18138 – [3.4/4.0 regression] libgcc_s.so.1 not found by 64-bit testsuite (original) (raw)
Description David Billinghurst 2004-10-25 03:33:51 UTC
The following patch breaks the 64-bit testsuite for mips-sgi-irix6.5. I have tracked it down in HEAD but the same patch causes a similar problem in the 3.4 branch.
2004-11-18 H.J. Lu <hongjiu.lu@intel.com>
[PR bootstrap/17684](show%5Fbug.cgi?id=17684 "RESOLVED FIXED - [3.4/4.0 Regression] /usr/ccs/bin/ld: Can't create libgcc_s.sl")
* Makefile.in (clean): Remove libgcc_s$(SHLIB_EXT).1.stage?.
(stage1-start): Remove and copy libunwind.a and
libunwind*$(SHLIB_EXT) instead of libunwind*.
(stage2-start): Likewise.
(stage3-start): Likewise.
(stage4-start): Likewise.
(stageprofile-start): Likewise.
(stagefeedback-start): Likewise.
* config/alpha/t-osf4 (SHLIB_LINK): Use a temporary file for
the shared library to be created and don't remove the existing
shared library.
* config/arm/t-netbsd (SHLIB_LINK): Likewise.
* config/i386/t-nwld (SHLIB_LINK): Likewise.
* config/mips/t-slibgcc-irix (SHLIB_LINK): Likewise.
* config/pa/t-hpux-shlib (SHLIB_LINK): Likewise.
* config/sh/t-linux (SHLIB_LINK): Likewise.
* config/t-libunwind-elf (SHLIBUNWIND_LINK): Likewise.
* config/t-slibgcc-darwin (SHLIB_LINK): Likewise.
* config/t-slibgcc-elf-ver (SHLIB_LINK): Likewise.
* config/t-slibgcc-sld (SHLIB_LINK): Likewise.
* mklibgcc.in (libgcc-stage-start): Also move "*${objext}s"
files.After this patch, the -mabi=64 gfortran testsuite reports the error
303790545:./assignment_1.exe: rld: Fatal Error: Cannot Successfully map soname 'libgcc_s.so.1' under any of the filenames /disk4/billingd/obj/gcc- f/gcc/libgcc_s.so.1:./libgcc_s.so.1:/disk4/billingd/obj/gcc-f/mips-sgi- irix6.5/mabi- 64/libgfortran/.libs/libgcc_s.so.1:/usr/local/lib/libgcc_s.so.1:/usr/lib64/libg cc_s.so.1:/usr/lib64/internal/libgcc_s.so.1:/lib64/libgcc_s.so.1:/opt/lib64/lib gcc_s.so.1:
and a similar problem occurs with the g++ testsuite.
Before the patch and configured with --enable-languages=c,f95 the follwoing shared libraries are built.
./gcc/libgcc_s.so ./gcc/libgcc_s_mabi-64.so ./gcc/libgcc_s_mabi-32.so ./gcc/libgmp.so.4 ./gcc/mabi-32/libgcc_s.so.1 ./gcc/mabi-64/libgcc_s.so.1 ./gcc/libgcc_s.so.1 ./mips-sgi-irix6.5/libgfortran/.libs/libgfortran.so.1.0 ./mips-sgi-irix6.5/libgfortran/.libs/libgfortran.so.1 ./mips-sgi-irix6.5/libgfortran/.libs/libgfortran.so ./mips-sgi-irix6.5/mabi-32/libgfortran/.libs/libgfortran.so.1.0 ./mips-sgi-irix6.5/mabi-32/libgfortran/.libs/libgfortran.so.1 ./mips-sgi-irix6.5/mabi-32/libgfortran/.libs/libgfortran.so ./mips-sgi-irix6.5/mabi-64/libgfortran/.libs/libgfortran.so.1.0 ./mips-sgi-irix6.5/mabi-64/libgfortran/.libs/libgfortran.so.1 ./mips-sgi-irix6.5/mabi-64/libgfortran/.libs/libgfortran.so
after the patch I get
find . -name ".so" ./gcc/libgcc_s.so ./gcc/libgcc_s.so.1.stage1 ./gcc/libgcc_s_mabi-32.so ./gcc/libgmp.so.4 ./gcc/libgcc_s.so.1.stage2 ./gcc/mabi-32/libgcc_s.so.1.stage2 ./gcc/mabi-32/libgcc_s.so.1 ./gcc/mabi-64/libgcc_s.so.1.stage2 ./gcc/mabi-64/libgcc_s.so.1 ./gcc/libgcc_s_mabi-64.so ./gcc/libgcc_s.so.1 ./mips-sgi-irix6.5/libgfortran/.libs/libgfortran.so.1.0 ./mips-sgi-irix6.5/libgfortran/.libs/libgfortran.so.1 ./mips-sgi-irix6.5/libgfortran/.libs/libgfortran.so ./mips-sgi-irix6.5/mabi-32/libgfortran/.libs/libgfortran.so.1.0 ./mips-sgi-irix6.5/mabi-32/libgfortran/.libs/libgfortran.so.1 ./mips-sgi-irix6.5/mabi-32/libgfortran/.libs/libgfortran.so ./mips-sgi-irix6.5/mabi-64/libgfortran/.libs/libgfortran.so.1.0 ./mips-sgi-irix6.5/mabi-64/libgfortran/.libs/libgfortran.so.1 ./mips-sgi-irix6.5/mabi-64/libgfortran/.libs/libgfortran.so
Comment 2 david.billinghurst@comalco.riotinto.com.au 2004-10-25 03:56:11 UTC
Subject: RE: [3.4/4.0 regression] libgcc_s.so.1 not found by 64-bit testsuite
The gfortran.exp part works for me on mips-sgi-irix6.5. Will try the rest overnight.
DavidNOTICE This e-mail and any attachments are private and confidential and may contain privileged information. If you are not an authorised recipient, the copying or distribution of this e-mail and any attachments is prohibited and you must not read, print or act in reliance on this e-mail or attachments. This notice should not be removed.
Comment 3 Eric Botcazou 2004-10-25 06:09:51 UTC
This is the same problem as on Solaris. I should have done my homework...