- Andrew Stubbs
- Chung-Lin Tang
- Dave Gilbert
- Ira Rosen
- Ken Werner
- Marcin Juszkiewicz
- Michael Hope
- Mounir Bsaibes
- Peter Maydell
- Ramana Radhakrishnan
- Revital Eres
- Richard Sandiford
- Ulrich Weigand
- The test results for the patch for lp:675347 on GCC 4.6 came back clean, so I merged it to Linaro GCC 4.6.
- The test results for lp:675347 on 4.5 had problems though, but they might be unrelated to the patch. The test results for the "discourage NEON on A8" patch had similar failures, and that's a 4.6 testsuite.
- Richard Earnshaw approved the Thumb register allocation patch. I've committed it upstream, and updated the patch trackers. It was already on the Linaro 4.6 branch.
- Now that GCC 4.6 is released, switched all the Linaro tracking tickets from 'Fix committed' to 'Fix released'.
- Merged from FSF 4.5 to Linaro 4.5 and submitted the patch for test. The tests came back clean, so I pushed it to the 4.5 branch. (Yay for Michael's new test service!)
- Merged more patches from SG++ to Linaro. Or, at least considered them for merge. Mostly I decided that they were not appropriate for Linaro, at least, not just yet. I have yet to push these patches to Launchpad.
Reviewed Richard Sandiford's patch for 714921.
- Retried the Android build with a view to integrating Android support in Linaro GCC 4.5 (4.6 should already support it). Eventually, after downloading many different git repositories and branches, and maxing out the memory on my machine a few times, I managed a successful build using the toolchain the Android team are using. I then backported Maxim's patches to Linaro GCC 4.5, and built and tested that, and got another successful Android build. I've pushed the patched toolchain to Launchpad at lp:~ams-codesourcery/gcc-linaro/android for testing. All being well, I'll merge Android support into the 4.5 trunk in time for the next release.
- PR48250 / CS Issue #9845 / Launchpad #723185. Unaligned DImode reloadunder NEON. Worked on new patch, submitted to gcc-patches after testing on Friday. Awaiting review.
CoreMark ARMv6/v7 regressions: working on new combine patch.
- We seem to have picked up a lot of ftbfs in the last couple of weeks - which is unfortunate because it may well be too close to the Natty release to do anything about them
Bug 745843 is a repeatable segfault in part of the build process of a package called vtk that is used by a few other things ; I've got this down to a particular call of one function - although gdb is getting rather confused (r0 & r1 changing as I single step across a branch)
- Bug 745861 petsc build failure; I'm getting one of two different link errors depending which mood it is in - mpi related?
- Bug 745873 - a meta package that just didn't have a list of packages to build with for armel; easy to do a simple fix (provided branch that built) for but the maintainer
says it's too late for natty anyway and some more thought is needed.
- Reading over some optimisation documents
- Tested weekly release on Beagle-c4 (still no OTG usb and hence no
networking for me)
- Also simple boot test on panda; not much time for more thoroug test. (seems to work)
- continued bringing patches upstream
- changing default vector size to 128 - resubmitted with changes
- according to comments, awaiting review
- if-conversion improvement - committed
- PR 48252 - bug in vzip/vuzp/vtrn implementation - patch submitted
- opened PR 48454 - a test failure with -mvectorize-with-neon-quad
- created a generic and local variant of the extbl parser
- continued to look into testsuite failures
down to 12 failures: https://wiki.linaro.org/KenWerner/Sandbox/libunwind
- continue to post patches upsteam
Cycle 11.11 planning
- cycle 11.05 blueprints reconciliation
- cycle 11.11 Technical Requirements sizing and allocation
- create a spreadsheet with macros to easy estimate and view who is over allocated and who is not.
- I spent a couple of days on initial cleanup of the omap3 patchstack in qemu-linaro. It's still some way from being upstreamable but at least now every patch in the stack compiles; this should make rebasing on upstream a bit less painful.
the board-ram-limits patchset is still stalled with upstream
- Aurelien applied lots of patches so the pipeline has drained again
- cleaning up/reworking patches which fix handling of Neon UNDEF cases. Not very exciting but it will get a large set of patches out of the qemu-linaro patchstack. == other ==
- meetings: toolchain, standup
Current qemu patch status is tracked here: https://wiki.linaro.org/PeterMaydell/QemuPatchStatus
- Sync'd with Andrew about T2 performance stuff.
- Spent some time investigating what could be done for divmodsi4 improvements . Working through the various phases. Not done tree level stuff for a while so my knowledge of the API is a a bit rusty.
- Some upstream bugzilla duty.
- Continue looking at divmodsi4 improvements.
- Continue looking at excessive VFP moves.
- Look at binutils + neon issue.
- Performance kick-off meeting
- Linaro toolchain meeting
- Andrew Stubbs meeting for 1 hour about T2 performance.
- April 13th - Internal ARM conference. Not at desk all day.
April 15 – 26 -> Booked.
- May 9-14 - LDS Budapest
- Sent a fix for PR target/46329 upstream.
- Discussed with Richard Guenther how to represent the interleaved load/store "functions" that we're adding to gimple. Sent a patch upstream for comments. Richard confirmed on IRC that he was happy with it, and no-one else has objected.
- Spent most of the week on the vectorisation itself, and on the testsuite.
- Finish work on vectorisation testsuite and submit.
WorkingGroups/ToolChain/ActivityReports/2011-04-08 (last modified 2011-04-13 14:23:35)