| ↑back Search ←Prev date Next date→ Show only urls | (Click on time to select a line by its url) |
| Who | What | When |
|---|---|---|
| dongs | mrec_: pretty sure chehab has you on ignore
at least, I dont see him replying to you :) | [02:26] |
| ........ (idle for 36mn) | ||
| mrec_ | because he has nothing to say heh
he knows he's a looser I don't depend on him since we have our own environment our solution is evolving and works with windows/mac/linux/freebsd (freebsd still has some issues but the freebsd developers are fixing it now) what should I do with an island solution for linux even that kaffeine shit only works with linux freebsd soso. but fbsd is only for marketing purpose anyway heh | [03:02] |
| ............. (idle for 1h1mn) | ||
| *** | awalls- has left "Leaving" | [04:08] |
| ........................................................................................................ (idle for 8h35mn) | ||
| Pepe | mrec_ still you canescalate your issue if you want. Try phoronix and other resources, too. | [12:43] |
| .... (idle for 17mn) | ||
| mchehab | Pepe, just do like I did: add mrec_ to your ignore list. he's a known troll | [13:00] |
| Pepe | I don't want to be between you or even understand relations between you two. I would like to have working driver for both my tuners, which I mention and if you would like I can or someone else borrow my tuners to get it fixed. :-)
That's why I joined this channel to get help or ask somebody if he has some problem like I do. or advices how to get it fixed. I know that community around Linux can't try all drivers. or even to have them | [13:01] |
| mchehab | Pepe, did you try upstream Kernel? | [13:06] |
| Pepe | I didn't. Will try then. | [13:06] |
| mchehab | there are some fixes that got merged on 4.10 and a few ones still pending
they change the drivers to not do DMA on stack, as this is not allowed anymore unfortunately, some old drivers do that Kernel 4.9 now forbids it, by default because of a patch meant to improve its security, placing the linux stack on a region of memory that doesn't allow DMA if you disable CONFIG_VMAP_STACK, the driver will work on Kernel 4.9 but patches addressing it are being posted and applied | [13:06] |
| Pepe | so I would disable it and it should work, right? and if it didnt I should get back to you, right?
or it's better to try upstream kernel? | [13:10] |
| mchehab | for me, it is better if you try upstream Kernel, as it helps to check if the fixes you need are there already
and did their job | [13:12] |
| Pepe | Ok. I will do that. | [13:13] |
| mchehab | hmm... you're trying to run it also on Odroid
on ARM, DMA on stack never worked there, you'll really need to work with the upstream fixes | [13:13] |
| Pepe | I tried it with old drivers and old kernel there. Didnt try there upstream drivers or even kernel.
but the issue with flooding system was there. | [13:14] |
| mchehab | ok, let's go by parts. first, test it on x86, with upstream Kernel (upstream from the media tree, I mean)
From your e-mail: "My tuner itself is disconnecting and connecting back in this case tvheadend doesn't seen it and I need to restart tvheadend." | [13:15] |
| Pepe | You confused me now. Media tree or upstream kernel? Upstream media tree I tried, but I didnt try upstream kernel.
yes, that's what upstream drivers did | [13:16] |
| mchehab | ok | [13:16] |
| Pepe | and sometimes they're flooding system. | [13:16] |
| mchehab | media tree
the above message looks more like a hardware problem are you connecting the USB stick directly or are you using some cable? | [13:16] |
| Pepe | Tuner has slot (I think output) for CAM, USB-B and external power. I connect there usb-b and to my computer USB A. | [13:20] |
| mchehab | are you using the external power supply? | [13:20] |
| Pepe | yes | [13:20] |
| mchehab | try to replace the USB A-B cable by another one
I found a lot of issues related to it yet, the driver should not have issues with too fast connect/disconnect, but avoiding this type of race conditions is not easy | [13:21] |
| Pepe | I tried that many times, but will do that. I think this can prevent disconnecting and connecting, but it works now on router (where I have them plugged and it seems stable, but I cant remove USBs, because it will flood system) | [13:22] |
| mchehab | the dvb-usb-v2 has less race conditions, as crope_ did lots of efforts to solve race conditions...
but not all drivers were migrated to it Btw, this message: sysfs: cannot create duplicate filename '/class/dvb/dvb0.ca0 | [13:23] |
| Pepe | was solved by upstream drivers, but disconnecting itself was caused I think with upstream driver, because as I mention on older kernel / driver it works w/o issues (except flooding) | [13:25] |
| mchehab | usually indicates that a device was removed, but the sysfs data was not removed yet when a new DVB device gets added with the same address
That's the flooding you're referring to? [ 4285.212518] usb read operation failed. (-71) #define EPROTO 71 /* Protocol error */ that usually happens when the device is not there anymore (e. g. the USB stack is not capable of talking with the device) uually, due to a physical condition from USB Kernel documentation: -EPROTO (*, **) a) bitstuff error b) no response packet received within the prescribed bus turn-around time c) unknown USB error | [13:26] |
| Pepe | yes this was. | [13:28] |
| mchehab | (*) Error codes like -EPROTO, -EILSEQ and -EOVERFLOW normally indicate
hardware problems such as bad devices (including firmware) or cables. | [13:28] |
| Pepe | yes this is it* | [13:28] |
| mchehab | either your hardware got damaged or your cables
perhaps some electrical discharge hit your cabling btw, I have a device here that returns this error if, instead of connecting to my LNBf cabling, I connect, by mistake, on a TV antenna that's because, on DVB/S, it sends a 13V or 18V DC power to feed the LNBf a normal TV cabling works as a short-circuit, with causes the power supply of the device to fail to feed it causing successive disconnections try to remove the antenna cable from your device and see if you notice disconnects | [13:29] |
| if the disconnects disapear, then there's something wrong with either your antenna cabling, or there's some other equipment short-circuiting the DC power
(that could happen, for example, if you connect two equipments/boards at the antena cabling without a DISEqC switch between them isolating DC) | [13:36] | |
| iive | Pepe: what is your device again? | [13:38] |
| Pepe | ok.. it's bad idea to trying it right now, when I use my internet :-D
I tried to remove antenna cables from both devices and there wasnt any disconnect Disconnect happened when I removed USB cable from router, which the other end is connect to my device | [13:38] |
| mchehab | ok, that's the normal behavior | [13:39] |
| Pepe | iive mainly: ID 0b48:300d TechnoTrend AG TT-connect CT-3650 CI | [13:39] |
| mchehab | so, it indeed seems to be a problem on your antenna cabling | [13:39] |
| Pepe | and I also mention usb read operation failed in last paragraph also in my other device ( ID 14f7:0003 TechniSat Digital GmbH CableStar Combo HD CI )
and what can I do when my devices are dual tuners? One for dvb-t and second on for dvb-c? In tvheadend I disabled dvb-t | [13:39] |
| iive | Pepe: some demodulators can handle dvb-t or dvb-c , it's not dual tuner unless you can use them both at once. | [13:42] |
| Pepe | ok, it's not, right. | [13:43] |
| iive | :) | [13:44] |
| mchehab | your ID 14f7:0003 TechniSat device seems to be using: drivers/media/usb/dvb-usb-v2/az6007.c
I'm reading its code to see if the USB stack issue was solved on it apparently yes, but let me do a review | [13:44] |
| iive | Pepe: does your device work on PC machine?
Pepe: if on PC, does it get all its power from the USB? | [13:45] |
| Pepe | iive I would like to mention *again* this my both tuners are working w/o any issues on my router. On C2 they were working, too. In both cases I can't unplug them otherwise it will flood system and everything i slow. I also try it on Intel J1900 (which I have as NAS) and they were working too on kernel 3.16. I really appreciate that you're trying to help me, but issues are related to kernel 4.9 (I cant have
old kernel , because OpenMediaVault and their support for ZFS type on file system) and upstream drivers, where you can see that CT-3650 sometimes is disconnecting itself and I need to restart tvheadend everytime (which I also report that there or atleast suggest to have there button for reload tuners) and as you can see I have two different bugtraces one seems to related to one device and second one to | [13:48] |
| mchehab | Pepe: on media tree, az6007 is doing the right thing with regards to DMA stack | [13:48] |
| Pepe | another.
iive and I can live w/o unplugging them or plugging them back, but it's annoying. :/ weird, because az6007 is doing this: http://pastebin.com/RMF67uez I googled it more but it seems related to this email: https://lkml.org/lkml/2014/2/28/428 | [13:48] |
| iive | Pepe: is your router an old PC, or arm/mips wifi router with usb port and custom linux? | [13:51] |
| mchehab | I have some az6007 devices here | [13:51] |
| iive | also, tvheadend should have support for plug-in/out of devices. | [13:52] |
| mchehab | let me see if I can reproduce the issue | [13:52] |
| iive | what version of tvheadend are you using? | [13:52] |
| Pepe | Take your time, please.
iive: https://tvheadend.org/issues/4267 was tested on Turris 1.1 (which is my router, it has CPU / SoC as you want powerpc P2020) and also on my J1900. iive mpmc told me to try also w_scan if it is working when CT-3650 itself disconnected, but I didnt try it, yet. | [13:52] |
| iive | ok, it's quite recent :) | [13:54] |
| Pepe | hm as I see also this one http://lkml.iu.edu/hypermail/linux/kernel/1312.0/05176.html it possible that I tried another usb port in my PC and I plugged it into usb 3.0. Will try it again on USB 2.0 (I mention some weird behaviour in my email, that I didnt notice that previously.) | [13:58] |
| mchehab | [ 7455.027440] usb 1-1: dvb_usb_v2: found a 'Terratec H7' in cold state
[ 7455.028545] usb 1-1: dvb_usb_v2: downloading firmware from file 'dvb-usb-terratec-h7-az6007.fw' [ 7455.034049] usb 1-1: dvb_usb_v2: found a 'Terratec H7' in warm state [ 7456.313162] usb 1-1: dvb_usb_v2: will pass the complete MPEG2 transport stream to the software demuxer [ 7456.313189] dvbdev: DVB: registering new adapter (Terratec H7) [ 7456.316067] usb 1-1: dvb_usb_v2: MAC address: c2:cd:0c:b4:10:00 [ 7456.367984] drxk: status = 0x639130d9 [ 7456.367987] drxk: detected a drx-3913k, spin A3, xtal 27.000 MHz [ 7458.597955] drxk: DRXK driver version 0.9.4300 [ 7458.690792] drxk: frontend initialized. [ 7458.690906] usb 1-1: DVB: registering adapter 0 frontend 0 (DRXK DVB-C DVB-T)... [ 7458.707424] mt2063_attach: Attaching MT2063 | [14:01] |
| iive | the device comes with 12V power adapter, so it always needs external power. | [14:01] |
| mchehab | so far, so good
let me see if I have the adapter for antenna | [14:01] |
| Pepe | Do you have it connected to usb 3.0 or usb 2.0? | [14:02] |
| mchehab | not sure | [14:03] |
| Pepe | Maybe it happens only on USB 3.0, which I will confirm later. I can't restart router now. :/ | [14:04] |
| mchehab | I connected to the other USB port
one is USB 2 the other is 3.0 on both, it started fine [ 7692.125894] usb 1-2: dvb_usb_v2: 'Terratec H7' successfully initialized and connected | [14:05] |
| Pepe | Start is fine, but try to watch something or record something and stop it after few sec or maybe unplug it. | [14:06] |
| mchehab | let me turn on my RF generator to see if it actualyl works
unpluging worked fine | [14:06] |
| Pepe | ok. :/
But I found same issue drxk: write_block: i2c write error at addr 0x831ffd in these two emails which are related to usb 3.0 ports maybe I thought that this should be fixed. | [14:06] |
| mchehab | scan worked fine
Scanning frequency #1 410000000 Lock (0x1f) Quality= Ok Signal= 100.00% C/N= 0.00dB UCB= 12 postBER= 5.39x10^-6 preBER= 101x10^-6 PER= 977x10^-6 Service Das Erste HD, provider MEDIA BROADCAST: reserved ... - Modulation Type : DVB-T nothing unusual at dmesg | [14:08] |
| Pepe | It's good for you, but not for me XD | [14:10] |
| mchehab | worked on both USB ports of my test machine
NUC5i7RYB frontal ports (one is USB 2, the other one USB 3) let update the Kernel to the latest upstream version... currently, it is with 4.10-rc7 + media stuff | [14:10] |
| iive | 4.10.1 has already been released :) | [14:12] |
| Pepe | I will try that. | [14:12] |
| mchehab | btw, I'm using a brand new USB A-B cable there
by coincidence, I bought one those days, and was under my desk :-) | [14:13] |
| Pepe | Iˇm using Belkin CU1000cp 3M or 1.8M(or maybe even 1.5M) | [14:14] |
| mchehab | I'm building 4.10 + media patches submitted for 4.11
should take a while... this is just an i5 machine I opted to remove the last patch from media (with merges from 4.11-rc1) in order to avoid any noise that usually happens on early -rc cycles so, my head is at: 3dc2046ca78b [media] media: ti-vpe: vpe: allow use of user specified stride | [14:15] |
| mrec_ | if I'm a troll mchehab is a twat ;-)
but he's a twat anyway breaking other's work we're not breaking anything and that's fact enough | [14:21] |
| Pepe | I know it's hard to fix something, which you can't reproduce. :/ I will try later if it works w/o any issues on USB 2.0 and when yes, I will be fine with that otherwise two bugtraces annoys me more and it seems they're related to tuners. WHen I dont have plugged them I dont see them.
according what I found in my log I have same head ( 3dc2046ca78b3cac6c8c9098a4e3024cd91abdb4 [media] media: ti-vpe: vpe: allow use of user specified stride) and kernel provided from Debian | [14:22] |
| mchehab | ok. I'll let you know what happens after having it compiled
with such version with regards to USB, some USB controllers have issues. That could indeed explain the differences, but before assuming it, let me test it here | [14:25] |
| Pepe | I will let you, dont worry. :-) I'm happy and also glad that you're trying to find the cause. | [14:27] |
| mchehab | kernel built. rebooting
worked on both ports | [14:29] |
| Pepe | hm it really seems like issue with USB controller. :/ | [14:33] |
| mchehab | it looks so | [14:33] |
| Pepe | and is there any way how to prevent CT-3650 from restarting or even how to get these bug traces fixed? | [14:33] |
| mchehab | let's go by parts:
1) the issue with your az6007-based board can be due to some trouble at the USB controller - or some bad cabling | [14:34] |
| Pepe | and sorry for asking again.. so if I dont want to have usb read operation failed. (-71) what I need to do or buy? I dont really know why it flodos sytem. Three or 5 messages would be enough, but it wont stop. Only what can I do is shutdown or restart | [14:34] |
| mchehab | 2) the issue with your DVB-S adapter seems to be an issue on your antenna cabling, with is short-circuiting the DC power | [14:35] |
| Pepe | It's DVB-C/T adapter
it doesnt matter i guess? | [14:35] |
| mchehab | it does | [14:36] |
| Pepe | I use DVB-C only | [14:36] |
| mchehab | ok. it is weird then that disconnecting the antenna cable fixed it | [14:38] |
| Pepe | wait, wait. I thought that disconnecting antenna cable should do error -71? | [14:39] |
| mchehab | either there's something really weird on your cabling (perhaps it is energized somehow)... or your board got damaged
let's reset :-) with DVB-S, the external power supply sends a DC voltage to power the LNBf located at the antena... | [14:39] |
| Pepe | ok. I got this part. | [14:41] |
| mchehab | on those devices, if you connect to a DVB-C or DVB-T cabling, it short-circuits the DC...
