Similar to PR target/7042, this bug is about reorg generating a return insn while current_function_epilogue_delay_list. The test-case is newlib/libc/string/wcscspn.c:1.1, which will be entered in the test-suite (including copyright/license note) with modified data-types to avoid included files. Like PR target/7042, this should only affect targets with ELIGIBLE_FOR_EPILOGUE_DELAY Noticed thanks to the sanity check in the CRIS port introduced with the fix for PR target/7042. The error looks like this, cutnpasted: /home/hp/cvs_areas/combined/cris-obj/gcc/xgcc -B/home/hp/cvs_areas/combined/cris-obj/gcc/ -nostdinc -B/home/hp/cvs_areas/combined\ /cris-obj/cris-unknown-elf/v10/newlib/ -isystem /home/hp/cvs_areas/combined/cris-obj/cris-unknown-elf/v10/newlib/targ-include -is\ ystem /home/hp/cvs_areas/combined/combined/newlib/libc/include -B/tmp/cris/cris-unknown-elf/bin/ -B/tmp/cris/cris-unknown-elf/lib\ / -isystem /tmp/cris/cris-unknown-elf/include -isystem /tmp/cris/cris-unknown-elf/sys-include -L/home/hp/cvs_areas/combined/cris-\ obj/ld -march=v10 -mbest-lib-options -DPACKAGE=\"newlib\" -DVERSION=\"1.13.0\" -I. -I/home/hp/cvs_areas/combined/combined/newli\ b/libc/string -O2 -DHAVE_RENAME -DHAVE_GETTIMEOFDAY -D_USE_WRITE -fno-builtin -O2 -g -O2 -O2 -g -O2 -march=v10 -mbest-lib-o\ ptions -c /home/hp/cvs_areas/combined/combined/newlib/libc/string/wcscspn.c /home/hp/cvs_areas/combined/combined/newlib/libc/string/wcscspn.c: In function 'wcscspn': /home/hp/cvs_areas/combined/combined/newlib/libc/string/wcscspn.c:85: internal compiler error: in output_149, at insn-output.c:1086 Please submit a full bug report, with preprocessed source if appropriate.
This bug was first spotted with "Fri Jan 14 01:28:39 UTC 2005", then confirmed on another system with "Fri Jan 14 10:58:15 CET 2005". Patch-testing in progress.
Any smaller testcase would be helpfull :).
Why was severity changed to minor? I changed it back to "normal"; this bug must be fixed, so maybe it should have been "critical". And I'll handle the test-case, don't worry.
Well there is no testcase in this bug.
Hm, I thought I had assigned this to me. Stupid bugzilla.
You know that there is a "view bug activity" which shows why something in the state. Again there is no testcase here, we really need one even though you are fixing it.
Of course we need a test-case. Still, when I've assigned the bug to me _I_will_take_care_of_that_. See also initial text. You're not helping.
Subject: Bug 19462 CVSROOT: /cvs/gcc Module name: gcc Changes by: hp@gcc.gnu.org 2005-01-15 21:30:42 Modified files: gcc/testsuite : ChangeLog Added files: gcc/testsuite/gcc.dg/torture: pr19462-1.c Log message: PR rtl-optimization/19462 * gcc.dg/torture/pr19462-1.c: New test. Patches: http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/ChangeLog.diff?cvsroot=gcc&r1=1.4892&r2=1.4893 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/gcc.dg/torture/pr19462-1.c.diff?cvsroot=gcc&r1=NONE&r2=1.1
Subject: Bug 19462 CVSROOT: /cvs/gcc Module name: gcc Changes by: hp@gcc.gnu.org 2005-01-19 16:39:27 Modified files: gcc : ChangeLog reorg.c Log message: PR rtl-optimization/19462 * reorg.c (find_end_label): Create return insn only if current_function_epilogue_delay_list is empty. Patches: http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/ChangeLog.diff?cvsroot=gcc&r1=2.7186&r2=2.7187 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/reorg.c.diff?cvsroot=gcc&r1=1.102&r2=1.103
Subject: Bug 19462 CVSROOT: /cvs/gcc Module name: gcc Changes by: hp@gcc.gnu.org 2005-01-19 17:04:25 Modified files: gcc/testsuite : ChangeLog gcc/testsuite/gcc.dg/torture: pr19462-1.c Log message: PR rtl-optimization/19462 * gcc.dg/torture/pr19462-1.c: Remove token xfail marker. Patches: http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/ChangeLog.diff?cvsroot=gcc&r1=1.4910&r2=1.4911 http://gcc.gnu.org/cgi-bin/cvsweb.cgi/gcc/gcc/testsuite/gcc.dg/torture/pr19462-1.c.diff?cvsroot=gcc&r1=1.1&r2=1.2
All committed to main trunk. (May reopen for branches.)