Open thonatron opened 3 years ago
Ive just had the a tuner die and me again with the other one still working, the only thing in dmesg between last known working and tuner been gone is this
[17558.772023] si2168 7-0064: downloading firmware from file 'dvb-demod-si2168-b40-01.fw'
[17559.459250] si2168 7-0064: firmware version: B 4.0.11
[17559.467044] si2157 8-0060: found a 'Silicon Labs Si2157-A30'
[17559.489318] si2157 8-0060: firmware version: 3.0.5
[17837.335797] perf: interrupt took too long (2503 > 2500), lowering kernel.perf_event_max_sample_rate to 79750
[23727.218983] kauditd_printk_skb: 19 callbacks suppressed
[23727.218985] audit: type=1400 audit(1612742401.084:31): apparmor="DENIED" operation="capable" profile="/usr/sbin/cups-browsed" pid=22052 comm="cups-browsed" capability=23 capname="sys_nice"
[28106.267154] si2168 2-0064: downloading firmware from file 'dvb-demod-si2168-b40-01.fw'
[28106.961138] si2168 2-0064: firmware version: B 4.0.11
[28106.968912] si2157 3-0060: found a 'Silicon Labs Si2157-A30'
[28106.991198] si2157 3-0060: firmware version: 3.0.5
[34135.110378] perf: interrupt took too long (3137 > 3128), lowering kernel.perf_event_max_sample_rate to 63750
[59399.157784] si2168 2-0064: downloading firmware from file 'dvb-demod-si2168-b40-01.fw'
[59399.851997] si2168 2-0064: firmware version: B 4.0.11
[59399.859771] si2157 3-0060: found a 'Silicon Labs Si2157-A30'
[59399.882056] si2157 3-0060: firmware version: 3.0.5
lspci shows the card still there but the rev is now FF instead of 04
05:00.0 Ethernet controller: Intel Corporation I210 Gigabit Network Connection (rev 03)
06:00.0 PCI bridge: Pericom Semiconductor PI7C9X2G304 EL/SL PCIe2 3-Port/4-Lane Packet Switch (rev 05)
07:01.0 PCI bridge: Pericom Semiconductor PI7C9X2G304 EL/SL PCIe2 3-Port/4-Lane Packet Switch (rev 05)
07:02.0 PCI bridge: Pericom Semiconductor PI7C9X2G304 EL/SL PCIe2 3-Port/4-Lane Packet Switch (rev 05)
08:00.0 Multimedia video controller: Conexant Systems, Inc. CX23887/8 PCIe Broadcast Audio and Video Decoder with 3D Comb (rev 04)
09:00.0 Multimedia video controller: Conexant Systems, Inc. CX23887/8 PCIe Broadcast Audio and Video Decoder with 3D Comb (rev ff)
lscpi -vvv
08:00.0 Multimedia video controller: Conexant Systems, Inc. CX23887/8 PCIe Broadcast Audio and Video Decoder with 3D Comb (rev 04)
Subsystem: Hauppauge computer works Inc. CX23887/8 PCIe Broadcast Audio and Video Decoder with 3D Comb
Control: I/O- Mem+ BusMaster+ SpecCycle- MemWINV- VGASnoop- ParErr- Stepping- SERR- FastB2B- DisINTx-
Status: Cap+ 66MHz- UDF- FastB2B- ParErr- DEVSEL=fast >TAbort- <TAbort- <MAbort- >SERR- <PERR- INTx-
Latency: 0, Cache Line Size: 64 bytes
Interrupt: pin A routed to IRQ 17
Region 0: Memory at f7400000 (64-bit, non-prefetchable) [size=2M]
Capabilities: <access denied>
Kernel driver in use: cx23885
Kernel modules: cx23885
09:00.0 Multimedia video controller: Conexant Systems, Inc. CX23887/8 PCIe Broadcast Audio and Video Decoder with 3D Comb (rev ff) (prog-if ff)
!!! Unknown header type 7f
Kernel driver in use: cx23885
Kernel modules: cx23885
rebooting gives
lspci
05:00.0 Ethernet controller: Intel Corporation I210 Gigabit Network Connection (rev 03)
06:00.0 PCI bridge: Pericom Semiconductor PI7C9X2G304 EL/SL PCIe2 3-Port/4-Lane Packet Switch (rev 05)
07:01.0 PCI bridge: Pericom Semiconductor PI7C9X2G304 EL/SL PCIe2 3-Port/4-Lane Packet Switch (rev 05)
07:02.0 PCI bridge: Pericom Semiconductor PI7C9X2G304 EL/SL PCIe2 3-Port/4-Lane Packet Switch (rev 05)
08:00.0 Multimedia video controller: Conexant Systems, Inc. CX23887/8 PCIe Broadcast Audio and Video Decoder with 3D Comb (rev 04)
Now something different happens, i thought id try to see what bios options are available, rebooted spammed F2 to enter Bios over ipmi, and the board just hangs at entering setup. Rebooted and this time let it fully boot while i figure out whats going on with entering setup, and both cards are back working, didnt require a cold boot. So thinking that the card is somehow stopping me entering the bios, i reboot again and im now in the bios. So crashed card -> reboot -> no bios not crashed card -> reboot -> bios crashed card -> reboot -> reboot whilst rebooting -> card fine
Ive found a setting in bios about logging pcie errors to either the bios or ipmi so ill see how that works out
Have you tested the card under Windows to see if exhibit any of the same problem by using PiP mode or with SageTV where you can test each of the 4 tuner.
