On Thu, Jun 02, 2005 at 09:37:51AM -0700, Ken Simpson wrote: > It _may_ (just because I have a gut feeling, not because of actual > consideration) be a socket options issue that may be related to your > platform. I had a very similar problem with OpenBSD last summer > wherein the socket blocking status could not be changed. The fix was a > very simple change to the configure setup for APR.
Yes, in all the tests that are failing for me, I see where it is checking for OpenBSD or Apache >2.0.52 which must have been when the bug was fixed. > Here's the patch that fixed that problem -- I haven't been able to > find the discussion thread yet. Thanks for the link. My version of apr_hints.m4 already has this code applied. In fact, in another thread back in February, Joe had asked me to set the ac_cv_o_nonblock_inherited value before running the tests but it did not help. Can you tell me more about the bug you had? What does a "socket options issue" mean? Do you know what ac_cv_o_nonblock_inherited is doing when it gets set? Thanks, William -- Knowmad Services Inc. http://www.knowmad.com
