Re: [PATCH, rs6000] Use direct moves for __builtin_signbit for 128-bit floating-point

Hi Segher, > On Jun 29, 2016, at 4:43 PM, Segher Boessenkool > wrote: > > Hi, > > On Tue, Jun 28, 2016 at 04:44:08PM -0500, Bill Schmidt wrote: >> +void >> +rs6000_split_signbit (rtx dest, rtx src) >> +{ >> + machine_mode d_mode = GET_MODE (dest);

Re: [PATCH, rs6000] Use direct moves for __builtin_signbit for 128-bit floating-point

d the signbit-3.c testcase to make use of the recently added __float128 builtins. Bootstrapped and tested on powerpc64le-unknown-linux-gnu with no regressions. Is this ok for trunk, and eventual 6.2 backport? Thanks! Bill [gcc] 2016-07-01 Michael Meissner Bill Sc

[PATCH, rs6000] Fix PR71297 (ICE on invalid calls to vec_ld and vec_st)

kick in. The patch is pretty obvious and I think adding a test case would be extraneous, though I can do so if desired. Bootstrapped and tested on powerpc64le-unknown-linux-gnu with no regressions, and the original failure is fixed. Is this ok for trunk? Thanks, Bill 2016-07-07 Bill Schmidt

Re: [PATCH, rs6000] Fix PR71297 (ICE on invalid calls to vec_ld and vec_st)

Resending since I thinkoed Segher's email address. Sorry for the noise. Bill On Thu, 2016-07-07 at 14:11 -0500, Bill Schmidt wrote: > Hi, > > PR71297 reports that we ICE when __builtin_vec_ld or __builtin_vec_st is > provided with an incorrect number of arguments. This

Re: [PATCH, rs6000] Fix PR71297 (ICE on invalid calls to vec_ld and vec_st)

> On Jul 8, 2016, at 12:14 AM, Segher Boessenkool > wrote: > > On Thu, Jul 07, 2016 at 03:40:28PM -0500, Bill Schmidt wrote: >>> PR71297 reports that we ICE when __builtin_vec_ld or __builtin_vec_st is >>> provided with an incorrect number of arguments. This p

Re: [PATCH 0/9] separate shrink-wrapping

Not that getting the terminology right isn't important, but it would be nice if Segher could get a review for the rest of the content, too. :) Bill > On Jul 8, 2016, at 8:45 AM, Segher Boessenkool > wrote: > > On Fri, Jul 08, 2016 at 09:16:03AM -0400, David Malcolm wrote: >> As far as I unders

Re: [PATCH, rs6000] Fix PR71297 (ICE on invalid calls to vec_ld and vec_st)

Fixed in trunk with r238168, test case included. Thanks! Bill > On Jul 8, 2016, at 7:29 AM, Bill Schmidt wrote: > >> >> On Jul 8, 2016, at 12:14 AM, Segher Boessenkool >> wrote: >> >> On Thu, Jul 07, 2016 at 03:40:28PM -0500, Bill Schmidt wrote: &

[PATCH, rs6000, libgcc] Implement temporary solution for __divkc3 and __mulkc3

apped and tested on powerpc64le-unknown-linux-gnu with no regressions. I've also asked the glibc team to verify that this serves their requirements. Is this ok for trunk, and for gcc-6-branch after a short burn-in period? Thanks, Bill [libgcc] 2016-07-11 Bill Schmidt * con

[PATCH, rs6000] Fix vec_construct vectorization cost to be somewhat more accurate

st suite have not changed). Is this ok for trunk? Thanks, Bill 2016-07-15 Bill Schmidt * config/rs6000/rs6000.c (rs6000_builtin_vectorization_cost): Improve vec_construct estimate. Index: gcc/config/rs6000/rs6000.c =

Re: [PATCH] Amend dump expectation in slsr-8.c (PR, tree-optimization/71490)

> On Jul 15, 2016, at 2:24 AM, Richard Biener > wrote: > > On Thu, Jul 14, 2016 at 6:10 PM, Martin Liška wrote: >> On 07/14/2016 01:21 PM, Richard Biener wrote: >>> On Thu, Jul 14, 2016 at 1:06 PM, Martin Liška wrote: On 07/13/2016 07:21 PM, Jeff Law wrote: > Isn't that a code qualit

[PATCH, rs6000] Fix {div,mul}kc3-1.c for older hardware

trunk and gcc-6-branch? Thanks, Bill 2016-07-15 Bill Schmidt * gcc.target/powerpc/divkc3-1.c: Require p8vector support. * gcc.target/powerpc/mulkc3-1.c: Likewise. Index: gcc/testsuite/gcc.target/powerpc/divkc3-1.c

[PATCH, rs6000, preapproved] Remove XFAILs from g++.dg/pr70098.C and gcc.target/powerpc/pr71763.c

Hi, A recent patch from Alan Modra removed the need to XFAIL these two test cases. Tested on powerpc64[le]-unknown-linux-gnu, committed to trunk and gcc-6-branch. Thanks, Bill 2016-07-15 Bill Schmidt * g++.dg/pr70098.C: Remove XFAIL for powerpc64_no_dm. * gcc.target

