Hello,
First of all, thanks to everyone who has made Knoppmyth possible- I am using R5C7 and it worked right out of the box. This is a great solution for my digital recording needs!

I am trying to get my box wake to record reliably. My hardware doesn't seem to support wake on alarm from S5 (even when I enter the time directly through the bios), but it does support wake on alarm from S3. Since I have to reboot to set the time, I had to modify some of the scripts I found on the wiki (http://www.knoppmythwiki.org/index.php? ... upToRecord). Essentially, I set the wake up time by nvram-wakeup, set up a cron job to run at reboot that puts the box into S3 (by way of the included hibernate script) and reboot; after the system is almost all the way up (I see the splash screen for the nvidia driver) it goes into hibernation mode. When I wake the system up (either manually or automatically) I modified the hibernate.conf file to remove the cron job. All of this has worked flawlessly when I wake it up manually, however when the system wakes up automatically, it records the program I have specified, but then it fails to go back to sleep.

I spent some time troubleshooting the problem, and I have determined that the problem is with the ivtv module. In order to get the system to drop into S3 properly, I have to remove the ivtv module before it hibernates. The friendly version 'modprobe -r ivtv' does not work; even with the front end and the back end shutdown, ivtv claims to be in use. So, I resorted to the scorched earth approach 'rmmod -f ivtv' and this works- when I have woken it up manually and sent it back to sleep manually. When the system wakes up automatically and records, it gets through the reboot procedure and then hangs. I can login remotely by ssh and dmesg puts out the following complaint (reconstructed from /var/log/kern.log):

Unable to handle kernel NULL pointer dereference at virtual address 00000000
printing eip:
e0c293b1
*pde = 00000000
Oops: 0000 [#1]
PREEMPT SMP
Modules linked in: ehci_hcd shpchp pci_hotplug 8250_pnp pcspkr cx8800 cx88xx ir_common ivtv bttv lirc_i2c lirc_dev wlan_wep ipv6 dvb_ttpci l64781 tda8083 stv0297 sp8870 ttpci_eeprom saa7146_vv saa7146 mt312 btcx_risc cx24110 tda1004x ves1820 stv0299 ves1x93 dvb_core video_buf v4l2_common v4l1_compat evdev usbhid skge autofs4 af_packet snd_ice1724 snd_ice17xx_ak4xxx snd_ak4114 snd_ak4xxx_adda snd_ac97_codec snd_ac97_bus snd_pcm snd_timer snd_page_alloc snd_mpu401_uart snd_rawmidi snd_seq_device snd soundcore tda9887 msp3400 saa7127 saa7115 tuner tveeprom i2c_algo_bit videodev ath_pci ath_rate_sample wlan ath_hal via_rhine mii agpgart i2c_viapro i2c_core parport_pc parport 8250 serial_core pcmcia firmware_class yenta_socket rsrc_nonstatic pcmcia_core video thermal processor fan container button battery ac rtc ohci1394 ieee1394 usb_storage ohci_hcd uhci_hcd usbcore a100u2w megaraid tmscsim tsdev
CPU: 0
EIP: 0060:[pg0+543769521/1067881472] Tainted: P VLI
EFLAGS: 00010217 (2.6.15-chw-2)
EIP is at i2c_del_adapter+0xc1/0x250 [i2c_core]
eax: db1f1bd8 ebx: 00000000 ecx: db1f19b4 edx: 00000000
esi: dfc32c44 edi: e0f5c8a8 ebp: db1f19b4 esp: d925fec8
ds: 007b es: 007b ss: 0068
Process rmmod (pid: 4537, threadinfo=d925e000 task=d9269a70)
Stack: db1f19b4 00000000 db1f0000 dfc32c44 e0f5c8a8 e0f5c8f8 e0f30bcd db1f19b4
db1f0000 dfc32c00 c02c5c6b dfc32c00 dfc32ccc c031d1f8 dfc32c44 dfc32c44
dfc32c84 e0f5c8a8 c031d2a6 dfc32c44 e0f5c900 e0f5c8a8 00000001 00000000
Call Trace:
[pg0+546945997/1067881472] ivtv_remove+0x1ed/0x4c0 [ivtv]
[pci_device_remove+43/48] pci_device_remove+0x2b/0x30
[__device_release_driver+104/128] __device_release_driver+0x68/0x80
[driver_detach+102/117] driver_detach+0x66/0x75
[bus_remove_driver+66/112] bus_remove_driver+0x42/0x70
[driver_unregister+11/32] driver_unregister+0xb/0x20
[pci_unregister_driver+14/32] pci_unregister_driver+0xe/0x20
[pg0+546947036/1067881472] module_cleanup+0x4c/0x50 [ivtv]
[pg0+546946960/1067881472] module_cleanup+0x0/0x50 [ivtv]
[sys_delete_module+316/384] sys_delete_module+0x13c/0x180
[syscall_call+7/11] syscall_call+0x7/0xb
Code: 93 50 ff ff ff 85 d2 0f 85 5f 01 00 00 89 c3 8b 00 0f 18 00 90 81 fb c8 ce c2 e0 75 e2 8b 9d 24 02 00 00 8d 85 24 02 00 00 39 c3 <8b> 3b 89 04 24 74 27 8d b3 a4 fe ff ff 8b 46 0c 56 ff 50 38 89
<6>ACPI: PCI interrupt for device 0000:00:09.0 disabled

Looking back through kern.log this is the same error I see on other occassions where the system has woken up automatically to record, then failed to go back to sleep.

I should also mention that this problem has happened intermittently in the past (sometimes the system would wake from S3 automatically, record and go back to sleep without a problem); but the last four times it has failed with the error above.

So, after this rather long post, my question is: does anyone know of a 'cleaner' way to unload ivtv? It seems that it doesn't like rmmod -f; I have searched other posts (and other message boards) but I can't find anything on this, I can't even determine what exactly would be using ivtv if the front end and backend are not up and running. Alternatively, if anyone out there knows how to stop ivtv from loading on boot I think I could make my scheme work. Any other solutions to my problem would also be welcome.

Thanks!
Jeremiah


Knoppmyth R5C7
Abit KV-85 mobo (Via K8M800/VT8237R)
AMD Sempron 3100+
512 MB Ram
Hauppage PVR-350
Chaintech AV-710 Sound Card
Generic GeForce MX4000 128 MB Nvidia video card
DLink DWL-G510 Wireless PCI card
Seagate 160 GB SATA-300 hard drive