Re: Fix streaming of target optimization/option nodes

Hi, actually this patch break fortran, I get streaming error in: lto1: internal compiler error: in streamer_get_pickled_tree apparently picking error_mark_node for variable constructor results in reading integer_type... Honza > Hi, > the testcase in PR ipa/61324 fails because it is compiled with

[PATCH] gcc/testsuite/gcc.dg/typeof-2.c

gcc/testsuite/ChangeLog 2014-12-15 Sebastian Huber * gcc.dg/typeof-2.c: Add checks for non-atomic types. --- gcc/testsuite/gcc.dg/typeof-2.c | 31 +++ 1 file changed, 31 insertions(+) diff --git a/gcc/testsuite/gcc.dg/typeof-2.c b/gcc/testsuite/gcc.dg/typeo

PING for gcc/flag-types.h (was: Re: [Patch, gcc/flag-types.h + Fortran] PR54687 - Fortran options cleanup)

The Fortran part of the .opt changes has been approved, but for Enum(), some bits had to be moved to gcc/flag-types.h – and review for that file is still missing. https://gcc.gnu.org/ml/gcc-patches/2014-12/msg01068.html Tobias Tobias Burnus wrote: This patch cleans up Fortran's option handli

PR 62053

Hi, this patch fixes type inconsistency in between main variant and its variant. This is caused by build_cplus_array_type first laying down the type and only then linking it into vairant chain. Bootstrapped/regtested x86_64-linux, OK? Honza PR tree-optimization/62053 * tree.c (

Re: [Patch, Fortran] PR 63674: procedure pointer and non/pure procedure

Janus Weil wrote: 2014-12-14 12:00 GMT+01:00 FX : Good point, thank you. Updated patch attached. I guess I still new formal approval by someone with reviewer status … OK Thanks, committed as r218717. Can you change "non-pure" to "impure"? That would better match the Fortran naming, where "i

Fix streaming of target optimization/option nodes

Hi, the testcase in PR ipa/61324 fails because it is compiled with -O0 and linked with -O2. This should not matter anymore if there wasn't the following problem in streamer that makes us to merge all default nodes across units. Bootstrapped/regtested x86_64-linux, plan to commit it after more te

PR ipa/61602 (ICE in single_use propagation)

Hi, this testcase is another place where we leave dead symbols in the symtab and thus ipa-single-use propagation fails. Bootstrapped/regtested x86_64-linux, comitted. PR ipa/61602 * gcc.dg/torture/pr61602.c: New testcase. * cgraph.h (ipa_discover_readonly_nonaddressable_

Fix ICE in insert_to_assembler_name_hash

Hi, this patch fixes ICE in symbol_table::insert_to_assembler_name_hash in cases where C++ FE produces a symbol but refuses to give it an assembler name (because it is abstract but w/o abstract flag set). Bootstrapped/regtested x86_64-linux, comitted. Index: ChangeLog

Fix ICE on missing binfo

Hi, this patch fixes more of PR64043: when we have OBJ_TYPE_REF without a BINFO, we should not try to devirtualize. This is most conveniently done by modifying virtual_method_call_p. I still need to figure out why -O0 / -O2 combination with LTO leads to -fdevirtualize being set. It should not.

[Patch, libstdc++/64302, libstdc++/64303] Fix match_results iterators and regex_token_iterator

Bootstraped and tested :) I'm not sure if it's safe to directly backport it to 4.9, since one inline function (_M_normalize_result) is added; but at least we can manually inline them by copy & paste. -- Regards, Tim Shen commit 6be0eb9cfb66c0229c7121890371f9d45189321d Author: timshen Date: S

[C++ patch] do not make extern inlines as needed when not optimizing

Hi, while looking into PR ipa/61324 I noticed that extern inline functions are produced by C++ FE even at -O0. This seems like considerable waste especially because these may trigger instantiations. Bootstrepped/regtested x86_64-linux, OK? Honza * decl2.c (decl_needed_p): When not optim

Do not build callgraph for external functions when inlining

Hi, this patch implements analogous change I did to ipa.c reachability to remove extern functions early when not optimizing. These are common in C++ code and it is not effective to process them until after inliner. Bootstrapped/regtested x86_64-linux, will commit it shortly. Honza * cgr

Re: [off-list] A fixincludes warning-removal patch

On 12/14/14 14:07, Uros Bizjak wrote: Hello Bruce! I have posted a small fixincludes patch [1] that fixes a warning during the fixincludes build, Please apply. However, I'm sure it is not "the last remaining warning" since new warnings get invented. :) Thank you - Bruce 2014-12-09 Uros Biz

flattening cfgloop.h

Hi, The attached patch flattens cfgloop.h. Created new file cfgloopanal.h which exports function prototypes from cfgloopanal.c. Rest of the patch involves restructuring includes. Boostrapped and tested on x86_64-unknown-linux-gnu with ISL enabled. Built on all targets in config-list.mk with langua

[match-and-simplify] set simplify::capture_max to 0 if pattern contains no captures

