Bug 3618

Summary:	ide-tape driver needs patch to be SMP-stable
Product:	[Retired] Red Hat Linux	Reporter:	Tony Kocurko <akocurko>
Component:	kernel	Assignee:	Cristian Gafton <gafton>
Status:	CLOSED NEXTRELEASE	QA Contact:
Severity:	medium	Docs Contact:
Priority:	medium
Version:	6.0
Target Milestone:	---
Target Release:	---
Hardware:	i386
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	1999-08-31 16:48:33 UTC	Type:	---
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:

Description Tony Kocurko 1999-06-21 16:28:36 UTC

Without fail, writing any kind of level 0 dump to /dev/nht0
hangs the system after a while under RedHat 6.0. This
includes tar and dump directly to /dev/nht0 or piped through
to dd.

I suspect ide-tape.c.

We're running RedHat Linux with the 2.2.5-22smp SMP kernel
writing to a Seagate STT28000A-RF ATAPI Travan-4 tape drive.

Backups worked under RedHat 5.2, since I made them in
preparation for full, fresh installs of RedHat 6.0.

I have seen many, many reports of this problem with IDE tape
drives. Gadi Oxman, the author of ide-tape.c, suggested in
reply to one such problem a patch to take ide-tape.c from
version 1.14 to 1.15, but the patch fails on several hunks.

Comment 1 Tony Kocurko 1999-06-24 12:54:59 UTC

Gadi Oxman, the maintainer of /usr/src/linux/drivers/block/ide-tape.c,
sent a patch to upgrade it from version 1.14 to 1.15. That patch
installed flawlessly on RedHat's kernel version 2.2.5-22. I was then
able to do a tar backup of all our file systems, which include a pair
of 1.37 gigabyte seismic data files.

According to Gadi, his patch resolves some problem between ide-tape.c
and the SMP, which we are using in our kernel.

Comment 2 Bill Nottingham 1999-06-24 17:03:59 UTC

Can you post the patch?

Comment 3 Michael K. Johnson 1999-07-13 17:22:59 UTC

We should ask Gadi for the patch if the next kernel we release
does not have the new ide-tape driver in it, or get it ourselves
if it is otherwise available.  Newer 2.2.x kernels might include
the newer driver; I have not checked yet.

We should certainly use a new ide-tape driver anyway because of
the bug that slab poisoning found; I don't know if this is that
same bug or another bug.

Comment 4 Bill Nottingham 1999-08-31 16:48:59 UTC

fixed in the latest kernels in Raw Hide.