[Bug middle-end/50638] [4.7 Regression] emulated TLS fails
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50638 Michael Matz matz at gcc dot gnu.org changed: What|Removed |Added CC||jojelino at gmail dot com --- Comment #14 from Michael Matz matz at gcc dot gnu.org 2011-10-11 11:08:17 UTC --- *** Bug 50658 has been marked as a duplicate of this bug. ***
[Bug middle-end/50638] [4.7 Regression] emulated TLS fails
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50638 --- Comment #12 from Michael Matz matz at gcc dot gnu.org 2011-10-10 11:59:33 UTC --- Author: matz Date: Mon Oct 10 11:59:29 2011 New Revision: 179745 URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=179745 Log: PR middle-end/50638 * tree-emutls.c (gen_emutls_addr): Call add_referenced_var. Modified: trunk/gcc/ChangeLog trunk/gcc/tree-emutls.c
[Bug middle-end/50638] [4.7 Regression] emulated TLS fails
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50638 Michael Matz matz at gcc dot gnu.org changed: What|Removed |Added Status|ASSIGNED|RESOLVED Resolution||FIXED --- Comment #13 from Michael Matz matz at gcc dot gnu.org 2011-10-10 12:54:30 UTC --- AFAIK Fixed.
[Bug middle-end/50638] [4.7 Regression] emulated TLS fails
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50638 --- Comment #11 from Dominique d'Humieres dominiq at lps dot ens.fr 2011-10-08 10:22:30 UTC --- On x86_64-apple-darwin10 the patch in http://gcc.gnu.org/ml/gcc-patches/2011-10/msg00597.html fixes the tls failures (over a thousand in my tests;-).
[Bug middle-end/50638] [4.7 Regression] emulated TLS fails
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50638 Richard Guenther rguenth at gcc dot gnu.org changed: What|Removed |Added Target Milestone|--- |4.7.0 --- Comment #3 from Richard Guenther rguenth at gcc dot gnu.org 2011-10-07 08:29:41 UTC --- I suppose emultls needs another add_referenced_var call.
[Bug middle-end/50638] [4.7 Regression] emulated TLS fails
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50638 Rainer Orth ro at gcc dot gnu.org changed: What|Removed |Added CC||ro at gcc dot gnu.org --- Comment #4 from Rainer Orth ro at gcc dot gnu.org 2011-10-07 16:40:09 UTC --- The same issue breaks bootstrap on Solaris 8 and 9/x86 (so far, SPARC probably too) with Sun as in libgo: /vol/gcc/src/hg/trunk/local/libgo/runtime/go-defer.c: In function '__go_defer': /vol/gcc/src/hg/trunk/local/libgo/runtime/go-defer.c:16:1: internal compiler error: Segmentation Fault
[Bug middle-end/50638] [4.7 Regression] emulated TLS fails
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50638 Michael Matz matz at gcc dot gnu.org changed: What|Removed |Added Status|NEW |ASSIGNED --- Comment #5 from Michael Matz matz at gcc dot gnu.org 2011-10-07 16:53:59 UTC --- Mine. See also http://gcc.gnu.org/ml/gcc-patches/2011-10/msg00597.html . What's the variable that causes the segfault?
[Bug middle-end/50638] [4.7 Regression] emulated TLS fails
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50638 --- Comment #6 from ro at CeBiTec dot Uni-Bielefeld.DE ro at CeBiTec dot Uni-Bielefeld.DE 2011-10-07 16:58:40 UTC --- --- Comment #5 from Michael Matz matz at gcc dot gnu.org 2011-10-07 16:53:59 UTC --- Mine. See also http://gcc.gnu.org/ml/gcc-patches/2011-10/msg00597.html . What's the variable that causes the segfault? Program received signal SIGSEGV, Segmentation fault. set_is_used (tp=0xbf674be4, Parma_Polyhedra_Library14Grid_Generator11set_divisorERK10__gmp_exprIA1_12__mpz_structS3_E=0x80475fc, data=0x0) at /vol/gcc/src/hg/trunk/local/gcc/tree-flow-inline.h:562 562 ann-used = true; (gdb) where #0 set_is_used (tp=0xbf674be4, Parma_Polyhedra_Library14Grid_Generator11set_divisorERK10__gmp_exprIA1_12__mpz_structS3_E=0x80475fc, data=0x0) at /vol/gcc/src/hg/trunk/local/gcc/tree-flow-inline.h:562 #1 mark_all_vars_used_1 (tp=0xbf674be4, Parma_Polyhedra_Library14Grid_Generator11set_divisorERK10__gmp_exprIA1_12__mpz_structS3_E=0x80475fc, data=0x0) at /vol/gcc/src/hg/trunk/local/gcc/tree-ssa-live.c:379 #2 0x087b479a in walk_tree_1 (tp=0xbf674be4, func=0x868f570 mark_all_vars_used_1, ZNK23Parma_Polyhedra_Library6Matrix8num_rowsEv=0x0, pset=0x0, lh=0) at /vol/gcc/src/hg/trunk/local/gcc/tree.c:10457 #3 0x08690152 in mark_all_vars_used () at /vol/gcc/src/hg/trunk/local/gcc/tree-ssa-live.c:595 #4 remove_unused_locals () at /vol/gcc/src/hg/trunk/local/gcc/tree-ssa-live.c:727 #5 0x084e8909 in execute_function_todo ( __gmp_exprIA1_12__mpq_structS3_EE6Status10ascii_loadERSi=0x826) at /vol/gcc/src/hg/trunk/local/gcc/passes.c:1695 #6 0x084e8f44 in mp_exprIA1_12__mpq_structS1_E7set_strEPKci ( _RKNS_17Constraint_SystemEPj=0x84e87b4 execute_function_todo, __gmp_exprIA1_12__mpq_structS3_EE6Status10ascii_loadERSi=0x826) at /vol/gcc/src/hg/trunk/local/gcc/passes.c:1548 #7 0x084e943e in execute_todo (2__mpz_structS7_E=2086) at /vol/gcc/src/hg/trunk/local/gcc/passes.c:1741 #8 0x084ec2da in execute_one_pass (essionENS_15Relation_SymbolES4_=0x923e6a0) at /vol/gcc/src/hg/trunk/local/gcc/passes.c:2087 #9 0x084ec601 in execute_pass_list (essionENS_15Relation_SymbolES4_=0x923e6a0) at /vol/gcc/src/hg/trunk/local/gcc/passes.c:2119 #10 0x084ec614 in execute_pass_list (essionENS_15Relation_SymbolES4_=0x8f6c420) at /vol/gcc/src/hg/trunk/local/gcc/passes.c:2120 #11 0x084ec614 in execute_pass_list (essionENS_15Relation_SymbolES4_=0x8f6bb20) at /vol/gcc/src/hg/trunk/local/gcc/passes.c:2120 #12 0x0860d603 in tree_rest_of_compilation ( 4NNC_PolyhedronEEESaIS3_EE6removeERKS3_=0xbf638f80) at /vol/gcc/src/hg/trunk/local/gcc/tree-optimize.c:420 #13 0x082927d3 in cgraph_expand_function ( ZN23Parma_Polyhedra_Library3BoxINS_8IntervalI10__gmp_exprIA1_12__mpq_structS4_ENS_25Interval_Restriction_NoneINS_20Interval_Info_BitsetIjNS_29Rational_Interval_Info_Policy24generalized_affine_imageENS_8VariableENS_15Relation_SymbolERKNS_17Linear_ExpressionERKS2_IA1_12__mpz_structSJ_E=0xbf5bf668) at /vol/gcc/src/hg/trunk/local/gcc/cgraphunit.c:1805 #14 0x08294a15 in Status8test_allEj () at /vol/gcc/src/hg/trunk/local/gcc/cgraphunit.c:1864 #15 cgraph_optimize () at /vol/gcc/src/hg/trunk/local/gcc/cgraphunit.c:2141 #16 0x08294e7b in cgraph_finalize_compilation_unit () at /vol/gcc/src/hg/trunk/local/gcc/cgraphunit.c:1312 #17 0x08166724 in c_write_global_declarations () at /vol/gcc/src/hg/trunk/local/gcc/c-decl.c:9936 #18 0x0859e922 in compile_file (argc=11, argv=0x804794c) at /vol/gcc/src/hg/trunk/local/gcc/toplev.c:581 #19 do_compile (argc=11, argv=0x804794c) at /vol/gcc/src/hg/trunk/local/gcc/toplev.c:1925 #20 toplev_main (argc=11, argv=0x804794c) at /vol/gcc/src/hg/trunk/local/gcc/toplev.c:2001 #21 0x08cacf9b in main (argc=11, argv=0x804794c) at /vol/gcc/src/hg/trunk/local/gcc/main.c:36 (gdb) up #1 mark_all_vars_used_1 (tp=0xbf674be4, Parma_Polyhedra_Library14Grid_Generator11set_divisorERK10__gmp_exprIA1_12__mpz_structS3_E=0x80475fc, data=0x0) at /vol/gcc/src/hg/trunk/local/gcc/tree-ssa-live.c:379 379 set_is_used (t); (gdb) p t $1 = (tree) 0xbf656e40 (gdb) pt var_decl bf656e40 __emutls_v.__go_panic_defer type record_type bf656de0 __emutls_object BLK size integer_cst bf5b5a48 constant 128 unit size integer_cst bf5b5a64 constant 16 align 32 symtab 0 alias set -1 canonical type bf656de0 fields field_decl bf64bf74 __size type integer_type bf5c73c0 unsigned int unsigned SI file Program received signal SIGSEGV, Segmentation fault. 0xbf6dd359 in strlen () from /usr/lib/libc.so.1 The program being debugged was signaled while in a function called from GDB. GDB remains in the frame where the signal was received. To change this behavior use set unwindonsignal on. Evaluation of the expression containing the function (debug_tree(tree_node*)) will be abandoned. When the function is done executing, GDB
[Bug middle-end/50638] [4.7 Regression] emulated TLS fails
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50638 --- Comment #7 from Michael Matz matz at gcc dot gnu.org 2011-10-07 17:00:59 UTC --- Hmm, I can't build go due to: ../../../gcc/libgo/runtime/sigqueue.goc:79:1: internal compiler error: in maybe_record_trace_start, at dwarf2cfi.c:2243 but that seems unrelated.
[Bug middle-end/50638] [4.7 Regression] emulated TLS fails
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50638 --- Comment #8 from Michael Matz matz at gcc dot gnu.org 2011-10-07 17:02:00 UTC --- Okay, so it's really the emutlv_v variables. That should be fixed by the patch at gcc-patches.
[Bug middle-end/50638] [4.7 Regression] emulated TLS fails
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50638 --- Comment #9 from ro at CeBiTec dot Uni-Bielefeld.DE ro at CeBiTec dot Uni-Bielefeld.DE 2011-10-07 17:03:31 UTC --- --- Comment #7 from Michael Matz matz at gcc dot gnu.org 2011-10-07 17:00:59 UTC --- Hmm, I can't build go due to: ../../../gcc/libgo/runtime/sigqueue.goc:79:1: internal compiler error: in maybe_record_trace_start, at dwarf2cfi.c:2243 but that seems unrelated. Doesn't happen for me on i386-pc-solaris2.[89] as of r179656. Rainer
[Bug middle-end/50638] [4.7 Regression] emulated TLS fails
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50638 --- Comment #10 from ro at CeBiTec dot Uni-Bielefeld.DE ro at CeBiTec dot Uni-Bielefeld.DE 2011-10-07 17:05:11 UTC --- --- Comment #8 from Michael Matz matz at gcc dot gnu.org 2011-10-07 17:02:00 UTC --- Okay, so it's really the emutlv_v variables. That should be fixed by the patch at gcc-patches. I'll try a bootstrap on a different machine, but will only be able to report results on monday. Rainer
[Bug middle-end/50638] [4.7 Regression] emulated TLS fails
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50638 Hans-Peter Nilsson hp at gcc dot gnu.org changed: What|Removed |Added Status|UNCONFIRMED |NEW Last reconfirmed||2011-10-07 CC||hp at gcc dot gnu.org Ever Confirmed|0 |1 --- Comment #1 from Hans-Peter Nilsson hp at gcc dot gnu.org 2011-10-07 00:34:21 UTC --- For cris-elf too, another emutls target. The full build-gcc.sh regression list for cris-elf at r179625 (last known working at r179611): g++.sum g++.dg/tls/static-1.C gcc.sum gcc.dg/debug/tls-1.c gcc.sum gcc.dg/tls/nonpic-1.c gcc.sum gcc.dg/tls/opt-13.c gcc.sum gcc.dg/tls/opt-14.c gcc.sum gcc.dg/tls/opt-5.c gcc.sum gcc.dg/tls/opt-6.c gcc.sum gcc.dg/tls/opt-8.c gcc.sum gcc.dg/tls/pr24428-2.c gcc.sum gcc.dg/tls/pr24428.c gcc.sum gcc.dg/tls/pr45870.c gcc.sum gcc.dg/tls/pr47715-5.c gcc.sum gcc.dg/tls/thr-cse-1.c gcc.sum gcc.dg/torture/tls/run-gd.c gcc.sum gcc.dg/torture/tls/run-ie.c gcc.sum gcc.dg/torture/tls/run-ld.c gcc.sum gcc.dg/torture/tls/run-le.c gcc.sum gcc.dg/torture/tls/thr-init-1.c gcc.sum gcc.dg/torture/tls/thr-init-2.c gfortran.sum gfortran.dg/select_type_12.f03 (the last one *may* be due to a different bug, see PR50640)
[Bug middle-end/50638] [4.7 Regression] emulated TLS fails
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=50638 --- Comment #2 from Hans-Peter Nilsson hp at gcc dot gnu.org 2011-10-07 00:39:03 UTC --- (In reply to comment #1) For cris-elf too, another emutls target. The full build-gcc.sh regression list for cris-elf at r179625 (last known working at r179611): Sorry, I forgot that list is truncated at 20 lines, here are the 3 ones left: objc.sum objc.dg/torture/tls/thr-init-2.m objc.sum objc.dg/torture/tls/thr-init-3.m objc.sum objc.dg/torture/tls/thr-init.m
