[Bug driver/52556] Ability to change GLIBC_DYNAMIC_LINKER
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=52556 Andrew Pinski changed: What|Removed |Added CC||hdante at gmail dot com --- Comment #7 from Andrew Pinski --- *** Bug 111084 has been marked as a duplicate of this bug. ***
[Bug driver/52556] Ability to change GLIBC_DYNAMIC_LINKER
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52556 Brooks Moses brooks at gcc dot gnu.org changed: What|Removed |Added CC||brooks at gcc dot gnu.org --- Comment #6 from Brooks Moses brooks at gcc dot gnu.org --- Richard Biener mentioned that Red Hat has a local patch that allows amending link_spec from an external file, which would sound more flexible here: http://patchwork.ozlabs.org/patch/80538/ It's not completely clear whether it applies to this case, however; he mentioned that it could be used to produce the equivalent of a -Wl,-rpath=... option. The Google-branch change that Christer mentions adds a -dynamic-linker=... option to the link spec rather than an -rpath option -- I would guess that the patch Richard mentions could cover that as well, but I'm not sure.
[Bug driver/52556] Ability to change GLIBC_DYNAMIC_LINKER
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52556
--- Comment #5 from Benda Xu heroxbd at sohu dot com ---
Dear Jakub,
I find your comment misleading for the --with-specs option.
It sets CONFIGURE_SPECS macro, which is used in gcc.c as
static const char *const driver_self_specs[] = {
%{fdump-final-insns:-fdump-final-insns=.} %fdump-final-insns,
DRIVER_SELF_SPECS, CONFIGURE_SPECS, GOMP_SELF_SPECS, GTM_SELF_SPECS
};
so it is a list of specs for the driver *itself*, should not be used to specify
link spec.
If you mean passing --with-specs=-Wl,-dynamic-linker=/whatever/ld.so to
configure, it is broken in multilib cases.
Therefore we lack of a convenient way to specify a default alternative location
of the dynamic linker.
Benda
[Bug driver/52556] Ability to change GLIBC_DYNAMIC_LINKER
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52556
Benda Xu heroxbd at sohu dot com changed:
What|Removed |Added
CC||heroxbd at sohu dot com
--- Comment #4 from Benda Xu heroxbd at sohu dot com ---
(In reply to Jakub Jelinek from comment #2)
Just use -Wl,-dynamic-linker=/whatever/ld.so
-Wl,-dynamic-linker=/whatever/ld.so cannot be set permanently. It requires a
wrapper to inject such argument all the time, which is ugly.
or you could use --with-specs configure option.
Builtin specs of GCC is already very complicated (output from gcc-4.7.3
-dumpspecs on amd64):
...
*link:
%{!static:--eh-frame-hdr} %{m32|mx32:;:-m elf_x86_64}
%{m32:-m elf_i386}%{mx32:-m elf32_x86_64}
%{shared:-shared} %{!shared: %{!static: %{rdynamic:-export-dynamic}
%{m32:-dynamic-linker
%{muclibc:/lib/ld-uClibc.so.0;:%{mbionic:/system/bin/linker;:/lib/ld-linux.so.2}}}
%{m32|mx32:;:-dynamic-linker
%{muclibc:/lib/ld64-uClibc.so.0;:%{mbionic:/system/bin/linker64;:/lib64/ld-linux-x86-64.so.2}}}
%{mx32:-dynamic-linker
%{muclibc:/lib/ldx32-uClibc.so.0;:%{mbionic:/system/bin/linkerx32;:/libx32/ld-linux-x32.so.2
%{static:-static}}
...
writing an equivalent specs with prefixed dynamic linker is horrible.
Please consider sopport dynamic linker prefix by an extra configure option.
[Bug driver/52556] Ability to change GLIBC_DYNAMIC_LINKER
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52556 --- Comment #1 from Christer Solskogen christer.solskogen at gmail dot com 2012-03-11 12:54:45 UTC --- I see now that the google gcc branch is having this kind of ability. Any plans of integrating them?
[Bug driver/52556] Ability to change GLIBC_DYNAMIC_LINKER
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52556 Jakub Jelinek jakub at gcc dot gnu.org changed: What|Removed |Added CC||jakub at gcc dot gnu.org --- Comment #2 from Jakub Jelinek jakub at gcc dot gnu.org 2012-03-11 14:14:28 UTC --- Just use -Wl,-dynamic-linker=/whatever/ld.so or you could use --with-specs configure option.
[Bug driver/52556] Ability to change GLIBC_DYNAMIC_LINKER
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=52556 --- Comment #3 from Christer Solskogen christer.solskogen at gmail dot com 2012-03-11 15:50:57 UTC --- I could, but is not a sexy(*) as --with-runtime-root-prefix :-) *) http://patchwork.ozlabs.org/patch/80538/
