Rockbox

Tasklist

FS#6193 - iPod Photo reboots while playing

Attached to Project: Rockbox
Opened by Matthias Urlichs (smurf) - Monday, 16 October 2006, 07:36 GMT
Last edited by Steve Bavin (pondlife) - Wednesday, 10 January 2007, 18:07 GMT
Task Type Bugs
Category Music playback
Status Closed
Assigned To No-one
Operating System iPod 4G Color
Severity High
Priority Normal
Reported Version
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 0
Private No

Details

During playback, about every half hour or so, my iPod 4G Color reboots. The boot sequence stops with "Rockbox loaded", i.e. before the RockBox splash screen, and I have to reboot it again manually. The problem seems to be unrelated to specific MP3 contents, as the second time around Rockbox will not die again at the same place.

I have tried the current daily build as well as the earliest one that's still available (2006-09-20). Both fail. The version I was using before, 2006-08-19, works correctly.
This task depends upon

Closed by  Steve Bavin (pondlife)
Wednesday, 10 January 2007, 18:08 GMT
Reason for closing:  Duplicate
Additional comments about closing:  Seems to be the same iPod CPU scaling crash as reported on  FS#5264  and  FS#5710  .
Comment by Barry Wardell (barrywardell) - Tuesday, 17 October 2006, 10:05 GMT
Hmm. The system_reboot() function was implemented for the iPods on the 20th of September (http://www.rockbox.org/viewcvs.cgi/firmware/system.c.diff?r1=1.119&r2=1.120). I wonder if your problem still happens with that change reversed? Could it be that somehow it is being called and has only been noticed now that it actually does something?
Comment by Matthias Urlichs (smurf) - Wednesday, 18 October 2006, 04:55 GMT
Hmm. Current CVS build (self-built) now does a "data abort at 4fb94".
Comment by Matthias Urlichs (smurf) - Wednesday, 18 October 2006, 08:08 GMT
... which is here:

0x0004fb74 <set_cpu_frequency+168>: add r1, r12, #24576 ; 0x6000
0x0004fb78 <set_cpu_frequency+172>: ldr r3, [r1, #32]
0x0004fb7c <set_cpu_frequency+176>: bic r3, r3, #-268435456 ; 0xf0000000
0x0004fb80 <set_cpu_frequency+180>: bic r3, r3, #240 ; 0xf0
0x0004fb84 <set_cpu_frequency+184>: mov r2, #536870912 ; 0x20000000
0x0004fb88 <set_cpu_frequency+188>: add r2, r2, #112 ; 0x70
0x0004fb8c <set_cpu_frequency+192>: orr r3, r3, r2
0x0004fb90 <set_cpu_frequency+196>: str r3, [r1, #32]
0x0004fb94 <set_cpu_frequency+200>: add r2, r12, #16384 ; 0x4000
0x0004fb98 <set_cpu_frequency+204>: ldr r3, [r2, #36]

and seems to be the interrupt-reenable going haywire. Somebody with more hardware knowledge please look at this ...
Comment by Barry Wardell (barrywardell) - Wednesday, 18 October 2006, 08:24 GMT
Ah, yes. It is indeed cpu frequency scaling. It's a known bug that it is currently causing crashes. There is a patch which is working on fixing this http://www.rockbox.org/tracker/task/6095 .

Loading...