Can I submit simple patches like this to the primary ML?
greg at kroah.com
Thu Jan 29 00:07:54 EST 2015
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 at kroah.com> 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 at gmail.com>
> >> 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 at gmail.com>
> >> ---
> >> 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
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
> 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 :)
More information about the Kernelnewbies