AIC-7890 problems with kernel build

2000-09-04 Thread Brandon Hume
I've had this problem for the past couple of days (I went a while between updates of my kernel since I was busy installing a new disk and moving OSes around). Just after the "waiting for SCSI devices to settle" message, I'll get a number of SCB errors (which I don't have written down,

Re: AIC-7890 problems with kernel build

2000-09-04 Thread Chris Hedley
On Mon, 4 Sep 2000, Brandon Hume wrote: Just after the "waiting for SCSI devices to settle" message, I'll get a number of SCB errors (which I don't have written down, unfortunately), and then eventually a panic. This is with ACPI enabled... if I don't enable ACPI, it will proceed

Re: AIC-7890 problems with kernel build

2000-09-04 Thread Mike Meyer
Chris Hedley writes: On Mon, 4 Sep 2000, Brandon Hume wrote: Just after the "waiting for SCSI devices to settle" message, I'll get a number of SCB errors (which I don't have written down, unfortunately), and then eventually a panic. This is with ACPI enabled... if I don't enable ACPI,

Re: AIC-7890 problems with kernel build

2000-09-04 Thread Brandon Hume
On Mon, 4 Sep 2000, Brandon Hume wrote: Sounds similar to what I'm getting; after a boot I get a load of the following: ahc0: ahc_intr - referenced scb not valid during SELTO (31, 255) That's what I get... although I get SELTO (31, 0) Similar config to yours; a Tyan Thunder SMP BX with

Re: AIC-7890 problems with kernel build

2000-09-04 Thread Mitsuru IWASAKI
Hi, And, of course, the kernel config... a fair amount of crap still not trimmed out, but the only difference between this config and the one used to build the functional kernel is apm commented out and ACPI added, which I've already tried reversing without much difference. All I can say is

Re: AIC-7890 problems with kernel build

2000-09-04 Thread Brandon Hume
Could you disable acpi and try again to make the problem clearer? Without `device acpi' line in your kernel config file, any ACPI code As I mentioned, I've already done so. I know ACPI isn't causing the panic, but it does seem to change the BEHAVIOR of the panic. (I get the errors mentioned

Re: AIC-7890 problems with kernel build

2000-09-04 Thread Mike Meyer
Mitsuru IWASAKI writes: All I can say is that acpi is initilized after pcib and its children are attached so I don't think ACPI code affects PCI stuff... # Power management support (see LINT for more options) #device apm device acpi Could you disable acpi and

Re: AIC-7890 problems with kernel build

2000-09-04 Thread Mitsuru IWASAKI
Could you disable acpi and try again to make the problem clearer? Without `device acpi' line in your kernel config file, any ACPI code isn't compiled in your kernel. That was the first thing I tried when mine failed. It didn't make any difference - I saw the same failure (the page