> On Apr 5, 2021, at 4:07 PM, Jason Gunthorpe <jgg@xxxxxxxxxx> wrote: > > On Mon, Apr 05, 2021 at 03:41:15PM +0200, Christoph Hellwig wrote: >> On Mon, Apr 05, 2021 at 08:23:54AM +0300, Leon Romanovsky wrote: >>> From: Leon Romanovsky <leonro@xxxxxxxxxx> >>> >>>> From Avihai, >>> >>> Relaxed Ordering is a PCIe mechanism that relaxes the strict ordering >>> imposed on PCI transactions, and thus, can improve performance. >>> >>> Until now, relaxed ordering could be set only by user space applications >>> for user MRs. The following patch series enables relaxed ordering for the >>> kernel ULPs as well. Relaxed ordering is an optional capability, and as >>> such, it is ignored by vendors that don't support it. >>> >>> The following test results show the performance improvement achieved >>> with relaxed ordering. The test was performed on a NVIDIA A100 in order >>> to check performance of storage infrastructure over xprtrdma: >> >> Isn't the Nvidia A100 a GPU not actually supported by Linux at all? >> What does that have to do with storage protocols? > > I think it is a typo (or at least mit makes no sense to be talking > about NFS with a GPU chip) Probably it should be a DGX A100 which is a > dual socket AMD server with alot of PCIe, and xptrtrdma is a NFS-RDMA > workload. We need to get a better idea what correctness testing has been done, and whether positive correctness testing results can be replicated on a variety of platforms. I have an old Haswell dual-socket system in my lab, but otherwise I'm not sure I have a platform that would be interesting for such a test. > AMD dual socket systems are well known to benefit from relaxed > ordering, people have been doing this in userspace for a while now > with the opt in. -- Chuck Lever
