Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:18 as permitted sender) client-ip=2620:137:e000::1:18;
Date:   Thu, 28 Apr 2022 17:06:44 +0200
From:   Christoph Hellwig <hch@lst.de>
To:     Thomas =?iso-8859-1?Q?Wei=DFschuh?= <linux@weissschuh.net>
Cc:     Christoph Hellwig <hch@lst.de>, Keith Busch <kbusch@kernel.org>,
        Jens Axboe <axboe@fb.com>, Sagi Grimberg <sagi@grimberg.me>,
        linux-kernel@vger.kernel.org, linux-nvme@lists.infradead.org
Subject: Re: [PATCH] nvme-pci: fix host memory buffer allocation size
Message-ID: <20220428150644.GA22685@lst.de>
References: <20220428101922.14216-1-linux@weissschuh.net> <20220428143603.GA20460@lst.de> <5060d75e-46c0-4d29-a334-62c7e9714fa7@t-8ch.de>
MIME-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
In-Reply-To: <5060d75e-46c0-4d29-a334-62c7e9714fa7@t-8ch.de>
User-Agent: Mutt/1.5.17 (2007-11-01)
Precedence: bulk

On Thu, Apr 28, 2022 at 04:44:47PM +0200, Thomas Wei?schuh wrote:
> Is the current code supposed to reach HMPRE? It does not for me.
> 
> The code tries to allocate memory for HMPRE in chunks.
> The best allocation would be to allocate one chunk for all of HMPRE.
> If this fails we half the chunk size on each iteration and try again.
> 
> On my hardware we start with a chunk_size of 4MiB and just allocate
> 8 (hmmaxd) * 4 = 32 MiB which is worse than 1 * 200MiB.

And that is because the hardware only has a limited set of descriptors.

Is there any real problem you are fixing with this?  Do you actually
see a performance difference on a relevant workload?