Fix GH-14506: Closing a userspace stream inside a userspace handler causes heap corruption #18797
Conversation
|
Test failures were unrelated and due to an issue at the time of branching on the master branch. Rebased for a CI run again now. |
|
Rebased because the userspace code changed in between opening this PR and now. |
|
You don't need to ping RM's for bug fix... Why is this master only? I can see the memory corruption and this seems like somthing that should not break real code if I'm not missing anything? |
We don't always apply those engine protections on lower branches if we can't fully gauge the consequences. |
|
Yeah I guess there is a small chance that some code could relay on stream being closed in this way so probably master only is better. |
If really necessary, it is always possible to backport in the future anyway once this code has sat into master long enough. |
…r causes heap corruption Use the PHP_STREAM_FLAG_NO_FCLOSE flag to prevent closing a stream while a handler is running. We already do this in some other places as well. Only handlers that do something with the stream afterwards need changes.
3 participants
Use the PHP_STREAM_FLAG_NO_FCLOSE flag to prevent closing a stream while a handler is running. We already do this in some other places as well. Only handlers that do something with the stream afterwards need changes.