linux-kernel - Re: [PATCH v6 03/13] PCI/P2PDMA: Add PCI p2pmem DMA mappings to adjust the bus offset

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

Date:   Fri, 21 Sep 2018 11:48:13 -0500
From:   Bjorn Helgaas <helgaas@...nel.org>
To:     Logan Gunthorpe <logang@...tatee.com>
Cc:     Jens Axboe <axboe@...nel.dk>, Keith Busch <keith.busch@...el.com>,
        Alex Williamson <alex.williamson@...hat.com>,
        Sagi Grimberg <sagi@...mberg.me>, linux-nvdimm@...ts.01.org,
        linux-rdma@...r.kernel.org, linux-pci@...r.kernel.org,
        linux-kernel@...r.kernel.org, linux-nvme@...ts.infradead.org,
        Stephen Bates <sbates@...thlin.com>,
        linux-block@...r.kernel.org,
        Jérôme Glisse <jglisse@...hat.com>,
        Jason Gunthorpe <jgg@...lanox.com>,
        Christian König <christian.koenig@....com>,
        Benjamin Herrenschmidt <benh@...nel.crashing.org>,
        Bjorn Helgaas <bhelgaas@...gle.com>,
        Max Gurtovoy <maxg@...lanox.com>,
        Dan Williams <dan.j.williams@...el.com>,
        Christoph Hellwig <hch@....de>
Subject: Re: [PATCH v6 03/13] PCI/P2PDMA: Add PCI p2pmem DMA mappings to
 adjust the bus offset

On Fri, Sep 21, 2018 at 08:15:50AM -0500, Bjorn Helgaas wrote:
> On Wed, Sep 12, 2018 at 06:11:46PM -0600, Logan Gunthorpe wrote:
> > The DMA address used when mapping PCI P2P memory must be the PCI bus
> > address. Thus, introduce pci_p2pmem_map_sg() to map the correct
> > addresses when using P2P memory. Memory mapped in this way does not
> > need to be unmapped.
> 
> I think the use of "map" in this context is slightly confusing because the
> general expectation is that map/unmap must be balanced.
> 
> I assume it's because the "mapping" consumes no resources, e.g., requires
> no page table entries.  Possibly there's a better verb than "map", e.g.,
> "convert", "convert_to_p2pdma", etc?
> 
> If you keep "map", maybe add a sentence or two about why there's no
> corresponding unmap?

Another wrinkle is that "map" usually takes an A and gives you back a
B.  Now the caller has both A and B and both are still valid.

Here we pass in an SGL and the SGL is transformed, so the caller only
has B and A has been destroyed, i.e., the SGL can no longer be used as
it was before, and there's no way to get A back.

Maybe this is pointless bikeshedding, so if your color is the best,
don't change anything.

Bjorn