[PATCH] Remove dead call to gimple_bb in slsr_process_phi

e-unknown-linux-gnu with no regressions. Is this ok for trunk? (The error is harmless, so I see no reason for a backport.) Thanks! Bill 2016-07-25 Bill Schmidt * gimple-ssa-strength-reduction.c (slsr_process_phi): Remove dead and unnecessary call to gimple_bb. Index: gcc/gimp

[PATCH, rs6000] Correct effective target in gcc.target/powerpc/pr63354.c

. Tested on powerpc64-unknown-linux-gnu, verified. Ok for trunk? Thanks, Bill 2016-07-25 Bill Schmidt * gcc.target/powerpc/pr63354.c: Require lp64 since -mprofile-kernel is not legal with -m32. Index: gcc/testsuite/gcc.target/powerpc/pr63354.c

[PATCH, rs6000] Add missing boilerplate

Hi, I forgot to add the usual licensing and copyright boilerplate to a couple of new files. Fixed here with attribution of the code origins. Bootstrapped, committed as obvious. Thanks, Bill 2016-07-29 Bill Schmidt * config/rs6000/_divkc3.c: Add copyright/license boilerplate

[PATCH, rs6000] Fix some test failures on older hardware

. Ok for trunk and for backport to gcc-6-branch? Thanks, Bill 2016-07-29 Bill Schmidt * gcc.target/powerpc/divkc3-1.c: Require POWER8 hardware. * gcc.target/powerpc/mulkc3-1.c: Likewise. Index: gcc/testsuite/gcc.target/powerpc/divkc3-1.c

Re: [RS6000] fix rtl checking internal compiler error

Good catch, Alan, this one is my fault. I'll handle the backports to the 5 and 6 branches. Bill > On Dec 1, 2016, at 12:34 AM, Alan Modra wrote: > > I'm committing this one as obvious once my powerpc64le-linux bootstrap > and regression check completes. It fixes hundreds of rtl checking > te

Re: [PATCH, rs6000] Fold vector addition built-ins in GIMPLE

What's your target triple? > On Dec 4, 2016, at 6:36 AM, Andreas Schwab wrote: > > On Nov 01 2016, Bill Schmidt wrote: > >> * gcc.target/powerpc/fold-vec-add-7.c: New. > > spawn -ignore SIGHUP /daten/gcc/gcc-20161203/Build/gcc/xgcc > -B/daten/gcc/gcc-2

[PATCH] Fix PR78646

. Bootstrapped and tested on powerpc64le-unknown-linux-gnu with no regressions. Tested with a powerpc64le->x86_64 cross to verify the correct type is now chosen. Committed. Thanks, Bill 2016-12-05 Bill Schmidt Stefan Freudenberger PR tree-optimization/78

[PATCH, rs6000] Fix 32-bit BE failure for gcc.target/powerpc/fold-vec-add-7.c

on LE. Tested on powerpc64le- unknown-linux-gnu and verified that the test runs correctly there. Is this ok for trunk? Thanks, Bill 2016-12-07 Bill Schmidt * gcc.target/powerpc/fold-vec-add-7.c: Require effective target to support __int128. Index: gcc/testsuite/gcc.target

[PATCH, rs6000] Use new pass registration framework for analyze_swaps

-p8-17.c.232r.dfinit -rw-rw-r-- 1 wschmidt wschmidt 5092 Dec 8 16:40 swaps-p8-17.c.233r.swaps -rw-rw-r-- 1 wschmidt wschmidt 9751 Dec 8 16:40 swaps-p8-17.c.234r.cse1 Is this ok for trunk? Thanks, Bill 2016-12-08 Bill Schmidt * config/rs6000/rs6000-passes.def: New file.

[PATCH, rs6000] Fix gcc.target/powerpc/pr78691-ppc.c

Hi, The fix for PR78691 added a new test case for 32-bit PowerPC. However, the test needs to be restricted to 32-bit targets to avoid a FAIL. This patch fixes that. Tested on powerpc64-unknown-linux-gnu. Ok for trunk? Thanks, Bill 2016-12-09 Bill Schmidt * gcc.target/powerpc

[PATCH, rs6000] Fix PR78695

patch fixes the problem in the obvious way. Bootstrapped and tested on powerpc64le-unknown-linux-gnu with no regressions. Is this ok for trunk? The problem was introduced in GCC 7, so no backports are needed. Thanks, Bill [gcc] 2016-12-11 Bill Schmidt PR target/78695

Re: [PATCH, rs6000] Fix PR78695

if it succeeds? Thanks, Bill [gcc] 2016-12-11 Bill Schmidt PR target/78695 * config/rs6000/rs6000.c (find_alignment_op): Discard from consideration any artificial definition. [gcc/testsuite] 2016-12-11 Bill Schmidt PR target/78695 * gcc.target/power

Re: [PATCH, rs6000] Fix PR78695

Hi Segher, This indeed bootstrapped on powerpc64le-unknown-linux-gnu with no regressions. Ok for trunk? Thanks for the review! Bill On 12/11/16 3:31 PM, Bill Schmidt wrote: > Hi Segher, > > On 12/11/16 2:00 PM, Segher Boessenkool wrote: >> Maybe this should use DF_REF_IS_ART

Re: [PATCH, rs6000] Fix PR78695

Ah, I misread your comment and went hunting for DF_REF_ARTIFICIAL. Sorry! Will fix. Thanks again, Bill > On Dec 11, 2016, at 5:29 PM, Segher Boessenkool > wrote: > > On Sun, Dec 11, 2016 at 03:31:35PM -0600, Bill Schmidt wrote: >> On 12/11/16 2:00 PM, Segher Boessenkoo

[PATCH, rs6000] Cost model adjustment

Hi, This patch makes a slight adjustment to the vectorization cost model that was previously overlooked. Bootstrapped and tested on powerpc64le-unknown-linux gnu with no regressions. Ok for trunk? Thanks, Bill 2016-12-13 Bill Schmidt * config/rs6000/rs600.c

Re: [PATCH, rs6000] Add support for vector element-reversal built-ins

pc64le-unknown-linux-gnu with no regressions. Is this revised version ok for trunk? Thanks! Bill [gcc] 2016-04-27 Bill Schmidt * config/rs6000/altivec.h: Change definitions of vec_xl and vec_xst. * config/rs6000/rs6000-builtin.def (LD_ELEMREV_V2DF): New. (L

[PATCH, rs6000] Backport some swap optimization improvements

. Is this ok for GCC 5.4? Thanks, Bill [gcc] 2016-04-28 Bill Schmidt PR target/69868 + swap optimization backports * config/rs6000/rs6000.c (swap_web_entry): Enlarge special_handling bitfield. (special_handling_values): Add SH_XXPERMDI, SH_CONCAT, SH_VPERM

Re: [PATCH,rs6000] Add built-in support for new Power9 darn (deliver a random number) instruction

On Mon, 2016-05-09 at 08:58 -0500, Segher Boessenkool wrote: > Hi Kelvin, > > On Thu, May 05, 2016 at 10:26:01AM -0600, Kelvin Nilsen wrote: > > (UNSPEC_DARN_32): New usnpec constant. > > Typo. > > > ("darn_32"): New instruction. > > We don't normally use quotes for insn names. > > >

[PATCH, rs6000] Fix PR70963: Wrong code for V2DF/V2DI vec_cts with zero scale factor

in the test case. Bootstrapped and tested on powerpc64le-unknown-linux-gnu with no regressions. Is this ok for trunk, and eventual backport to 6 and 5? Thanks, Bill [gcc] 2016-05-09 Bill Schmidt * config/rs6000/vsx.md (vsx_xvcvdpsxds_scale): Generate correct code for a zero

Re: [PATCH, rs6000] Fix PR70963: Wrong code for V2DF/V2DI vec_cts with zero scale factor

> On May 10, 2016, at 3:25 AM, Segher Boessenkool > wrote: > > Hi Bill, > > On Mon, May 09, 2016 at 10:00:31PM -0500, Bill Schmidt wrote: >> PR70963 reports a problem with vec_cts when used to convert vector double to >> vector long long. >> This is due t

[PATCH, rs6000] Fix pr70963.c test case for older hardware

patch remedies this by ensuring the test will not run unless p8vector support is available. Verified on powerpc64-unknown-linux-gnu and powerpc64le-unknown-linux-gnu. Is this ok for trunk? I will wait a bit longer before deploying to earlier releases… Thanks, Bill 2016-05-11 Bill Schmidt

Re: [PATCH][RFC] Introduce BIT_FIELD_INSERT

Hi Richard, (Sorry for duplication to your personal email, I had new-mailer issues.) The new vector-6 test produces very good code for powerpc64le with this patch: addis 9,2,.LC0@toc@ha sldi 3,3,32 addi 9,9,.LC0@toc@l rldicl 9,9,0,32 or 3,9,3 blr

Re: [PATCH][RFC] Introduce BIT_FIELD_INSERT

On May 15, 2016, at 7:55 PM, Bill Schmidt wrote: > > Hi Richard, > > (Sorry for duplication to your personal email, I had new-mailer issues.) > > The new vector-6 test produces very good code for powerpc64le with this patch: > >addis 9,2,.LC0@toc@ha >

[PATCH, rs6000] Add some missing vector built-ins

PERMQ. [gcc/testsuite] 2015-08-18 Bill Schmidt * gcc.target/powerpc/altivec-35.c: New test. * gcc.target/powerpc/p8vector-builtin-8.c: New test. * gcc.target/powerpc/vsx-vector-7.c: New test. Index: gcc/config/rs6000/alti

[PATCH, rs6000] A few more vector interfaces

, Bill [gcc] 2015-08-20 Bill Schmidt * config/rs6000/altivec.h (vec_pmsum_be): New #define. (vec_shasigma_be): New #define. * config/rs6000/rs6000-builtin.def (VPMSUMB): New BU_P8V_AV2_2. (VPMSUMH): Likewise. (VPMSUMW): Likewise. (VPMSUMD): Likewise

[PATCH, testsuite] Enable vect_double for PowerPC targets with VSX support

own-linux-gnu with no regressions. Is this ok for trunk? Thanks, Bill 2015-08-27 Bill Schmidt * lib/target-supports.exp (check-effective_target_vect_double): Enable for Power targets with VSX hardware available. * gfortran.dg/vect/O3-pr49957.f: Replace vect_

[PATCH, rs6000] Improve swap optimization to modify general xxpermdi patterns

multiplications, which I've added here. Bootstrapped and tested on powerpc64le-unknown-linux-gnu with no regressions. Is this ok for trunk? Thanks, Bill [gcc] 2015-08-30 Bill Schmidt * config/rs6000/rs6000.c (swap_web_entry): Enlarge special_handling bit

[PATCH] Fix ICE when generating a vector shift by scalar

at the shift amount is truncated to the inner mode of the vector when necessary. I've added a test case verifying correct PowerPC code generation in this case. Bootstrapped and tested on powerpc64le-unknown-linux-gnu with no regressions. Is this ok for trunk? Thanks, Bill [gcc] 2015-08

Re: [PATCH] Fix ICE when generating a vector shift by scalar

On Tue, 2015-09-01 at 11:01 +0200, Richard Biener wrote: > On Mon, Aug 31, 2015 at 10:28 PM, Bill Schmidt > wrote: > > Hi, > > > > The following simple test fails when attempting to convert a vector > > shift-by-scalar into a vector shift-by-vector. > &g

Re: [PATCH] Fix ICE when generating a vector shift by scalar

On Wed, 2015-09-02 at 14:44 +0200, Richard Biener wrote: > On Tue, Sep 1, 2015 at 5:53 PM, Bill Schmidt > wrote: > > On Tue, 2015-09-01 at 11:01 +0200, Richard Biener wrote: > >> On Mon, Aug 31, 2015 at 10:28 PM, Bill Schmidt > >> wrote: > >> > Hi, >

[PATCH, rs6000] Use hardware support for vector character multiply

m the check_effective_target_vect_char_mult target support test. I resolved this by adding check_effective_target_powerpc_altivec to that test. Bootstrapped and tested on powerpc64le-unknown-linux-gnu with no regressions. Is this ok for trunk? Thanks, Bill [gcc] 2015-09-03 Bill Sc

Re: [PATCH, rs6000] Use hardware support for vector character multiply

On Thu, 2015-09-03 at 23:26 +0800, Andrew Pinski wrote: > On Thu, Sep 3, 2015 at 11:20 PM, Bill Schmidt > wrote: > > Hi, > > > > It was pointed out to me recently that multiplying two vector chars is > > performed using scalarization, even though we have hardware s

Re: [PATCH, rs6000] Use hardware support for vector character multiply

On Thu, 2015-09-03 at 11:36 -0400, David Edelsohn wrote: > On Thu, Sep 3, 2015 at 11:20 AM, Bill Schmidt > wrote: > > Hi, > > > > It was pointed out to me recently that multiplying two vector chars is > > performed using scalarization, even though we have hardware s

[PATCH, rs6000] Extend LE swap optimization to handle vector permutes with constant masks

wo new tests, one to verify that swaps are removed, and one to verify that the vperm transformation produces correct results. Bootstrapped and tested on powerpc64le-unknown-linux-gnu with no regressions. Is this ok for trunk? Thanks, Bill [gcc] 2015-09-10 Bill Schmidt * co

Re: [PATCH, rs6000] Extend LE swap optimization to handle vector permutes with constant masks

On Thu, 2015-09-10 at 19:55 +0200, Bernhard Reutner-Fischer wrote: > On September 10, 2015 7:48:10 PM GMT+02:00, David Edelsohn > wrote: > >On Thu, Sep 10, 2015 at 1:41 PM, Bill Schmidt > > wrote: > >> Currently the little-endian swap optimization is disabled for >

Re: [PATCH] vectorizing conditional expressions (PR tree-optimization/65947)

Hi Alan, The cost modeling of the epilogue code seems pretty target-specific ("An EQ stmt and an AND stmt, reduction of the max index and a reduction of the found values, a broadcast of the max value," resulting in two vector_stmts, one vec_to_scalar, and two scalar_to_vecs). On powerpc, this wil

Re: [PATCH] vectorizing conditional expressions (PR tree-optimization/65947)

need some code in tree-vect-generic.c to implement on emulated max > reduction, which would then require updates to the costs modelling of > anything that uses max reduction (not just cond reduction). All of that is > outside the scope of this patch. > > > Thanks, > Alan. &g

Re: [PATCH] vectorizing conditional expressions (PR tree-optimization/65947)

On Fri, 2015-09-11 at 15:29 +0100, Richard Sandiford wrote: > Ramana Radhakrishnan writes: > > On Fri, Sep 11, 2015 at 2:19 PM, Bill Schmidt > > wrote: > >> Hi Alan, > >> > >> I probably wasn't clear enough. The implementation in the vectorizer is

Re: [PATCH] vectorizing conditional expressions (PR tree-optimization/65947)

On Fri, 2015-09-11 at 16:28 +0100, Richard Sandiford wrote: > Bill Schmidt writes: > > On Fri, 2015-09-11 at 15:29 +0100, Richard Sandiford wrote: > >> Ramana Radhakrishnan writes: > >> > On Fri, Sep 11, 2015 at 2:19 PM, Bill Schmidt > >> > wrote: &

Re: [PATCH] vectorizing conditional expressions (PR tree-optimization/65947)

On Mon, 2015-09-14 at 10:47 +0100, Alan Lawrence wrote: > On 11/09/15 14:19, Bill Schmidt wrote: > > > > A secondary concern for powerpc is that REDUC_MAX_EXPR produces a scalar > > that has to be broadcast back to a vector, and the best way to implement > > it for us a

Re: [PATCH, RFC] Fix PR71915, PR71490: Handle casts on strides consistently

On Nov 4, 2016, at 9:10 AM, Christophe Lyon wrote: > > > Hi, > > Since this was committed, I'm seeing a failure in > gcc.dg/tree-ssa/slsr-8.c scan-tree-dump-times optimized " \\* " 9 > on aarch64 targets. Hi Christophe, Best open a bug report or I will likely lose this in the crush. If you

Re: [PATCH, rs6000] Fold vector addition built-ins in GIMPLE

Just a note that the "-std=gnu11" option in the test cases in this patch is a leftover from a previous version of the patch, which I'll plan to remove. Sorry for the oversight! Bill > On Nov 1, 2016, at 9:05 PM, Bill Schmidt wrote: > > Hi, > > As Jakub sug

[PATCH, committed] Fix PR78210

Hi, On some targets (AArch64 in particular) test gcc.dg/tree-ssa/slsr-8.c fails because the scan of the SLSR dump is too strict. On these targets, a multiply may be a widening multiply, and that wasn't accounted for. Now it is. Committed as obvious. Thanks, Bill 2016-11-04 Bill Sc

[PATCH] Fix PR77848

tatements itself. Bootstrapped and tested on powerpc64le-unknown-linux-gnu, with only the specified regression. Is this ok for trunk? Thanks, Bill [gcc] 2016-11-15 Bill Schmidt PR tree-optimization/77848 * tree-if-conv.c (version_loop_for_if_conversion): When versioning

Re: [PATCH] Fix PR77848

Thanks, Richard! I'll follow up with these changes over the next day or two. Appreciate all the help! Bill On Wed, 2016-11-16 at 16:08 +0100, Richard Biener wrote: > On Tue, Nov 15, 2016 at 9:03 PM, Bill Schmidt > wrote: > > Hi, > > > > https://gcc.gnu.org/b

Re: [PATCH] Fix PR77848

On 11/16/16 9:08 AM, Richard Biener wrote: > On Tue, Nov 15, 2016 at 9:03 PM, Bill Schmidt > wrote: >> - if ((any_pred_load_store || any_complicated_phi) >> - && !version_loop_for_if_conversion (loop)) >> + /* Since we have no cost model, always version lo

Re: [PATCH] Fix PR77848

> On Nov 17, 2016, at 3:42 AM, Richard Biener > wrote: > > On Thu, Nov 17, 2016 at 4:14 AM, Bill Schmidt > wrote: >> >> Yes, the new predicate works fine. New patch below, bootstrapped and tested >> on powerpc64le-unknown-linux-gnu, with only the bb-slp-c

[PATCH] Fix PR78413

latches have a single predecessor before versioning an outer loop. Bootstrapped and tested on powerpc64le-unknown-linux-gnu with no regressions. Is this ok for trunk? Thanks, Bill [gcc] 2016-11-18 Bill Schmidt PR tree-optimization/78413 * tree-if-conv.c (versionable_outer_loop_p

Re: [PATCH] Fix PR78413

> On Nov 18, 2016, at 10:33 AM, Markus Trippelsdorf > wrote: > > On 2016.11.18 at 10:27 -0600, Bill Schmidt wrote: >> === >> --- gcc/testsuite/gcc.dg/tree-ssa/pr78413.c (revision 0) >> +++ gcc/tests

Re: Request permission to delete gcc.dg/vect/costmodel/ppc/costmodel-fast-math-vect-pr29925.c

On Fri, 2015-12-11 at 10:47 +0100, Richard Biener wrote: > On Thu, Dec 10, 2015 at 8:33 PM, David Edelsohn wrote: > > On Thu, Dec 10, 2015 at 2:23 PM, Bill Schmidt > > wrote: > >> Hi, > >> > >> The subject test case has been failing as follows: > >

Re: [PATCH 4/4] Un-XFAIL ssa-dom-cse-2.c for most platforms

On Mon, 2015-12-21 at 09:10 -0500, David Edelsohn wrote: > On Mon, Dec 21, 2015 at 8:13 AM, Alan Lawrence wrote: > > ...the test passes with --param sra-max-scalarization-size-Ospeed. > > > > Verified on aarch64 and with stage1 compiler for hppa, powerpc, sparc, s390. > > > > On alpha, tree-optimi

