On Tue, 2005-09-20 at 21:45 +1000, ibm-main wrote: Shane,
There was a small problem with the patch Maeda-san sent. It doesn't take into account that kill_sb is called even in the cases where fill_super failed. The patchset f0.1 addresses this. Let me know if you see any problems. Thanks, chandra > From: "MAEDA Naoaki" > > > > Sorry to interrupt the discussion. I found a different rcfs > > mount/umount problem on f0. > > > > I tried to run the rapid mount/unmount script on f0, and then > > stopped the script by SIGINT. After that, I've sometimes observed > > that further rcfs mount request is failed by EPERM permanently. > > > > I guess the reason is that rcfs_mount_count is wrongly > > incremented in rcfs_fill_super() on error return case. > > > > Following patch fix the problem. > > I have also observed the EPERM behaviour on f0 after some further testing. > Then I applied the patch, and following make/reboot, some *VERY* strange > things started happening. > I ran the "rapid" script, and interrupted it; "df -a" showed 2 (yes *two*) > entries for rcfs (both showing all zeroes). I issued a umount, and the > "df -a" showed just the one entry, but with usage figures that exactly > matched those for / (root) - not zeroes as I would have expected. > A subsequent umount for rcfs generated the following (all 4) messages ; > umount: none: not found > umount: /rcfs: not mounted > umount: none: not found > umount: /rcfs: not mounted > > I was subsequently able to replicate this to some extent after another > reboot; ran the script and interrupted it. Only one entry in the "df -a", > but it showed usage numbers as per /, and the 4 messges above issued after a > umount. The "echo" trace from the loop showed a umount happened to be the > last command issued, by there was still that entry in the "df". > > Food for thought maybe. > > As a side-note, I now have two (different) systems that consistently display > the umount hang at e19. > > Shane ... > > > > ------------------------------------------------------- > SF.Net email is sponsored by: > Tame your development challenges with Apache's Geronimo App Server. Download > it for free - -and be entered to win a 42" plasma tv or your very own > Sony(tm)PSP. Click here to play: http://sourceforge.net/geronimo.php > _______________________________________________ > ckrm-tech mailing list > https://lists.sourceforge.net/lists/listinfo/ckrm-tech > -- ---------------------------------------------------------------------- Chandra Seetharaman | Be careful what you choose.... - [EMAIL PROTECTED] | .......you may get it. ---------------------------------------------------------------------- ------------------------------------------------------- SF.Net email is sponsored by: Tame your development challenges with Apache's Geronimo App Server. Download it for free - -and be entered to win a 42" plasma tv or your very own Sony(tm)PSP. Click here to play: http://sourceforge.net/geronimo.php _______________________________________________ ckrm-tech mailing list https://lists.sourceforge.net/lists/listinfo/ckrm-tech
