Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 22 Jan 2001 21:54:42 +0800
From:      <derrenl@synology.com>
To:        iain@research.canon.com.au;, gibbs@scsiguy.com;, cheen@synology.com;;
Cc:        freebsd-scsi@FreeBSD.ORG
Subject:   RE: More target mode observations
Message-ID:  <200101221354.VAA12927@synology.com>

next in thread | raw e-mail | index | archive | help
>Following up to myself with a bit more info,  
>
>I played around a little bit more, and discovered that yes, we do have
>some atio's waiting, and if we get a bus reset one of these gets sent
>back.
>
>I also discovered that ahc_intr() never gets called during the rescan of
>the bus whilst in target mode. To me at least this would suggest that
>the controller is never being selected, or failing that the controller
>is never telling the device driver that it has been selected.
>
>Unfortunately we do not have any hardware here for examining SCSI
>busses, otherwise I could find out whether the controller was responding
>to the select with BSY (well, we have the logic analyser, but nothing to
>plug it in with).
>
>The other thing is that I only get bus reset intermittantly, which
>worries me. By this I mean that I can do a camcontrol reset <bus> on one
>machine, and the other machine doesn't always pick it up (I don't get
>he "Somebody reset bus %d" message).
>
 
Hi:

I am experiencing exactly what you described in your mail.
And I am also looking for the answers.

I have two AHA-2940U2W card on two different machine respectively.
I named them brain-test and idb-test respectively. They are connected by
an external scsi cable. The chip on the scsi card is aic-7890 (it should 
work well according to Justin.) And the FreeBSD version is 4.1-20000812-STABLE.

Unfortunately, when I triggered the target mode operation on the idb-test 
machine and rescanned the bus from the brain-test machine, it found nothing. 
On the one way, the brain-test did select the target mode controller, but it 
got no response and returned with time out message. On the other way, the
target controller didn't capture any select request. It looks like that chip on
idb-test is not properly triggered. I mean it doesn't know it should act
as a target mode controller. But the interesting thins is it can catch the 
channel reset signal.

I had also tested the aic-7899, and put in on the idb-test machine. 
But it still didn't work.

I'm including a copy of the /var/log/messages with diagnostic output.
I wish it helps for verifying the situations.

If anyone has any ideas or suggestions, I would like to hear from you

	Derren L.
	derrenl@synology.com


Here is the /var/log/messages output:
#
# the log on brain-test machine
#
brain-test /kernel: ahc1: <Adaptec 2940 Ultra2 SCSI adapter> port 0x1c00-0x1cff mem 0xfb400000-0xfb400fff irq 10 at device 14.0 on pci1
brain-test /kernel: ahc1: Reading SEEPROM...done.
brain-test /kernel: ahc1: BIOS eeprom is present
brain-test /kernel: ahc1: Secondary High byte termination Enabled
brain-test /kernel: ahc1: Secondary Low byte termination Enabled
brain-test /kernel: ahc1: Primary Low Byte termination Enabled
brain-test /kernel: ahc1: Primary High Byte termination Enabled
brain-test /kernel: ahc1: aic7890/91 Wide Channel A, SCSI Id=7, 16/255 SCBs
brain-test /kernel: ahc1: hardware scb 64 bytes; kernel scb 32 bytes; ahc_dma 8 bytes
brain-test /kernel: ahc1: Downloading Sequencer Program... 403 instructions downloaded

# Enable the debug information
brain-test /kernel: (xpt0:xpt0:0:-1:-1): debugging flags now 7

# Reset the bus. And the idb-test did get the reset request.
brain-test /kernel: (noperiph:ahc1:0:-1:-1): SCSI bus reset delivered. 0 SCBs aborted.
 
# Rescan the target mode controller
# It got a select time out status
brain-test /kernel: (probe0:ahc1:0:9:0): probestart   
brain-test /kernel: (probe0:ahc1:0:9:0): xpt_action   
brain-test /kernel: (probe0:ahc1:0:9:0): ahc_action
brain-test /kernel: (probe0:ahc1:0:9:0): start scb(0xc0a41000)
brain-test /kernel: ahc_handle_scsiint: SELTO
brain-test /kernel: (probe0:ahc1:0:9:0): ahc_done - scb 0
brain-test /kernel: (probe0:ahc1:0:9:0): xpt_done
brain-test /kernel: (probe0:ahc1:0:9:0): camisr(probe0:ahc1:0:9:0): probedone


#
# the log on idb-test machine
#
idb-test /kernel: ahc1: <Adaptec 2940 Ultra2 SCSI adapter> port 0x1c00-0x1cff mem 0xfb400000-0xfb
idb-test /kernel: ahc1: Reading SEEPROM...done.
idb-test /kernel: ahc1: BIOS eeprom is present
idb-test /kernel: ahc1: Secondary High byte termination Enabled
idb-test /kernel: ahc1: Secondary Low byte termination Enabled 
idb-test /kernel: ahc1: Primary Low Byte termination Enabled   
idb-test /kernel: ahc1: Primary High Byte termination Enabled   
idb-test /kernel: ahc1: aic7890/91 Wide Channel A, SCSI Id=9, 16/255 SCBs
idb-test /kernel: ahc1: hardware scb 64 bytes; kernel scb 32 bytes; ahc_dma 8 bytes
idb-test /kernel: ahc1: Downloading Sequencer Program... 392 instructions downloaded
 
# Enable the debug information
idb-test /kernel: (xpt0:xpt0:0:-1:-1): debugging flags now 7

# The target mode was enabled here
idb-test /kernel: (targ0:ahc1:0:9:0): Lun now enabled for target mode

# idb-test can capture the channel reset signal
idb-test /kernel: ahc1: Someone reset channel A
idb-test /kernel: (targ0:ahc1:0:9:0): camisrSaw event 4e:0






To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-scsi" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200101221354.VAA12927>