------- Comment #2 from slyfox at inbox dot ru 2010-03-21 20:11 ------- Created an attachment (id=20154) --> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=20154&action=view) valgrind.log - sheds some light on SIGSEGV reason
... ==10526== Invalid write of size 8 ==10526== at 0x105DC344: schedule_block (haifa-sched.c:1963) ==10526== by 0x1029F5B7: schedule_insns (sched-rgn.c:2930) ==10526== by 0x1029F6BF: rest_of_handle_sched (sched-rgn.c:3441) ==10526== by 0x10230E87: execute_one_pass (passes.c:1277) ==10526== by 0x102310C7: execute_pass_list (passes.c:1326) ==10526== by 0x102310DF: execute_pass_list (passes.c:1327) ==10526== by 0x10330A6F: tree_rest_of_compilation (tree-optimize.c:420) ==10526== by 0x1047C99F: cgraph_expand_function (cgraphunit.c:1047) ==10526== by 0x1047E4CB: cgraph_optimize (cgraphunit.c:1106) ==10526== by 0x1001B1E7: c_write_global_declarations (c-decl.c:8111) ==10526== by 0x102E83D3: toplev_main (toplev.c:981) ==10526== by 0x1007C5C3: main (main.c:35) ==10526== Address 0x10 is not stack'd, malloc'd or (recently) free'd ... -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=43470
