]> git.ipfire.org Git - thirdparty/gcc.git/log
thirdparty/gcc.git
11 years agoDaily bump.
GCC Administrator [Fri, 6 Jun 2014 00:16:28 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r211291

11 years agoDaily bump.
GCC Administrator [Thu, 5 Jun 2014 00:16:35 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r211250

11 years agobackport: re PR tree-optimization/57230 (tree-ssa-strlen incorrectly optimizes a...
Richard Biener [Wed, 4 Jun 2014 11:59:45 +0000 (11:59 +0000)] 
backport: re PR tree-optimization/57230 (tree-ssa-strlen incorrectly optimizes a strlen to 0)

2014-06-04  Richard Biener  <rguenther@suse.de>

Backport from mainline
2013-05-13  Jakub Jelinek  <jakub@redhat.com>

PR tree-optimization/57230
* tree-ssa-strlen.c (handle_char_store): Add missing integer_zerop
check.

* gcc.dg/strlenopt-23.c: New test.

From-SVN: r211225

11 years agoDaily bump.
GCC Administrator [Wed, 4 Jun 2014 00:16:30 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r211207

11 years agoDaily bump.
GCC Administrator [Tue, 3 Jun 2014 00:16:31 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r211154

11 years agoDaily bump.
GCC Administrator [Mon, 2 Jun 2014 00:16:40 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r211122

11 years agoDaily bump.
GCC Administrator [Sun, 1 Jun 2014 00:16:42 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r211107

11 years agoDaily bump.
GCC Administrator [Sat, 31 May 2014 00:16:35 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r211098

11 years agore PR c++/56947 (Bogus 'XX' was not declared in this scope)
Jason Merrill [Fri, 30 May 2014 15:09:29 +0000 (11:09 -0400)] 
re PR c++/56947 (Bogus 'XX' was not declared in this scope)

PR c++/56947
* pt.c (instantiate_decl): Don't defer instantiation of a nested
function.

From-SVN: r211083

11 years agoDaily bump.
GCC Administrator [Fri, 30 May 2014 00:16:33 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r211068

11 years agobool_set: Use UTF-8 for accented characters.
Jonathan Wakely [Thu, 29 May 2014 14:54:42 +0000 (15:54 +0100)] 
bool_set: Use UTF-8 for accented characters.

* include/tr2/bool_set: Use UTF-8 for accented characters.
* scripts/run_doxygen: Handle Doxygen 1.8.x change.

From-SVN: r211056

11 years agoDaily bump.
GCC Administrator [Thu, 29 May 2014 00:16:29 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r211038

11 years agore PR target/61044 (Computed goto on AVR fails to use word-addressing)
Georg-Johann Lay [Wed, 28 May 2014 08:50:18 +0000 (08:50 +0000)] 
re PR target/61044 (Computed goto on AVR fails to use word-addressing)

PR target/61044
* doc/extend.texi (Local Labels): Note that label differences are
not supported for AVR.

From-SVN: r211002

11 years agoDaily bump.
GCC Administrator [Wed, 28 May 2014 00:16:37 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r210991

11 years ago* gnat.dg/aliasing1.adb (dg-final): Robustify pattern matching.
Eric Botcazou [Tue, 27 May 2014 20:39:22 +0000 (20:39 +0000)] 
* gnat.dg/aliasing1.adb (dg-final): Robustify pattern matching.

From-SVN: r210985

11 years agoDaily bump.
GCC Administrator [Tue, 27 May 2014 00:16:37 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r210952

11 years agoPR 61310 Rewrite implementation of CTIME and FDATE intrinsics.
Janne Blomqvist [Mon, 26 May 2014 19:44:24 +0000 (22:44 +0300)] 
PR 61310 Rewrite implementation of CTIME and FDATE intrinsics.

2014-05-26  Janne Blomqvist  <jb@gcc.gnu.org>

Backport from mainline
PR libfortran/61310
* intrinsics.texi (CTIME): Remove mention of locale-dependent
behavior.

2014-05-26  Janne Blomqvist  <jb@gcc.gnu.org>

Backport from mainline
PR libfortran/61310
* intrinsics/ctime.c (strctime): Rename to gf_ctime, use snprintf
instead of strftime.
(fdate): Use gf_ctime.
(fdate_sub): Likewise.
(ctime): Likewise.
(ctime_sub): Likewise.

From-SVN: r210947

11 years agoDaily bump.
GCC Administrator [Mon, 26 May 2014 00:16:39 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r210915

11 years agoDaily bump.
GCC Administrator [Sun, 25 May 2014 00:16:06 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r210903

11 years agobackport: [multiple changes]
Dominique d'Humieres [Sat, 24 May 2014 15:45:02 +0000 (17:45 +0200)] 
backport: [multiple changes]

2014-05-24  Dominique d'Humieres <dominiq@lps.ens.fr>

Backport r195492 and r195815
2013-01-27  Paul Thomas  <pault@gcc.gnu.org>

PR fortran/55789
* gfortran.dg/associate_14.f90: New test.

PR fortran/56047
* gfortran.dg/associate_13.f90: New test.

2013-02-06  Paul Thomas <pault@gcc.gnu.org>

PR fortran/55789
* gfortran.dg/array_constructor_41.f90: New test.

From-SVN: r210894

11 years agobackport: [multiple changes]
Dominique d'Humieres [Sat, 24 May 2014 15:36:14 +0000 (17:36 +0200)] 
backport: [multiple changes]

2014-05-24  Dominique d'Humieres <dominiq@lps.ens.fr>

Backport r195492 and r195815
2013-01-27  Paul Thomas  <pault@gcc.gnu.org>

PR fortran/55789
PR fortran/56047
* gfortran.h : Add associate_var to symbol_attr.
* resolve.c (resolve_assoc_var): Set associate_var attribute.
If the target class_ok is set, set it for the associate
variable.
* check.c (allocatable_check): Associate variables should not
have the allocatable attribute even if their symbols do.
* class.c (gfc_build_class_symbol): Symbols with associate_var
set will always have a good class container.

2013-02-06  Paul Thomas  <pault@gcc.gnu.org>

PR fortran/55789
* trans-array.c (trans_array_constructor): Remove condition
'dynamic' = true if the loop ubound is a VAR_DECL.

From-SVN: r210893

11 years agoDaily bump.
GCC Administrator [Sat, 24 May 2014 00:16:29 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r210882

11 years agobackport: [multiple changes]
Richard Biener [Fri, 23 May 2014 09:46:18 +0000 (09:46 +0000)] 
backport: [multiple changes]

2014-05-23  Richard Biener  <rguenther@suse.de>

Backport from mainline
2013-08-30  Richard Biener  <rguenther@suse.de>

PR tree-optimization/58228
* tree-vect-data-refs.c (vect_analyze_data_ref_access): Do not
allow invariant loads in nested loop vectorization.

* gcc.dg/torture/pr58228.c: New testcase.

2013-09-26  Richard Biener  <rguenther@suse.de>

PR tree-optimization/58539
* tree-vect-loop.c (vect_create_epilog_for_reduction): Honor
the fact that debug statements are not taking part in loop-closed
SSA construction.

* gcc.dg/torture/pr58539.c: New testcase.

From-SVN: r210847

11 years agoDaily bump.
GCC Administrator [Fri, 23 May 2014 00:16:32 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r210831

11 years agore PR target/61208 (armhf: generated asm code produces "branch out of range" error...
Richard Earnshaw [Thu, 22 May 2014 15:56:34 +0000 (15:56 +0000)] 
re PR target/61208 (armhf: generated asm code produces "branch out of range" error in gas with -Os)

PR target/61208
* arm.md (arm_cmpdi_unsigned): Fix length calculation for Thumb2.

From-SVN: r210816

11 years agoDaily bump.
GCC Administrator [Thu, 22 May 2014 00:16:24 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r210730

11 years agoDaily bump.
GCC Administrator [Wed, 21 May 2014 00:16:34 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r210662

11 years agobackport: re PR target/60991 ([avr] Stack corruption when using 24-bit integers __int...
Senthil Kumar Selvaraj [Tue, 20 May 2014 08:37:50 +0000 (08:37 +0000)] 
backport: re PR target/60991 ([avr] Stack corruption when using 24-bit integers __int24 or __memx pointers in large stack frame)

gcc/
2014-05-20  Senthil Kumar Selvaraj  <senthil_kumar.selvaraj@atmel.com>

Backport from mainline r210325
2014-05-12  Senthil Kumar Selvaraj  <senthil_kumar.selvaraj@atmel.com>

PR target/60991
* config/avr/avr.c (avr_out_store_psi): Use correct constant
to restore Y.

gcc/testsuite/
2014-05-20  Senthil Kumar Selvaraj  <senthil_kumar.selvaraj@atmel.com>

Backport from mainline r210325
2014-05-12  Senthil Kumar Selvaraj  <senthil_kumar.selvaraj@atmel.com>

PR target/60991
* gcc.target/avr/pr60991.c: New testcase.

