Re: HEADS UP: UNIX domain socket locking changes merged to CVS HEAD

From: Robert Watson <rwatson_at_FreeBSD.org>
Date: Thu, 1 Mar 2007 09:33:33 +0000 (GMT)
On Thu, 1 Mar 2007, Scott Robbins wrote:

>> So the good news is once you can build .. it seems to be the problem is 
>> solved.. of course I still could lock up.. but only using X will tell :-0
>
> Yes, I'm almost afraid to startx.  I won't really get a chance till tonight, 
> but Robert, thank you.  (Also, I've lost track of who was the one who was 
> able to get the trace, but thanks are also due there.)

No problem, and sorry about the deadlocks.  I'm was a bit surprised they 
didn't turn up in the rather heavy UNIX domain socket testing with MySQL, but 
it occurs to me that traffic with MySQL is largely synchronous due to it being 
a query transmitted completely, processing, and then a reply transmitted 
completely.  As a result, there was likely very little simulteneous 
transmitting and receiving on the same socket at the same time.  It sounds 
like a new regression test is called for...

Robert N M Watson
Computer Laboratory
University of Cambridge
Received on Thu Mar 01 2007 - 08:33:34 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:39:06 UTC