The Linaro Toolchain Working Group is pleased to announce the release
of both Linaro GCC 4.6 and Linaro GCC 4.5.
Linaro GCC 4.6 is the fourth release in the 4.6 series. Based off the
latest GCC 4.6.0+svn174261, it adds new optimisations and vectoriser
improvements.
Interesting changes include:
* Updates to 4.6.0+r174261
* Blocks can now vectorise into ORN and BIC instructions
* Support for half word to double word multiply and accumulate operations
* Better support for other widening multiply operations
* Further performance improvements in NEON strided loads and stores
* Performance improvements targeted at EEMBC CoreMark
Fixes:
* PR target/48454: Set the lengths correctly for the case with Quad vectors.
Known issues:
* Building Python 2.7 with -mfpu=neon exposes a bug in vmov.i64 in
binutils 2.20.51. Please use 2.21 or later.
The strided load/store improvements allow the vectoriser to
efficiently access values that occur at every n'th address, such as
all of the red values in a RGB image or all of the left channel
samples in a interleaved audio array. For example, a plain C function
that converts between RGB and CYMK now runs 7.3 x faster on an A9.
Linaro GCC 4.5 2011.06 is the eleventh release in the 4.5 series.
Based off the latest GCC 4.5.3+svn174250, it is a maintenance focused
release.
Interesting changes in 4.5 include:
* Updates to 4.5.3+r174250
Fixes:
* LP: #744754: ICE in reload_cse_simplify_operands, at
postreload.c:402 with neon optimized code
* LP: #748138: ICE in redirect_jump, at jump.c:1443
The source tarball is available from:
https://launchpad.net/gcc-linaro/+milestone/4.6-2011.06-0https://launchpad.net/gcc-linaro/+milestone/4.5-2011.06-0
Downloads are available from the Linaro GCC page on Launchpad:
https://launchpad.net/gcc-linaro
Mailing list: http://lists.linaro.org/mailman/listinfo/linaro-toolchain
Bugs: https://bugs.launchpad.net/gcc-linaro/
Questions? https://ask.linaro.org/
Interested in commercial support? inquire at support(a)linaro.org
-- Michael
Hi there,
I tried to use android-toolchain-eabi-4.5-2011.05-0-linux-x86.tar.bz2
to enable hard float ABI.
But it built failed on Android 2.3.3. Most errors are like address
alignment issues.
Did you have any patch for that? Successfully tested?
Thanks
Richard
My testing for the widening multiplies patches came back clean. I've
committed it upstream, and merged it to Linaro GCC 4.6.
The testing for my Thumb2 constants patches failed (bootstrap failure on
ARM), so can't commit that right away. The failure is a SIGSEGV in the
stage 2 compiler building it's own libgcc. This is going to be tricky to
pin down!
Continued work on widening operations in the GCC middle-end. I now have
it doing all the arbitrary width widening operations that I want. I've
tested that it didn't do the wrong thing for just about all permutations
of input and output types, but I've yet to do anything like a bootstrap
test or anything. There's still quite a lot more work to do in tidying
it up.
----
Upstream patched requiring review:
* NEON scheduling patch
http://gcc.gnu.org/ml/gcc-patches/2011-02/msg01431.html
Tracked a bootstrap failure with SMS flags on ARM machine exposed in
recent trunk.
Fixed and tested a patch for that.
Tested another SMS patch following comments received from gcc ml@
(http://gcc.gnu.org/ml/gcc-patches/2011-05/msg02294.html)
Following conversation with Micheal, added SPEC2006 analysis info to
the benchmarks wiki page
-https://wiki.linaro.org/Internal/ToolChain/Benchmarks
Discussed with Ramana weird RTL pattern generated for
thumb2_movhi_insn with -march=armv7-a -mthumb using trunk. Probably
will open a PR for that.
