OS4 with recoverable error requester?

I'm getting some sporadic recoverable errors (OS4 Final). They are not very recoverable, though, since they appear again and again. The solution is a reboot.

My question is: shouldn't the Grim Reapers appear, instead of these old school requesters (that never appeared on previous updates)?

I have a strong feeling these are internet or MUI-related, since they appear mainly when using IBrowse (having said that, I use IB all the time, so it could be other thing...).
The text on the requester seems to always be the same. Is that info important in anyone helping me track out what's happening?
It's more annoying than serious, since it only happens every other session, but I'd like to see it solved.

Any hints or sugestions?

@BrandonLee

If you only do a softreset, you could try this:
Open a shell and type dumpdebugbuffer ram:crashlog.txt .
Maybe there is something usefull in it.

@ZeroG

I didn't know about that command. It will be very useful in the future (that's what I get never reading all docs...):

Here's the log, which is chinese to me. Maybe you can sort it out. ;)


a1ide.device 52.1 (09.12.2006)
[a1ide/ata_read_drive_properties] After-reset signature invalid for unit 1
it8212ide.device 52.1 (09.12.2006)
lsi53c8xx.device 52.1 (09.12.2006)
sii0680ide.device 52.1 (09.12.2006)
sii3112ide.device 52.1 (09.12.2006)
sii3114ide.device 52.1 (09.12.2006)
sii3512ide.device 52.1 (09.12.2006)
Mode 1280x1024x 8@75 79.976khz out of range
Mode 1280x1024x24@75 79.976khz out of range
Mode 1280x1024x16@75 79.976khz out of range
No SB128 found!
FILESAVE DriverInit
[VIA-AC97] Debug: AC-Link reset, primary codec not ready!
[VIA-AC97] Error: sorry, you don't seem to have the AC'97 codec!
Unable to initialize Card subsystem.
****************** Recoverable Alert ******************
Error: 0100000F Task: 0x6BA4EDA0
********************************************************
WARNING (UHCI Driver)!! Host controller 0 stopped (status=28)!
****************** Recoverable Alert ******************
Error: 0100000F Task: 0x6BA4EDA0
********************************************************
Dump of context at 0x01C86EE0
Trap type: DSI exception
Machine State (raw): 0x0000F030
Machine State (verbose): [ExtInt on] [User] [FPU on] [IAT on] [DAT on]
Temporary stack trace:
#0: in module Kickstart/kernel+0x000155AC (0x014155AC)
#1: in module Kickstart/kernel+0x0002FD00 (0x0142FD00)
#2: in module Kickstart/kernel+0x0001794C (0x0141794C)
#3: in module Kickstart/kernel+0x00017548 (0x01417548)
#4: in module Kickstart/kernel+0x00015F74 (0x01415F74)
#5: in module Kickstart/kernel+0x000160D0 (0x014160D0)
#6: in module Kickstart/dos.library.kmod+0x00001394 (0x014D9714)
#7: in module Kickstart/dos.library.kmod+0x000160A4 (0x014EE424)
#8: in module Kickstart/dos.library.kmod+0x0001615C (0x014EE4DC)
#9: in module Kickstart/dos.library.kmod+0x000165B0 (0x014EE930)
#10: 0x7FF555B8
#11: in module Kickstart/dos.library.kmod+0x00006BBC (0x014DEF3C)
#12: in module Kickstart/dos.library.kmod+0x00025E04 (0x014FE184)
#13: 0x800003EB
#14: 0x24404A80

Crashed process: IBrowse (0x6B25A1D0)
0: 00000000 65041F00 00000000 65D3CFDC 65D3CFDC 00000000 000000A0 00000000
8: 6B55129C 00000003 014189F4 65D3C5A0 28422024 00000000 00000000 FFFFFFFF
16: 194D2E81 01DA0000 00000000 00000001 01DA3B4A 00000000 00000000 00000000
24: 00000010 01DA0000 00000000 EFFF9B50 00000000 65D3CFDC 65D3CFD0 01C927D0
CR: 88222048 XER: 20000000 CTR: 014155B8 LR: 01429E00
DSISR: 42000000 DAR: 00000007
DSI verbose error description: Access not found in hash or BAT (page fault)
Access was a store operation
a1ide.device 52.1 (09.12.2006)
[a1ide/ata_read_drive_properties] After-reset signature invalid for unit 1
it8212ide.device 52.1 (09.12.2006)
lsi53c8xx.device 52.1 (09.12.2006)
sii0680ide.device 52.1 (09.12.2006)
sii3112ide.device 52.1 (09.12.2006)
sii3114ide.device 52.1 (09.12.2006)
sii3512ide.device 52.1 (09.12.2006)
Mode 1280x1024x 8@75 79.976khz out of range
Mode 1280x1024x24@75 79.976khz out of range
Mode 1280x1024x16@75 79.976khz out of range
No SB128 found!
FILESAVE DriverInit
[VIA-AC97] Debug: AC-Link reset, primary codec not ready!
[VIA-AC97] Error: sorry, you don't seem to have the AC'97 codec!
Unable to initialize Card subsystem.

@BrandonLee

I am no expert in reading crashlogs, but i would say that IB crashes.
Disable the 68k-JIT and see what happens.

Maybe it is a good idea to post the crashlogs in your other thread on AW.net too.

@BrandonLee

Quote:
The alert number can be found in the include file
<exec/alerts.h>:
Quote:
This doesnt say very much, just that the "memory lists"
seem to be corrupted and that doing a reset after saving
the current work to a non-important, temporary partition
may be a good idea...

Using the debug kernel with debuglevel 10 would allow you
to identify the task from the entries of the debug log,
but its not guaranteed that its the task which tried to
free the memory who has a bug.

BTW, any idea where the "Unable to initialize Card subsystem."
message comes from?

@ZeroG

IBrowse is already blacklisted. Previously, I also had quite a few Grim Reapers. Black-listing it solved that, but had no effect on the damn "recoverables"... :(

The crashlog has been posted on AW now.

@TetiSoft

The card initialization bits were (I think) from a couple of soundcard files I had on DEVS:, for some reason. I removed those now.

And about the debug kernel, I'm not very experienced using it. Can you give me some pointers on how to proceed?
On Grim Reapers, I used to have lockups when trying the debug option (doesn't seem to happen anyomre after OS4Final, though).

@BrandonLee

Boot your system loading "kernel.debug" instead of "kernel". You must either change your actual configuration in SYS:Kickstart/Kicklayout or add one with this kernel.debug

When the OS is loaded, type this into a Shell:

kdebug "console memory debuglevel 10"

At this point, the kernel will spit out lines of debug (slowing down the system a bit)
When the crash occurs, use the dumpdebugbuffer command to save the crashlog.

@Elwood

I'll give it a try. Since the crash is intermitent, it may take a while. Basically, I'll stick with a slower system until I catch the damn cause.
Is it just a case of changing the line where it says "kernel" to "kernel.debug" or do I need to do anything else?

This problem is really annoying me, I want to go to the bottom of it! ;)

@BrandonLee

The pre4 install cd contained a debug kicklayout in AmigaOS40_CD:Installation-Files/Kickstart.

To whom this may be of interest:

My problems seem to point to DepthToFront. Not sure if this tool is buggy or if it only happens in conjunction with other program (mainly IBrowse), but since removing it, I no longer get recoverable errors (knocks on wood...).

Anyone using DepthToFront without trouble?