Hello all.. one of my empegs has been locking up with the error "Sigkill Error" in a box over top of visuals... the music stops and then the player reboots quickly... that usually clears it up and it picks up where it left off. but sometimes it will stay in this loop unless I press forward to next song repeatedly while the unit boots and that will clear things up for awhile. It appears to be doing this a bit more these days.. and was wondering if anyone knew what could be the issue? drive? bad data? I am running v3.00-alpha11...

Thanks!!

Bit-rot on one of the tracks, or elsewhere on the drive.. or just low on memory (v3a11 wants a lot, and doesn't play as nicely as it could when other apps are also loaded), or ... ??

But first, upgrade your Hijack.

-ml

Thank you sir...
I will try the hijack upgrade first...
I would guess it might be related to memory... because the issue does not seem related to any specific tracks.. the same track may play many times just fine but crash another... and when it happens it is almost never on the same track... It does seem to happen more in the first few tunes it plays after a cold startup.
I will try and update this thread in a couple of days...

Thanks again.

Well.. I had Hijack v461 on it, so I upgraded to v488 and it reproduced the error within 5 mins but this time the error has changed a bit.
It now reports: "0000.42 Sigkill err"

Should I clean it out and start over?

Thanks again for your time...

Bad IDE cable?

Bad IDE header? (Look 6 threads down )

Ok so after playing songs to and from work today, it reported 0000.51 Sigkill err and rebooted shortly ater startup... I noticed after it does this and reboots the visual goes away and the display changes to track after track info mode... once in this mode, the error seems to occur a lot less frequently if at all. So I pulled it out of the dash and did a cold reboot and the visual came back as it is the default but again produced the error fairly quickly. So I cold booted again and changed the visual to the simple o-scope from the previous 3d cube (not sure if that was the name) and it has not happened so far. I do wonder if some visuals use more memory and it could be possibly related to memory explaining why it seems to fail more frequently with complex visuals as opposed to simpler or none at all.
Just thinking out loud.

Does adding this to the config.ini file help at all?

[Startup]
ReserveCache=4

Hey Mark..
I tried your config settings and got 0000.73 & 0000.83 sigkill err.. (if I flip through visuals it happens the moment I hit Borgbeat)
But only with the Borgbeat visual... [3d cube] so far I can't seem to get the other visuals to produce this...

Do the error codes translate to anything meaningful?

I am ok with not running borgbeat if that is the culprit... but I think I need to keep watching it closely... a new behavior I have seen is after the error the visual now comes back... this is different and harder to catch as if I don't see it fail and reload, I will miss the error altogether.

Here is the bootlog from the terminal session in case it indicates something...

SA1100 serial driver version 4.27 with no serial options enabled
ttyS00 at 0xf8010000 (irq = 15) is a SA1100 UART
ttyS01 at 0xf8050000 (irq = 17) is a SA1100 UART
ttyS02 at 0xf8030000 (irq = 16) is a SA1100 UART
Signature is 636f6972 'rioc'
Tuner: loopback=0, ID=-1
show_message("Hijack v488 by Mark Lord")
empeg display initialised.
empeg dsp audio initialised
empeg dsp mixer initialised
empeg dsp initialised
empeg audio-in initialised, CS4231A revision a0
empeg remote control/panel button initialised.
empeg usb initialised, PDIUSBD12 id 1012
empeg state support initialised 0089/88c1 (save to d0004380).
empeg RDS driver initialised
empeg power-pic driver initialised (first boot)
RAM disk driver initialized: 16 RAM disks o
empeg single channel IDE
Probing primary interface...
ide_data_test: wrote 0x0000 read 0x0080
ide_data_test: wrote 0xffff read 0x0080
ide_data_test: wrote 0xaaaa read 0x0080
ide_data_test: wrote 0x5555 read 0x0080
ide_data_test: wrote 0x0000 read 0x0080
ide_data_test: wrote 0xffff read 0x0080
ide_data_test: wrote 0xaaaa read 0x0080
ide_data_test: wrote 0x5555 read 0x0080
ide_data_test: wrote 0x0000 read 0x0080
ide_data_test: wrote 0xffff read 0x0080
ide_data_test: wrote 0xaaaa read 0x0080
ide_data_test: wrote 0x55
ide_data_test: wrote 0x0000 read 0x0080
ide_data_test: wrote 0xffff read 0x0080
ide_data_test: wrote 0xaaaa read 0x0080
ide_data_test: wrote 0x5555 read 0x0080
ide_data_test: wrote 0x0000 read 0x0000
ide_data_test: wrote 0xffff read 0xffff
ide_data_test: wrote 0xaaaa read 0xaaaa
ide_data_test: wrote 0x5555 read 0x5555
hda: IBM-DJSA-220, ATA DISK drive
ide0 at 0x000-0x007,0x038 on irq 6
hda: IBM-DJSA-220, 19077MB w/1874kB Cache, CHS=38760/16/63
empeg-flash driver initialized
smc chip id/revision 0x3349
smc9194.c:v0.12 03/06/96 by Erik Stahlman ([email protected])

SMC9194: SMC91C94(r:9) at 0x4008000 IRQ:7 INTF:TP MEM:6144b MAC 00:02:d7:12:03:f
c
Partition check:
hda: hda1 < hda5 hda6 > hda2 hda3 hda4
RAMDISK: ext2 filesystem found at block 0
RAMDISK: Loading 320 blocks [1 disk] into ram disk... done.
EXT2-fs warning: checktime reached, running e2fsck is recommended
VFS: Mounted root (ext2 filesystem).
empeg-pump v0.03 (19980601)
Press Ctrl-A to enter pump...VFS: Mounted root (ext2 filesystem) readonly.
change_root: old root has d_count=1
Trying to unmount old root ... Freeing unused kernel memory: 4k initempeg init 0
.8
I see this is a developer image!
Mounting proc
Mounting first music partition
Tried to mount /dev/hda4 as reiserfs but got error 19
Mounting second music partition
Tried to mount /dev/hdc4 but got error 6
Error mounting partitions (possibly already mounted)
Remounting first music partition read-only
Remounting second music partition read-only
No secondary hard disk
Press 'q' now to go into development mode. You Have Zero Seconds To ComðStarting
player
Timezone: US/Eastern
Hijack: intercepting config.ini

hijack: removed menu entry: "Serial Port Assignment"
khttpd: listening on port 80
kftpd: listening on port 21
player.cpp : 587:empeg-car 3.00-alpha11 2005/07/25.
Prolux 4 empeg car - 2.1485 Jul 25 2005
Vcb: 0x4041a000

They are "playlist_fid" dot "track_number". Playlist 0000 is your "master" (down-down-down) playlist, and the track numbers seem to be just about everything.

So something else is wrong, nobody has that many bad tracks.

A few possibilities:

• Bit errors in the empeg SRAM chips (dubious).
• IDE header issues (again doubtful in this case, because it should cause even more severe issues).
• Not enough available RAM for the player (still possible, I guess: try ReserveCache=64 ?).
• Bad data in the player's dynamic data partition.

Several others here have been unable to run v3alpha successfully, and simply went back to v2.01 without the same issues. You could do that, I suppose. But if most of us *can* run v3, then there's gotta be something fixable.

The variable is the dynamic data partition, aka /dev/hda3. You could zero that out and see if things improve.

So, what I would do in your situation is the following:

• Run set_empeg_max_fid to patch the player for large numbers of tracks. Regardless.
• Then, nuke the dynamic data partition: connect via serial, and hit Control^C to kill off the player.
• Now nuke the dynamic data, and force a database rebuild, using the following sequence of commands:
  Code:

  cat /dev/zero > /dev/hda3
  rwm
  rm /empeg/var/[dtp]*
  player -i
  ## wait for the player to rebuild databases and start up,
  ## then hit Control^C again, and continue with:
  rom
  exit

Good luck!

I thought there were known crash bugs with certain 3d visuals on the player?

Exactly. My v3alpha player works great except it crashes with Borgbeat. All other visuals (at least the ones in my favorites) are ok.

Does anyone know if this crash is restricted to V3 alpha?

Flashman, is there a reason you're running the alpha? The main feature that V3 alpha gives you are OGG and FLAC playback. If your collection isn't in those formats, you'd probably encounter fewer problems with 2.0.

V3 also gives superb cross-fading and gapless playback. I run it on all of my players.

But it can be flakey. We used it tonight just for FM reception, and the code was still doing sigkills at regular intervals. So that probably rules out the dynamic data partition as the cause.

Maybe a visual -- the spinning cube thingie was up on this one here while it crashed every 3-5 minutes, until I switched to one of the bar graph visuals. Was stable after the switch.

That doesn't prove anything, but it does suggest that more experimentation might beef up some evidence of visual malfunctions..

Cheers

Well, Before I tried the additional changes I decided to let it run quite a bit over the weekend on different visuals...

It never failed.

However, the moment I go back to Borgbeat... BOOM!...
Ok, so seeing as how I ended up on Borgbeat by accident and it is not my favorite visual anyway... It really is no big deal to me... Guess I should have put the two together when I first saw the error that the only thing that had changed on the player was the visual.
I run V3 on all my players.. Doh! I just realized I should try that visual on the other players...

I will still try your suggestions sometime over the week out of sheer curiosity.

Ok and the last question... do you guys suppose we may ever see anything post V3???

Thanks again for everyone's comments!!

That would be Borgbeat.

Perhaps a FAQ entry with known issues on the latest v3alpha would be appropriate?

Ahh.. Okay, so to save me a day or three reverse-engineering things..

Peter, Roger, Toby?

Is there some easy byte to twiddle in a file somewhere to either (1) fix this, or (2) prevent Borgbeat from being selected by the Autochanger ?

Just knowing the file layout of visuals.bin would likely be enough information for this..

Mmm... the offending visual, BorgBeat, seems to be the very first named visual inside that file.

So it's probably "visual 0" or some such thing, which may have something to do with why that particular visual trashes things.

??

.. and I suppose one thing to try would be the v2.01 visuals.bin file with the v3a11 player..

The visuals.bin file is written (mostly) in ARM assembler, so it won't be obfuscated by the compiler, although it might have been obfuscated by Toby's coding style

I only saw the API documentation for the visuals once (yeah, we did have some), but ISTR that there's a table (or a table of tables) near the front of the file that points to the visuals. It might be easy to knock out the Borgbeat one from the list.

Mike or Toby would be the go-to people for the info, though. Or maybe John G -- ISTR that he and Toby reworked some of the visuals for v3.

I know they removed or disabled the "Lavatory Floor" visual because it was causing crashes (thread here). A version of hijack was released to recover from the same problem (thread here).

Yes, I could do a similar "crash recovery" fix for this one, too.

But what we'd really like is a way to just eliminate the troublesome visual altogether, so there would be fewer crashes that need recovering from.

Cheers

I experience the same problem Sigkill error while on the bar graph visual. I can not remeber the exact name of it but it is one of the 32 ones. I would get the SIGKILL error and then it would reboot and when it came back up it would be on the Now and Next info screen. It would do it on any track and at any time. I would also get the error on the Now and Next info screen every now and then. While I was trying to diagnose what my problem was I took my player apart, disconnected and then reconnected the IDE cable, and that seemed to make it go away for awhile. So I had conculded that I have a bad IDE cable or connector. It has started to do it a lot now so when ever I get a chance I was going to open it back up and take a good look at the IDE header and maybe go ahead and replace the IDE cable.

Just my 1/2cent

Only on the Alpha player software, correct?

Congrats to all! I was able to fix my problem without bothering anyone... I concur with everyone who feels like this is THE BEST forum on the web!



These instructions worked to stop a segfault/segkill error that was restarting the player right after it was "Loading Visuals". I'm using V3 Alpha and the visual looks like it was the (stock, not modified) "Image Pan" visual that was crashing the player.

At first, I tried loading the "nukesavearea2.zImage" (from this post) but that didn't work.

So thank you!!!

Well done!