ext/standard: stream_filter_register() orphaned user_filter_map on sh… - #22838
Open
devnexen wants to merge 2 commits into
Open
ext/standard: stream_filter_register() orphaned user_filter_map on sh…#22838devnexen wants to merge 2 commits into
devnexen wants to merge 2 commits into
Conversation
…utdown re-registration. Fix phpGH-22818 During request shutdown the user_filter_map is torn down by the user_filters RSHUTDOWN before the streams referencing it are flushed. A user filter whose filter() callback re-registers the filter recreated the now-NULL map, then, since the volatile factory was still present in FG(stream_filters), deleted the freshly added entry and left an empty orphaned map behind. The following stream_filter_append() located the factory but no matching fdat, tripping ZEND_ASSERT(fdat), and the recreated map leaked. Register the volatile factory first and only create and populate user_filter_map on success, so a re-registration during the shutdown window fails without recreating the map. The existing NULL-map guard in user_filter_factory_create() then handles the append gracefully.
Member
|
Asking Nora to have a look as she fixed a bug around this code recently. |
Sjord
reviewed
Jul 30, 2026
| zend_string_release_ex(classname, 0); | ||
| efree(fdat); | ||
| } | ||
| zend_hash_add_ptr(BG(user_filter_map), filtername, fdat); |
Contributor
There was a problem hiding this comment.
Can zend_hash_add_ptr fail here and return NULL? Should the return value be checked?
Member
Author
There was a problem hiding this comment.
It can only fail on a duplicate key, and the factory registration above already rejects those with the same key, so I'll use zend_hash_add_new_ptr() to make that invariant explicit.
The volatile factory registration above already rejects a duplicate filter name, so the name cannot be in the map either. The add_new variant asserts that invariant in debug builds.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
…utdown re-registration.
Fix GH-22818
During request shutdown the user_filter_map is torn down by the user_filters RSHUTDOWN before the streams referencing it are flushed. A user filter whose filter() callback re-registers the filter recreated the now-NULL map, then, since the volatile factory was still present in FG(stream_filters), deleted the freshly added entry and left an empty orphaned map behind. The following stream_filter_append() located the factory but no matching fdat, tripping ZEND_ASSERT(fdat), and the recreated map leaked.
Register the volatile factory first and only create and populate user_filter_map on success, so a re-registration during the shutdown window fails without recreating the map. The existing NULL-map guard in user_filter_factory_create() then handles the append gracefully.