GCC build failure, HEAD@171947 on native
regress
regress@apple.com
Mon Apr 4 18:46:00 GMT 2011
With your recent patch, GCC HEAD revision 171947 had problems on:
native: build (about the same as the previous build)
Attached is build output for those targets.
The previous build was of revision 171943.
Log information for changes since the last build:
------------------------------------------------------------------------
r171944 | steven | 2011-04-04 11:23:27 -0700 (Mon, 04 Apr 2011) | 8 lines
Changed paths:
M /trunk/gcc/ChangeLog
M /trunk/gcc/cprop.c
* cprop.c (oprs_not_set_p): Remove.
(mark_set, mark_clobber): Remove.
(mark_oprs_set): Rewrite using DF_INSN_INFO_DEFS cache.
(reg_not_set_p): New function.
(find_avail_set): Use reg_not_set_p instead of oprs_not_set_p.
(cprop_insn): Likewise.
(cprop_jump): Use FOR_EACH_EDGE.
------------------------------------------------------------------------
r171945 | steven | 2011-04-04 11:24:05 -0700 (Mon, 04 Apr 2011) | 13 lines
Changed paths:
M /trunk/gcc/ChangeLog
M /trunk/gcc/cprop.c
* cprop.c: s/gcse/cprop/ everywhere except for flag_gcse.
(gcse_obstack): Renamed to cprop_obstack.
(GNEW, GNEWVEC, GNEWVAR): Remove.
(gmalloc): Remove.
(alloc_hash_table): Use XNEWVAR instead of GNEWVAR.
(GOBNEW, GOBNEWVAR): Adjust for gcse_obstack renaming.
(gcse_alloc): Likewise, and rename to cprop_alloc.
(alloc_gcse_men, free_gcse_mem): Remove.
(gcse_constant_p): Rename to cprop_constant_p, and adjust all callers.
(compute_hash_table_work): Allocate and free reg_set_bitmap here.
(one_cprop_pass): Likewise. Adjust for gcse_obstack renaming.
------------------------------------------------------------------------
r171946 | steven | 2011-04-04 11:24:50 -0700 (Mon, 04 Apr 2011) | 9 lines
Changed paths:
M /trunk/gcc/ChangeLog
M /trunk/gcc/cprop.c
* cprop.c (implicit_set_cond_p): Assume nothing about COND, move
checks on form of COND from find_implicit_sets to here.
(find_implicit_sets): Cleanup control flow. Split critical edges
if it exposes implicit sets. Allocate/resize implicit_sets as
necessary.
(one_cprop_pass): Only delete unreachable blocks if local_cprop_pass
changed something. Run df_analyze after find_implicit_sets if any
edges were split. Do not allocate implicit_sets here.
------------------------------------------------------------------------
r171947 | steven | 2011-04-04 11:27:17 -0700 (Mon, 04 Apr 2011) | 15 lines
Changed paths:
M /trunk/gcc/ChangeLog
M /trunk/gcc/cprop.c
* cprop.c (struct expr): Split 'expr' field in 'dest' and 'src'.
(expr_equiv_p): Remove.
(insert_set_in_table): Look at <dest, src> pair instead of expr.
(hash_scan_set): Update call to insert_set_in_table.
(dump_hash_table): Dump <dest, src> pair.
(lookup_set): Simplify. Lookup <dest, src> pair.
(compute_transp): Remove, fold heavily simplified code into...
(compute_local_properties): ...here. Expect COMP and TRANSP
unconditionally.
(find_avail_set): Take set directly from struct expr.
(find_bypass-set): Likewise.
(bypass_block): Likewise.
(cprop_insn): Likewise. Remove redundant INSN_P test.
------------------------------------------------------------------------
For more information, see <http://gcc.gnu.org/regtest/HEAD/>.
-------------- next part --------------
/Users/regress/tbox/svn-gcc/gcc/config/darwin-crt3.c: In function 'atexit':
/Users/regress/tbox/svn-gcc/gcc/config/darwin-crt3.c:529:29: warning: taking address of expression of type 'void' [enabled by default]
# Early copyback; see "all" above for the rationale. The
# early copy is necessary so that the gcc -B options find
# the right startup files when linking shared libgcc.
/bin/sh /Users/regress/tbox/svn-gcc/libgcc/../mkinstalldirs ../.././gcc
parts="crt2.o crt3.o"; \
for file in $parts; do \
rm -f ../.././gcc/$file; \
/usr/bin/install -c -m 644 $file ../.././gcc/; \
case $file in \
*.a) \
ranlib -c ../.././gcc/$file ;; \
esac; \
done
make[2]: *** [all-stage1-target-libgcc] Error 2
make[1]: *** [stage1-bubble] Error 2
make: *** [bootstrap] Error 2
+ '[' -s .bad_compare ']'
+ exit 1
-------------- next part --------------
--
Geoffrey Keating <geoffk@apple.com>
(via an automated GCC regression-testing script.)
More information about the Gcc-regression
mailing list