]> git.ipfire.org Git - thirdparty/gcc.git/log
thirdparty/gcc.git
14 years agoDaily bump.
GCC Administrator [Mon, 6 Sep 2010 00:16:41 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r163890

14 years agoDaily bump.
GCC Administrator [Sun, 5 Sep 2010 00:16:58 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r163870

14 years agoDaily bump.
GCC Administrator [Sat, 4 Sep 2010 00:16:58 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r163851

14 years agoabi-avx.exp: Link with gcc.target/i386/clearcap.map if supported.
Rainer Orth [Fri, 3 Sep 2010 15:07:55 +0000 (15:07 +0000)] 
abi-avx.exp: Link with gcc.target/i386/clearcap.map if supported.

* gcc.target/x86_64/abi/avx/abi-avx.exp: Link with
gcc.target/i386/clearcap.map if supported.

From-SVN: r163829

14 years agoDaily bump.
GCC Administrator [Fri, 3 Sep 2010 00:17:22 +0000 (00:17 +0000)] 
Daily bump.

From-SVN: r163795

14 years agobackport: re PR middle-end/45423 (#pragma omp atomic on bool has issues)
Jakub Jelinek [Thu, 2 Sep 2010 08:00:55 +0000 (10:00 +0200)] 
backport: re PR middle-end/45423 (#pragma omp atomic on bool has issues)

Backport from mainline
2010-08-30  Jakub Jelinek  <jakub@redhat.com>

PR middle-end/45423
* gimplify.c (goa_stabilize_expr): Handle TRUTH_NOT_EXPR
and TRUTH_{AND,OR,XOR}_EXPR.
* c-parser.c (c_parser_omp_atomic): Handle boolean
{PRE,POST}_{INC,DEC}REMENT.
cp/
* parser.c (cp_parser_omp_atomic): Handle boolean
{PRE,POST}_INCREMENT.
testsuite/
* gcc.dg/gomp/atomic-12.c: New test.
* gcc.dg/gomp/atomic-13.c: New test.
* gcc.dg/gomp/atomic-14.c: New test.
* g++.dg/gomp/atomic-11.C: New test.
* g++.dg/gomp/atomic-12.C: New test.
* g++.dg/gomp/atomic-13.C: New test.
* g++.dg/gomp/atomic-14.C: New test.

From-SVN: r163762

14 years agoDaily bump.
GCC Administrator [Thu, 2 Sep 2010 00:16:58 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r163751

14 years agobackport: re PR rtl-optimization/45353 (ICE: RTL check: expected elt 3 type 'B',...
Eric Botcazou [Wed, 1 Sep 2010 12:24:35 +0000 (12:24 +0000)] 
backport: re PR rtl-optimization/45353 (ICE: RTL check: expected elt 3 type 'B', have '0' (rtx barrier) in sel_bb_head, at sel-sched-ir.c:4329 with -fselective-scheduling and __builtin_unreachable())

Backport from mainline
2010-08-20  Jakub Jelinek  <jakub@redhat.com>

PR rtl-optimization/45353
* sel-sched-ir.c (sel_bb_head): Return NULL even if next_nonnote_insn
after bb_note is a BARRIER.

From-SVN: r163731

14 years agogimplify.c (gimplify_init_constructor): Do not create a temporary for a volatile...
Eric Botcazou [Wed, 1 Sep 2010 12:19:11 +0000 (12:19 +0000)] 
gimplify.c (gimplify_init_constructor): Do not create a temporary for a volatile LHS if...

* gimplify.c (gimplify_init_constructor): Do not create a temporary for
a volatile LHS if the constructor has only one element.

From-SVN: r163729

14 years agoDaily bump.
GCC Administrator [Wed, 1 Sep 2010 00:16:38 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r163710

14 years agoDaily bump.
GCC Administrator [Tue, 31 Aug 2010 00:16:44 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r163663

14 years agore PR c++/44991 (default argument with '<' cause compilation error)
Jason Merrill [Mon, 30 Aug 2010 13:38:09 +0000 (09:38 -0400)] 
re PR c++/44991 (default argument with '<' cause compilation error)

PR c++/44991
* parser.c (cp_parser_parameter_declaration): Pop parameter decls
after tentative parsing.

From-SVN: r163644

14 years agoDaily bump.
GCC Administrator [Mon, 30 Aug 2010 00:16:33 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r163633

14 years agoDaily bump.
GCC Administrator [Sun, 29 Aug 2010 00:16:57 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r163622

14 years agore PR target/41484 (Please add memory forms of pmovzx* (SSE4.1))
Uros Bizjak [Sat, 28 Aug 2010 14:27:33 +0000 (16:27 +0200)] 
re PR target/41484 (Please add memory forms of pmovzx* (SSE4.1))

PR target/41484
* config/i386/sse.md (sse4_1_extendv8qiv8hi2): Also accept memory
operands for operand 1.
(sse4_1_extendv4qiv4si2): Ditto.
(sse4_1_extendv2qiv2di2): Ditto.
(sse4_1_extendv4hiv4si2): Ditto.
(sse4_1_extendv2hiv2di2): Ditto.
(sse4_1_extendv2siv2di2): Ditto.
(sse4_1_zero_extendv8qiv8hi2): Ditto.
(sse4_1_zero_extendv4qiv4si2): Ditto.
(sse4_1_zero_extendv2qiv2di2): Ditto.
(sse4_1_zero_extendv4hiv4si2): Ditto.
(sse4_1_zero_extendv2hiv2di2): Ditto.
(sse4_1_zero_extendv2siv2di2): Ditto.
(*sse4_1_extendv8qiv8hi2): Remove insn pattern.
(*sse4_1_extendv4qiv4si2): Ditto.
(*sse4_1_extendv2qiv2di2): Ditto.
(*sse4_1_extendv4hiv4si2): Ditto.
(*sse4_1_extendv2hiv2di2): Ditto.
(*sse4_1_extendv2siv2di2): Ditto.
(*sse4_1_zero_extendv8qiv8hi2): Ditto.
(*sse4_1_zero_extendv4qiv4si2): Ditto.
(*sse4_1_zero_extendv2qiv2di2): Ditto.
(*sse4_1_zero_extendv4hiv4si2): Ditto.
(*sse4_1_zero_extendv2hiv2di2): Ditto.
(*sse4_1_zero_extendv2siv2di2): Ditto.

