[davilla]

since xbmc is up, you should be able to ssh in.

/root/.xbmc/temp/xbmc.log

[noggin]

Thanks Davilla - your patience is amazing.

I have been using a copy of "output/target" in my buildroot-AmLogic directory as my NFS share rather than the rootfs.tar - which is probably my mistake.

[noggin]

Untar-ing the rootfs.tar and using that instead doesn't seem to make a difference, other than the usr.sqsh being there. XBMC is still crashing out. Will paste logs via PasteBin in case they are of interest.

In other news, connecting an LNB to TV Headend was not a total success. It's talking to the tuner, but I don't think it's connecting properly, as it never finds any services. I have a feeling that the AV6211 is connected to the AMLogic SoC by one of the three Transport Stream interfaces, so I suspect I need to enable one or all of TS0-TS2 in my kernel as well?

That's the next job.

[noggin]

This is my xbmc.log file : http://pastebin.com/30GTVMvN
This is my xbmc.old.log file : http://pastebin.com/qQUpTSeL (This seems to have more in it.)

I ended up getting these by typing cat xbmc.log from the UART. (Whilst I could SSH into my box, and cat the directory there, I couldn't use Filezilla to access the box, and can't log to a file using Putty over SSH?)

Wonder if anything sticks out - doesn't seem to log the crash?

[davilla]

nothing jumps out

run xbmc under gdb, then you can backtrace the crash

gdb /opt/xbmc-aml/usr/lib/xbmc/xbmc.bin
gdb> run --standalone

[noggin]

Not near my box at the moment, but have the boot logs with me.

When I try to run xbmc from the command line via UART - by simply typing xbmc at the busybox # prompt I get :

Code:

sh: glxinfo: not found
XBMC needs hardware accelerated OpenGL rendering.
Install an appropriate graphics driver.

Please consult XBMC Wiki for supported hardware
http://wiki.xbmc.org/?title=Supported_hardware

Is this also likely to be the case under gdb?

Another thought I had was whether the two video output bits of the box (HDMI and CVBS) may be relevant? (I only have a cable connected to the HDMI output - and don't have a CVBS 3.5mm jack cable at the moment)

Also - is this relevant?

http://forum.xbmc.org/showthread.php?tid=35830

Seems to be a possible cause of :

Code:

xbmc.bin: /home/stephen/SteveDev/buildroot-AmLogic/output/build/xbmc-0987bea35050a3f5d352c3205d75b7173a7d126d/xbmc/utils/MathUtils.h:166: int MathUtils::truncate_int(double): Assertion `x > static_cast<double>((-2147483647 - 1) / 2) - 1.0' failed.

Or am I barking up totally the wrong tree?

[davilla]

you have to setup the paths and env vars

that Assertion is because you have something wrong

[noggin]

you have to setup the paths and env vars

that Assertion is because you have something wrong

Paths to what? Env vars for what? Is it a problem in the build on my Intel box making the buildroot - or is it because there is something odd about how my AMLogic box is configured.

I see that xbmc.bin and tvheadend are all inside the usr.sqsh SquashFs bit - which appears to be mounted OK, as I can run tvheadend and xbmc (albeit xbmc fails?).

However I'm not sure my /tmp/userdata is happening as I think I'm getting yaffs2 errors during s10setup in init.d? Can I mount this somewhere else? (SD card?)

Do I need to fiddle with advancedsettings.xml to change my paths - or is it something more fundamental and OS-related?

Am I right that my NFS rootfs should be R/W - if so is there any reason to have the usr.sqsh in the loop - can't I just have this as part of my target rootfs?

[noggin]

nothing jumps out

run xbmc under gdb, then you can backtrace the crash

gdb /opt/xbmc-aml/usr/lib/xbmc/xbmc.bin
gdb> run --standalone

I'm definitely doing something wrong somewhere...
Here's what BusyBox reports as my environment after boot-up (with s95xbmc renamed to stop it running)

Code:

/ # env
USER=root
HOME=/
a9_clk=600M
clk81=187500000
logo=osd1
TERM=vt102
PATH=/sbin:/usr/sbin:/bin:/usr/bin
SHELL=/bin/sh
PWD=/
resolution_size=1280,720

xbmc.bin appears to be in /usr/lib/xbmc - which is my mounted usr.sqsh?
(In your path above it seems to be referred to by it's position in the NFS share on the NFS server not the local mount of the usr.sqsh - or am I missing something here? I'm sure I am!

For development should I have my /usr folder within my NFS share and somehow stop the mounting of usr.sqsh?

