[Bug 936563] New: openssl build failure due to debugedit segfault on ppc64/ppc64le
http://bugzilla.suse.com/show_bug.cgi?id=936563 Bug ID: 936563 Summary: openssl build failure due to debugedit segfault on ppc64/ppc64le Classification: openSUSE Product: openSUSE Factory Version: 201505* Hardware: PowerPC-64 OS: SUSE Other Status: NEW Severity: Normal Priority: P5 - None Component: Printing Assignee: jsmeix@suse.com Reporter: normand@linux.vnet.ibm.com QA Contact: jsmeix@suse.com Found By: --- Blocker: --- openssl build failure due to debugedit segfault on ppc64/ppc64le The failure is reported in OBS (1) and doing an osc local build adding traces is pointing to a segfault of debugedit (2) Using gdb gives the associated backtrace in rpm/ (1) https://build.opensuse.org/package/live_build_log/openSUSE:Factory:PowerPC/o... === related log extract: [ 400s] + /usr/lib/rpm/find-debuginfo.sh /home/abuild/rpmbuild/BUILD/openssl-1.0.2a [ 400s] extracting debug info from /home/abuild/rpmbuild/BUILDROOT/openssl-1.0.2a-3.4.ppc64le/lib64/engines/libgost.so [ 400s] extracting debug info from /home/abuild/rpmbuild/BUILDROOT/openssl-1.0.2a-3.4.ppc64le/lib64/libcrypto.so.1.0.0 [ 400s] error: Bad exit status from /var/tmp/rpm-tmp.yS67uz (%install) [ 400s] === (2) === ++ /usr/lib/rpm/debugedit -b /home/abuild/rpmbuild/BUILD -d /usr/src/debug -i -l /home/abuild/rpmbuild/BUILD/openssl-1.0.2a/debugsources.list /home/abuild/rpmbuild/BUILDROOT/openssl-1.0.2a-0.ppc64le/lib64/libcrypto.so.1.0.0 + id= + exit + exit abuild@twppc64le:~/rpmbuild/BUILD/openssl-1.0.2a> echo $? 139 === (3) === abuild@twppc64le:~/rpmbuild/BUILDROOT/openssl-1.0.2a-0.ppc64le/usr/lib64> gdb /usr/lib/rpm/debugedit ... (gdb) r -b /home/abuild/rpmbuild/BUILD -d /usr/src/debug -i -l /home/abuild/rpmbuild/BUILD/openssl-1.0.2a/debugsources.list /home/abuild/rpmbuild/BUILDROOT/openssl-1.0.2a-0.ppc64le/lib64/libcrypto.so.1.0.0 Starting program: /usr/lib/rpm/debugedit -b /home/abuild/rpmbuild/BUILD -d /usr/src/debug -i -l /home/abuild/rpmbuild/BUILD/openssl-1.0.2a/debugsources.list /home/abuild/rpmbuild/BUILDROOT/openssl-1.0.2a-0.ppc64le/lib64/libcrypto.so.1.0.0 Missing separate debuginfos, use: zypper install glibc-debuginfo-2.21-6.25.ppc64le [Thread debugging using libthread_db enabled] Using host libthread_db library "/lib64/libthread_db.so.1". Program received signal SIGSEGV, Segmentation fault. edit_attributes (phase=<optimized out>, t=<optimized out>, ptr=0x3fffb738987f "", dso=<optimized out>) at tools/debugedit.c:890 890 tools/debugedit.c: No such file or directory. Missing separate debuginfos, use: zypper install libbz2-1-debuginfo-1.0.6-35.81.ppc64le libelf1-debuginfo-0.161-2.53.ppc64le liblua5_1-debuginfo-5.1.5-12.6.ppc64le liblzma5-debuginfo-5.2.1-1.74.ppc64le libpopt0-debuginfo-1.16-28.58.ppc64le libz1-debuginfo-1.2.8-7.68.ppc64le (gdb) bt #0 edit_attributes (phase=<optimized out>, t=<optimized out>, ptr=0x3fffb738987f "", dso=<optimized out>) at tools/debugedit.c:890 #1 edit_dwarf2 (dso=<optimized out>) at tools/debugedit.c:1345 #2 main (argc=<optimized out>, argv=<optimized out>) at tools/debugedit.c:1685 === -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=936563
Dinar Valeev
http://bugzilla.suse.com/show_bug.cgi?id=936563
http://bugzilla.suse.com/show_bug.cgi?id=936563#c1
Johannes Meixner
http://bugzilla.suse.com/show_bug.cgi?id=936563
http://bugzilla.suse.com/show_bug.cgi?id=936563#c2
--- Comment #2 from Michel Normand
This is no "Printing" issue.
Oups, sorry wrong selection in pull-down :( * a trial of local build of Base:System/openssl with last req (1) is still failing the same way. * a trial of local build Base:System/rpm with last req (2) and using it for openssl local build (-p option) is still failing the same way. (rpm package provides the debugedit that segfault). (1) https://build.opensuse.org/request/show/311830 (2) https://build.opensuse.org/request/show/312944 -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=936563
http://bugzilla.suse.com/show_bug.cgi?id=936563#c3
Dinar Valeev
http://bugzilla.suse.com/show_bug.cgi?id=936563
http://bugzilla.suse.com/show_bug.cgi?id=936563#c4
--- Comment #4 from Marcus Meissner
http://bugzilla.suse.com/show_bug.cgi?id=936563
http://bugzilla.suse.com/show_bug.cgi?id=936563#c5
--- Comment #5 from Marcus Meissner
http://bugzilla.suse.com/show_bug.cgi?id=936563
http://bugzilla.suse.com/show_bug.cgi?id=936563#c6
--- Comment #6 from Michel Normand
openssl build failure due to debugedit segfault on ppc64/ppc64le
in fact the problem described by this bug is only for ppc64le. the other failure for ppc64 was a completely different problem tracked by https://build.opensuse.org/request/show/314686 -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=936563
Dinar Valeev
http://bugzilla.suse.com/show_bug.cgi?id=936563
http://bugzilla.suse.com/show_bug.cgi?id=936563#c7
--- Comment #7 from Michel Normand
/dev/null readelf: Warning: DW_FORM_strp offset too big: d000012 readelf: Warning: DW_FORM_strp offset too big: d690287 readelf: Warning: DW_FORM_strp offset too big: 83102000 readelf: Warning: DW_FORM_strp offset too big: 89010000 readelf: Warning: DIE at offset 77ff4 refers to abbreviation number 112 which does not exist === (2) else if ((t->tag == DW_TAG_compile_unit || t->tag == DW_TAG_partial_unit) && t->attr[i].attr == DW_AT_name && form == DW_FORM_strp && debug_sections[DEBUG_STR].data) { char *name;
name = (char *) debug_sections[DEBUG_STR].data + do_read_32_relocated (ptr); if (*name == '/' && comp_dir == NULL) <=== line 890 { === (3) [michel@twppc64le:~/work/Base:System/openssl] $readelf -wi /mnt/disk3/build-root/standard-ppc64le/home/abuild/rpmbuild/BUILD/openssl-1.0.2c/crypto/ec/ecp_nistp224.o
/dev/null readelf: Warning: DW_FORM_strp offset too big: d000012 readelf: Warning: DW_FORM_strp offset too big: d690287 readelf: Warning: DW_FORM_strp offset too big: 2092000 readelf: Warning: DW_FORM_strp offset too big: 89010000 readelf: Warning: DIE at offset 1283 refers to abbreviation number 112 which does not exist === === (4) [ 84s] gcc -I.. -I../.. -I../modes -I../asn1 -I../evp -I../../include -fPIC -DOPENSSL_PIC -DZLIB -DOPENSSL_THREADS -D_REENTRANT -DDSO_DLFCN -DHAVE_DLFCN_H -m64 -DL_ENDIAN -O3 -Wall -fmessage-length=0 -grecord-gcc-switches -O2 -Wall -D_FORTIFY_SOURCE=2 -fstack-protector -funwind-tables -fasynchronous-unwind-tables -g -O3 -std=gnu99 -Wa,--noexecstack -fno-common -DTERMIO -DPURIFY -DSSL_FORBID_ENULL -D_GNU_SOURCE -DOPENSSL_NO_BUF_FREELISTS -Wall -Wa,--noexecstack -DOPENSSL_BN_ASM_MONT -DSHA1_ASM -DSHA256_ASM -DSHA512_ASM -DAES_ASM -DVPAES_ASM -c -o ecp_nistp224.o ecp_nistp224.c ===
-- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=936563
http://bugzilla.suse.com/show_bug.cgi?id=936563#c8
--- Comment #8 from Michel Normand
ecp_nistp224.readelf_wi.txt 2>&1
-- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=936563
Ulrich Weigand
http://bugzilla.suse.com/show_bug.cgi?id=936563
http://bugzilla.suse.com/show_bug.cgi?id=936563#c9
--- Comment #9 from Ulrich Weigand
http://bugzilla.suse.com/show_bug.cgi?id=936563
Hanns-Joachim Uhl
http://bugzilla.suse.com/show_bug.cgi?id=936563
http://bugzilla.suse.com/show_bug.cgi?id=936563#c10
--- Comment #10 from Ulrich Weigand
http://bugzilla.suse.com/show_bug.cgi?id=936563
Marcus Meissner
http://bugzilla.suse.com/show_bug.cgi?id=936563
http://bugzilla.suse.com/show_bug.cgi?id=936563#c11
--- Comment #11 from Ulrich Weigand
This happens with current mainline and current gcc-5-branch on ppc64le, but for some reason not on x86_64.
Actually, it does happen on x86_64 as well, but only on mainline, not in the GCC 5 branch. The bug happens only when CONST_WIDE_INT is being used. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=936563
http://bugzilla.suse.com/show_bug.cgi?id=936563#c12
--- Comment #12 from Ulrich Weigand
http://bugzilla.suse.com/show_bug.cgi?id=936563
http://bugzilla.suse.com/show_bug.cgi?id=936563#c13
--- Comment #13 from Michel Normand
I have opened a GCC bugzilla to track this problem: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66728
Only for information, there is still pending discussion upstream for a proposed patch as per: https://gcc.gnu.org/cgi-bin/search.cgi?form=extended&m=all&q=66728&s=D&ul=/ml/gcc-patches/%25&wm=wrd -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=936563
http://bugzilla.suse.com/show_bug.cgi?id=936563#c14
Michel Normand
participants (1)
-
bugzilla_noreply@novell.com