Hi Richard,
Your patch a076632e274abe344ca7648b7c7f299273d4cbe0 appears to have broken bootstrap-O3 for 32-bit armhf. Do you have an AArch32-capable machine to reproduce/investigate this on? Let me know if not, and I'll make a proper bug report with a testcase.
ICE:
00:33:32 In function ‘syscall.forkExec’:
00:33:32 go1: error: address taken, but ADDRESSABLE bit not set
00:33:32 PHI argument
00:33:32 &go..C479;
00:33:32 for PHI node
00:33:32 err$__object_78 = PHI <err$__object_76(58), &go..C479(59)>
00:33:32 during GIMPLE pass: fre
00:33:32 go1: internal compiler error: verify_ssa failed
00:33:32 0x9c18d7 verify_ssa(bool, bool)
00:33:32 /home/tcwg-buildslave/workspace/tcwg_gnu_0/abe/snapshots/gcc.git~master/gcc/tree-ssa.c:1214
00:33:32 0x6f8d5b execute_function_todo
00:33:32 /home/tcwg-buildslave/workspace/tcwg_gnu_0/abe/snapshots/gcc.git~master/gcc/passes.c:2049
00:33:32 0x6f9abf do_per_function
00:33:32 /home/tcwg-buildslave/workspace/tcwg_gnu_0/abe/snapshots/gcc.git~master/gcc/passes.c:1687
00:33:32 0x6f9abf execute_todo
00:33:32 /home/tcwg-buildslave/workspace/tcwg_gnu_0/abe/snapshots/gcc.git~master/gcc/passes.c:2096
00:33:32 Please submit a full bug report,
00:33:32 with preprocessed source if appropriate.
00:33:32 Please include the complete backtrace with any bug report.
00:33:32 See <https://gcc.gnu.org/bugs/> for instructions.
00:33:32 Makefile:3001: recipe for target 'syscall.lo' failed
Full build log: https://ci.linaro.org/job/tcwg_gcc-bisect-gnu-master-arm-bootstrap_O3/16/ar…
Regards,
--
Maxim Kuvyrkov
https://www.linaro.org
> On May 11, 2021, at 8:15 AM, tcwg-jira (Jira) <projects(a)linaro.org> wrote:
>
> There is 1 comment.
>
>
> GNU Toolchain / <Mail Attachment.png> GNU-692 IN PROGRESS
> Regressions from tcwg_binutils/tcwg_cross/tcwg_gnu CI
>
> View issue · Add comment
>
> 1 comment
>
> tcwg-jira on 11/May/21 5:04 AM
>
> Successfully identified regression in gcc in CI configuration tcwg_gnu/gnu-master-arm-bootstrap_O3. So far, this commit has regressed CI configurations:
> • tcwg_gnu/gnu-master-arm-bootstrap_O3
> Culprit:
> <cut>
> commit a076632e274abe344ca7648b7c7f299273d4cbe0
> Author: Richard Biener <rguenther(a)suse.de>
> Date: Fri May 7 09:51:18 2021 +0200
> middle-end/100464 - avoid spurious TREE_ADDRESSABLE in folding debug stmts
> canonicalize_constructor_val was setting TREE_ADDRESSABLE on bases
> of ADDR_EXPRs but that's futile when we're dealing with CTOR values
> in debug stmts. This rips out the code which was added for Java
> and should have been an assertion when we didn't have debug stmts.
> To not regress g++.dg/tree-ssa/array-temp1.C we have to adjust the
> testcase to not look for a no longer applied invalid optimization.
> 2021-05-10 Richard Biener <rguenther(a)suse.de>
> PR middle-end/100464
> PR c++/100468
> gcc/
> • gimple-fold.c (canonicalize_constructor_val): Do not set
> TREE_ADDRESSABLE.
> gcc/cp/
> • call.c (set_up_extended_ref_temp): Mark the temporary
> addressable if the TARGET_EXPR was.
> gcc/testsuite/
> • gcc.dg/pr100464.c: New testcase.
> • g++.dg/tree-ssa/array-temp1.C: Adjust.
> </cut>
> Details: https://ci.linaro.org/job/tcwg_gcc-bisect-gnu-master-arm-bootstrap_O3/16/ar…
> Even more details: https://ci.linaro.org/job/tcwg_gcc-bisect-gnu-master-arm-bootstrap_O3/16/ar…
>
>
> This message was sent by Atlassian Jira (v8.11.1#811002-sha1:94cd716)
> Jira is improving email notifications, share your feedback!
> Get Jira notifications on your phone! Download the Jira Server app for Android or iOS.