Re: [PATCH 4/4] Un-XFAIL ssa-dom-cse-2.c for most platforms

On Mon, 2015-12-21 at 15:22 +, Alan Lawrence wrote: > On 21/12/15 14:59, Bill Schmidt wrote: > >>> > >>> On powerpc64, the test passes with -mcpu=power8 (the loop is vectorized > >>> as a > >>> reduction); however, without that, similar code

Re: [PATCH 0/4 v3] Fix PR/63679 when --param sra-max-scalarization-size specified

On Tue, 2015-12-22 at 15:54 +, Alan Lawrence wrote: > On 21/12/15 13:13, Alan Lawrence wrote: > > This is a respin of previous patch series: > > https://gcc.gnu.org/ml/gcc-patches/2015-10/msg03271.html > > Minus three of the smaller patches having already been committed; with the > > updated ve

Re: [PATCH 4/4] Un-XFAIL ssa-dom-cse-2.c for most platforms

On Tue, 2015-12-22 at 16:00 +, Alan Lawrence wrote: > On 21/12/15 15:33, Bill Schmidt wrote: > >> > >> Not on a stage1 compiler - check_p8vector_hw_available itself requires > >> being > >> able to run executables - I'll check on gcc112. However, bo

Re: [PATCH 0/4 v3] Fix PR/63679 when --param sra-max-scalarization-size specified

On Tue, 2015-12-22 at 17:36 +, Alan Lawrence wrote: > On 22/12/15 16:05, Bill Schmidt wrote: > > On Tue, 2015-12-22 at 15:54 +, Alan Lawrence wrote: > >> On 21/12/15 13:13, Alan Lawrence wrote: > >>> This is a respin of previous patch series: > >>&g

[PATCH, rs6000] Add support for lxvx and stxvx P9 instructions

owerpc64le-unknown-linux-gnu with no regressions. Ok for trunk, and then for backport to GCC 5? Thanks, Bill [gcc] 2015-12-27 Bill Schmidt * config/rs6000/rs6000.c (rs6000_emit_le_vsx_move): Verify that this is never called when lxvx/stxvx are available. (pass_analyze_s

[PATCH, rs6000] Use lxvx and stxvx for 128-bit float, etc., with -mcpu=power9

these cases, and disables those other patterns when P9 vector support is available. Bootstrapped and tested on powerpc64le-unknown-linux-gnu with no regressions. Ok for trunk? Thanks, Bill [gcc] 2015-01-06 Bill Schmidt * config/rs6000/vsx.md (*p9_vecload_): Replace VSX_M mode

[PATCH, rs6000] Handle vector reductions in swap optimization

ter a settling period. Thanks, Bill [gcc] 2016-01-06 Bill Schmidt * config/rs6000/rs6000.c (v2df_reduction_p): New function. (rtx_is_swappable_p): Reductions are swappable. (insn_is_swappable_p): V2DF reductions are swappable. [gcc/testsuite] 2016-01-06 Bill Sc

[PATCH] Fix PR68799

64le-unknown-linux-gnu with no regressions for GCC 5.3. Currently doing the same for trunk. Assuming no regressions is this ok for mainline, and thereafter for GCC 5 and GCC 4.9? Thanks, Bill 2016-01-15 Bill Schmidt PR tree-optimization/68799 * gimple-ssa-strength-reduction.

[PING] SLSR for conditional candidates

Half-hearted ping for http://gcc.gnu.org/ml/gcc-patches/2013-03/msg01291.html ... I promise this is the last major code dump for SLSR. ;) Thanks, Bill

Re: [PING] SLSR for conditional candidates

On Thu, 2013-05-02 at 13:29 +0200, Richard Biener wrote: > On Mon, 29 Apr 2013, Bill Schmidt wrote: > > > Half-hearted ping for > > http://gcc.gnu.org/ml/gcc-patches/2013-03/msg01291.html ... > > > > I promise this is the last major code dump for SLSR. ;) > >

Re: [PATCH] SLSR for conditional candidates

On Fri, 2013-05-03 at 13:58 -0700, H.J. Lu wrote: > On Fri, Mar 29, 2013 at 2:15 PM, Bill Schmidt > wrote: > > This is a resubmission of a patch I sent to the list last August, > > re-based for the C++ changes since then. It implements the last big > > chunk of straight

[PATCH, committed] Avoid bootstrap failure for i686-pc-linux-gnu due to SLSR

, Bill gcc: 2013-05-03 Bill Schmidt * gimple-ssa-strength-reduction.c (slsr_process_phi): Disable. (find_candidates_in_block): Disable slsr_process_phi. gcc/testsuite: 2013-05-03 Bill Schmidt * gcc.dg/tree-ssa/slsr-32.c: Skip test for now. * gcc.dg/tree

[PATCH] Fix thinko in SLSR that caused x86 bootstrap failure

d code. Bootstrapped and tested on powerpc-unknown-linux-gnu and i686-pc-linux-gnu with no new regressions. Ok for trunk? Thanks, Bill gcc: 2013-05-05 Bill Schmidt * gimple-ssa-strength-reduction.c (slsr_process_phi): Re-enable. (find_candidates_in_block): R

