From mboxrd@z Thu Jan  1 00:00:00 1970
Received: from eggs.gnu.org ([2001:4830:134:3::10]:33736)
	by lists.gnu.org with esmtp (Exim 4.71)
	(envelope-from <cota@braap.org>) id 1b8lFd-00016B-NF
	for qemu-devel@nongnu.org; Fri, 03 Jun 2016 05:18:46 -0400
Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71)
	(envelope-from <cota@braap.org>) id 1b8lFZ-0008QX-CJ
	for qemu-devel@nongnu.org; Fri, 03 Jun 2016 05:18:44 -0400
Received: from out2-smtp.messagingengine.com ([66.111.4.26]:42330)
	by eggs.gnu.org with esmtp (Exim 4.71)
	(envelope-from <cota@braap.org>) id 1b8lFX-0008N6-0O
	for qemu-devel@nongnu.org; Fri, 03 Jun 2016 05:18:41 -0400
Date: Fri, 3 Jun 2016 05:18:28 -0400
From: "Emilio G. Cota" <cota@braap.org>
Message-ID: <20160603091828.GA5251@flamenco>
References: <1464138802-23503-1-git-send-email-cota@braap.org>
	<1464138802-23503-11-git-send-email-cota@braap.org>
	<574B487B.5080200@gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <574B487B.5080200@gmail.com>
Subject: Re: [Qemu-devel] [PATCH v6 10/15] qht: QEMU's fast,
 resizable and scalable Hash Table
List-Id: <qemu-devel.nongnu.org>
List-Unsubscribe: <https://lists.nongnu.org/mailman/options/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=unsubscribe>
List-Archive: <http://lists.nongnu.org/archive/html/qemu-devel/>
List-Post: <mailto:qemu-devel@nongnu.org>
List-Help: <mailto:qemu-devel-request@nongnu.org?subject=help>
List-Subscribe: <https://lists.nongnu.org/mailman/listinfo/qemu-devel>,
	<mailto:qemu-devel-request@nongnu.org?subject=subscribe>
To: Sergey Fedorov <serge.fdrv@gmail.com>
Cc: QEMU Developers <qemu-devel@nongnu.org>, MTTCG Devel <mttcg@listserver.greensocs.com>, Alex =?iso-8859-1?Q?Benn=E9e?= <alex.bennee@linaro.org>, Paolo Bonzini <pbonzini@redhat.com>, Richard Henderson <rth@twiddle.net>

On Sun, May 29, 2016 at 22:52:27 +0300, Sergey Fedorov wrote:
> I was just wondering if it could be worthwhile to pass a hash function
> when initializing a QHT. Then we could have variants of qht_insert(),
> qht_remove() and qht_lookup() which does not require a computed hash
> value but call the function by themselves. This could make sense since a
> hash value passed the the functions should always be exactly the same
> for the same object.

I considered this when designing the API. I think it's not worth having
in qht; callers could have their own wrapper to do this though.

For the only caller of qht that we have so far I don't see this
as being worth the hassle.

For instance, we couldn't use the same function for lookups and
inserts/removals, since the hash function would look like:

uint32_t hash_func(void *p)
{
	TranslationBlock *tb = p;
	return tb_hash_func(tb->phys_pc, ...);
}

But for lookups we don't yet know *tb (that's what we're looking for!).
All we have is the tb_desc struct that we use for comparisons.

Thanks,

		Emilio