My empeg has now started to do very strange things.
This was one of the units that had the bad batch of initial memory that Hugo replaced by hand.

Anyway, this is what happens when the empeg player is started manually from the command-line:


empeg:/empeg/bin# logout
Shell exit
Starting player
empeg-car 1.02.
3a8e70f8 DHCP Our mac is 00:02:d7:10:01:c1
3a8e70f8 DHCP Our address is 255.255.255.255
3a8e70f8 DHCP Broadcast address is 255.255.255.255
3a8e70f8 DHCP Current IP address is 255.255.255.255
3a8e70f8 DHCP Sending DHCP discover
player: memory violation at pc=0x02063b40, lr=0x00000003 (bad address=0x0fc6f0b0
, code 2)
pc : [<02063b40>] lr : [<00000003>]
sp : be3ff994 ip : 02160714 fp : be3ffd48
r10: be3ffd00 r9 : 00009800 r8 : 06fe0000
r7 : 020ea610 r6 : 00000002 r5 : 02162b20 r4 : 020e87a0
r3 : 036e12a8 r2 : 00db84a8 r1 : 020e06ae r0 : 00000000
Flags: nZCv IRQs on FIQs on Mode USER_32 Segment user
Control: D030117D Table: D030117D DAC: 00000015
Function entered at [<0204a3dc>] from [<0204e84c>]
r10 = BFFFFF64 r9 = 02114EA8 r8 = 02104FB4 r7 = 00002C0C
r6 = 02114EC0 r5 = 02114EA8 r4 = 00000000
Function entered at [<0204e830>] from [<0200a484>]
r4 = BE3FFD64
Function entered at [<0200a3d4>] from [<02068904>]
r5 = 0000000C r4 = BE3FFE40
Function entered at [<02068844>] from [<0209f69c>]
r4 = 00000000
Function entered at [<0209e690>] from [<02068688>]
r8 = 02111EBC r7 = 02111EA8 r6 = 0200A3D4 r5 = 00000003
r4 = 00000001
Function entered at [<020685d8>] from [<0209f69c>]
r5 = 00000000 r4 = 00000000
Function entered at [<0206c2d4>] from []
Function entered at [<0206c2d4>] from []
Function entered at [] from []
Function entered at [<0000ee15>] from [<00000000>]
Unable to handle kernel paging request at virtual address 0000ee19
memmap = D0300000, pgd = c0b00000
*pgd = d0305001, *pmd = d0305001, *pte = 00000000, *ppte = 00000000
Internal error: Oops: 2
CPU: 0
pc : [] lr : []
sp : c035ff48 ip : c035ff04 fp : c035ff90
r10: 00000002 r9 : c0bb60e0 r8 : 0000000c
r7 : 00000000 r6 : 0000ee15 r5 : 00000000 r4 : e1a0c00d
r3 : 60000013 r2 : c01021e4 r1 : 00000001 r0 : e1a0bffd
Flags: nZCv IRQs on FIQs on Mode SVC_32 Segment user
Control: D030117D Table: D030117D DAC: 00000015
Process player (pid: 689, stackpage=c035f000)
Stack:
c035ff20: c001a5e0 c00e3b
38
c035ff40: 60000013 ffffffff c0bb60f8 c0bb5a20 c035ffb8 c035e000 0fc6f0b0 c00129
38
c035ff60: 0fc6f0b0 00000002 c035ffb8 00000005 0fc6f0b0 00000000 00000002 000098
00
c035ff80: be3ffd00 c035ffb4 c035ff94 c0012e28 c00126f0 c0100978 02162b20 000000
02
c035ffa0: 020ea610 06fe0000 00000000 c035ffb8 c000b008 c0012d50 00000000 020e06
ae
c035ffc0: 00db84a8 036e12a8 020e87a0 02162b20 00000002 020ea610 06fe0000 000098
00
c035ffe0: be3ffd00 be3ffd48 02160714 be3ff994 00000003 02063b40 60000010 ffffff
ff
Backtrace:
Function entered at [] from []
r10 = BE3FFD00 r9 = 00009800 r8 = 00000002 r7 = 00000000
r6 = 0FC6F0B0 r5 = 00000005 r4 = C035FFB8
Function entered at [] from []
r8 = 06FE0000 r7 = 020EA610 r6 = 00000002 r5 = 02162B20
r4 = C0100978
Code: ebfcda25 e2440010 (e5961004) e1a035a1 e59f20cc
Restored terminal settings
Remounting first music partition read-only
Remounting second music partition read-only
Abnormal player termination
Player received signal 11
Starting bash.
empeg:/empeg/bin#


And on a power-on process, the unit simply recycles that process above and the front drisplay goes from the Empeg Startup screen, Loading Music Databases... bunch of text, then the player display pops up, then it jumps batch to the startup screen again, and does this forever.

