Repository navigation
gh-160324: Fix io.StringIO state after MemoryError - #160325
Merged
Merged
Conversation
Allocate before consuming the accumulated writes in realize() and make_intermediate(), and reset the size when the contents are lost, so that later calls do not index outside the internal buffer.
gpshead
marked this pull request as ready for review
October 22, 2026 19:15
Shrink the buffer of the accumulating writer before finishing it, so that getvalue(), read() and the other methods leave the contents unchanged when they fail with MemoryError instead of dropping them.
Member
Author
|
more |
|
Thanks @gpshead for the PR 🌮🎉.. I'm working now to backport this PR to: 3.14, 3.15. |
|
GH-161041 is a backport of this pull request to the 3.15 branch. |
|
GH-161042 is a backport of this pull request to the 3.14 branch. |
gpshead
added a commit
that referenced
this pull request
Nov 9, 2026
…0325) (#161041) * gh-160324: Keep `io.StringIO` state after `MemoryError` (GH-160325) Allocate before consuming the accumulated writes in realize() and make_intermediate(), and reset the size when the contents are lost, so that later calls do not index outside the internal buffer. Shrink the buffer of the accumulating writer before finishing it, so that getvalue(), read() and the other methods leave the contents unchanged when they fail with MemoryError instead of dropping them. ANT-2026-MPHGX2FL (cherry picked from commit 83b047e) Co-authored-by: Gregory P. Smith <68491+gpshead@users.noreply.github.com> * Avoid _PyUnicodeWriter_SetBuffer(); it does not exist in 3.15 Update the writer fields directly instead. * Do not copy the intermediate string in make_intermediate() In 3.15, writing a string to a new PyUnicodeWriter copies it, so a MemoryError there lost the contents of the StringIO. Disable overallocation for this write so the writer shares the string (copy-on-write) instead, as it does in main. --------- Co-authored-by: Gregory P. Smith <68491+gpshead@users.noreply.github.com> Co-authored-by: Gregory P. Smith <greg@krypto.org>
gpshead
added a commit
that referenced
this pull request
Nov 9, 2026
…0325) (#161042) * gh-160324: Keep `io.StringIO` state after `MemoryError` (GH-160325) Allocate before consuming the accumulated writes in realize() and make_intermediate(), and reset the size when the contents are lost, so that later calls do not index outside the internal buffer. Shrink the buffer of the accumulating writer before finishing it, so that getvalue(), read() and the other methods leave the contents unchanged when they fail with MemoryError instead of dropping them. ANT-2026-MPHGX2FL (cherry picked from commit 83b047e) Co-authored-by: Gregory P. Smith <68491+gpshead@users.noreply.github.com> * Avoid _PyUnicodeWriter_SetBuffer() and _PyUnicode_ResizeCompact() They do not exist in 3.14. Use PyUnicode_Resize(), which leaves the buffer unchanged on failure, and update the writer fields directly. * Do not copy the intermediate string in make_intermediate() In 3.14, writing a string to a new PyUnicodeWriter copies it, so a MemoryError there lost the contents of the StringIO. Disable overallocation for this write so the writer shares the string (copy-on-write) instead, as it does in main. --------- Co-authored-by: Gregory P. Smith <68491+gpshead@users.noreply.github.com> Co-authored-by: Gregory P. Smith <greg@krypto.org>
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
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.
realize()andmake_intermediate()inModules/_io/stringio.cconsumed the accumulated writes before allocating what they needed next, and on failure leftstring_sizeat the old length over an empty buffer. Both now do everything that can fail first: they allocate the buffer or the new writer, and shrink the old writer's buffer to its contents so thatPyUnicodeWriter_Finish()has nothing left to allocate. AMemoryErrorin any of these steps leaves theStringIOwith its previous contents. The shrink is the onePyUnicodeWriter_Finish()did anyway, so there is no extra copy;shrink_writer()reaches into_PyUnicodeWriterfor it, aswrite_str()already does.The test injects allocation failures with
_testcapi.set_nomemory, so it only runs where_testcapiis available.Fixes #160324
ANT-2026-MPHGX2FL
io.StringIOcrashes or returns garbage when used after aMemoryError#160324