• 1
  • 34
  • 35
  • 36(current)
  • 37
  • 38
  • 168
OpenELEC Testbuilds for RaspberryPi (Kodi 16.0)
We are talking about the first ploint, it get dry.

And I will install 802 and tested overnight.
Okay, does only the EPG window gets dry or all places where EPG info gets displayed?

Just trying to find out whether your problems could be related to some recent changes to the EPG window code.
@ksooo maybe you Can test this on your set up too ?

Find similar wont work from channel guide epg ( but works from timeline epg) search opens but closes direct
Using Tvh with last nights build on rpi fom Milhouse

Fails in all the latest build i try
Sad 
(2015-08-04, 02:04)zaphod24 Wrote: Edit: Deleted Epg10.db and rebooted once or twice and thought it was working...

On #0806 I deleted Epg10.db rebooted ... then epg correctly populated, however trying on a subsequent 3rd reboot (leaving refreshed epg) had a blank epg guide that is multichannel epg.SadSad
Location UK; Media server Windows 7 with ArgusTV 2.3 with TBS6981 DVB-S2 x4 and DVB-T x 2:All network connections cabled on 1Gb router Raspberry Pi2 1GB x 2; RPI 3 x1; PiB+512MB x 3; TV Samsung 55" C8000; AV Denon AVR X2200W
(2015-08-08, 10:57)Milhouse Wrote: http://forum.kodi.tv/showthread.php?tid=...pid2005396

Thanks, sdhost looks good. Do I understand correctly from observation 4 that force_turbo is no longer needed to use sdhost at higher core_freq?
(2015-08-08, 19:20)doveman2 Wrote: Thanks, sdhost looks good. Do I understand correctly from observation 4 that force_turbo is no longer needed to use sdhost at higher core_freq?

Yes, force_turbo=1 no longer needed for maximum sdhost performance.
Texture Cache Maintenance Utility: Preload your texture cache for optimal UI performance. Remotely manage media libraries. Purge unused artwork to free up space. Find missing media. Configurable QA check to highlight metadata issues. Aid in diagnosis of library and cache related problems.
(2015-08-08, 19:22)Milhouse Wrote: Yes, force_turbo=1 no longer needed for maximum sdhost performance.

Great, thanks.
New OpenELEC Jarvis build #0808: RPi / RPi2
(Supercedes previous build)

Code:
# uname -a
Linux rpi512 4.1.4 #1 Sat Aug 8 23:59:38 BST 2015 armv6l GNU/Linux

# vcgencmd version
Aug  3 2015 14:56:05
Copyright (c) 2012 Broadcom
version 4b51d81eb0068a875b336f4cc2c468cbdd06d0c5 (clean) (release)

# lsb_release
OpenELEC (Milhouse) - Version: devel-20150808235847-#0808-gc12e7a9 [Build #0808]

# vcdbg log msg 2>&1 | grep DTOK
001579.289: Kernel trailer DTOK property says yes

# Kernel device tree status: Enabled

Based on tip of OpenELEC master (c12e7a96, changelog) and tip of XBMC master (d42e33ce, changelog) with the following modifications: Build Highlights:
  1. EPG: Possible fix (PR7734) for update problems - please test/confirm
  2. EPG: Sync EPG update on startup (PR7367)
  3. libcec: Holding buttons fix for Philips TVs (PR:125) - please post if odd CEC button behaviour is observed
  4. systemd: Update to systemd-224
Build Details:
  1. OpenELEC:
    • brcmfmac_sdio-firmware: fix service type (PR:4267, 1 commit, 2 files changed)
    • Update libinput and xf86-input-libinput (PR:4265, 2 commits, 2 files changed)
    • linux/patches/xpad-fix: update patch for kernel v4.1.4 (PR:4264, 1 commit, 1 file changed)
    • libjpeg-turbo: update to libjpeg-turbo-1.4.1 (2b511cca)
  2. XBMC:
    • changed: For memory cache don't create a bigger cache than filesize (PR:7547, 1 commit, 5 files changed)
    • [gitignore] Add libdvdcss artifact (PR:7741, 1 commit, 1 file changed)
    • [addons] extend context menu system (PR:7544, 7 commits, 21 files changed)
    • [python] deprecate thumb/icon methods and args in favour of setArt (PR:7690, 4 commits, 2 files changed)
    • [pydocs] sortMethod list is already at the bottom of the page (PR:7758, 1 commit, 1 file changed)
    • ADD: [keyboard] allow to map long key presses (PR:7346, 2 commits, 7 files changed)
    • [epg][fix] trigger epgs create after we started the epg container (PR:7734, 1 commit, 1 file changed)
  3. Additional commits/pull requests/changes not yet merged upstream:
    • Added: PR:125: Fix for holding buttons on remotes of Philips TVs (libcec)
    • Added: PR:4266: [systemd] update to systemd-224
    • Added: PR:7367: [epg] Sync epg update from client / db on startup
Texture Cache Maintenance Utility: Preload your texture cache for optimal UI performance. Remotely manage media libraries. Purge unused artwork to free up space. Find missing media. Configurable QA check to highlight metadata issues. Aid in diagnosis of library and cache related problems.
I decided to expand the FAT32 partition on my brother's SD card, just in case I need to install a debug build for him in the future. I used MiniTool Partition Wizard and had to delete the ext4 partition before I could resize the FAT32 partition (it wouldn't let me decrease the size of the ext4 partition to make room to expand the FAT32 one), which is now 415MB (it defaulted to that after I deleted the ext4 partition, so I just left it at that). I figured it would be good to start afresh anyway, as I saw a couple of things that didn't look quite right.

I always get confused about whether OE should be able to boot with a blank ext4 partition and create any files it needs on first boot, or if it needs files installed there before it can boot but anyway, now it fails to boot and says "Could not mount /dev/mmcblk0p2", so is there something I need to put on there? The only other thing I can think of is that maybe the second partition isn't recognised as mmcblk0p2 after the changes I made but I've double-checked with MinTool and it's only showing two partitions on there.

EDIT: Checking the ext4 partition, it just has an empty "lost+found" folder on there but maybe that was created when I formatted the partition and doesn't indicate that the RPi was able to access it.
The second partition (/dev/mmcblk0p2) must be formatted with ext4 (or ext2), and when OpenELEC boots it will populate the partition with any required files and directories so yes, a completely empty second partition is fine, and you don't need to copy any files to it before the first boot.

As the partition can't be mounted, are you sure your second partition has been formatted with an ext4/ext2 filesystem? I'm not familiar with MiniTool, but just creating the partition isn't enough - once the partition is created you'll also need to create a valid filesystem in the new partition.
Texture Cache Maintenance Utility: Preload your texture cache for optimal UI performance. Remotely manage media libraries. Purge unused artwork to free up space. Find missing media. Configurable QA check to highlight metadata issues. Aid in diagnosis of library and cache related problems.
(2015-08-09, 02:50)doveman2 Wrote: EDIT: Checking the ext4 partition, it just has an empty "lost+found" folder on there but maybe that was created when I formatted the partition and doesn't indicate that the RPi was able to access it.

Yes, that's always created when the filesystem is added. Did you label the partition/filesystem with the name "Storage", just in case that's an issue.

Assuming the filesystem is valid I'm not sure why it wouldn't boot, is your cmdline.txt referencing the partition as "disk=/dev/mmcblk0p2" ?

When you drop into the debug shell after a failed boot you could try mounting /dev/mmcblk0p2 manually and you might see what the problem is.
Texture Cache Maintenance Utility: Preload your texture cache for optimal UI performance. Remotely manage media libraries. Purge unused artwork to free up space. Find missing media. Configurable QA check to highlight metadata issues. Aid in diagnosis of library and cache related problems.
(2015-08-09, 01:48)Milhouse Wrote: [*]libcec: Holding buttons fix for Philips TVs (PR:125) - please post if odd CEC button behaviour is observed

Great, now it works, thanks
(2015-08-09, 03:07)Milhouse Wrote: Yes, that's always created when the filesystem is added. Did you label the partition/filesystem with the name "Storage", just in case that's an issue.

Assuming the filesystem is valid I'm not sure why it wouldn't boot, is your cmdline.txt referencing the partition as "disk=/dev/mmcblk0p2" ?

When you drop into the debug shell after a failed boot you could try mounting /dev/mmcblk0p2 manually and you might see what the problem is.

Thanks, I looked in /dev and saw that there was p1, p2, and p5 and then realised that I formatted the second partition as Logical, not Primary, which causes it to be p5 as p2 is occupied by an invisible (or at least, not obvious) partition. I changed cmdline.txt to refer to p5 and then it booted fine, so I changed it back to p2 and changed the partition from Logical to Primary with MiniTool (almost instant) but although it would get past the bootscreen, it then just showed a black screen, so I deleted the partition and recreated it as Primary and now it's fine. I guess something that was created when I booted it as p5 referenced p5 and thus couldn't find what it was looking for after I changed it to p2.
I'm a bit confused with TVheadend (yet again!).

I've enabled the service and PVR client and enabled TV, gone to the config and it shows both tuners on my dual-tuner. I've got Initial scan and Idle scan ticked for both and tuner #0 is linked to #1 and vice-versa.

I've added a DVB-T network for Freeview - Crystal Palace and Muxes shows 7 but they seem to stuck on PEND, even after a reboot.

What's even more confusing is that neither Confluence or AppTV show a menu entry for LiveTV and I can't see anything in the settings for either skin to enable one (although I'm sure there used to be such a setting). Is the entry meant to automatically appear when Kodi detects a properly configured tuner? I'm not getting any "Connection lost" popups, so it seems to be setup correctly.

EDIT: With debug logging, I see a lot of this in the log:

06:14:03 460.793457 T:1556083776 DEBUG: AddOnLog: Tvheadend HTSP Client: pvr.hts - sending message [getDiskSpace : 152]
06:14:03 460.794037 T:1738531904 DEBUG: AddOnLog: Tvheadend HTSP Client: pvr.hts - received response [152]

EDIT2: It seems the default keymapping has my power button shutdown Kodi, which isn't ideal as it just looks like it's frozen! Maybe have it display the power menu would be better?

06:30:20 1438.410278 T:1966985216 DEBUG: LIRC: Update - NEW at 1430582:e2 0 KEY_MEDIA_UP devinput (KEY_MEDIA_UP)
06:30:23 1441.148804 T:1966985216 DEBUG: LIRC: Update - NEW at 1433320:74 0 KEY_POWER devinput (KEY_POWER)
06:30:23 1441.149048 T:1966985216 DEBUG: OnKey: rewind (0xc4) pressed, action is ShutDown()
(2015-08-08, 15:47)Heiko123 Wrote: We are talking about the first ploint, it get dry.

And I will install 802 and tested overnight.

So I tested with it, but the same results.
The epg ended today 21:00 (last scan yesterday)
And the records task wont be updated,
Normally if I goto recordings I see the "working..." event for short time, but now never.

I'll testing the actually version.
  • 1
  • 34
  • 35
  • 36(current)
  • 37
  • 38
  • 168

Logout Mark Read Team Forum Stats Members Help
OpenELEC Testbuilds for RaspberryPi (Kodi 16.0)10