[PATCH] Use anonymous SSA name support in SLSR

This removes lazy_create_slsr_reg and replaces uses of make_ssa_name with make_temp_ssa_name, removing the need for a bunch of unnecessary vars. Bootstrapped and tested on powerpc64-unknown-linux-gnu with no new regressions. Ok for trunk? Thanks, Bill 2013-05-06 Bill Schmidt

Re: [PATCH] Fix thinko in SLSR that caused x86 bootstrap failure

On Mon, 2013-05-06 at 21:25 +0200, Jakub Jelinek wrote: > On Sun, May 05, 2013 at 03:45:17PM -0500, Bill Schmidt wrote: > > 2013-05-05 Bill Schmidt > > > > * gimple-ssa-strength-reduction.c (slsr_process_phi): Re-enable. > > (find_candidates_in_block):

[PATCH] Fix SLSR conditional candidate bug with commuted operands

This fixes the following bug: On Mon, 2013-05-06 at 21:25 +0200, Jakub Jelinek wrote: > On Sun, May 05, 2013 at 03:45:17PM -0500, Bill Schmidt wrote: > > 2013-05-05 Bill Schmidt > > > > * gimple-ssa-strength-reduction.c (slsr_process_phi): Re-enable. > > (fi

Re: [PATCH] Fix thinko in SLSR that caused x86 bootstrap failure

On Mon, 2013-05-06 at 17:28 -0700, H.J. Lu wrote: > On Mon, May 6, 2013 at 12:29 PM, Bill Schmidt > wrote: > > On Mon, 2013-05-06 at 21:25 +0200, Jakub Jelinek wrote: > >> On Sun, May 05, 2013 at 03:45:17PM -0500, Bill Schmidt wrote: > >> > 2013-05-05 Bill Sch

Re: [PATCH] Fix thinko in SLSR that caused x86 bootstrap failure

On Tue, 2013-05-07 at 11:34 +0200, Jakub Jelinek wrote: > On Mon, May 06, 2013 at 08:18:27PM -0500, Bill Schmidt wrote: > > > AIL: gcc.dg/vect/vect-28.c -flto execution test > > > FAIL: gcc.dg/vect/vect-28.c execution test > > > FAIL: gfortran.dg/array_constructor_9

[PATCH] Limit SLSR increment vector size to avoid quadratic behavior

increments that do not appear in the increment vector are not processed. Bootstrapped and tested on powerpc64-unknown-linux-gnu with no new regressions. Ok for trunk? Thanks, Bill 2013-05-06 Bill Schmidt * gimple-ssa-strength-reduction.c (MAX_INCR_VEC_LEN): New constant

Re: [PATCH] Fix thinko in SLSR that caused x86 bootstrap failure

On Tue, 2013-05-07 at 11:34 +0200, Jakub Jelinek wrote: > On Mon, May 06, 2013 at 08:18:27PM -0500, Bill Schmidt wrote: > > > AIL: gcc.dg/vect/vect-28.c -flto execution test > > > FAIL: gcc.dg/vect/vect-28.c execution test > > > FAIL: gfortran.dg/array_constructor_9

[PATCH] Restrict conditional SLSR candidates to CAND_MULTs

ux-gnu, -m32/-m64, with no new regressions. Ok for trunk? Thanks, Bill 2013-05-07 Bill Schmidt * gimple-ssa-strength-reduction.c (find_phi_def): Revert former "fix." (alloc_cand_and_find_basis): Restrict conditional candidate processing to CAND_MULTs. In

[PATCH, committed] Correct signedness warning in gimple-ssa-strength-reduction.c

This corrects a signed/unsigned comparison. Regstrapped on powerpc64-unknown-linux-gnu, committed as obvious. Thanks, Bill 2013-05-08 Bill Schmidt * gimple-ssa-strength-reduction.c (count_candidates): Change return value to int. (analyze_candidates_and_replace

[PATCH, rs6000] Increase MALLOC_ABI_ALIGNMENT for 32-bit PowerPC

g the information. Bootstrapped and tested on powerpc64-unknown-linux-gnu with no new regressions. Verified that SPEC CPU2006 degradations are fixed with no new degradations. Ok for trunk? Also, do you want any backports? Thanks, Bill 2013-05-16 Bill Schmidt * config/rs6000/rs6

Re: [PATCH, rs6000] Increase MALLOC_ABI_ALIGNMENT for 32-bit PowerPC

On Fri, 2013-05-17 at 01:11 -0400, David Edelsohn wrote: > On Thu, May 16, 2013 at 10:40 PM, Bill Schmidt > wrote: > > This removes two degradations in CPU2006 for 32-bit PowerPC due to lost > > vectorization opportunities. Previously, GCC treated malloc'd arrays as

Re: [patch, powerpc] increase array alignment for Altivec

On Tue, 2013-05-21 at 21:57 -0400, David Edelsohn wrote: > On Tue, May 21, 2013 at 7:13 PM, Sandra Loosemore > wrote: > > On 05/21/2013 04:04 PM, David Edelsohn wrote: > >> > >> > >> There are three issues here: > >> > >> 1) Someone in the LTC toolchain team needs to benchmark this patch on > >> P

