I've been running Myth for almost a year now without any major issues. In May pulled a 120 GB Seagate HD and replaced it with two 200 GB Seagate (XFS LVM for /myth) drives and a new Promise ATA 133 controller. I thought the machine was working fine but recently it has become extremely unstable.

I can watch 3-4 streams on differnt frontend machines fine, but once i try to watch live tv on the main backend machine or record a new program, ivtv has DMA errors, which seem to cascade to all of my PCI devices.

I've tried installing a newer version of ivtv (0.4.6) and compiled a new kernel without SMP (2.6.15).

From some other lists, I've tried to disable PIO on the drives and downgrade them to UDMA4 (from 5). I've increased the latency of the IDE controller, all without solving the problem. I don't believe there are any problems with the drives, since I ran badblocks on both without any errors.

The old Seagate was connected to the motherboard, which is ATA100, w/o the 48-bit addressing, so I had to buy and use a new card for the new drives.

Any help would be appreciated.

Thanks,
imolloy

System:
ASUS P4B266 Mobo
2.0 GHz P4
512 MB RAM
NVidia GeForce 4 440 MX
Hauppauge WinTV PVR-250
Promise ULTRA133TX2
2x ST3200822A Seagate Barracude 200GB 7200.7

From messages
Jul 30 19:08:54 mythtv kernel: ivtv0 warning: IRQ: IVTV_IRQ_DEC_DMA_ERR
Jul 30 19:09:15 mythtv kernel: hdf: dma_timer_expiry: dma status == 0x61
Jul 30 19:09:25 mythtv kernel: hdf: DMA timeout error
Jul 30 19:09:25 mythtv kernel: hdf: dma timeout error: status=0x58 { DriveReady SeekComplete DataRequest }
Jul 30 19:09:25 mythtv kernel: ide: failed opcode was: unknown
Jul 30 19:09:45 mythtv kernel: hde: dma_timer_expiry: dma status == 0x21
Jul 30 19:09:55 mythtv kernel: hde: DMA timeout error
Jul 30 19:09:55 mythtv kernel: hde: dma timeout error: status=0x58 { DriveReady SeekComplete DataRequest }
Jul 30 19:09:55 mythtv kernel: ide: failed opcode was: unknown
Jul 30 19:09:55 mythtv kernel: hdf: status error: status=0x50 { DriveReady SeekComplete }
Jul 30 19:09:55 mythtv kernel: ide: failed opcode was: unknown
Jul 30 19:09:55 mythtv kernel: hde: status error: status=0x50 { DriveReady SeekComplete }
Jul 30 19:09:55 mythtv kernel: ide: failed opcode was: unknown
Jul 30 19:09:55 mythtv kernel: hdf: status timeout: status=0xd0 { Busy }
Jul 30 19:09:55 mythtv kernel: ide: failed opcode was: unknown
Jul 30 19:09:55 mythtv kernel: PDC202XX: Primary channel reset.
Jul 30 19:09:55 mythtv kernel: ide2: reset: success
Jul 30 19:09:59 mythtv kernel: NETDEV WATCHDOG: eth0: transmit timed out


IVTV from dmsg
ivtv: ==================== START INIT IVTV ====================
ivtv: version 0.4.6 (tagged release) loading
ivtv: Linux version: 2.6.15-chw-nosmp preempt 586 gcc-3.3
ivtv: In case of problems please include the debug info between
ivtv: the START INIT IVTV and END INIT IVTV lines, along with
ivtv: any module options, when mailing the ivtv-users mailinglist.
ivtv0: Autodetected WinTV PVR 250 card (cx23416 based)
PCI: Enabling device 0000:02:0a.0 (0014 -> 0016)
PCI: IRQ 0 for device 0000:02:0a.0 doesn't match PIRQ mask - try pci=usepirqmask
PCI: setting IRQ 9 as level-triggered
PCI: Assigned IRQ 9 for device 0000:02:0a.0
ivtv0: Unreasonably low latency timer, setting to 64 (was 32)
tveeprom: ivtv version
tveeprom: Hauppauge: model = 32062, rev = C199, serial# = 8054556
tveeprom: tuner = TCL 2002N 5H (idx = 99, type = 50)
tveeprom: tuner fmt = NTSC(M) (eeprom = 0x08, v4l2 = 0x00001000)
tveeprom: audio processor = MSP3445 (type = c)
tveeprom: decoder processor = SAA7115 (type = 13)
ivtv0: i2c attach to card #0 ok [client=tveeprom, addr=50]
tuner (ivtv): chip found at addr 0xc2 i2c-bus ivtv i2c driver #0
ivtv0: i2c attach to card #0 ok [client=(tuner unset), addr=61]
msp3400 1-0040: chip=MSP3445G-B8 +nicam +simple +simpler +radio mode=simpler
msp3400 1-0040: msp34xxg daemon started
ivtv0: i2c attach to card #0 ok [client=MSP3445G-B8, addr=40]
saa7115 1-0021: saa7115 found @ 0x42 (ivtv i2c driver #0)
ivtv0: i2c attach to card #0 ok [client=saa7115, addr=21]
ivtv0: loaded v4l-cx2341x-enc.fw firmware (262144 bytes)
ivtv0: Encoder revision: 0x02050032
ivtv0: Allocate DMA encoder MPEG stream: 128 x 32768 buffers (4096KB total)
ivtv0: Allocate DMA encoder YUV stream: 194 x 10800 buffers (2048KB total)
ivtv0: Allocate DMA encoder VBI stream: 120 x 17472 buffers (2048KB total)
ivtv0: Allocate DMA encoder PCM audio stream: 455 x 4608 buffers (2048KB total)
tuner: type set to 50 (TCL 2002N) by ivtv i2c driver #0
ivtv0: Initialized WinTV PVR 250, card #0
ivtv: ==================== END INIT IVTV ====================

