Re: [PATCH] Fix warnings for profiledbootstrap (PR bootstrap/77788)

> Note the input.c change may not be necessary once the input.c cleanups are > done, but I don't think it'd be actively harmful. Yes, thanks, apparently I am not yet finished with that cleanup. But this highlights the data flow of this parameter is just too obfuscated yet. Bernd.

Re: [C++ PATCH] P0258R2 - add __has_unique_object_representations

On Wed, Oct 05, 2016 at 10:43:00PM -0400, Jason Merrill wrote: > On Wed, Oct 5, 2016 at 4:02 PM, Jakub Jelinek wrote: > > +case INTEGER_TYPE: > > + return TYPE_PRECISION (t) == GET_MODE_PRECISION (TYPE_MODE (t)); > > + > > +case BOOLEAN_TYPE: > > + /* For bool

[PATCH, i386]: Remove remaining -Wimplicit-fallthrough warnings from sse.md

... and introduce some related cleanups. 2016-10-06 Uros Bizjak * config/i386/sse.md (andnot3): Add FALLTHRU comments. Introduce ssesuffix variable. (3): Ditto. (*3): Ditto. Bootstrapped and regression tested on x86_64-linux-gnu {,-m32}. Will commit later

Re: fold x ^ y to 0 if x == y

On Wed, 5 Oct 2016, Jeff Law wrote: > On 07/08/2016 05:17 AM, Richard Biener wrote: > > On Fri, 8 Jul 2016, Prathamesh Kulkarni wrote: > > > > > Hi Richard, > > > For the following test-case: > > > > > > int f(int x, int y) > > > { > > >int ret; > > > > > >if (x == y) > > > ret =

[PATCH] Fix PR77839

The following fixes PR77839 by implementing some more lattice-transition constraints in SCCVN -- basically make it TOP -> constant value -> value -> BOTTOM. This avoids the iteration seen for the testcase in question. Bootstrapped and tested on x86_64-unknown-linux-gnu, applied to trunk.

[PATCH] Assert on invalid bitmap iterations

The following guards against (some) remove-current-bit cases. It would have ICEd for PR77855 instead of producing wrong code. Bootstrap / regtest running on x86_64-unknown-linux-gnu. Comments? Thanks, Richard. 2016-10-06 Richard Biener * bitmap.c

Re: [PATCH 3/3] Extend -falign-FOO=N to N[,M[,N2[,M2]]]

On 09/30/2016 07:54 PM, Denys Vlasenko wrote: +struct target_flag_state { + /* Each falign-foo can generate up to two levels of alignment: + -falign-foo=N,M[,N2,M2] */ + struct align_flags x_align_loops[2]; + struct align_flags x_align_jumps[2]; + struct align_flags x_align_labels[2]; +

Re: [C++ PATCH] P0258R2 - add __has_unique_object_representations

On 06/10/16 12:15 +0200, Jakub Jelinek wrote: On Thu, Oct 06, 2016 at 11:05:19AM +0100, Jonathan Wakely wrote: >I would think that we want the same answer for bool and enums: Either >the whole type size participates in the value representation, but only >certain values have defined behavior (so

Re: [RFA] [PR tree-optimization/71661] Fix forwarder removal when new loops are exposed

On Wed, Oct 5, 2016 at 5:58 PM, Jeff Law wrote: > > Removal of forwarder blocks is a two step process. > > First we build a worklist of potential forwarder blocks, Then we iterate > over that worklist attempting to remove each potential forwarder block in > the worklist. > > The

Re: [gomp4] correct a tile bug with combined loops in fortran

On 10/05/16 23:14, Cesar Philippidis wrote: Nathan spotted a bug where the fortran FE wasn't lowering the tile clause properly in combined parallel/kernels loops. The problem was that gfc_trans_omp_d has two sources of clauses, code->ext.omp_clauses and do_clauses. Because of the way that the

[PATCH] Fix PR77855

PRE in prune_clobbered_mems iterates over a bitmap, eventually removing the current bit. That's not safe in case the current element gets removed (that will terminate the iteration as the next element is then NULL). The following fixes this by keeping a queue of size one (it's safe to remove

Re: [C++ PATCH] P0258R2 - add __has_unique_object_representations

On 05/10/16 22:43 -0400, Jason Merrill wrote: On Wed, Oct 5, 2016 at 4:02 PM, Jakub Jelinek wrote: +case INTEGER_TYPE: + return TYPE_PRECISION (t) == GET_MODE_PRECISION (TYPE_MODE (t)); + +case BOOLEAN_TYPE: + /* For bool values other than 0 and 1 should

Re: [PATCH][fold-const] Fix native_encode_real for HFmode constants

On Wed, Oct 5, 2016 at 4:06 PM, Kyrill Tkachov wrote: > Hi all, > > I encountered a wrong-code issue with my WIP store merging pass when it was > trying to encode HFmode constants. > I am using native_encode_real to write the constants to a byte array and > it's

Re: [PATCH] Fix ICE for sparc targets in function_arg_record_value (PR target/77759)

> gcc/ > PR target/77759 > * config/sparc/sparc.c (classify_data_t): Remove unused int_regs field. > (classify_registers): Don't set int_regs. > (function_arg_slotno): Don't initialise int_regs. Check slotno is > within range for empty structs, just like int

Re: [C++ PATCH] P0258R2 - add __has_unique_object_representations

On Thu, Oct 06, 2016 at 11:05:19AM +0100, Jonathan Wakely wrote: > >I would think that we want the same answer for bool and enums: Either > >the whole type size participates in the value representation, but only > >certain values have defined behavior (so we should return true), or > >only certain

Re: [C++/66443] deleted ctor and vbase construction

On 10/05/16 19:49, Jason Merrill wrote: But DR 1658 says that B::B is *not* deleted (because A is not a potentially constructed subobject). oo, I'd not noticed that. It certainly might make things much simpler. Essentially a check for ABSTRACT_TYPE_P somewhere in that code. nathan

Re: [PR77862] ICE in add_equivalence

On Thu, Oct 6, 2016 at 6:45 AM, kugan wrote: > Hi, > > In VRP we can generate new SSA variables while performing the propagation, > at least in EVRP. Therefore, when we access array vr_values we have to check > num_vr_values. This caused ICE for the PR77862.

[PATCH, C++] Warn on redefinition of builtin functions (PR c++/71973)

Hi! Currently C++ does not warn at all when built-in functions are re-defined with a different signature, while C does warn on that even without -Wall. Thus I'd like to propose a -Wall enabled warning for that in C++ only. Initially I tried to warn unconditionally but that made too many tests

[PATCH, ARM] PR71607: New approach to arm_disable_literal_pool

Hello, This patch tackles the issue reported in PR71607. This patch takes a different approach for disabling the creation of literal pools. Instead of disabling the patterns that would normally transform the rtl into actual literal pools, it disables the creation of this literal pool rtl by

Re: [PATCH v3] gcc/config/tilegx/tilegx.c (tilegx_function_profiler): Save r10 to stack before call mcount

Hello Maintainers: Is this patch OK? Please help check it when you have time (at least for me, it passes all related test and comparation). And I shall continue to find and fix another issues about tilegx. Thanks. On 6/4/16 21:25, cheng...@emindsoft.com.cn wrote: > From: Chen Gang

Re: [PATCH, C++] Warn on redefinition of builtin functions (PR c++/71973)

Hi Bernd, Not familiar with this area but one comment below... On 06/10/16 15:12, Bernd Edlinger wrote: Hi! Currently C++ does not warn at all when built-in functions are re-defined with a different signature, while C does warn on that even without -Wall. Thus I'd like to propose a -Wall

Re: [PATCH 15/16] RTL frontend (rtl1), on top of dump reader

On 10/05/2016 06:15 PM, David Malcolm wrote: +;; MEM[(struct isl_obj *)] = _obj_map_vtable; +(insn 1045 0 1046 2 (set (reg:SI 480) +(high:SI (symbol_ref:SI ("isl_obj_map_vtable") +[flags 0xc0] +))) + y.c:12702 -1 + (nil)) +(insn 1046 1045

Re: [C++ PATCH] P0258R2 - add __has_unique_object_representations

On Thu, Oct 6, 2016 at 6:54 AM, Jonathan Wakely wrote: > On 06/10/16 12:15 +0200, Jakub Jelinek wrote: >> >> On Thu, Oct 06, 2016 at 11:05:19AM +0100, Jonathan Wakely wrote: >>> >>> >I would think that we want the same answer for bool and enums: Either >>> >the whole type size

