[llvm-dev] Pass is not initialized

[Mike Aizatsky via llvm-dev]

I need some help figuring out existing pass problem.  I've submitted a CL
on Friday (http://reviews.llvm.org/rL262103) that broke _only_ armv7 &
aarch64 bots: "Pass 'SanitizerCoverageModule' is not initialized." (it was
rolled back).

Today I've obtained access to aarch64 machine today, applied the patch and
check-all passes without any issues (terminal capture at the end).

Could there be something wrong with these bots? Do you have any idea what
to try?

=== Build Failure ===

It was failing with this error (
http://lab.llvm.org:8011/builders/clang-cmake-aarch64-full/builds/1493/steps/ninja%20check%201/logs/stdio
):

Pass 'SanitizerCoverageModule' is not initialized.
Verify if there is a pass dependency cycle.
Required Passes:
clang-3.9:
/home/renato.golin/buildslave/clang-cmake-aarch64-full/llvm/lib/IR/LegacyPassManager.cpp:668:
void llvm::PMTopLevelManager::schedulePass(llvm::Pass*): Assertion `PI &&
"Expected required passes to be initialized"' failed.
#0 0x00000000014de8a8 llvm::sys::PrintStackTrace(llvm::raw_ostream&)
(/home/renato.golin/buildslave/clang-cmake-aarch64-full/stage1/bin/clang-3.9+0x14de8a8)
Stack dump:
0. Program arguments:
/home/renato.golin/buildslave/clang-cmake-aarch64-full/stage1/bin/clang-3.9
-cc1 -triple aarch64-unknown-linux-gnu -emit-obj -mrelax-all -disable-free
-main-file-name coverage-caller-callee.cc -mrelocation-model static
-mthread-model posix -mdisable-fp-elim -mdisable-tail-calls -fmath-errno
-masm-verbose -mconstructor-aliases -munwind-tables -fuse-init-array
-target-cpu generic -target-feature +neon -target-abi aapcs
-dwarf-column-info -debug-info-kind=line-tables-only -dwarf-version=4
-debugger-tuning=gdb -resource-dir
/home/renato.golin/buildslave/clang-cmake-aarch64-full/stage1/bin/../lib/clang/3.9.0
-internal-isystem
/usr/lib/gcc/aarch64-linux-gnu/4.9/../../../../include/c++/4.9
-internal-isystem
/usr/lib/gcc/aarch64-linux-gnu/4.9/../../../../include/aarch64-linux-gnu/c++/4.9
-internal-isystem
/usr/lib/gcc/aarch64-linux-gnu/4.9/../../../../include/aarch64-linux-gnu/c++/4.9
-internal-isystem
/usr/lib/gcc/aarch64-linux-gnu/4.9/../../../../include/c++/4.9/backward
-internal-isystem /usr/local/include -internal-isystem
/home/renato.golin/buildslave/clang-cmake-aarch64-full/stage1/bin/../lib/clang/3.9.0/include
-internal-externc-isystem /usr/include/aarch64-linux-gnu
-internal-externc-isystem /include -internal-externc-isystem /usr/include
-fdeprecated-macro -fdebug-compilation-dir
/home/renato.golin/buildslave/clang-cmake-aarch64-full/stage1/projects/compiler-rt/test/asan/AARCH64LinuxConfig/TestCases/Posix
-ferror-limit 19 -fmessage-length 0 -fsanitize-coverage-type=3
-fsanitize-coverage-indirect-calls -fsanitize=address
-fsanitize-blacklist=/home/renato.golin/buildslave/clang-cmake-aarch64-full/stage1/bin/../lib/clang/3.9.0/asan_blacklist.txt
-fno-assume-sane-operator-new -fallow-half-arguments-and-returns
-fno-signed-char -fobjc-runtime=gcc -fcxx-exceptions -fexceptions
-fdiagnostics-show-option -o /tmp/coverage-caller-callee-72b4a0.o -x c++
/home/renato.golin/buildslave/clang-cmake-aarch64-full/llvm/projects/compiler-rt/test/asan/TestCases/Posix/coverage-caller-callee.cc
1. <eof> parser at end of file
clang-3.9: error: unable to execute command: Aborted
clang-3.9: error: clang frontend command failed due to signal (use -v to
see invocation)
clang version 3.9.0 (trunk 262106)
Target: aarch64-unknown-linux-gnu
Thread model: posix
InstalledDir:
/home/renato.golin/buildslave/clang-cmake-aarch64-full/stage1/./bin
clang-3.9: note: diagnostic msg: PLEASE submit a bug report to
http://llvm.org/bugs/ and include the crash backtrace, preprocessed source,
and associated run script.
clang-3.9: note: diagnostic msg:


=== Successful Run ===

% uname -a
Linux <host_name> 3.13.0-45-generic #74-Ubuntu SMP Tue Jan 13 19:38:11 UTC
2015 aarch64 aarch64 aarch64 GNU/Linux

% ninja check-llvm
[24/24] Running the LLVM regression tests
Testing Time: 110.50s
  Expected Passes    : 15324
  Expected Failures  : 127
  Unsupported Tests  : 511
keplerarm1% ninja check-all
[1/1] Running all regression tests
Testing Time: 109.94s
  Expected Passes    : 15324
  Expected Failures  : 127
  Unsupported Tests  : 511
% git status
On branch master
Your branch is up-to-date with 'origin/master'.

Changes not staged for commit:
  (use "git add <file>..." to update what will be committed)
  (use "git checkout -- <file>..." to discard changes in working directory)

modified:   lib/Transforms/Instrumentation/SanitizerCoverage.cpp





On Sat, Feb 27, 2016 at 9:00 AM Mike Aizatsky <aizatsky at chromium.org> wrote:

> Renato,
>
> I did. And I thought I fixed it in http://reviews.llvm.org/rL262111 (as soon
> as I got breakage notification).
>
> There's something wrong with those architectures. There are no problems on
> others. I'll take better look on Monday.
>
> On Sat, Feb 27, 2016 at 6:25 AM Renato Golin <renato.golin at linaro.org>
> wrote:
>
>> On 27 February 2016 at 02:10, Mike Aizatsky via llvm-commits
>> <llvm-commits at lists.llvm.org> wrote:
>> > Author: aizatsky
>> > Date: Fri Feb 26 20:10:27 2016
>> > New Revision: 262103
>> >
>> > URL: http://llvm.org/viewvc/llvm-project?rev=262103&view=rev
>> > Log:
>> > [sancov] do not instrument nodes that are full pre-dominators
>> >
>> > Summary:
>> > Without tree pruning clang has 2,667,552 points.
>> > Wiht only dominators pruning: 1,515,586.
>> > With both dominators & predominators pruning: 1,340,534.
>> >
>> > Differential Revision: http://reviews.llvm.org/D17671
>>
>> I've just reverted this patch in r262139, as it broke all ARM and
>> AArch64 buildbots. Please, pay attention to the bots and contact the
>> owner or revert on the first sign of trouble.
>>
>> cheers,
>> --renato
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20160301/214b57b5/attachment-0001.html>