Re: [patch, powerpc] increase array alignment for Altivec

On Tue, 2013-05-21 at 21:57 -0400, David Edelsohn wrote: > On Tue, May 21, 2013 at 7:13 PM, Sandra Loosemore > wrote: > > On 05/21/2013 04:04 PM, David Edelsohn wrote: > >> > >> > >> There are three issues here: > >> > >> 1) Someone in the LTC toolchain team needs to benchmark this patch on > >> P

Re: [patch, powerpc] increase array alignment for Altivec

On Thu, 2013-05-23 at 08:54 -0600, Sandra Loosemore wrote: > On 05/23/2013 06:29 AM, Bill Schmidt wrote: > > > > Sandra and David, > > > > The array-alignment patch is performance-neutral with respect to > > CPU2006. All variations were in the noise range. >

[PATCH] Fix PR57441 (memory error in SLSR)

ntial overflow and heap corruption. This patch removes the no longer valid assumption and allocates a fixed-size vector. Bootstrap and regtest in progress on powerpc64-linux-gnu. Ok for trunk if everything checks out? Thanks, Bill 2013-05-29 Bill Schmidt * gimple-ssa-strength-reduct

[PATCH, rs6000] Disable analyze_swaps for UNSPEC_VSX_CVDPSPN

owerpc64le-unknown-linux-gnu with no regressions. Ok for trunk? Thanks, Bill [gcc] 2014-09-09 Bill Schmidt * config/rs6000/rs6000.c (rtx_is_swappable_p): Add UNSPEC_VSX_CVDPSPN as an unswappable operand, and add commentary on how to make it legal in future. [gcc/testsuite

Re: [PATCH 12/14][Vectorizer] Redefine VEC_RSHIFT_EXPR and vec_shr_optab as endianness-neutral

.def (VEC_RSHIFT_EXPR, VEC_LSHIFT_EXPR): Comment shift > > direction. > > > > * doc/md.texi (vec_shr_m, vec_shl_m): Document shift direction. > > > > Testing Done: > > > > Bootstrap and check-gcc on x86_64-none-linux-gnu; check-gcc on > > aarch64-no

[PATCH, rs6000] Teach analyze_swaps to avoid vec_ste

I've added a new compile-time test case to verify the fix. The test ICEs on existing trunk but passes with the new changes. Bootstrapped and tested on powerpc64le-unknown-linux-gnu with no regressions. Is this ok for trunk? Thanks, Bill [gcc] 2014-09-24 Bill Schmidt * con

[PATCH, rs6000] Fix PR63335 (vec_any/all_nge/nle)

ix to 4.8 and 4.9 as well. Thanks, Bill [gcc] 2014-09-24 Bill Schmidt PR target/63335 * config/rs6000/rs6000-c.c (altivec_build_resolved_builtin): Exclude VSX_BUILTIN_XVCMPGEDP_P from special handling. [gcc/testsuite] 2014-09-24 Bill Schmidt PR target/

[PATCH, rs6000, committed] Fix effective target in gcc.target/powerpc/pr63335.c

2014-09-26 Bill Schmidt * gcc.target/powerpc/pr63335.c: Change effective target to vsx_hw. Index: gcc/testsuite/gcc.target/powerpc/pr63335.c === --- gcc/testsuite/gcc.target/powerpc/pr63335.c (revision 215645

[PATCH, rs6000] Remove splat calls with out-of-range arguments from gcc.dg/vmx/ops.c

-unknown-linux-gnu. Ok to commit? Thanks, Bill 2014-09-29 Bill Schmidt * gcc.dg/vmx/ops.c: Remove calls to vec_splat, vec_vsplth, vec_vspltw, and vec_vspltb for which the second argument is out of range. Index: gcc/testsuite/gcc.dg/vmx/ops.c

[PATCH, rs6000, libcpp] Revise search_line_fast to avoid old unaligned load sequences

is an excellent test of search_line_fast(), and that appears to be all we do at present for the existing implementations. Bootstrapped and tested on powerpc64le-unknown-linux-gnu and powerpc64-unknown-linux-gnu with no new regressions. Is this ok for trunk? Thanks, Bill 2014-09-29 Bill Sc

[PATCH, rs6000] Warn for deprecated use of vec_lvsl and vec_lvsr for little endian

egressions. Is this ok for trunk? Thanks, Bill [gcc] 2014-09-29 Bill Schmidt * config/rs6000/rs6000-c.c (altivec_resolve_overloaded_builtin): Issue a warning message when vec_lvsl or vec_lvsr is used with a little endian target. [gcc/testsuite] 2014-09-29 Bi

[PATCH, rs6000] Generate LE code for vec_lvsl and vec_lvsr that is compatible with BE code

. Bootstrapped and tested on powerpc64le-unknown-linux-gnu with no regressions. Is this ok for trunk? Thanks, Bill [gcc] 2014-09-29 Bill Schmidt * altivec.md (altivec_lvsl): New define_expand. (altivec_lvsl_direct): Rename define_insn from altivec_lvsl. (altivec_lvsr): New

<    3   4   5   6   7   8   9   10   11   12   >