2.4.3 + aic7xxx-6.1.9 doesn't boot

2.4.3 + aic7xxx-6.1.9 doesn't boot

[Igor Mozetic]

The latest aic7xxx-6.1.9 doesn't boot, I can see something like:

scsi1:0:0:0: Attempting to queue an ABORT message 
scsi1:0:0:0: Command found on device queue 
aic7xxx_abort returns 8194 

2.4.3 + aic7xxx-6.1.8 works fine (ignoring problems with
renaming devices and not beeing able to read from CD RW).
C440GX+ UP, 2G RAM, working dmesg shows:
  
SCSI subsystem driver Revision: 1.00
scsi0 : SCSI host adapter emulation for IDE ATAPI devices
  Vendor:           Model: ATAPI CDROM       Rev: 120N
  Type:   CD-ROM                             ANSI SCSI revision: 02
Detected scsi CD-ROM sr0 at scsi0, channel 0, id 0, lun 0
sr0: scsi3-mmc drive: 116x/0x cd/rw xa/form2 cdda tray
Uniform CD-ROM driver Revision: 3.12
scsi1 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 6.1.8
        <Adaptec aic7896/97 Ultra2 SCSI adapter>
        aic7896/97: Wide Channel A, SCSI Id=7, 32/255 SCBs

scsi2 : Adaptec AIC7XXX EISA/VLB/PCI SCSI HBA DRIVER, Rev 6.1.8
        <Adaptec aic7896/97 Ultra2 SCSI adapter>
        aic7896/97: Wide Channel B, SCSI Id=7, 32/255 SCBs

  Vendor: IBM       Model: DNES-309170W      Rev: SA30
  Type:   Direct-Access                      ANSI SCSI revision: 03
Detected scsi disk sda at scsi1, channel 0, id 0, lun 0
  Vendor: SONY      Model: CD-R   CDU948S    Rev: 1.0j
  Type:   CD-ROM                             ANSI SCSI revision: 02
Detected scsi CD-ROM sr1 at scsi1, channel 0, id 6, lun 0
scsi1:0:0:0: Tagged Queuing enabled.  Depth 8
(scsi1:A:6): 10.000MB/s transfers (10.000MHz, offset 15)
sr1: scsi-1 drive
(scsi1:A:0): 40.000MB/s transfers (20.000MHz, offset 31, 16bit)
SCSI device sda: 17916240 512-byte hdwr sectors (9173 MB)

-Igor Mozetic

Re: 2.4.3 + aic7xxx-6.1.9 doesn't boot

[Justin T. Gibbs]

>The latest aic7xxx-6.1.9 doesn't boot, I can see something like:
>
>scsi1:0:0:0: Attempting to queue an ABORT message 
>scsi1:0:0:0: Command found on device queue 
>aic7xxx_abort returns 8194 

Either disable the initial bus reset in SCSI-Select or lower
the bus settle delay from 15000ms to something like 5000ms
in your kernel config.  Because I enforce the bus settle
delay by blocking incoming commands from the SCSI layer, yet
the low level drivers are not responsible for timeout handling,
the top level starts timeing out probe requests with short timeouts.
The current strategy allows the system to finish inializing other
devices while the bus settle expires, but it has this vulnerability.

It is also my fault for not properly testing the default bus settle
delay.  The system I tested it on had initial bus resets disabled
from a prior test of that scenario.  Oops.

--
Justin