home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!ogicse!uwm.edu!ux1.cso.uiuc.edu!mp.cs.niu.edu!fnnews!fncrd6.fnal.gov!rohrer
- From: rohrer@fncrd6.fnal.gov (Keith Rohrer)
- Newsgroups: comp.os.linux
- Subject: Re: HArd drive timeout
- Message-ID: <2173@fnnews.fnal.gov>
- Date: 12 Aug 92 15:32:22 GMT
- Article-I.D.: fnnews.2173
- References: <1992Aug12.123856.1635@capitol.com>
- Sender: news@fnnews.fnal.gov
- Organization: Fermilab
- Lines: 55
- Nntp-Posting-Host: fncrd6.fnal.gov
-
- In article <1992Aug12.123856.1635@capitol.com> moore@capitol.com (Jim Moore) writes:
- >I just recently bought a Talon 120Mb IDE drive for my
- >Gateway 2000 33Mhz 386 specifically to use Linux.
- >I have the disc partitioned into :
- > /
- > swap
- > /usr
- > /home
- >And I have even compiled and booted a new kernel (0.96c).
- >However, whenever I do a big compile/link I start getting
- >"HD-timeout" "HD reset" "HD ready" mesasges and have to power cycle.
- >The Talon is the second IDE drive in the system. The first is a 200Mb
- >Maxtor one huge DOS partition.
- >I browsed around the kernel hd.c code a little, but I really don't
- >know much about the PC disc hardware.
- >Any ideas/fixes would be most welcome.
- >Jim Moore
- >moore@capitol.com
- >--
- >Jim Moore Capitol Disc Interactive
- >moore@capitol.com
- > =============================================
- > If you lie to a computer, it will get you.
-
- I've got a similar situation. I've got a 386-40, Forex chipset, AMI BIOS (5-5-
- 91 date), a Maxtor 7213A hard drive, and hard drive timeouts regularly under
- disk-intensive situations, especially when I try to do things on two VC's
- at once. I've had this problem on 0.96c, all patchlevels, and 0.97; I haven't
- had time to apply pl1 to 0.97 quite yet. When I do something disk-intensive
- such as fsck/efsck, or gcc, and sometimes lha or tar | tar, I'll get the
- hard drive timeout, resetting message, usually several times. I modified
- the printk in question in hd.c, and the sectors it times out on are similar,
- but not always the same (some sectors time out more often than others).
- In the case of fsck/efsck, though, it will give me a huge number of errors
- which are not consistent with one another, then if I run it again, everything
- will check out just fine. The problems seem to come up far more often when
- I start something going in one VC, then switch to another VC and do something
- else that uses the disk (including, but not limited to, logging in and
- starting top!). I made that modification someone mentioned a few days ago
- including setting HD_DELAY to 120. No help. I did, however, add to the
- printk the buffer address in question; I will see if that changes.
-
- Does anybody have any idea what to do? Is it the Maxtor? The partition
- is physically perfect, having been used and NDD'd many times under MS-DOG.
-
- This has been plaguing me for weeks now...
-
- Keith
- (rohrer@fncrd0.fnal.gov)
-
- --
- Disclaimer: None of Grinnell College, URA, Fermilab, and any other affiliated
- persons or orginizations have licensed my ideas or opinions, and thus are
- not entitled to any which may appear above.
-
-