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
There's an issue with fibers unrelated to JIT or opcache that causes memory corruption in certain cases if a fiber is resumed in the shutdown handler after an uncaught exception triggered execution of said shutdown handler.
This memory corruption usually manifests as impossible TypeErrors when returning from random functions.
The only way I found of sometimes reproducing the issue is running the test.php script from https://paste.daniil.it/jit_1.tar.xz (without JIT or opcache) and randomly hitting ctrl-c multiple times during the handshake, but still it's super hard to reproduce reliably...
Description
There's an issue with fibers unrelated to JIT or opcache that causes memory corruption in certain cases if a fiber is resumed in the shutdown handler after an uncaught exception triggered execution of said shutdown handler.
This memory corruption usually manifests as impossible TypeErrors when returning from random functions.
I've been seeing a lot of reports including in the amphp chat (join https://t.me/+RO_PpIt0DDZhvPPs then https://t.me/c/1156566948/47988, https://t.me/c/1156566948/47730, https://t.me/c/1156566948/47048), and got the issue myself several times, but haven't managed to reliably reproduce yet.
The only way I found of sometimes reproducing the issue is running the test.php script from https://paste.daniil.it/jit_1.tar.xz (without JIT or opcache) and randomly hitting ctrl-c multiple times during the handshake, but still it's super hard to reproduce reliably...
The issue itself is triggered by MadelineProto's shutdown handler, which resumes the revoltphp event loop (with a bit of reflection trickery) and continues execution of all fibers currently in scope, also spawning some new ones: https://github.com/danog/MadelineProto/blob/v8/src/Shutdown.php#L55
PHP Version
8.2.12
Operating System
No response
The text was updated successfully, but these errors were encountered: