Skip to contentSkip to conversation

glibc 2.7 fails to build

PATCH In review Roman I Khimov 2 participants 3 messages Feb 13, 2009

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 glibc 2.7 fails to build

Hello.

As I've mentioned some days ago, I'd submitted several bugs via bugzilla and as there is zero reaction on it and in general direction to completely stop using it, I'm reposting that patches here.

This one is bug 5000 (http://bugs.openembedded.net/show_bug.cgi?id=5000):

glibc 2.7 fails to build with
===========
In file included from ../include/limits.h:125,

                 from ../nptl/sysdeps/pthread/allocalim.h:21,
                 from ../include/alloca.h:20,
                 from ../stdlib/stdlib.h:497,
                 from ../include/stdlib.h:8,
                 from ../nptl/sysdeps/i386/i686/../tls.h:28,
                 from ../nptl/sysdeps/i386/i686/tls.h:34,
                 from ../include/tls.h:6,
                 from ../sysdeps/unix/sysv/linux/i386/sysdep.h:30,
                 from <stdin>:1:

/home/rik/work/altell/distro/oe/tmp/staging/i686-linux/usr/include/limits.h:125:26:
error: no include path in which to search for limits.h
===============
Configuration tried was for x86-32-nocona.

Fixed easily by applying same glibc-use-isystem-include-fixed.patch as used with 2.6.1, like in patch attached

packages/glibc/glibc_2.7.bb +2 -1
index 3349ce5..2bd4fa4 100644
@@ -65,7 +65,8 @@ SRC_URI = "ftp://ftp.gnu.org/pub/gnu/glibc/glibc-${PV}.tar.bz2 \
 	   file://generic-bits_typesizes.h \
 	   file://generic-bits_time.h \
            file://etc/ld.so.conf \
-           file://generate-supported.mk"
+           file://generate-supported.mk \
+           file://glibc-use-isystem-include-fixed.patch;patch=1"
 
 
 # Build fails on sh3 and sh4 without additional patches
Signature · 4 lines
                            Roman
 http://roman.khimov.ru
mailto: [email protected]
gpg --keyserver hkp://subkeys.pgp.net --recv-keys 0xE5E055C3
#1 lore raw
Quoted 5 lines · Roman I Khimov

Hello.

As I've mentioned some days ago, I'd submitted several bugs via bugzilla and as there is zero reaction on it and in general direction to completely stop using it, I'm reposting that patches here.

Your patch looks good to go in, but I have a general remark, which will save you a lot of trouble:

Quoted 1 line

/home/rik/work/altell/distro/oe/tmp/staging/i686-linux/usr/include/limits.h:125:26:

Set TARGET_VENDOR (or use angstrom, it does that for you) so the native system (i686-linux) and the target (i686-linux) don't share the same staging area.

regards,

Koen

#2 lore raw

Koen Kooi:

Quoted 7 lines · Roman I Khimov
Quoted 2 lines

/home/rik/work/altell/distro/oe/tmp/staging/i686-linux/usr/include/limi ts.h:125:26:

Set TARGET_VENDOR (or use angstrom, it does that for you) so the native system (i686-linux) and the target (i686-linux) don't share the same staging area.

Understood, thanks. We're building on x86_64, so they're not shared in this case, but for our distro in general it is a bug.

Signature · 4 lines
                            Roman
 http://roman.khimov.ru
mailto: [email protected]
gpg --keyserver hkp://subkeys.pgp.net --recv-keys 0xE5E055C3