On Tue, 2010-01-05 at 00:08 +0100, Maxime Henrion wrote: > On Mon, 2010-01-04 at 09:36 +0000, Simon Marlow wrote: > > On 03/01/10 23:19, Maxime Henrion wrote: > > > At that point I feel like this patch would be worth committing, minus > > > the libffi/ghc.mk hack and possibly name changes, so I'm even more eager > > > to hear about the advice of the GHC community. > > > > I'd prefer to use something like elf_OBJ_FORMAT rather than > > elf_TARGET_OS. We already have an OBJFORMAT_ELF in rts/Linker.c that > > could be commoned up. Apart from this, the patch looks fine to me, I'd > > be happy to see it go in. > > That was clearly a poor named, I now changed it to elf_OBJ_FORMAT per > your suggestion. > > > Regarding the warnings, I think you may want to tweak this #ifdef in > > nativeGen/X86/Ppr.hs: > > > > pprTypeAndSizeDecl :: CLabel -> Doc > > #if linux_TARGET_OS > > pprTypeAndSizeDecl lbl > > | not (externallyVisibleCLabel lbl) = empty > > | otherwise = ptext (sLit ".type ") <> > > pprCLabel_asm lbl <> ptext (sLit ", @object") > > #else > > pprTypeAndSizeDecl _ > > = empty > > #endif > > Thanks! I fixed that one and I no longer see any warnings when linking > code. > > I'm attaching an updated patch to this mail.
Should I file a ticket to trac so that we don't forget about this? Cheers, Maxime _______________________________________________ Cvs-ghc mailing list Cvs-ghc@haskell.org http://www.haskell.org/mailman/listinfo/cvs-ghc