Hi,
* learning more on andoid (repo tool, some branches, the basics of the
android build system)
* finished to setup my environment to build the android sources
* sucessfully build linaro-android using linaro gcc 4.4, 4.5 and 4.6
* tracked down the libgui.so linaro android issue to the apriori
prelinker
* workaround: disable prelinking for this module (patch -> linaro-dev)
* I still don't understand what exactly is supposed to be prelinked
and what's exactly causing the fail
* Note: next Monday is a public holiday in Germany
Regards
Ken
== String Routines ==
* Completed gathering the SPEC2k6 memcpy results, graphed them, sent them out
* Gathered SPEC2k6 memset results, graphed them, sent them out
== 64bit Atomics ==
* Modified gcc backend to do 64bit Atomic ops - the code looks good,
but I've not
done much testing yet.
== Other ==
* Upstreamed a small ltrace patch
Next week:
Plan is to get gcc tests done and attack libgcc for the pre-v7
fallbacks (the tricky
bit there is runtime deciding what to use)
Also run spec and denbench for strlen and some other string routines
== Progress ==
* Spent some time on the VFP moves and look at ivopts for a bit.
Analysing a couple of options here.
* Committed the DImode moves patch upstream.
* Fixed PR49335 where GCC was generating rsb ip, sp, ip lsl #2
* Proposed a fix for PR48454 (finally being able to reproduce it) -
was a case of a missing length attribute for vec_pack_trunc.
* Investigated some regressions in v1 of a popular embedded benchmark
- leads to BRANCH_COST tuning.
* Merged by neon-vorn-vbic patch into linaro-gcc :4.6
== Plans ==
* Spend some time on the VFP moves and look at ivopts for a bit.
* Merge the fix for lengths to linaro-gcc:4.6 and 4.5 if applicable.
* Merge fix for PR49335 to linaro-gcc:4.5 and 4.6 .
* Find some time for some upstream patch review.
* T2 performance review meeting/
Meetings:
* 1-1s
* TCWG call.
RAG:
Red:
Amber:
Green: USB kbd+mouse finally working on QEMU beagle model
Current Milestones:
|| || Planned || Estimate || Actual ||
||qemu-linaro-2011-06 || 2011-06-16 || 2011-06-16 || ||
Historical Milestones:
||finish qemu-cont-integn || 2011-01-25 || 2011-01-25 || handed off ||
||first qemu-linaro release || 2011-02-08 || 2011-02-08 || 2011-02-08 ||
||qemu-linaro 2011-03 || 2011-03-08 || 2011-03-08 || 2011-03-08 ||
||qemu-linaro 2011-04 || 2011-04-21 || 2011-04-21 || 2011-04-21 ||
||qemu-linaro 2011-05 || 2011-05-19 || 2011-05-19 || n/a ||
||close out 1105 blueprints || 2011-05-28 || 2011-05-28 || 2011-05-19 ||
||complete 1111 planning || 2011-05-28 || 2011-05-28 || 2011-05-27 ||
== upstream-omap3-patches ==
* started on disentangling the patchstack: submitted patches upstream
for a few standalone fixes. First few steps in a big job...
== omap3-usb-model ==
* added QEMU's USB OHCI model to the omap3/beagle, and (after some
debugging and submitting a couple of OHCI bugfixes upstream)
got USB keyboard and mouse working
== linaro-qemu-11.11 ==
* rebased on master and identified gcc 4.6 compiler patches we need
== other ==
* added and tested patches for an overo board model
* added Beagle board support for returning EDID data from a fake
monitor so the kernel will actually turn the display on
* discussions about Android emulator (which looks likely to take the
upstreamed ARMv7 translator with the fixes we've worked on over
the last six months)
* office move
* QEMU 0.15 is not too far in the future: need to make sure all the
ARM stuff we want is in it
Meetings: standup, GSoC student, Christoffer Dall (working on KVM for ARM)
Current qemu patch status is tracked here:
https://wiki.linaro.org/PeterMaydell/QemuPatchStatus
Absences:
1-5 August: Linaro sprint 1111
(maybe) 15-16 August: QEMU/KVM strand at LinuxCon NA, Vancouver
[LinuxCon proper follows on 17-19th]
Hi,
- vectorization of widening multiplication of unsigned types and
constants - committed to mainline
- fix vectorizer testsuite failures on ARM - submitted
- testing a patch to fix a bug in the vectorizer revealed by the
widen-mult patch
- testing a patch to fix bad peeling heuristic that causes
gcc.dg/vect/vect-72.c to fail on ARM
Ira
