Re: [ECOS] STM3210E crash

[Alan Bowman <alan dot michael dot bowman at gmail dot com>]

> Any ideas would be appreciated, particularly evidence that
> applications built in ROM mode work/don't work on other people's
> boards.

I have managed to get JTAG debugging of the failures, and it's not
very enlightening.  I have gone back to the code as it existed in the
repository, except that I have left the flash adjusted to 384kB.  The
configuration is as per the default, except that it is set to ROM
startup, and asserts are enabled.  I have flashed the board with the
tm_basic test, and this partially runs when I simply reset the board.

When I connect GDB and place a breakpoint at cyg_assert_fail, it
mostly triggers almost as soon as the test starts (use JTAG to reset
the board, then allow GDB to continue).  When the new thread is
created to run the test, the stack check usually fails.  This does not
seem to always be the case, although it seems to happen much more
frequently when the JTAG is connected than when I just let the program
run.

When I try to disable optimisation (replace -O2 in the global compiler
flags with -O0) I get a test that does not crash, but sits perpetually
in the wait_for_tick() function, waiting for a tick value to change
from 0.  It never does.

I am mystified.  Any clues would be more than appreciated...

Alan

-- 
Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos
and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss