Re: PING: PATCH [9/n]: Prepare x32: PR middle-end/47383: ivopts miscompiles Pmode != ptr_mode

On Tue, Jul 5, 2011 at 7:07 PM, H.J. Lu wrote: > On Tue, Jul 5, 2011 at 8:24 AM, Richard Guenther > wrote: >> On Tue, Jul 5, 2011 at 4:56 PM, Ulrich Weigand wrote: >>> H.J. Lu wrote: >>> > However, this still seems odd to me, as I had understood the address in > a TARGET_MEM_REF needs

Re: Ping^2: TARGET_HAVE_NAMED_SECTIONS cleanup

On Tue, Jul 5, 2011 at 9:16 PM, Joseph S. Myers wrote: > Ping^2.  The patch > is still pending > review. Ok if there are no objections from target maintainers in 24h. Thanks, Richard. > -- > Joseph S. Myers > jos...@codesourcery.com >

Fix PR testsuite/49542

This fixes the couple of failures that have appeared recently in the vectorizer testsuite on SPARC 64-bit. Ira diagnosed that they are legitimate fallouts of a recent improvement to the vectorizer: http://gcc.gnu.org/ml/gcc-patches/2011-04/msg02210.html so the attached patch simply XFAILs them

[PATCH] Fix configure --with-cloog

Hello This patch fix an issue while building with cloog and gmp installed in a custom separate directories. How to reproduce : - Make sure you've installed cloog and gmp in separate directories (ie ${WITH-CLOOG-PATH}/lib doesn't contain libgmp) - Make sure neither gmp nor cloog is not installed

[PATCH, ARM, iWMMXt][2/5]: intrinsic head file change

Hi, It is the second part of iWMMXt maintenance. *config/arm/mmintrin.h: Revise the iWMMXt intrinsics head file. Fix some intrinsics and add some new intrinsics. Thanks, Xinyu 2_mmintrin.diff Description: 2_mmintrin.diff

[PATCH, ARM, iWMMXt][0/5]: iWMMXt intrinsics maintenance and pipeline description

Hi, Since the patch of iWMMXt intrinsics maintenance and pipeline description is too big to review, I subdivide it into five parts. 1 ARM generic code change 2 iWMMXt intrinsic head file change. 3 iWMMXt built in define and expand 4 WMMX machine description 5 WMMX pipeline description These fiv

[PATCH, ARM, iWMMXt][1/5]: ARM code generic change

Hi, It is the first part of iWMMXt maintenance. *config/arm/arm.c (arm_option_override): Enable iWMMXt with VFP. iWMMXt and NEON are incompatible. iWMMXt unsupported under Thumb-2 mode. (arm_expand_binop_builtin): Accept immediate op (with mode VOID) *config/arm/arm.md: Resettle include locat

[PATCH, ARM, iWMMXt][3/5]: built in define and expand

Hi, It is the third part of iWMMXt maintenance. *config/arm/arm.c (enum arm_builtins): Add/fix iWMMXt built-in fcode. (builtin_description bdesc_2arg): Add/fix iWMMXt built in. (builtin_description bdesc_1arg): Same. (arm_init_iwmmxt_builtins): Define new built-in function type. Add/fix iWMMX

[PATCH, ARM, iWMMXt][4/5]: WMMX machine description

Hi, It is the fourth part of iWMMXt maintenance. *config/arm/arm.c (arm_output_iwmmxt_shift_immediate): New function. Serving for WMMX immediate shift insn output. (arm_output_iwmmxt_tinsr): New function. Serving for tinsr insn output. *config/arm/arm-protos.h: Add new functions protos. *config

[PATCH, ARM, iWMMXt][5/5]: pipeline description

Hi, It is the fifth part of iWMMXt maintenance. *config/arm/marvell-f-iwmmxt.md: New file. Add Marvell WMMX pipeline description. Thanks, Xinyu 5_pipeline.diff Description: 5_pipeline.diff

Cleanup Solaris ASM_SPEC handling

As already noted in the Solaris configuration cleanup patch, the ASM_SPEC handling on Solaris can be simplified. This patch does this, also as a prerequisite for a followup to provide a 64-bit default Solaris/x86 configuration. The basic observation is that there's a common part handled by both S

Provide 64-bit default Solaris/x86 configuration (PR target/39150)

There has long been some clamoring for a amd64-*-solaris2 configuration similar to sparcv9-sun-solaris2. I've resisted this for quite some time, primarily because it doubles the maintenance effort of testing both the 32-bit default and 64-bit default configurations. After the recent cleanup patch

[testsuite] Don't XFAIL gcc.dg/tree-ssa/20030807-7.c (PR tree-optimization/49647)

As described in the PR, gcc.dg/tree-ssa/20030807-7.c seems to XPASS everywhere. This patch removes the xfail. Tested with the appropriate runtest invocation on i386-pc-solaris2.10. Ok for mainline? Rainer 2011-07-06 Rainer Orth PR tree-optimization/49647 * gcc.dg/t

[Path,AVR]: Improve loading of 32-bit constants

For loading a 32-bit constant in a register, there is room for improvement: * SF can be handled the same way as SI and therefore the patch adds a peep2 to produce a *reload_insf analogon to *reload_insi. * If the destination register overlaps NO_LD_REGS, values already loaded into some other

Re: [testsuite] Don't XFAIL gcc.dg/tree-ssa/20030807-7.c (PR tree-optimization/49647)

On Wed, 6 Jul 2011, Rainer Orth wrote: > As described in the PR, gcc.dg/tree-ssa/20030807-7.c seems to XPASS > everywhere. This patch removes the xfail. > > Tested with the appropriate runtest invocation on i386-pc-solaris2.10. > > Ok for mainline? Ok. Thanks, Richard. > Rainer > > >

Re: [PATCH] Fix configure --with-cloog

> Hello > > This patch fix an issue while building with cloog and gmp installed in > a custom separate directories. > > How to reproduce : >  - Make sure you've installed cloog and gmp in separate directories > (ie ${WITH-CLOOG-PATH}/lib doesn't contain libgmp) >  - Make sure neither gmp nor cloog

PATCH TRUNK: [gcc/configure.ac] Generate GCCPLUGIN_VERSION_* macros

Hello All, The fie plugin-version.h is now generated by gcc/configure.ac. It contains version information (about the GCC supposed to load the plugin compiled with it) as constant strings. But I think it will also help some plugin developers if that file (which is packaged in gcc-4.6-plugin-dev on

[Patch,testsuite]: target-supports.exp: Disable -fprofile-generate for AVR

AVR tests will fail if -fprofile-generate is given because that is not (yet) implemented. CCed avr port maintainer in the case they have objections. Ok to commit? Johann * lib/target-supports.exp (check_profiling_available): Disable profiling with -fprofile-generate for target a

Re: [PATCH] Address lowering [1/3] Main patch

On Tue, Jul 5, 2011 at 3:59 PM, William J. Schmidt wrote: > (Sorry for the late response; yesterday was a holiday here.) > > On Mon, 2011-07-04 at 16:21 +0200, Richard Guenther wrote: >> On Thu, Jun 30, 2011 at 4:39 PM, William J. Schmidt >> wrote: >> > This is the first of three patches related

Re: PATCH TRUNK: [gcc/configure.ac] Generate GCCPLUGIN_VERSION_* macros

On Wed, Jul 6, 2011 at 2:50 PM, Basile Starynkevitch wrote: > Hello All, > > The fie plugin-version.h is now generated by gcc/configure.ac. It contains > version information (about the GCC supposed to load the plugin compiled with > it) as constant strings. > > But I think it will also help some p

[PATCH] Fix PR49645, with C FE pieces

This fixes PR49645 - with MEM_REF the value-numbering machinery to look through aggregate copies wasn't working reliably as we have two representations for X, X and MEM[&X]. The following patch fixes that by internally always using the more complicated representation. The patch needs consistent

Re: [Patch,testsuite]: target-supports.exp: Disable -fprofile-generate for AVR

Georg-Johann Lay writes: > Index: lib/target-supports.exp > === > --- lib/target-supports.exp (revision 175811) > +++ lib/target-supports.exp (working copy) > @@ -497,6 +497,13 @@ proc check_profiling_available { test_wh > >

Re: [PATCH] Fix PR49645, with C FE pieces

Hi, On Wed, 6 Jul 2011, Richard Guenther wrote: > *** copy_reference_ops_from_ref (tree ref, V > *** 579,585 > > memset (&temp, 0, sizeof (temp)); > /* We do not care for spurious type qualifications. */ > ! temp.type = TYPE_MAIN_VARIANT (TREE_TYPE (ref)

Re: PATCH TRUNK: [gcc/configure.ac] Generate GCCPLUGIN_VERSION_* macros

On Wed, Jul 06, 2011 at 03:21:47PM +0200, Richard Guenther wrote: > On Wed, Jul 6, 2011 at 2:50 PM, Basile Starynkevitch > wrote: > > > > I belive it can help to make plugin code more robust. A serious plugin > > developper could then add in his plugin code something like > > > > #if GCCPLUGIN_VE

Re: [Path,AVR]: Improve loading of 32-bit constants

2011/7/6 Georg-Johann Lay : > For loading a 32-bit constant in a register, there is room for > improvement: > > * SF can be handled the same way as SI and therefore the patch >  adds a peep2 to produce a *reload_insf analogon to *reload_insi. > > * If the destination register overlaps NO_LD_REGS, v

Re: [PATCH] Fix PR49645, with C FE pieces

On Wed, 6 Jul 2011, Richard Guenther wrote: > * c-decl.c (finish_decl): Also set DECL_HARD_REGISTER for global > register variables. OK. -- Joseph S. Myers jos...@codesourcery.com

--enable-gnu-indirect-function patch

This patch: http://gcc.gnu.org/ml/gcc-patches/2010-09/msg02070.html for x86_64-*-linux* sets the default for --enable-gnu-indirect-function to "glibc-2011". This string is not used anywhere else in gcc as far as I can see. What is the purpose of that? The original version of the patch http://

patch committed: Correct configure option name in docs

This patch corrects the name of the configure option --enable-gnu-indirect-function in the docs to correspond to the source. Committed as obvious. Ian 2011-07-06 Ian Lance Taylor * doc/install.texi (Configuration): It's --enable-gnu-indirect-function, not --enable-indirect-fu

Re: PATCH TRUNK: [gcc/configure.ac] Generate GCCPLUGIN_VERSION_* macros

On Wed, Jul 6, 2011 at 3:46 PM, Basile Starynkevitch wrote: > On Wed, Jul 06, 2011 at 03:21:47PM +0200, Richard Guenther wrote: >> On Wed, Jul 6, 2011 at 2:50 PM, Basile Starynkevitch >> wrote: >> > >> > I belive it can help to make plugin code more robust. A serious plugin >> > developper could

Re: --enable-gnu-indirect-function patch

On 07/06/11 14:57, Ian Lance Taylor wrote: This patch: http://gcc.gnu.org/ml/gcc-patches/2010-09/msg02070.html for x86_64-*-linux* sets the default for --enable-gnu-indirect-function to "glibc-2011". This string is not used anywhere else in gcc as far as I can see. What is the purpose of that

Define WORDS_BIG_ENDIAN in rs6000/vxworks.h

At I enumerated the cases in GCC where WORDS_BIG_ENDIAN and BYTES_BIG_ENDIAN may differ. The ARM -mwords-little-endian case has now had a deprecation patch submitted and approved (wi

Re: [PATCH] Address lowering [1/3] Main patch

On Wed, 2011-07-06 at 15:16 +0200, Richard Guenther wrote: > On Tue, Jul 5, 2011 at 3:59 PM, William J. Schmidt > wrote: > > (Sorry for the late response; yesterday was a holiday here.) > > > > On Mon, 2011-07-04 at 16:21 +0200, Richard Guenther wrote: > >> On Thu, Jun 30, 2011 at 4:39 PM, William

Re: PATCH TRUNK: [gcc/configure.ac] Generate GCCPLUGIN_VERSION_* macros

On Wed, Jul 06, 2011 at 04:02:48PM +0200, Richard Guenther wrote: > On Wed, Jul 6, 2011 at 3:46 PM, Basile Starynkevitch > wrote: > > On Wed, Jul 06, 2011 at 03:21:47PM +0200, Richard Guenther wrote: > >> On Wed, Jul 6, 2011 at 2:50 PM, Basile Starynkevitch > >> wrote: > >> > > >> > I beleive it

Re: PATCH [1/n] X32: Add initial -x32 support

Hi Paolo, DJ, Nathanael, Alexandre, Ralf, Is the change . * configure.ac: Support --enable-x32. * configure: Regenerated. diff --git a/gcc/configure.ac b/gcc/configure.ac index 5f3641b..bddabeb 100644 --- a/gcc/configure.ac +++ b/gcc/configure.ac @@ -611,6 +611,11 @@ AC_ARG_ENABLE

Re: [Patch,testsuite]: target-supports.exp: Disable -fprofile-generate for AVR

Rainer Orth wrote: > Georg-Johann Lay writes: > >> Index: lib/target-supports.exp >> === >> --- lib/target-supports.exp (revision 175811) >> +++ lib/target-supports.exp (working copy) >> @@ -497,6 +497,13 @@ proc check_profiling_ava

Re: [PATCH] Address lowering [1/3] Main patch

On Wed, Jul 6, 2011 at 4:28 PM, William J. Schmidt wrote: > On Wed, 2011-07-06 at 15:16 +0200, Richard Guenther wrote: >> On Tue, Jul 5, 2011 at 3:59 PM, William J. Schmidt >> wrote: >> > (Sorry for the late response; yesterday was a holiday here.) >> > >> > On Mon, 2011-07-04 at 16:21 +0200, Ric

Re: PATCH [1/n] X32: Add initial -x32 support

On Wed, Jul 6, 2011 at 4:48 PM, H.J. Lu wrote: > Hi Paolo, DJ, Nathanael, Alexandre, Ralf, > > Is the change > . >        * configure.ac: Support --enable-x32. >        * configure: Regenerated. > > diff --git a/gcc/configure.ac b/gcc/configure.ac > index 5f3641b..bddabeb 100644 > --- a/gcc/config

Re: PATCH TRUNK: [gcc/configure.ac] Generate GCCPLUGIN_VERSION_* macros

On Wed, Jul 6, 2011 at 10:37, Basile Starynkevitch wrote: > On Wed, Jul 06, 2011 at 04:02:48PM +0200, Richard Guenther wrote: >> On Wed, Jul 6, 2011 at 3:46 PM, Basile Starynkevitch >> wrote: >> > On Wed, Jul 06, 2011 at 03:21:47PM +0200, Richard Guenther wrote: >> >> On Wed, Jul 6, 2011 at 2:50

Re: [Patch,testsuite]: target-supports.exp: Disable -fprofile-generate for AVR

Georg-Johann Lay writes: > Here is a revised patch. > > Ok? I'd like to defer to the target maintainers here: they know their port, while I don't. Rainer -- - Rainer Orth, Center for Biotechnology, Bielefeld U

Re: Provide 64-bit default Solaris/x86 configuration (PR target/39150)

> * In the sparcv9-sun-solaris2.11 builds, the 32-bit libgo tests fail to > link since they have unresolved references to > __sync_bool_compare_and_swap_8 and __sync_add_and_fetch_8. I could > trace this to -mv8plus being missing in that configuration. I'm > uncertain where best to handle

Re: Provide 64-bit default Solaris/x86 configuration (PR target/39150)

Eric Botcazou writes: >> * In the sparcv9-sun-solaris2.11 builds, the 32-bit libgo tests fail to >> link since they have unresolved references to >> __sync_bool_compare_and_swap_8 and __sync_add_and_fetch_8. I could >> trace this to -mv8plus being missing in that configuration. I'm >> u

Re: [Patch,testsuite]: target-supports.exp: Disable -fprofile-generate for AVR

2011/7/6 Rainer Orth : > Georg-Johann Lay writes: > >> Here is a revised patch. >> >> Ok? > > I'd like to defer to the target maintainers here: they know their port, > while I don't. > Approved. Denis.

Re: [Path,AVR]: Improve loading of 32-bit constants

Denis Chertykov wrote: > 2011/7/6 Georg-Johann Lay : >> For loading a 32-bit constant in a register, there is room for >> improvement: >> >> * SF can be handled the same way as SI and therefore the patch >> adds a peep2 to produce a *reload_insf analogon to *reload_insi. >> >> * If the destination

[PATCH, testsuite] Fix for PR49519, miscompiled 447.dealII in SPEC CPU 2006

Hi, I've prepared a patch for: http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49519 I've also prepared a test which reproduces the error. ChangeLog entry: 2011-07-06 Kirill Yukhin PR tailcall-optimization/49519 * calls.c (mem_overlaps_already_clobbered_arg_p): Additional

Re: Improve Solaris mudflap support (PR libmudflap/49550)

Hi, Rainer - > > It addresses a couple of testsuite failures: > > [...] > > where the registration of __iob has been done automatically by the > > compiler. I avoid this problem by not registering stdin, stdout, and > > stderr separately on Solaris. OK. > > * Some tests were failing whi

Re: Improve Solaris mudflap support (PR libmudflap/49550)

Hi Frank, >> > * Some tests were failing while calling unregister in munmap. It turned >> > out that there had been no corresponding mmap registration before. >> > This occurs because Solaris has mmap64 for largefile-aware programs >> > instead. Fixed by wrapping mmap64, too. What I don't

Re: Provide 64-bit default Solaris/x86 configuration (PR target/39150)

> Thanks, I'll give it a try. I just seemd weird to have MASK_V8PLUS and > MASK_V9 at the same time. Yes, that's why the existing comment should also be enhanced. I'll fix it. -- Eric Botcazou

Re: PATCH [1/n] X32: Add initial -x32 support

On Wed, Jul 6, 2011 at 8:02 AM, Richard Guenther wrote: > On Wed, Jul 6, 2011 at 4:48 PM, H.J. Lu wrote: >> Hi Paolo, DJ, Nathanael, Alexandre, Ralf, >> >> Is the change >> . >>        * configure.ac: Support --enable-x32. >>        * configure: Regenerated. >> >> diff --git a/gcc/configure.ac b/

[v3] Correctly determine baseline_subdir for 64-bit default Solaris gcc

As alluded to in Provide 64-bit default Solaris/x86 configuration (PR target/39150) http://gcc.gnu.org/ml/gcc-patches/2011-07/msg00327.html (which was meant to be Cc'ed to libstdc++, but bounced due to a stupid typo), there are now to variant bi-arch gcc configurations for Solaris

Re: C++ mangling, function name to mangled name (or tree)

On 06/07/2011 18:25, Kevin André wrote: On Wed, Jul 6, 2011 at 18:00, Pierre Vittet wrote: I would like user of the plugin to give in arguments the name of the functions on which he would like a test to be run. That means that I must convert the string containing a function name (like "myc

Re: Improve Solaris mudflap support (PR libmudflap/49550)

Hi, Rainer - > >> > If so, I'd rather do it by adding some MFWRAP_OS_SPEC to avoid > >> > having to duplicate the whole spec in the Solaris config > >> > headers. > > > > Why would solaris have to duplicate MFWRAP_SPEC if mmap64 is added to > > the default gcc.c one? > > I assumed that you

Re: C++ mangling, function name to mangled name (or tree)

Le 6 juil. 2011 à 18:40, Pierre Vittet a écrit : > On 06/07/2011 18:25, Kevin André wrote: >> On Wed, Jul 6, 2011 at 18:00, Pierre Vittet wrote: >> >> >>> I would like user of the plugin to give in arguments the name of the >>> functions on which he would like a test to be run. That means tha

Re: Improve Solaris mudflap support (PR libmudflap/49550)

Hi Frank, >> > Why would solaris have to duplicate MFWRAP_SPEC if mmap64 is added to >> > the default gcc.c one? >> >> I assumed that you wanted to keep the default generic, and meant to >> separate target specific additions from the generic part. > > I don't have a strong opinion on this, but if

Re: [RFC PATCH, go]: Port to ALPHA arch - sysinfo.go fixup

On Tue, Jul 5, 2011 at 7:17 PM, Mike Stump wrote: >> Please note that we set >> -mieee flag to compile .go files from library and also we add this >> flag to default testsuite compile flags. > >> > Ick, I think this patch might be expedient, but, wrong.  Ian will have to > think about it and dec

Re: [pph] Stream and restore static_aggregates (issue4626096)

On Tue, Jul 5, 2011 at 9:08 PM, Diego Novillo wrote: >        * pph-streamer-in.c (pph_add_bindings_to_namespace): > diff --git a/gcc/cp/pph-streamer-in.c b/gcc/cp/pph-streamer-in.c > index 72536a5..0bab93b 100644 > --- a/gcc/cp/pph-streamer-in.c > +++ b/gcc/cp/pph-streamer-in.c > @@ -1167,11 +11

DOC patch: about gengtype & plugins

Hello All, The attached documentation patch is nearly trivial, I was tempted to apply it without review. ### gcc/ChangeLog entry ### 2011-07-06 Basile Starynkevitch * doc/plugins.texi (Building GCC plugins): gengtype needs its gtype.state ### end gcc/ChangeLog entry ### O

Re: [Path,AVR]: Improve loading of 32-bit constants

2011/7/6 Georg-Johann Lay : > Denis Chertykov wrote: >> 2011/7/6 Georg-Johann Lay : >>> For loading a 32-bit constant in a register, there is room for >>> improvement: >>> >>> * SF can be handled the same way as SI and therefore the patch >>>  adds a peep2 to produce a *reload_insf analogon to *rel

[Patch,testsuite]: Filter more test cases to fit target capabilities

Hi, I am struggling against hundreds of fails in the testsuite because many cases are not carefully written, e.g. stull like shifting an int by 19 bits if int is only 16 bits wide. This patch adds some additional tests to avoid FAILs that are confusing. Sorry for gathering it in one patch, other

[PATCH, go]: Compile go checks with $(GOCFLAGS)

Hello! IMO, it makes sense to compile go tests with the same pack of flags as the library. Additionally, this solves an issue with extra compile flags (i.e. -mieee) that needs to be added to handle NaN/Inf. Patch was tested on x86_64-pc-linux-gnu {, -m32} and alphaev68-pc-linux-gnu (where the pat

Re: [testsuite] ARM wmul tests: require arm_dsp_multiply

On 06/29/2011 06:25 AM, Richard Earnshaw wrote: > On 23/06/11 22:38, Janis Johnson wrote: >> Tests wmul-[1234].c and mla-2.c in gcc.target/arm require support that >> the arm backend identifies as TARGET_DSP_MULTIPLY. The tests all >> specify a -march option with that support, but it is overridden

Re: [RFC PATCH, go]: Port to ALPHA arch - sysinfo.go fixup

Uros Bizjak writes: > On Tue, Jul 5, 2011 at 7:17 PM, Mike Stump wrote: > >>> Please note that we set >>> -mieee flag to compile .go files from library and also we add this >>> flag to default testsuite compile flags. >> >>> >> Ick, I think this patch might be expedient, but, wrong.  Ian will ha

Re: [RFC PATCH, go]: Port to ALPHA arch - sysinfo.go fixup

On Wed, Jul 6, 2011 at 7:34 PM, Ian Lance Taylor wrote: > This seems like a reasonable patch to me, but technically speaking it is > incomplete.  Go should have IEEE floating point behaviour by default.  I > believe Java is the same.  Ideally there would be a target-independent > way for a fronte

[testsuite] fixes for gcc.target/arm/mla-1.c

Test gcc.target/arm/mla-1.c scans the assembly file for a string that is part of the name, which always succeeds. This patch adds a tab to the search target to avoid that. It also removes the -march option and pruning of warnings about conflicts, and restricts the tests to targets that support Th

Re: [Path,AVR]: Improve loading of 32-bit constants

Denis Chertykov wrote: > 2011/7/6 Georg-Johann Lay : >> Denis Chertykov wrote: >>> 2011/7/6 Georg-Johann Lay : For loading a 32-bit constant in a register, there is room for improvement: * SF can be handled the same way as SI and therefore the patch adds a peep2 to produce

Re: [1/11] Use targetm.shift_truncation_mask more consistently

Bernd Schmidt writes: > At some point we've grown a shift_truncation_mask hook, but we're not > using it everywhere we're masking shift counts. This patch changes the > instances I found. The documentation reads: Note that, unlike @code{SHIFT_COUNT_TRUNCATED}, this function does @emph{not} app

Re: [7/11] rtl optimizer changes

On 07/01/2011 10:35 AM, Bernd Schmidt wrote: > * explow.c (trunc_int_for_mode): Use GET_MODE_PRECISION > instead of GET_MODE_BITSIZE where appropriate. > * rtlanal.c (subreg_lsb_1, subreg_get_info, nonzero_bits1, > num_sign_bit_copies1, canonicalize_condition, low_bitmask_le

Re: [8/11] Expander changes

On 07/01/2011 10:36 AM, Bernd Schmidt wrote: > * optabs.c (expand_binop): Use GET_MODE_PRECISION instead of > GET_MODE_BITSIZE where appropriate. > (widen_leading, expand_parity, expand_ctz, expand_ffs, > expand_unop, expand_abs_nojump, expand_one_cmpl_abs_nojump, > ex

Re: [PATCH] Fix PR49645, with C FE pieces

On Wed, Jul 6, 2011 at 6:26 AM, Richard Guenther wrote: > > This fixes PR49645 - with MEM_REF the value-numbering machinery > to look through aggregate copies wasn't working reliably as > we have two representations for X, X and MEM[&X].  The following > patch fixes that by internally always using

Re: [9/11] Fix units mismatch in comparison

On 07/01/2011 10:38 AM, Bernd Schmidt wrote: > * rtlanal.c (nonzero_bits1): Don't compare GET_MODE_SIZE against > a bitsize. Ok. r~

Re: [pph] Stream and restore static_aggregates (issue4626096)

On 11-07-06 13:12 , Gabriel Charette wrote: I had removed these two lines because that pretty much does nothing (I think, at least removing them didn't make anything fail...)... When we It started causing an ICE in the test case that I marked fixed and others (because we now stream static_agg

Re: [10/11] Expander fixes for 40-bit integers

On 07/01/2011 10:41 AM, Bernd Schmidt wrote: > * optabs.c (expand_binop): Tighten conditions for doubleword > expansions. > (widen_bswap): Assert that mode bitsize and precision are the > same. > * stor-layout.c (get_best_mode): Skip modes that have lower > preci

Re: [11/11] Fix get_mode_bounds

On 07/01/2011 10:42 AM, Bernd Schmidt wrote: > get_mode_bounds should also use GET_MODE_PRECISION, but this exposes a > problem on ia64 - BImode needs to be handled specially here to work > around another preexisting special case in gen_int_mode. Would it be better to remove the trunc_int_for_mode

libgo patch committed: Fix json test when rand returns 0

This patch is necessary when compiling the test cases with optimization, which changes the order of the calls to rand. I proposed the same patch to the upstream library. Bootstrapped and tested on x86_64-unknown-linux-gnu. Committed to mainline. Ian diff -r 48658f7ed377 libgo/go/json/scanner_t

Re: [PATCH] Use ira_reg_class_max_nregs array instead of CLASS_MAX_NREGS macro

Hi. The patch http://gcc.gnu.org/ml/gcc-patches/2011-06/msg02405.html is still pending review. Anatoly.

Re: [Path,AVR]: Improve loading of 32-bit constants

Denis Chertykov wrote: > I have asked about example of *d instead of !d. > Just svn GCC with *d vs svn GCC !d. > > Denis. Is the patch ok with the original !d instead of *d ? It still improves and the !d vs. *d don't matter because there's always r I think. Johann

[patch] PR other/49658 fix typo in extend.texi

Index: extend.texi === --- extend.texi (revision 175887) +++ extend.texi (working copy) @@ -1629,7 +1629,7 @@ char **foo = (char *[]) @{ "x", "y", "z" @}; @end smallexample -Compound literals for scalar types and union types are is

Re: [PATCH, testsuite] Fix for PR49519, miscompiled 447.dealII in SPEC CPU 2006

> 2011-07-06 Kirill Yukhin > > PR tailcall-optimization/49519 Please do not invent components, this will disable the automatic xref of the commit in bugzilla. Copy the "Component" field of the PR, middle-end here. > * calls.c (mem_overlaps_already_clobbered_arg_p): Additional

[PATCH] Fix dead_debug_insert_before ICE (PR debug/49522, take 2)

On Tue, Jul 05, 2011 at 10:06:51PM +0200, Jakub Jelinek wrote: > On Tue, Jul 05, 2011 at 10:35:11AM +0200, Eric Botcazou wrote: > > > There are two kinds of changes we do on the debug insns without immediate > > > rescanning: > > > 1) reset the debug insn > > > 2) replace a reg use with DEBUG_EXPR

[wwwdocs] Buildstat update for 4.5

Latest results for 4.5.x -tgc Testresults for 4.5.3: powerpc-apple-darwin8.11.0 sparc-sun-solaris2.7 Testresults for 4.5.2 powerpc-apple-darwin8.11.0 Testresults for 4.5.1 powerpc-apple-darwin8.11.0 Index: buildstat.html ==

[wwwdocs] Buildstat update for 4.4

Latest results for 4.4.x. -tgc Testresults for 4.4.6: powerpc-apple-darwin8.11.0 Testresults for 4.4.4: powerpc-apple-darwin8.11.0 Index: buildstat.html === RCS file: /cvs/gcc/wwwdocs/htdocs/gcc-4.4/buildstat.html,v retrieving

Re: [PATCH, go]: Compile go checks with $(GOCFLAGS)

Uros Bizjak writes: > IMO, it makes sense to compile go tests with the same pack of flags as > the library. Additionally, this solves an issue with extra compile > flags (i.e. -mieee) that needs to be added to handle NaN/Inf. > > Patch was tested on x86_64-pc-linux-gnu {, -m32} and > alphaev68-pc

Remove unused t-* fragments

This patch removes three unused t-* makefile fragments. (t-pa is unused because no target uses it explicitly and all PA targets define nonempty tmake_file; t-$cpu_type is is only used implicitly if tmake_file is empty after config.gcc.) Bootstrapped with no regressions on x86_64-unknown-linux-gnu

[lra] initial support of debug info and some fixes

The following patch contains some of my work since the last lra patch: o initial support of debug info in LRA. o code size improvement for i386 (changes in register banks). o code improvement by better reallocation of non-reload pseudos in LRA. by using allocno class which is usually wider than

Re: Remove unused t-* fragments

On Wed, Jul 6, 2011 at 10:14 PM, Joseph S. Myers wrote: > This patch removes three unused t-* makefile fragments.  (t-pa is > unused because no target uses it explicitly and all PA targets define > nonempty tmake_file; t-$cpu_type is is only used implicitly if > tmake_file is empty after config.gc

Re: [PATCH] Fix dead_debug_insert_before ICE (PR debug/49522)

> So you prefer something like this (untested) instead? I think that, ideally, we should avoid leaving the dead_debug chain in the semi-broken state that we currently have. > Without the second loop I have no idea how to make it work in > dead_debug_reset, the other dead_debug_use referencing th

Re: [PATCH] Fix dead_debug_insert_before ICE (PR debug/49522, take 2)

> And here is a version that passed bootstrap/regtest on x86_64-linux and > i686-linux: > > 2011-07-06 Jakub Jelinek > > PR debug/49522 > * df-problems.c (dead_debug_reset): Remove dead_debug_uses > referencing debug insns that have been reset. > (dead_debug_insert_before

[pph] Add FIXME comment to avoid finalizing decls when generating pph image. (issue4626099)

We do not need to finalize decls and add them to the varpool when generating the pph image as we will do this when streaming in (lto also already does it this way). I simply added a comment for now, because this will not fix anything, it will simply avoid streaming out unecessary stuff. Since t

Re: [Patch, Fortran] Register allocatable coarrays.

Daniel Carrera wrote: On 07/05/2011 09:57 AM, Tobias Burnus wrote: On 07/04/2011 11:34 PM, Daniel Carrera wrote: The test compiles, but there are "expected failures" because gcc doesn't think that allocatable scalar coarrays are supported. [...] I really don't want to add knowingly a failing

Re: [Patch, Fortran] Register allocatable coarrays.

On Wed, Jul 06, 2011 at 10:57:35PM +0200, Tobias Burnus wrote: > > I have now committed it as Rev. 175937 > Does Daniel have write-after-approval svn access? If not, we should probably get him access. -- Steve

PING: PATCH [8/n]: Prepare x32: PR other/48007: Unwind library doesn't work with UNITS_PER_WORD > sizeof (void *)

PING. On Thu, Jun 30, 2011 at 1:47 PM, H.J. Lu wrote: > On Thu, Jun 30, 2011 at 12:02 PM, Richard Henderson wrote: >> On 06/30/2011 11:23 AM, H.J. Lu wrote: >>> +#ifdef REG_VALUE_IN_UNWIND_CONTEXT >>> +typedef _Unwind_Word _Unwind_Context_Reg_Val; >>> +/* Signal frame context.  */ >>> +#define S

C++ PATCH for c++/49568 (unneeded thunks emitted on xcoff target)

This issue was exposed by honza's change to how thunks are represented, but the underlying problem was that we weren't setting DECL_COMDAT on them properly. Tested that the bug is fixed with an alpha-dec-osf5.1b cross compiler, regression tested x86_64-pc-linux-gnu, applying to trunk. commit d

Re: [pph] Test cleanup (issue4572050)

After having a look at how pph.exp works last Friday I think I could do this myself easily enough. Or are you still modifying the tests and want me to avoid touching this for now? Gab On Fri, Jul 1, 2011 at 5:51 PM, Lawrence Crowl wrote: > On 7/1/11, Gabriel Charette wrote: >> One problem now

[PATCH] [Annotalysis] Change to get_virtual_function_decl

This patch modifies the behavior of cp_get_virtual_function_decl in gcc/cp/class.c so that it returns NULL if the function declaration cannot be found. The previous behavior was to fail with a segmentation fault. The method-not-found case may occur when Annotalysis uses the function to look up a

[PATCH] Add -mno-r11 option to suppress load of ppc64 static chain in indirect calls

This patch adds an option to not load the static chain (r11) for 64-bit PowerPC calls through function pointers (or virtual function). Most of the languages on the PowerPC do not need the static chain being loaded when called, and adding this instruction can slow down code that calls very short fu

Re: [PATCH] Add -mno-r11 option to suppress load of ppc64 static chain in indirect calls

On Wed, Jul 6, 2011 at 6:29 PM, Michael Meissner wrote: > This patch adds an option to not load the static chain (r11) for 64-bit > PowerPC > calls through function pointers (or virtual function).  Most of the languages > on the PowerPC do not need the static chain being loaded when called, and >

Re: [11/11] Fix get_mode_bounds

On 07/06/11 20:37, Richard Henderson wrote: > On 07/01/2011 10:42 AM, Bernd Schmidt wrote: >> get_mode_bounds should also use GET_MODE_PRECISION, but this exposes a >> problem on ia64 - BImode needs to be handled specially here to work >> around another preexisting special case in gen_int_mode. >

Re: [CFT][PATCH 0/6] Move dwarf2 cfi creation to a new pass

On 07/03/11 22:01, Richard Henderson wrote: > Bernd's original patch to optimize dwarf2 cfi for shrink-wrapping > is difficult to analyze because that optimization was done via a > random debugging hook during final, and the cfi notes are deleted > at the end of final so that we don't get debug com

Re: [PATCH] [Annotalysis] Change to get_virtual_function_decl

On Wed, Jul 6, 2011 at 17:56, Delesley Hutchins wrote: > This patch modifies the behavior of cp_get_virtual_function_decl in > gcc/cp/class.c so that it returns NULL if the function declaration > cannot be found.  The previous behavior was to fail with a > segmentation fault.  The method-not-found

[testsuite] arm tests: remove -march= and dg-prune-output from 3 tests

For three tests in gcc.target/arm that don't depend on processor-specific behavior, don't specify the -march option. This makes dg-prune-output for warnings about conflicts unnecessary, so remove it. Two of these tests are for internal compiler errors that showed up with particular values of -mar

[testsuite] arm thumb tests: remove -march= and dg-prune-output from 9 tests

This patch removes -march= from nine tests that also check for relevant effective targets. If -march is removed there is no need to ignore compiler warnings about conflicting options with dg-prune-output, so the patch removes that from the tests. OK for trunk, and for 4.6 in a few days? 2011-07-0

  1   2   >