In article <19990402111937.B746@antioche.eu.org> you write:
>On Thu, Apr 01, 1999 at 07:50:21PM -0600, John Darrow wrote:
>> Recently, I've been having trouble with boot hangs on -current on i386.
>> I've got two similarly-configured machine, one a P5/233 MMX with 64MB
>> RAM, the other a PII/333 with 128MB RAM.  Both have an IDE hard drive
>> as master on the primary channel, an ATAPI ZIP as master on the secondary,
>> and an ATAPI CD-ROM as secondary slave.  Both booted fine with kernels from
>> 19990325 and before, but hang before detecting the boot device on
>> kernels since 19990326.  I am able to drop into ddb using ctrl-alt-esc.
 
>> My prime suspect is changes to the following three source files:
>> sys/dev/ata/ata_wdc.c           v1.17 -> 1.18
>> sys/dev/ic/wdc.c                v1.62 -> 1.63
>> sys/dev/scsipi/atapi_wdc.c      v1.18 -> 1.19
 
>> However, as I'm leaving for the weekend in about ten minutes, I haven't had
>> time to do any further isolation of the problem.
 
>Could you try a more recent kernel ? We have now:
>sys/dev/ata/ata_wdc.c v1.19
>sys/dev/ic/wdc.c v1.66
>sys/dev/scsipi/atapi_wdc.c v1.20

>I just fixed a problem in atapi_wdc.c yesterday that could cause similar
>behavior.

Well, after missing my flight, I was able to do some more experimenting.
I was able to get a kernel to boot with 19990401 (pre-branch) sources of
everything except scsipi/atapi_wdc.c, which I had to slightly patch with
a couple of rollbacks of the v1.18-1.19 patches.  As I'm now 1000 miles
away from both computers and about to hit the road again, I won't be
able to take a look at the new versions until at least Sunday night,
though, and probably won't try any more boot tests until I get back home
Monday night.

jdarrow

--
John Darrow
Computing Services, Wheaton College, Wheaton, IL
John.P.Darrow@wheaton.edu