xavi/android_kernel_m2note/arch/arm64/kernel/io.c, branch o-8.1

FROMLIST: arm64: Avoid aligning normal memory pointers in __memcpy_{to,from}io

2017-12-14T16:04:49+00:00

(cherry picked from arm64/for-next/core commit 9ca255bf041ddc7698b6906dbd846c0ba64b1fe1) __memcpy_{to,from}io fall back to byte-at-a-time copying if both the source and destination pointers are not 8-byte aligned. Since one of the pointers always points at normal memory, this is unnecessary and detrimental to performance, so only do byte copying until we hit an 8-byte boundary for the device pointer. This change was motivated by performance issues in the pstore driver. On a test platform, measuring probe time for pstore, console buffer size of 1/4MB and pmsg of 1/2MB, was in the 90-107ms region. Change managed to reduce it to 10-25ms, an improvement in boot time. Cc: Kees Cook Cc: Anton Vorontsov Cc: Tony Luck Cc: Catalin Marinas Cc: Will Deacon Cc: Anton Vorontsov Cc: Robin Murphy Signed-off-by: Mark Salyzyn Signed-off-by: Will Deacon Bug: 63716230 Change-Id: I245545e8243a54b44d30fbb0d0c71a9b8a77ef63

UPSTREAM: arm64: optimize memcpy_{from,to}io() and memset_io()

2017-12-14T16:04:48+00:00

(cherry picked from commit 70ddb63a88bfd45eb6abe36e2bf4f8f351a447d7) Optimize memcpy_{from,to}io() and memset_io() by transferring in 64 bit as much as possible with minimized barrier usage. This simplest optimization brings faster throughput compare to current byte-by-byte read and write with barrier in the loop. Code's skeleton is taken from the powerpc. Link: http://lkml.kernel.org/p/20141020133304.GH23751@e104818-lin.cambridge.arm.com Reviewed-by: Catalin Marinas Reviewed-by: Trilok Soni Signed-off-by: Joonwoo Park Signed-off-by: Will Deacon Bug: 63716230 Change-Id: I27a3ecfcdb8d1e80ceac5eb5f71609283e0c901b

first commit

2016-08-15T02:19:42+00:00