Unfortunately i don't have any windows machines, but mythtv does allow me to switch tuners, havent tried PiP yet. My bios pcie debugging option did bring some good(?) news
Event ID | Time Stamp | Sensor Name | Sensor Type | Description
-- | -- | -- | -- | --
1 | 2021/02/07 17:26:10 | OEM | OOB | Activate Node Product Key - Assertion
2 | 2021/02/09 03:58:06 | #0x00 | Critical Interrupt | PCI SERR - Assertion@Bus08(DevFn00)
3 | 2021/02/09 08:27:22 | #0x00 | Critical Interrupt | PCI SERR - Assertion@Bus07(DevFn10)
At least i now have a timestamp of what time each one error-ed out, not that i know what it means yet. At 03:58 and 08:27 mythtv wasnt doing anything to do with tuning, the cards are set to open on demand
Feb 9 04:32:29 Thor mythbackend: mythbackend[2776]: N Expire autoexpire.cpp:241 (CalcParams) AutoExpire: CalcParams(): Max required Free Space: 100.0 GB w/freq: 15 min
Feb 9 04:47:29 Thor mythbackend: mythbackend[2776]: N Expire autoexpire.cpp:241 (CalcParams) AutoExpire: CalcParams(): Max required Free Space: 100.0 GB w/freq: 15 min
Feb 9 05:02:29 Thor mythbackend: mythbackend[2776]: N Expire autoexpire.cpp:241 (CalcParams) AutoExpire: CalcParams(): Max required Free Space: 100.0 GB w/freq: 15 min
...
Feb 9 08:06:55 Thor mythbackend: mythbackend[2776]: I HouseKeeping housekeeper.cpp:159 (Run) HouseKeeperTask 'RecordedArtworkUpdate' Finished Successfully.
Feb 9 08:17:30 Thor mythbackend: mythbackend[2776]: N Expire autoexpire.cpp:241 (CalcParams) AutoExpire: CalcParams(): Max required Free Space: 100.0 GB w/freq: 15 min
Feb 9 08:32:30 Thor mythbackend: mythbackend[2776]: N Expire autoexpire.cpp:241 (CalcParams) AutoExpire: CalcParams(): Max required Free Space: 100.0 GB w/freq: 15 min
Feb 9 08:47:30 Thor mythbackend: mythbackend[2776]: N Expire autoexpire.cpp:241 (CalcParams) AutoExpire: CalcParams(): Max required Free Space: 100.0 GB w/freq: 15 min
Feb 9 09:02:30 Thor mythbackend: mythbackend[2776]: N Expire autoexpire.cpp:241 (CalcParams) AutoExpire: CalcParams(): Max required Free Space: 100.0 GB w/freq: 15 min
dmesg still doesnt give anything which is not surprising since since it still thinks the cards are there
ls -al /dev/dvb/
total 0
drwxr-xr-x 6 root root 120 Feb 8 11:46 .
drwxr-xr-x 20 root root 5000 Feb 8 11:46 ..
drwxr-xr-x 2 root root 120 Feb 8 11:46 adapter0
drwxr-xr-x 2 root root 120 Feb 8 11:46 adapter1
drwxr-xr-x 2 root root 120 Feb 8 11:46 adapter2
drwxr-xr-x 2 root root 120 Feb 8 11:46 adapter3
but both are dead by the (rev ff) in lspci
lspci
06:00.0 PCI bridge: Pericom Semiconductor PI7C9X2G304 EL/SL PCIe2 3-Port/4-Lane Packet Switch (rev 05)
07:01.0 PCI bridge: Pericom Semiconductor PI7C9X2G304 EL/SL PCIe2 3-Port/4-Lane Packet Switch (rev 05)
07:02.0 PCI bridge: Pericom Semiconductor PI7C9X2G304 EL/SL PCIe2 3-Port/4-Lane Packet Switch (rev 05)
08:00.0 Multimedia video controller: Conexant Systems, Inc. CX23887/8 PCIe Broadcast Audio and Video Decoder with 3D Comb (rev ff)
09:00.0 Multimedia video controller: Conexant Systems, Inc. CX23887/8 PCIe Broadcast Audio and Video Decoder with 3D Comb (rev ff)
Ive put in a request with supermicro about the PCI SERR to try and find out what that means, hopefully ill get some good news but the board is now EOL. Im going to see if i can put the card in another computer, but this might take a bit of time
Keep in mind that consumer TV tuner where never meant for server system I don't know if this well help but in the bios try disable PCI SERR Maybe when brad come back in he will have some better suggestion for you.
ive got in in one of the PCIE3x8 slots now, it wasnt visible, in fact i couldnt get in to the bios, until i removed it and set the slot to gen1. Now i can get in the bios see the card and linux can see the card, ill just have to wait to see if it falls again, but its not ideal as my hba is in a slower PCIE2x4 slot, luckily ive only got 4 HDD attached to it at the moment. The card is described as
PCI bridge: Pericom Semiconductor PI7C9X2G304 EL/SL PCIe2 3-Port/4-Lane Packet Switch (rev 05)
which im thinking means its PCIe2 so it should work fine in a gen3 slot unless im missing something really obvious
If it's pcie2 it will be perfectly fine in a pcie3 slot. On intel systems, for whatever strange reason, their implementation of pcie requires to be set to gen1. The card should continue to work without issue for you that way.
is there anyway to make it work in the pcie2 slot if i cant set the gen on the slot, ive only got the option to set gen on the 2 pcie3 slots
This is specifically an Intel issue in this regard. I don't know what they are doing electrically or systematically to make the boards behave as such. As you've seen when the pcie chipset becomes unhappy the board just disappears. This is not a driver issue, but some inherent issue in the Intel chipset.
little update, it eventually managed to error out again, The log in the ipmi did come up with about 15 of the serr# errors before it did drop off so i guess thats an improvement of sorts. Supermicro got back to me, but as the board is EOL not a lot could be done. In the original manual it mentions Serr and perr disable, but the latest bios doesn't have those options, supermicro provided me with some old bios files but the i got to a point where the m.2 drive in a pcie caddy wasn't recognized in so it wouldn't boot for me. For the moment ive pulled the card, looking to test it in this desktop, but the pciex1 slot is blocked by a graphics card. so for now it looks like supermicro x10slm-f and a quadhd is a no go.
Ive been struggling to to get this card working for about a week now, from a cold boot it works fine for a while then eventually it silently disappears from the computer. A warm reboot of the computer leaves the card not found on the next boot, but the pcie splitter/bridge is still seen, just nothing that is attached to it i have
ive tried setting this to 0 or 2 and letting the driver handle it automatically but at some point it always seams to crash out and have no tuners found. Dmesg gives nothing useful, nothing about why the card just disappears, i must need debugging enabled somewhere. The card is in a gen2 pciex4 slot, (i cant set it gen1), i could try the pcie3 slots but i have a nvme drive and a hba card in them so its not realistic option to move it to one of those slots I moved over to this kernel from the default one in ubuntu 20.04LTS due to the errors about mpeg risc op code that kept crashing the card. The main use of the card is with mythtv, The problem usually starts with a failure to get a signal lock, but interestingly if i switch to a different tuner pair from (1,2) -> (3,4) or the other way, that tuner is usually working, until it doesnt work, then i cant get a signal at all. Then to server need at cold start like mentioned above. I think i have the intel Vt-d extensions enabled as this computer is going to be running some vms once i get this card and mythtv playing nice, so id rather not have to disable those
cpu motherboard
dmesg when card is working
lspci from cold boot
lspci from warm boot: