[Lldb-commits] [PATCH] D87637: [lldb/test] Enable faulthandler in dotest

2020-11-16 Thread Jordan Rupprecht via Phabricator via lldb-commits
rupprecht added a comment. In D87637#2395603 , @teemperor wrote: > Yeah I don't think this revision is to blame as just disabling the code here > doesn't bring back the backtraces. Doesn't explain what happened to them, but > that's probably on me to fig

[Lldb-commits] [PATCH] D87637: [lldb/test] Enable faulthandler in dotest

2020-11-14 Thread Raphael Isemann via Phabricator via lldb-commits
teemperor accepted this revision. teemperor added a comment. This revision is now accepted and ready to land. Yeah I don't think this revision is to blame as just disabling the code here doesn't bring back the backtraces. Doesn't explain what happened to them, but that's probably on me to figure

[Lldb-commits] [PATCH] D87637: [lldb/test] Enable faulthandler in dotest

2020-11-13 Thread Jordan Rupprecht via Phabricator via lldb-commits
rupprecht added a comment. My cmake config appears to be correct. I tried adding this to `CommandInterpreter::HandleCommand`: lldbassert(ENABLE_BACKTRACES && "back traces are not enabled!"); lldbassert(false && "crash!"); Running LLDB directly does what you describe: $ bin/lldb (lldb)

[Lldb-commits] [PATCH] D87637: [lldb/test] Enable faulthandler in dotest

2020-11-13 Thread Jordan Rupprecht via Phabricator via lldb-commits
rupprecht added a comment. At head: $ ninja check-lldb-api-commands-apropos-basic [2/3] Running lit suite /home/rupprecht/src/llvm-project/lldb/test/API/commands/apropos/basic FAIL: lldb-api :: commands/apropos/basic/TestApropos.py (1 of 1) TEST 'lldb-api :: commands

[Lldb-commits] [PATCH] D87637: [lldb/test] Enable faulthandler in dotest

2020-11-13 Thread Jordan Rupprecht via Phabricator via lldb-commits
rupprecht added a comment. Thanks, I'll take a look at that. For SIGTERM (issued by test runners to handle timeouts), the stack trace is printed via `faulthandler.register(signal.SIGTERM, chain=True)`. The `chain=True` //should// cause previous signal handlers registered for SIGTERM to execute

[Lldb-commits] [PATCH] D87637: [lldb/test] Enable faulthandler in dotest

2020-11-13 Thread Raphael Isemann via Phabricator via lldb-commits
teemperor added a comment. +1 to what Jonas said. You can reproduce this by just adding an `abort();` call at the start of `CommandInterpreter::HandleCommand` and then for example like the `TestApropos.py` test. > If the LLDB process that this calls crashes, it shouldn't affect that. LLDB is

[Lldb-commits] [PATCH] D87637: [lldb/test] Enable faulthandler in dotest

2020-11-13 Thread Jonas Devlieghere via Phabricator via lldb-commits
JDevlieghere added a comment. I think Raphael is referring to the LLVM's `PrettyStackTrace` which prints a backtrace when you crash. Other than the driver I don't think we enable that in LLDB, so maybe LLVM was registering them by default if you had no signal handler installed? By default the

[Lldb-commits] [PATCH] D87637: [lldb/test] Enable faulthandler in dotest

2020-11-13 Thread Jordan Rupprecht via Phabricator via lldb-commits
rupprecht added a comment. In D87637#2395246 , @teemperor wrote: > I'm late to the party, but I actually don't think this is a good change as it > disables the normal LLDB backtraces. The current test errors we see on Green > Dragon look now like this: >

[Lldb-commits] [PATCH] D87637: [lldb/test] Enable faulthandler in dotest

2020-11-13 Thread Raphael Isemann via Phabricator via lldb-commits
teemperor added a comment. I'm late to the party, but I actually don't think this is a good change as it disables the normal LLDB backtraces. The current test errors we see on Green Dragon look now like this: Assertion failed: (size() >= N && "Dropping more elements than exist"), function dr

[Lldb-commits] [PATCH] D87637: [lldb/test] Enable faulthandler in dotest

2020-09-16 Thread Jordan Rupprecht via Phabricator via lldb-commits
This revision was automatically updated to reflect the committed changes. Closed by commit rGdd67581407c1: [lldb/test] Enable faulthandler in dotest (authored by rupprecht). Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D87637/new/ https://reviews.ll

[Lldb-commits] [PATCH] D87637: [lldb/test] Enable faulthandler in dotest

2020-09-16 Thread Jonas Devlieghere via Phabricator via lldb-commits
JDevlieghere accepted this revision. JDevlieghere added a comment. This revision is now accepted and ready to land. I think this will be very useful. LGTM! Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D87637/new/ https://reviews.llvm.org/D87637 __

[Lldb-commits] [PATCH] D87637: [lldb/test] Enable faulthandler in dotest

2020-09-14 Thread Jordan Rupprecht via Phabricator via lldb-commits
rupprecht created this revision. rupprecht added a reviewer: JDevlieghere. Herald added a project: LLDB. Herald added a subscriber: lldb-commits. rupprecht requested review of this revision. Register the `faulthandler` module so we can see what lldb tests are doing when they misbehave (e.g. run u