--- Additional Comments From pluto at pld-linux dot org 2005-04-15 07:20
---
(In reply to comment #11)
full link command:
/usr/bin/ld --eh-frame-hdr -m elf_i386 -shared -o .libs/wmfthumbnail.so
-z combreloc
-L/usr/lib/gcc/i686-pld-linux/4.0.0
--- Additional Comments From pluto at pld-linux dot org 2005-04-14 16:05
---
print *h-u.undef.abfd
$15 = {id = 246, filename = 0x80cf5a0 /usr/X11R6/lib/libX11.so,
xvec = 0x400adf60, iostream = 0x84ee0a0, iovec = 0x400aa6f0,
cacheable = 1, target_defaulted = 0,
--- Additional Comments From amodra at bigpond dot net dot au 2005-04-15
02:52 ---
No, (b) is wrong. elf.c:1177 won't result in the sym being marked as defined in
the as-needed shared lib. I need to dig some more (or have those readelf -sD
results!)
--
--- Additional Comments From j at bitron dot ch 2005-04-13 09:15 ---
Problem appears when using --as-needed as e.g. gcc inserts automatically around
-lgcc_s. It seems to have been introduced by
http://sourceware.org/ml/binutils-cvs/2005-02/msg8.html and happens with FSF
binutils
--- Additional Comments From gschafer at zip dot com dot au 2005-04-13
11:04 ---
This seems to be the cause of build failure when compiling xorg-6.8.2:
gcc -m32 -o glxinfo -O2 -march=i686 -fno-strict-aliasing -ansi -pedantic
-Wall -Wpointer-arith -Wundef -L../../exports/lib
--- Additional Comments From j at bitron dot ch 2005-04-13 16:39 ---
Part of the problem seems to have been solved in latest binutils mainline, i.e.
the first test case works fine now.
But linking glxinfo of Xorg 6.8.2 still fails although with other error
messages:
gcc -m32 -o
--- Additional Comments From pluto at pld-linux dot org 2005-04-13 22:24
---
(In reply to comment #4)
Please try latest mainline binutils. My 2005-04-11 ld/emultempl/elf32.em
patch
may well have fixed this problem.
I've built the binutils-2.16.90.0.1 with patch from ML:
--- Additional Comments From amodra at bigpond dot net dot au 2005-04-14
03:34 ---
comment #5 just seems to be a simple case of ld not finding a library, probably
because the right options were not passed to ld.
comment #6 is more worrying. Can you provided a reduced testcase?