From-SVN: r163614

14 years agoDaily bump.
GCC Administrator [Sat, 28 Aug 2010 00:17:55 +0000 (00:17 +0000)] 
Daily bump.

From-SVN: r163604

14 years agoDaily bump.
GCC Administrator [Fri, 27 Aug 2010 00:16:52 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r163577

14 years agoDaily bump.
GCC Administrator [Thu, 26 Aug 2010 00:16:30 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r163558

14 years agoDaily bump.
GCC Administrator [Wed, 25 Aug 2010 00:16:53 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r163534

14 years agobackport: re PR rtl-optimization/44691 (ICE: RTL check: expected code 'reg', have...
Andrey Belevantsev [Tue, 24 Aug 2010 09:11:48 +0000 (13:11 +0400)] 
backport: re PR rtl-optimization/44691 (ICE: RTL check: expected code 'reg', have 'plus' in rhs_regno, at rtl.h:1050)

Backport from mainline:
        PR rtl-optimization/44691
        * sel-sched.c (count_occurrences_1): Also punt when SUBREG_REG
        is not a register.

        * gfortran.dg/pr44691.f: New test.

From-SVN: r163516

14 years agosel-sched.c: Add 2010 to copyright years.
Andrey Belevantsev [Tue, 24 Aug 2010 09:10:40 +0000 (13:10 +0400)] 
sel-sched.c: Add 2010 to copyright years.

        * sel-sched.c: Add 2010 to copyright years.
        * sel-sched-ir.c: Likewise.
        * sel-sched-ir.h: Likewise.

From-SVN: r163515

14 years agobackport: sel-sched-ir.c (sel_restore_other_notes): Rename to sel_restore_notes.
Andrey Belevantsev [Tue, 24 Aug 2010 09:09:36 +0000 (13:09 +0400)] 
backport: sel-sched-ir.c (sel_restore_other_notes): Rename to sel_restore_notes.

Backport from mainline:
 2010-01-14  Alexander Monakov  <amonakov@ispras.ru>

        * sel-sched-ir.c (sel_restore_other_notes): Rename to sel_restore_notes.
        Update all callers.  Call reemit_notes for all insns.

From-SVN: r163514

14 years agobackport: re PR rtl-optimization/42246 (ICE in init_seqno for 186.crafty with sel...
Andrey Belevantsev [Tue, 24 Aug 2010 09:08:23 +0000 (13:08 +0400)] 
backport: re PR rtl-optimization/42246 (ICE in init_seqno for 186.crafty with sel-sched)

Backport from mainline:
 PR rtl-optimization/42246
         * sel-sched-ir.h (get_all_loop_exits): Include exits from inner
         loops.

From-SVN: r163513

14 years agobackport: sel-sched.c (compute_av_set_at_bb_end): Do not test that number of all...
Andrey Belevantsev [Tue, 24 Aug 2010 09:06:59 +0000 (13:06 +0400)] 
backport: sel-sched.c (compute_av_set_at_bb_end): Do not test that number of all successors is the same as number of successors...

Backport from mainline:
        * sel-sched.c (compute_av_set_at_bb_end): Do not test that number of
         all successors is the same as number of successors in current region.

From-SVN: r163512

14 years agobackport: sel-sched.c (maybe_emit_renaming_copy): Exit early when expression to renam...
Andrey Belevantsev [Tue, 24 Aug 2010 09:05:46 +0000 (13:05 +0400)] 
backport: sel-sched.c (maybe_emit_renaming_copy): Exit early when expression to rename is not separable.

Backport from mainline:
 2010-01-14  Alexander Monakov  <amonakov@ispras.ru>

        * sel-sched.c (maybe_emit_renaming_copy): Exit early when expression
        to rename is not separable.  Otherwise check that its LHS is not NULL.

From-SVN: r163511

14 years agobackport: sel-sched.c (choose_best_reg_1): Loop over all regs for mode.
Andrey Belevantsev [Tue, 24 Aug 2010 09:04:29 +0000 (13:04 +0400)] 
backport: sel-sched.c (choose_best_reg_1): Loop over all regs for mode.

Backport from mainline:
 2010-01-14  Alexander Monakov  <amonakov@ispras.ru>

        * sel-sched.c (choose_best_reg_1):  Loop over all regs for mode.

From-SVN: r163510

14 years agobackport: sel-sched.c (mark_unavailable_hard_regs): Do not try to search available...
Andrey Belevantsev [Tue, 24 Aug 2010 09:03:31 +0000 (13:03 +0400)] 
backport: sel-sched.c (mark_unavailable_hard_regs): Do not try to search available registers when failed to discover LHS register class.

Backport from mainline:
 2010-01-14  Alexander Monakov  <amonakov@ispras.ru>

        * sel-sched.c (mark_unavailable_hard_regs): Do not try to search
        available registers when failed to discover LHS register class.
        Fix indentation.  Update comment.