From-SVN: r210635

11 years agoDaily bump.
GCC Administrator [Tue, 20 May 2014 00:16:21 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r210626

11 years agoDaily bump.
GCC Administrator [Mon, 19 May 2014 00:17:06 +0000 (00:17 +0000)] 
Daily bump.

From-SVN: r210593

11 years agoDaily bump.
GCC Administrator [Sun, 18 May 2014 00:16:34 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r210570

11 years agoDaily bump.
GCC Administrator [Sat, 17 May 2014 00:16:35 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r210539

11 years agoPR 61187 Fix use of uninitialized data.
Janne Blomqvist [Fri, 16 May 2014 20:42:56 +0000 (23:42 +0300)] 
PR 61187 Fix use of uninitialized data.

2014-05-16  Janne Blomqvist  <jb@gcc.gnu.org>

Backport from trunk:
PR libfortran/61187
* io/unix.c (raw_close): Check if s->fd is -1.
(fd_to_stream): Check return value of fstat(), handle error.

From-SVN: r210529

11 years agoDaily bump.
GCC Administrator [Fri, 16 May 2014 00:16:31 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r210481

11 years agoDaily bump.
GCC Administrator [Thu, 15 May 2014 00:16:30 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r210452

11 years agoDaily bump.
GCC Administrator [Wed, 14 May 2014 00:16:35 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r210407

11 years agobackport: re PR tree-optimization/60382 (ICE on valid code at -O3 on x86_64-linux...
Richard Biener [Tue, 13 May 2014 13:21:47 +0000 (13:21 +0000)] 
backport: re PR tree-optimization/60382 (ICE on valid code at -O3 on x86_64-linux-gnu (in vect_create_epilog_for_reduction, at tree-vect-loop.c:4352))

2014-05-13  Richard Biener  <rguenther@suse.de>

Backport from mainline
2014-03-04  Richard Biener  <rguenther@suse.de>

PR tree-optimization/60382
* tree-vect-loop.c (vect_is_simple_reduction_1): Do not consider
dead PHIs a reduction.

* gcc.dg/vect/pr60382.c: New testcase.

From-SVN: r210371

11 years agoDaily bump.
GCC Administrator [Tue, 13 May 2014 00:16:32 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r210346

11 years agoDaily bump.
GCC Administrator [Mon, 12 May 2014 00:16:38 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r210316

11 years agoDaily bump.
GCC Administrator [Sun, 11 May 2014 00:16:18 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r210302

11 years agoDaily bump.
GCC Administrator [Sat, 10 May 2014 00:16:25 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r210295

11 years agobackport: re PR target/61055 ([avr] wrong test instruction after increment with -O1)
Georg-Johann Lay [Fri, 9 May 2014 11:34:46 +0000 (11:34 +0000)] 
backport: re PR target/61055 ([avr] wrong test instruction after increment with -O1)

gcc/
Backport from 2014-05-09 trunk r210267
PR target/61055
* config/avr/avr.md (cc): Add new attribute set_vzn.
(addqi3, negqi2) [cc]:
Set cc insn attribute to set_vzn instead of set_zn for alternatives
with INC, DEC or NEG.
* config/avr/avr.c (avr_notice_update_cc): Handle SET_VZN.
(avr_out_plus_1): ADIW sets cc0 to CC_SET_CZN.
INC, DEC set cc0 to CC_CLOBBER.

gcc/testsuite/
Backport from 2014-05-09 trunk r210267
PR target/61055
* gcc.target/avr/torture/pr61055.c: New test.

From-SVN: r210270

11 years agoDaily bump.
GCC Administrator [Fri, 9 May 2014 00:16:29 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r210250

11 years agore PR libstdc++/61117 (Manual uses term open source but should say free software)
Joshua Gay [Thu, 8 May 2014 19:57:06 +0000 (19:57 +0000)] 
re PR libstdc++/61117 (Manual uses term open source but should say free software)

2014-05-08  Joshua Gay  <jgay@gnu.org>

PR libstdc++/61117
* doc/xml/faq.xml (faq.license.what_restrictions): Replace "open
source" with "free software".
* doc/html/faq.html: Likewise.

From-SVN: r210242

11 years agobackport: re PR target/60609 (Error: value of 256 too large for field of 1 bytes...
Charles Baylis [Thu, 8 May 2014 17:06:04 +0000 (17:06 +0000)] 
backport: re PR target/60609 (Error: value of 256 too large for field of 1 bytes at 68242)

2014-05-08  Charles Baylis  <charles.baylis@linaro.org>

        Backport from mainline
        2014-04-07  Charles Baylis  <charles.baylis@linaro.org>

        PR target/60609
        * config/arm/arm.h (ASM_OUTPUT_CASE_END): Remove.
        (LABEL_ALIGN_AFTER_BARRIER): Align barriers which occur after
        ADDR_DIFF_VEC.

From-SVN: r210227

11 years agoDaily bump.
GCC Administrator [Thu, 8 May 2014 00:16:25 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r210194

11 years agobackport: re PR middle-end/58809 (ICE with complex variable in OpenMP reduction clause)
Jakub Jelinek [Wed, 7 May 2014 16:11:42 +0000 (18:11 +0200)] 
backport: re PR middle-end/58809 (ICE with complex variable in OpenMP reduction clause)

Backported from mainline
2013-10-21  Marek Polacek  <polacek@redhat.com>

PR middle-end/58809
* fold-const.c (fold_range_test): Return 0 if the type is not
an integral type.

* gcc.dg/gomp/pr58809.c: New test.

From-SVN: r210180

11 years agobackport: re PR middle-end/59011 (ICE in make_decl_rtl, at varasm.c:1147)
Jakub Jelinek [Wed, 7 May 2014 16:10:35 +0000 (18:10 +0200)] 
backport: re PR middle-end/59011 (ICE in make_decl_rtl, at varasm.c:1147)

Backported from mainline
2013-12-03  Jakub Jelinek  <jakub@redhat.com>

PR middle-end/59011
* gimplify.c (nonlocal_vla_vars): New variable.
(gimplify_var_or_parm_decl): Put VAR_DECLs for VLAs into
nonlocal_vla_vars chain.
(gimplify_body): Call declare_vars on nonlocal_vla_vars chain
if outer_bind has DECL_INITIAL (current_function_decl) block.

* gcc.dg/pr59011.c: New test.

From-SVN: r210179

11 years agobackport: re PR c++/59297 (ICE: openmp atomic with indirect LHS)
Jakub Jelinek [Wed, 7 May 2014 16:09:41 +0000 (18:09 +0200)] 
backport: re PR c++/59297 (ICE: openmp atomic with indirect LHS)

Backported from mainline
2013-11-28  Jakub Jelinek  <jakub@redhat.com>

PR c++/59297
* semantics.c (finish_omp_atomic): Call finish_expr_stmt
rather than add_stmt.

* g++.dg/gomp/pr59297.C: New test.

2013-05-31  Jason Merrill  <jason@redhat.com>

PR c++/56930
* semantics.c (potential_constant_expression_1): Handle OMP_ATOMIC*.

From-SVN: r210178

11 years agobackport: re PR tree-optimization/59014 (wrong code at -Os and above on x86_64-linux...
Jakub Jelinek [Wed, 7 May 2014 16:08:28 +0000 (18:08 +0200)] 
backport: re PR tree-optimization/59014 (wrong code at -Os and above on x86_64-linux-gnu)

Backported from mainline
2013-11-27  Jakub Jelinek  <jakub@redhat.com>

PR tree-optimization/59014
* gcc.c-torture/execute/pr59014-2.c: New test.

2013-11-26  Jakub Jelinek  <jakub@redhat.com>

PR tree-optimization/59014
* tree-vrp.c (register_edge_assert_for_1): Don't look
through conversions from non-integral types or through
narrowing conversions.

* gcc.c-torture/execute/pr59014.c: New test.

