> > "Mark A. Grondona": > > Could I ask what the bug in the previous patch was? > > Also the new enhancement? thanks :-) > > The bug is in aufs_mmap(), a race condition. > When two threads issues mmap(2) for the same file descriptor first time, > the condition based upon the local var named 'mmapped' may be incorrect. > The enhancement is disacarding the newly added mutex lock in the > previous patch, and re-use the rw_semaphore in struct au_finfo.
Thanks, Junjiro! I tested your new patch overnight with no failures, so I think it is good. Thanks again for your help, and the explanation. mark > > Junjiro Okajima > ------------------------------------------------------------------------- This SF.net email is sponsored by: Microsoft Defy all challenges. Microsoft(R) Visual Studio 2008. http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
