apr-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Philip Martin <phi...@codematters.co.uk>
Subject Re: svn commit: rev 2017 - trunk/subversion/libsvn_subr
Date Mon, 27 May 2002 01:13:10 GMT
striker@tigris.org writes:

> Author: striker
> Date: 2002-05-26 09:27 GMT
> New Revision: 2017
> Modified:
>    trunk/subversion/libsvn_subr/svn_error.c
> Log:
> * subversion/libsvn_subr/svn_error.c
>   (svn_pool_create): Use the new functionality in APR to set
>     a maximum on the total amount of memory kept in the freelist.
>     I've choosen to set it to 4MB without any justification, so
>     it might very well be that this number needs to be tweaked.
>     For a bit of background on this change, see:
>       http://subversion.tigris.org/servlets/ReadMsg?msgId=73898&listName=dev

Since this change the Subversion regression tests are failing when
compiled with APR_POOL_DEBUG=1.  They hang in svn_pool_clear() when it
recreates the Subversion error pool. Essentially the Subversion code
does this

  apr_pool_clear_debug (some_pool, file_line);
  apr_pool_create_ex(error_pool, some_pool, abort_on_pool_failure, NULL);

The problem is that when some_pool was originally created by
apr_pool_create_ex() a mutex was created which was allocated within
some_pool itself.  When apr_pool_clear_debug is called it free's all
the allocated memory including the memory used for mutex.  Thus any
further attempt to use some_pool, such as for the parent of the new
error_pool, may fail when attempting the lock some_pool's mutex. The
Subversion regression tests hang attempting to lock this mutex.


View raw message