Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]


Groups > linux.kernel > #1616242

Re: [RFC 6/8] nvmet: Be careful about using iomem accesses when dealing with p2pmem

From Logan Gunthorpe <logang@deltatee.com>
Newsgroups linux.kernel
Subject Re: [RFC 6/8] nvmet: Be careful about using iomem accesses when dealing with p2pmem
Date 2017-04-04 19:30 +0200
Message-ID <tsArg-7Vm-33@gated-at.bofh.it> (permalink)
References <tqQA9-4sB-3@gated-at.bofh.it> <tqQAa-4sB-5@gated-at.bofh.it> <tsuvv-43D-17@gated-at.bofh.it> <tsySu-6KL-25@gated-at.bofh.it>
Organization linux.* mail to news gateway

Show all headers | View raw



On 04/04/17 04:59 AM, Sagi Grimberg wrote:
> We can never ever get here from an IO command, and that is a good thing
> because it would have been broken if we did, regardless of what copy
> method we use...

Yes, I changed this mostly for admin commands. I did notice connect
commands do end up reading from the p2mem and this patchset correctly
switches it to iomemcpy. However, based on Cristoph's comment, I hope to
make it more general such that iomem is hidden within sgls and any
access will either be correct or create a warning.


On 04/04/17 09:46 AM, Jason Gunthorpe wrote:
> Transactions might not complete at the NVMe device before the CPU
> processes the RDMA completion, however due to the PCI-E ordering rules
> new TLPs directed to the NVMe will complete after the RMDA TLPs and
> thus observe the new data. (eg order preserving)
> 
> It would be very hard to use P2P if fabric ordering is not preserved..

Yes, my understanding is the same, the PCI-E ordering rules save us here.

Thanks,

Logan

Back to linux.kernel | Previous | NextPrevious in thread | Next in thread | Find similar | Unroll thread


Thread

[RFC 6/8] nvmet: Be careful about using iomem accesses when dealing with p2pmem Logan Gunthorpe <logang@deltatee.com> - 2017-03-31 00:20 +0200
  Re: [RFC 6/8] nvmet: Be careful about using iomem accesses when  dealing with p2pmem Sagi Grimberg <sagi@grimberg.me> - 2017-04-04 13:10 +0200
    Re: [RFC 6/8] nvmet: Be careful about using iomem accesses when  dealing with p2pmem Jason Gunthorpe <jgunthorpe@obsidianresearch.com> - 2017-04-04 17:50 +0200
      Re: [RFC 6/8] nvmet: Be careful about using iomem accesses when  dealing with p2pmem Logan Gunthorpe <logang@deltatee.com> - 2017-04-04 19:30 +0200
      Re: [RFC 6/8] nvmet: Be careful about using iomem accesses when  dealing with p2pmem Sagi Grimberg <sagi@grimberg.me> - 2017-04-06 07:40 +0200
        Re: [RFC 6/8] nvmet: Be careful about using iomem accesses when  dealing with p2pmem Logan Gunthorpe <logang@deltatee.com> - 2017-04-06 18:10 +0200
        Re: [RFC 6/8] nvmet: Be careful about using iomem accesses when  dealing with p2pmem Jason Gunthorpe <jgunthorpe@obsidianresearch.com> - 2017-04-06 18:40 +0200
        Re: [RFC 6/8] nvmet: Be careful about using iomem accesses when  dealing with p2pmem "Stephen  Bates" <sbates@raithlin.com> - 2017-04-07 13:30 +0200
          Re: [RFC 6/8] nvmet: Be careful about using iomem accesses when  dealing with p2pmem Sagi Grimberg <sagi@grimberg.me> - 2017-04-10 10:40 +0200

csiph-web