[VIRT-198 # QEMU: SVE Emulation Support ]
All softfp prerequisites merged upstream.
First 25 of 70 patches of SVE decode merged upstream.
[Upstream]
Review and testing of Alex's check-tcg revival.
[STOP]
r~
== Progress ==
o GNU release transition
* Provided support to handle infra issues
* Meetings on GCC 8
o LLVM
* Benchmarking effort with HPC team [HPC-183/184]
- Added OpenMP support to LLVM scripts and jobs
- Generate toolchain tarballs as job artifact.
* Thumbv8 bot failures [TCWG-1400]:
- Needs a deeper cleanup of LIT configuration.
o Misc
* Various meetings and discussions.
== This Week ==
* GCC bugs (8/10)
a) Committed patches for:
i) TCWG-1413 / PR83648: missing -Wsuggest-attribute=malloc on a
trivial malloc-like function
ii) TCWG-1414 / PR85817: ICE in expand_call at gcc/calls.c:4291
iii) TCWG-1415 / PR85734: --suggest-attribute=malloc misdiagnoses
static functions
b) Work in progress for:
i) TCWG-125: static cast from float to int not working on ARM hardfp
ii) TCWG-1416 / PR85787: malloc_candidate_p fails to detect malloc
attribute on nested phis
* TCWG-1234: Code hoisting and register pressure (1/10)
i) Looking into Bin Cheng's patches for computing register pressure on GIMPLE
* Misc (1/10)
- Meetings
- Gerrit reviews 25367 and 25371
== Next Week ==
- Continue with GCC bugs
- TCWG-1234
SVE Support ([VIRT-198])
========================
RISU Support for SVE ([VIRT-199])
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
- get back to merging/preparing the SVE series :todo
[VIRT-199] https://projects.linaro.org/browse/VIRT-199
SVE Reviews
~~~~~~~~~~~
- reviewed {PATCH 0/9} target/arm: Fixups for ARM_FEATURE_V8_FP16
Message-Id: <20180425012300.14698-1-richard.henderson(a)linaro.org>
SoftFloat Bugs ([VIRT-69])
==========================
- CLOSING NOTE [2018-05-18 Fri 16:41]
I think all known bugs have fixes now merged.
[VIRT-69] https://projects.linaro.org/browse/VIRT-69
Upstream Work ([VIRT-109])
==========================
- catch-up on review queue :todo
- posted {PATCH v4 00/49} fix building of tests/tcg Message-Id:
<20180517174718.10107-1-alex.bennee(a)linaro.org>
- spin v5 and get merged :todo
- posted {PATCH v2 0/2} support reading of CNT{VCT|FRQ}_EL0 from
user-space Message-Id:
<20180518114424.18054-1-alex.bennee(a)linaro.org>
- posted {RFC PATCH 0/2} Travis Stability Patches Message-Id:
<20180518091440.1559-1-alex.bennee(a)linaro.org>
[VIRT-109] https://projects.linaro.org/browse/VIRT-109
[a gdb bug] https://sourceware.org/bugzilla/show_bug.cgi?id=23127
Fixing up tests/tcg
~~~~~~~~~~~~~~~~~~~
- posted {PATCH v4 00/49} fix building of tests/tcg Message-Id:
<20180517174718.10107-1-alex.bennee(a)linaro.org>
KVM CI Loop ([VIRT-2])
======================
- need to sync-up on the current state of this work :todo
[VIRT-2] https://projects.linaro.org/browse/VIRT-2
[our wiki] https://wiki.linaro.org/Core/Virtualization
Other Tasks
===========
- Investigate weird QEMU CI OBS build issue :todo
- the OBS build should be failing the RISU tests but isn't, despite
the commit id it reports
- bringing up my SyncQuacer DevBox with Gentoo
- needed one patch for Nouveau w.r.t 32/64 bit PCIE lanes
- needed to manually patch LVM to build initrd Message-Id:
<20180516201903.24309-1-alex.bennee(a)linaro.org>
Completed Reviews [7/7]
=======================
{PATCH v6 0/3} target/arm: Add a dynamic XML-description of the cp-registers to GDB
Message-Id: <1524153386-3550-1-git-send-email-abdallah.bouassida(a)lauterbach.com>
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
- CLOSING NOTE tested with new gdb - some weirdness on timer register
{PATCH 00/19} softfloat: Clean up NaN handling
Message-Id: <20180511004345.26708-1-richard.henderson(a)linaro.org>
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
- CLOSING NOTE [2018-05-11 Fri 20:55]
Found a few minor bugs, rth re-spinning with my float-to-float stuff
{PATCH v4 00/11} target/arm: Fixups for ARM_FEATURE_V8_FP16
Message-Id: <20180512003217.9105-1-richard.henderson(a)linaro.org>
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
- CLOSING NOTE [2018-05-15 Tue 11:42]
Despite initial confusion on branch tested fine on both, all good.
{Qemu-devel} {PATCH 0/9} target/arm: Implement v8.1-Atomics
Message-Id: <20180427002651.28356-1-richard.henderson(a)linaro.org>
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
- CLOSING NOTE [2018-05-15 Tue 11:43]
Already merged, others have reviewed
{PATCH v2 00/27} softfloat patch roundup
Message-Id: <20180512004311.9299-1-richard.henderson(a)linaro.org>
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
- CLOSING NOTE [2018-05-15 Tue 11:45]
v5 released
{PATCH v5 00/28} softfloat patch roundup
Message-Id: <20180514221219.7091-1-richard.henderson(a)linaro.org>
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
- CLOSING NOTE [2018-05-15 Tue 15:10]
Looking good.
{PATCH v6 00/28} softfloat patch roundup
Message-Id: <20180515222540.9988-1-richard.henderson(a)linaro.org>
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
- CLOSING NOTE [2018-05-18 Fri 16:16]
Merged upstream
Absences
========
- CBG Beer Festival (half day sometime next week)
- VNC18 Connect (17th-21st September 2018)
- KVM Forum 2018 (24th-26th October 2018)
Current Review Queue
====================
* {Qemu-arm} {PATCH v3-a 00/27} target/arm: Scalable Vector Extension
Message-Id: <20180516223007.10256-1-richard.henderson(a)linaro.org>
* {Qemu-devel} {PATCH} gdbstub: Clarify what gdb_handlesig() is doing
Message-Id: <20180515181958.25837-1-peter.maydell(a)linaro.org>
* {Qemu-devel} {PATCH 0/9} Honor CPU_DUMP_FPU
Message-Id: <20180511035240.4016-1-richard.henderson(a)linaro.org>
* {Qemu-devel} {RFC PATCH v2 00/12} iommu: add MemTxAttrs argument to IOMMU translate function
Message-Id: <CAFEAcA_Dei48HtKk6GnEgXYcXdY2htCXej4pSfY+q9V6f=abdA(a)mail.gmail.com>
* {RFC PATCH 00/12} tests/tcg: Add TriCore tests
Message-Id: <20180501142222.19154-1-kbastian(a)mail.uni-paderborn.de>
* {PATCH v3 00/15} fp-test + hardfloat
Message-Id: <1522883475-27858-1-git-send-email-cota(a)braap.org>
--
Alex Bennée
[Activity]
[TCWG-1319] Link AOSP with LLD.
Closed this out as I've gone as far as I can profitably go without
being an Android expert.
- With trunk LLD I can now link and boot AOSP with LLD without any
use_clang_lld=false on both Arm and AArch64 emulators. I also managed
to get it to boot on a Hikey960 (aarch64).
- Communicated recent additions to LLD in aid of building AOSP, and
extra command line options needed to the Google team.
Landed support for GCC inline assembly constraint S in LLVM.
Some comments on LTO + linker script proposal on llvm-dev mailing list.
Started thinking in more detail about ICF=safe.
- Decided that LLVM was a better place to do this than clang after
attempting to use libtooling.
- There is an existing LLVM pass called merge functions that does what
ICF does but at the LLVM IR level
-- This is off by default, maybe worth looking at as part of LTO and
code-size reductions.
- Started to write a prototype pass that uses
Function::isAddressTaken() although I think that this may give me too
pessimistic a result.
Peter
Progress: [short week, 3 days]
* VIRT-65 [QEMU upstream maintainership]
+ code review
- make system registers visible via gdbstub
- Richard's patches to fix various softfloat issues
* VIRT-164 [improve Cortex-M emulation]
+ MPC emulation:
- rebased and fixed up conflicts with a different series that
touched the same parts of the memory system
- realised that passing txattrs to IOMMUs means we need to
change the notifier API; came up with a plan to handle this
by using "IOMMU indexes" analogous to our existing TCG MMU
indexes. Updated everything to the new API that comes out
of that approach.
- started on tidying up the final loose ends; hope to have
something I can send to the list early next week
thanks
-- PMM
o 3 days off
== Progress ==
o LLVM
* Thumbv8 bot failures:
- Reproduced the issue on TX1
- These test cases need to be disabled for ARM targets (patch on-going).
o Misc
* Various meetings and discussions.
[VIRT-243 # ARMv8.1-Atomic instructions ]
v3 posted and merged.
[Upstream]
(Upgraded my laptop to Fedora 28, which lead to...)
Posted a patch for gcc 8 Werrors.
Posted patches for clang 6 Werrors.
Sent pull for Emilio's TranslateOps patches.
Sent pull for my openrisc decodetree patches.
Sent pull for tcg-next.
Reviewed microblaze patches.
Posted two rounds of softfloat snan patches.
Posted two rounds of fp16 patches.
