I have a MPC5200 system with Flash Filesystem. I am running a test from the flash that loads some diagnostics into the file in the same directory. I can run the test only once. The second time my system “gracefullyâ€
Hum, the program that cause the crash is devf-generic. That’s the flash driver, what every your application is doing this should never happend. Unless there is a hardware bug (bad memory, power supply) you probably will have to report this to QSS as they are the only one that can fix it.
When I start the network drivers I can get a different error.
testMain
/bin/ksh: testMain: Not enough memory
Shutdown[0,0] S/C/F=11/1/11 C/D=00038028/0008d854 state(c0)= now lock
[0]PID-TID=315407-1? P/T FL=00088014/00000000
[0]ACTIVE PID=1 PF=00019001 “proc/boot/procnto-600”
[0]ASPACE PID=1 PF=00019001 “proc/boot/procnto-600”
ppcbe context[03fedeb0]:
0000: 48002103 03fedf60 00092568 00088bb8 00000000 00000000 48002103 48002103
0020: 00000021 80021030 00000000 00000002 00000000 0009451c 00000000 00000000
0040: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 0008d620
0060: 00008000 00000000 00000001 03a906d8 0334b908 00000000 0334b8d8 00000000
0080: 0006d520 000417dc 40089032 00041800 40000000 20000000 00000001 00000000
00a0: 0334b718
instruction[00041800]:
81 29 ff f8 48 01 6f 71 3c 60 00 09 38 63 8b a4 38 80 04 7b 48 01 c3 01 a1 3f
stack[03fedf60]:
0000: 03fedfa0 000417dc 40089032 00041800 40000000 20000000 0334b4À
This error seems different but if I have some resource shortage it may give a clue. For example a stack corruption can produce any bogus error.
No a stack corruption will result in a segment violation, not a kernel crash.
There is something wrong with the hardware or you BSP is not perfectly adapted with the hardware.