GeezerVison is showing github.com — gh-129005: Remove copies from _pyio using take_bytes by cmaloney · Pull Request #141539 · python/cpython
Conversation
cmaloney
commented
Nov 13, 2025
•
edited
Loading
Uh oh!
There was an error while loading. Please reload this page.
Memory usage now matches that of _io for large files.
$ ./python -Wd -I -m test test_io.test_largefile -uall,largefile,walltime -M32G -v
== CPython 3.15.0a1+ (heads/gh129005_take_bytes-dirty:d8b88498ebf, Nov 13 2025, 15:43:51) [Clang 21.1.5 ]
== Linux-6.17.7-arch1-1-x86_64-with-glibc2.42 little-endian
== Python build: debug
== cwd: <build_dir>/build/test_python_worker_116910æ
== CPU count: 32
== encodings: locale=UTF-8 FS=utf-8
== resources: all
Using random seed: 365280675
0:00:00 load avg: 9.60 Run 1 test sequentially in a single process
0:00:00 load avg: 9.60 [1/1] test_io.test_largefile
test_large_read (test.test_io.test_largefile.CLargeFileTest.test_large_read) ...
... expected peak memory use: 2.3G
... process data size: 2.3G
ok
test_lseek (test.test_io.test_largefile.CLargeFileTest.test_lseek) ... ok
test_osstat (test.test_io.test_largefile.CLargeFileTest.test_osstat) ... ok
test_seek_read (test.test_io.test_largefile.CLargeFileTest.test_seek_read) ... ok
test_seek_readall (test.test_io.test_largefile.CLargeFileTest.test_seek_readall) ...
... expected peak memory use: 2.3G
... process data size: 2.3G
... process data size: 2.3G
ok
test_seekable (test.test_io.test_largefile.CLargeFileTest.test_seekable) ... ok
test_truncate (test.test_io.test_largefile.CLargeFileTest.test_truncate) ... ok
test_large_read (test.test_io.test_largefile.PyLargeFileTest.test_large_read) ...
... expected peak memory use: 2.3G
... process data size: 2.3G
ok
test_lseek (test.test_io.test_largefile.PyLargeFileTest.test_lseek) ... ok
test_osstat (test.test_io.test_largefile.PyLargeFileTest.test_osstat) ... ok
test_seek_read (test.test_io.test_largefile.PyLargeFileTest.test_seek_read) ... ok
test_seek_readall (test.test_io.test_largefile.PyLargeFileTest.test_seek_readall) ...
... expected peak memory use: 2.3G
... process data size: 2.3G
... process data size: 2.3G
ok
test_seekable (test.test_io.test_largefile.PyLargeFileTest.test_seekable) ... ok
test_truncate (test.test_io.test_largefile.PyLargeFileTest.test_truncate) ... ok
test_it (test.test_io.test_largefile.TestCopyfile.test_it) ... ok
test_it (test.test_io.test_largefile.TestSocketSendfile.test_it) ... ok
----------------------------------------------------------------------
Ran 16 tests in 6.950s
OK
0:00:06 load avg: 9.72 [1/1] test_io.test_largefile passed
== Tests result: SUCCESS ==
1 test OK.
Total duration: 7.0 sec
Total tests: run=16
Total test files: run=1/1
Result: SUCCESS
For reference; this test goes from 8.5sec overall duration on my AMD 64 bit Arch Linux box to 7 seconds.
Sorry, something went wrong.
Uh oh!
There was an error while loading. Please reload this page.
Memory usage now matches that of _io for large files
|
Optimization enabled by gh-139871 (Trying to keep that one issue referencing all the things which I use it for; planning to expand the What's New with a single "here's all the modules that got faster" list once optimize)
|
Sorry, something went wrong.
Uh oh!
There was an error while loading. Please reload this page.
cmaloney
changed the title
gh-129005: Remove copy from _pyio using take_bytes
gh-129005: Remove copies from _pyio using take_bytes
Nov 14, 2025
|
|
@@ -620,15 +620,15 @@ def read(self, size=-1): |
|
|
if n < 0 or n > len(b):
|
|
|
raise ValueError(f"readinto returned {n} outside buffer size {len(b)}")
|
|
|
del b[n:]
|
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.
Is there a difference between b.resize(n) and del b[n:]?
Sorry, something went wrong.
Uh oh!
There was an error while loading. Please reload this page.
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.
.resize() could grow while the del b[n:] never will. There's some code which does a = bytearray(b'12'); del a[3:] (intentionally del a slice past the end) as part of its buffer management. Generally I prefer resize because there's a lot less code involved than the slice code (and the slice code eventually calls resize anyways)
Sorry, something went wrong.
Uh oh!
There was an error while loading. Please reload this page.
vstinner reacted with thumbs up emoji
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.
Sorry, something went wrong.
Uh oh!
There was an error while loading. Please reload this page.
Uh oh!
There was an error while loading. Please reload this page.
StanFromIreland
pushed a commit
to StanFromIreland/cpython
that referenced
this pull request
Dec 6, 2025
…1539)
Memory usage now matches that of _io for large files.
Reviewers
vstinner
vstinner approved these changes
Uh oh!
There was an error while loading. Please reload this page.
Assignees
No one assigned
Milestone
No milestone
Development
Successfully merging this pull request may close these issues.
Uh oh!
There was an error while loading. Please reload this page.
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.
|
vstinner
left a comment