Hi,
apt-get sometimes fails on HTTP downloads with EINTR from a select() call.
It was not looping over select() as it does at other places.
I filed a bug report, and now Jason says that there shouldn't be any signals
to the HTTP process causing an EINTR.
I am not fit in this area. How can I trace what signal was causing the
EINTR? Are there any expected extra signals in the Hurd, or is there a bug
that is causing spurious EINTR failures in select(), without a signal actually
being raised? Or a wrong signal being raised?
The existing code works fine in Linux, so there is obviously
something subtle going on.
Thanks,
Marcus
_______________________________________________
Bug-hurd mailing list
[EMAIL PROTECTED]
http://mail.gnu.org/mailman/listinfo/bug-hurd
