- Fixed LLD segfault when exceptions.
- Implement large code model TLSLD relocations in LLD.
- More thoughts on MOVK prel relocations in the ABI as there looks to
be a use case for them in HWASAN that isn't covered by the existing
relocations.
- Some thoughts on how TCWG LLVM Jira initiatives could be worded and
how we may want to word them in the next cycle.
- Initial thoughts written down on Connect presentation, need to flesh
out more this week.
- Upgraded machine to 18.04, almost managed to break it in the process
but luckily it was recoverable. Will likely have some teething
problems.
o LLVM:
* Machine Outliner:
- Improved stack fixup handling (remove cnadidates when it is not
bebeficial)
- Fixing an issue related to R12 usage.
o Misc
* Various meetings and discussions.
[VIRT-327 # Richard's upstream QEMU work ]
Posted v4 of arm hflags reorg.
Split out 3 minor patch sets from the larger aa32 decodetree set.
Reviewed v6 of invert-endian tlb patch set.
Reviewed ajb's fpu header reorg.
Posted an RFC vs Andrew Jones' SVE-in-KVM patch set.
r~
QEMU Tooling ([VIRT-252])
=========================
- working through v4 comments on [next version branch]
[next version branch]
https://github.com/stsquad/qemu/tree/plugins/plugins-v5
Upstream Work ([VIRT-109])
==========================
- spent some more time looking at TCG EL2 behaviour
- there is something not working properly even without the VHE
patches
- need to do some tooling work to be able to properly debug HYP code
- posted {PATCH v1 0/7} softfloat header cleanups Message-Id:
<20190808164117.23348-1-alex.bennee(a)linaro.org>
- posted {PATCH v2 0/7} softfloat includes cleanup Message-Id:
<20190809091940.1223-1-alex.bennee(a)linaro.org>
- posted {PATCH v1 0/2} docker DEF_TARGET_LIST cleanup Message-Id:
<20190809155047.24526-1-alex.bennee(a)linaro.org>
[VIRT-109] https://projects.linaro.org/browse/VIRT-109
Completed Reviews [6/6]
=======================
{Qemu-arm} {PATCH 0/2} target/arm: Fix routing of singlestep exceptions
Message-Id: <20190805130952.4415-1-peter.maydell(a)linaro.org>
- CLOSING NOTE [2019-08-07 Wed 11:47]
Looks good, would be nice to integrate a better testcase
{PATCH} gdbstub: Fix handling of '!' packet with new infra
Message-Id: <20190805190901.14072-1-ramiro.polla(a)gmail.com>
- CLOSING NOTE [2019-08-07 Wed 11:59]
Queued to my tree
{PATCH 0/3} tests/tcg: disentangle makefiles
Message-Id: <20190730123759.21723-1-pbonzini(a)redhat.com>
{Qemu-devel} {PATCH v2 00/29} Tame a few "touch this, recompile the world" headers
Message-Id: <20190806151435.10740-1-armbru(a)redhat.com>
- CLOSING NOTE [2019-08-07 Wed 17:32]
Looks OK but breaks a lot of cross compiles. More subtly to fix.
{Qemu-devel} {PATCH untested for-4.2} memory: fix race between TCG and accesses to dirty bitmap
Message-Id: <20190729214717.6616-1-pbonzini(a)redhat.com>
{PATCH v3 00/29} Tame a few "touch this, recompile the world" headers
Message-Id: <87k1bmpn7y.fsf(a)dusky.pond.sub.org>
Current Review Queue
====================
* {PATCH v3 00/34} target/arm: Implement ARMv8.1-VHE
Message-Id: <20190803184800.8221-1-richard.henderson(a)linaro.org>
* {Qemu-devel} {PATCH v4 00/22} target/arm: Implement ARMv8.5-MemTag, system mode
Message-Id: <20190307170440.3113-1-richard.henderson(a)linaro.org>
* {Qemu-devel} {RFC PATCH} Implement qemu_thread_yield for posix, use it in mttcg to handle EXCP_YIELD
Message-Id: <20190717054655.14104-1-npiggin(a)gmail.com>
* {PATCH v3 0/6} tests/docker: add podman support
Message-Id: <20190713143311.17620-1-marcandre.lureau(a)redhat.com>
* {PATCH 0/2} tests/acceptance: Add test of NeXTcube framebuffer using OCR
Message-Id: <20190629150056.9071-1-f4bug(a)amsat.org>
* {Qemu-devel} {PATCH 0/4} Introduce the microvm machine type
Message-Id: <20190628115349.60293-1-slp(a)redhat.com>
--
Alex Bennée
Progress:
* VIRT-65 [QEMU upstream maintainership]
+ we needed an rc4 for 4.1.0 (predictably), so herded the necessary
fixes into it
+ bug fixing:
- LP:1838913 : single-step exceptions were always being taken to EL1,
even if the guest configured debug exceptions to go to EL2 or EL3
- LP:1815423 : x86 TCG bug where we were giving the wrong result for
SSE float-to-int conversions that raised the 'invalid' exception
- LP:1796520 and LP:1839325 : started investigating some sh4 linux-user bugs
+ code review:
- patch to implement aspeed SD controller model
- Eric Auger's patchset of minor SMMUv3 fixes
- Aspeed GPIO controller model
- v3 of the GreenSocs reset-handling refactoring
- RTH's preliminary cleanup of handling of PC in arm decoder
thanks
-- PMM
[VIRT-263 # ARMv8.1-VHE Virtual Host Extensions ]
Posted a couple of revisions. Good feedback between this
and MemTag, both of which need to adjust the set of TLBs.
