Re: Drop frequencies from basic blocks

On 2017.11.05 at 11:55 +0100, Jan Hubicka wrote: > > On 2017.11.03 at 16:48 +0100, Jan Hubicka wrote: > > > this is updated patch which I have comitted after profiledbootstrapping > > > x86-64 > > > > Unfortunately, compiling tramp3d-v4.cpp is 6-7% slower after this patch. > > This happens with

Re: [Patch, testcase] PR82767 Fix scan-assembler patterns in i386/pr71321.c

On Mon, Nov 6, 2017 at 12:19 AM, Peryt, Sebastian wrote: >> On Sun, Nov 5, 2017 at 12:14 PM, Peryt, Sebastian >> wrote: >> > Hi, >> > >> > After r253934 gcc.target/i386/pr71321.c started to fail due to the wrong >> number of scan-assembler -

Re: [RFA][PATCH][PR target/82788] Remove uses of PROBE_INTERVAL in x86 target files

On Mon, Nov 6, 2017 at 5:47 AM, Jeff Law wrote: > On 11/03/2017 02:44 AM, Uros Bizjak wrote: >> Hello! >> >> -ENOCHANGELOG > Arggh. Downside of doing all the work on one machine, but mail > elsewhere. Attached with ChangeLog this time :-) > >> >> diff --git

[PATCH] Fix recent store-merging ICE (PR tree-optimization/82838)

Hi! force_gimple_operand_1 clears the *seq first and then adds statements there if any are needed. So calling force_gimple_operand_1 twice on the same seq is throwing away the earlier statements if any, rather than appending new statements to those. Fixed thusly, bootstrapped/regtested on

Re: [RFA][PATCH][PR target/82788] Remove uses of PROBE_INTERVAL in x86 target files

On 11/03/2017 02:44 AM, Uros Bizjak wrote: > Hello! > > -ENOCHANGELOG Arggh. Downside of doing all the work on one machine, but mail elsewhere. Attached with ChangeLog this time :-) > > diff --git a/gcc/testsuite/gcc.c-torture/execute/pr82788.c >

[PATCH, rs6000] Add support for usadv16qi and usadv8hi standard patterns

Hi, This patch adds support for vectorization of unsigned SAD expressions. SAD vectorization uses the usad pattern to represent a widening accumulation of SADs performed on a narrower type. The two cases in this patch operate on V16QImode and V8HImode, respectively, accumulating into V4SImode.

RE: [Patch, testcase] PR82767 Fix scan-assembler patterns in i386/pr71321.c

> On Sun, Nov 5, 2017 at 12:14 PM, Peryt, Sebastian > wrote: > > Hi, > > > > After r253934 gcc.target/i386/pr71321.c started to fail due to the wrong > number of scan-assembler - 2 instead of 3. This patch is fixing that. > > Are you sure that there is no problem with

Re: [Patch, testcase] PR82767 Fix scan-assembler patterns in i386/pr71321.c

On Sun, Nov 5, 2017 at 12:14 PM, Peryt, Sebastian wrote: > Hi, > > After r253934 gcc.target/i386/pr71321.c started to fail due to the wrong > number of scan-assembler - 2 instead of 3. This patch is fixing that. Are you sure that there is no problem with the code

Re: [patch] Do not report non-executed blocks in Ada coverage

On 11/03/2017 12:51 PM, Eric Botcazou wrote: Hi, as explained in https://gcc.gnu.org/ml/gcc-patches/2017-11/msg00105.html we don't (necessarily) want to report non-executed blocks in Ada. The couple of attached patches prevent this from happening by detecting whether we are dealing with an

Re: [PATCH] Zero vptr in dtor for -fsanitize=vptr.

