HP M820e CDRW - nsp32 driver.

[cga2000]

On Wed, Oct 10, 2007 at 11:56:04AM EDT, Matthew Wilcox wrote:

[..]

> They do suggest that, but after looking into it, it seems like a small
> mistake in the driver that won't actually affect operation.

----------------------- /var/log/messages ----------------------------
Oct 10 18:11:44 turki -- MARK --
Oct 10 18:19:25 turki kernel: pccard: CardBus card inserted into slot 1
Oct 10 18:19:25 turki kernel: SCSI subsystem initialized
Oct 10 18:19:25 turki kernel: nsp32: loading...
Oct 10 18:19:25 turki kernel: PCI: Enabling device 0000:06:00.0 (0000 -> 0003)
Oct 10 18:19:25 turki kernel: ACPI: PCI Interrupt 0000:06:00.0[A] -> Link [LNKA] -> GSI 11 (level, low) -> IRQ 11
Oct 10 18:19:25 turki kernel: nsp32: No EEPROM detected: 0x0
Oct 10 18:19:25 turki kernel: nsp32: term power on
Oct 10 18:19:25 turki kernel: scsi0 : NinjaSCSI-32Bi/UDE: irq 11, io 0x3000+0x80
Oct 10 18:19:26 turki kernel: scsi 0:0:4:0: CD-ROM            hp       CD-Writer+ M820  1.23 PQ: 0 ANSI: 2
Oct 10 18:19:27 turki kernel: nsp32: irq: 11 mmio: d90ac000+0x1000 slot: 0000:06:00.0 model: KME SCSI CardBus card
Oct 10 18:19:27 turki kernel: nsp32: Bus Reset
Oct 10 18:19:37 turki kernel: nsp32: Host Reset
Oct 10 18:19:37 turki kernel: nsp32: term power on
Oct 10 18:19:47 turki kernel: sr 0:0:4:0: scsi: Device offlined - not ready after error recovery
Oct 10 18:19:47 turki kernel: sr0: scsi3-mmc drive: 0x/0x caddy
Oct 10 18:19:47 turki kernel: sr 0:0:4:0: Attached scsi generic sg0 type 5
----------------------- /var/log/messages ----------------------------

As far as I can tell, nothing seems really wrong till we get to the
"nsp32: term power on" message.  Not sure about the two previous
messages from the driver indicating a "Bus Reset" and much later a
"Host Reset"..

Are these normal scsi actions under the circumstances?

And then there is another ten-second wait and this time the "kernel"
complains that the device had to be "offlined". 

In this scenario, I don't see anything that would suggest an IRQ
conflict.

Just from the look of it I would even suspect something is wrong with
my hardware!

Would that make sense?

I still have an old copy of Win98 on this machine, so I guess I could
try to run some tests from there.

> This procedure does suggest some kind of interrupt problem though.
> IRQ 11 makes sense; it's shared with yenta.  

Not sure if, as shown above, the absence of messages suggesting an IRQ
conflict when I "hotplug" the device (as opposed to booting with the
card in place) is significant. 

> I don't know why you're not seeing the nsp32 interrupt handler run
> though.

That's another difficulty especially in my state of total ignorance:  I
don't know what I should be seeing if things were working normally.

I've seen that there are more current versions of this driver and I'm
beginning to think I should give them a try.  

Here's what I see after hotplugging the device:

________________________ lsmod ________________________________________
Module                  Size  Used by
sg                     31132  0 
sr_mod                 15652  0 
nsp32                  22764  0 
scsi_mod               98248  3 sg,sr_mod,nsp32
ppdev                   8452  0 
lp                     10660  0 
thermal                13608  0 
fan                     4548  0 
button                  7600  0 
processor              16572  1 thermal
ac                      4996  0 
battery                 9700  0 
eeprom                  6928  0 
serial_cs              20964  1 
3c574_cs               12264  1 
pcmcia                 34500  2 serial_cs,3c574_cs
firmware_class          9248  1 pcmcia
snd_es1968             24544  0 
snd_ac97_codec         87424  1 snd_es1968
ac97_bus                2144  1 snd_ac97_codec
snd_pcm_oss            38336  0 
snd_mixer_oss          15264  1 snd_pcm_oss
snd_pcm                68168  3 snd_es1968,snd_ac97_codec,snd_pcm_oss
snd_timer              19524  1 snd_pcm
snd_page_alloc          9512  2 snd_es1968,snd_pcm
snd_mpu401_uart         7360  1 snd_es1968
pcspkr                  2592  0 
psmouse                34344  0 
snd_rawmidi            21472  1 snd_mpu401_uart
snd_seq_device          7532  1 snd_rawmidi
8250_pnp                9376  0 
8250                   21476  2 serial_cs,8250_pnp
serial_core            18400  1 8250
parport_pc             31780  1 
parport                32296  3 ppdev,lp,parport_pc
uhci_hcd               20684  0 
i2c_piix4               8012  0 
rtc                    11376  0 
snd                    46212  9 snd_es1968,snd_ac97_codec,snd_pcm_oss,snd_mixer_oss,snd_pcm,snd_timer,snd_mpu401_uart,snd_rawmidi,snd_seq_device
i2c_core               20304  2 eeprom,i2c_piix4
soundcore               7104  1 snd
usbcore               117528  2 uhci_hcd
yenta_socket           24716  6 
rsrc_nonstatic         11936  1 yenta_socket
pcmcia_core            36656  5 serial_cs,3c574_cs,pcmcia,yenta_socket,rsrc_nonstatic
________________________ lsmod ________________________________________

Any leads as to what I should do next .. even advice as to where I could
find current documents that explain how things should work highly
appreciated.

Can't help but think that the main issue here is my ignorance.

:-)

Thanks,
cga