pthread concurrent thread limit

Robert Collins robert.collins@itdomain.com.au
Tue Jun 26 07:06:00 GMT 2001


I've got a problem! So I'm asking for some assistance...

I'm working through the win32-pthreads test suite, trying to ascertain
the current practical quality of the pthreads code.

The attached 30kb (I figure on a small list that is small enough)
tarball has a testsuite. If someone of you folk could run the suite and
report back... I'd appreciate it.

The tarball only tests the number of threads that can be created.
(Actually it does 2 pre-requisite tests to prevent erroneous results).

(./configure && make check) on my system gives 1 failure, and the last
lines of the output look like:
my threadresult = 0 i=50
my thread
result = 11 i=52
FAIL: condvar2_1

If anyone who has the time to try this gets different results (higher or
lower 'i', or a different result to 11) knowing that + your system
configuration (winNT/2k etc, and installed memory) would be useful to
me.

Details:
The win32 CreateThread call is failing with error 8 - Not enough storage
is available to process this command. - and I've got no idea why. It
doesn't seem to be a Tls problem, and the c++ object has been created.
The process is using 6Mb of ram (virtual + physical) at the failure
point... which is _always_ 52. I've tried closing all my process's,
pulling my hair out, and no difference was observed.

thoughts, hints, suggestions, peanuts?

Rob
-------------- next part --------------
A non-text attachment was scrubbed...
Name: pthreads_test_suite-0.1.tar.gz
Type: application/x-gzip
Size: 37454 bytes
Desc: not available
URL: <http://cygwin.com/pipermail/cygwin-developers/attachments/20010626/e91804cc/attachment.bin>


More information about the Cygwin-developers mailing list