From-SVN: r163509

14 years agobackport: re PR rtl-optimization/42389 (ICE in advance_state_on_fence with sel-schd...
Andrey Belevantsev [Tue, 24 Aug 2010 09:02:30 +0000 (13:02 +0400)] 
backport: re PR rtl-optimization/42389 (ICE in advance_state_on_fence with sel-schd for 175.vpr)

Backport from mainline:
        PR rtl-optimization/42389
         * sel-sched.c (advance_one_cycle): Set FENCE_ISSUE_MORE
         to can_issue_more.
         (advance_state_on_fence): Likewise.
         (sel_target_adjust_priority): Print debug output only when
         sched_verbose >= 4, not 2.
         (get_expr_cost): Do not issue all unique insns on the next cycle.
         (fill_insns): Initialize can_issue_more from the value saved
         with the fence.
         * sel-sched-ir.c (flist_add): New parameter issue_more.
         Init FENCE_ISSUE_MORE with it.
         (merge_fences): Likewise.
         (init_fences): Update call to flist_add.
         (add_to_fences, add_clean_fence_to_fences)
         (add_dirty_fence_to_fences): Likewise.
         (move_fence_to_fences): Update call to merge_fences.
         (invoke_reorder_hooks): Do not reset can_issue_more on insns from
         sched groups.
         * sel-sched-ir.h (struct _fence): New field issue_more.
         (FENCE_ISSUE_MORE): New accessor macro.

From-SVN: r163508

14 years agobackport: re PR rtl-optimization/42388 (ICE in move_bb_info with sel-sched and modulo...
Andrey Belevantsev [Tue, 24 Aug 2010 09:01:18 +0000 (13:01 +0400)] 
backport: re PR rtl-optimization/42388 (ICE in move_bb_info with sel-sched and modulo-sched for 176.gcc)

Backport from mainline:
        PR rtl-optimization/42388
         * sel-sched-ir.c (maybe_tidy_empty_bb): Do not delete empty blocks
         that have no predecessors nor successors.  Do not call move_bb_info
         for empty blocks outside of current region.

From-SVN: r163507

14 years agobackport: re PR rtl-optimization/42294 (ICE in code_motion_path_driver for 416.gamess)
Andrey Belevantsev [Tue, 24 Aug 2010 08:59:47 +0000 (12:59 +0400)] 
backport: re PR rtl-optimization/42294 (ICE in code_motion_path_driver for 416.gamess)

Backport from mainline:
 2010-01-14  Alexander Monakov  <amonakov@ispras.ru>

        PR rtl-optimization/42294
        * sel-sched-ir.h (struct _sel_insn_data): Update comment.
        * sel-sched.c (move_exprs_to_boundary): Transitively add all
        originators' originators.

From-SVN: r163506

14 years agobackport: re PR rtl-optimization/39453 (ICE : in init_seqno, at sel-sched.c:6433)
Andrey Belevantsev [Tue, 24 Aug 2010 08:58:36 +0000 (12:58 +0400)] 
backport: re PR rtl-optimization/39453 (ICE : in init_seqno, at sel-sched.c:6433)

Backport from mainline:
 2010-01-14  Alexander Monakov  <amonakov@ispras.ru>

        PR rtl-optimization/39453
        PR rtl-optimization/42246
        * sel-sched-ir.c (considered_for_pipelining_p): Do not test
        for pipelining_p.
        (sel_add_loop_preheaders): Add preheader to last_added_blocks.

        * gcc.dg/pr39453.c: New.
        * gcc.dg/pr42246.c: New.

From-SVN: r163505

14 years agobackport: re PR middle-end/42245 (ICE in verify_backedges for 197.parser with sel...
Andrey Belevantsev [Tue, 24 Aug 2010 08:57:18 +0000 (12:57 +0400)] 
backport: re PR middle-end/42245 (ICE in verify_backedges for 197.parser with sel-sched)

Backport from mainline:
 2010-01-14  Andrey Belevantsev  <abel@ispras.ru>
            Alexander Monakov  <amonakov@ispras.ru>

        PR middle-end/42245
        * sel-sched-ir.c (sel_recompute_toporder): New.  Use it...
        (maybe_tidy_empty_bb): ... here.  Make static.  Add new
        argument.  Update all callers.
        (tidy_control_flow): ... and here.  Recompute topological order
        of basic blocks in region if necessary.
        (sel_redirect_edge_and_branch): Change return type.  Return true
        if topological order might have been invalidated.
        (purge_empty_blocks): Export and move from...
        * sel-sched.c (purge_empty_blocks): ... here.
        * sel-sched-ir.h (sel_redirect_edge_and_branch): Update prototype.
        (maybe_tidy_empty_bb): Delete prototype.
        (purge_empty_blocks): Declare.

From-SVN: r163504

14 years agobackport: re PR rtl-optimization/42294 (ICE in code_motion_path_driver for 416.gamess)
Andrey Belevantsev [Tue, 24 Aug 2010 08:55:33 +0000 (12:55 +0400)] 
backport: re PR rtl-optimization/42294 (ICE in code_motion_path_driver for 416.gamess)

Backport from mainline:
        PR rtl-optimization/42294
         * sel-sched.c (try_replace_dest_reg): When chosen register
         and original register is the same, do not bail out early, but
         still check all original insns for validity of replacing destination
         register.  Set EXPR_TARGET_AVAILABLE to 1 before leaving function
         in this case.

From-SVN: r163503

14 years agobackport: re PR rtl-optimization/41697 (ICE on gcc.c-torture/compile/20090917-1.c)
Andrey Belevantsev [Tue, 24 Aug 2010 08:54:02 +0000 (12:54 +0400)] 
backport: re PR rtl-optimization/41697 (ICE on gcc.c-torture/compile/20090917-1.c)

Backport from mainline:
 PR rtl-optimization/41697
         * sel-sched-ir.c (fallthru_bb_of_jump): Bail out when a block with
         a conditional jump has a single successor.

