Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 28 Jul 2005 00:34:56 +0100 (BST)
From:      Robert Watson <rwatson@FreeBSD.org>
To:        Pawel Jakub Dawidek <pjd@FreeBSD.org>
Cc:        cvs-src@FreeBSD.org, src-committers@FreeBSD.org, cvs-all@FreeBSD.org
Subject:   Re: cvs commit: src/sys/kern kern_malloc.c
Message-ID:  <20050728003334.V54330@fledge.watson.org>
In-Reply-To: <200507272317.j6RNHVHb086993@repoman.freebsd.org>
References:  <200507272317.j6RNHVHb086993@repoman.freebsd.org>

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

On Wed, 27 Jul 2005, Pawel Jakub Dawidek wrote:

>  Fix the way how "InUse" column in 'vmstat -m' output works:
>  - increase number of allocations count only on successfull malloc(9),
>    so it doesn't confuse people;
>  - because we need to check if 'size > 0', hide 'mtsp->mts_memalloced += size;'
>    under the check as well, as for size=0 it is of course a no-op;
>  - avoid critical_enter()/critical_exit() in case of failure in
>    malloc_type_allocated() as there will be nothing to do.

I think it would be quite useful to track malloc failures by type, not 
just successes; libmemstat(3) incudes a memstat_get_failures() call to 
retrieve failure statistics if they are supported by the underlying 
allocator.  Adding a per-cpu failure account should be straight forward; 
also, I wouldn't worry to much about the cost of the critical enter/exit 
in the failure case, as optimizing for failures in malloc(9) will have 
little impact on system performance (especially since the removal of 
administrative limits on malloc types).

Robert N M Watson



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20050728003334.V54330>