Firewire problem 2: solved
Trying to analyse why my firewirecard does not detect my camera:
the ieee card seems to be recognised: I used lspci
(man lspci describes lspci as “a utility for displaying information about all PCI buses in the system and all devices connected to them. It shows a brief list of devices”).
~$ lspci | grep IEEE1394
0000:00:0a.0 FireWire (IEEE 1394): NEC Corporation uPD72874 IEEE1394 OHCI 1.1 3-port PHY-Link Ctrlr (rev 01)
That the card is recognised does not mean it is working.
Then made sure the necessary capture / export modules were loaded:
~$ modprob raw1394
~$ modprob dv1394
~$ modprob ohci1394
lsmod gives following:
( man lsmod: “lsmod is a program which formats the contents of the /proc/modules, showing what kernel modules are currently loaded.”)
~$ lsmod | grep 1394
raw1394 30956 0
ohci1394 35124 0
ieee1394 299832 4 raw1394,sbp2,ohci1394,dv1394
scsi_mod 139496 6 sg,sd_mod,usb_storage,sr_mod,sbp2,libata
then i check the bootup messages with dmesg
(man dmesg says: “dmesg is used to examine or control the kernel ring buffer. The program helps users to print out their bootup messages.”)
:~$ dmesg | grep 1394
[17179580.296000] ieee1394: Initialized config rom entry `ip1394'
[17179580.348000] ohci1394: fw-host0: Get PHY Reg timeout [0x00000000/0x00000000/100]
[17179580.452000] ohci1394: fw-host0: OHCI-1394 1.1 (PCI): IRQ=[177] MMIO=[ffe7e000-ffe7e7ff] Max Packet=[2048] IR/IT contexts=[4/4]
[17179580.452000] ohci1394: fw-host0: Get PHY Reg timeout [0x00000000/0x00000000/100]
[17179581.552000] ohci1394: fw-host0: Get PHY Reg timeout [0x00000000/0x00000000/100]
[17179598.484000] ieee1394: sbp2: Driver forced to serialize I/O (serialize_io=1)
[17179598.484000] ieee1394: sbp2: Try serialize_io=0 for better performance
[17179748.480000] ieee1394: raw1394: /dev/raw1394 device initialized
[17181055.968000] dv1394: stop_dma: DMA still going after 100 ms!
There are things in here, I don’t understand. “DMA still going after 100 ms” sounds not great … But raw1394 is initialised and also sbp2, (Serial Bus Protocol, for storage devices http://www.linux1394.org/sbp2.php)
A tip from http://linuxfr.org/forums/9/18451.html
To check if any signal from devices connected to the firewire ports is detected:
Switch of any firewire devices, then start tail,
~$ tail -f /var/log/messages
and switch the devices back on. Tail will log all activity on the busses. It will show the devices you connected.
In my case tail just blanks .... no registration of any activity.
But traffic on the USB busses is registered. When I pull my keyboard, and put it back in it is detected, so I am sure the program works.
At this stage, I now the video modules work, I know the firewire card is seen by the computer, also that there is no data traffic through the ieee card. ... is the card broken?
Next step, burn a Knoppix live cd (http://www.knoppix.org/) switch camera and firewire disk on, and boot from cd. On boot, I get the message: telling me that
Autodetection hangs. Udev hot-plug hardware detection failed.
Boot is interrupted.
Stefan Richter linux1394-user mailinglist adds advise:
(...) one easy thing to check is the contents of
/proc/interrupts. The second column is the number of interrupts
received. It should increase by one or more on ohci1394's line if a
(powered up) device is plugged in or out of a FireWire port. If it
doesn't, there is either a problem between hardware and Linux' generic
interrupt routing, or the physical bus interface was damaged.
I did not try this, but went straight to the shop.
After replacing the ieee1394 card by a new one, everything works again.
I can capture and export, and my disks are detected. Great.
My old card is a Idream Firewire Pci Kit. Firewire IEEE 1394A PCI Host Controller card with 4 Ports (3 external + 1 Internal) At the time I choose this one because it is clearly marked as compliant to IEEE 1394 OpenHCI Specifications.
The new card is a La Cie FireWire 400 PCI Card. Not marked on package as open standards compliant. only: "compatible with Windows". Not very reassuring ... In the user guide finally, compliancy is mentioned.
Posted on: Friday, October 27, 2006
by: Peter in category: Hardware, Problems and solutions
Recent Comments