Re: kmemleak - percpu reliability?

From	Johannes Berg <johannes@sipsolutions.net>
Newsgroups	linux.kernel
Subject	Re: kmemleak - percpu reliability?
Date	2016-04-27 11:50 +0200
Message-ID	<rsugz-7q8-31@gated-at.bofh.it> (permalink)
References	<rstDQ-779-19@gated-at.bofh.it> <rsu6S-7lj-9@gated-at.bofh.it> <rsugz-7q8-33@gated-at.bofh.it> <rsu6S-7lj-9@gated-at.bofh.it>
Organization	linux.* mail to news gateway

Show all headers | View raw

On Wed, 2016-04-27 at 10:38 +0100, Catalin Marinas wrote:
> 
> Kmemleak tries to reduce the false positives to the detriment of more
> false negatives. 

:)

> One way it does this is by having to scan the memory
> twice and no changes to the leaked object (crc32) should have
> happened. It also scans the task stacks which is another source of
> false/stale pointers. The leak may eventually be reported but you
> can't really be precise on when this would be.
> 

Ok, fair enough. I don't remember if I asked it to scan twice, but
anyway, I did convince myself separately (with prints) that it was
leaked :)

Thanks for the explanation!

johannes

Back to linux.kernel | Previous | Next — Previous in thread | Find similar | Unroll thread

Thread

kmemleak - percpu reliability? Johannes Berg <johannes@sipsolutions.net> - 2016-04-27 11:10 +0200
  Re: kmemleak - percpu reliability? Catalin Marinas <catalin.marinas@arm.com> - 2016-04-27 11:40 +0200
    Re: kmemleak - percpu reliability? Johannes Berg <johannes@sipsolutions.net> - 2016-04-27 11:50 +0200

csiph-web