Re: TCP stalls in current git, possibly splice related

[Jens Axboe <jens.axboe@oracle.com>]

On Wed, Jul 18 2007, Johannes Berg wrote:
> On Mon, 2007-07-16 at 14:02 +0200, Jens Axboe wrote:
> 
> > Yep, it's a sender thing, so upgrading the receiver will not change
> > anything.
> 
> Ok, I upgraded, but that didn't help. And in fact, I don't see how it
> could have since synergy doesn't use splice or sendfile. I should've
> thought of that right away, sorry.
> 
> It seems that packets are actually coming in during the time that my
> mouse hangs though (ran wireshark in parallel and saw no pauses in the
> timeline.) Hence, it actually seems to be on the receiver side, and
> running the synergy client under strace reveals that during the time my
> mouse hangs it's in poll() waiting for input on the tcp socket. sysrg-t
> doesn't show anything useful, it's just scheduling waiting for data.
> According to wireshark data is sent, but it never shows up at the
> application layer.

OK, then we can put splice off the hook at least :-)

If it's easily reproducible (and it sounds like it), then a git bisect
might be the easiest way forward.

-- 
Jens Axboe