All of lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH 2.6.17-rc5 0/7] Kernel memory leak detector 0.2
@ 2006-05-27 12:07 Catalin Marinas
  2006-05-27 12:23 ` [PATCH 2.6.17-rc5 1/7] Base support for kmemleak Catalin Marinas
                   ` (6 more replies)
  0 siblings, 7 replies; 11+ messages in thread
From: Catalin Marinas @ 2006-05-27 12:07 UTC (permalink / raw)
  To: linux-kernel

This is a new version (0.2) of the kernel memory leak detector based
on the tracing garbage collection technique. See the
Documentation/kmemleak.txt file for a more detailed description. The
patches are also available from
http://homepage.ntlworld.com/cmarinas/kmemleak/.

Thanks to all who contributed useful suggestions for these patches.

What's new in this version:

- cleaned up the code according to the suggestions on LKML
- implemented the buffering of allocation/freeing calls before
  kmemleak is properly initialised (many allocations would be missed
  without this feature)
- moved the pointer leaks information from /proc/memleak to
  /sys/kernel/debug/memleak (debugfs)
- the task stacks are not scanned by default because this can lead
  to many false negatives
- fixed the locking and tested on SMP (ARM)
- fixed other bugs

To do:

- better testing
- better support for modules (module static variables, new pointer
  aliases found in modules)
- test Ingo's suggestion on task stack scanning
- NUMA support

-- 
Catalin

^ permalink raw reply	[flat|nested] 11+ messages in thread
* [PATCH 2.6.17-rc5 0/7] Kernel memory leak detector 0.3
@ 2006-05-30 13:50 Catalin Marinas
  2006-05-30 14:07 ` [PATCH 2.6.17-rc5 1/7] Base support for kmemleak Catalin Marinas
  0 siblings, 1 reply; 11+ messages in thread
From: Catalin Marinas @ 2006-05-30 13:50 UTC (permalink / raw)
  To: linux-kernel

This is a new version (0.3) of the kernel memory leak detector. See
the Documentation/kmemleak.txt file for a more detailed
description. The patches are also available from
http://homepage.ntlworld.com/cmarinas/kmemleak/ (to be uploaded
shortly).

What's new in this version:

- more clean-up
- initial kernel modules support. It scans the data sections in a
  module while ignoring the text areas
- minor optimisation to avoid radix-tree lookup when updating
  the pointer information
- DEBUG_FS selected automatically (until/if a different method of
  triggering the memory scanning is implemented)

To do:

- better testing
- support for pointer aliases in kernel modules (i.e. container_of
  usage in modules)
- test Ingo's suggestion on task stacks scanning
- NUMA support

-- 
Catalin

^ permalink raw reply	[flat|nested] 11+ messages in thread

end of thread, other threads:[~2006-05-30 14:07 UTC | newest]

Thread overview: 11+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2006-05-27 12:07 [PATCH 2.6.17-rc5 0/7] Kernel memory leak detector 0.2 Catalin Marinas
2006-05-27 12:23 ` [PATCH 2.6.17-rc5 1/7] Base support for kmemleak Catalin Marinas
2006-05-27 18:41   ` Anne Thrax
2006-05-27 20:20     ` Catalin Marinas
2006-05-27 12:23 ` [PATCH 2.6.17-rc5 2/7] Some documentation " Catalin Marinas
2006-05-27 12:23 ` [PATCH 2.6.17-rc5 3/7] Add the memory allocation/freeing hooks " Catalin Marinas
2006-05-27 12:23 ` [PATCH 2.6.17-rc5 4/7] Add kmemleak support for i386 Catalin Marinas
2006-05-27 12:23 ` [PATCH 2.6.17-rc5 5/7] Add kmemleak support for ARM Catalin Marinas
2006-05-27 12:23 ` [PATCH 2.6.17-rc5 6/7] Remove some of the kmemleak false positives Catalin Marinas
2006-05-27 12:24 ` [PATCH 2.6.17-rc5 7/7] Simple testing for kmemleak Catalin Marinas
  -- strict thread matches above, loose matches on Subject: below --
2006-05-30 13:50 [PATCH 2.6.17-rc5 0/7] Kernel memory leak detector 0.3 Catalin Marinas
2006-05-30 14:07 ` [PATCH 2.6.17-rc5 1/7] Base support for kmemleak Catalin Marinas

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.