➜

This issue tracker has been migrated to GitHub, and is currently read-only.
For more information, see the GitHub FAQs in the Python's Developer Guide.

Author vstinner
Recipients vstinner
Date 2019-06-18.13:05:20
SpamBayes Score -1.0
Marked as misclassified Yes
Message-id <1560863121.22.0.720622336463.issue37329@roundup.psfhosted.org>
In-reply-to
Content 
$ ./configure -C --enable-unicode=ucs4 --with-valgrind
$ make
$ valgrind --suppressions=Misc/valgrind-python.supp ./python -m test.regrtest test___all__
(...)
==18497== LEAK SUMMARY:
==18497==    definitely lost: 324 bytes in 2 blocks
==18497==    indirectly lost: 0 bytes in 0 blocks
==18497==      possibly lost: 5,665,418 bytes in 27,453 blocks
==18497==    still reachable: 8,530,840 bytes in 57,856 blocks
==18497==         suppressed: 32 bytes in 1 blocks
==18497== Reachable blocks (those to which a pointer was found) are not shown.
==18497== To see them, rerun with: --leak-check=full --show-leak-kinds=all
==18497== 
==18497== For lists of detected and suppressed errors, rerun with: -s
==18497== ERROR SUMMARY: 4082 errors from 4082 contexts (suppressed: 0 from 0)

I'm concerned by "324 bytes in 2 blocks". Attached PR fix it.

Downstream issue: https://bugzilla.redhat.com/show_bug.cgi?id=1696322


Fixing "definitely lost" false alarm should help to catch real memory leaks.
History
Date User Action Args
2019-06-18 13:05:21vstinnersetrecipients: + vstinner
2019-06-18 13:05:21vstinnersetmessageid: <1560863121.22.0.720622336463.issue37329@roundup.psfhosted.org>
2019-06-18 13:05:21vstinnerlinkissue37329 messages
2019-06-18 13:05:20vstinnercreate