You can already override the timeout with -timeout. I don't think it makes sense to just multiply the -timeout by -count because you'd very quickly get into unfeasibly long timeouts.

@quentinmit In the case of benchmark, I really do want to wait the full length. Some benchmarks have to be run with high counts to get good accuracy. It's annoying to add the timeout every time.

Please reconsider opening this bug.

Duplicate of #12446

A possible workaround is to use go test -c, then run the resulting
$PKG.test -test.timeout=NNNs -test.count=NNN binary.

On Wed, Oct 5, 2016 at 9:54 AM, Carl Mastrangelo notifications@github.com
wrote:

@quentinmit https://github.com/quentinmit In the case of benchmark, I
really do want to wait the full length. Some benchmarks have to be run with
high counts to get good accuracy. It's annoying to add the timeout every
time.

Please reconsider opening this bug.

—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
#17337 (comment), or mute
the thread
https://github.com/notifications/unsubscribe-auth/AAAcA7CuTa7m5TAVkc3qT8D8R9OaoQCvks5qwtk0gaJpZM4KOEgv
.

@davecheney it's risky, because if the timer is even off by a little, it would fire too early (and also ignores the -test.benchtime). It would be better to not have a work around at all, and the test runner do the right thing by default.

My own version was just to set the timeout to 20 hours, which is arguably not good either. @dsnet mentioned in his bug that it isn't straightforward, but I think it should still be solved (by me maybe, if there is agreement that it's a problem).