While driving today, the fuse blew in the empeg's wire harness....

Later tonight, I was playing the empeg for about the same time (1 hour or so) and while I was out of the room, I smelled something burning. The empeg was pulsing it's LED but did not resond. It was set to play on repeat, so something was wrong (Capt. Obvious here).

I opened the player and saw this part looked like burnt toast... any idea what it is? I know it's on the fan controller, but that's about it. Why is it getting so hot and by cooking like this, why would the player stop at "probing primary interface"? I assume it's not gettting beyond the kernal?

I swapped my drives to my backup player and SEG FAULT appeard on the screen over and over...

TeraTerm Pro isn't storing much of the boot log, but I was able to copy this:
Code:

---

  player.cpp        : 579:empeg-car 3.00-alpha10 2005/07/09.
! memory_stream.cpp :  42:MemoryStream failed to allocate memory.
! tunedb_disk.cpp   : player(21): user memory violation at pc=0x0206687c, lr=0x0
0000010 (bad address=0x000004c0, code 2)
878:Failepc : [<0206687c>]    lr : [<00000010>]
sp : bffffa78  ip : bffffa90  fp : bffffa8c
d to builr10: 021d543c  r9 : 0217c144  r8 : bffffacc
d database. Br7 : 00000010  r6 : 0220b350  r5 : 00000000  r4 : bffffa90
ad. 0xc00r3 : 00000010  r2 : bffffa90  r1 : 00000000  r0 : 000004c0
7000e
Flags: nZCv  IRQs on  FIQs on  Mode USER_32  Segment user
Control: C0EF917D  Table: C0EF917D  DAC: 00000015
Function entered at [<02066864>] from [<02067268>]
 r5 = BFFFFACC  r4 = BFFFFA90
Function entered at [<0206722c>] from [<020681a0>]
 r5 = 000004C0  r4 = 0220B340
Function entered at [<02068138>] from [<0203c38c>]
 r10 = 021D543C  r9 = 0217C144  r8 = 00000000  r7 = 021D543C
 r6 = 00000000  r5 = BFFFFB74  r4 = 0220B340
Function entered at [<0203c2a0>] from [<0200d940>]
 r8 = BFFFFDB8  r7 = BFFFFCD0  r6 = 021D52C4  r5 = 021D5308
 r4 = BFFFFB74
Function entered at [<0200d900>] from [<02039c68>]
 r6 = 0217C144  r5 = 00000000  r4 = BFFFFB74
Function entered at [<020399c0>] from [<020398d0>]
 r10 = 0202ADAC  r9 = 00000000  r8 = 00000001  r7 = BFFFFF64
 r6 = BFFFFE08  r5 = BFFFFCD0  r4 = BFFFFCD0
Function entered at [<02039890>] from [<0203b080>]
 r6 = BFFFFE08  r5 = FFFFFFFF  r4 = BFFFFCD0
Function entered at [<0203af18>] from [<0203b1dc>]
 r8 = 00000001  r7 = BFFFFF64  r6 = 021D5374  r5 = 00000001
 r4 = 021D529C
Function entered at [<0203b13c>] from [<0202adbc>]
 r8 = 0217B568  r7 = 00000000  r6 = 020000C0  r5 = BFFFFF64
 r4 = 00000001
Function entered at [<0202adac>] from [<02131bb0>]
Function entered at [<02131ad4>] from [<02000104>]
 r10 = 00000000  r8 = 00000000  r7 = 00000000  r6 = 00000000
 r5 = 00000000  r4 = 00000000
Restored terminal settings
Remounting first music partition read-only
Remounting second music partition read-only
Abnormal player termination
Player received signal 11
Starting playTimezone: Atlantic/South_Georgia
er
Hijack: intercepting config.ini

