From: John Hubbard <jhubbard@nvidia.com>
To: Logan Gunthorpe <logang@deltatee.com>,
<linux-kernel@vger.kernel.org>, <linux-nvme@lists.infradead.org>,
<linux-block@vger.kernel.org>, <linux-pci@vger.kernel.org>,
<linux-mm@kvack.org>, <iommu@lists.linux-foundation.org>
Cc: "Stephen Bates" <sbates@raithlin.com>,
"Christoph Hellwig" <hch@lst.de>,
"Dan Williams" <dan.j.williams@intel.com>,
"Jason Gunthorpe" <jgg@ziepe.ca>,
"Christian König" <christian.koenig@amd.com>,
"Don Dutile" <ddutile@redhat.com>,
"Matthew Wilcox" <willy@infradead.org>,
"Daniel Vetter" <daniel.vetter@ffwll.ch>,
"Jakowski Andrzej" <andrzej.jakowski@intel.com>,
"Minturn Dave B" <dave.b.minturn@intel.com>,
"Jason Ekstrand" <jason@jlekstrand.net>,
"Dave Hansen" <dave.hansen@linux.intel.com>,
"Xiong Jianxin" <jianxin.xiong@intel.com>,
"Bjorn Helgaas" <helgaas@kernel.org>,
"Ira Weiny" <ira.weiny@intel.com>,
"Robin Murphy" <robin.murphy@arm.com>
Subject: Re: [PATCH 09/16] dma-direct: Support PCI P2PDMA pages in dma-direct map_sg
Date: Mon, 3 May 2021 17:01:29 -0700 [thread overview]
Message-ID: <2a08b685-53c1-3ee4-0ce1-315554d63685@nvidia.com> (raw)
In-Reply-To: <8de928ab-2842-dac9-07ad-a098124f791f@deltatee.com>
On 5/3/21 10:04 AM, Logan Gunthorpe wrote:
> Oops missed a comment:
>
> On 2021-05-02 5:28 p.m., John Hubbard wrote:
>>> int dma_direct_map_sg(struct device *dev, struct scatterlist *sgl, int nents,
>>> enum dma_data_direction dir, unsigned long attrs)
>>> {
>>> - int i;
>>> + struct pci_p2pdma_map_state p2pdma_state = {};
>>
>> Is it worth putting this stuff on the stack--is there a noticeable
>> performance improvement from caching the state? Because if it's
>> invisible, then simplicity is better. I suspect you're right, and that
>> it *is* worth it, but it's good to know for real.
>
> I haven't measured it (it would be hard to measure), but I think it's
> fairly clear here. Without the state, xa_load() would need to be called
> on *every* page in an SGL that maps only P2PDMA memory from one device.
> With the state, it only needs to be called once. xa_load() is cheap, but
> it is not that cheap.
OK, thanks for spelling it out for me. :)
>
> There's essentially the same optimization in get_user_pages for
> ZONE_DEVICE pages. So, if it is necessary there, it should be necessary
> here.
>
Right, that's a pretty solid example.
thanks,
--
John Hubbard
NVIDIA
next prev parent reply other threads:[~2021-05-04 0:01 UTC|newest]
Thread overview: 99+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-04-08 17:01 [PATCH 00/16] Add new DMA mapping operation for P2PDMA Logan Gunthorpe
2021-04-08 17:01 ` [PATCH 01/16] PCI/P2PDMA: Pass gfp_mask flags to upstream_bridge_distance_warn() Logan Gunthorpe
2021-05-02 3:58 ` John Hubbard
2021-05-03 15:57 ` Logan Gunthorpe
2021-05-03 18:17 ` John Hubbard
2021-05-03 18:20 ` Logan Gunthorpe
2021-05-03 18:23 ` John Hubbard
2021-05-03 18:24 ` Christoph Hellwig
2021-05-11 16:05 ` Don Dutile
2021-05-11 16:12 ` Logan Gunthorpe
2021-05-11 16:23 ` Don Dutile
2021-04-08 17:01 ` [PATCH 02/16] PCI/P2PDMA: Avoid pci_get_slot() which sleeps Logan Gunthorpe
2021-05-02 5:35 ` John Hubbard
2021-05-03 16:08 ` Logan Gunthorpe
2021-05-03 18:20 ` John Hubbard
2021-05-03 18:25 ` Christoph Hellwig
2021-05-11 16:05 ` Don Dutile
2021-05-11 16:16 ` Logan Gunthorpe
2021-05-11 16:05 ` Don Dutile
2021-05-11 16:14 ` Logan Gunthorpe
2021-04-08 17:01 ` [PATCH 03/16] PCI/P2PDMA: Attempt to set map_type if it has not been set Logan Gunthorpe
2021-05-02 19:58 ` John Hubbard
2021-05-03 16:17 ` Logan Gunthorpe
2021-05-03 18:22 ` John Hubbard
2021-05-03 18:35 ` Christoph Hellwig
2021-05-03 18:46 ` Logan Gunthorpe
2021-05-11 16:05 ` Don Dutile
2021-04-08 17:01 ` [PATCH 04/16] PCI/P2PDMA: Refactor pci_p2pdma_map_type() to take pagmap and device Logan Gunthorpe
2021-05-02 20:41 ` John Hubbard
2021-05-03 16:30 ` Logan Gunthorpe
2021-05-03 18:31 ` John Hubbard
2021-05-03 18:56 ` Logan Gunthorpe
2021-05-03 21:54 ` John Hubbard
