Re: [fortran, patch] IEEE intrinsic modules (ping)

FX writes: > Index: gcc/testsuite/gfortran.dg/ieee/ieee_6.f90 > === > --- gcc/testsuite/gfortran.dg/ieee/ieee_6.f90 (revision 0) > +++ gcc/testsuite/gfortran.dg/ieee/ieee_6.f90 (revision 0) > @@ -0,0 +1,78 @@ > +! { dg-do run } > +!

Re: Warn when returning the address of a temporary (middle-end) v2

( https://gcc.gnu.org/ml/gcc-patches/2014-06/msg01692.html ) On Sun, 22 Jun 2014, Marc Glisse wrote: Hello, I followed the advice in this discussion: https://gcc.gnu.org/ml/gcc-patches/2014-04/msg00269.html and here is a new patch. I made an effort to isolate a path in at least one subcase s

Re: [i386] Replace builtins with vector extensions

On Sat, Jun 28, 2014 at 6:53 PM, Marc Glisse wrote: > There is always a risk, but then even with builtins I think there was a > small risk that an RTL optimization would mess things up. It is indeed > higher if we expose the operation to the optimizers earlier, but it would be > a bug if an "optim

Re: [fortran, patch] IEEE intrinsic modules (ping)

> This may raise inexact, see C11 7.6.2.3. Installed as obvious. Thanks! FX

Re: [i386] Replace builtins with vector extensions

On Sun, 29 Jun 2014, Ulrich Drepper wrote: I think the Arm definitions come from a different angle. It's new, there is no assumed semantics. Is it that new? I thought it was implemented based on a rather precise specification by ARM. Again, I don't really know arm. For the x86 intrinsics

Re: Optimize type streaming

In addition of pr61644 and pr61646, this commit breaks a lot of fortran tests with -flto -O0. TIA Dominique FAIL: gfortran.dg/actual_array_constructor_1.f90 -g -flto (internal compiler error) FAIL: gfortran.dg/alloc_comp_assign_6.f90 -g -flto (internal compiler error) FAIL: gfortran.dg/allo

[wwwdocs] Clean up references to Bugzilla, use https,...

Consistently use bug tracker versus bug database, Bugzilla (where possible), or the like. Avoid some references altogether and refer to our bugs/ pages instead of directly linking to Bugzilla in some. Use https for all remaining links to our Bugzilla instance. Committed. Gerald Index: news.html

[wwwdocs,libstdc++] Remove libstdc++/lib3styles.css

We don't use this any more, so let's remove it. Gerald 2014-06-29 Gerald Pfeifer * lib3styles.css: Remove. * index.html: Remove reference to lib3styles.css. Index: index.html === RCS file: /cvs/gcc/wwwdocs/htdoc

[Patch, Fortran] Coarray fixes for select type/associate and type of derived components

This patch fixes some issues with polymorphic coarrays. I still have to fix at least one issue. Fixed by the patch: a) The temporary pointer generated with SELECT TYPE has to be a coarray. That's fixed with the resolve.c patch. The comment is also bogus: The comment is correct – and gfortran

[wwwdocs] Remove disabled http://gcc.gnu.org/benchresults/

This has been commented for more than twelve years. Yank it for real. Applied. Gerald Index: index.html === RCS file: /cvs/gcc/wwwdocs/htdocs/benchmarks/index.html,v retrieving revision 1.31 diff -u -r1.31 index.html --- index.html

[PATCH] gcc/c/c-decl.c: Let NEWDECL dereference shared data to fix segment fault issue.

NEWDECL dereferences shared data of OLDDECL, or when free NEWDECL, also have effect with OLDDECL. At present, only fix the related reference for current issue. If find another related issue, can follow this fix. The related git number is 71e19e54060804493e13748613077b0e69c0cfd9, and the related in

Re: [PATCH] gcc/c/c-decl.c: Let NEWDECL dereference shared data to fix segment fault issue.

On Sun, 29 Jun 2014, Chen Gang wrote: > NEWDECL dereferences shared data of OLDDECL, or when free NEWDECL, also > have effect with OLDDECL. At present, only fix the related reference for > current issue. If find another related issue, can follow this fix. I'm afraid I can't make any sense of this

Re: [fortran,patch] Binding label can be any initialisation expression

> Works as advertized and even fixes pr38838. > Thanks for the patch. Thanks for the review, committed to trunk as rev. 212123 FX

[patch libstdc++] Add xmethods for std::vector and std::unique_ptr

GDB now supports xmethods in its Python API: https://sourceware.org/gdb/current/onlinedocs/gdb/Xmethods-In-Python.html This feature will be available in GDB starting version 7.8 (which has not yet been released, but has been branched). The attached patch adds xmethods to the classes std::vector an

Re: [Committed] Fix lto.c compiling

Andrew Pinski writes: > I committed this as obvious. The changelog says it all. I think the problem is that LTO bootstraps frequently error out with -Werror. That incentives LTO users to bootstrap with --disable-werror. [e.g. currently the graphite libraries don't build this way] The main pro

Re: [wwwdocs,Patch] Update Fortran part in gcc-4.9/changes.html

On Tue, 25 Jun 2013, Tobias Burnus wrote: Comments are welcome. Otherwise, I intent to commit it soon. (Well, comments are also welcome after committal ;-) Since you are offering that ;-), I find the the wording "(but inexact)" in When STOP or ERROR STOP are used to terminate the executio

Re: [Patch] Patch set for regex instantiation

On 11/01/14 19:48 -0500, Tim Shen wrote: Here're 4 patches that finally led the _Compiler's instantiation and some other optimization for compiling time. 1) Create class _ScannerBase to make _Scanner pithier. Move const static members to src/c++11/regex.cc. I think it might be time to revisit

Re: [GSoC] Addition of ISL AST generation to Graphite

I'm pretty it's the following that causes bootstrap to fail for me: 2014-06-29 Roman Gareev * Makefile.in: Add the compilation of graphite-isl-ast-to-gimple.o. * common.opt: Add new switch fgraphite-code-generator=[isl|cloog]. * flag-types.h: A

Re: [GSoC] Addition of ISL AST generation to Graphite

On 29/06/2014 19:31, Gerald Pfeifer wrote: I'm pretty it's the following that causes bootstrap to fail for me: 2014-06-29 Roman Gareev * Makefile.in: Add the compilation of graphite-isl-ast-to-gimple.o. * common.opt: Add new switch fgraphite-code-genera

[committed-PATCH] Add missing HAVE_cloog guards

I just committed the attached patch to fix the gcc bootstrap problem reported by Gerald. At the moment my bootstrap already passed the point where it previously failed, but did not yet complete. I committed it is very likely to be both correct and to unbreak the build. I will report back aft

Re: [PATCH, alpha]: Wrap {un,}aligned_store sequence with memory blockages.

On Fri, Jun 27, 2014 at 10:04 PM, Richard Henderson wrote: > On 06/27/2014 10:04 AM, Uros Bizjak wrote: >> This happened due to the way stores to QImode and HImode locations are >> implemented on non-BWX targets. The sequence reads full word, does its >> magic to the part and stores the full word

Re: [Patch] Patch set for regex instantiation

On Sun, Jun 29, 2014 at 10:18 AM, Jonathan Wakely wrote: > Alternatively, we could just make them local statics instead of member > data and still remove the duplciation (see attached). This is a good idea. Can we use constexpr for the functions/static variables? -- Regards, Tim Shen

[wwwdocs] testing/: Remove reference to obsolete autobuilder; use https

Amazing what collateral damage^Wroom for improvement I'm running into as part of the http->https conversion. Applied. Gerald Index: testing/index.html === RCS file: /cvs/gcc/wwwdocs/htdocs/testing/index.html,v retrieving revision 1.

[PATCH v2, rtl]: Teach _.barriers and _.eh_range passes to not split a call and its corresponding CALL_ARG_LOCATION note.

On Fri, Jun 27, 2014 at 9:11 PM, Richard Henderson wrote: > On 06/26/2014 02:15 PM, Uros Bizjak wrote: >> * except.c (emit_note_eh_region_end): New helper function. >> (convert_to_eh_region_ranges): Use emit_note_eh_region_end to >> emit EH_REGION_END note. > > This bit looks good. > >

Re: Optimize type streaming

> In addition of pr61644 and pr61646, this commit breaks a lot of > fortran tests with -flto -O0. Hello, the problem here is that we have POINTER_TYPE that points to array of variable length (not sure why it happens only with -O0). The ICE is introduced by the fact that we stream the type inside of

Re: [PATCH] gcc/c/c-decl.c: Let NEWDECL dereference shared data to fix segment fault issue.

> On Sun, 29 Jun 2014, Chen Gang wrote: > > > NEWDECL dereferences shared data of OLDDECL, or when free NEWDECL, also > > have effect with OLDDECL. At present, only fix the related reference for > > current issue. If find another related issue, can follow this fix. > > I'm afraid I can't make any

[wwwdocs,Java] Make links in faq.html and java/faq.html relative

Ever since 2007-10-01 Gerald Pfeifer * gcc_release: Do not include copies of faq.html in releases, and remove all the explicit support code to pull in web pages. there was no reason faq.html (formerly included as FAQ) needed absolute links unless wrt. services only available

Re: [Patch] Patch set for regex instantiation

2014-06-29 20:47 GMT+02:00 Tim Shen : > On Sun, Jun 29, 2014 at 10:18 AM, Jonathan Wakely wrote: >> Alternatively, we could just make them local statics instead of member >> data and still remove the duplciation (see attached). > > This is a good idea. Can we use constexpr for the functions/static

Re: [GSoC] Addition of ISL AST generation to Graphite

On Sun, 29 Jun 2014, Tobias Grosser wrote: sorry for the breakage. This seems to be caused by forgotten include guards. I attached a patch (not bootstrapped yet), that should fix the issue. I will start a gcc (non-graphite) bootstrap to see if it works and will commit it immediately after. Pl

Re: [committed-PATCH] Add missing HAVE_cloog guards

On Sun, 29 Jun 2014, Tobias Grosser wrote: I just committed the attached patch to fix the gcc bootstrap problem reported by Gerald. Thanks! I closed PR bootstrap/61650 and added that reference to your ChangeLog entry. Gerald

Re: [PATCH] gcc/c/c-decl.c: Let NEWDECL dereference shared data to fix segment fault issue.

