Re: Linux RDMA mini-conf at Plumbers 2018

From: Mike Rapoport
Date: Sun Sep 23 2018 - 11:17:54 EST


On Thu, Sep 20, 2018 at 09:54:28PM +0300, Leon Romanovsky wrote:
> On Thu, Sep 20, 2018 at 12:19:23PM -0600, Jason Gunthorpe wrote:
> > This is just a friendly reminder that registration deadlines are
> > approaching for this conference. Please see
> >
> > https://www.linuxplumbersconf.org/event/2/page/7-attend
> >
> > For details.
> >
> > This year we expect to have close to a day set aside for RDMA related
> > topics. Including up to half a day for the thorny general kernel issues
> > related to get_user_pages(), particularly as exasperated by RDMA.
> >
> > We have been working on the following concepts for sessions, I've
> > roughly marked names based on past participation in related email
> > threads. As we get closer to the conference date we will be organizing
> > leaders for each section based on these lists, please let us know of
> > any changes, or desire to be a leader!
> >
> > RDMA and get_user_pages
> > =======================
> > Dan Williams <dan.j.williams@xxxxxxxxx>
> > Matthew Wilcox <willy@xxxxxxxxxxxxx>
> > John Hubbard <jhubbard@xxxxxxxxxx>
> > Nicholas Piggin <npiggin@xxxxxxxxx>
> > Jan Kara <jack@xxxxxxx>
> >
> > RDMA, DAX and persistant memory co-existence.
> >
> > Explore the limits of what is possible without using On
> > Demand Paging Memory Registration. Discuss 'shootdown'
> > of userspace MRs
> >
> > Dirtying pages obtained with get_user_pages() can oops ext4
> > discuss open solutions.
> >
> > RDMA and PCI peer to peer
> > =========================
> > Don Dutile <ddutile@xxxxxxxxxx>
> > Alex Williamson <alex.williamson@xxxxxxxxxx>
> > Christoph Hellwig <hch@xxxxxx>
> > Stephen Bates <sbates@xxxxxxxxxxxx>
> > Logan Gunthorpe <logang@xxxxxxxxxxxx>
> > Jérôme Glisse <jglisse@xxxxxxxxxx>
> > Christian König <christian.koenig@xxxxxxx>
> > Bjorn Helgaas <bhelgaas@xxxxxxxxxx>
> >
> > RDMA and PCI peer to peer transactions. IOMMU issues. Integration
> > with HMM. How to expose PCI BAR memory to userspace and other
> > drivers as a DMA target.
> >
> > Improving testing of RDMA with syzkaller, RXE and Python
> > ========================================================
> > Noa Osherovich <noaos@xxxxxxxxxxxx>
> > Don Dutile <ddutile@xxxxxxxxxx>
> > Jason Gunthorpe <jgg@xxxxxxxxxxxx>
> >
> > Problem solve RDMA's distinct lack of public tests.
> > Provide a better framework for all drivers to test with,
> > and a framework for basic testing in userspace.
> >
> > Worst remaining unfixed syzkaller bugs and how to try to fix them
> >
> > How to hook syzkaller more deeply into RDMA.
> >
> > IOCTL conversion and new kABI topics
> > ====================================
> > Jason Gunthorpe <jgg@xxxxxxxxxxxx>
> > Alex Rosenbaum <alexr@xxxxxxxxxxxx>
> >
> > Attempt to close on the remaining tasks to complete the project
> >
> > Restore fork() support to userspace
> >
> > Container and namespaces for RDMA topics
> > ========================================
> > Parav Pandit <parav@xxxxxxxxxxxx>
> > Doug Ledford <dledford@xxxxxxxxxx>
> >
> > Remaining sticky situations with containers
> >
> > namespaces in sysfs and legacy all-namespace operation
> >
> > Remaining CM issues
> >
> > Security isolation problems
> >
> > Very large Contiguous regions in userspace
> > ==========================================
> > Christopher Lameter <cl@xxxxxxxxx>
> > Parav Pandit <parav@xxxxxxxxxxxx>
> >
> > Poor performance of get_user_pages on very large virtual ranges
> >
> > No standardized API to allocate regions to user space
> >
> > Carry over from last year
> >
> > As we get closer to the conference date the exact schedule will be
> > published on the conference web site. I belive we have the Thursday
> > set aside right now.
> >
> > If there are any last minute topics people would like to see please
> > let us know.
>
> I want to remind you that Mike wanted to bring the topic of enhancing
> remote page faults during post-copy container migration in CRIU over
> RDMA.

It's more Joel's topic, but thanks for the reminder anyway :)

> Thanks
>
> >
> > See you all in Vancouver!
> >
> > Thanks,
> > Jason & Leon
> >

--
Sincerely yours,
Mike.