On 2/5/20 10:04 pm, Anders Montonen wrote:
A couple of times I’ve had source-builder hang seemingly randomly, with the
following Python backtrace:
Do you see this problem with the bootstrap script 'rtems-bootstrap' in
the top of the rtems.git rep?
I think the RSB version could be retired now the rtems.git repo has a
decent version, plud this is the last autotools release of RTEMS so the
need to bootstrap will go and both will go.
However fixing execute.py would be nice ...
Exception in thread _stderr[]:
Traceback (most recent call last):
File "/usr/lib/python3.8/threading.py", line 932, in _bootstrap_inner
self.run()
File "/usr/lib/python3.8/threading.py", line 870, in run
self._target(*self._args, **self._kwargs)
File "/home/anders/work/rtems/rsb/source-builder/sb/execute.py", line 204,
in _readthread
data = data.decode(sys.stdout.encoding)
UnicodeDecodeError: 'utf-8' codec can't decode bytes in position 4094-4095:
unexpected end of data
My guess is that since the loop in execute._readthread() is reading in fixed
4096-byte chunks, occasionally the boundary will fall in the middle of an UTF-8
code point, causing the decode call to throw an error. Since source-builder
stops reading the file handle at that point, it will then deadlock the child
process. Unfortunately, I haven’t yet been able to reliably trigger the issue.
Usually, re-running the same build command succeeds.
Interesting and your analysis sounds right. I am not sure how this could
be handled. A test case with execute.py would be a good start, see the
end of the file.
Chris
_______________________________________________
users mailing list
users@rtems.org
http://lists.rtems.org/mailman/listinfo/users
