[Milhouse]

Would it be worthwhile for me to load 0505 and generate some logs for you?

I'm running a Braswell Pentium NUC.

Might as well get a 4.6.0-rc6 dmesg for completeness, I guess...

[Milhouse]

@Roby77: One-off test build with 4.4.7 kernel: Build #0505x

Let me know what effect this has on Braswell stability, and/or remote auto-repeat. Thanks.

[Milhouse]

So, we definitely have a deadlock, which is bad (possibly worse than a crash actually) but no evidence of any crash, other than your word. If you are experiencing a crash then post your crashlog, or stop claiming you are experiencing a crash.

I've opened trac #16718 for the "Search YouTube" GUI deadlock - it is being caused by PR9664.

[SmallwoodDR82]

[ 2.057769] Hardware name: Shuttle Inc. XH61V/FH61V, BIOS 1.05 05/03/2013
[ 2.057771] Workqueue: events output_poll_execute
[ 2.057773] 0000000000000000 ffffffff812b6f16 ffff880119f57b98 0000000000000000
[ 2.057774] ffffffff8103a756 0000000000000000 ffff880119f57bf0 0000000000000000
[ 2.057775] 000000000000003c ffff8800d0c54e08 0000000000000001 ffffffff8103a7b5
[ 2.057775] Call Trace:
[ 2.057778] [<ffffffff812b6f16>] ? dump_stack+0x5a/0x6f
[ 2.057780] [<ffffffff8103a756>] ? __warn+0xc8/0xe1
[ 2.057781] [<ffffffff8103a7b5>] ? warn_slowpath_fmt+0x46/0x4e
[ 2.057783] [<ffffffff81848087>] ? _raw_spin_lock_irqsave+0xe/0x29
[ 2.057784] [<ffffffff81378fed>] ? drm_wait_one_vblank+0x123/0x13c
[ 2.057786] [<ffffffff810636ce>] ? wake_up_atomic_t+0x1f/0x1f
[ 2.057788] [<ffffffff814e01e7>] ? ironlake_crtc_enable+0xac9/0xb90
[ 2.057789] [<ffffffff814c56d7>] ? hsw_write64+0x5/0x5
[ 2.057790] [<ffffffff814dddbb>] ? intel_atomic_commit+0x5c2/0x101c
[ 2.057792] [<ffffffff8138ae6c>] ? drm_atomic_check_only+0x273/0x52a
[ 2.057794] [<ffffffff8137127d>] ? restore_fbdev_mode+0x109/0x223
[ 2.057795] [<ffffffff81372a3d>] ? drm_fb_helper_restore_fbdev_mode_unlocked+0x2b/0x5e
[ 2.057796] [<ffffffff81372aa6>] ? drm_fb_helper_set_par+0x36/0x41
[ 2.057797] [<ffffffff81372a0a>] ? drm_fb_helper_hotplug_event+0xab/0xb3
[ 2.057798] [<ffffffff81368cc1>] ? output_poll_execute+0x12b/0x15a
[ 2.057800] [<ffffffff8104b188>] ? process_one_work+0x139/0x1d4
[ 2.057801] [<ffffffff8104b6d3>] ? worker_thread+0x1d3/0x29e
[ 2.057802] [<ffffffff8104b500>] ? rescuer_thread+0x2b6/0x2b6
[ 2.057803] [<ffffffff8104f1ff>] ? kthread+0x95/0x9d
[ 2.057804] [<ffffffff81848382>] ? ret_from_fork+0x22/0x40
[ 2.057805] [<ffffffff8104f16a>] ? init_completion+0x1d/0x1d
[ 2.057808] ---[ end trace 8421c749f3f31a64 ]---

Kernel bug with your hardware. Post your Xorg.0.log please.

Here is Xorg.0.log - http://pastebin.com/thaAyeCg

Please let me know if you need anything else!

Still seeing this Kernel bug in latest build #0505

[smitopher]

Would it be worthwhile for me to load 0505 and generate some logs for you?

I'm running a Braswell Pentium NUC.

Might as well get a 4.6.0-rc6 dmesg for completeness, I guess...

Here you go

##############################################
# LibreELEC #
# http://libreelec.tv #
##############################################

LibreELEC (community) Version: devel-20160505223738-#0505-g8ed15fc
LibreELEC git: 8ed15fc817243d9b799883a9535727c516a99310
media-center:~ # dmesg | pastebinit
http://sprunge.us/JAYI
media-center:~ # journalctl -a | pastebinit
http://sprunge.us/RdhL
media-center:~ # cat /storage/.kodi/temp/kodi.log | pastebinit
http://sprunge.us/ZhRR
media-center:~ # vainfo | pastebinit
libva info: VA-API version 0.39.0
libva info: va_getDriverName() returns 0
libva info: Trying to open /usr/lib/va/i965_drv_video.so
libva info: Found init function __vaDriverInit_0_39
libva info: va_openDriver() returns 0
http://sprunge.us/LBjh
media-center:~ #

[adb76]

Let me know what effect this has on Braswell stability, and/or remote auto-repeat. Thanks.

So I've directly tested the Build #0505x with my Braswell NUC Pentium system and this one seems to be stable compared to the #05050 build which gets after few minutes an

Code:

[  142.351740] [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... render ring idle

error and the UI hangs completely.

If it's useful for you I can go back to #0505 and provide the logs.

[Roby77]

@Roby77: One-off test build with 4.4.7 kernel: Build #0505x

Let me know what effect this has on Braswell stability, and/or remote auto-repeat. Thanks.

Thank you, will test tonight

[cypher242]

@Roby77: One-off test build with 4.4.7 kernel: Build #0505x

Let me know what effect this has on Braswell stability, and/or remote auto-repeat. Thanks.

Build 0505 crashes/freezes on my intel nuc NUC5PPYH - 0505x is now running for about 45 minutes without freeze or crash.

[Roby77]

Good news, i see in intel mailing list that several people are reporting gpu crash issue with same error like mine

https://bugs.freedesktop.org/show_bug.cgi?id=95196

Hope that will not need a year to fix

[m_gl]

@Roby77: One-off test build with 4.4.7 kernel: Build #0505x

Let me know what effect this has on Braswell stability, and/or remote auto-repeat. Thanks.

505 crashes and auto-repeat is also broken.

505x works fine, no more crashes, also auto-repeat is working again.

[Milhouse]

also auto-repeat is working again.

Interesting, thanks. Since the only change between #0505 and #0505x is the kernel (all other packages remained the same) this would suggest the 4.6.0 kernel is breaking auto-repeat somehow... Argh...

[schumi2004]

@Milhouse

I run your builds with great pleasure on both x86 and RPi2 but was wondering why DVD playback is disabled?
I couldn't find a answer on why this is disabled for Krypton builds?

[Roby77]

don't know why but

https://forum.libreelec.tv/thread-119-po...tml#pid843

[noggin]

@Milhouse

I run your builds with great pleasure on both x86 and RPi2 but was wondering why DVD playback is disabled?
I couldn't find a answer on why this is disabled for Krypton builds?

Some Libre/OpenElec packaging issues for the components required for DVD replay AIUI.

[sunnyfunny]

One-off test build with 4.4.7 kernel: Build #0505x

All fine on my NUC5PPYH Thanks for this build