Re: bug#32452: 26.1; gnutls_try_handshake maxes out cpu retrying when server is a bit busy

[Lars Ingebrigtsen]

Paul Eggert <eggert@cs.ucla.edu> writes:

> This diff is wrong, as it omits a line
> "GL_COND_OBJ_NANOSLEEP_CONDITION = @GL_COND_OBJ_NANOSLEEP_CONDITION@".

Yup, looks like it.

> I ran what should have been something like your commands and got the
> attached patch. One way forward is for you to simply install the
> attached patch and move on from there. Or we can continue to look into
> why things work for me and not for you. I suppose it could be an
> Autoconf bug on your platform, but it'd be an odd one.

As far as I can tell, I'm running exactly the same commands you are, but
I end up with a result that misses the line up there.  

Let's see...  I've been doing this on my development laptop (which uses
debian/bookworm (i.e., "testing").  I now re-ran the merge-gnulib on a
debian/bullseye (i.e., "stable"), and I get the same results as you now
(i.e., things work).

So either there's something weird on my laptop, or it sounds like
there's an Autoconf bug in debian/bookworm?

-- 
(domestic pets only, the antidote for overdose, milk.)
   bloggy blog: http://lars.ingebrigtsen.no