[VIRT-198 # QEMU: SVE Emulation Support ]
Most of this week's work was on indirectly related patch sets. In particular,
fp16 and atomics. The last round of check-gcc testing had
FAIL: gcc.dg/vect/slp-multitypes-2.c execution test
FAIL: gcc.target/aarch64/advsimd-intrinsics/vldX.c -O0 execution test
FAIL: gcc.target/aarch64/advsimd-intrinsics/vqtbX.c -O0 execution test
as the only remaining (relevant) failures. I believe that the latter two are
generic tcg failures that are cured by this week's tcg-next pull, but have not
yet re-based everything to be sure.
r~
SVE Support ([VIRT-198])
========================
RISU Support for SVE ([VIRT-199])
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
- regenerate more half-precision test cases
[VIRT-199] https://projects.linaro.org/browse/VIRT-199
SVE Reviews
~~~~~~~~~~~
- started reviewing {PATCH 0/9} target/arm: Fixups for
ARM_FEATURE_V8_FP16 Message-Id:
<20180425012300.14698-1-richard.henderson(a)linaro.org>
- fixed a couple of other problems testing picked up, see [this
branch]
[this branch] https://github.com/stsquad/qemu/tree/review/rth-fp16-fixes
SoftFloat Bugs ([VIRT-69])
==========================
- posted {PATCH v3 0/5} refactor float-to-float and fix AHP
Message-Id: <20180510094206.15354-1-alex.bennee(a)linaro.org>
- fixes up the current cases, AArch32 is lagging in later rev
support
- rth had a go at a canonical NaN solution
- reviewed {PATCH 00/19} softfloat: Clean up NaN handling Message-Id:
<20180511004345.26708-1-richard.henderson(a)linaro.org>
[VIRT-69] https://projects.linaro.org/browse/VIRT-69
Upstream Work ([VIRT-109])
==========================
- catch-up on review queue :todo
- reviewed {PATCH v6 0/3} target/arm: Add a dynamic XML-description of
the cp-registers to GDB Message-Id:
<1524153386-3550-1-git-send-email-abdallah.bouassida(a)lauterbach.com>
- finished testing - some wierdness with register values
[VIRT-109] https://projects.linaro.org/browse/VIRT-109
[a gdb bug] https://sourceware.org/bugzilla/show_bug.cgi?id=23127
Fixing up tests/tcg
~~~~~~~~~~~~~~~~~~~
- posted {PATCH v3 00/46} fix building of tests/tcg Message-Id:
<20180424152405.10304-1-alex.bennee(a)linaro.org>
- series looking pretty good, more comments incoming
- working on [v4 of the series]
[v4 of the series]
https://github.com/stsquad/qemu/tree/testing/tcg-tests-revival-v4
KVM CI Loop ([VIRT-2])
======================
- need to sync-up on the current state of this work :todo
[VIRT-2] https://projects.linaro.org/browse/VIRT-2
[our wiki] https://wiki.linaro.org/Core/Virtualization
Other Tasks
===========
- Investigate weird QEMU CI OBS build issue :todo
- the OBS build should be failing the RISU tests but isn't, despite
the commit id it reports
- bringing up my SyncQuacer, did some serial port debugging to debug
DIMMS
- turns out supported DIMMs are "4GB UDIMM and 16GB RDIMM"
- will need new firmware to get the 16GB UDIMMs working
[ARMv8.2-LPA support] https://projects.linaro.org/browse/TCWG-1395
[ARMv8.2-LVA] https://projects.linaro.org/browse/TCWG-1396
Completed Reviews [5/5]
=======================
{PATCH v6 0/3} target/arm: Add a dynamic XML-description of the cp-registers to GDB
Message-Id: <1524153386-3550-1-git-send-email-abdallah.bouassida(a)lauterbach.com>
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
- CLOSING NOTE tested with new gdb - some weirdness on timer register
{PATCH 00/19} softfloat: Clean up NaN handling
Message-Id: <20180511004345.26708-1-richard.henderson(a)linaro.org>
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
- CLOSING NOTE [2018-05-11 Fri 20:55]
Found a few minor bugs, rth re-spinning with my float-to-float stuff
Absences
========
- Bank Holiday (May 7th)
- VNC18 Connect (17th-21st September 2018)
- KVM Forum 2018 (24th-26th October 2018)
Current Review Queue
====================
* {Qemu-devel} {PATCH 0/9} target/arm: Implement v8.1-Atomics
Message-Id: <20180427002651.28356-1-richard.henderson(a)linaro.org>
* {RFC PATCH 00/12} tests/tcg: Add TriCore tests
Message-Id: <20180501142222.19154-1-kbastian(a)mail.uni-paderborn.de>
* {RFC PATCH v2 00/19} reverse debugging
Message-Id: <20180428123627.12445.9923.stgit@pasha-VirtualBox>
* {PATCH 0/2} arm64: Report signal frame size to userspace via auxv
Message-Id: <1525776211-28169-1-git-send-email-Dave.Martin(a)arm.com>
* {Qemu-devel} {PATCH v4 0/4} qemu-thread: support --enable-debug-mutex
Message-Id: <20180423053927.13715-1-peterx(a)redhat.com>
* {PATCH v3 00/15} fp-test + hardfloat
Message-Id: <1522883475-27858-1-git-send-email-cota(a)braap.org>
--
Alex Bennée
