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 can simply turn down this kind of warning. >> 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? > thanks, > > greg k-h -- Simplicity is the ultimate sophistication _______________________________________________ Kernelnewbies mailing list Kernelnewbies@xxxxxxxxxxxxxxxxx http://lists.kernelnewbies.org/mailman/listinfo/kernelnewbies