root@mythtv:/var/log# cat /proc/interrupts
CPU0
0: 811765 XT-PIC timer
1: 10 XT-PIC i8042
2: 0 XT-PIC cascade
3: 4569 XT-PIC eth0
4: 0 XT-PIC uhci_hcd:usb2
5: 0 XT-PIC CMI8738-MC6
8: 1 XT-PIC rtc
9: 0 XT-PIC ivtv0
10: 2770 XT-PIC ide2
11: 44283 XT-PIC nvidia
12: 100 XT-PIC uhci_hcd:usb1
14: 6737 XT-PIC ide0
15: 79 XT-PIC ide1
NMI: 0
ERR: 0

A couple things to try for diagnostic purposes. Possibly taking coals to Newcastle...

- Check what hdparm has to say about the drives. Both basic settings using "-v" and timings using "-Tt".

- Check your memory, either using memtest or booting from the CD and use the memory checking utility there (the name esacpes me). The later is preferable because it operates at the lowest level.

I also seem to remember postings from folks who had problems with the Promise controllers before. Searching for that as a keyword might turn up something...

I tried searching a lot, and most of the solutions involved turning off PIO, downgrading the UDMA to 4, or the thread ended without any solution. The other thing I noticed is all six of my PCI cards are on the same bus. I heard some people say this caused DMA problems. Any truth to this?

Both hde and hdf are identical.

root@mythtv:/myth/tmp# hdparm -v /dev/hde

/dev/hde:
multcount = 16 (on)
IO_support = 0 (default 16-bit)
unmaskirq = 0 (off)
using_dma = 1 (on)
keepsettings = 0 (off)
readonly = 0 (off)
readahead = 256 (on)
geometry = 24321/255/63, sectors = 200049647616, start = 0
root@mythtv:/myth/tmp# hdparm -i /dev/hdf

/dev/hdf:

Model=ST3200822A, FwRev=3.01, SerialNo=3LJ05KBX
Config={ HardSect NotMFM HdSw>15uSec Fixed DTR>10Mbs RotSpdTol>.5% }
RawCHS=16383/16/63, TrkSize=0, SectSize=0, ECCbytes=4
BuffType=unknown, BuffSize=8192kB, MaxMultSect=16, MultSect=16
CurCHS=16383/16/63, CurSects=16514064, LBA=yes, LBAsects=268435455
IORDY=on/off, tPIO={min:240,w/IORDY:120}, tDMA={min:120,rec:120}
PIO modes: pio0 pio1 pio2 pio3 pio4
DMA modes: mdma0 mdma1 mdma2
UDMA modes: udma0 udma1 udma2 udma3 udma4 *udma5
AdvancedPM=no WriteCache=enabled
Drive conforms to: ATA/ATAPI-6 T13 1410D revision 2:

* signifies the current active mode
root@mythtv:/myth/tmp# hdparm -Tt /dev/hde

/dev/hde:
Timing cached reads: 1376 MB in 2.00 seconds = 687.07 MB/sec
Timing buffered disk reads: 166 MB in 3.03 seconds = 54.72 MB/sec
root@mythtv:/myth/tmp# hdparm -Tt /dev/hdf

/dev/hdf:
Timing cached reads: 1464 MB in 2.00 seconds = 730.65 MB/sec
Timing buffered disk reads: 184 MB in 3.02 seconds = 60.86 MB/sec

I booted the machine off a memtest cd and all of my RAM appears to be perfectly fine.

Also, my machine appears to run fine (but very, very choppy) if I completely turn DMA off on hde and hdf.

Is it possible to tell ivtv to run my PVR-250 in PIO mode, and would that help the situation and be fast enough to capture SD at full resolution?

There has got to be a way to solve these DMA errors...

Anyone successfully resolved DMA problems like this?

Make sure that the fans in your case are working.

Because it was working then became unstable, I don't think that it is a tuning issue. It looks
to me like a hardware issue.. like one of your drives is dying.

I would make sure that I had good a backup.