Thread VC 7.1 failures (Bronek Kozicki)
Any idea why some tests fail in this environment? They do all pass on my machine and appearantly also on Victor's. Stefan
Stefan Slapeta wrote:
Any idea why some tests fail in this environment? They do all pass on my machine and appearantly also on Victor's.
I do not know what's wrong. All tests in libs/thread/test succeeds when started from interactive account. Regular regression tests are run in background (on non-interactive desktop), as job scheduled under special account. Neither interactive account nor the one used by jobs have administrator priviledges. B.
At Thursday 2004-09-30 11:14, you wrote:
Stefan Slapeta wrote:
Any idea why some tests fail in this environment? They do all pass on my machine and appearantly also on Victor's.
I do not know what's wrong. All tests in libs/thread/test succeeds when started from interactive account. Regular regression tests are run in background (on non-interactive desktop), as job scheduled under special account. Neither interactive account nor the one used by jobs have administrator priviledges.
sometimes it's really baffling isn't it. Wish I could offer you some suggestions, but I can't think of any.
B.
_______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost
Victor A. Wagner Jr. http://rudbek.com The five most dangerous words in the English language: "There oughta be a law"
"Stefan Slapeta" <stefan@slapeta.com> wrote in message news:cjgsv2$gql$1@sea.gmane.org...
Any idea why some tests fail in this environment? They do all pass
on my
machine and appearantly also on Victor's.
It could well be a timing issue. Many of the Boost.Thread tests make the (usually valid) assumption that certain operations will finish in a specified "long" amount of time (e.g. one second). If the machine is running under heavy load, this assumption could be invalid, causing sporadic test failures. It's perhaps not the ideal way to run the tests, but that's the way I found them and I haven't been able to do anything about it yet. Mike
Michael Glassford wrote:
It could well be a timing issue. Many of the Boost.Thread tests make the (usually valid) assumption that certain operations will finish in
Right. I often (but not always) run regression tests under lower CPU priority than interactive processes. Sometimes it makes some tests run slower than expected. B.
participants (3)
-
Bronek Kozicki -
Michael Glassford -
Stefan Slapeta -
Victor A. Wagner Jr.