svn r128671 causes a bootstrap failure on *apple-darwin: (gdb) r Starting program: /Volumes/development/gcc/head/objdir/gcc/f951 /Volumes/development/gcc/head/gcc/libgfortran/intrinsics/selected_int_kind.f90 -feliminate-unused-debug-symbols -fPIC -quiet -dumpbase selected_int_kind.f90 -mmacosx-version-min=10.4 -auxbase-strip .libs/selected_int_kind.o -g -O2 -Wall -version -fno-repack-arrays -fno-underscoring -fallow-leading-underscore -fno-common -I . -fintrinsic-modules-path finclude -o selected_int_kind.s Reading symbols for shared libraries .++ done GNU F95 (GCC) version 4.3.0 20070923 (experimental) [trunk revision 128695] (powerpc-apple-darwin8.10.0) compiled by GNU C version 4.3.0 20070923 (experimental) [trunk revision 128695], GMP version 4.2.2, MPFR version 2.3.0. GGC heuristics: --param ggc-min-expand=100 --param ggc-min-heapsize=131072 Program received signal EXC_BAD_ACCESS, Could not access memory. Reason: KERN_PROTECTION_FAILURE at address: 0x0000000c 0x00177360 in dbxout_end_source_file (line=4) at /Volumes/development/gcc/head/gcc/gcc/dbxout.c:1181 1181 if (current_file->bincl_status == BINCL_PROCESSED) (gdb) bt #0 0x00177360 in dbxout_end_source_file (line=4) at /Volumes/development/gcc/head/gcc/gcc/dbxout.c:1181 #1 0x00065b28 in gfc_advance_line () at /Volumes/development/gcc/head/gcc/gcc/fortran/scanner.c:322 #2 0x000675b8 in gfc_next_char_literal (in_string=4) at /Volumes/development/gcc/head/gcc/gcc/fortran/scanner.c:754 #3 0x000678d8 in gfc_next_char () at /Volumes/development/gcc/head/gcc/gcc/fortran/scanner.c:926 #4 0x000587f8 in gfc_match_literal_constant (result=0xbfffe900, signflag=0) at /Volumes/development/gcc/head/gcc/gcc/fortran/primary.c:785 #5 0x000449fc in match_mult_operand (result=0xbfffe968) at /Volumes/development/gcc/head/gcc/gcc/fortran/matchexp.c:161 #6 0x00044cf8 in match_add_operand (result=0xbfffe9e8) at /Volumes/development/gcc/head/gcc/gcc/fortran/matchexp.c:382 #7 0x00044f84 in match_level_2 (result=0xbfffea68) at /Volumes/development/gcc/head/gcc/gcc/fortran/matchexp.c:501 #8 0x000450f8 in match_level_3 (result=0xbfffead8) at /Volumes/development/gcc/head/gcc/gcc/fortran/matchexp.c:572 #9 0x00045240 in match_and_operand (result=0xbfffeb58) at /Volumes/development/gcc/head/gcc/gcc/fortran/matchexp.c:620 #10 0x000454ac in match_or_operand (result=0xbfffebc8) at /Volumes/development/gcc/head/gcc/gcc/fortran/matchexp.c:743 #11 0x000455d0 in match_equiv_operand (result=0xbfffec38) at /Volumes/development/gcc/head/gcc/gcc/fortran/matchexp.c:786 #12 0x000456f4 in match_level_5 (result=0xbfffeca8) at /Volumes/development/gcc/head/gcc/gcc/fortran/matchexp.c:832 #13 0x0004489c in gfc_match_expr (result=0x42510c24) at /Volumes/development/gcc/head/gcc/gcc/fortran/matchexp.c:891 #14 0x00008ef8 in gfc_match_array_spec (asp=0xbfffedb4) at /Volumes/development/gcc/head/gcc/gcc/fortran/array.c:297 #15 0x0001a078 in gfc_match_data_decl () at /Volumes/development/gcc/head/gcc/gcc/fortran/decl.c:1456 #16 0x0004fee0 in match_word (str=0x4 <Address 0x4 out of bounds>, subr=0x19e18 <gfc_match_data_decl>, old_locus=0xbfffef0c) at /Volumes/development/gcc/head/gcc/gcc/fortran/parse.c:64 #17 0x000505e8 in decode_statement () at /Volumes/development/gcc/head/gcc/gcc/fortran/parse.c:133 #18 0x000512f4 in next_statement () at /Volumes/development/gcc/head/gcc/gcc/fortran/parse.c:505 #19 0x0005319c in parse_spec (st=ST_DATA_DECL) at /Volumes/development/gcc/head/gcc/gcc/fortran/parse.c:1958 #20 0x00054d18 in parse_progunit (st=ST_BLOCK_DATA) at /Volumes/development/gcc/head/gcc/gcc/fortran/parse.c:2965 #21 0x000552b4 in gfc_parse_file () at /Volumes/development/gcc/head/gcc/gcc/fortran/parse.c:3285 #22 0x0007f684 in gfc_be_parse_file (set_yydebug=4) at /Volumes/development/gcc/head/gcc/gcc/fortran/f95-lang.c:258 #23 0x000df3a8 in toplev_main (argc=6665944, argv=0x406c6380) at /Volumes/development/gcc/head/gcc/gcc/toplev.c:1045 #24 0x00002430 in _start () #25 0x00002134 in start () Darwin is an arch which uses stabs, at least for Tiger and below. Aka, 10.8 and lower.
See PR33502 and http://gcc.gnu.org/ml/gcc-patches/2007-09/msg01721.html for the patch which causes the regression. See also http://gcc.gnu.org/ml/fortran/2007-09/msg00394.html
Subject: Bug 33538 Author: fxcoudert Date: Mon Sep 24 20:24:11 2007 New Revision: 128724 URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=128724 Log: PR fortran/33538 * scanner.c, parse.c, gfortran.h: Revert revision 128671. Modified: trunk/gcc/fortran/ChangeLog trunk/gcc/fortran/gfortran.h trunk/gcc/fortran/parse.c trunk/gcc/fortran/scanner.c
Patch reverted, trunk should be back on track. Sorry.