Messages in this thread |  | | From | (Alexander Sanda) | Subject | Weird problem | Date | 8 Jul 1996 19:30:20 GMT |
| |
Well, it's really weird, but it exists and I can reproduce it.
Situation:
I tried to mount a HPFS partiton with mount -thpfs /dev/sda8 /mnt Unfortunately, I made a typo, sda8 is not hpfs, it's ext2 (sda9 is the HPFS). Allright, mount told me about the error, and I corrected it. The partition got mounted, as I expected.
After a while, I noticed a increasing load (have xload running), uptime showed me a load average about 2.0. I did a ps and saw, that sendmail got stuck, while it was running the queue (it does it every 15 minutes).
A line such as: /usr/lib/sendmail: running queue ...
appeared in the ps output. The funny thing is, that I can't kill this stuck sendmail process. It just sits there and eats cpu load as hell, (after a while the loadavg increased to 4.0).
I have no clue, what sendmail has to do with filesystem code, but I'am not a kernel hacker, so I don't want to say any nonsense here :-)
Here is, what the kernel wrote to the log:
darkstar kernel: HPFS: hpfs_read_super: Not HPFS darkstar kernel: HPFS: bplus_lookup: sector not found darkstar last message repeated 4 times darkstar kernel: ll_rw_block: device 08:08: only 512-char blocks implemented (4096) darkstar kernel: ll_rw_block: device 08:08: only 512-char blocks implemented (4096) darkstar kernel: ll_rw_block: device 08:08: only 512-char blocks implemented (4096)
I can reproduce this behaviour both on 2.0.3 and 2.0.4. System is a P133/Triton/64megs/NCR (with the new driver, but it happens with the "original" driver aswell), IBM DPES-31080. kernel compiled with networking support, almost all as modules (also hpfs)
-- # /AS/ mailto:alex@darkstar.ping.at beware of the Penguin ! # # http://www.ping.at/members/alexa/ powered by Linux 2.0 # -----------------------------------------------------------------------
|  |