https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81968
Bug ID: 81968
Summary: [8 regression] early lto debug objects make Solaris ld
SEGV
Product: gcc
Version: 8.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: lto
Assignee: unassigned at gcc dot gnu.org
Reporter: ro at gcc dot gnu.org
CC: marxin at gcc dot gnu.org, rguenth at gcc dot gnu.org
Target Milestone: ---
Target: *-*-solaris2.11
The Early LTO Debug patches caused a large number of testsuite regressions on
Solaris when using /bin/ld:
+UNRESOLVED: g++.dg/lto/20101010-4 cp_lto_20101010-4_0.o-cp_lto_20101010-4_0.o
execute -std=c++0x -flto -g -r -nostdlib
+FAIL: g++.dg/lto/20101010-4 cp_lto_20101010-4_0.o-cp_lto_20101010-4_0.o link,
-std=c++0x -flto -g -r -nostdlib
They all are like
FAIL: c-c++-common/cilk-plus/CK/cilk-for-2.c -O3 -flto -g (test for excess
errors)
Excess errors:
collect2: fatal error: ld terminated with signal 11 [Segmentation Fault]
compilation terminated.
ld: warning: file /var/tmp//ccOFDEXadebugobjtem: section [1] has invalid type [
SHT_NULL ]
ld: warning: file /var/tmp//ccOFDEXadebugobjtem: section [5] contains both
SHF_EXCLUDE and SHF_ALLOC flags: SHF_EXCLUDE ignored
The same problem happens when gcc is configured to use gas and ld instead of
as/ld, sparc and x86, 32 and 64-bit.
Clearly a SEGV isn't the best error handling, and I'm not yet certain if the
errors
are benign. OTOH, I don't yet know why the objects are created this way.
Rainer