[PATCH] Fix PR77879

The following fixes a bug in PTA analysis manifesting in wrong pointer equality folding. Bootstrap and regtest running on x86_64-unknown-linux-gnu. Richard. 2016-10-06 Richard Biener PR tree-optimization/77879 * tree-ssa-structalias.c (handle_const_call):

Re: [PATCH] Assert on invalid bitmap iterations

On Thu, 6 Oct 2016, Richard Biener wrote: > > The following guards against (some) remove-current-bit cases. It > would have ICEd for PR77855 instead of producing wrong code. > > Bootstrap / regtest running on x86_64-unknown-linux-gnu. > > Comments? Found one additional issue. Bootstrapped

[PATCH] Clear DECL_CHAIN of local vars

The following patch makes sure to clear DECL_CHAIN of gimple_bind_vars once we dissolve that (but makes sure to preserve the BLOCK_VARS tail). This should allow better GC in the face of remove-unused-locals eventually removing the last reference to a decl (and not keep it live via DECL_CHAIN of

Re: [PATCH, RFC] gcov: dump in a static dtor instead of in an atexit handler

On 09/30/2016 02:31 PM, Rainer Orth wrote: this would be i386-pc-solaris2.12. I'm not sure if the constructor priority detection works in a cross scenario. Hi. By the way, I tried to test the cross-compiler: $ ../configure --disable-bootstrap --enable-languages=c,c++,fortran

Re: [PATCH 1/3] rs6000: Use "dg-warning ... 0" in a few places

On Thu, Oct 06, 2016 at 07:57:35PM +, Segher Boessenkool wrote: > GCC used to give warnings for line # 1 for these, but it prints no > line number anymore. Adjust. LGTM. > > > 2016-10-06 Segher Boessenkool > > gcc/testsuite/ > *

Re: C++ PATCH for P0135, C++17 guaranteed copy elision

Here's an update that handles more cases. commit 726447a210cf92a85e8f9d014b85d958a0af62b5 Author: Jason Merrill Date: Wed Oct 5 19:31:09 2016 -0400 C++17 copy elision improvements. * call.c (build_temp, convert_like_real): Don't re-copy

Re: [PATCH 3/3] testsuite: Use -Wno-psabi instead of pruning compiler messages

On Thu, Oct 06, 2016 at 11:08:52PM +0200, Jakub Jelinek wrote: > On Thu, Oct 06, 2016 at 07:57:37PM +, Segher Boessenkool wrote: > > It is much nicer to avoid a warning than to manually prune it from the > > compiler output. This patch adds -Wno-psabi to various testcases, and > > removes

Re: [PATCH, C++] Warn on redefinition of builtin functions (PR c++/71973)

On 10/06/16 16:14, Kyrill Tkachov wrote: > > @@ -1553,7 +1588,7 @@ duplicate_decls (tree newdecl, tree olddecl, bool > > /* Whether or not the builtin can throw exceptions has no >bearing on this declarator. */ > - TREE_NOTHROW (olddecl) = 0; > + TREE_NOTHROW (olddecl) =

[PATCH] Optimize a >= 0 && a < b if VR info says b >= 0 (PR tree-optimization/77664)

Hi! For signed a, if we see a >= 0 && a < b and from VR know that b >= 0, we can optimize those 2 comparisons into one unsigned - (unsigned) a < (unsigned) b. Similarly for a < 0 || a > b (and also for a <= b in the first and a >= b in the second). The patch implements it in the reassoc

Re: [PATCH/AARCH64] Have the verbose cost model output output be controllable

On Fri, Sep 23, 2016 at 10:46 PM, Andrew Pinski wrote: > Hi, > As reported in PR 61367, the aarch64 back-end is too verbose when it > is dealing with the cost model. I tend to agree, no other back-end is > this verbose. So I decided to add an option to enable this verbose >

Re: [PATCH] accept flexible arrays in struct in unions (c++/71912 - [6/7 regression])

I'm asking you to clarify the logic. It seems that your change to fldtype affects these two tests: if (eltype == fldtype || TYPE_UNNAMED_P (eltype)) if (TREE_CODE (fldtype) != ARRAY_TYPE) ...but this is extremely subtle. It would be a lot clearer to check fld for

Re: [RFC] Extend ipa-bitwise-cp with pointer alignment propagation

On 6 October 2016 at 18:51, Jan Hubicka wrote: >> > >> > What do you mean by "for instance?" What are the other cases when it >> > happens? >> Well ipa_get_type() returned NULL for 481.wrf, and I assumed it was a >> fortran-only >> code-base but apparently it's a mix of C and

[EVRP] Register ranges for y in (x COND y) for Early VRP.

