------- Comment #11 from rguenth at gcc dot gnu dot org 2009-09-03 14:51 ------- Created an attachment (id=18476) --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=18476&action=view) testcase
testcase, _umoddi_s.i from libgcc2.c > valgrind ./cc1 -quiet -O2 -m32 t.i ==1232== Invalid read of size 1 ==1232== at 0x84CA158: reg_preferred_class (reginfo.c:916) ==1232== by 0x8415E19: update_equiv_regs (ira.c:2397) ==1232== by 0x8417A8F: ira (ira.c:3135) ==1232== by 0x8418335: rest_of_handle_ira (ira.c:3343) ==1232== by 0x84911CB: execute_one_pass (passes.c:1292) ==1232== by 0x84913FF: execute_pass_list (passes.c:1341) ==1232== by 0x849141B: execute_pass_list (passes.c:1342) ==1232== by 0x85EEB4B: tree_rest_of_compilation (tree-optimize.c:390) ==1232== by 0x8801650: cgraph_expand_function (cgraphunit.c:1111) ==1232== by 0x8801801: cgraph_expand_all_functions (cgraphunit.c:1170) ==1232== by 0x8801E72: cgraph_optimize (cgraphunit.c:1429) ==1232== by 0x8801355: cgraph_finalize_compilation_unit (cgraphunit.c:1042) ==1232== Address 0x4ac5707 is 0 bytes after a block of size 495 alloc'd ==1232== at 0x40240D2: calloc (vg_replace_malloc.c:397) ==1232== by 0x8B3E40F: xcalloc (xmalloc.c:162) ==1232== by 0x84CA23E: allocate_reg_info (reginfo.c:950) ==1232== by 0x84CA411: reginfo_init (reginfo.c:1003) ==1232== by 0x84911CB: execute_one_pass (passes.c:1292) ==1232== by 0x84913FF: execute_pass_list (passes.c:1341) ==1232== by 0x849141B: execute_pass_list (passes.c:1342) ==1232== by 0x85EEB4B: tree_rest_of_compilation (tree-optimize.c:390) ==1232== by 0x8801650: cgraph_expand_function (cgraphunit.c:1111) ==1232== by 0x8801801: cgraph_expand_all_functions (cgraphunit.c:1170) ==1232== by 0x8801E72: cgraph_optimize (cgraphunit.c:1429) ==1232== by 0x8801355: cgraph_finalize_compilation_unit (cgraphunit.c:1042) not sure if this is the cause of this bug. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=41241