| Weekly edition | Kernel | Security | Distributions | Search |
| Archives | Calendar | Subscribe | Write for LWN | LWN.net FAQ |
Nastiness explained (sort of)
Nastiness explained (sort of)
Posted May 7, 2008 23:29 UTC (Wed) by csamuel (subscriber, #2624)In reply to: Nastiness explained (sort of) by viro
Parent article: Stable kernel updates (security fix)
Al, thanks for that description! Could this be related to a problem we're seeing on Barcelona (B3) based systems where open(2) occasionally fails with ENOENT when creating a file in a directory that most certainly exists ? We're seeing it under 2.6.24 and 2.6.25, on both NFS and local (ext3) filesystems. We did think it was the queueing system occasionally corrupting the mode variable and dropping O_CREAT until a user forwarded us a message showing /bin/cp having an identical problem. :-(
(Log in to post comments)