Hi, Attached patch Register ranges for y in (x COND y) for Early VRP. Bootstrapped and regression tested on x86_64-linux-gnu with no new regressions. Is this OK for trunk? Thanks, Kugan gcc/ChangeLog: 2016-10-06 Kugan Vivekanandarajah * tree-vrp.c

[PATCH, RS6000] improve builtin expansion of memcmp for p7

I've improved the builtin memcmp expansion so it avoids a couple of things that p7 and previous processors don't like. Performance on p7 is now never worse than glibc memcmp(). Bootstrap/regtest in progress on power7 ppc64 BE.  OK for trunk if testing passes? gcc/ChangeLog: 2016-10-06 Aaron

Re: [PATCH, RS6000] improve builtin expansion of memcmp for p7

Hi Aaron, On Thu, Oct 06, 2016 at 04:12:31PM -0500, Aaron Sawdey wrote: > I've improved the builtin memcmp expansion so it avoids a couple of > things that p7 and previous processors don't like. Performance on > p7 is now never worse than glibc memcmp(). Bootstrap/regtest in progress > on power7

[PATCH v2] aarch64: Add split-stack initial support

From: Adhemerval Zanella Changes from previous version: - Rewrite how to setup variadic argument: instead of using the hard_fp_offset value to setup the x10, save a fp value before stack allocation instead. This allows linker/gold to not require scan

Re: [PATCH] Optimize a >= 0 && a < b if VR info says b >= 0 (PR tree-optimization/77664)

On Thu, 6 Oct 2016, Jakub Jelinek wrote: For signed a, if we see a >= 0 && a < b and from VR know that b >= 0, we can optimize those 2 comparisons into one unsigned - (unsigned) a < (unsigned) b. Similarly for a < 0 || a > b (and also for a <= b in the first and a >= b in the second). I

Re: [PATCH] Optimize a >= 0 && a < b if VR info says b >= 0 (PR tree-optimization/77664)

On Thu, Oct 06, 2016 at 11:55:00PM +0200, Marc Glisse wrote: > On Thu, 6 Oct 2016, Jakub Jelinek wrote: > > >For signed a, if we see a >= 0 && a < b and from VR know that b >= 0, > >we can optimize those 2 comparisons into one unsigned - > >(unsigned) a < (unsigned) b. Similarly for a < 0 || a >

[RFC][VRP] Improve intersect_ranges

Hi, In vrp intersect_ranges, Richard recently changed it to create integer value ranges when it is integer singleton. Maybe we should do the same when the other range is a complex ranges with SSA_NAME (like [x+2, +INF])? Attached patch tries to do this. There are cases where it will be

Re: Patch, Split powerpc -mfloat128 into 2 parts

On Wed, Oct 05, 2016 at 06:18:50PM -0400, Michael Meissner wrote: > I have removed the #define of __float128, and instead created a keyword that > points to the _Float128 type. I added a little commentary on the use of > __ieee128 (which is needed to create the type when we aren't using >

Re: [PATCH 3/3] testsuite: Use -Wno-psabi instead of pruning compiler messages

On Thu, Oct 06, 2016 at 07:57:37PM +, Segher Boessenkool wrote: > It is much nicer to avoid a warning than to manually prune it from the > compiler output. This patch adds -Wno-psabi to various testcases, and > removes dg-prune-output, for the testcases where x86 and/or PowerPC had > those

Re: Possible patch for fortran/57910

> Le 6 oct. 2016 à 19:35, Louis Krupp a écrit : > > Dominique, > > Vous avez raison. I attached the wrong patch. I've resent the message with > the correct patch. Which works as expected. Thanks > > I tried to make pr69955.f90 run only on 64-bit Linux: > > ! {

Re: [PATCH] - improve sprintf buffer overflow detection (middle-end/49905)

On Thu, 6 Oct 2016, Martin Sebor wrote: > > The problem is that extracting via an integer of the opposite signedness > > is *not* defined unless the argument has a value representable in both > > types - both as a matter of ISO C rules on variadic functions, and as a > > practical matter of what

