[darcs-users] Darcs 2.4 - performance testing (stage 0)
me at worldmaker.net
Wed Jan 13 03:52:17 UTC 2010
On 1/12/2010 22:10, Jason Dagit wrote:
> Obviously the darcs-stats file is in pretty high demand and
> darcs-benchmark doesn't seem to be waiting long enough between runs
> to close the file correctly (or otherwise doesn't have the proper
> sharing permissions setup for the file).
> The haskell run-time created it for the duration that darcs is executing
> and I bet darcs tried to access while it was running. I think there are
> essentially two solutions:
> 1) Create the file elsewhere
> 2) make it a non-fatal error
It's not a fatal error for the benchmark, it keeps on going. I think
that is directly and indirectly related to the later issues (such as the
second one I posted). darcs-benchmark probably needs some sort of:
"failed with error? then sleep a few seconds and retry".
It's not in the repository, its in darcs-benchmark's "_playground"
parent directory. The error is from the haskell run-time, even it looks
like it might be darcs reporting it (what my earlier snippet didn't show
is that its followed immediately by GHC dumping its RTS options help
It seems that GHC itself is fighting for access rights to that file. I'm
not sure if its worth reporting as a GHC bug, but it the easy solution
is probably some sort of "cooldown" in darcs-benchmark between darcs
calls to help insure that GHC has relinquished that file between
> Well, that's also interesting because it couldn't find darcs-2.3.1
> in its place in Haskell's bin directory (despite accessibility in
> PATH), but it found this other Bin directory in my PATH.
> Is it possible that the order of your paths is interfering?
Well, now that I think of it, one is in my account PATH (which works
from darcs-benchmark) and one in the system-wide PATH (which apparently
> Anyway, it doesn't look like darcs-benchmark wants to run correctly
> on Windows Vista just yet.
> I'm still troubleshooting the OS X case too. Thanks for trying to
> figure this out!
Interestingly enough, I gave it some time to get further in the process
and it seems to hit the same "hanging" issue that you've seen, appearing
to just fall asleep upon hitting the second big get test.
It would probably be helpful if darcs-benchmark run --verbose would
pass-through darcs' progress reports.
More information about the darcs-users