Hi
I'm sure people tracking linaro-dev@ will be interested about the
ongoing discussion on a new ARM hard-float ABI port which is taking
place on the debian-arm@ mailing-list:
http://lists.debian.org/debian-arm/2010/07/threads.html#00019
Cheers,
--
Loïc Minier
Hi,
A new Linaro Toolchain mailing-list was just created.
This list will serve Linaro development, discussions, and planning
around the usual bits: GCC, Binutils, GNU C Library / EGLIBC, and GDB
bits -- but the list of tools is very open, so LLVM, Valgrind or QEMU
are fine topics too!
(Discussions around upstream patches or new upstream features should
still be preferably done on the respective upstream mailing-lists.)
Once they see the new list, I bet the folks from the Linaro Toolchain
Working Group will be rushing to subscribe. ;-)
You can subscribe from:
http://lists.linaro.org/mailman/listinfo/linaro-toolchain
Read you there!
Cheers,
--
Loïc Minier
Hello everyone!
I'd like to ask for early testing of a new launch-control branch adding
support for generating system profile information on ARM systems. I need
you to find all the ARM hardware you have access to and run a bit of
python code and send me the result. See below for details.
== Prerequisites ==
1) Debian-based system
2) Installed packages:
* python-2.6 (2.5 support is coming)
* python-apt
* python-debian
* lsb-release
* usbutils
== Instructions ==
The branch is at: lp:~zkrynicki/launch-control/hw-profile
Instructions:
$ bzr get lp:~zkrynicki/launch-control/hw-profile \
launch-control.hw-profile
$ cd launch-control.hw-profile
$ ./get-system-profile -i -o system-profile-with-a-sensible-name.json
If something goes wrong please attach the output of the following
commands:
* lsb_release -a
* cat /proc/cpuinfo
* cat /proc/meminfo
* lsusb
* uname -m
Report bugs at: https://bugs.edge.launchpad.net/launch-control/+filebug
== About launch-control ==
Launch control is project that aims to implement the Validation
Dashboard spec. See:
https://blueprints.launchpad.net/ubuntu/+spec/arm-m-validation-dashboard
== System Detection Features ==
Hardware:
* CPU (for x86, x86_64 and arm, PPC is *not* supported yet)
* Board detection
* Memory
* USB
Software:
* distribution
* installed debian packages (via apt)
Best regards
Zygmunt
PS: re-sent from @linaro.org account, ignore the @canonical.com dupe if
it shows up please.
--
Zygmunt Bazyli Krynicki <zygmunt.krynicki(a)canonical.com>
Hi folks
Just a note that gcc-linaro branches have been moved. They used to be
owned by the gcc-linaro-dev team, and are now owned by the
linaro-toolchain-dev team (used also for binutils, gdb etc.). You can
still branch the official aliases:
lp:gcc-linaro/4.4
lp:gcc-linaro/4.5
However, if you had pushed to ~gcc-linaro-dev/gcc-linaro/4.4, you have
to explicitly push once to ~linaro-toolchain-dev/gcc-linaro/4.4
instead:
bzr push --remember lp:~linaro-toolchain-dev/gcc-linaro/4.4
Sorry for the inconvenience.
Thanks,
--
Loïc Minier
Hi
These are the notes from today's standup call, also at:
https://wiki.linaro.org/WorkingGroups/ToolChain/Meetings/2010-07-09
== Attendess ==
* Marcin Juszkiewicz
* Ulrich Weigand
* Yao Qi
* Andrew Stubbs
* Loic Minier
== Minutes ==
* Yao
* ADA patches cause regressions in the testsuite; deferred
* doko to get back to Yao on the issue
* Merged Ubuntu patches
* ACTION update wiki page
* Worked on reverting the license patch; DONE;
* ACTION update wiki page
* Merging some CodeSourcery 4.4 patches to 4.5 now
* Ulrich
* Working on https://bugs.launchpad.net/gcc-linaro/+bug/602287
* Caused by the inclusion of updated backend features, raising the
test expectations
* No actual regression since 4.4.0
* Final changes are very different and rework the infrastructure
* Decision not to backport the new infrastructure
* Recommendation to ignore failure and mark as wontfix in 4.4
* This is fixed in 4.5
* HW breakpoints and watchpoints
* Expand wiki page
* Request merging in arm_next.git to Nicolas Pitre
* Follow new versions of the patches
* Test gdb part
* Marcin
* working on intermediate stages for linux, eglibc, gcc
* linux patch got rev-ed and is pending a new review
* gcc and eglibc: hackish versions ready, need to cleanup
* Andrew
* Did a release notes wiki page; Loic reviewed
* Updated version in source tree
* Where to host the tarball?
