-
-
Notifications
You must be signed in to change notification settings - Fork 30.5k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
gh-117657: Make frame clearing thread-safe #120542
base: main
Are you sure you want to change the base?
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
What is the test case that triggered the TSAN error?
Objects/genobject.c
Outdated
@@ -158,6 +159,15 @@ gen_dealloc(PyGenObject *gen) | |||
PyObject_GC_Del(gen); | |||
} | |||
|
|||
static void | |||
gen_dealloc(PyGenObject *gen) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
This doesn't seem right to me. tp_dealloc
can't be called concurrently on an object.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
In this case the racing part is mainly _PyFrame_ClearExceptCode
and setting the generator's flags. _PyEval_FrameClearAndPop
can race with this as it also clears a generator, but instead it's done by ceval.c
.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I don't understand -- how can _PyFrame_ClearExceptCode
race with gen_dealloc
?
It looks like gen_dealloc
is only called from the tp_dealloc
handler. If an object is being destroyed, then no other thread has access to it (or there's a separate serious bug).
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Hmm I'm not too sure either. Let me revert this fix for now and try to hunt it down again.
I first saw this getting triggered spuriously on our main test suite (I can't find the test now). But the TSAN logs I posted here are for the test case I wrote in this PR. |
Data race reported by TSAN on the test: