Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 31 Jan 2006 19:07:05 +0000 (GMT)
From:      Robert Watson <rwatson@FreeBSD.org>
To:        John Baldwin <jhb@freebsd.org>
Cc:        Peter Jeremy <peterjeremy@optushome.com.au>, src-committers@freebsd.org, cvs-all@freebsd.org, cvs-src@freebsd.org
Subject:   Re: cvs commit: src/sys/conf files options src/sys/kern kern_rwlock.c subr_lock.c src/sys/sys lock.h rwlock.h
Message-ID:  <20060131190547.Y95776@fledge.watson.org>
In-Reply-To: <200601311340.51250.jhb@freebsd.org>
References:  <200601272313.k0RNDQHI064747@repoman.freebsd.org> <20060128141445.GC2341@turion.vk2pj.dyndns.org> <200601311340.51250.jhb@freebsd.org>

next in thread | previous in thread | raw e-mail | index | archive | help

On Tue, 31 Jan 2006, John Baldwin wrote:

> On Saturday 28 January 2006 09:14, Peter Jeremy wrote:
>> On Fri, 2006-Jan-27 23:13:26 +0000, John Baldwin wrote:
>>>  Log:
>>>  Add a basic reader/writer lock implementation to the kernel.
>>
>> Thank you.  This looks interesting.
>>
>>>  Tested on:      i386 (4 cpu box with a kernel module that used 4 threads
>>>                  that randomly chose between read locks and write locks
>>>                  that ran w/o panicing for over a day solid.  It usually
>>>                  panic'd within a few seconds when there were bugs during
>>>                  testing. :)  The kernel module source is available on
>>>                  request.)
>>
>> Can I suggest that this module be committed into tools/regression or
>> similar so that it is generally available.
>
> I can I guess.  I have 3 currently (crash, evtest, and crash2 (which is the 
> latest one that uses multiple threads, crash just has a single thread)).

It would be nice of we had something on the order of src/sys/test, where we 
could put test kernel code that is never intended to see production.  The 
problem with putting things in src/tools is that the code gets very stale, 
very quickly, as kernel APIs move and the implementation in that tree doesn't. 
I have some synchronization micro-benchmarks and memory allocation 
micro-benchmarks I use as part of the netperf work, and they would similarly 
benefit from a central but non-productionable place in the src/sys tree.

Robert N M Watson



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20060131190547.Y95776>