On 11/03/2017 03:31 PM, Marek Polacek wrote: I think the last dg-output shouldn't have any regexps at the end, so: // { dg-output "0x\[0-9a-fA-F]*: note: object has invalid vptr" } Thanks for that. I'll fix it. Martin

Update comment in tree-vrp.h

Hello, looking at the history, it seems that this comment should be updated as in the following patch (the definition of struct value_range is right below and explains VR_RANGE and VR_ANTI_RANGE). I am not committing this as obvious because the text below does *not* actually describe

Re: [PATCH 4/7] [fortran, committed] Remove semicolon after do {} while (0) in match macros

On 11/05/2017 03:32 PM, Steve Kargl wrote: On Sun, Nov 05, 2017 at 11:05:02AM +0100, Tom de Vries wrote: this patch removes a semicolon after "do {} while (0)" in match macros. The patch contains this part. This allows the macros to be used in if-then-elses without curly braces. The

Re: [patch, fortran] Do loop (or index) interchange for FORALL and DO CONCURRENT, take 2

Hi Steve, it is assumed + expensive, and that loops which access later indices (which access memory + in bigger strides) should be moved to the first loops. Looks good to me with the typo fix. Committed as r254430. Thanks for the review (and for catching the typo; my fingers appear to

Re: [patch, fortran] Do loop (or index) interchange for FORALL and DO CONCURRENT, take 2

On Sun, Nov 05, 2017 at 12:03:29PM +0100, Thomas Koenig wrote: > + > +/* Code for index interchange for loops which are grouped together in DO > + CONCURRENT or FORALL statements. This is currently only applied if the > + iterations are grouped together in a single statement. > + > + For

Re: [Patch, fortran] PR81447 - [7/8] gfortran fails to recognize the exact dynamic type of a polymorphic entity that was allocated in a external procedure

Committed to 7-branch as revision 254429. Paul On 5 November 2017 at 12:44, Paul Richard Thomas wrote: > Hi Andre and Thomas, > > Thanks for looking at this. > > I left the condition as it is because it is the same practice as all > sorts of other parts of

Re: [Patch, fortran] PR78641 - [6/7/8 Regression] [OOP] ICE on polymorphic allocatable function in array constructor

Hi Steve, Committed to trunk as revision 254428. I will follow with 6- and 7-branches later in the week. Thanks Paul On 4 November 2017 at 14:46, Steve Kargl wrote: > On Sat, Nov 04, 2017 at 12:02:58PM +, Paul Richard Thomas wrote: >> Dear All, >> >>

Re: [PATCH 4/7] [fortran, committed] Remove semicolon after do {} while (0) in match macros

On Sun, Nov 05, 2017 at 11:05:02AM +0100, Tom de Vries wrote: > > this patch removes a semicolon after "do {} while (0)" in match macros. The patch contains this part. > This allows the macros to be used in if-then-elses without curly braces. The patch does not contain any changes that

Re: [Patch, fortran] PR81447 - [7/8] gfortran fails to recognize the exact dynamic type of a polymorphic entity that was allocated in a external procedure

Hi Andre and Thomas, Thanks for looking at this. I left the condition as it is because it is the same practice as all sorts of other parts of gfortran. That said, Thomas's suggestion is I think the right one. Committed revision as revision 254427. 7-branch will come later. Regards Paul On 4

[wwwdocs] Tweak subscription information at projects/gupc.html

The double colon in the mailto:: must have been a typo, and the other links sometimes gives troubles, so shorten this overall. Applied. Gerald Index: projects/gupc.html === RCS file: /cvs/gcc/wwwdocs/htdocs/projects/gupc.html,v

[wwwdocs] Remove blackfin.uclinux.org from readings.html

blackfin.uclinux.org has been resulting in Firefox complaining for months The owner of blackfin.uclinux.org has configured their website improperly. To protect your information from being stolen, Firefox has not connected to this website. and on top of that now their certificate has

[wwwdocs] Adjust link to gcc.opensuse.org

We actually do not need the trailing slash, and the server now redirects to https. (It also redirects further, but that may not be an ultimately stable URL, so I'm adding this to my exception list.) Applied. Gerald Index: benchmarks/index.html

[libstdc++,doc] docs.oracle.com has switched to https

I just applied the patch below. Gerald 2017-11-05 Gerald Pfeifer * doc/xml/manual/abi.xml: Move docs.oracle.com references to https. Index: doc/xml/manual/abi.xml === --- doc/xml/manual/abi.xml

[Patch, testcase] PR82767 Fix scan-assembler patterns in i386/pr71321.c

Hi, After r253934 gcc.target/i386/pr71321.c started to fail due to the wrong number of scan-assembler - 2 instead of 3. This patch is fixing that. 2017-11-05 Sebastian Peryt PR testsuite/82767 * gcc.target/i386/pr71321.c: Fix invalid testcase. Is it ok

[wwwdocs] Adjust link to Knuth's page in projects/prefetch.html

Applied. Gerald Index: projects/prefetch.html === RCS file: /cvs/gcc/wwwdocs/htdocs/projects/prefetch.html,v retrieving revision 1.30 diff -u -r1.30 prefetch.html --- projects/prefetch.html 17 May 2017 16:49:01 - 1.30

[patch, fortran] Do loop (or index) interchange for FORALL and DO CONCURRENT, take 2

Hello world, the attached patch now includes a new option to warn about a loop interchange, plus a test case using that option. Regression-tested. OK for trunk? Regards Thomas 2017-11-05 Thomas Koenig PR fortran/82471 * lang.opt

Re: Drop frequencies from basic blocks

> On 2017.11.03 at 16:48 +0100, Jan Hubicka wrote: > > this is updated patch which I have comitted after profiledbootstrapping > > x86-64 > > Unfortunately, compiling tramp3d-v4.cpp is 6-7% slower after this patch. > This happens with an LTO/PGO bootstrapped gcc using --enable-checking=release.

[PATCH 7/7][committed] Remove semicolon after do {} while (0) in DEF_SANITIZER_BUILTIN

Hi, Consider this code in gcc/asan.c: ... #define DEF_SANITIZER_BUILTIN(ENUM, NAME, TYPE, ATTRS) \ do { \ decl = add_builtin_function ("__builtin_" NAME, TYPE, ENUM, \ BUILT_IN_NORMAL, NAME, \

[PATCH 6/7] [libcpp,committed] Remove semicolon after do {} while (0) in BUF_APPEND

Hi, this patch removes a semicolon after "do {} while (0)" in BUF_APPEND. This allows the macro to be used in if-then-elses without curly braces. Committed as obvious. Thanks, - Tom 2017-11-02 Tom de Vries PR other/82784 * lex.c (BUF_APPEND):

[PATCH 5/7][committed] Remove semicolon after ASM_OUTPUT_BEFORE_CASE_LABEL macro body

Hi, this patch removes a semicolon after ASM_OUTPUT_BEFORE_CASE_LABEL macro bodies. This allows the macro to be used in if-then-elses without curly braces. Bootstrapped and reg-tested on x86_64. Committed as obvious. Thanks, - Tom 2017-11-02 Tom de Vries

[PATCH 3/7] [graphite, committed] Remove semicolon after do {} while (0) in DEBUG_PRINT

Hi, this patch removes a semicolon after "do {} while (0)" in DEBUG_PRINT. This allows the macro to be used in if-then-elses without curly braces. Committed as obvious. Thanks, - Tom 2017-11-02 Tom de Vries PR other/82784 *

[PATCH 4/7] [fortran, committed] Remove semicolon after do {} while (0) in match macros

Hi, this patch removes a semicolon after "do {} while (0)" in match macros. This allows the macros to be used in if-then-elses without curly braces. Thanks, - Tom 2017-11-02 Tom de Vries PR other/82784 * parse.c (match, matcha, matchs, matcho,

[PATCH 2/7] [libquadmath, committed] Remove semicolon after do {} while (0) in MPN_MUL_N_RECURSE

Hi, this patch removes a semicolon after "do {} while (0)" in MPN_MUL_N_RECURSE. This allows the macro to be used in if-then-elses without curly braces. Committed as obvious. Thanks, - Tom 2017-11-02 Tom de Vries PR other/82784 * printf/gmp-impl.h

[PATCH 1/7] [libsanitizer, committed] Remove semicolon after do {} while (0) in macro body

Hi, this removes a semicolon after "do {} while (0)" in several macro bodies in libsanitizer. This allows the macros to be used in if-then-elses without curly braces. Committed as obvious. Thanks, - Tom 2017-11-02 Tom de Vries PR other/82784 *

Re: Drop frequencies from basic blocks

On 2017.11.03 at 16:48 +0100, Jan Hubicka wrote: > this is updated patch which I have comitted after profiledbootstrapping x86-64 Unfortunately, compiling tramp3d-v4.cpp is 6-7% slower after this patch. This happens with an LTO/PGO bootstrapped gcc using --enable-checking=release. On X86_64: