LinHES Forums
http://forums.linhes.org/

saa7134-alsa kernel bug
http://forums.linhes.org/viewtopic.php?f=2&t=18500
Page 1 of 1

Author:  silliam [ Thu Jul 10, 2008 12:13 am ]
Post subject:  saa7134-alsa kernel bug

Just did a full install of R5.5 and everything works except when i modprobe saa7134-alsa. It works fine until I try to watch tv from my saa7134 tuner. The screen just turns black for a minute or two and then I am returned to the main mythtv screen. I looked at the dmesg output and I get this kernel bug. saa7134-alsa worked fine in mythbuntu and in a past gentoo install so I am stumped as to how to proceed.

Does anyone have any suggestions as to what I should do? or should I report this to linux kernel bugzilla.

Code:
kernel BUG at /KNOPPIX.build/src/v4l/xc-test-d4f7804a393c/v4l/videobuf-dma-sg.c:282!
invalid opcode: 0000 [#1]
PREEMPT SMP
Modules linked in: nvidia(P) autofs4 lirc_pvr150 lirc_dev wlan_ccmp saa7134_alsa ipv6 af_packet fuse usbhid ff_memless pcmcia yenta_socket rsrc_nonstatic pcmcia_core video output sbs fan dock container battery ac aufs usb_storage ohci_hcd nvram wlan_scan_sta ath_rate_sample wm8775 cx25840 ivtv saa7115 msp3400 tuner tea5767 tda8290 tda18271 tda827x tuner_xc2028 tda9887 tuner_simple mt20xx tea5761 saa7134 compat_ioctl32 videobuf_dma_sg firmware_class videobuf_core i2c_algo_bit ir_kbd_i2c ir_common cx2341x tveeprom ath_pci videodev v4l2_common v4l1_compat wlan ath_hal(P) snd_intel8x0 snd_ac97_codec ac97_bus e1000 snd_pcm_oss snd_mixer_oss snd_pcm snd_timer button thermal processor parport_pc parport i2c_i801 8250_pnp 8250 serial_core intel_agp snd agpgart i2c_core ehci_hcd shpchp pci_hotplug uhci_hcd soundcore snd_page_alloc pcspkr rtc_cmos rtc_core rtc_lib tsdev evdev usbcore sbp2 ohci1394 ieee1394
CPU:    0
EIP:    0060:[<f8eb23f8>]    Tainted: P        VLI
EFLAGS: 00010296   (2.6.23-chw-4 #1)
EIP is at videobuf_dma_unmap+0x58/0x60 [videobuf_dma_sg]
eax: 00000029   ebx: f6c740cc   ecx: 00000002   edx: 00000000
esi: f6d187c0   edi: f72f0800   ebp: f2d9bf44   esp: f2d9be34
ds: 007b   es: 007b   fs: 00d8  gs: 0033  ss: 0068
Process mythbackend (pid: 3775, ti=f2d9a000 task=c19cf870 task.ti=f2d9a000)
Stack: f8eb31b0 00000000 19721112 f6c74000 f8eb2411 00000000 00000000 f6d187c0
       f6cde644 00000000 00000af5 f58e9a74 f6cde400 f6cde800 c19d8400 00000af5
       00000000 f27848a4 f2784958 00000000 00000000 00000000 00000000 00000000
Call Trace:
 [<f8eb2411>] videobuf_pci_dma_unmap+0x11/0x20 [videobuf_dma_sg]
 [<f8e3353b>] snd_pcm_oss_get_active_substream+0x5b/0x70 [snd_pcm_oss]
 [<c0567c1d>] __mutex_lock_slowpath+0x12d/0x290
 [<f99c3561>] snd_card_saa7134_hw_free+0x41/0x70 [saa7134_alsa]
 [<f8e13f36>] snd_pcm_release_substream+0x36/0x70 [snd_pcm]
 [<f8e33709>] snd_pcm_oss_release_file+0x19/0x30 [snd_pcm_oss]
 [<f8e34ee4>] snd_pcm_oss_release+0x44/0xa0 [snd_pcm_oss]
 [<c01854d2>] __fput+0xa2/0x1d0
 [<c01824b7>] filp_close+0x47/0x80
 [<c01838b9>] sys_close+0x69/0xc0
 [<c0104412>] syscall_call+0x7/0xb
 =======================
Code: c7 43 18 00 00 00 00 c7 43 1c 00 00 00 00 83 c4 0c 5b c3 c7 44 24 08 12 11 72 19 89 44 24 04 c7 04 24 b0 31 eb f8 e8 f8 75 27 c7 <0f> 0b eb fe 0f 0b eb fe 81 ec dc 00 00 00 89 44 24 24 89 e0 e8
EIP: [<f8eb23f8>] videobuf_dma_unmap+0x58/0x60 [videobuf_dma_sg] SS:ESP 0068:f2d9be34

Author:  langelgjm [ Tue Jul 22, 2008 3:22 pm ]
Post subject: 

Sorry, I don't have a solution for you, but I get the same error message and behavior. I have a Lifeview FlyVideo 3000. saa7134-alsa loads fine, but when I try to watch TV, this bug pops up. Here's my output:

Code:
magic mismatch: 0 (expected 19721112)
------------[ cut here ]------------
kernel BUG at /KNOPPIX.build/src/v4l/xc-test-d4f7804a393c/v4l/videobuf-dma-sg.c:282!
invalid opcode: 0000 [#1]
PREEMPT SMP
Modules linked in: saa7134_alsa saa7134 nvidia(P) autofs4 nfsd exportfs cx8800 cx88xx bttv btcx_risc lirc_i2c lirc_dev ipv6 af_packet fuse usbhid ff_memless pcmcia yenta_socket rsrc_nonstatic pcmcia_core aufs sbp2 ohci1394 ieee1394 usb_storage ohci_hcd ehci_hcd nvram wm8775 cx25840 snd_via82xx gameport snd_ac97_codec ac97_bus snd_pcm_oss snd_mixer_oss snd_pcm ivtv snd_page_alloc snd_mpu401_uart hostap_pci snd_seq_dummy snd_seq_oss hostap ieee80211_crypt_rtl saa7115 snd_seq_midi snd_seq_midi_event msp3400 snd_seq snd_timer tuner tea5767 snd_rawmidi tda8290 tda18271 tda827x orinoco_pci via686a tuner_xc2028 tda9887 tuner_simple mt20xx tea5761 hwmon orinoco hermes snd_seq_device compat_ioctl32 8250_pnp 8250 videobuf_dma_sg firmware_class serial_core videobuf_core i2c_algo_bit ir_kbd_i2c cx2341x ir_common i2c_viapro tveeprom via_agp videodev v4l2_common uhci_hcd snd agpgart i2c_core v4l1_compat shpchp pci_hotplug usbcore parport_pc parport prism2_pci p80211 soundcore pcspkr rtc_cmos rtc_core rtc_lib tsdev evdev
CPU:    0
EIP:    0060:[<f11293f8>]    Tainted: P        VLI
EFLAGS: 00010296   (2.6.23-chw-4 #1)
EIP is at videobuf_dma_unmap+0x58/0x60 [videobuf_dma_sg]
eax: 00000029   ebx: edfd00cc   ecx: ffffffff   edx: 00000000
esi: e4c8e160   edi: df856a00   ebp: e0aa5f44   esp: e0aa5e34
ds: 007b   es: 007b   fs: 00d8  gs: 0033  ss: 0068
Process mythbackend (pid: 5275, ti=e0aa4000 task=effb56d0 task.ti=e0aa4000)
Stack: f112a1b0 00000000 19721112 edfd0000 f1129411 00000000 00000000 e4c8e160
       ee0b6244 c14dd2f8 c14dd2e0 c068bc80 ee0b6000 eb5f8000 eff23400 e96a3c14
       00000000 ee3d30a4 ee3d3158 00000000 00000000 00000000 00000000 00000000
Call Trace:
 [<f1129411>] videobuf_pci_dma_unmap+0x11/0x20 [videobuf_dma_sg]
 [<f124e53b>] snd_pcm_oss_get_active_substream+0x5b/0x70 [snd_pcm_oss]
 [<c0567c1d>] __mutex_lock_slowpath+0x12d/0x290
 [<f1033561>] snd_card_saa7134_hw_free+0x41/0x70 [saa7134_alsa]
 [<f1b03f36>] snd_pcm_release_substream+0x36/0x70 [snd_pcm]
 [<f124e709>] snd_pcm_oss_release_file+0x19/0x30 [snd_pcm_oss]
 [<f124fee4>] snd_pcm_oss_release+0x44/0xa0 [snd_pcm_oss]
 [<c01854d2>] __fput+0xa2/0x1d0
 [<c01824b7>] filp_close+0x47/0x80
 [<c01838b9>] sys_close+0x69/0xc0
 [<c0104412>] syscall_call+0x7/0xb
 =======================
Code: c7 43 18 00 00 00 00 c7 43 1c 00 00 00 00 83 c4 0c 5b c3 c7 44 24 08 12 11 72 19 89 44 24 04 c7 04 24 b0 a1 12 f1 e8 f8 05 00 cf <0f> 0b eb fe 0f 0b eb fe 81 ec dc 00 00 00 89 44 24 24 89 e0 e8
EIP: [<f11293f8>] videobuf_dma_unmap+0x58/0x60 [videobuf_dma_sg] SS:ESP 0068:e0aa5e34

