Merge lp:~uweigand/gcc-linaro/vecsetextractmem-4.7 into lp:gcc-linaro/4.7

Proposed by Ulrich Weigand
Status: Rejected
Rejected by: Ulrich Weigand
Proposed branch: lp:~uweigand/gcc-linaro/vecsetextractmem-4.7
Merge into: lp:gcc-linaro/4.7
To merge this branch: bzr merge lp:~uweigand/gcc-linaro/vecsetextractmem-4.7
Reviewer Review Type Date Requested Status
Linaro Toolchain Developers Pending
Review via email: mp+100293@code.launchpad.net

This proposal supersedes a proposal from 2012-03-29.

Description of the change

Allow memory scalar operands with vec_set/vec_extract, using vld1/vst1.

Updated version: fixed V2DImode patterns.

Not yet upstream; merge request for test and benchmark purposes.

To post a comment you must log in.
Revision history for this message
Michael Hope (michaelh1) wrote : Posted in a previous version of this proposal

cbuild has taken a snapshot of this branch at r114975 and queued it for build.

The diff against the ancestor r114974 is available at:
 http://builds.linaro.org/toolchain/snapshots/gcc-linaro-4.7+bzr114975~uweigand~vecsetextractmem-4.7.diff

and will be built on the following builders:
 a9-builder armv5-builder i686 x86_64

You can track the build queue at:
 http://ex.seabright.co.nz/helpers/scheduler

cbuild-snapshot: gcc-linaro-4.7+bzr114975~uweigand~vecsetextractmem-4.7
cbuild-ancestor: lp:gcc-linaro/4.7+bzr114974
cbuild-state: check

Revision history for this message
Michael Hope (michaelh1) wrote : Posted in a previous version of this proposal
Revision history for this message
Michael Hope (michaelh1) wrote : Posted in a previous version of this proposal
Revision history for this message
Michael Hope (michaelh1) wrote : Posted in a previous version of this proposal
Revision history for this message
Michael Hope (michaelh1) wrote : Posted in a previous version of this proposal
Download full text (4.1 KiB)

cbuild successfully built this on armv7l-natty-cbuild280-ursa1-cortexa9r1.

The build results are available at:
 http://ex.seabright.co.nz/build/gcc-linaro-4.7+bzr114975~uweigand~vecsetextractmem-4.7/logs/armv7l-natty-cbuild280-ursa1-cortexa9r1

-PASS: gcc.c-torture/execute/vshuf-v2df.c compilation, -O2
-PASS: gcc.c-torture/execute/vshuf-v2df.c compilation, -O2 -flto -fno-use-linker-plugin -flto-partition=none
-PASS: gcc.c-torture/execute/vshuf-v2df.c compilation, -O2 -flto -fuse-linker-plugin -fno-fat-lto-objects
-PASS: gcc.c-torture/execute/vshuf-v2df.c compilation, -O3 -fomit-frame-pointer
-PASS: gcc.c-torture/execute/vshuf-v2df.c compilation, -O3 -g
+FAIL: gcc.c-torture/execute/vshuf-v2df.c compilation, -O2
+FAIL: gcc.c-torture/execute/vshuf-v2df.c compilation, -O2 -flto -fno-use-linker-plugin -flto-partition=none
+FAIL: gcc.c-torture/execute/vshuf-v2df.c compilation, -O2 -flto -fuse-linker-plugin -fno-fat-lto-objects
+FAIL: gcc.c-torture/execute/vshuf-v2df.c compilation, -O3 -fomit-frame-pointer
+FAIL: gcc.c-torture/execute/vshuf-v2df.c compilation, -O3 -g
-PASS: gcc.c-torture/execute/vshuf-v2df.c execution, -O2
-PASS: gcc.c-torture/execute/vshuf-v2df.c execution, -O2 -flto -fno-use-linker-plugin -flto-partition=none
-PASS: gcc.c-torture/execute/vshuf-v2df.c execution, -O2 -flto -fuse-linker-plugin -fno-fat-lto-objects
-PASS: gcc.c-torture/execute/vshuf-v2df.c execution, -O3 -fomit-frame-pointer
-PASS: gcc.c-torture/execute/vshuf-v2df.c execution, -O3 -g
+UNRESOLVED: gcc.c-torture/execute/vshuf-v2df.c execution, -O2
+UNRESOLVED: gcc.c-torture/execute/vshuf-v2df.c execution, -O2 -flto -fno-use-linker-plugin -flto-partition=none
+UNRESOLVED: gcc.c-torture/execute/vshuf-v2df.c execution, -O2 -flto -fuse-linker-plugin -fno-fat-lto-objects
+UNRESOLVED: gcc.c-torture/execute/vshuf-v2df.c execution, -O3 -fomit-frame-pointer
+UNRESOLVED: gcc.c-torture/execute/vshuf-v2df.c execution, -O3 -g
-PASS: gcc.c-torture/execute/vshuf-v2di.c compilation, -Os
+FAIL: gcc.c-torture/execute/vshuf-v2di.c compilation, -Os
-PASS: gcc.c-torture/execute/vshuf-v2di.c execution, -Os
+UNRESOLVED: gcc.c-torture/execute/vshuf-v2di.c execution, -Os
-PASS: gcc.dg/torture/pr52407.c -O3 -fomit-frame-pointer (test for excess errors)
-PASS: gcc.dg/torture/pr52407.c -O3 -fomit-frame-pointer execution test
-PASS: gcc.dg/torture/pr52407.c -O3 -fomit-frame-pointer -funroll-all-loops -finline-functions (test for excess errors)
-PASS: gcc.dg/torture/pr52407.c -O3 -fomit-frame-pointer -funroll-all-loops -finline-functions execution test
-PASS: gcc.dg/torture/pr52407.c -O3 -fomit-frame-pointer -funroll-loops (test for excess errors)
-PASS: gcc.dg/torture/pr52407.c -O3 -fomit-frame-pointer -funroll-loops execution test
-PASS: gcc.dg/torture/pr52407.c -O3 -g (test for excess errors)
-PASS: gcc.dg/torture/pr52407.c -O3 -g execution test
+FAIL: gcc.dg/torture/pr52407.c -O3 -fomit-frame-pointer (test for excess errors)
+UNRESOLVED: gcc.dg/torture/pr52407.c -O3 -fomit-frame-pointer compilation failed to produce executable
+FAIL: gcc.dg/torture/pr52407.c -O3 -fomit-frame-pointer -funroll-all-loops -finline-f...

Read more...

Revision history for this message
Michael Hope (michaelh1) wrote :

cbuild has taken a snapshot of this branch at r114976 and queued it for build.

The diff against the ancestor r114974 is available at:
 http://builds.linaro.org/toolchain/snapshots/gcc-linaro-4.6+bzr114976~uweigand~vecsetextractmem-4.7.diff

and will be built on the following builders:
 a9-builder armv5-builder i686 x86_64

You can track the build queue at:
 http://ex.seabright.co.nz/helpers/scheduler

cbuild-snapshot: gcc-linaro-4.6+bzr114976~uweigand~vecsetextractmem-4.7
cbuild-ancestor: lp:gcc-linaro+bzr114974
cbuild-state: check

Revision history for this message
Michael Hope (michaelh1) wrote :

Re-spawning due to a bug in the taker. The devfocus was still 4.6 so the build completed and tested but has '4.6' everywhere across it.

cbuild-state: queue

Revision history for this message
Michael Hope (michaelh1) wrote :

cbuild has taken a snapshot of this branch at r114976 and queued it for build.

The diff against the ancestor r114974 is available at:
 http://builds.linaro.org/toolchain/snapshots/gcc-linaro-4.7+bzr114976~uweigand~vecsetextractmem-4.7.diff

and will be built on the following builders:
 a9-builder armv5-builder i686 x86_64

You can track the build queue at:
 http://ex.seabright.co.nz/helpers/scheduler

cbuild-snapshot: gcc-linaro-4.7+bzr114976~uweigand~vecsetextractmem-4.7
cbuild-ancestor: lp:gcc-linaro+bzr114974
cbuild-state: check

Revision history for this message
Michael Hope (michaelh1) wrote :
Revision history for this message
Michael Hope (michaelh1) wrote :
Revision history for this message
Michael Hope (michaelh1) wrote :
Revision history for this message
Michael Hope (michaelh1) wrote :
Revision history for this message
Michael Hope (michaelh1) wrote :
Revision history for this message
Michael Hope (michaelh1) wrote :
Revision history for this message
Michael Hope (michaelh1) wrote :
Revision history for this message
Michael Hope (michaelh1) wrote :
Revision history for this message
Loïc Minier (lool) wrote :

Is this still active?

Revision history for this message
Ulrich Weigand (uweigand) wrote :

Yes, but moving forward on this (just like mov-vld1.64-4.7) is blocked pending a fix to GCC's currently incorrect setting of alignment requirements for vector types. Without this fix, applying this patch will introduce latent bugs (just like mov-vld1.64-4.7, except that there the bugs are not just latent).

There is a patch currently under discussion on gcc-patches:
http://gcc.gnu.org/ml/gcc-patches/2012-06/msg00589.html
which will fix the alignment requirements. Once this is accepted upstream, I'll move forward with those two merge requests.

Revision history for this message
Loïc Minier (lool) wrote :

Ok, thanks for the update!

--
Loïc Minier

Subscribers

People subscribed via source and target branches