From-SVN: r210177

11 years agobackport: re PR target/59101 (integer wrong code bug)
Jakub Jelinek [Wed, 7 May 2014 16:07:21 +0000 (18:07 +0200)] 
backport: re PR target/59101 (integer wrong code bug)

Backported from mainline
2013-11-14  Jakub Jelinek  <jakub@redhat.com>
    Uros Bizjak  <ubizjak@gmail.com>

PR target/59101
* config/i386/i386.md (*anddi_2): Only allow CCZmode if
operands[2] satisfies_constraint_Z that might have bit 31 set.

* gcc.c-torture/execute/pr59101.c: New test.

From-SVN: r210176

11 years agobackport: re PR middle-end/58564 (possible wrong code bug at -O0)
Jakub Jelinek [Wed, 7 May 2014 16:05:38 +0000 (18:05 +0200)] 
backport: re PR middle-end/58564 (possible wrong code bug at -O0)

Backported from mainline
2013-09-30  Jakub Jelinek  <jakub@redhat.com>

PR middle-end/58564
* fold-const.c (fold_ternary_loc): For A < 0 : <sign bit of A> : 0
optimization, punt if sign_bit_p looked through any zero extension.

* gcc.c-torture/execute/pr58564.c: New test.

From-SVN: r210174

11 years agobackport: re PR rtl-optimization/58365 (likely wrong code bug)
Jakub Jelinek [Wed, 7 May 2014 16:04:44 +0000 (18:04 +0200)] 
backport: re PR rtl-optimization/58365 (likely wrong code bug)

Backported from mainline
2013-09-10  Jakub Jelinek  <jakub@redhat.com>

PR rtl-optimization/58365
* cfgcleanup.c (merge_memattrs): Also clear MEM_READONLY_P
resp. MEM_NOTRAP_P if they differ, or set MEM_VOLATILE_P if
it differs.

* gcc.c-torture/execute/pr58365.c: New test.

From-SVN: r210173

11 years agobackport: re PR c++/58325 (Spurious unused-but-set-variable warning on delete[] of...
Jakub Jelinek [Wed, 7 May 2014 16:03:49 +0000 (18:03 +0200)] 
backport: re PR c++/58325 (Spurious unused-but-set-variable warning on delete[] of volatile pointer)

Backported from mainline
2013-09-09  Jakub Jelinek  <jakub@redhat.com>

PR c++/58325
* init.c (build_vec_delete): Call mark_rvalue_use on base.

* g++.dg/warn/Wunused-var-21.C: New test.

From-SVN: r210172

11 years agobackport: re PR tree-optimization/58277 (wrong code at -O3)
Jakub Jelinek [Wed, 7 May 2014 16:01:24 +0000 (18:01 +0200)] 
backport: re PR tree-optimization/58277 (wrong code at -O3)

Backported from mainline
2013-08-30  Jakub Jelinek  <jakub@redhat.com>

PR tree-optimization/58277
* tree-ssa-strlen.c (strlen_enter_block): If do_invalidate gave up
after seeing too many stmts with vdef in between dombb and current
bb, invalidate everything.

* gcc.c-torture/execute/pr58277-1.c: New test.
* gcc.c-torture/execute/pr58277-2.c: New test.

From-SVN: r210171

11 years agobackport: re PR tree-optimization/58209 (ICE in extract_range_from_binary_expr, at...
Jakub Jelinek [Wed, 7 May 2014 16:00:33 +0000 (18:00 +0200)] 
backport: re PR tree-optimization/58209 (ICE in extract_range_from_binary_expr, at tree-vrp.c:2294)

Backported from mainline
2013-08-23  Jakub Jelinek  <jakub@redhat.com>

PR tree-optimization/58209
* tree-tailcall.c (find_tail_calls): Give up for pointer result types
if m or a is non-NULL.

* gcc.c-torture/execute/pr58209.c: New test.

From-SVN: r210170

11 years agobackport: re PR target/57777 (Python module fails compilation with "-march=core-avx2...
Jakub Jelinek [Wed, 7 May 2014 15:59:37 +0000 (17:59 +0200)] 
backport: re PR target/57777 (Python module fails compilation with "-march=core-avx2 -O3")

Backported from mainline
2013-07-03  Jakub Jelinek  <jakub@redhat.com>

PR target/57777
* config/i386/predicates.md (vsib_address_operand): Disallow
SYMBOL_REF or LABEL_REF in parts.disp if TARGET_64BIT && flag_pic.