On Sun, Jun 29, 2014 at 10:00:34PM +0200, Jan Hubicka wrote: > > On Sun, 29 Jun 2014, Chen Gang wrote: > > > > > NEWDECL dereferences shared data of OLDDECL, or when free NEWDECL, also > > > have effect with OLDDECL. At present, only fix the related reference for > > > current issue. If find anoth

Re: [DOC Patch] Attribute 'naked'

On Thu, 26 Jun 2014, Gerald Pfeifer wrote: >> 2014-06-17 David Wohlferd >> >> * doc/extend.texi (Function Attributes): Update 'naked' attribute >> doc. > This patch makes sense to me and I've seen feedback on an earlier > iteration that lead to this now. So, I am planning to commit this

Re: [PATCH] gcc/c/c-decl.c: Let NEWDECL dereference shared data to fix segment fault issue.

On Sun, 29 Jun 2014, Jan Hubicka wrote: > So Cheng is right we need to clear DECL_STRUCT_FUNCTION or we end up > ggc_freeing it that later ICEs in push_cfun. We do not need to clear > the other fields. Thanks for the explanation. This patch is OK with the typo fixes Marek has noted. -- Jos

Re: [PATCH 1/7] Add missing documentation of four IPA-CP params

On Tue, 10 Jun 2014, Gerald Pfeifer wrote: > On Wed, 21 May 2014, Martin Jambor wrote: >> +@item ipa-cp-loop-hint-bonus >> +When IPA-CP determines that a cloning candidate would make the number >> +of iterations of a loop known, it adds a bonus of >^

SRA: don't drop clobbers

Hello, with this patch on top of https://gcc.gnu.org/ml/gcc-patches/2014-06/msg02315.html we finally warn for the testcase of PR 60517. The new function is copied from init_subtree_with_zero right above. I guess it might be possible to merge them into a single function, if desired. I don't und

Re: [PATCH] gcc/c/c-decl.c: Let NEWDECL dereference shared data to fix segment fault issue.

On 06/30/2014 05:23 AM, Joseph S. Myers wrote: > On Sun, 29 Jun 2014, Jan Hubicka wrote: > >> So Cheng is right we need to clear DECL_STRUCT_FUNCTION or we end up >> ggc_freeing it that later ICEs in push_cfun. We do not need to clear >> the other fields. > > Thanks for the explanation. Thi

[PATCH, rs6000, testsuite] Skip gfortran.dg/round_4.f90 for PowerPC Linux

Hi, The test in gfortran.dg/round_4.f90, introduced in GCC 4.9, checks for correct behavior of different rounding modes. However, for quad-precision floating-point, it requires that the number 0.100481 be exactly represented. Since the PowerPC long double implemen

RE: [PATCH] Fix PR61306: improve handling of sign and cast in bswap

> From: Richard Biener [mailto:richard.guent...@gmail.com] > Sent: Friday, June 27, 2014 4:49 PM > > FIne with me now. Commited. Best regards, Thomas

[Committed] [PATCH, loop2_invariant, 1/2] Check only one register class

On 26 June 2014 05:30, Jeff Law wrote: > On 06/11/14 04:05, Zhenqiang Chen wrote: >> >> On 10 June 2014 19:06, Steven Bosscher wrote: >>> >>> On Tue, Jun 10, 2014 at 11:22 AM, Zhenqiang Chen wrote: Hi, For loop2-invariant pass, when flag_ira_loop_pressure is enabled, func

Re: [PATCH] Pass correct memory attributes for build constant

>>test.c: >>extern bar(unsigned char p[3][2]); >>void foo(int i) >>{ >> unsigned char data[3][2] = {{1,1}, {1,0}, {1,1}}; >> >> bar(data); >>} > First, note, I'm not an ARM expert. However, the first question I > have is are we sure the initializer is always going to be suitably > aligned?

[tree-optimization/61607] Look through SSA_NAME_VALUE chains

SSA_NAME_VALUE is, in effect, a chain of values. ie, it's possible for SSA_NAME_VALUE of any given SSA_NAME to refer to another SSA_NAME. In many cases it is advantageous to look deeper into those chains, particularly when simplifying conditionals for jump threading. The problem with simpl