| Submitter | Patrick Georgi |
|---|---|
| Date | 2010-06-22 15:16:52 |
| Message ID | <4C20D3E4.7080606@georgi-clan.de> |
| Download | mbox | patch |
| Permalink | /patch/1540/ |
| State | Accepted |
| Headers | show |
Comments
Patrick Georgi wrote: > attached patch tells lpgcc to add two defines to the compiler it > invokes: __LIBPAYLOAD__ and an architecture specific one (__i386__ or > __powerpc__) > > > Signed-off-by: Patrick Georgi <patrick.georgi@coresystems.de> Acked-by: Peter Stuge <peter@stuge.se>
On 6/22/10 5:16 PM, Patrick Georgi wrote: > Hi, > > attached patch tells lpgcc to add two defines to the compiler it > invokes: __LIBPAYLOAD__ and an architecture specific one (__i386__ or > __powerpc__) > > > Signed-off-by: Patrick Georgi <patrick.georgi@coresystems.de> > Isn't __i386__ or __powerpc__ already set by the compiler? Anyways, Acked-by: Stefan Reinauer <stepan@coresystems.de>
On Wed, Jun 23, 2010 at 5:46 AM, Stefan Reinauer <stepan@coresystems.de> wrote: > On 6/22/10 5:16 PM, Patrick Georgi wrote: > > Hi, > > attached patch tells lpgcc to add two defines to the compiler it > invokes: __LIBPAYLOAD__ and an architecture specific one (__i386__ or > __powerpc__) Yes it is. Or something is. You should use what the compiler sets IMHO. ron
On 23.06.2010 14:52, ron minnich wrote: > On Wed, Jun 23, 2010 at 5:46 AM, Stefan Reinauer <stepan@coresystems.de> wrote: > >> On 6/22/10 5:16 PM, Patrick Georgi wrote: >> >>> attached patch tells lpgcc to add two defines to the compiler it >>> invokes: __LIBPAYLOAD__ and an architecture specific one (__i386__ or >>> __powerpc__) >> >> Isn't __i386__ or __powerpc__ already set by the compiler? >> > Yes it is. Or something is. You should use what the compiler sets IMHO. > I just checked and my gcc 4.2.1 always knows the following #defines on i386, even if you specify -nostdlib -nostdinc. #define __i386 1 #define __i386__ 1 #define i386 1 Patrick, which gcc version needs the additional __i386__ #define? Regards, Carl-Daniel
Am 24.06.2010 16:26, schrieb Carl-Daniel Hailfinger:
> Patrick, which gcc version needs the additional __i386__ #define?
Hmm.. Tried it, none.. No idea why that didn't work last time.
Shall I revert it?
Patrick
On 24.06.2010 16:31, Patrick Georgi wrote: > Am 24.06.2010 16:26, schrieb Carl-Daniel Hailfinger: > >> Patrick, which gcc version needs the additional __i386__ #define? >> > Hmm.. Tried it, none.. No idea why that didn't work last time. > > Shall I revert it? > I think the rest of your patch is OK, so I'm against a complete revert. If you revert only the __i386__ hunk, you have my Acked-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006@gmx.net> Thanks. Regards, Carl-Daniel
Patch
Index: bin/lpgcc =================================================================== --- bin/lpgcc (Revision 5638) +++ bin/lpgcc (Arbeitskopie) @@ -103,14 +103,16 @@ if [ "$CONFIG_TARGET_I386" = "y" ]; then _ARCHINCDIR=$_INCDIR/i386 _ARCHLIBDIR=$_LIBDIR/i386 + _TARGETCFLAGS="-D__i386__=1" fi if [ "$CONFIG_TARGET_POWERPC" = "y" ]; then _ARCHINCDIR=$_INCDIR/powerpc _ARCHLIBDIR=$_LIBDIR/powerpc + _TARGETCFLAGS="-D__powerpc__=1" fi -_CFLAGS="-m32 -nostdinc -nostdlib -I$_INCDIR -I$_ARCHINCDIR" +_CFLAGS="-m32 -nostdinc -nostdlib -I$_INCDIR -I$_ARCHINCDIR -D__LIBPAYLOAD__=1 $_TARGETCFLAGS" # Check for the -fno-stack-protector silliness
Hi, attached patch tells lpgcc to add two defines to the compiler it invokes: __LIBPAYLOAD__ and an architecture specific one (__i386__ or __powerpc__) Signed-off-by: Patrick Georgi <patrick.georgi@coresystems.de>