Kernel memory leak detector 0.10
Get series with b4
b4 am [email protected]Versions newest first
- v1 Latest10 patches 9 repliesIn review
Patches in v1 No applies recorded yet
| # | Subject and files | Size | Replies | Applied |
|---|---|---|---|---|
| 1/10 | Base support for kmemleakinclude/ include/ init/ +3 more | +1728 | 0 replies | |
| 2/10 | Some documentation for kmemleakDocumentation/ | +157 | 0 replies | |
| 3/10 | Add the memory allocation/freeing hooks for kmemleakinclude/ mm/ mm/ +1 more | +45 | 0 replies | |
| 4/10 | Modules support for kmemleakkernel/ | +41 | 0 replies | |
| 5/10 | Add kmemleak support for i386arch/ include/ include/ | +25 | 0 replies | |
| 6/10 | Add kmemleak support for ARMarch/ include/ | +19 | 0 replies | |
| 7/10 | Remove some of the kmemleak false positivesdrivers/ drivers/ drivers/ +10 more | +56 | 0 replies | |
| 8/10 | Keep the __init functions after initializationinclude/ lib/ | +17 | 0 replies | |
| 9/10 | Simple testing for kmemleaklib/ mm/ mm/ | +101 | 0 replies | |
| 10/10 | Update the MAINTAINERS file for kmemleakMAINTAINERS | +6 | 0 replies |
Reviews trailers per patch
No Reviewed-by, Acked-by or Tested-by trailers for v1 yet.
Cover letter [PATCH 2.6.18-rc6 00/10] Kernel memory leak detector 0.10
This is a new version (0.10) of the kernel memory leak detector. See the Documentation/kmemleak.txt file for a more detailed description. The patches are downloadable from (the whole patch or the broken-out series):
http://homepage.ntlworld.com/cmarinas/kmemleak/patch-2.6.18-rc6-kmemleak-0.10.bz2
http://homepage.ntlworld.com/cmarinas/kmemleak/broken-out/patches-kmemleak-0.10.tar.bz2
What's new in this version:
- replaced the pointers radix tree with a hash table to avoid the locking dependencies caused by the radix tree memory allocations
- fixed locking dependency problems by no longer holding the memleak_lock when allocating/freeing memory from kmemleak and also using RCU
- changed the naming of tracked memory blocks from "pointer" to "object"
- code clean-up
To do:
- testing on a wider range of platforms and configurations
- support for ioremap tracking (once the generic ioremap patches are merged)
- eliminate the task stacks scanning (if possible, by marking the allocated blocks as temporary until the return to user-space - Ingo's suggestion)
- precise type identification (after first assessing the efficiency of the current method as it requires changes to the kernel API)
Signature · 1 line
Catalin