Here is an included hw test capture:


empeg-car bootstrap v1.00 20000601 (hugo@empeg.com)
If there is anyone present who wants to upgrade the flash, let them speak now,
or forever hold their peace...
0000 empeg-car board test version 0.04
0100 ram test starting
0110 testing ic 1 (0-3mb, low word)
0120 testing ic 2 (0-3mb, high word)
0130 testing ic 3 (4-7mb, low word)
0140 testing ic 4 (4-7mb, high word)
0150 testing ic 19 (8-11mb, low word)
0160 testing ic 32 (8-11mb, high word)
01f0 ram test done
0200 ide detect
0210 command issued
0220 drive id IBM-DARA-218000
02f0 ide ok
0300 dsp/i2c test
03f0 dsp/i2c ok
0400 temperature detect
0410 temperature is 33
04f0 temperature ok
0500 usb detect
0510 usb rev 1012
05f0 usb ok
0600 ethernet detect
0610 ethernet rev 334b
06f0 ethernet ok
0700 cs4231 detect
0710 cs4231 rev a0
07f0 cs4231 ok
00f0 tests complete


Also this is being sent in as an email.

Any ideas?
Thanks

Smoker_Man
#080000449 MkII - 36Gb Blue

At a first guess, it's upset by something in your current playlist. Try this:

dd if=/dev/zero of=/dev/empeg_state bs=128 count=1

...then ctrl-Ding to re-run the player. You will have lost your volume settings, current visual, etc.

If that still behaves the same, try this:

dd if=/dev/zero of=/dev/hda3 bs=1k count=16

...this is a bit more severe, but I *think* (from memory) it won't affect EQ, just the current playlist.

Hugo

At a first guess, it's upset by something in your current playlist.
Exactely what happened to me. Same symptoms. Was able to fix it thanks to Hugo's wisedom expressed in the previous post. However my question is what possibly could have been wrong with the current playlist.
My first though was a corrupted mp3 file but the file/playlist in question appeared to play just fine afterwards.

753, did your problem happen when running version 1.02 of the software, or had you already upgraded to 1.03?

___________
Tony Fabris

When the problem did occur, i was still running 1.02. I upgraded to 1.03 (dev img), in the hope that this would solve the problem. It didn't.
However I can't say wheter or not the problem would have existed if I had ran 1.03 in the first place.

Interesting. I wonder if there's a way to find out how the database is getting corrupted? Or at least add some bounds-checking code to the player software at the failure point? So at least it would crash gracefully and tell you that the playlist got lost, instead of going haywire.

___________
Tony Fabris

It just happened again. I was listen to an Audiobook (Lord of the Rings, BBC Radio Show) as the in the middle of episode 7 Gandalf was suddenly interrupted and the player started a Startup-Loop i had already seen once.

However altough the symptoms are practically the same as in the above case(s), the cause seem to be a different. When this first happened to my player, it seemed to be 'upset by something in the current playlist'. Just as Hugo diagnosed when Smoker_man had the very same problem. After having reset the empeg state the player booted and when the very same playlist was selected, it played just fine. The problem was not reproducable.

This time it is. The player seem to be upset about something within the mp3 file itsself.
(21:04) Gandalf speaks.
"If we fail, we fall. If we suceed, then we will face the next ta... " (21:12)
Every single time it gets to that point the player terminates abnormally and the startup loop is initiated.
At the end of each startup Gandalf get the change to say "ta..", before the player reboots.
As long as Gandalfs "ta..." lasts the player will respond to commands. If i pause it or skip the track, then this breaks the loop cycle. I can even fast-forward over 21:12 to prevent the player from entering the loop.

If you've got that MP3 file on your PC somewhere, check it for corruption, for instance, with MP3Trim.

___________
Tony Fabris

...and send it to us for analysis if you can.

Hugo

I attempted to do just so a while back, but unfortunately the file exeeds MP3 Trim's (Freeware Version) maximum file size.
However i can tell that there must be something nasty about that file. (I admit, i didn't rip it myself, its a file i got of a p2p filesharing network. )
If i open the file with Cool Edit and save it unchanged, it will play fine on the empeg, indicating that there is something wrong with the original file.
Winamp however plays the original file fine with only a scrachty noise noticeable at 21:12

Uhm.. its a 25 meg file. You sure dont want me to trash your support mailbox with it.
Wait a second ...
Okay, i split the file using TrackMager down to 70k roughly and verified it still induces the startup loop.
File is attached to this post and sent to carbugs@riohome.com.

MP3Trim seems to think that file is fine.

The last frame is incomplete, but I don't think that's where the bug lies, I think that's just an error in the software you used to trim the file.

I'm sure Hugo will take a look at it and find out why XAudio burps on the file. More importantly, he'll check to make sure the ARM decoder (the one that's being used in the upcoming 1.1) doesn't burp on that file.

___________
Tony Fabris

MP3Trim seems to think that file is fine

Well, that file is definitely sick. Out of curiosity I tried to play it from Netscape twice. Both times 'crescendo' choked on the last byte, taking Netscape down with it.

It's not just the empeg that doesn't like this episode of the hitchhikers . . .

Henno
mk2 6 nr 6