From-SVN: r163502

14 years agobackport: sched-deps.c (init_deps): New parameter lazy_reg_last.
Andrey Belevantsev [Tue, 24 Aug 2010 08:53:11 +0000 (12:53 +0400)] 
backport: sched-deps.c (init_deps): New parameter lazy_reg_last.

Backport from mainline:
 2009-11-13  Andrey Belevantsev  <abel@ispras.ru>

        * sched-deps.c (init_deps): New parameter lazy_reg_last.  Don't
        allocate reg_last when in case lazy_reg_last is true.
        (init_deps_reg_last): New.
        (free_deps): When max_reg is 0, this context is already freed.
        * sched-int.h (init_deps_reg_last): Export.
        (init_deps): Update prototype.
        * sched-ebb.c (schedule_ebb): Update the call to init_deps.
        * sched-rgn.c (sched_rgn_compute_dependencies): Likewise.
        * ddg.c (build_intra_loop_deps): Likewise.
        * sel-sched-ir.c (copy_deps_context, create_deps_context,
        reset_deps_context, deps_init_id): Likewise.
        (init_first_time_insn_data): Lazy allocate INSN_DEPS_CONTEXT.
        (free_data_for_scheduled_insn): New, break down from ...
        (free_first_time_insn_data): ... here.
        (has_dependence_p): Allocate reg_last now, when it is needed.
        (extend_insn_data): When maximal LUID is big enough, allocate
        per-insn data in smaller chunks.
        * sel-sched-ir.h (free_data_for_scheduled_insn): Export.
        * sel-sched.c (update_seqnos_and_stage): Free INSN_DEPS_CONTEXT
        in scheduled insn.

From-SVN: r163500

14 years agobackport: re PR rtl-optimization/41697 (ICE on gcc.c-torture/compile/20090917-1.c)
Andrey Belevantsev [Tue, 24 Aug 2010 08:51:56 +0000 (12:51 +0400)] 
backport: re PR rtl-optimization/41697 (ICE on gcc.c-torture/compile/20090917-1.c)

Backport from mainline:
 2009-10-15  Steve Ellcey  <sje@cup.hp.com>

        PR rtl-optimization/41697
        * sel-sched-ir.h (_eligible_successor_edge_p): Check successor count.

From-SVN: r163499

14 years agobackport: re PR rtl-optimization/40101 (200.sixtrack ICEs in get_seqno_by_preds,...
Andrey Belevantsev [Tue, 24 Aug 2010 08:50:50 +0000 (12:50 +0400)] 
backport: re PR rtl-optimization/40101 (200.sixtrack ICEs in get_seqno_by_preds, at sel-sched-ir.c:3752)

Backport from mainline:
        PR rtl-optimization/40101
         * sel-sched-ir.c (get_seqno_by_preds): Allow returning negative
         seqno.  Adjust comment.
         * sel-sched.c (find_seqno_for_bookkeeping): Assert that when
         inserting bookkeeping before a jump, the jump is not scheduled.
         When no positive seqno found, provide a value.  Add comment.

From-SVN: r163498

14 years agobackport: 20071210-2.c: New testcase.
Andrey Belevantsev [Tue, 24 Aug 2010 08:49:50 +0000 (12:49 +0400)] 
backport: 20071210-2.c: New testcase.

Backport from mainline:

gcc/testsuite/Changelog:

2009-04-28  Alexander Monakov  <amonakov@ispras.ru>
        * gcc.target/ia64/20071210-2.c: New testcase.

gcc/Changelog:

2009-04-28  Alexander Monakov  <amonakov@ispras.ru>
        * sel-sched-ir.c (maybe_tidy_empty_bb): Do not attempt to delete a
        block if there are complex incoming edges.
        (sel_merge_blocks): Remove useless assert.
        (sel_redirect_edge_and_branch): Check that edge was redirected.
        * sel-sched-ir.h (_eligible_successor_edge_p): Remove assert.
        (sel_find_rgns): Delete declaration.
        * sel-sched.c (purge_empty_blocks): Attempt to remove first block of
        the region when it is not a preheader.

From-SVN: r163497

14 years agoDaily bump.
GCC Administrator [Tue, 24 Aug 2010 00:16:48 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r163492

14 years agore PR c++/45315 (ICE: tree check: expected aggr_init_expr, have call_expr in build_va...
Jason Merrill [Mon, 23 Aug 2010 06:39:47 +0000 (02:39 -0400)] 
re PR c++/45315 (ICE: tree check: expected aggr_init_expr, have call_expr in build_value_init, at cp/init.c:317)

PR c++/45315
* init.c (build_new_1): Don't use build_value_init in a template.

From-SVN: r163467

14 years agoDaily bump.
GCC Administrator [Mon, 23 Aug 2010 00:16:46 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r163463

14 years agoDaily bump.
GCC Administrator [Sun, 22 Aug 2010 00:17:01 +0000 (00:17 +0000)] 
Daily bump.

From-SVN: r163451

14 years agoDaily bump.
GCC Administrator [Sat, 21 Aug 2010 00:16:34 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r163435

14 years agoDaily bump.
GCC Administrator [Fri, 20 Aug 2010 00:16:44 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r163391

14 years agoDaily bump.
GCC Administrator [Thu, 19 Aug 2010 00:16:30 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r163360

15 years agoDaily bump.
GCC Administrator [Wed, 18 Aug 2010 00:16:31 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r163321

15 years agore PR target/45296 (register long double ICE at -O2, -Os, -O3)
Uros Bizjak [Tue, 17 Aug 2010 14:25:52 +0000 (16:25 +0200)] 
re PR target/45296 (register long double ICE at -O2, -Os, -O3)

PR target/45296
* reginfo.c (globalize_reg): Reject stack registers.

testsuite/ChangeLog:

PR target/45296
* gcc.target/i386/pr45296.c: New test.

From-SVN: r163307

15 years agoDaily bump.
GCC Administrator [Tue, 17 Aug 2010 00:16:41 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r163289

15 years agobackport: re PR tree-optimization/44632 (wrong code for complex division)
John David Anglin [Mon, 16 Aug 2010 20:24:54 +0000 (20:24 +0000)] 
backport: re PR tree-optimization/44632 (wrong code for complex division)

Backport from mainline:
2010-08-09  Richard Guenther  <rguenther@suse.de>

PR middle-end/44632
* function.c (gimplify_parameters): Do not clear addressable
bit of the original parameter.

From-SVN: r163285

15 years agoDaily bump.
GCC Administrator [Mon, 16 Aug 2010 00:16:35 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r163273

15 years agocast.C: Change fields of structures to "long" to pass the test on targets with...
Uros Bizjak [Sun, 15 Aug 2010 17:14:56 +0000 (19:14 +0200)] 
cast.C: Change fields of structures to "long" to pass the test on targets with...

* g++.dg/cast.C: Change fields of structures to "long" to pass
the test on targets with STRUCTURE_SIZE_BOUNDARY != BITS_PER_UNIT.
* g++.dg/cpp0x/iop.C: Ditto.
* g++.dg/cpp0x/named_refs.C: Ditto.
* g++.dg/cpp0x/rv1p.C: Ditto.
* g++.dg/cpp0x/rv2p.C: Ditto.
* g++.dg/cpp0x/rv3p.C: Ditto.
* g++.dg/cpp0x/rv4p.C: Ditto.
* g++.dg/cpp0x/rv5p.C: Ditto.
* g++.dg/cpp0x/rv6p.C: Ditto.
* g++.dg/cpp0x/rv7p.C: Ditto.
* g++.dg/cpp0x/rv8p.C: Ditto.
* g++.dg/cpp0x/unnamed_refs.C: Ditto.

From-SVN: r163265

15 years agoDaily bump.
GCC Administrator [Sun, 15 Aug 2010 00:16:35 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r163256

15 years agoRevert:
John David Anglin [Sat, 14 Aug 2010 14:42:43 +0000 (14:42 +0000)] 
Revert:

2010-08-08  John David Anglin  <dave.anglin@nrc-cnrc.gc.ca>

PR boehm-gc/34544
* gthr-posix.h (__gthread_start): Delete.
(__gthread_active_init): Use pthread_default_stacksize_np instead of
pthread_create to determine if hpux pthreads are active.
* gthr-posix95.h (__gthread_start): Delete.
(__gthread_active_init): Likewise use pthread_default_stacksize_np.

From-SVN: r163243

15 years agoDaily bump.
GCC Administrator [Sat, 14 Aug 2010 00:16:37 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r163233

15 years agoDaily bump.
GCC Administrator [Fri, 13 Aug 2010 00:16:43 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r163216

15 years agore PR middle-end/45262 (Optimization results in wrong result on expression x>>31...
Jakub Jelinek [Thu, 12 Aug 2010 15:28:40 +0000 (17:28 +0200)] 
re PR middle-end/45262 (Optimization results in wrong result on expression x>>31||(-x)>>31)

PR middle-end/45262
* fold-const.c (make_range) <case NEGATE_EXPR>: Punt if
-a overflows.  Normalize the range.

* gcc.c-torture/execute/pr45262.c: New test.

From-SVN: r163195

15 years agoDaily bump.
GCC Administrator [Thu, 12 Aug 2010 00:17:36 +0000 (00:17 +0000)] 
Daily bump.

From-SVN: r163180

15 years agoDaily bump.
GCC Administrator [Wed, 11 Aug 2010 00:16:51 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r163083

15 years agobackport: re PR middle-end/41551 (ia64: ICE: in instantiate_virtual_regs_in_insn...
Steve Ellcey [Tue, 10 Aug 2010 15:40:14 +0000 (15:40 +0000)] 
backport: re PR middle-end/41551 (ia64: ICE: in instantiate_virtual_regs_in_insn, at function.c:1630)

2010-08-10  Steve Ellcey  <sje@cup.hp.com>
    Jakub Jelinek <jakub@redhat.com>

Backport from mainline:
PR middle-end/41551
* function.c (instantiate_virtual_regs_in_insn): Copy to new reg
before forcing mode.
* testsuite/gcc.dg/pr41551.c: New test.

Co-Authored-By: Jakub Jelinek <jakub@redhat.com>
From-SVN: r163072

15 years agoDaily bump.
GCC Administrator [Tue, 10 Aug 2010 00:16:38 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r163044

15 years agoDaily bump.
GCC Administrator [Mon, 9 Aug 2010 00:16:48 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r163022

15 years agobackport: re PR target/44805 (FAIL: gcc.c-torture/execute/builtins/strpcpy-2.c execut...
John David Anglin [Sun, 8 Aug 2010 19:57:46 +0000 (19:57 +0000)] 
backport: re PR target/44805 (FAIL: gcc.c-torture/execute/builtins/strpcpy-2.c execution, at -O1 and above)

Backport from mainline.
2010-07-17  John David Anglin  <dave.anglin@nrc-cnrc.gc.ca>

PR target/44805
* config/pa/pa.h (CONSTANT_ALIGNMENT): Align strings to BITS_PER_WORD
on all targets.

From-SVN: r163018

15 years agore PR tree-optimization/45109 (ICE: in get_constraint_for_component_ref, at tree...
Richard Guenther [Sun, 8 Aug 2010 15:51:14 +0000 (15:51 +0000)] 
re PR tree-optimization/45109 (ICE: in get_constraint_for_component_ref, at tree-ssa-structalias.c:2932)

2010-08-08  Richard Guenther  <rguenther@suse.de>

PR tree-optimization/45109
* tree-ssa-structalias.c (get_constraint_for_component_ref):
Explicitly strip handled components and indirect references.

* gcc.c-torture/compile/pr45109.c: New testcase.

From-SVN: r163010

15 years agore PR boehm-gc/34544 (pthread_default_stacksize_np failed.)
John David Anglin [Sun, 8 Aug 2010 15:11:20 +0000 (15:11 +0000)] 
re PR boehm-gc/34544 (pthread_default_stacksize_np failed.)

PR boehm-gc/34544
* gthr-posix.h (__gthread_start): Delete.
(__gthread_active_init): Use pthread_default_stacksize_np instead of
pthread_create to determine if hpux pthreads are active.
* gthr-posix95.h (__gthread_start): Delete.
(__gthread_active_init): Likewise use pthread_default_stacksize_np.

From-SVN: r163005

15 years agothunk3.C: Skip for targets with generic thunk support.
Uros Bizjak [Sun, 8 Aug 2010 13:05:43 +0000 (15:05 +0200)] 
thunk3.C: Skip for targets with generic thunk support.

* g++.old-deja/g++.jason/thunk3.C: Skip for targets with
generic thunk support.

From-SVN: r163000

15 years agore PR fortran/31588 (gfortran should be able to output Makefile dependencies with...
Daniel Franke [Sun, 8 Aug 2010 01:59:15 +0000 (21:59 -0400)] 
re PR fortran/31588 (gfortran should be able to output Makefile dependencies with -M* options)

2010-08-07  Daniel Franke  <franke.daniel@gmail.com>

PR fortran/31588
PR fortran/43954
Backport from mainline:
* gfortranspec.c (lang_specific_driver): Removed deprecation
warning for -M.
* lang.opt: Add options -M, -MM, -MD, -MMD, -MF, -MG, -MP, -MT, -MQ.
* lang-specs.h (CPP_FORWARD_OPTIONS): Add -M* options.
* cpp.h (gfc_cpp_makedep): New.
(gfc_cpp_add_dep): New.
(gfc_cpp_add_target): New.
* cpp.c (gfc_cpp_option): Add deps* members.
(gfc_cpp_makedep): New.
(gfc_cpp_add_dep): New.
(gfc_cpp_add_target): New.
(gfc_cpp_init_options): Initialize new options.
(gfc_cpp_handle_option): Handle new options.
(gfc_cpp_post_options): Map new options to libcpp-options.
(gfc_cpp_init): Handle deferred -MQ and -MT options.
(gfc_cpp_done): If requested, write dependencies to file.
* module.c (gfc_dump_module): Add a module filename as target.
* scanner.c (open_included_file): New parameter system; add the
included file as dependency.
(gfc_open_included_file): Add the included file as dependency.
(gfc_open_intrinsic_module): Likewise.
* invoke.texi: Removed deprecation warning for -M.
* gfortran.texi: Removed Makefile-dependencies project.

From-SVN: r162990

15 years agoDaily bump.
GCC Administrator [Sun, 8 Aug 2010 00:16:40 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r162987

15 years agoDaily bump.
GCC Administrator [Sat, 7 Aug 2010 00:16:44 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r162971

15 years agore PR target/44942 (Bug in argument passing of long double)
Eric Botcazou [Fri, 6 Aug 2010 23:23:29 +0000 (23:23 +0000)] 
re PR target/44942 (Bug in argument passing of long double)

PR target/44942
* config/sparc/sparc.c (function_arg_advance): Always take into account
the padding, if any.

From-SVN: r162969

15 years agore PR fortran/44660 (ICE in resolve_equivalence())
Mikael Morin [Fri, 6 Aug 2010 17:17:37 +0000 (17:17 +0000)] 
re PR fortran/44660 (ICE in resolve_equivalence())

2010-08-06  Mikael Morin  <mikael@gcc.gnu.org>

PR fortran/44660
* gfortran.h (gfc_namespace): New field old_equiv.
(gfc_free_equiv_until): New prototype.
* match.c (gfc_free_equiv_until): New, renamed from gfc_free_equiv with
a parameterized stop condition.
(gfc_free_equiv): Use gfc_free_equiv_until.
* parse.c (next_statement): Save equivalence list.
(reject_statement): Restore equivalence list.

From-SVN: r162949

15 years agoexpmed.c (expand_mult_const): Expand shift into temporary.
Uros Bizjak [Fri, 6 Aug 2010 07:52:04 +0000 (09:52 +0200)] 
expmed.c (expand_mult_const): Expand shift into temporary.

2010-08-06  Uros Bizjak  <ubizjak@gmail.com>

* expmed.c (expand_mult_const) <case alg_shift>: Expand shift into
temporary. Emit move from temporary to accum, so REG_EQUAL note will
be attached to this insn in correct mode.

2010-08-06  Uros Bizjak  <ubizjak@gmail.com>

* config/i386/i386.c (ix86_decompose_address): Check for SI_REG
using REGNO of base_reg directly.

From-SVN: r162940

15 years agoDaily bump.
GCC Administrator [Fri, 6 Aug 2010 00:16:49 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r162926

15 years agoDaily bump.
GCC Administrator [Thu, 5 Aug 2010 00:17:33 +0000 (00:17 +0000)] 
Daily bump.

From-SVN: r162894

15 years agospu.c (spu_emit_branch_hint): Do not access NOTE_KIND of non-NOTE insns.
Ulrich Weigand [Wed, 4 Aug 2010 18:40:05 +0000 (18:40 +0000)] 
spu.c (spu_emit_branch_hint): Do not access NOTE_KIND of non-NOTE insns.

* config/spu/spu.c (spu_emit_branch_hint): Do not access NOTE_KIND of
non-NOTE insns.

From-SVN: r162870

15 years agoDaily bump.
GCC Administrator [Wed, 4 Aug 2010 00:16:53 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r162851

15 years agoDaily bump.
GCC Administrator [Tue, 3 Aug 2010 00:16:37 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r162832

15 years agobackport: JobStateReasons.java (add): Fix infinite recursion with call to super.
Andrew Haley [Mon, 2 Aug 2010 14:25:47 +0000 (14:25 +0000)] 
backport: JobStateReasons.java (add): Fix infinite recursion with call to super.

Merge from trunk:

2010-07-30  Andrew Haley  <aph@redhat.com>

        * javax/print/attribute/standard/JobStateReasons.java (add): Fix
        infinite recursion with call to super.

2010-04-27  Andrew Haley  <aph@redhat.com>

        * java/util/concurrent/CopyOnWriteArrayList.java: Fix for empty
        list.

2010-04-27  Andrew Haley  <aph@redhat.com>

        * gnu/javax/print/ipp/IppResponse.java (parseAttributes): Handle
        IppValueTag.UNKNOWN.
        * gnu/javax/print/ipp/IppRequest.java (writeOperationAttributes):
        Handle RequestedAttributes.
        * gnu/javax/print/ipp/IppPrintService.java (processResponse): Add
        DocFlavor.SERVICE_FORMATTED.PAGEABLE and
        DocFlavor.SERVICE_FORMATTED.PRINTABLE.

From-SVN: r162820

15 years agoDaily bump.
GCC Administrator [Mon, 2 Aug 2010 00:16:48 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r162807

15 years agoDaily bump.
GCC Administrator [Sun, 1 Aug 2010 00:16:43 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r162790

15 years agoDaily bump.
GCC Administrator [Sat, 31 Jul 2010 00:16:52 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r162728

15 years agoDaily bump.
GCC Administrator [Fri, 30 Jul 2010 00:17:00 +0000 (00:17 +0000)] 
Daily bump.

From-SVN: r162699

15 years agoDaily bump.
GCC Administrator [Thu, 29 Jul 2010 00:16:45 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r162664

15 years agobackport: re PR middle-end/42509 (nonoverlapping_memrefs_p misinterprets NULL MEM_OFF...
Ulrich Weigand [Wed, 28 Jul 2010 18:00:08 +0000 (18:00 +0000)] 
backport: re PR middle-end/42509 (nonoverlapping_memrefs_p misinterprets NULL MEM_OFFSET as const0_rtx)

Backport from mainline:
2010-04-03  Richard Guenther  <rguenther@suse.de>

PR middle-end/42509
* alias.c (nonoverlapping_memrefs_p): For spill-slot accesses
require a non-NULL MEM_OFFSET.

From-SVN: r162650

15 years agoDaily bump.
GCC Administrator [Wed, 28 Jul 2010 00:16:36 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r162613

15 years agore PR c/44555 (Pointer evalutions, is that expected ?)
Richard Guenther [Tue, 27 Jul 2010 13:07:28 +0000 (13:07 +0000)] 
re PR c/44555 (Pointer evalutions, is that expected ?)

2010-07-27  Richard Guenther  <rguenther@suse.de>

PR c/44555
* c-common.c (c_common_truthvalue_conversion): Remove
premature and wrong optimization concering ADDR_EXPRs.

* gcc.c-torture/execute/pr44555.c: New testcase.

From-SVN: r162566

15 years agore PR tree-optimization/44977 (ice in propagate_rhs_into_lhs, at tree-ssa-dom.c:2728)
Richard Guenther [Tue, 27 Jul 2010 13:05:17 +0000 (13:05 +0000)] 
re PR tree-optimization/44977 (ice in propagate_rhs_into_lhs, at tree-ssa-dom.c:2728)

2010-07-27  Richard Guenther  <rguenther@suse.de>

PR tree-optimization/44977
* tree-ssa-dom.c (propagate_rhs_into_lhs): Do not create invalid
SSA form.

* gcc.dg/torture/pr44977.c: New testcase.

From-SVN: r162565

15 years agoDaily bump.
GCC Administrator [Tue, 27 Jul 2010 00:16:51 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r162552

15 years agoDaily bump.
GCC Administrator [Mon, 26 Jul 2010 00:16:42 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r162523

15 years agoDaily bump.
GCC Administrator [Sun, 25 Jul 2010 00:17:13 +0000 (00:17 +0000)] 
Daily bump.

From-SVN: r162505

15 years agoDaily bump.
GCC Administrator [Sat, 24 Jul 2010 00:16:44 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r162494

15 years agore PR fortran/45019 (Aliasing of TARGET dummy argument not detected correctly)
Tobias Burnus [Fri, 23 Jul 2010 08:40:00 +0000 (10:40 +0200)] 
re PR fortran/45019 (Aliasing of TARGET dummy argument not detected correctly)

2010-07-23  Tobias Burnus  <burnus@net-b.de>

        PR fortran/45019
        * dependency.c (gfc_check_dependency): Add argument alising
        * check.
        * symbol.c (gfc_symbols_could_alias): Add argument alising
        * check.

