[Date Prev][Date Next] [Chronological] [Thread] [Top]

RE: back-bdb IDL cache leak (ITS#2367)



My mistake, I wasn't looking at the right section of code. Yes, shutdown
seems to be the only problem.

  -- Howard Chu
  Chief Architect, Symas Corp.       Director, Highland Sun
  http://www.symas.com               http://highlandsun.com/hyc
  Symas: Premier OpenSource Development and Support

> -----Original Message-----
> From: Jonghyuk Choi [mailto:jongchoi@us.ibm.com]

>
> On IDL cache entry replacement, both of AVL and LRU are being freed.
> Does avl_delete() not work as intended in the code ?
> I just added error checking code right after avl_delete() for testing.
>
> Upon shutdown, yes. Only LRU list was freed.
> avl_delete() is now called by bdb_db_close().
> Thanks.
>
> - Jong
>
> -----------------------------------
> Full_Name: Howard Chu
> Version: 2.1
> OS:
> URL:
> Submission from: (NULL) (24.126.120.178)
> Submitted by: hyc
>
>
> The IDL cache code maintains both an LRU list and an AVL tree for its
> cached
> IDLs. When it replaces an entry in the cache, it only
> replaces it in the
> LRU
> list. The AVL tree is left with stale data/leaks. Also, at database
> shutdown,
> only the LRU list is freed, so the AVL tree structures are
> leaked further.
>
> ------------------------
> Jong Hyuk Choi
> IBM Thomas J. Watson Research Center - Enterprise Linux Group
> P. O. Box 218, Yorktown Heights, NY 10598
> email: jongchoi@us.ibm.com
> (phone) 914-945-3979    (fax) 914-945-4425   TL: 862-3979
>
>