Author:  langelgjm [ Wed Jul 23, 2008 7:00 pm ]
Post subject:  kernel bug in videobuf-dma-sg

Just to give another data point, I got the newest version of all the v4l drivers (including saa7134, saa7134-alsa, and videobuf-dma-sg) and installed them by following the instructions here http://linuxtv.org/v4lwiki/index.php/How_to_build_from_Mercurial.

This had no effect, other than to change the line number on which the error occurs. Also, if I understand the error message correctly, the error isn't with saa7134-alsa, but with videobuf-dma-sg.

I thought about submitting a kernel bug report, but they have rules about only using vanilla kernels, and I don't really have the patience to put a new kernel on my machine. Instead I think I am going to take a look at what's changed between different version of this videobuf-dma-sg file, though I'm way out of my league here, so any help would be appreciated.

Code:
------------[ cut here ]------------
kernel BUG at /home/gjm/v4l-dvb/v4l/videobuf-dma-sg.c:272!
invalid opcode: 0000 [#1]
PREEMPT SMP
Modules linked in: saa7134_alsa autofs4 nfsd exportfs cx8800 cx88xx bttv btcx_risc lirc_i2c lirc_dev ipv6 af_packet fuse usbhid ff_memless pcmcia yenta_socket rsrc_nonstatic pcmcia_core aufs sbp2 ohci1394 ieee1394 usb_storage ohci_hcd ehci_hcd nvram hostap_pci hostap ieee80211_crypt_rtl tuner_simple tuner_types wm8775 cx25840 via686a hwmon snd_via82xx gameport snd_ac97_codec ac97_bus snd_pcm_oss snd_mixer_oss snd_pcm orinoco_pci orinoco hermes snd_page_alloc snd_mpu401_uart snd_seq_dummy snd_seq_oss ivtv snd_seq_midi snd_seq_midi_event saa7115 snd_seq msp3400 snd_timer snd_rawmidi tuner snd_seq_device 8250_pnp 8250 serial_core firmware_class i2c_algo_bit saa7134 ir_common cx2341x videodev v4l1_compat compat_ioctl32 v4l2_common i2c_viapro videobuf_dma_sg videobuf_core tveeprom uhci_hcd via_agp agpgart prism2_pci snd i2c_core p80211 shpchp pci_hotplug usbcore parport_pc parport soundcore pcspkr rtc_cmos rtc_core rtc_lib tsdev evdev
CPU:    0
EIP:    0060:[<f0c53438>]    Not tainted VLI
EFLAGS: 00010292   (2.6.23-chw-4 #1)
EIP is at videobuf_dma_unmap+0x58/0x60 [videobuf_dma_sg]
eax: 00000029   ebx: eb0150cc   ecx: ffffffff   edx: 00000000
esi: e1a23ae0   edi: df4bca00   ebp: d6b11f44   esp: d6b11e24
ds: 007b   es: 007b   fs: 00d8  gs: 0033  ss: 0068
Process mythbackend (pid: 3639, ti=d6b10000 task=d7fe0110 task.ti=d6b10000)
Stack: f0c541d8 00000000 19721112 eb015000 f0c53451 d6b11e9d f0d525a3 00000000
       e2e98620 00000000 00000000 e1a23ae0 ebd8d644 00000000 eff23448 e6f85ed4
       ebd8d400 e32bf400 e2e98620 00000af5 00000000 e25c98a4 e25c9958 00000000
Call Trace:
 [<f0c53451>] videobuf_sg_dma_unmap+0x11/0x20 [videobuf_dma_sg]
 [<f0d525a3>] snd_pcm_oss_change_params+0x1f3/0xd80 [snd_pcm_oss]
 [<f0d5353b>] snd_pcm_oss_get_active_substream+0x5b/0x70 [snd_pcm_oss]
 [<c0567c1d>] __mutex_lock_slowpath+0x12d/0x290
 [<f1743564>] snd_card_saa7134_hw_free+0x44/0x70 [saa7134_alsa]
 [<f0db2f36>] snd_pcm_release_substream+0x36/0x70 [snd_pcm]
 [<f0d53709>] snd_pcm_oss_release_file+0x19/0x30 [snd_pcm_oss]
 [<f0d54ee4>] snd_pcm_oss_release+0x44/0xa0 [snd_pcm_oss]
 [<c01854d2>] __fput+0xa2/0x1d0
 [<c01824b7>] filp_close+0x47/0x80
 [<c01838b9>] sys_close+0x69/0xc0
 [<c0104412>] syscall_call+0x7/0xb
 =======================
Code: c7 43 14 00 00 00 00 c7 43 18 00 00 00 00 83 c4 0c 5b c3 c7 44 24 08 12 11 72 19 89 44 24 04 c7 04 24 d8 41 c5 f0 e8 b8 65 4d cf <0f> 0b eb fe 0f 0b eb fe 81 ec ec 00 00 00 89 44 24 24 89 e0 e8
EIP: [<f0c53438>] videobuf_dma_unmap+0x58/0x60 [videobuf_dma_sg] SS:ESP 0068:d6b11e24

Author:  cliffsjunk [ Thu Jul 24, 2008 9:51 am ]
Post subject: 

Does this affect the Hauppauge PVR-XXX like the PVR-350?

Under R5F27, lsmod shows several saa7134 devices loaded, but
no saa7134-alsa. I suspect that this is because alsa controls
the ESS sound card that the PVR-350 feeds into (line-out on
the PVR-350 loops to line-in on the ESS sound card).

Can I install R5.5 on this PVR-350 system without causing a big
down time?

Edit: I did a test Auto Install of R5.5 on this PVR-350 system
and the sound worked fine (there were other issues though
that I have posted elsewhere).

Cliff

Author:  langelgjm [ Thu Jul 24, 2008 9:56 am ]
Post subject: 

Well, I have a Hauppauge PVR-150 that works without issue. I think some of the Hauppauge cards are based on the saa7134, but I'm pretty sure the decent ones (hardware MPEG-2 encoding) aren't, so they shouldn't have any problems with this.

My issue is specifically with saa7134-alsa, not saa7134, so if you don't have the former loaded, I'm guessing you don't have anything to worry about.

I have also reported the bug on the v4l mailing list. I've tried building a bunch of different versions of the v4l drivers, but no luck yet.

Page 1 of 1 All times are UTC - 6 hours
Powered by phpBB® Forum Software © phpBB Group
http://www.phpbb.com/