* gcc.target/i386/pr57777.c: New test.

From-SVN: r210169

11 years agobackport: re PR target/57623 (BEXTR intrinsic has memory operands switched around...
Jakub Jelinek [Wed, 7 May 2014 15:57:48 +0000 (17:57 +0200)] 
backport: re PR target/57623 (BEXTR intrinsic has memory operands switched around (fails to compile code))

Backported from mainline
2013-06-27  Jakub Jelinek  <jakub@redhat.com>

PR target/57623
* config/i386/i386.md (bmi_bextr_<mode>): Swap predicates and
constraints of operand 1 and 2.

* gcc.target/i386/bmi-bextr-3.c: New test.

From-SVN: r210168

11 years agobackport: re PR target/57623 (BEXTR intrinsic has memory operands switched around...
Jakub Jelinek [Wed, 7 May 2014 15:54:34 +0000 (17:54 +0200)] 
backport: re PR target/57623 (BEXTR intrinsic has memory operands switched around (fails to compile code))

Backported from mainline
2013-06-27  Jakub Jelinek  <jakub@redhat.com>

PR target/57623
* config/i386/i386.md (bmi2_bzhi_<mode>3): Swap AND arguments
to match RTL canonicalization.  Swap predicates and
constraints of operand 1 and 2.

* gcc.target/i386/bmi2-bzhi-1.c: New test.

From-SVN: r210166

11 years agobackport: re PR tree-optimization/59164 (ice: tree check: expected tree that contains...
Richard Biener [Wed, 7 May 2014 11:51:36 +0000 (11:51 +0000)] 
backport: re PR tree-optimization/59164 (ice: tree check: expected tree that contains ‘decl minimal’ structure, have ‘integer_cst’ in get_var_info, at tree-into-ssa.c:380)

2014-05-07  Richard Biener  <rguenther@suse.de>

Backport from mainline
2013-11-19  Richard Biener  <rguenther@suse.de>

PR tree-optimization/59164
* tree-vect-loop.c (vect_analyze_loop_operations): Adjust
check whether we can create an epilogue loop to reflect the
cases where we create one.

* gcc.dg/torture/pr59164.c: New testcase.

From-SVN: r210154

11 years agobackport: [multiple changes]
Richard Biener [Wed, 7 May 2014 10:07:09 +0000 (10:07 +0000)] 
backport: [multiple changes]

2014-05-07  Richard Biener  <rguenther@suse.de>

Backport from mainline
2013-11-29  Richard Biener  <rguenther@suse.de>

PR tree-optimization/59334
* tree-ssa-dce.c (eliminate_unnecessary_stmts): Fix bug
in previous commit.

2013-11-28  Richard Biener  <rguenther@suse.de>

PR tree-optimization/59330
* tree-ssa-dce.c (eliminate_unnecessary_stmts): Simplify
and fix delayed marking of free calls not necessary.

* gcc.dg/torture/pr59330.c: New testcase.

2014-01-09  Richard Biener  <rguenther@suse.de>

PR tree-optimization/59715
* tree-cfg.h (split_critical_edges): Declare.
* tree-cfg.c (split_critical_edges): Export.
* tree-ssa-sink.c (execute_sink_code): Split critical edges.

* gcc.dg/torture/pr59715.c: New testcase.

From-SVN: r210147

11 years agobackport: [multiple changes]
Richard Biener [Wed, 7 May 2014 10:01:36 +0000 (10:01 +0000)] 
backport: [multiple changes]

2014-05-07  Richard Biener  <rguenther@suse.de>

Backport from mainline
2013-11-29  Richard Biener  <rguenther@suse.de>

PR tree-optimization/59334
* tree-ssa-dce.c (eliminate_unnecessary_stmts): Fix bug
in previous commit.

2013-11-28  Richard Biener  <rguenther@suse.de>

PR tree-optimization/59330
* tree-ssa-dce.c (eliminate_unnecessary_stmts): Simplify
and fix delayed marking of free calls not necessary.

* gcc.dg/torture/pr59330.c: New testcase.

