uninitialized whenever 'if' condition is false
>> [-Wsometimes-uninitialized]
>
> I suggest you file a bug with clang, gcc doesn't have this problem at
> all as obviously, if you look at the code, that variable can never be
> used uninitialized.
>> I think there is not a problem on the current code but to avoid
>> further problems I believe it is worth to initialize this function
>> with NULL.
>> What do you think?
>
> Don't paper over bugs in the compiler with kernel code changes for no
> good reason :)
Agreed. But whenever I find a warning in GCC during the build what
should I do with it?
Can I simply send it to the main ml?
If you believe your patch is correct, and if checkpatch.pl doesn't complain about your patch, then get_maintainers.pl will tell you the people and mailing lists you should send your patch to.
> thanks,
>
> greg k-h
--
Simplicity is the ultimate sophistication
_______________________________________________
Kernelnewbies mailing list
Kernelnewbies@xxxxxxxxxxxxxxxxx
http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
Peter
_______________________________________________ Kernelnewbies mailing list Kernelnewbies@xxxxxxxxxxxxxxxxx http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