=> Launchpad; ACTION: Loic to create a linaro-toolchain-drivers
team which can upload tarballs
* Need to review the ~4 patches in yellow; thinks it's doable before
release
* Will skip ada ones for now
* Loic
* Created milestone for Tuesday release
* Proof-read various stuff
* Sprint
* Would like to spend time with Michael Hope and Loic to discuss how
we're going to work going forward
* Andrew to discuss multiarch topic
* Ulrich would like to discuss cross-compilation
* Yao will think of some sprint topics
* Marcin would like to discuss the cross-toolchain package and the
automated bootstrap spec
* Loic would like to cover patch tracking; how do we ensure stuff is
upstream and we don't pile up delta
* Multiarch
* Andrew would like to know which advice to follow in the multiarch path bug
=> for now, wait for Steve Langasek's input and implement that
first, as it covers the existing multiarch spec
https://wiki.ubuntu.com/MultiarchSpec
* Michael Hope: stuck on a few things, but more process than code, so
will catchup after call
Thanks,
--
Loïc Minier
Hi,
A summary from this weeks kernel consolidation working group meeting
is now available. It can be viewed online at the following link and is
reproduced below.
https://wiki.linaro.org/WorkingGroups/KernelConsolidation/Meetings/2010-07-…
A list of all meetings can be found at:
https://wiki.linaro.org/WorkingGroups/KernelConsolidation/Meetings
== Wed 7th July 2010 ==
=== Attendees ===
Jamie Bennett
Matt Wadde
Jeremy Kerr
Nicolas Pitre
Amit Kucheria
Loïc Minier
David Rusling
John Rigby
Arnd Bergmann
=== Agenda ===
* Action items from last meeting.
* Matt to check that the arm_next.git tree has everything we need and
works
* Nicolas to review latest round of patches from Jeremy, then can
include in arm_next and push to Russell
* Matt checkout status of device tree support for his board
* Jeremy to chat with Nicolas about documenting boot interface
* Loic create Launchpad project and processes around kernel
consolidation
* npitre to find a way to expose the list of trees merged in
arm_next.git via a wiki page
* npitre tested his stack-overflow module on x86 and the issue is
present as well and needs to find out whether it's a bug in the compiler
or in his code
* lool to invite jcrigby explicitly
* Blueprint status.
https://blueprints.launchpad.net/ubuntu/+spec/arm-m-using-device-tree-on-armhttps://blueprints.launchpad.net/ubuntu/+spec/arm-m-kernel-version-alignmenthttps://blueprints.edge.launchpad.net/linaro-pm-wg/+spec/arm-m-kernel-power…https://blueprints.launchpad.net/ubuntu/+spec/arm-m-missing-security-featur…https://blueprints.launchpad.net/ubuntu/+spec/arm-m-debugging-with-oprofilehttps://blueprints.launchpad.net/ubuntu/+spec/arm-m-alsa-soc-fdt-bindingshttps://blueprints.edge.launchpad.net/ubuntu/+spec/kernel-maverick-arm-sing…https://blueprints.launchpad.net/ubuntu/+spec/arm-m-versatile-express-a9-en…
=== Action Items ===
* Matt to code review the arm_next.git tree and check that all patches
are present.
* jk to report on progress of talking to Grant WRT upstreaming his
patches.
* mwaddel to talk to ARM about device-tree work.
* Jeremy to chat with Nicolas about documenting boot interface.
* npitre to find a way to expose the list of trees merged in
arm_next.git via a wiki page.
* jcrigby to talk to slangasek about the list of kernel flavours to
produce.
* amitk to report back the status on linux-omap's patches.
* jcrigby to determine if versatile is still needed.
* lool to check that -fPIE support is currently enabled in the Linaro
toolchain.
* npitre and arnd to review ericm's single zImage patches.
* ericm to send single zImage patches to russelking after they have
been reviewed.
* jcrigby and lool to arrange a uboot specific call for interested
team members.
=== Minutes ===
==== Action item review ====
* mwaddel is still working on checking arm_next.git tree. He hasn't
managed to boot it yet due to issues. A code review is needed.
* [ACTION] Matt to code review the arm_next.git tree and check that
all patches are present.
* npitre has been working with jk to review his latest round of
patches. The process is ongoing and some clean-up is needed. The main
problem is that jk's patches depend on some patches from Grant Likely
and Grants aren't upstream yet.
* jk explained that Grant is working on upstreaming his patches and he
foresees no problems.
* [ACTION] jk to report on progress of talking to Grant WRT
upstreaming his patches.
* mwaddel committed to talking to ARM to discuss device-tree work.
* [ACTION] mwaddel to talk to ARM about device-tree work.
* 3 patches still remain on the boot interface and jk is still working
on them. npitre needs to review.
* [ACTION] Jeremy to chat with Nicolas about documenting boot
interface.
* lool asked the working group to email him activity reports weekly to
facilitate the production of a weekly report.
* npitre didn't manage to complete an action item on the list of trees
merged in arm_next.git. This was carried over.
* [ACTION] npitre to find a way to expose the list of trees merged in
arm_next.git via a wiki page.
* [ACTION] npitre to describe the problem he is having with the
stack-overflow module and pass this onto a tool chain person for
investigation.
* jcrigby will take up the packaging of the kernel tree.
* [ACTION] jcrigby to talk to slangasek about the list of kernel
flavours to produce.
==== Blueprint Status ====
* jk explained that the last week has been spent rebasing his current
patches and working with npitre to review them.
* He also mentioned that no new hardware had been enabled and will not
be until the current patchset is approved.
* OMAP3 and OMAP4 is being worked by Linaro and its partners, patches
will be fed into the Linaro kernel.
* Some partners want complete tree branches merging into the Linaro
kernel to enable their new hardware to boot 'out-of-the-box' with
Linaro.
* davidrusling asked about the status of OMAP WRT to upstream. amitk
explained that about 700 patches are not upstream for OMAP4 but OMAP3 is
in a much better position. TI have recently added patches for the Panda
board to the linux-omap tree. It is understood that around half of the
700 patches will need some work to upstream. npitre said that he is
willing to take large chunks of code as long as it merges nicely.
* [ACTION] amitk to report back the status on linux-omap's patches.
* The issue of keeping a stable kernel tree was discussed. One possible
solution was to have two tree's, one that will constantly rebase with
upstream and continue to fold new code in with a second tree being more
'tried and tested' which could be used as a basis for release. The
details still need working out.
* jcrigby is currently packaging Linux versatile but the patches to
enable QEMU support for versatile aren't available in Linaro ATM. lool
asked if we need to keep versatile support, jk explained that keeping it
would be a good idea as its currently used. A final decision needs to be
made after investigations.
* [ACTION] jcrigby to determine if versatile is still needed.
* Arnd Bergmann joined the work group this week, he introduced himself
and explained that he would be working on QEMU initially.
* Kernel power management discussion will be removed from this work
group as it has its own now.
* npitre explained that one of his work items on the missing security
features was blocking on toolchain support. lool asked npitre to file a
bug.
* [ACTION] lool to check that -fpie support is currently enabled in the
Linaro kernel.
* npitre is monitoring ericm's patches for a single zImage kernel. He
believes that some patches are simple enough to get upstream.
* [ACTION] npitre and arnd to review ericm's single zImage patches.
* [ACTION] ericm to send single zImage patches to russelking after
they have been reviewed.
* mwaddel explained that there is no support for mmc on the versatile
express boards. This needed adding.
* Future tracking of uboot work will be discussed on the call from the
future.
* [ACTION] jcrigby and lool to arrange a uboot specific call for
interested team members.
Regards,
Jamie.
--
Linaro Release Manager
Hi
We're started having standup calls in the Toolchain WG as an
experiment. The goal is to identify blockers and resolve
issues/questions as soon as they appear, checking the health of the
work queues of people in the group etc. They are scheduled on
Wednesdays and Fridays complementing the weekly WG calls on Mondays.
These are the notes from today's first call, also at:
https://wiki.linaro.org/WorkingGroups/ToolChain/Meetings/2010-07-07
== Attendees ==
* Yao Qi
* Andrew Stubbs
* Michael Hope
* Loic Minier
== Minutes ===
=== Andrew ===
* has work to do :-)
* will attack review of the armel regressions now that Loic's beagle
is up again
* pbrook assigned himself to perl + mysql bugs and has a preliminary
patch for one of them
=== Michael ===
* needs more work since the getfem++ failure is unreproducible
* should do the first backport(s) from
https://wiki.linaro.org/WorkingGroups/ToolChain/UbuntuRegressions
* some people on holiday, but should try to get upstream patches
(best effort) reviewed before backporting them from the upstream
mailing-list
* assigned to LP #602171 now
We reviewed some of the issues at the top of
https://wiki.linaro.org/WorkingGroups/ToolChain/UbuntuRegressions and
updated status, some of the issues need to be clarified with Matthias
Klose (doko).
=== Yao ===
* moving patches from his local CodeSourcery tree with Ubuntu patches
into bzr
* will send a merge request for the Ubuntu patches today
* faces issues with ada patches; we probably failed to identify
prerequisite patches; need to do test build
=== Loic ===
* slacking, as usual ;-)
* fixed beagleboard
* offers himself for help on the packaging and bzr issues which might
popup; will help Michael with bzr + process documentation and Yao
with package builds
Thanks,
--
Loïc Minier
A test rebuild of maverick (main) on amd64 and i386 did finish last weekend [1].
- There are a few failures which I could not reproduce on
plain maverick (using the maverick GCC). Bug reports are filed
for the linaro-gcc project, tagged with `armel toolchain'.
- Bug reports for packages which fail to build in maverick [2] are
filed for ubuntu with priority high, milestoned to alpha-3.
- Some packages are unchecked (uninstallable build dependencies,
when the snapshot was taken) [3].
Some issues with our build infrastructure to do test rebuilds on armel will be
fixed this week, and the armel test rebuild should start this week.
Matthias
[1] https://launchpad.net/ubuntu/+archive/test-rebuild-20100628
[2]
https://launchpad.net/ubuntu/+archive/test-rebuild-20100628/+builds?build_t…
[3]
https://launchpad.net/ubuntu/+archive/test-rebuild-20100628/+builds?build_t…
Hi there
I should have sent out such weekly notes much earlier, but it's never
too late to set it right! So please find below the minutes from
today's Toolchain Working Group meeting and the activity reports from
previous week.
You can find the full meeting notes at
https://wiki.linaro.org/WorkingGroups/ToolChain/Meetings/2010-07-05
== Agenda ==
* New starter: Michael Hope
* Action items from last meeting
* Blueprints status
* cross-compilers blueprint (hrw)
* maverick armel toolchain
* testsuite regressions
* regression FTBFSes
* merging Ubuntu patches (low)
* Proposed release date, next Monday
* Assigning important package build failures (FTBFSes) in Ubuntu with
Linaro GCC to Linaro folks
* MySQL LP #600936
* Perl LP #600951
* Upcoming new mailing-list
* Conflict with GCC Summit
== Action Items from this Meeting ==
* Loic and Andrew to extend wiki page to track progress of the merging
of Ubuntu patches, assignees etc.
* Loic to setup and run a call on merge workflow with bzr
* Matthias to sort autotools issues with new cell-4.4 Linaro patch or
to ask Ulrich for help :-)
* Marcin to fix -gcc symlink (LP #600927) and update-alternatives
installability issue in his public cross-compiler packages
== Action Items from Previous Meeting ==
* Loic to create wiki page tracking merging of Ubuntu patches starting
from Andrew's review: DONE (by Andrew)
* Ulrich to merge the cell-4.4 patch with the Linaro branch: DONE
* Julian and Andrew to review remaining testsuite failures
* Loic to check whether we can resolve conflicts between next GCC
Summit and Linaro Developers Summit
* Paul Brook to assign himself to the FTBFS bug he is working on
* Ulrich to pick up the other FTBFS bug
* Loic to setup gcc-linaro milestone for July release
== Minutes ==
* New starter from Canonical: Michael Hope; round of introductions
* Ubuntu patches being reviewed and tested within CodeSourcery
* Can be merged in Linaro branches right now
* Loic and Andrew to extend wiki page to track progress of the
merging
* Loic, Yao and Andrew to have a bzr call on using the bzr
review/merge process; Loic to organize; same time tomorrow
(Tuesday)
* Loic and Andrew to discuss binutils patches this week with doko
* Ulrich sent a new cell-4.4 patch on top of the Linaro branch; doesn't
work for doko due to some autotools issues; Ulrich believes this is
minor stuff which Matthias will fix
* Andrew had no time to review the remaining testsuite failures; Julian
was on leave next week
* Andrew to review the results when Loic tells him his beagle is up
again
* Blueprint status
* cross-compilers
* hrw wants to check that everything is merged in the gcc-4.5
packaging SVN
* doko working on 4.5.1 merge
* hrw will fix the regression raised on debian-embedded@
* plans to rewrite his internal stage patches
* -gcc symlink LP #600927 is about to be fixed, needs merging in
packaging SVN
* update-alternatives issue
* linaro-m / Ubuntu maverick toolchain blueprint
* see above for testsuite failures and merging of Ubuntu patches
* regression FTBFSes
* Paul Brook looking into at least one of the two failure to
build issues
* Andrew to ask pbrook to assign himself to the bug
* Ulrich to poke the other bug once we know which one that is
* Initial release
* Proposed release date: next Monday; next Tuesday is actually best
to discussthis next Monday; Andrew raises that the perl+mysql bugs
might be hard, we dont know
* Release version: 4.x + 2010.07 (month of the release); monthly releases
* Loic to setup a milestone to track this
* New mailing-list: new linaro-toolchain public mailing-list on its
way, stay tuned
* Conflict with GCC Summit; Loic checked with Mark Mitchell and Kiko
what we could do about the conflict, not an option to move LDS / UDS,
GCC Summit date was probably just agreed upon (implies reservations
were made :-/) so probably not an option to move it either; Mark
seemed to think it wasn't an issue to send some folks to each summit
Activity reports from last week:
= Ulrich Weigand =
== GCC ==
* Resolved merge conflicts between cell-branch.diff and gcc444-linaro.diff
== Education ==
* Started getting familiar with Debian package build process
* Started getting familiar with Launchpad infrastructure
* Started reading ARMv7 Architecture Reference Manual
== Infrastructure ==
* Set up ARM root filesystem for use under QEMU
* Investigated options for ARM development hardware
== Admin ==
* Completed NewStarters tasks
* Started participating in toolchain WG calls
* Made travel arrangments for Prague meeting
= Yao Qi =
== Linaro GCC ==
* Import Ubuntu GCC patches [1] to Linaro GCC 4.4.
Patches on C/C++/fortran are applied, natively built and tested.
Check them in internal git repo.
* Fixing bootstrapping error when ada patches in [1] are applied.
More ada patches in [1] might be applied.
== Misc ==
* Set up account on linaro.org, and have an interview meeting with
Loïc and Andrew.
= Andrew Stubbs =
== CS work plan ==
* Attended many meetings and wrote many emails, documents, and
blueprints etc. towards completing the CodeSourcery Work Plan.
* This has now been agreed, so now we can get on with real work. :)
= Loic Minier =
* Finished gcc-4.4 + Linaro build on local beagleboard; tried to
* upgrade to maverick kernels and broke boot in the process; LP #591941
* Wrote up activity reporting guidelines
https://wiki.linaro.org/WorkingGroups/ToolChain/Reports
* Compiled Bzr tips page
https://wiki.linaro.org/WorkingGroups/ToolChain/BzrTips
* Interviews with candidates and introductory calls with new starters
* Toolchain WG planning, CodeSourcery work plan review
Cheers
--
Loïc Minier