Hi, is there any particular reason why httpd does not automatically fall back to read/write if sendfile failed [1]? Or is the only problem that nobody has written the code yet? I have googled a bit but have not found any discussion about this.
Cheers, Stefan [1] The linux sendfile man page has: Applications may wish to fall back to read(2)/write(2) in the case where sendfile() fails with EINVAL or ENOSYS.