Re: Linux hates my camcorder
Brought to you by:
aeb,
bencollins
From: Stefan R. <st...@s5...> - 2016-08-15 19:56:20
|
On Jul 09 Adam Goldman wrote: > I'm trying to capture video from a Sony DCR-TRV120. When I turn it on, I > get a continuous loop of the following messages on the console: > > Jun 24 19:00:06 starship kernel: [ 845.491993] firewire_ohci 0000:09:00.0: node ID not valid, new bus reset in progress > Jun 24 19:00:07 starship kernel: [ 846.632134] firewire_core 0000:09:00.0: rediscovered device fw0 > Jun 24 19:00:07 starship kernel: [ 846.718717] firewire_core 0000:09:00.0: rediscovered device fw1 > Jun 24 19:00:07 starship kernel: [ 846.718799] firewire_core 0000:09:00.0: phy config: new root=ffc1, gap_count=5 > Jun 24 19:00:07 starship kernel: [ 846.718950] firewire_ohci 0000:09:00.0: node ID not valid, new bus reset in progress > Jun 24 19:00:08 starship kernel: [ 847.860130] firewire_core 0000:09:00.0: rediscovered device fw0 > Jun 24 19:00:08 starship kernel: [ 847.946744] firewire_core 0000:09:00.0: rediscovered device fw1 > Jun 24 19:00:08 starship kernel: [ 847.946834] firewire_core 0000:09:00.0: phy config: new root=ffc1, gap_count=5 > > This is with a fresh Debian 8.5 install and the vendor kernel. uname > output is: > > Linux starship 3.16.0-4-686-pae #1 SMP Debian 3.16.7-ckt25-2 (2016-04-08) i686 Linux > > I'm using an IBM 19K5680 (FRU 19K5686) CardBus host adapter with PCI ID > 104c:8019 and the following lspci output: > > 09:00.0 FireWire (IEEE 1394): Texas Instruments TSB12LV23 IEEE-1394 Controller This link layer controller is quite old. Alas I am not in possession of an errata document for this model. (There is TI literature slls312: TSB12LV22 errata; perhaps some of them have been fixed in TSB12LV23.) Could you install linux-firewire-utils and run lsfirewirephy to read out the chip ID of the physical layer device of the card? With a bit of luck, you may even be able to run lsfirewirephy while the camcorder is attached, and collect the camcorder's phy identifiers too. > This camcorder worked OK with the old drivers and a TI-based PCI card 10 > years ago. I confirmed that its FireWire port is still functional and > that this host adapter works with other video devices. > > Any suggestions? Have you had a chance to test this camcorder on a different machine recently? (Different host adapter, or/and different OS.) I am undecided whether your kernel log above is showing a problem with software/firmware interaction, or a problem at the physical layer. You get very brief intervals in which firewire-core can read something on the bus before the next reset occurs, so maybe a software/firmware problem is more likely. If you can, please collect a more verbose log by entering # echo 7 > /sys/module/firewire_ohci/parameters/debug or $ sudo dd > /sys/module/firewire_ohci/parameters/debug <<< 7 then plug in/ switch on the camcorder. After you got the kernel log from a few bus reset loops, you can switch verbose logging off again by writing 0 instead of 7. -- Stefan Richter -======----- =--- -==== http://arcgraph.de/sr/ |