"Excuse me please. My ear is full of milk..."

– Oliver Hardy, Going Bye Bye (1934)

Skip on down to the menu.

Problems With EAGLE and pam_mount in openSUSE 11.3/11.4

Wow, this one's taken a really long time to find a solution to.

The problem occurs when running EAGLE pcb cad version 4.x from a smb share on a Windows server. The share is mounted using pam_mount so that it shows up in each user's ~/ directory.

What happens is that after the user logs in and the share is mounted, EAGLE won't start. It complains that the directory does not exist. Sometimes just letting the computer sit overnight seems to solve the problem. I suspect it has something to do with the server being shut down and re-started over night for backup. Once I got EAGLE up and running I would keep it running for weeks at a time so I could get work done when I needed to instead of waiting until the next morning.

Other than that nothing seemed to solve the problem, until I ran across this mail list archive. It describes the following problem.

on 32-bit systems using plain stat() (not stat64()) is problematic on CIFS mount points (kernel 2.6.32)

Later on it says:

Is your server guaranteed to return unique inode numbers? If not as a workaround, try mounting with "noserverino" mount option and see whether the problem goes away.

So... I added "noververino" to the mount options in pam_mount and problem solved! Woo hoo!!! Boy am I glad to be rid of that problem.

By the way, EAGLE version 5.x works just fine. I suspect the issue stems from the Qt tool kit used to develop the program. Version 4 is not current any more and I suspect more and more of these strange compatibility issues will be surfacing as the current Linux code base changes over time. It's not like the Windows world where backward compatibility is guaranteed for an almost insane period of time.