Red Hat Bugzilla – Bug 52333
Compaq AP500 integrated SCSI will not warm-boot with Red Hat Linux 7.1.
Last modified: 2007-04-18 12:36:13 EDT
From Bugzilla Helper:
User-Agent: Mozilla/4.0 (compatible; MSIE 5.5; Windows 98)
Description of problem:
Compaq AP500 with integrated SCSI controller(Smbios 53c875). BIOS is the
latest (Compaq 686W V7.14), Symbios 53C875 PCI UW SCSI(Don't have its BIOS
version since it's integrated). The problem is that the unit will fail to
recognize the SCSI controller during a warm boot. Cold boot works fine.
We've tried the native driver that came with the OS, the latest drivers
from the symbios website and none of them fixed this. This issue is
similar to a known issue in Red Hat version 6.2, which was resolved with a
driver. That driver was also tried, with no success. We've tried version
Red Hat 6.2 on the unit, and IT WORKS FINE--WARM BOOTS WITH NO PROBLEM.
Possibler causes (without hooking up the arium): SCSI controller left in
an invalid state during the reboot because the driver does not
reinitialize it properly (PCI reset not given?).
Version-Release number of selected component (if applicable):
Steps to Reproduce:
1. Put together the configuration in the description field.
2. Reboot the computer without a shutdown.
Actual Results: Will not boot.
Expected Results: It should boot.
There is a service request open (188375) for this issue that Joseph Tejano
Is this with the 2.4.2-2 or the 2.4.3-12 kernel ?
Pretty sure it is the 2.4.2-2 version.
The 2.4.2-2 875 scsi driver had a bug; we have made available an updated disk
for during the installation, and the 2.4.3-12 security update kernel has this
bug fixed. It would be useful to know if the 2.4.3-12 fixes your specific
problem (so that it would be the already known bug)
How do I get the disk, or do I just upgrade the kernel to version 2.4.3-12?
Recommended is upgrading to 2.4.3-12; the disk is only intended for during
installation (although it has the updated module for normal use too)
Since this hasn't been touched for months, I assume this is fixed by the
disk Arjan pointed to, so I'm closing the bug. Re-open if this is not