Re: [PATCH 2/3] rs6000: Warn for "GCC vector passed by reference" only with -Wpsabi

On Thu, Oct 06, 2016 at 07:57:36PM +, Segher Boessenkool wrote: > As suggested by Jakub in > https://gcc.gnu.org/ml/gcc-patches/2016-10/msg00305.html . > The warning name isn't so super for us, but the option is undocumented > anyway; maybe it means "warn for particularly surprising abi

Re: [PATCH] - improve sprintf buffer overflow detection (middle-end/49905)

On 10/05/2016 05:54 PM, Joseph Myers wrote: On Wed, 5 Oct 2016, Martin Sebor wrote: On 10/05/2016 05:11 PM, Joseph Myers wrote: On Wed, 5 Oct 2016, Martin Sebor wrote: may have been subjected to. Issuing a warning for a safe piece of code only on the basis that there might be some other

[PATCH] Make std::scoped_allocator_adaptor's OUTERMOST recursive

This fixes a TODO in which is one of our remaining C++11 conformance gaps. I also found that the constructors need to be constrained, otherwise is_constructible gives false positives, and then that causes uses_allocator to give false positives, which makes uses-allocator take a branch that

Re: [PATCH] 77864 Fix noexcept conditions for map/set default constructors

On 06/10/16 22:17 +0200, François Dumont wrote: Another approach is to rely on existing compiler ability to compute conditional noexcept when defaulting implementations. This is what I have done in this patch. The new default constructor on _Rb_tree_node_base is not a problem as it is not

Re: Patch, Split powerpc -mfloat128 into 2 parts

On Thu, 6 Oct 2016, Segher Boessenkool wrote: > On Wed, Oct 05, 2016 at 06:18:50PM -0400, Michael Meissner wrote: > > I have removed the #define of __float128, and instead created a keyword that > > points to the _Float128 type. I added a little commentary on the use of > > __ieee128 (which is

Re: [PATCH] [PR libcpp/71681] Fix handling header.gcc in subdirectories

On 09/08/2016 12:09 PM, Thomas Schwinge wrote: Hi! A few review comments: On Wed, 7 Sep 2016 20:19:20 +0300, Andris Pavenis wrote: This patch fixes handling header.gcc in subdirectories when command line option -remap has been used. (I have not yet looked up what

Re: [PATCH 3/3] Extend -falign-FOO=N to N[,M[,N2[,M2]]]

On 10/06/2016 11:47 AM, Bernd Schmidt wrote: On 09/30/2016 07:54 PM, Denys Vlasenko wrote: +struct target_flag_state { + /* Each falign-foo can generate up to two levels of alignment: + -falign-foo=N,M[,N2,M2] */ + struct align_flags x_align_loops[2]; + struct align_flags

[PATCH 2/3] rs6000: Warn for "GCC vector passed by reference" only with -Wpsabi

As suggested by Jakub in https://gcc.gnu.org/ml/gcc-patches/2016-10/msg00305.html . The warning name isn't so super for us, but the option is undocumented anyway; maybe it means "warn for particularly surprising abi differences". 2016-10-06 Segher Boessenkool

Re: Possible patch for fortran/57910

Dear Louis, > PR fortran/57910 > * trans-expr.c (gfc_add_interface_mapping): Don't try to > dereference call-by-value scalar argument. > > The patch seems to work without breaking other tests. From the patch, I think the PR number is wrong and should be 69955. The test fails on darwin with At

Re: Possible patch for fortran/57910

On Thu, Oct 06, 2016 at 07:04:36PM +0200, Dominique d'Humières wrote: > Dear Louis, > > > PR fortran/57910 > > * trans-expr.c (gfc_add_interface_mapping): Don't try to > > dereference call-by-value scalar argument. > > > > The patch seems to work without breaking other tests. > >From the patch, I

Re: Possible patch for fortran/57910

Dominique, Vous avez raison. I attached the wrong patch. I've resent the message with the correct patch. I tried to make pr69955.f90 run only on 64-bit Linux: ! { dg-do run { target x86_64-*-linux* } } I'm not sure there's a portable way to query virtual memory usage, and testing this on

Re: [PATCH] Delete GCJ

On Oct 6, 2016, at 9:56 AM, Rainer Orth wrote: > I wouldn't hard-fail, but completely disable objc-gc with an appropriate > warning. The Objective-C maintainers may have other preferences, though. gcc historically is fairly weak at complex configurations. I need

RTL frontend input format again (was Re: [PATCH 15/16] RTL frontend (rtl1), on top of dump reader)

On Thu, 2016-10-06 at 15:30 +0200, Bernd Schmidt wrote: > On 10/05/2016 06:15 PM, David Malcolm wrote: > > +;; MEM[(struct isl_obj *)] = _obj_map_vtable; > > +(insn 1045 0 1046 2 (set (reg:SI 480) > > +(high:SI (symbol_ref:SI ("isl_obj_map_vtable") > > +[flags 0xc0] > >

Re: [PATCH] Delete GCJ

> On 6 Oct 2016, at 17:56, Rainer Orth wrote: > this assumption may not hold, though: in Solaris 11+ where libgc is bundled, both 32 and 64-bit libs are present, as always. I'd also claim that for multilib testing in general, it's bad to test

Re: [gofrontend-dev] Go patch committed: avoid allocation of varargs parameter with optimize-allocs

On Thu, Oct 6, 2016 at 9:35 AM, Chris Manghane wrote: > > I noticed that this patch does not include the change go-gcc.cc that caused > stack-allocated memory to be incorrectly initialized. Can you add this patch > as well? Whoops, sorry for missing that. Committed with the

[PATCH 3/3] testsuite: Use -Wno-psabi instead of pruning compiler messages

It is much nicer to avoid a warning than to manually prune it from the compiler output. This patch adds -Wno-psabi to various testcases, and removes dg-prune-output, for the testcases where x86 and/or PowerPC had those (or did warn, so the testcase failed). Tested on powerpc64-linux {-m32,-m64},

Re: RTL frontend input format again (was Re: [PATCH 15/16] RTL frontend (rtl1), on top of dump reader)

On Thu, 2016-10-06 at 15:53 -0400, David Malcolm wrote: > On Thu, 2016-10-06 at 15:30 +0200, Bernd Schmidt wrote: > > On 10/05/2016 06:15 PM, David Malcolm wrote: > > > +;; MEM[(struct isl_obj *)] = _obj_map_vtable; > > > +(insn 1045 0 1046 2 (set (reg:SI 480) > > > +(high:SI

Re: [PATCH] 77864 Fix noexcept conditions for map/set default constructors

On 05/10/2016 14:13, Jonathan Wakely wrote: On 05/10/16 14:10 +0200, Marc Glisse wrote: On Wed, 5 Oct 2016, Jonathan Wakely wrote: I added conditional noexcept to maps and sets, but forgot to account for the comparison function, which could throw when constructed. IMO you are fighting a

Re: [C++ PATCH] P0258R2 - add __has_unique_object_representations

OK.

[PATCH 1/3] rs6000: Use "dg-warning ... 0" in a few places

GCC used to give warnings for line # 1 for these, but it prints no line number anymore. Adjust. 2016-10-06 Segher Boessenkool gcc/testsuite/ * gcc.target/powerpc/warn-1.c: Change line number in dg-warning. * gcc.target/powerpc/warn-2.c: Ditto.

Re: [PATCH] Fix PR77826

On Wed, 5 Oct 2016, Richard Biener wrote: The following will fix PR77826, the issue that in match.pd matching up two things uses operand_equal_p which is too lax about the type of the toplevel entity (at least for integer constants). Bootstrap / regtest pending on x86_64-unknown-linux-gnu.

Re: [PATCH] Delete GCJ

> On 6 Oct 2016, at 17:42, Rainer Orth wrote: > >>> Here's what I tested. This requires a boehm-gc version 7.0 or later >>> (having the >>> header files in a gc subdirectory). Depending on your available library, it >>> only builds the GC enabled library for the

Re: [PATCH] Delete GCJ

On 06.10.2016 18:46, Iain Sandoe wrote: > >> On 6 Oct 2016, at 17:42, Rainer Orth wrote: >> > Here's what I tested. This requires a boehm-gc version 7.0 or later (having the header files in a gc subdirectory). Depending on your available library,