Code:

/ # cd /usr
/usr # ls
bin      lib      libexec  sbin     share
/usr # cd lib
/usr/lib # cd xbmc
/usr/lib/xbmc # ls
addons    system    xbmc.bin

Code:

/usr/lib/xbmc # gdb /usr/lib/xbmc/xbmc.bin
GNU gdb (GDB) 7.1
Copyright (C) 2010 Free Software Foundation, Inc.
License GPLv3+: GNU GPL version 3 or later <http://gnu.org/licenses/gpl.html>
This is free software: you are free to change and redistribute it.
There is NO WARRANTY, to the extent permitted by law.  Type "show copying"
and "show warranty" for details.
This GDB was configured as "arm-unknown-linux-gnueabi".
For bug reporting instructions, please see:
<http://www.gnu.org/software/gdb/bugs/>...
Reading symbols from /usr/lib/xbmc/xbmc.bin...(no debugging symbols found)...done.
(gdb) run --standalone
Starting program: /usr/lib/xbmc/xbmc.bin --standalone
[Thread debugging using libthread_db enabled]
Could not init logging classes. Permission errors on ~/.xbmc (/.xbmc/temp/)
ERROR: Unable to create application. Exiting

Program exited with code 0377.

So there is something wrong. (Well that's obvious!)

~/.xbmc is in the "root" folder on my NFS share isn't it? Is there a read-only or permission problem do you think? Would I change that on my Ubuntu machine that is running the NFS Server?

By the way - I don't see anything after the A9 boot-up logo (and don't always see that)

Advice etc.? Is this because I don't appear to have any NAND Flash stuff mounted?

Any of this relevant ?

Code:

[    2.217110] nand env: nand_env_probe.
[    2.220612] set nand parts for chip 4096MB
[    2.244401] Creating 8 MTD partitions on "B revision 26nm NAND 4GiB H27UBG8T2B":
[    2.246243] 0x000001400000-0x000001800000 : "ubootenv"
[    2.253028] 0x000001800000-0x000002800000 : "aml_logo"
[    2.261890] 0x000002800000-0x000003800000 : "recovery"
[    2.267308] 0x000003800000-0x000004c00000 : "boot"
[    2.274158] 0x000004c00000-0x000024c00000 : "system"
[    2.436218] 0x000024c00000-0x000030c00000 : "cache"
[    2.494767] 0x000030c00000-0x000050c00000 : "userdata"
[    2.651134] 0x000050c00000-0x000100000000 : "NFTL_Part"
[    4.759168] nftl initilize completely dev size: 0xa1e00000 5
[    4.759853]  avnftl8: unknown partition table
[    4.763687] nftl release flush cache data: 0

Code:

[    9.856337] yaffs: dev is 32505859 name is "mtdblock3" rw
[    9.856371] yaffs: passed flags ""
mount: mounting /dev/mtdblock3 on /recovery failed: Invalid argument
chown: unknown user/group default:default
S10setup: init fb0, fb1
2880+0 records in
2880+0 records out
3686400 bytes (3.5MB) copied, 0.013209 seconds, 266.2MB/s
2880+0 records in
2880+0 records out
3686400 bytes (3.5MB) copied, 0.013288 seconds, 264.6MB/s
S10setup: remount / rw
S10setup: checking keys
generating rsa key... generating dsa key... S10setup: moving /etc
S10setup: moving /media
S10setup: mounting /tmp/userdata
[   12.348418] yaffs: dev is 32505863 name is "mtdblock7" rw
[   12.348465] yaffs: passed flags ""
[   12.351754] yaffs2 check point failed.
[   13.346526] aml nand read data ecc failed at page:116728  blk 455 chip 0, readretry_failed_cnt:1
[   13.349754] HYNIX NAND set partmeters here and hynix_read_retry_cnt:0
[   13.356191] Enter aml_nand_set_reg_value_hynix
[   13.360589] aml_nand_set_reg_value_hynix, REG(0xa7):     value:0x21 for chip[0]
[   13.367620] aml_nand_set_reg_value_hynix, REG(0xad):     value:0x27 for chip[0]
[   13.374627] aml_nand_set_reg_value_hynix, REG(0xae):     value:0x7c for chip[0]
[   13.381632] aml_nand_set_reg_value_hynix, REG(0xaf):     value:0xc5 for chip[0]
[   14.022418] aml nand read data ecc failed at page:114074  blk 445 chip 0, readretry_failed_cnt:1
[   14.025668] HYNIX NAND set partmeters here and hynix_read_retry_cnt:1
[   14.032061] Enter aml_nand_set_reg_value_hynix
[   14.036499] aml_nand_set_reg_value_hynix, REG(0xa7):     value:0x0 for chip[0]
[   14.043416] aml_nand_set_reg_value_hynix, REG(0xad):     value:0x1e for chip[0]
[   14.050422] aml_nand_set_reg_value_hynix, REG(0xae):     value:0x6b for chip[0]
[   14.057447] aml_nand_set_reg_value_hynix, REG(0xaf):     value:0xb7 for chip[0]
S10setup: remount / ro
S10setup: switching to ondemand
S10setup: done

Does the above suggest that something is wrong in NAND-land?

Code:

ifdown: interface eth0 not configured
ifdown: interface wlan0 not configured
udhcpc (v1.17.4) started
Failed to kill daemon: No such file or directory
Sending discover...
[   21.603098] nftl creat stucture completely free blk: 5 erased blk: 1388
Sending select for 192.168.1.131...
Lease of 192.168.1.131 obtained, lease time 86400
Failed to kill daemon: No such file or directory
[   22.675545] set hash low=2,high=800001
[   22.677596] set hash low=2,high=810001
deleting routers
route: SIOCDELRT: No such process
adding dns 192.168.1.254

And here it looks as if XBMC is reconfiguring my eth0 from static IP to DHCP by dropping eth0 - wouldn't this do nasty things to an NFS mounted file system - or am I barking up another wrong tree?

[davilla]

whao, jumping around to much, I can't follow

#1, disable auto launch of xbmc
#2, enable dhcp by changing /etc/network/interface from manual to dhcp
#3, export HOME=/root
#4 disable network management in xbmc, ie. in xbmc/advancedsettings.xml, remove <enablenetworkmanager>1</enablenetworkmanager>

now get boot, setup of rootfs right. once that's right, then you can launch xbmc manually.

[noggin]

whao, jumping around to much, I can't follow

Oops - sorry...

#1, disable auto launch of xbmc

Think I've done this by renaming /etc/init.d/s95xbmc as old-s95xbmc (Looking at rcS in init.d it appears to run all the S?? scripts in numerical order - so renaming any script so that it doesn't start with S stops it being executed?)

#2, enable dhcp by changing /etc/network/interface from manual to dhcp

Do I need to enable dhcp - if I pass a static ip from uBoot via my bootargs, can I keep with this - makes SSHing and TVHeadend access a lot easier if I know the IP address of the box without having to type ifconfig?

#3, export HOME=/root

Aah. I think I see that this is important. At the moment HOME=/ according to my busybox environment. Can I send HOME=/root as a bootargs bit from uBoot (asking as I'm away from my dev set-up today)

#4 disable network management in xbmc, ie. in xbmc/advancedsettings.xml, remove <enablenetworkmanager>1</enablenetworkmanager>

Is this file in my usr.sqsh? How is it best to deal with this. I'm not sure I know where usr.sqsh is being mounted (or is it bound?) to /usr. Would it be better, for dev purposes, to run with a /usr directory populated on my NFS share and not mount the usr.sqsh? I'm not sure how to edit the contents of usr.sqsh (Mount it, copy the contents to a directory, edit them, and resquash?)

Alternatively where is the default advanced settings file being made so I can stop it being enabled? Looking at https://github.com/Pivosgroup/xbmc/blob/...ttings.cpp it appears that m_enableNetworkManager = false; in the standard settings? So is the advanced settings xml file created later?

now get boot, setup of rootfs right. once that's right, then you can launch xbmc manually.

Hope so!

[davilla]

2) or static, either is fine, just stop using xbmc's net manager until you get everything sorted out.

4) see buildroot-linux -> package/thirdparty/xbmc/advancedsettings.xml

[noggin]

4) see buildroot-linux -> package/thirdparty/xbmc/advancedsettings.xml

Aagh - obvious. Thanks.

(I'm struggling to do any meaningful search within github.com - is there a knack?)

Meanwhile looking at the AVL6211 driver code - it appears to be quite 'odd' (particularly in reported capabilities - some of them appear to be DVB-T options, and it doesn't flag max/min symbolrates - which explains why TV Headend is getting confused...) Time to add the v4lutils package as well and see what's what. I'm enjoying this - in a strange sort of way...

[davilla]

I tend not to try and use github.com for searching, this works much faster

find . -name advancedsettings.xml

[noggin]

whao, jumping around to much, I can't follow

#1, disable auto launch of xbmc

done

#2, enable dhcp by changing /etc/network/interface from manual to dhcp

Have left on static IP

#3, export HOME=/root

Did this by typing export HOME=/root at busybox # prompt

Code:

/ # export HOME=/root
/ # env
USER=root
HOME=/root
a9_clk=600M
clk81=187500000
logo=osd1
TERM=vt102
PATH=/sbin:/usr/sbin:/bin:/usr/bin
SHELL=/bin/sh
PWD=/
resolution_size=1280,720

#4 disable network management in xbmc, ie. in xbmc/advancedsettings.xml, remove <enablenetworkmanager>1</enablenetworkmanager>

Done

now get boot, setup of rootfs right. once that's right, then you can launch xbmc manually.

Hmm - whether I try to run xbmc as the login user or after typing su I get the same message

Code:

/ # xbmc
sh: glxinfo: not found
XBMC needs hardware accelerated OpenGL rendering.
Install an appropriate graphics driver.

Please consult XBMC Wiki for supported hardware
http://wiki.xbmc.org/?title=Supported_hardware
/ # su
~ # xbmc
sh: glxinfo: not found
XBMC needs hardware accelerated OpenGL rendering.
Install an appropriate graphics driver.

Please consult XBMC Wiki for supported hardware
http://wiki.xbmc.org/?title=Supported_hardware
~ # glxinfo
sh: glxinfo: not found
~ #

I can't find a glxinfo anywhere. Should it be there or is the fact that xbmc is trying to run it a sign of something else odd?

Is there anything I can do to confirm my NFS rootfs is correctly configured and set-up?

Is there anything I can do to confirm any other stuff (is there some NAND stuff going on?) is correctly working (or relocate it to my NFS server if it isn't?)

I notice I get

Code:

[   20.858039] nftl creat stucture completely free blk: 7 erased blk: 1385

and this error

Code:

[   44.899369] aml nand write failed at 102916