== This Week ==
* GCC
- PR97903 (missed lowering to vtst): Committed fix to trunk,
- PR66791: Submitted patch to replace vtst* (a, b) by (a & b) != 0 in
arm_neon.h.
- Committed fix for build failure in trunk.
* Validation
- vect metric: Pushed all pending patches.
== Next Week ==
- PR66791: Continue working on more intrinsics
- Validation: Monitor vect metric
== Progress ==
* GCC upstream validation:
- Reported a few regressions
- discussing update of the list of configs
- tried qemu-6.0, issue with hwasan testing on aarch64
* GCC
- MVE/vectorization: waiting for feedback on patches for vcmp,
vld2/vst2, vld4/st4
* Misc
- scripts patch reviews
- looking at gdb issue with register names in target description
== Next ==
* MVE auto-vectorization/intrinsics improvements
* GCC/cortex-M testing improvements & fixes
Progress (short week, 2 days):
* UM-2 [QEMU upstream maintainership]
- sent some small patches: fixing Coverity issues; converting a
few users of a legacy API to the new function
* QEMU-406 [QEMU support for MVE (M-profile Vector Extension; Helium)]
- Implemented more insns: VMUL, VSTRB/H/W, VMAX, VMIN,
VABD, VHADD, VHSUB, VMULL, VRMULH. Progress meter: 36/210 (17%)
-- PMM
QEMU Device and Machine Models ([QEMU-418])
===========================================
- posted [kvm-unit-tests PATCH v1 0/4] enable LPI and ITS for TCG
Message-Id: <20210428101844.22656-1-alex.bennee(a)linaro.org>
[QEMU-418] <https://projects.linaro.org/browse/QEMU-418>
QEMU Upstream Work ([UM-2])
===========================
- updated [testing/next branch] we some more fixes
[UM-2] <https://projects.linaro.org/browse/UM-2>
[testing/next branch]
<https://github.com/stsquad/qemu/tree/testing/next>
Other
=====
- Spent more time on blog post for [LBO-99]
[LBO-99] <https://projects.linaro.org/browse/LBO-99>
Completed Reviews [3/3]
=======================
[PATCH v3 0/2] tests/docker: tests/tcg/ppc64le: Newer toolchain to build tests for PowerISA v3.1 instructions
Message-Id: <20210423205757.1752480-1-matheus.ferst(a)eldorado.org.br>
[PATCH 0/2] plugins: Freeing allocated values in hash tables.
Message-Id: <20210421140934.7561-1-ma.mandourr(a)gmail.com>
[PATCH v4 00/12] qtests: Check accelerator available at runtime via QMP 'query-accels'
Message-Id: <20210415163304.4120052-1-philmd(a)redhat.com>
[PATCH v2 0/8] GICv3 LPI and ITS feature implementation
Message-Id: <20210401024152.203896-1-shashi.mallela(a)linaro.org>
Absences
========
- Bank Holiday on Monday
Current Review Queue
====================
TODO [PATCH v4 00/12] qtests: Check accelerator available at runtime via QMP 'query-accels'
Message-Id: <20210415163304.4120052-1-philmd(a)redhat.com>
===================================================================================================================================================
TODO [RFC PATCH v2 0/6] hw/arm/virt: Introduce cpu topology support
Message-Id: <20210413080745.33004-1-wangyanan55(a)huawei.com>
==============================================================================================================================
TODO [PATCH v2 00/12] virtio-gpu: Add support for Blob resources feature
Message-Id: <20210420065347.2685768-1-vivek.kasireddy(a)intel.com>
========================================================================================================================================
TODO [RFC v14 00/80] arm cleanup experiment for kvm-only build
Message-Id: <20210416162824.25131-1-cfontana(a)suse.de>
===================================================================================================================
--
Alex Bennée
Progress:
* UM-2 [QEMU upstream maintainership]
- release work: rc5 Monday, and then 6.0 final release on Thursday.
- sent a pullreq with some accumulated arm patches now 6.1 is open
- fixed a trivial bug where WFI in the usermode emulator crashed
* QEMU-406 [QEMU support for MVE (M-profile Vector Extension; Helium)]
- Implemented more instructions: logic ops, VCLS, VCLZ, VREV, VMVN,
VABS, VNEG, VADD, VDUP, VSUB, VMULH, and widening VLDR.
(These are all simple "always call an out-of-line helper" versions.
We'll want an optimised "we know we have no predication" version
for many of them, but I plan to do that afterwards.)
- Worked through the v8M Arm ARM identifying which insns we must
implement. There are 210 insn patterns in total, of which I have
currently implemented 23, so we can produce a rough-and-ready
progress meter: 23/210 (11%)
(There is of course work to do beyond just the insns, but they
form the bulk of it, so the meter has some utility.)
-- PMM