Re: [PATCH] Delete GCJ

Hi Matthias, >>> this assumption may not hold, though: in Solaris 11+ where libgc is >>> bundled, both 32 and 64-bit libs are present, as always. I'd also claim >>> that for multilib testing in general, it's bad to test different >>> multilibs with different configurations, so I'd rather have

Re: Patch, Split powerpc -mfloat128 into 2 parts

On Thu, Oct 06, 2016 at 10:26:34PM +, Joseph Myers wrote: > On Thu, 6 Oct 2016, Segher Boessenkool wrote: > > > On Wed, Oct 05, 2016 at 06:18:50PM -0400, Michael Meissner wrote: > > > I have removed the #define of __float128, and instead created a keyword > > > that > > > points to the

Re: Possible patch for fortran/57910

On 10/06/2016 03:52 PM, Louis Krupp wrote: I've attached an updated patch for pr69955. It works just as you said. Please let me know if this or my patch for pr57910 is OK to check in. Louis Both are OK. Thanks. Jerry

Re: Set nonnull attribute to ptr_info_def based on VRP

Hi Richard, Thanks for the review. On 09/08/16 18:58, Richard Biener wrote: On Tue, Aug 9, 2016 at 12:58 AM, kugan wrote: Hi Jakub, Thanks for the review. On 08/08/16 16:40, Jakub Jelinek wrote: On Mon, Aug 08, 2016 at 01:36:51PM +1000, kugan wrote:

Re: [PATCH] [ARC] Disable compact casesi patterns for arcv2

* Claudiu Zissulescu [2016-09-30 15:52:03 +0200]: > Please find the updated patch, > Claudiu > > gcc/ > 2016-05-09 Claudiu Zissulescu > > * common/config/arc/arc-common.c (arc_option_optimization_table): > Remove compact

Re: Fix ICE due to map typespecs with different sized charlens being copied

On Wed, Oct 05, 2016 at 04:40:50PM -0400, Fritz Reese wrote: > When union type symbols are compared and they contain maps containing > characters with different lengths, their type symbols should have > different backend declarations otherwise the gimple tree explodes. > Strangely enough the

Re: Possible patch for fortran/57910

I've attached an updated patch for pr69955. It works just as you said. Please let me know if this or my patch for pr57910 is OK to check in. Louis On Thu, 06 Oct 2016 14:30:29 -0700 Dominique d'Humières wrote > > > Le 6 oct. 2016 à 19:35, Louis Krupp

Re: [PATCH] Delete GCJ

On 06.10.2016 20:00, Mike Stump wrote: > On Oct 6, 2016, at 9:56 AM, Rainer Orth wrote: >> I wouldn't hard-fail, but completely disable objc-gc with an appropriate >> warning. The Objective-C maintainers may have other preferences, though. I think I can't do that

[Committed/AARCH64] Add comment before each set of cores in aarch64-cores.def

Like https://gcc.gnu.org/ml/gcc-patches/2015-11/msg02145.html but updated for the new cores that were added and add the character which represents the implementors. Committed as obvious after a bootstrap and test. Thanks, Andrew * config/aarch64/aarch64-cores.def: Add a comment before each set

Fortran, committed: ICE (segfault) with deferred-length strings (pr 57910)

Fixed in revision 240850.

Go patch committed: allocate temporary slices on stack

This patch by Than McIntosh changes the Go frontend to allocate the temporary slice created to pass to the append call on the stack. This is a step toward https://golang.org/issue/17304. Bootstrapped and ran Go testsuite on x86_64-pc-linux-gnu. Committed to mainline. Ian Index:

[PATCH, RS6000, Committed] increase buf size in rs6000_elf_asm_out_{constructor,destructor}

It seems we now have analysis that concludes these buffers may possibly overflow. This broke bootstrap on ppc64 BE. Bootstrap passed on ppc64 BE power7. Committing as pre-approved by Segher. 2016-10-06 Aaron Sawdey * config/rs6000/rs6000.c

Re: PR35503 - warn for restrict pointer

On 22 September 2016 at 23:15, Joseph Myers wrote: > On Thu, 22 Sep 2016, Prathamesh Kulkarni wrote: > >> Would that be acceptable ? I am not sure how to make %Z check if the >> argument has type vec * >> since vec is not really a builtin C type. >> Could you suggest me a

Fortran, committed: Memory leak with array constructor and derived type (pr 69955).

Fixed in revision 240851.

