You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
For some bugs, like TSan or ASan Nightly issues, we usually don't have a bisection (esp. for issues that have been in the codebase for a long time or where we don't have a standalone testcase but a stack.
Using this stack, it would be nice if we could automatically trigger an initial needinfo? based on the first relevant frame. The challenges here are
Finding the first non-generic frame. Sometimes there are a few generic frames on top of the stack that aren't relevant. My gut feeling tells me that we can find heuristics to work around this, e.g. pick the first frame that belongs to a .c|.cpp|.rs file (not .h) that is in our codebase.
If the hg annotate turns up a developer no longer working at Mozilla (which is common for old code), we might have to check the surroundings for other developers that might be able to help.
We have to deal with reformat commits (I believe we have a hg ext to do this already?)
The text was updated successfully, but these errors were encountered:
For some bugs, like TSan or ASan Nightly issues, we usually don't have a bisection (esp. for issues that have been in the codebase for a long time or where we don't have a standalone testcase but a stack.
Using this stack, it would be nice if we could automatically trigger an initial needinfo? based on the first relevant frame. The challenges here are
Finding the first non-generic frame. Sometimes there are a few generic frames on top of the stack that aren't relevant. My gut feeling tells me that we can find heuristics to work around this, e.g. pick the first frame that belongs to a
.c|.cpp|.rs
file (not.h
) that is in our codebase.If the hg annotate turns up a developer no longer working at Mozilla (which is common for old code), we might have to check the surroundings for other developers that might be able to help.
We have to deal with reformat commits (I believe we have a hg ext to do this already?)
The text was updated successfully, but these errors were encountered: