linux-kernel - Re: [PATCH v2 2/2] SLUB: Mark merged slab caches in /proc/slabinfo

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Date:	Thu, 16 Sep 2010 15:08:23 -0700
From:	Tony Luck <tony.luck@...el.com>
To:	Linus Torvalds <torvalds@...ux-foundation.org>
Cc:	Christoph Lameter <cl@...ux.com>,
	David Rientjes <rientjes@...gle.com>,
	"Ted Ts'o" <tytso@....edu>, Pekka Enberg <penberg@...nel.org>,
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH v2 2/2] SLUB: Mark merged slab caches in /proc/slabinfo

On Thu, Sep 16, 2010 at 10:49 AM, Linus Torvalds
<torvalds@...ux-foundation.org> wrote:
> That would allow slabtop etc to show the number of actual allocations
> at the granularity that we want (ie you'd see who is the real user),
> but still then use the same slab pages and avoid the slab
> fragmentation.

It might be hard to interpret the data. Suppose "foo" and "bar"
are slab caches that have separate accounting, but a merged
allocator. Now we do 20 allocations from foo, and free 19 of them.
Then we do 19 allocations from bar.  The accounts would show
that foo has 1 object in use and 19 free. While bar would show
19 objects in use.

But the 19 "free" foo objects aren't free at all, they've been
handed out to bar.

-Tony
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/