Summary: | segfault in aggregate_value_p while bootstrapping fortran | ||
---|---|---|---|
Product: | gcc | Reporter: | Bernhard Reutner-Fischer <aldot> |
Component: | target | Assignee: | Not yet assigned to anyone <unassigned> |
Status: | RESOLVED FIXED | ||
Severity: | normal | CC: | buytenh, gcc-bugs, k.shutemov, P.Schaffnit, pbrook, tbm |
Priority: | P3 | Keywords: | build, ice-on-valid-code |
Version: | 4.2.0 | ||
Target Milestone: | 4.3.0 | ||
Host: | Target: | arm-linux-uclibcgnueabi,arm-linux-gnueabi | |
Build: | Known to work: | ||
Known to fail: | 4.1.0 4.2.0 | Last reconfirmed: | 2007-01-23 17:53:28 |
Description
Bernhard Reutner-Fischer
2007-01-16 20:10:57 UTC
My backtrace looks slightly different but I can confirm this bug for arm-linux-gnueabi. Lennert Buytenhek reported on debian-arm that fortran fails to bootstrap when building natively and I verified that the testcase also segfaults for me - both with current GCC 4.1 and 4.2 from SVN. Program received signal SIGSEGV, Segmentation fault. aggregate_value_p (exp=0x0, fntype=0x0) at /home/tbm/scratch/gcc-4.2/gcc/function.c:1756 1756 tree type = (TYPE_P (exp)) ? exp : TREE_TYPE (exp); (gdb) where #0 aggregate_value_p (exp=0x0, fntype=0x0) at /home/tbm/scratch/gcc-4.2/gcc/function.c:1756 #1 0x00000000004d9b7b in emit_library_call_value_1 (retval=1, orgfun=0x2b407a3468e0, value=0x0, fn_type=LCT_CONST, outmode=TImode, nargs=<value optimized out>, p=0x7fff30ef3e20) at /home/tbm/scratch/gcc-4.2/gcc/calls.c:3312 #2 0x00000000004da125 in emit_library_call_value (orgfun=0x0, value=0x0, fn_type=LCT_THROW, outmode=TImode, nargs=2) at /home/tbm/scratch/gcc-4.2/gcc/calls.c:3967 #3 0x00000000005d49ba in expand_twoval_binop_libfunc (binoptab=<value optimized out>, op0=0x2b407a363f20, op1=0x2b407a363f00, targ0=0x0, targ1=0x2b407a380320, code=MOD) at /home/tbm/scratch/gcc-4.2/gcc/optabs.c:2283 #4 0x00000000005289cf in expand_divmod (rem_flag=1, code=<value optimized out>, mode=DImode, op0=0x2b407a363f20, op1=0x2b407a363f00, target=0x2b407a363ee0, unsignedp=0) at /home/tbm/scratch/gcc-4.2/gcc/expmed.c:4914 #5 0x000000000052e887 in expand_expr_real_1 (exp=0x2b407a2d4280, target=0x2b407a363ee0, tmode=<value optimized out>, modifier=<value optimized out>, alt_rtl=0x7fff30ef4270) at /home/tbm/scratch/gcc-4.2/gcc/expr.c:8189 #6 0x00000000005327c9 in expand_expr_real (exp=0x2b407a2d4280, target=0x2b407a363ee0, tmode=DImode, modifier=EXPAND_NORMAL, alt_rtl=0x7fff30ef4270) at /home/tbm/scratch/gcc-4.2/gcc/expr.c:6706 #7 0x000000000053783d in store_expr (exp=0x0, target=0x2b407a363ee0, call_param_p=0) at /home/tbm/scratch/gcc-4.2/gcc/expr.c:4370 #8 0x0000000000539147 in expand_assignment (to=0x2b407a37e160, from=<value optimized out>) at /home/tbm/scratch/gcc-4.2/gcc/expr.c:4249 #9 0x000000000052e709 in expand_expr_real_1 (exp=0x2b407a2d4460, target=0x2b407a2d3400, tmode=<value optimized out>, modifier=<value optimized out>, alt_rtl=<value optimized out>) at /home/tbm/scratch/gcc-4.2/gcc/expr.c:8604 #10 0x00000000005326b9 in expand_expr_real (exp=0x2b407a2d4460, target=0x2b407a2d3400, tmode=VOIDmode, modifier=EXPAND_NORMAL, alt_rtl=0x0) at /home/tbm/scratch/gcc-4.2/gcc/expr.c:6700 #11 0x0000000000629049 in expand_expr_stmt (exp=0x0) at /home/tbm/scratch/gcc-4.2/gcc/expr.h:493 #12 0x000000000064fed0 in expand_gimple_basic_block (bb=0x2b407a37f100) at /home/tbm/scratch/gcc-4.2/gcc/cfgexpand.c:1383 #13 0x0000000000650eda in tree_expand_cfg () at /home/tbm/scratch/gcc-4.2/gcc/cfgexpand.c:1649 #14 0x000000000064d4b8 in execute_one_pass (pass=0x9db040) at /home/tbm/scratch/gcc-4.2/gcc/passes.c:870 #15 0x000000000064d60c in execute_pass_list (pass=0x9db040) at /home/tbm/scratch/gcc-4.2/gcc/passes.c:917 #16 0x000000000048c22e in tree_rest_of_compilation (fndecl=0x2b407a37b1c0) Please backport to the 4_2-branch, too. Looks like this was only applied to trunk. (In reply to comment #3) > Please backport to the 4_2-branch, too. Looks like this was only applied to > trunk. So this is not a regression. |