Using non-standard cache size -80 (bonus 0Mb, adjustment 180)
  player.cpp        : 579:empeg-car 3.00-alpha10 2005/07/09.
! memory_stream.cpp :  42:MemoryStream failed to allocate memory.
! tunedb_disk.cpp   : player(25): user memory violation at pc=0x0206687c, lr=0x0
0000010 (bad address=0x000004c0, code 2)
878:Failepc : [<0206687c>]    lr : [<00000010>]
sp : bffffa78  ip : bffffa90  fp : bffffa8c
d to build dar10: 021d543c  r9 : 0217c144  r8 : bffffacc
tabase. Br7 : 00000010  r6 : 0220b350  r5 : 00000000  r4 : bffffa90
ad. 0xc00r3 : 00000010  r2 : bffffa90  r1 : 00000000  r0 : 000004c0
7000e
Flags: nZCv  IRQs on  FIQs on  Mode USER_32  Segment user
Control: C0EF917D  Table: C0EF917D  DAC: 00000015
Function entered at [<02066864>] from [<02067268>]
 r5 = BFFFFACC  r4 = BFFFFA90
Function entered at [<0206722c>] from [<020681a0>]
 r5 = 000004C0  r4 = 0220B340
Function entered at [<02068138>] from [<0203c38c>]
 r10 = 021D543C  r9 = 0217C144  r8 = 00000000  r7 = 021D543C
 r6 = 00000000  r5 = BFFFFB74  r4 = 0220B340
Function entered at [<0203c2a0>] from [<0200d940>]
 r8 = BFFFFDB8  r7 = BFFFFCD0  r6 = 021D52C4  r5 = 021D5308
 r4 = BFFFFB74
Function entered at [<0200d900>] from [<02039c68>]
 r6 = 0217C144  r5 = 00000000  r4 = BFFFFB74
Function entered at [<020399c0>] from [<020398d0>]
 r10 = 0202ADAC  r9 = 00000000  r8 = 00000001  r7 = BFFFFF64
 r6 = BFFFFE08  r5 = BFFFFCD0  r4 = BFFFFCD0
Function entered at [<02039890>] from [<0203b080>]
 r6 = BFFFFE08  r5 = FFFFFFFF  r4 = BFFFFCD0
Function entered at [<0203af18>] from [<0203b1dc>]
 r8 = 00000001  r7 = BFFFFF64  r6 = 021D5374  r5 = 00000001
 r4 = 021D529C
Function entered at [<0203b13c>] from [<0202adbc>]
 r8 = 0217B568  r7 = 00000000  r6 = 020000C0  r5 = BFFFFF64
 r4 = 00000001
Function entered at [<0202adac>] from [<02131bb0>]
Function entered at [<02131ad4>] from [<02000104>]
 r10 = 00000000  r8 = 00000000  r7 = 00000000  r6 = 00000000
 r5 = 00000000  r4 = 00000000
Restored terminal settings
Remounting first music partition read-only
Remounting second music partition read-only
Abnormal player termination
Player received signal 11
Starting playTimezone: Atlantic/South_Georgia
er
Hijack: intercepting config.ini

Using non-standard cache size -80 (bonus 0Mb, adjustment 180)
  player.cpp        : 579:empeg-car 3.00-alpha10 2005/07/09.  

---

Is this due to my config.ini having a ReserveCache of 180 on a stock Mk2a?

I'm going about removing my config.ini as per here.

I also noticed that Mark Lord mentinoed SegFault indicates some internal error here.

I'm trying to make a new config.ini file by Configuring the player in emplode, but emplode crashes on sync. On boot, I can browse the player (segfaults are not on every boot now) and skip through playlists, but the player WILL NOT PLAY .

In addition to the removal the config.ini, I did a "./player -i" but no luck. I'm still not able to get a song to play. It stays at 0:00.

Well... I finally decided to revert to 2.00final and all is well again. I'm hoping that the install of 2.00 helped more than it being "not alpha10" because I'd really like to get 3a10 to work - or at least 2.01.

Sorry about the panic. I value my hard drives and their contents (years of tweaking playlists) more than I value my empeg. I don't know how long the empeg was smoking before I came back into the room or if anything on the drives got fried. Since swapping the drives to another player brought about new problems, I feared the the drives were fubared along with the fan controller.

Maybe the smoke clouded my judgement.

This is the only thing I noticed during the standard database rebuilds that go with switching to 3 to 2: Code:

---

  ext2fs_check_if_mount: No such file or directory while determining whether /dev/
hdc4 is mounted.
! local_audio_component.cpp:1831:Invalid bitrate tag for FID 0x643d0.

---

Sorry, I replied in the other thread.

2.00 is unaware of extra memory, so you do not have to worry about the caching subsystem grabbing gobs of space and stuffing audio in it to the exclusion of all else. Now that you have stability at 2.00, you might want to move to 2.01. Memory caching will be an issue here, and you probably want your reservecache variable at a high number to lock out as much memory as needed to keep your player responding.

I would recommend deleting and rebuilding the databases, along with an fsck, just to be safe. The errors you mentioned are not dangerous, to my knowledge. It just looks like a bad tag and that your mount table doesn't have the second drive in it (this is normal unless you tweak it).

Okay, I guess I didn't understand how the 2.01 and 3.x handle extra memory.

Basically, 2.01 and 3.0 are aware of the memory and try to use it. The issue (at least one that I know of and have hit before) is the aggressive caching of the player application. It is great for limited memory, filling the buffer with audio and keeping the drives spun down. The problem is that the cache tries to fill as much as it can. With large amounts of memory available, the caching routine can choke out other applications and cause them to fail due to memory resource starvation.

There are better descriptions in other threads, but it comes down to the fact that the bulk disk reads are done in kernel memory. Competing applications (player, telnet, other daemons) cannot allocate the RAM needed while these reads are done, so those applications give memory allocation failures.

The reservecache option forces the player software not to use the extra space for caching, keeping cache operations quick and small enough not to interfere with the kernel (and other software) operations. The practical result is that 2.01 can use the extra memory, but you have to force it not to use that memory for cache beyond a certain "breakpoint".

Note that the memory is still very useful for hijack and other purposes.

Most likely bad connections on the extra memory..