causing the device to reboot producing lots of error -71 (and other similar error codes) | [14:41] |
| Pepe | ok.. so I need to disconnect and connect antenna cable and see if I got -71 error? I only disconnect it and then unplug cable. | [14:41] |
| mchehab | on those devices, if the antenna is disconnected, you won't get error -71 (if the USB cable is plugged)
when you unplug the USB cable, though, you may receive some random errors for a while, until the USB stack detects the disconnection. That's normal. For example: [ 101.967361] usb read operation failed. (-32) [ 101.967600] usb write operation failed. (-32) [ 101.967737] usb write operation failed. (-32) [ 101.967738] usb 2-2: dvb_usb_v2: found a 'Terratec H7' in cold state that happened when I disconnected (and later reconnected) the USB cable on my device | [14:42] |
| Pepe | for a while? It won't stop in my case. :/ | [14:43] |
| mchehab | in my case, just 3 messages
until the disconnect message comes from the USB core and the driver handles it (if the driver doesn't handle, it can happen forever) due to the RC and CAM code, with polls the device from time to time | [14:44] |
| Pepe | ok, so something is bad in driver! :-D | [14:45] |
| mchehab | it could, but error 32 is fatal; error 71 isn't
error 71 means a timeout at the messages sent through the USB bus | [14:45] |
| Pepe | and error 5 means what? | [14:46] |
| mchehab | #define EPIPE 32 /* Broken pipe */
#define EIO 5 /* I/O error */ see include/uapi/asm-generic/errno-base.h and include/uapi/asm-generic/errno.h at the Kernel tree | [14:47] |
| iive | Pepe: if you are using a cable (dvb-c) i do highly recommend getting an galvanic separator. | [14:47] |
| mchehab | or /usr/include/asm-generic/errno.h and /usr/include/asm-generic/errno-base.h | [14:48] |
| Pepe | ok. I will look at it. What about these two bug traces which I mention in mail? | [14:48] |
| iive | some TV's return current into the cables. | [14:49] |
| mchehab | Pepe, as iive mentioned: there are some devices that you could use to isolate your board from DC and low frequencies (like 50Hz AC) | [14:49] |
| Pepe | galvanic separator is cheap. I will look for one. | [14:50] |
| mrec_ | linuxtv software developers have 0 knowledge about HW :-) | [14:50] |
| mchehab | if disconnecting the antenna cable prevents your board to reboot, a galvanic separator will solve the issue | [14:50] |
| Pepe | Would be amplifier better? Cant use galvanic separator with amplifier what I found. I didnt use amplifier, but if I would want more devices I would need it. | [14:51] |
| mrec_ | no need to get a galvanic separator I can tell you | [14:52] |
| *** | awalls_ has left | [14:52] |
| mchehab | amplifiers are usually a bad idea, as they add noise
you shouldn't use, except if the signal is already too weak | [14:52] |
| Pepe | In my shop I see three galvanic separator one cost 4€, other 6€ and the most expensive 11€. I will try one and if it wont work I will return it.
what should I use then? Because I have cable from ISP which is splitter into two different outlets and one outlet has 2 TV and second outlet 2 tuners and one TV. So when I will need more I will need amplifier, right? because one tuner has issue with signals | [14:53] |
| mchehab | Pepe, you can use amplifiers, but you need to take care with them as they add noise | [14:56] |
| iive | Pepe: i got an cheap chinese amplifier, it boosts signal and lowers quality ... | [14:57] |
| mchehab | if you're already using good quality splitters and cable, then you could use a low-noise good quality amplifier | [14:58] |
| iive | Pepe: usually the signal you get is quite strong already. stronger than what you get by antena. | [14:58] |
| Pepe | ok, thanks. | [14:58] |
| mchehab | but, before doing that, better to check if your cabling is good enough. Here, I use only RG6 cables
as I had already too many bad experiences with RG5 and RG59 also, sometimes, solving a bad signal on some point is just a matter of replacing the connectors | [14:58] |
| Pepe | Thanks. I will check everything before buying amplifier. But let's go again back to my email. Why my tuner it self is restarting when I didnt do anything to it and these bugtraces which cause upstream drivers on old kernel? | [15:01] |
| iive | mrec_: feel free to help Pepe at any time. | [15:02] |
| mchehab | tuner self restarting is usually not a Kernel issue, but, instead, a hardware one
it could be a broken USB port or a broken board/cable or a bad power supply OK, kernel should not be creating duplicated sysfs nodes due to that - but solving it won't solve your issue | [15:03] |
| iive | mchehab: the device uses its own 12V power, so this should not be it. | [15:05] |
| mchehab | iive: except if the power supply is damaged | [15:05] |
| iive | mchehab: it works in one case and then not another? | [15:05] |
| Pepe | Will try another USB port on my pc. I doubt about board/cable/PSU, because it was working on old kernel and I didnt notice any restart or disappearing my tuner from tvheadend and when I have connected to router for month and I plugged to PC (and I saw that it have some issues on 4.9) I moved it back to router it's 2 days right now and there's not disappearing. :/ | [15:06] |
| mchehab | OK, so, it is not the power supply
nor board it could still be the cable - if it doesn't provide a good contact with the USB port you're connecting | [15:06] |
| Pepe | will try another and let you know | [15:07] |
| mchehab | it looks more like a problem at the USB ports on your PC
testing it is hard, as usually all USB ports are powered by the same circuit logic so, when such circuit is not working fine, the same problem may happen on multiple ports | [15:08] |
| Pepe | I will also try again kernel 3.16 on same machine if I will notice some disconnect | [15:09] |
| mchehab | OK
that is a good test | [15:09] |
| iive | imho, the box should not get power from USB if it uses its own power | [15:10] |
| mchehab | iive: true, but some hardware issue may still be affecting the USB circuit - for example, the USB chipset could be damaged by some overcharge | [15:11] |
| iive | hum. | [15:12] |
| mchehab | testing with the 3.16 kernel (with works with this device on other equipaments) will help to eliminate this | [15:12] |
| iive | i got a bit confused, could we summrise it a bit. | [15:12] |
| mchehab | I mean: some other device, connected to that USB port in the past damaged it
causing its chipset to malfunction | [15:13] |
| iive | i got too many variables for a single problem. | [15:14] |
| mchehab | well, let's try to isolate it | [15:14] |
| iive | aka, kernel version, host device and input signal cable. | [15:14] |
| mchehab | if an old Kernel version works on the same machine, with the same hardware, then there was a regressions
on such case, Pepe can use git bisect to get the Kernel change that caused the failure but, as I said before, error -71 usually means that the hardware got damaged | [15:15] |
| Pepe | oh there will so many commits between 3.16 and 4.10 | [15:16] |
| iive | Pepe: you know what binary search is, don't you? | [15:16] |
| mchehab | log2(n) | [15:17] |
| Pepe | this one, right | [15:17] |
| mchehab | usually, 16-17 interactions would be enough to get when the bug happened
but first, please check if Kernel 3.16 really works on this machine, with this hardware | [15:17] |
| iive | and with signal cable plugged in. | [15:18] |
| mchehab | yes | [15:19] |
| Pepe | will do it. I will report it in evening, when no one will use internet. | [15:19] |
| mchehab | OK | [15:19] |
| iive | what's the issue with the internet? does it run on the same cable? | [15:19] |
| Pepe | iive told you I have tunners plugged into routers and I cant unplug it:-)
Ok, guys thanks for your time and really thanks that you're trying to help. I will get things ready and will report it soon. WIll try also another usb cable | [15:20] |
| mchehab | glad to help | [15:21] |
| ........................ (idle for 1h57mn) | ||
| *** | awalls_ has left "Leaving" | [17:18] |
| ........... (idle for 50mn) | ||
| Pepe | Ok, guys so I can test it now. CT-3650 I did reboot after flooding and even when GUI went off you can see this: https://ctrlv.cz/oZgZ , https://ctrlv.cz/TYQV right now I'm waiting for restart on kernel 4.9. Will try also Combostar on USB 3.0. | [18:08] |
| .... (idle for 18mn) | ||
| 1st bugtrace same as I mention in email: http://pastebin.com/6qjGVGEB | [18:26] | |
| dongs | screenshots of monitor, really? lol | [18:27] |
| Pepe | what should I do? I was logout from it. | [18:27] |
| ....... (idle for 33mn) | ||
| CT-3650 restarted, but it didnt initializate CAM http://pastebin.com/h5Jg1aja tvheadend says slot empty but atleast now I can see tuner there.
power save was disabled on both tuners. No sign of bad drxk (yet, I plugged to USB 2.0) gonna try it on 3.16.0-4 | [19:00] | |
| ..... (idle for 21mn) | ||
| No issues with drxk even here. Driver which use CT-3650 is dvb_usb_ttusb2 and I dont really know how this is happen here I have signal strength 79% and less CE (Continuity Errors) when I'm really sure I have orange bar in signal strength and it was around 59% on latest kernel. Screenshot: https://ctrlv.cz/dS8q (I will do it also on kernel 4.9) dmesg is clear and I had issue with conax card that's why there
so many DVB cam detected: http://pastebin.com/VcpFgwXb I will leave it run for one hour atleast to detect any problems here on this kernel. | [19:23] | |
| ..... (idle for 23mn) | ||
| No restart, yet. Everything is running. | [19:47] | |
| mchehab | so, 3.16 worked; 4.9 didn't right?
you'll need to bisect it then | [19:58] |
| Pepe | yea
maybe I can try driver from 3.16 on 4.9? first? | [19:58] |
| mchehab | yes, you could do that
try 4.8 with shouldn't have issues with the DMA stack although the type of problem seems unrelated to it | [19:58] |
| Pepe | I think bisect will take so much time, so I rather try old kernel from 3.16 on 4.9. and if it won't work I will try 4.8 | [20:00] |
| mchehab | the closest to upstream, the best
testing all kernels from 3.16 to 4.9 will likely be slower than bisect this article explains git bisect: https://www.kernel.org/pub/software/scm/git/docs/git-bisect-lk2009.html the Kernel's admin guide contains a quick howto: https://static.lwn.net/kerneldoc/admin-guide/bug-bisect.html | [20:05] |
| Pepe | Thanks for arctiles. I think it will be good start for me. Let's wait for a while. Now I will try old drivers on latest kernel hopefully it will work and it will be enough for me. (Then I will gonna try also 4.8) | [20:10] |
| ......... (idle for 42mn) | ||
| I couldnt get work old drivers on 4.9. It wont load them. Never mind. I tried also to have provided kernel for 4.8 (build there was package linux-kbuild which is no longer available for that kernel). I dont want to compile kernel, so much. It will take so much time on Celeron. I think it would be better to test it on my notebook, where I will put ubuntu (which provides also kernel 4.8)
What I can do is use really old media tree to kernel 4.9. same as was for 3.16 in this case I will see if there's mistake in kernel (DMA stack or others) or in driver. | [20:52] | |
| ........... (idle for 54mn) | ||
| or would be better to install linux-image-4.9.0-0.bpo.2-amd64-dbg to debug it? | [21:47] | |
| ........ (idle for 38mn) | ||
| Tomorrow I look at my notebook. Should be faster and better for me. I compiled kernel on server many times, but not on desktop with GUI. :-) | [22:25] | |
| ↑back Search ←Prev date Next date→ Show only urls | (Click on time to select a line by its url) |