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 :) thanks, greg k-h _______________________________________________ Kernelnewbies mailing list Kernelnewbies@xxxxxxxxxxxxxxxxx http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
