183 replies to this topic

[pieterg]

This way, you'd end up with config from patches, in the defconfig.
I usually try to avoid that, use a defconfig with settings only from the 'clean' kerneltree.
And all patches add their stuff with the right defaults (usually 'm')

[athoik]

I notice that. First i tried on original vu kernel, but then i noticed that some entries (eg CONFIG_DVB_RTL2830,CONFIG_DVB_AF903) where missing.

So i will recreate them without patching kernel.

Edited by athoik, 3 November 2013 - 13:58.

[athoik]

Dear @OpenPLi,

Here is the patch to enable EM28xx for vusolo2 and vuduo2.

 0001-Enable-EM28xx-for-vuduo2-and-vusolo2.patch.txt   19.53KB   29 downloads

The procedure to create new config files was the following:

tar -xjf stblinux-3.3.6-2.0.tar.bz2
cd linux
cp ../vuXXXX2_defconfig .config
ARCH=mips make menuconfig
Device Drivers --->
Multimedia support --->
< > Video For Linux (NEW) was empty changed to <*> Video For Linux
Video capture adapters (NEW) --->
V4L USB devices(NEW) --->
< > Empia EM28xx USB video capture support (NEW) was empty changed to <M> Empia EM28xx USB video capture support
< > Empia EM28xx ALSA audio module was empty changed to <M> Empia EM28xx ALSA audio module
< > DVB/ATSC Support for em28xx based TV cards was empty changed to <M> DVB/ATSC Support for em28xx based TV cards
Exit and Save Configuration.

[pieterg]

shouldn't we just add the EM28xx options?
The defconfigs have been a bit of a mess it seems, now that you re-generated them, a lot of settings have changed.
And I'm not sure about the consequences, did you test the resulting kernels?

[pieterg]

Also, the kernel PR's have to be bumped by the way.

[athoik]

shouldn't we just add the EM28xx options?
The defconfigs have been a bit of a mess it seems, now that you re-generated them, a lot of settings have changed.
And I'm not sure about the consequences, did you test the resulting kernels?

The defconfings have been a mess, because probably they didn't generated with the correct kernel. But the generated with copy pasta...

Maybe @Persian Prince can enlight us how those deconfigs generated.

I build those kernels on clean system (that's why i forgot to bump kernel PR) and everything worked. I will try once again with the kernels bumped too.

This clean up of the defconfig soon or late must happen.

[pieterg]

Yes, if you are sure most things still work with this config, I will just apply your commit.
But possibly with a different commit message, indicating the defconfigs have been regenerated. (and the new EM28xx options split into a seperate commit)

[athoik]

I just compiled solo2 and those are the differences:

+./kernel-module-au8522_3.3.8-r0_vusolo2.ipk
-./kernel-module-cdfs_2.6.27-r1.4_vusolo2.ipk
+./kernel-module-em28xx_3.3.8-r0_vusolo2.ipk
+./kernel-module-em28xx-alsa_3.3.8-r0_vusolo2.ipk
+./kernel-module-em28xx-dvb_3.3.8-r0_vusolo2.ipk
+./kernel-module-gspca-main_3.3.8-r0_vusolo2.ipk
+./kernel-module-tuner_3.3.8-r0_vusolo2.ipk
+./kernel-module-videobuf-core_3.3.8-r0_vusolo2.ipk
+./kernel-module-videobuf-dvb_3.3.8-r0_vusolo2.ipk
+./kernel-module-videobuf-vmalloc_3.3.8-r0_vusolo2.ipk

(cdfs probably is build from feeds, not from make image)

Now i am build duo2, soon i will post results.

BTW. I will recreate patch also with correct message and bumped kernel.

Edited by athoik, 9 November 2013 - 16:38.

[athoik]

Same results for duo2. (only cdfs is missing)

+./kernel-module-au8522_3.3.8-r0_vuduo2.ipk
-./kernel-module-cdfs_2.6.27-r0.4_vuduo2.ipk
+./kernel-module-em28xx_3.3.8-r0_vuduo2.ipk
+./kernel-module-em28xx-alsa_3.3.8-r0_vuduo2.ipk
+./kernel-module-em28xx-dvb_3.3.8-r0_vuduo2.ipk
+./kernel-module-gspca-main_3.3.8-r0_vuduo2.ipk
+./kernel-module-tuner_3.3.8-r0_vuduo2.ipk
+./kernel-module-videobuf-core_3.3.8-r0_vuduo2.ipk
+./kernel-module-videobuf-dvb_3.3.8-r0_vuduo2.ipk
+./kernel-module-videobuf-vmalloc_3.3.8-r0_vuduo2.ipk

Here is the the with proper message (kernel also bumped to PR_INC = ".4").

 0001-Regenerate-vusolo2-and-vuduo2-defconfig.patch.txt   20.1KB   10 downloads

Edited by athoik, 9 November 2013 - 16:53.

[pieterg]

thx, applied

[n0dder]

thx, applied

 

Thanks - much appreciated

I can now install the em28xx plugin, and the 290e is detected, but on a solo2 flashed with todays image, there still seems to be something missing - at least the tuner doesn't show up in the list of installed tuners, and I get the following in dmesg:

 

em28xx: New device PCTV Systems PCTV 290e @ 480 Mbps (2013:024f, interface 0, class 0)
em28xx: DVB interface 0 found
em28xx #0: chip ID is em28174
em28xx #0: Identified as PCTV nanoStick T2 290e (card=78)
Registered IR keymap rc-pinnacle-pctv-hd
input: em28xx IR (em28xx #0) as /devices/platform/ehci-brcm.0/usb1/1-1/rc/rc0/input1
rc0: em28xx IR (em28xx #0) as /devices/platform/ehci-brcm.0/usb1/1-1/rc/rc0
em28xx #0: v4l2 driver version 0.1.3
em28xx #0: V4L2 video device registered as video0
usbcore: registered new interface driver em28xx
DVB: Unable to find symbol cxd2820r_attach()
em28xx #0: /2: frontend initialization failed
Em28xx: Initialized (Em28xx dvb Extension) extension

I'll try to dig around to see if it is something I can fix myself, so consider this as a status update :-)

 

Thanks again all,

nodder

[pieterg]

Looks like the frontend driver customization setting might be wrong (dvb drivers should depend on all possible frontend drivers)

[athoik]

On what "menu" those must be enabled?

 

There were missing before or caused by the regeneration of defconfig?

[pieterg]

Perhaps this has been wrong from the start, I haven't looked into it in detail. Dvb-usb config has never been paid much attention to, by vu+.

[n0dder]

Looks like the frontend driver customization setting might be wrong (dvb drivers should depend on all possible frontend drivers)

 

After installing kernel-module-cxd2820r and kernel-module-tda18271 the adapter seems to be registered correctly:

 

em28xx: New device PCTV Systems PCTV 290e @ 480 Mbps (2013:024f, interface 0, class 0)
em28xx: DVB interface 0 found
em28xx #0: chip ID is em28174
em28xx #0: Identified as PCTV nanoStick T2 290e (card=78)
Registered IR keymap rc-pinnacle-pctv-hd
input: em28xx IR (em28xx #0) as /devices/platform/ehci-brcm.0/usb1/1-1/rc/rc0/input1
rc0: em28xx IR (em28xx #0) as /devices/platform/ehci-brcm.0/usb1/1-1/rc/rc0
em28xx #0: v4l2 driver version 0.1.3
em28xx #0: V4L2 video device registered as video0
usbcore: registered new interface driver em28xx
tda18271 4-0060: creating new instance
TDA18271HD/C2 detected @ 4-0060
DVB: registering new adapter (em28xx #0)
DVB: registering adapter 1 frontend 0 (Sony CXD2820R)...
em28xx #0: Successfully loaded em28xx-dvb
Em28xx: Initialized (Em28xx dvb Extension) extension 

 

I was able to configure the tuner for European DVB-T frequencies, and I could also complete an Automatic scan which found all expected channels, but when I pushed the green button ("Ok") after completing the scan, I had a GSOD.  Crashlog attached (I can't deduce anything meaningful from it though).

Attached Files

•  enigma2_crash_1384117296.log   16.98KB   2 downloads

[n0dder]

In my first attempts to do Automatic scans, I disabled scanning for the 2 builtin DVB-S2 tuners, but as I kept getting GSOD on completion, I tried a full Automatic Scan of all tuners, and I was then able to press OK without GSOD.   So I can now watch DVB-T channels, but for some reason (I'm guessing it may be related to what causes the GSOD), the channel lists for the DVB-T tuner is corrupted, so I can only find the channels when browsing All services.  Thats not a big issue - I have added the channels I need to my Favorites - but I thought I'd mention it in case it can help debugging.

The tuner has gone "blind" once - ie. black screen for all DVB-T channels, and it complained about timeout/PAT not found.  Hasn't happened again after a reboot so far.

 

All in all, pretty good progress the last few days :-)

 

I've reported a bug to VU+ regarding HE-AAC passthrough not working (I checked the latest vuplus image, and that has the same problem)  - hopefully they will look at it...

 

Regards,

n0dder

Edited by n0dder, 11 November 2013 - 00:54.

[n0dder]

The tuner has gone "blind" once - ie. black screen for all DVB-T channels, and it complained about timeout/PAT not found.  Hasn't happened again after a reboot so far.

 

Hmm, happened again.  The message that shows up 20-30 seconds after zapping to a DVB-T channel is "No data on transponder! (Timeout reading PAT)"

DVB-S channels keep working, but so far only a reboot has been able to fix the DVB-T tuner.  Nothing in messages.

By the way, restarting Enigma causes the DVB-T(2) capability to disappear - the tuner can then only be configured as DVB-C or DVB-S.  Rebooting the box fixes that.

 

/n0dder

 

[Last posting today, I promise :]

Edited by n0dder, 11 November 2013 - 02:01.

[n0dder]

[...] the channel lists for the DVB-T tuner is corrupted, so I can only find the channels when browsing All services.

 

This problem was probably unrelated, and is likely already fixed by Littlesat.

Edited by n0dder, 11 November 2013 - 14:28.

[athoik]

So if you still have problems you can follow up here.

I read here and there for issues regarding DVB-T not recongized after Enigma restart and they mention usbtunerhelper that helps!

http://openpli.org/forums/topic/25328-usb-tuner-stops-working-after-enigma-restart/
http://www.world-of-satellite.com/archive/index.php/t-30143.html
http://www.world-of-satellite.com/archive/index.php/t-18596.html

Usbtunerhelper. solves that problem. At least for VU+. Install it, reboot and usb tuner does not disappear after a GUI restart.

support DVB-T2 auto scan
http://code.vuplus.com/gitweb/?p=vuplus_dvbapp;a=commit;h=c9269f3fc682d55d4fda55795b14a4224467aa43


http://code.vuplus.com/gitweb/?p=vuplus_openvuplus;a=tree;f=meta-openvuplus/recipes-vuplus/usbtunerhelper

vtuner git by vuplus
git://code.vuplus.com/git/vtuner.git

[pieterg]

What that does, is that it avoids the device to be reinitialized.
The vtuner driver does not properly handle that, it seems.
When you restart the usbtunerhelper, you will still run into the same issue.