diff options
author | Will Deacon <will.deacon@arm.com> | 2016-02-02 12:46:25 +0000 |
---|---|---|
committer | Jeff Vander Stoep <jeffv@google.com> | 2016-09-22 13:38:22 -0700 |
commit | 297bf83fa4e7804c28c6d950f39dbd3860552cb9 (patch) | |
tree | c6c026fa6f112973d0ca708c3b3b4276dceeab21 /.gitignore | |
parent | b5924f376a221ffa0ee64a4410aff54416c2fd2d (diff) |
UPSTREAM: arm64: lib: improve copy_page to deal with 128 bytes at a time
We want to avoid lots of different copy_page implementations, settling
for something that is "good enough" everywhere and hopefully easy to
understand and maintain whilst we're at it.
This patch reworks our copy_page implementation based on discussions
with Cavium on the list and benchmarking on Cortex-A processors so that:
- The loop is unrolled to copy 128 bytes per iteration
- The reads are offset so that we read from the next 128-byte block
in the same iteration that we store the previous block
- Explicit prefetch instructions are removed for now, since they hurt
performance on CPUs with hardware prefetching
- The loop exit condition is calculated at the start of the loop
Signed-off-by: Will Deacon <will.deacon@arm.com>
Tested-by: Andrew Pinski <apinski@cavium.com>
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Bug: 30369029
Patchset: kaslr-arm64-4.4
(cherry picked from commit 223e23e8aa26b0bb62c597637e77295e14f6a62c)
Signed-off-by: Jeff Vander Stoep <jeffv@google.com>
Change-Id: Icabd86bbecc60ad0d730ab796e33b8762cecb1fb
Diffstat (limited to '.gitignore')
0 files changed, 0 insertions, 0 deletions