[VIRT-339 # ARMv8.5-BTI, Branch Target Identification ]
Posted v7.
[VIRT-344 # ARMv8.5-MemTag, Memory Tagging Extension ]
Rebased upon current VHE+BTI work. Updated from beta manuals
to the ARM ARM issue E.a manual.
[VIRT-327 # Richard's upstream QEMU work ]
Reviewed v1 x86 gen_sse rewrite.
Reviewed Alex's v4 plugin patchset.
r~
o LLVM:
* Bots babysitting
* Machine Outliner:
- Rebased branch on upstream
- Working on stack fixup limits test cases
o Misc
* Various meetings and discussions.
Progress:
* VIRT-65 [QEMU upstream maintainership]
+ QEMU 4.1.0 rc3 sent out. As usual, we've found some last minute
bugs and we'll need an rc4 next week...
+ patch review:
- series from Philippe making some cleanups to use object_initialize_child()
and friends
- Alex's patch "generate a custom MIDR for -cpu max"
- made a start on reviewing RTH's 32-bit arm decodetree conversion
+ bug fixing:
- LP:1838277 : we were taking exceptions caused by BRK instructions at
EL2 to EL1, rather than to EL2; sent a patch fixing this
- LP:1838475 : FPU register stacking in M-profile CPUs without the
Security Extensions was incorrectly failing an NSACR check and
taking a bogus exception: sent a patch to fix
+ wrote a patchset that converts the sparc target away from the
deprecated and broken do_unassigned_access hook to use the new-in-2017
do_transaction_failure hook instead; one step closer to being able
to remove the old hook entirely...
+ did the same for the mips target
thanks
-- PMM
== This Week ==
* PR86753
- Working on approach suggested by Richard.
* PR90724
- Pinged upstream.
* Validation
- Merged patch to jenkins-scripts to add testsuite comparison to tcwg_gnu.
- Submitted patch for separating build and test steps in tcwg_gnu.
== Next Week ==
- Continue ongoing tasks.
== Progress ==
* LLVM 9.0.0 rc1 binaries uploaded
- Ran into one cross-platform issue (x86, AArch64, ARM etc)
- Opened a bug for libfuzzer tests on AArch64
* Use ninja in release job [LLVM-536]
- Done
* Investigate running benchmarks in containers [TCWG-1513]
- Seems to work, except for mcf which errors out; will investigate next week
* IR SVE reviews [LLVM-545]
- Another round of feedback to D53137
== Plan ==
* Review D47770
* Upcoming vacation: 6 - 13 August
* Friday off
== Progress ==
* GCC:
- FDPIC: started taking feedback into account. Changes under test.
- noinit attribute: posted a new version
* GNU-583 (Fix Linux kernel built for Thumb-2 with GCC using LTO)
- no progress this week
* GCC upstream validation:
- no new issue this week.
* Binutils:
- Non-contiguous memory regions support in the BFD linker: not started
yet. Received a few "warnings" as feedback.
* misc:
- infra fixes / troubleshooting
== Next ==
GCC:
- handle feedback on FDPIC and noinit patches
- binutils/linker support for non-contiguous memory regions
- GNU-583
- GCC upstream validation: Add a config for cortex-m33 (v8-m)
== Holidays ==
Aug 2-11 (next week)
[PR42719] BasicAA UnitTest failure on AArch64 when compiled with GCC
- Patch submitted upstream
[PR42853] LLD support for TPREL_G0 relocations
- In theory a simple change, but llvm-mc seems to be deliberately
doing something strange with MOVZ encodings fixMOVZ and I haven't
worked out why it is needed.
Investigations with respect to inline assembler constraints.
- Some patch review comments for my ABE (Linaro GCC build system)
patch for GCC RM multilibs
- LLD review for smaller AArch64 images.
- Wrote a summary of what TCWG do for the upstream community, may turn
into a Linaro blogpost
- Raised GCC PR91299 to cover incorrect weak definition inlining in
GCC LTO in presence of non-weak definition.
On holiday Friday 2nd August, back in the office on Monday.
[VIRT-263 # ARMv8.1-VHE Virtual Host Extensions ]
Merged some fixes from Alex.
Starting to test kvm-inside-tcg.
[VIRT-327 # Richard's upstream QEMU work ]
Posted v1 of aa32 base isa conversion to decodetree.
Reviewed v5 of the sparc64 invert endian tlb bit patches.
r~
[ACTIVITY] On buildbot and Linux kernel CI duty
- One obscure alias analysis bug that only failed on GCC on AArch64
that suddenly started failing a unit test. Tracked down to an
unitialised variable when API called directly via unit-test. Test has
been reverted, but I've put enough information in a PR for an AA
expert to either fix the test or initialise class member pointer to
nullptr.
- A long review of an LLD patch that changes page-alignment rules and
this has impact on the TLS address. Spent way too long trying to
reverse engineer why the formula used would work, needed to find the
corresponding loader code to find the other side.
- Assertion added to LLVM triggered a failure in Linux kernel build,
reproduced example and forwarded on
- Code size investigation Linaro Connect proposal accepted, now have
to write it.
Progress:
* VIRT-65 [QEMU upstream maintainership]
+ QEMU 4.1 rc2 tagged
+ investigated and sent patches for a problem reported by Mark Rutland
where we still were sometimes putting the initrd on top of the kernel
in our built-in bootloader code
+ spent a couple of days tracking down the cause of LP:1696773, an
intermittent bug when running AArch64 Go programs under QEMU linux-user
mode. This turned out to be an incorrect implementation of 'sigaltstack'
as setting a process-wide signal stack rather than a per-thread one.
+ prompted by a patch from Greensocs fixing a vmstate migration bug
in the pl330 model, improved the vmstate macros to catch that category
of bug; this detected one other bug lurking in a different device.
thanks
-- PMM
== Progress ==
* LLVM 8.0.1 final binaries uploaded
* Use ninja in release job [LLVM-536]
- Patches ready, but waiting for 9.0.0-rc1 so I can test (was
supposed to come out this week but got postponed to Monday)
* Investigate running benchmarks in containers [TCWG-1513]
- Still having trouble, but working on it
* IR SVE reviews [LLVM-545]
- Posted some feedback to D53137
* A bit of fuzzing of GlobalISel CallLowering for AArch32
== Plan ==
* Upcoming vacation: 6 - 13 August
[LLVM-478] Clang and GCC code size comparison
- Looked at CMSIS-DSP, clang consistently produces smaller code than
GCC in contrast to Zephyr. Recorded some specific areas
- Tidied up scripts and build system patches to post for review
- Wrote up a readme for how to reproduce the results
[Misc]
- LLD reviews for reducing image-size using some VA/PA address alignment tricks.
- MC and LLD changes to add MOV[WNZ] relocations.
- Looked into why our container might be giving Resource unavailable
during ninja check-all
Some thoughts on ABI process.
Progress:
* VIRT-65 [QEMU upstream maintainership]
+ QEMU 4.1 rc1 tagged
+ investigating various bugs:
- LP:1836501 (assert using vexpress-a15 board with KVM) : couldn't
reproduce; this reminded me I needed to do a debian upgrade on my
cubietruck devboard, though.
- LP:1830864 (assert using -cpu host,aarch64=off with KVM on a host
kernel older than 4.15): diagnosed problem and sent a patch to fix it
- sent a patch to fix a URL in our configure script which pointed
users at the binary downloads page of the project website when it
wanted the source downloads page
- sent a patch fixing a problem building elf2dmp when libcurl's header
files are in a non-standard location
thanks
-- PMM
== Progress ==
* LLVM 8.0.1 rc4 binaries uploaded
* Buildbot and kernel builds monitoring
- Investigated/reported/fixed a couple of issues
- Tried to reproduce a clang-native-arm-lnt-perf failure that's been
keeping the bot red since the 3rd of July, but it turns out to be very
tricky
* Investigate running benchmarks in containers [TCWG-1513]
- Got quite far with this, but got a bit stuck with a file that
cannot be accessed for reasons that are not clear to me yet
* IR SVE reviews [LLVM-545]
- Looking at Graham's rebased size queries patch
== Plan ==
* Upcoming vacation: 6 - 13 August