2010-07-23  Tobias Burnus  <burnus@net-b.de>

        PR fortran/45019
        * gfortran.dg/aliasing_dummy_5.f90: New.

From-SVN: r162448

15 years agoDaily bump.
GCC Administrator [Fri, 23 Jul 2010 00:17:15 +0000 (00:17 +0000)] 
Daily bump.

From-SVN: r162435

15 years agobackport: re PR target/44942 (Bug in argument passing of long double)
Jakub Jelinek [Thu, 22 Jul 2010 06:46:28 +0000 (08:46 +0200)] 
backport: re PR target/44942 (Bug in argument passing of long double)

Backport from mainline
2010-07-16  Jakub Jelinek  <jakub@redhat.com>

PR target/44942
* config/i386/i386-protos.h (ix86_function_arg_boundary): Change second
argument to const_tree.
* config/i386/i386.c (function_arg_advance): If padding needs to be
inserted before argument, increment cum->words by number of padding
words as well.
(contains_aligned_value_p): Change argument to const_tree.
(ix86_function_arg_boundary): Change second argument to const_tree.

* gcc.c-torture/execute/pr44942.c: New test.
* gcc.target/i386/pr44942.c: New test.

From-SVN: r162399

15 years agoDaily bump.
GCC Administrator [Thu, 22 Jul 2010 00:16:45 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r162394

15 years agore PR target/42869 (GOMP_critical_start wrong on Itanium due to __sync miscompilation)
Steve Ellcey [Wed, 21 Jul 2010 22:37:53 +0000 (22:37 +0000)] 
re PR target/42869 (GOMP_critical_start wrong on Itanium due to __sync miscompilation)

2010-07-21  Steve Ellcey  <sje@cup.hp.com>

        PR target/42869
        * config/ia64/sync.md (sync_compare_and_swap): Move memory fence.

From-SVN: r162387

15 years agobackport: target-supports.exp (check_sse_os_support_available): New proc.
Rainer Orth [Wed, 21 Jul 2010 14:37:33 +0000 (14:37 +0000)] 
backport: target-supports.exp (check_sse_os_support_available): New proc.

Backport from mainline:
2010-07-19  Rainer Orth  <ro@CeBiTec.Uni-Bielefeld.DE>

* lib/target-supports.exp (check_sse_os_support_available): New
proc.
(check_sse_hw_available): New proc.
(check_effective_target_sse_runtime): New proc.
(check_effective_target_sse2_runtime): New proc.
* lib/fortran-torture.exp (get-fortran-torture-options): Only add
-msse2 if check_sse_os_support_available.
* g++.dg/vect/vect.exp: Only run -msse2 tests if
check_sse_os_support_available.
* gcc.dg/vect/vect.exp: Likewise.
* gfortran.dg/vect/vect.exp: Likewise.
* gcc.target/i386/sol2-check: Renamed to ...
* gcc.target/i386/sse-os-support.h: ... this.
(sol2_check): Renamed to ...
(sse_os_support): ... this.
Only test movss with xmm registers.
* gcc.target/i386/sse-check.h: Reflect new header and function names.
Removed ILL_INSN, ILL_INSN_LEN.
* gcc.target/i386/sse2-check.h: Likewise.
* gcc.target/i386/sse3-check.h: Likewise.
* gcc.dg/pr40550.c: Use dg-require-effective-target sse_runtime.
Removed cpuid.h, __get_cpuid test.
* g++.dg/other/i386-1.C: Use dg-require-effective-target sse2_runtime.
Removed cpuid.h, __get_cpuid test.
* g++.dg/other/pr40446.C: Likewise.
* gcc.dg/compat/union-m128-1_main.c: Likewise.
* gcc.dg/compat/vector-1a_main.c: Likewise.
* gcc.dg/compat/vector-2a_main.c: Likewise.
* gcc.dg/pr36584.c: Likewise.
* gcc.dg/pr37544.c: Likewise.
* gcc.dg/torture/pr16104-1.c: Likewise.
* gcc.dg/torture/stackalign/alloca-2.c: Likewise.
* gcc.dg/torture/stackalign/alloca-3.c: Likewise.
* gcc.dg/torture/stackalign/push-1.c: Likewise.
* gcc.dg/torture/stackalign/vararg-3.c: Likewise.
* gcc.dg/torture/pr35771.h: Removed cpuid.h, __get_cpuid test.
* gcc.dg/torture/pr35771-1.c: Use dg-require-effective-target
sse2_runtime.
* gcc.dg/torture/pr35771-2.c: Likewise.
* gcc.dg/torture/pr35771-3.c: Likewise.

From-SVN: r162376

15 years agobackport: re PR testsuite/38946 (gcc trunk 143562 - Testsuite - gfortran failing...
Rainer Orth [Wed, 21 Jul 2010 09:06:42 +0000 (09:06 +0000)] 
backport: re PR testsuite/38946 (gcc trunk 143562 - Testsuite - gfortran failing tests that worked previously)

Backport from mainline:
2010-06-25  Jerry DeLisle  <jvdelisle@gcc.gnu.org>

PR testsuite/38946
* gfortran.dg/array_constructor_23.f: Update test to allow for small
error in comparing reals.

From-SVN: r162367

15 years agoDaily bump.
GCC Administrator [Wed, 21 Jul 2010 00:16:43 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r162358

15 years agoDaily bump.
GCC Administrator [Tue, 20 Jul 2010 00:16:52 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r162320

15 years agoDaily bump.
GCC Administrator [Mon, 19 Jul 2010 00:16:32 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r162289