1. 13 Jun, 2016 18 commits
  2. 12 Jun, 2016 4 commits
  3. 11 Jun, 2016 8 commits
  4. 10 Jun, 2016 10 commits
    • fold-const: Don't access bit fields with too big mode (PR71310) · 95c18dd0
      Currently, optimize_bit_field_compare reads the bitfield in word_mode
      if it can.  If the bit field is normally accessed in a smaller mode,
      this might be a violation of the memory model, although the "extra"
      part of the read is not used.  But also, previous stores to the bit
      field will have been done in the smaller mode, and then bigger loads
      from it cause a LHS problem.
      
      
      	PR middle-end/71310
      	* fold-const.c (optimize_bit_field_compare): Don't try to use
      	word_mode unconditionally for reading the bit field, look at
      	DECL_BIT_FIELD_REPRESENTATIVE instead.
      
      gcc/testsuite/
      	PR middle-end/71310
      	* gcc.target/powerpc/pr71310.c: New testcase.
      
      From-SVN: r237319
      Segher Boessenkool committed
    • re PR tree-optimization/71478 (ICE in tree-ssa-reassoc.c after r236564) · b97d37b4
      gcc/testsuite/ChangeLog:
      
      2016-06-11  Kugan Vivekanandarajah  <kuganv@linaro.org>
      
      	PR middle-end/71478
      	* gcc.dg/pr71478.c: New test.
      
      gcc/ChangeLog:
      
      2016-06-11  Kugan Vivekanandarajah  <kuganv@linaro.org>
      
      	PR middle-end/71478
      	* tree-ssa-reassoc.c (reassociate_bb): Remove (-1) from ops list for
      	vector integer type.
      
      From-SVN: r237318
      Kugan Vivekanandarajah committed
    • re PR middle-end/71494 (label as value in nested function) · ec1c20db
      	PR middle-end/71494
      	* tree-nested.c (convert_nonlocal_reference_stmt): For GIMPLE_GOTO
      	without LABEL_DECL, set *handled_ops_p to false instead of true.
      
      	* gcc.c-torture/execute/pr71494.c: New test.
      
      From-SVN: r237317
      Jakub Jelinek committed
    • re PR c/68657 ("gcc -Werror=sign-conversion test.i" shows invalid:… · 50b15873
      re PR c/68657 ("gcc -Werror=sign-conversion test.i" shows invalid: -Wsign-conversion is not an option that controls warnings)
      
      	PR c/68657
      	* c.opt (Wpsabi): Add Warning flag.
      
      	* gcc.target/i386/pr68657.c: New test.
      
      From-SVN: r237316
      Jakub Jelinek committed
    • re PR inline-asm/68843 (ICE with "u" input constraint) · 900bcda3
      	PR inline-asm/68843
      	* gcc.target/i386/pr68843-2.c: Add dg-do run and empty dg-options.
      	(test): Add -masm=intel alternatives.
      
      From-SVN: r237315
      Jakub Jelinek committed
    • PR c/71392 - SEGV calling integer overflow built-ins with a null pointer · 4d926e34
      gcc/ChangeLog:
      2016-06-10  Martin Sebor  <msebor@redhat.com>
      
      	PR c/71392
      	* builtin-attrs.def (ATTR_NOTHROW_NONNULL_LEAF_LIST): New macro.
      	(ATTR_NOTHROW_NONNULL_TYPEGENERIC_LEAF): Same.
      	* builtins.def (BUILT_IN_SADD_OVERFLOW, BUILT_IN_SADDL_OVERFLOW): Use
      	them.
      	(BUILT_IN_SADDLL_OVERFLOW, BUILT_IN_SSUB_OVERFLOW): Same.
      	(BUILT_IN_SSUBL_OVERFLOW, BUILT_IN_SSUBLL_OVERFLOW): Same.
      	(BUILT_IN_SMUL_OVERFLOW, BUILT_IN_SMULL_OVERFLOW): Same.
      	(BUILT_IN_SMULLL_OVERFLOW, BUILT_IN_UADD_OVERFLOW): Same.
      	(BUILT_IN_UADDL_OVERFLOW, BUILT_IN_UADDLL_OVERFLOW): Same.
      	(BUILT_IN_USUB_OVERFLOW, BUILT_IN_USUBL_OVERFLOW): Same.
      	(BUILT_IN_USUBLL_OVERFLOW, BUILT_IN_UMUL_OVERFLOW): Same.
      	(BUILT_IN_UMULL_OVERFLOW, BUILT_IN_UMULLL_OVERFLOW):
      
      gcc/ada/ChangeLog:
      2016-06-10  Martin Sebor  <msebor@redhat.com>
      
      	PR c/71392
      	* gcc/ada/gcc-interface/utils.c (handle_nonnull_attribute): Accept
      	the nonnull attribute in type-generic builtins.
      
      gcc/c-family/ChangeLog:
      2016-06-10  Martin Sebor  <msebor@redhat.com>
      
      	PR c/71392
      	* gcc/c-family/c-common.c (handle_nonnull_attribute): Accept
      	the nonnull attribute in type-generic builtins.
      
      gcc/lto/ChangeLog:
      2016-06-10  Martin Sebor  <msebor@redhat.com>
      
      	PR c/71392
      	* gcc/lto/lto-lang.c (handle_nonnull_attribute): Accept the nonnull
      	attribute in type-generic builtins.
      
      gcc/testsuite/ChangeLog:
      2016-06-10  Martin Sebor  <msebor@redhat.com>
      
      	PR c/71392
      	* c-c++-common/builtin-arith-overflow-1.c: Add test cases.
      
      From-SVN: r237314
      Martin Sebor committed
    • arm.h (pool_vector_label, [...]): Remove. · a102ee4b
      2016-06-10  Bernd Edlinger  <bernd.edlinger@hotmail.de>
      
              * config/arm/arm.h (pool_vector_label,
              return_used_this_function): Remove.
      
      From-SVN: r237313
      Bernd Edlinger committed
    • re PR tree-optimization/71335 (wrong code at -O2 and -O3 in 32-bit and 64-bit… · ad071b2b
      re PR tree-optimization/71335 (wrong code at -O2 and -O3 in 32-bit and 64-bit modes on x86_64-linux-gnu)
      
      	PR tree-optimization/71335
      	* tree-ssa-threadbackward.c (profitable_jump_thread_path): Filter out
      	zero length paths here.
      	(convert_and_register_jump_thread_path): Remove hacks related to
      	duplicated blocks in the jump thread path.
      	(fsm_find_control_statement_thread_paths): Avoid putting the same
      	block on the thread path twice, but ensure the thread path is
      	unchanged from the caller's point of view.
      
      	PR tree-optimization/71335
      	* gcc.c-torture/execute/pr71335.c: New test.
      
      From-SVN: r237312
      Jeff Law committed
    • predict.c (predict_loops): Remove PRED_LOOP_BRANCH. · 66b4d261
      	* predict.c (predict_loops): Remove PRED_LOOP_BRANCH.
      	* predict.def (PRED_LOOP_BRANCH): Remove.
      
      From-SVN: r237311
      Jan Hubicka committed
    • Remove platform-specific details from must-tail-call-2.c messages · 2704efe8
      gcc/testsuite/ChangeLog:
      	* gcc.dg/plugin/must-tail-call-2.c: Remove all details from
      	the various "cannot tail-call: " messages.
      
      From-SVN: r237310
      David Malcolm committed