[Backtracing] Add an indication that we're working on a backtrace. #69775
Conversation
When we crash, emit a message straight away that says we're working on a backtrace. If starting the backtracer fails, report that also. Finally, add a duration to the messages output by the backtracer, so that we can see how long it took. rdar://118055527
We shouldn't try to erase the message when in non-color mode; mostly that's intended for redirected output scenarios anyway, and in that case we don't really want garbage in the output. Also, improve the failure messages when the backtracer itself goes wrong or can't be executed. Finally, change the behaviour slightly so that *if* we're explicitly enabled, *and* the backtracer path wasn't explicitly specified, *and* we can't find the backtracer, we print a warning on start-up. We don't do that in any other case because we don't want spurious warnings everywhere. rdar://118055527
Add tests for the new immediate on-crash message and the backtrace timing message. rdar://118055527
al45tair
added
🍒 release cherry pick
|
Explanation: Generating backtraces can take a significant amount of time, particularly for statically linked binaries on Linux, because it involves scanning through all of the debug information, which means parsing DWARF data, which is just slow. This is particularly problematic in Kubernetes, because pods often have health check timeouts set such that Kubernetes will kill the pod if it doesn't respond within a certain time; if a container crashes, it won't be responding to the health checks, and if the backtrace takes longer than the timeout, it'll be killed before it completes. Previously we output the entire backtrace at the end, so in that case you'd get no useful information about the crash. This PR changes things so that (a) we output the crash location itself straight-away along with a message indicating that we're backtracing, and (b) we say at the end how long it took, to give people something they can use to tell if their timeouts are close to the time it'll take to generate a trace. |
|
@swift-ci Please test |
|
@swift-ci Please test |
Apparently I forgot to code sign the test binaries. rdar://118055527
al45tair
force-pushed
the
eng/PR-118055527-5.10
branch
from
75512a1 to
e068482
Compare
|
@swift-ci Please test |
When we crash, emit a message straight away that says we're working on a backtrace. If starting the backtracer fails, report that also.
Finally, add a duration to the messages output by the backtracer, so that we can see how long it took.
rdar://118055527