2021-05-03 22:57 ` Jason Gunthorpe
2021-05-03 23:40 ` John Hubbard
2021-04-08 17:01 ` [PATCH 05/16] dma-mapping: Introduce dma_map_sg_p2pdma() Logan Gunthorpe
2021-04-27 19:22 ` Jason Gunthorpe
2021-04-27 22:49 ` Logan Gunthorpe
2021-04-27 19:31 ` Jason Gunthorpe
2021-04-27 22:55 ` Logan Gunthorpe
2021-04-27 23:01 ` Jason Gunthorpe
2021-05-03 18:28 ` Christoph Hellwig
2021-05-03 18:31 ` Logan Gunthorpe
2021-05-02 21:23 ` John Hubbard
2021-05-03 16:38 ` Logan Gunthorpe
2021-05-11 16:05 ` Don Dutile
2021-04-08 17:01 ` [PATCH 06/16] lib/scatterlist: Add flag for indicating P2PDMA segments in an SGL Logan Gunthorpe
2021-05-02 22:34 ` John Hubbard
2021-04-08 17:01 ` [PATCH 07/16] PCI/P2PDMA: Make pci_p2pdma_map_type() non-static Logan Gunthorpe
2021-05-02 22:44 ` John Hubbard
2021-05-03 16:39 ` Logan Gunthorpe
2021-05-11 16:06 ` Don Dutile
2021-05-11 16:17 ` Logan Gunthorpe
2021-04-08 17:01 ` [PATCH 08/16] PCI/P2PDMA: Introduce helpers for dma_map_sg implementations Logan Gunthorpe
2021-05-02 22:52 ` John Hubbard
2021-05-03 0:50 ` John Hubbard
2021-05-03 17:15 ` Logan Gunthorpe
2021-04-08 17:01 ` [PATCH 09/16] dma-direct: Support PCI P2PDMA pages in dma-direct map_sg Logan Gunthorpe
2021-04-27 19:33 ` Jason Gunthorpe
2021-04-27 19:40 ` Jason Gunthorpe
2021-04-27 22:56 ` Logan Gunthorpe
2021-05-02 23:28 ` John Hubbard
2021-05-02 23:32 ` John Hubbard
2021-05-03 17:06 ` Logan Gunthorpe
2021-05-03 16:55 ` Logan Gunthorpe
2021-05-04 0:12 ` John Hubbard
2021-05-03 17:04 ` Logan Gunthorpe
2021-05-04 0:01 ` John Hubbard [this message]
2021-04-08 17:01 ` [PATCH 10/16] dma-mapping: Add flags to dma_map_ops to indicate PCI P2PDMA support Logan Gunthorpe
2021-05-03 0:32 ` John Hubbard
2021-05-03 17:09 ` Logan Gunthorpe
2021-05-11 16:06 ` Don Dutile
2021-05-11 16:19 ` Logan Gunthorpe
2021-04-08 17:01 ` [PATCH 11/16] iommu/dma: Support PCI P2PDMA pages in dma-iommu map_sg Logan Gunthorpe
2021-04-27 19:43 ` Jason Gunthorpe
2021-04-27 22:59 ` Logan Gunthorpe
2021-05-03 1:14 ` John Hubbard
2021-05-06 23:59 ` Logan Gunthorpe
2021-05-11 16:06 ` Don Dutile
2021-05-11 16:35 ` Logan Gunthorpe
2021-04-08 17:01 ` [PATCH 12/16] nvme-pci: Check DMA ops when indicating support for PCI P2PDMA Logan Gunthorpe
2021-05-03 1:29 ` John Hubbard
2021-05-03 17:17 ` Logan Gunthorpe
2021-05-04 0:17 ` John Hubbard
2021-04-08 17:01 ` [PATCH 13/16] nvme-pci: Convert to using dma_map_sg_p2pdma for p2pdma pages Logan Gunthorpe
2021-05-03 1:34 ` John Hubbard
2021-05-03 17:19 ` Logan Gunthorpe
2021-05-04 0:26 ` John Hubbard
2021-04-08 17:01 ` [PATCH 14/16] nvme-rdma: Ensure dma support when using p2pdma Logan Gunthorpe
2021-04-27 19:47 ` Jason Gunthorpe
2021-04-27 22:59 ` Logan Gunthorpe
2021-05-03 1:37 ` John Hubbard
2021-04-08 17:01 ` [PATCH 15/16] RDMA/rw: use dma_map_sg_p2pdma() Logan Gunthorpe
2021-04-08 17:01 ` [PATCH 16/16] PCI/P2PDMA: Remove pci_p2pdma_[un]map_sg() Logan Gunthorpe
2021-04-27 19:28 ` [PATCH 00/16] Add new DMA mapping operation for P2PDMA Jason Gunthorpe
2021-04-27 20:21 ` John Hubbard
2021-04-27 20:48 ` Dan Williams
2021-05-02 1:22 ` John Hubbard
2021-05-11 16:05 ` Don Dutile
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=2a08b685-53c1-3ee4-0ce1-315554d63685@nvidia.com \
--to=jhubbard@nvidia.com \
--cc=andrzej.jakowski@intel.com \
--cc=christian.koenig@amd.com \
--cc=dan.j.williams@intel.com \
--cc=daniel.vetter@ffwll.ch \
--cc=dave.b.minturn@intel.com \
--cc=dave.hansen@linux.intel.com \
--cc=ddutile@redhat.com \
--cc=hch@lst.de \
--cc=helgaas@kernel.org \
--cc=iommu@lists.linux-foundation.org \
--cc=ira.weiny@intel.com \
--cc=jason@jlekstrand.net \
--cc=jgg@ziepe.ca \
--cc=jianxin.xiong@intel.com \
--cc=linux-block@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=linux-nvme@lists.infradead.org \
--cc=linux-pci@vger.kernel.org \
--cc=logang@deltatee.com \
--cc=robin.murphy@arm.com \
--cc=sbates@raithlin.com \
--cc=willy@infradead.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).