Caused segfault for pattern containing no captures at: info.safe_grow_cleared(capture_max + 1); in capture_info::capture_info artificial test-case: (define_predicates integer_zerop) (simplify (bit_not integer_zerop) { build_zero_cst (type); }) * genmatch.c (simplify::simplify): Set simplify

Re: [match-and-simplify] print capture name

On 15 December 2014 at 02:03, Prathamesh Kulkarni wrote: > Print name of capture in patterns written by user. > > * genmatch.c > (capture::name): New. > (capture::capture): New default argument. > (parse_capture): Pass id to capture::capture. > (print_operand): Print name of capture if ava

[match-and-simplify] print capture name

Print name of capture in patterns written by user. * genmatch.c (capture::name): New. (capture::capture): New default argument. (parse_capture): Pass id to capture::capture. (print_operand): Print name of capture if available. Thanks, Prathamesh

Go patch committed: Fix crash when tracked field used in global init

My recent patch for tracked fields had a bug that caused a compiler crash when they were used in a global initializer. This patch fixes it. Bootstrapped and ran Go testsuite on x86_64-unknown-linux-gnu. Committed to mainline. Ian diff -r 5f3aca373b7d go/expressions.cc --- a/go/expressions.cc Sat

[PATCH] Remove dead code in omp-low.c

From: Andi Kleen Since the accelerator changes omp_expand_local is not used anymore. Remove the dead function and another function used by it. gcc/: 2014-12-14 Andi Kleen * omp-low.c (build_omp_regions_root): Remove dead function. (omp_expand_local): Dito. * omp-low.

[C++ Patch] PR 58882

Hi, in order to handle correctly the testcase, thus do not ICE on it, I think we have simply to replace the existing gcc_unreachable with a normal error call. Additionally, I noticed that we aren't handling correctly things like constexpr user-defined type-conversion operators, thus I added t

Re: PATCH: PR rtl-optimization/64037: Miscompilation with -Os and enum class : char parameter

On December 14, 2014 3:08:28 PM CET, "H.J. Lu" wrote: >On Mon, Dec 08, 2014 at 11:16:58PM +0100, Uros Bizjak wrote: >> Hello! >> >> > On Tue, Nov 25, 2014 at 5:05 PM, H.J. Lu >wrote: >> >> On Tue, Nov 25, 2014 at 7:01 AM, Richard Biener >> >> wrote: >> >>> On Tue, Nov 25, 2014 at 1:57 PM, H.J.

[gomp4] Merge trunk r218679 (2014-12-12) into gomp-4_0-branch

Hi! In r218719, I have committed a merge from trunk r218679 (2014-12-12) into gomp-4_0-branch. Grüße, Thomas pgp_QoHKRWj1T.pgp Description: PGP signature

Re: PATCH: PR rtl-optimization/64037: Miscompilation with -Os and enum class : char parameter

On Mon, Dec 08, 2014 at 11:16:58PM +0100, Uros Bizjak wrote: > Hello! > > > On Tue, Nov 25, 2014 at 5:05 PM, H.J. Lu wrote: > >> On Tue, Nov 25, 2014 at 7:01 AM, Richard Biener > >> wrote: > >>> On Tue, Nov 25, 2014 at 1:57 PM, H.J. Lu wrote: > Hi, > > The enclosed testcase fails

Re: [Patch, Fortran] PR 63674: procedure pointer and non/pure procedure

2014-12-14 12:00 GMT+01:00 FX : >> Good point, thank you. Updated patch attached. >> I guess I still new formal approval by someone with reviewer status … > > OK Thanks, committed as r218717. Cheers, Janus

Re: [PATCH][testsuite] Mark as UNSUPPORTED tests that don't fit into tiny memory model

domi...@lps.ens.fr (Dominique Dhumieres) writes: >> Here it is. Committed as r218662 to get things going again. > > The same change should be done for libitm and libatomic. And boehm-gc. Andreas. -- Andreas Schwab, sch...@linux-m68k.org GPG Key fingerprint = 58CA 54C7 6D53 942B 1756 01D3 44D5

Re: [Patch, Fortran] PR 63674: procedure pointer and non/pure procedure

> Good point, thank you. Updated patch attached. > I guess I still new formal approval by someone with reviewer status … OK

Re: [patch c++]: Fix PR/63996

++.dg/cpp1y/pr63996.C -std=c++14 (test for excess errors) Excess errors: /usr/local/gcc/gcc-20141214/gcc/testsuite/g++.dg/cpp1y/pr63996.C:6:10: error: ISO C++ forbids variable length array 'a' [-Wvla] Andreas. -- Andreas Schwab, sch...@linux-m68k.org GPG Key fingerprint = 58CA 54C7

Re: [Patch, Fortran] PR 63674: procedure pointer and non/pure procedure

>> Regtested on x86_64-unknown-linux-gnu. Ok for trunk? > > s/'%s'/%qs/g > nowadays. Good point, thank you. Updated patch attached. I guess I still new formal approval by someone with reviewer status ... Cheers, Janus Index: gcc/fortran/resolve.c =