2014-01-09  Richard Biener  <rguenther@suse.de>

PR tree-optimization/59715
* tree-cfg.h (split_critical_edges): Declare.
* tree-cfg.c (split_critical_edges): Export.
* tree-ssa-sink.c (execute_sink_code): Split critical edges.

* gcc.dg/torture/pr59715.c: New testcase.

From-SVN: r210146

11 years agore PR tree-optimization/57864 (ICE in bitmap_set_replace_value, at tree-ssa-pre.c...
Richard Biener [Wed, 7 May 2014 08:05:57 +0000 (08:05 +0000)] 
re PR tree-optimization/57864 (ICE in bitmap_set_replace_value, at tree-ssa-pre.c:862)

2014-05-07  Richard Biener  <rguenther@suse.de>

PR tree-optimization/57864
* tree-ssa-pre.c (phi_translate_1): Backport NAME case
simplification from mainline.  Do not lookup the VN
value-number here.

* gcc.dg/torture/pr57864.c: New testcase.

From-SVN: r210144

11 years agobackport: re PR tree-optimization/58246 (wrong code at -O1 and above)
Richard Biener [Wed, 7 May 2014 07:59:26 +0000 (07:59 +0000)] 
backport: re PR tree-optimization/58246 (wrong code at -O1 and above)

2014-05-07  Richard Biener  <rguenther@suse.de>

Backport from mainline
2013-08-29  Richard Biener  <rguenther@suse.de>

PR tree-optimization/58246
* tree-ssa-dce.c (mark_aliased_reaching_defs_necessary_1): Properly
handle the dominance check inside a basic-block.

* gcc.dg/torture/pr58246.c: New testcase.

From-SVN: r210141

11 years agoDaily bump.
GCC Administrator [Wed, 7 May 2014 00:16:43 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r210129

11 years agobackport: re PR tree-optimization/57417 (hang on volatile int array)
Richard Biener [Tue, 6 May 2014 14:22:41 +0000 (14:22 +0000)] 
backport: re PR tree-optimization/57417 (hang on volatile int array)

2014-05-06  Richard Biener  <rguenther@suse.de>

Backport from mainline
2013-05-27  Richard Biener  <rguenther@suse.de>

PR tree-optimization/57417
* tree-ssa-sccvn.c (set_ssa_val_to): Compare addresses using
get_addr_base_and_unit_offset.

* gcc.dg/torture/pr57417.c: New testcase.

From-SVN: r210110

11 years agoDaily bump.
GCC Administrator [Tue, 6 May 2014 00:16:37 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r210089

11 years agoDaily bump.
GCC Administrator [Mon, 5 May 2014 00:16:04 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r210050

11 years agoDaily bump.
GCC Administrator [Sun, 4 May 2014 00:16:21 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r210044

11 years agoPR 56919 Check _POSIX_MONOTONIC_CLOCK.
Janne Blomqvist [Sat, 3 May 2014 21:03:10 +0000 (00:03 +0300)] 
PR 56919 Check _POSIX_MONOTONIC_CLOCK.

2014-05-04  Janne Blomqvist  <jb@gcc.gnu.org>

Backport from 4.9
PR libfortran/56919
* intrinsics/system_clock.c (GF_CLOCK_MONOTONIC): Check
_POSIX_MONOTONIC_CLOCK as well.

From-SVN: r210042

11 years agoDaily bump.
GCC Administrator [Sat, 3 May 2014 00:16:22 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r210023

11 years agoDaily bump.
GCC Administrator [Fri, 2 May 2014 00:16:40 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r209993

11 years agoDaily bump.
GCC Administrator [Thu, 1 May 2014 00:16:31 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r209963

11 years agoDaily bump.
GCC Administrator [Wed, 30 Apr 2014 00:16:51 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r209919

11 years agoDaily bump.
GCC Administrator [Tue, 29 Apr 2014 00:16:20 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r209884

11 years agoDaily bump.
GCC Administrator [Mon, 28 Apr 2014 00:17:00 +0000 (00:17 +0000)] 
Daily bump.

From-SVN: r209845

11 years agoDaily bump.
GCC Administrator [Sun, 27 Apr 2014 00:16:33 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r209831

11 years agoDaily bump.
GCC Administrator [Sat, 26 Apr 2014 00:16:26 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r209815

11 years agore PR target/60941 (miscompilation of firefox javascript interpreter)
Eric Botcazou [Fri, 25 Apr 2014 10:47:44 +0000 (10:47 +0000)] 
re PR target/60941 (miscompilation of firefox javascript interpreter)

PR target/60941
* config/sparc/sparc.md (ashlsi3_extend): Delete.