Re: [PATCH] Assert on invalid bitmap iterations

On 10/06/2016 03:15 AM, Richard Biener wrote: The following guards against (some) remove-current-bit cases. It would have ICEd for PR77855 instead of producing wrong code. Bootstrap / regtest running on x86_64-unknown-linux-gnu. Comments? Thanks, Richard. 2016-10-06 Richard Biener

Re: [PATCH 15/16] RTL frontend (rtl1), on top of dump reader

Let me just make a first pass over this for minor/obvious issues. +we have little control of the input to that specific pass. We "control over" maybe? +The testsuite is below @file{gcc/testsuite/rtl.dg}. Not sure this needs to be in the manual (I have similar doubts about the entire

[PATCH][WIP] Add install-dvi Makefile targets

Currently the build machinery handles install-pdf and install-html targets, but no install-dvi target. This patch is a step towards fixing that. Note that I have only tested with --enable-languages=c,c++,lto,objc,obj-c++. Thus, target hooks will probably also have to be added for the languages I

Re: fold x ^ y to 0 if x == y

On 10/06/2016 01:18 AM, Richard Biener wrote: It ought to be easy to fold x ^ y to zero when x == y (famous last words). I'm sure I'll regret saying that when I go to look at how to twiddle DOM appropriately. Interesting idea. Though it get's (theoretically) interesting for ternary ops where

Re: [C++ PATCH] P0258R2 - add __has_unique_object_representations

On Thu, Oct 06, 2016 at 08:46:26AM -0400, Jason Merrill wrote: > >>> >I would think that we want the same answer for bool and enums: Either > >>> >the whole type size participates in the value representation, but only > >>> >certain values have defined behavior (so we should return true), or > >>>

[PING] [PATCH, ARM] Reduce stack usage in sha512 (PR target/77308)

Hi, I'd like to ping for this patch: https://gcc.gnu.org/ml/gcc-patches/2016-09/msg02362.html Thanks, Bernd. On 09/30/16 15:33, Bernd Edlinger wrote: > On 09/30/16 12:14, Bernd Edlinger wrote: >> Eric Botcazou wrote: A comment before the SETs and a testcase would be nice. IIRC we do

Re: [RFC] Extend ipa-bitwise-cp with pointer alignment propagation

> > > > What do you mean by "for instance?" What are the other cases when it > > happens? > Well ipa_get_type() returned NULL for 481.wrf, and I assumed it was a > fortran-only > code-base but apparently it's a mix of C and fortran. Yep, I also have expreinece that the K style declarations are

[PATCH] Allow `make tags` to work from top-level directory

The libdecnumber, libgcc, and libobjc subdirectories are missing TAGS targets in their Makefiles. The attached patch causes them to be skipped when running `make tags`. ChangeLog entry: 2016-10-06 Eric Gallager * Makefile.def: Mark libdecnumber, libgcc, and

Re: [PATCH] Delete GCJ

Hi Matthias, >> Here's what I tested. This requires a boehm-gc version 7.0 or later >> (having the >> header files in a gc subdirectory). Depending on your available library, it >> only builds the GC enabled library for the default multilib library, and just >> skips over building the

Re: [RFA] [PR tree-optimization/71661] Fix forwarder removal when new loops are exposed

On 10/06/2016 04:11 AM, Richard Biener wrote: The remove_forwarder_block hunk is redundant (I've fixed a similar bug there by checking in tree_forwarder_block_p). This function doesn't operate from a worklist and thus shouldn't have the issue. Agreed. I hadn't really looked closely at

Re: Use version namespace in normal mode

On 01/10/16 14:22 +0200, François Dumont wrote: On 30/09/2016 17:40, Jonathan Wakely wrote: On 29/09/16 21:59 +0200, François Dumont wrote: Hi I think _GLIBCXX_BEGIN_NAMESPACE_ALGO should default to _GLIBCXX_BEGIN_NAMESPACE_VERSION when parallel mode is not active. Same for

Re: [PATCH] Delete GCJ

On 06.10.2016 18:14, Matthias Klose wrote: > On 05.10.2016 18:28, Jeff Law wrote: >> On 10/04/2016 12:39 PM, Iain Sandoe wrote: I don't know who wants to review it, but if people want me to, Ok. The idea is that if ObjC is the last remaining user in tree for boehm-gc, then