[RFC] avoid indirect calls for DMA direct mappings

From: Christoph Hellwig <hch@lst.de>
To: iommu@lists.linux-foundation.org,
	Linus Torvalds <torvalds@linux-foundation.org>,
	Jesper Dangaard Brouer <brouer@redhat.com>
Cc: "Tariq Toukan" <tariqt@mellanox.com>,
	"Ilias Apalodimas" <ilias.apalodimas@linaro.org>,
	"Toke Høiland-Jørgensen" <toke@toke.dk>,
	"Robin Murphy" <robin.murphy@arm.com>,
	linux-kernel@vger.kernel.org
Subject: [RFC] avoid indirect calls for DMA direct mappings
Date: Thu,  6 Dec 2018 07:37:19 -0800	[thread overview]
Message-ID: <20181206153720.10702-1-hch@lst.de> (raw)

Hi all,

a while ago Jesper reported major performance regressions due to the
spectre v2 mitigations in his XDP forwarding workloads.  A large part
of that is due to the DMA mapping API indirect calls.

It turns out that the most common implementation of the DMA API is the
direct mapping case, and now that we have merged almost all duplicate
implementations of that into a single generic one is easily feasily to
direct calls for this fast path.

This patch adds a check if we are using dma_direct_ops in each fast path
DMA operation, and just uses a direct call instead.  For the XDP workload
this increases the number of packets per second from 7,438,283 to
9,610,088, so it provides a very significant speedup.

Note that the patch depends on a lot of work either queued up in the
DMA mapping tree, or still out on the list from review, so to actually
try the patch you probably want this git tree:

    git://git.infradead.org/users/hch/misc.git dma-direct-calls

Gitweb:

    http://git.infradead.org/users/hch/misc.git/shortlog/refs/heads/dma-direct-calls