From-SVN: r209793

11 years agoDaily bump.
GCC Administrator [Fri, 25 Apr 2014 00:16:18 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r209769

11 years agoDaily bump.
GCC Administrator [Thu, 24 Apr 2014 00:16:32 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r209728

11 years agobackport: re PR target/60909 (ICE: in extract_insn, at recog.c:2202 (unrecognizable...
Uros Bizjak [Wed, 23 Apr 2014 19:05:33 +0000 (21:05 +0200)] 
backport: re PR target/60909 (ICE: in extract_insn, at recog.c:2202 (unrecognizable insn) with -mrdrnd and __builtin_ia32_rdrand32_step())

        Backport from mainline
        2014-04-21  Uros Bizjak  <ubizjak@gmail.com>

        PR target/60909
        * config/i386/i386.c (ix86_expand_builtin)
        <case IX86_BUILTIN_RDRAND{16,32,64}_STEP>: Use temporary
        register for target RTX.

testsuite/ChangeLog:

        Backport from mainline
        2014-04-21  Uros Bizjak  <ubizjak@gmail.com>

        PR target/60909
        * gcc.target/i386/pr60909.c: New test.

From-SVN: r209718

11 years agoDaily bump.
GCC Administrator [Wed, 23 Apr 2014 00:16:37 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r209666

11 years agoDaily bump.
GCC Administrator [Tue, 22 Apr 2014 00:16:04 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r209550

11 years agoDaily bump.
GCC Administrator [Mon, 21 Apr 2014 00:16:25 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r209539

11 years agoDaily bump.
GCC Administrator [Sun, 20 Apr 2014 00:16:25 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r209531

11 years agoDaily bump.
GCC Administrator [Sat, 19 Apr 2014 00:16:22 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r209527

11 years agoDaily bump.
GCC Administrator [Fri, 18 Apr 2014 00:16:14 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r209510

11 years agoDaily bump.
GCC Administrator [Thu, 17 Apr 2014 00:16:15 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r209453

11 years agoDaily bump.
GCC Administrator [Wed, 16 Apr 2014 00:16:31 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r209436

11 years agoDaily bump.
GCC Administrator [Tue, 15 Apr 2014 00:16:18 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r209396

11 years agoDaily bump.
GCC Administrator [Mon, 14 Apr 2014 00:16:39 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r209349

11 years agobackport: re PR fortran/60810 (list directed io from array results in end of file)
Jerry DeLisle [Sun, 13 Apr 2014 02:05:02 +0000 (02:05 +0000)] 
backport: re PR fortran/60810 (list directed io from array results in end of file)

2014-04-12  Jerry DeLisle  <jvdelisle@gcc.gnu>

Backport from mainline
PR libfortran/60810
gfortran.dg/arrayio_13.f90: New test.

PR libfortran/60810
io/unit.c (is_trim_ok): If internal unit is array, do not trim.

From-SVN: r209345

11 years agoDaily bump.
GCC Administrator [Sun, 13 Apr 2014 00:16:27 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r209341

11 years agoDaily bump.
GCC Administrator [Sat, 12 Apr 2014 00:16:28 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r209330

11 years agoDaily bump.
GCC Administrator [Fri, 11 Apr 2014 00:16:34 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r209287

11 years agoDaily bump.
GCC Administrator [Thu, 10 Apr 2014 00:16:37 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r209258

11 years agoCheck if GCC uses assembler cfi support
Rainer Orth [Wed, 9 Apr 2014 14:57:48 +0000 (14:57 +0000)] 
Check if GCC uses assembler cfi support

* config/generic/asmcfi.h: Also check for
__GCC_HAVE_DWARF2_CFI_ASM.

From-SVN: r209245

11 years agoDaily bump.
GCC Administrator [Wed, 9 Apr 2014 00:16:33 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r209231

11 years agoDaily bump.
GCC Administrator [Tue, 8 Apr 2014 00:16:37 +0000 (00:16 +0000)] 
Daily bump.

From-SVN: r209206