linux/arch/powerpc/lib
Mark Nelson 57dda6ef5b powerpc: New copy_4K_page()
This new copy_4K_page() function was originally tuned for the best
performance on the Cell processor, but after testing on more 64bit
powerpc chips it was found that with a small modification it either
matched the performance offered by the current mainline version or
bettered it by a small amount.

It was found that on a Cell-based QS22 blade the amount of system
time measured when compiling a 2.6.26 pseries_defconfig decreased
by 4%. Using the same test, a 4-way 970MP machine saw a decrease of
2% in system time. No noticeable change was seen on Power4, Power5
or Power6.

The 4096 byte page is copied in thirty-two 128 byte strides. An
initial setup loop executes dcbt instructions for the whole source
page and dcbz instructions for the whole destination page. To do
this, the cache line size is retrieved from ppc64_caches.

A new CPU feature bit, CPU_FTR_CP_USE_DCBTZ, (introduced in the
previous patch) is used to make the modification to this new copy
routine - on Power4, 970 and Cell the feature bit is set so the
setup loop is executed, but on all other 64bit chips the setup
loop is nop'ed out.

Signed-off-by: Mark Nelson <markn@au1.ibm.com>
Signed-off-by: Paul Mackerras <paulus@samba.org>
2008-09-15 11:07:42 -07:00
..
alloc.c [POWERPC] Limit range of __init_ref_ok somewhat 2007-10-03 11:48:44 +10:00
checksum_32.S powerpc: Rename files to have consistent _32/_64 suffixes 2005-10-10 21:52:43 +10:00
checksum_64.S powerpc: Rename files to have consistent _32/_64 suffixes 2005-10-10 21:52:43 +10:00
code-patching.c PAGE_ALIGN(): correctly handle 64-bit values on 32-bit architectures 2008-07-24 10:47:21 -07:00
copy_32.S Remove obsolete #include <linux/config.h> 2006-06-30 19:25:36 +02:00
copypage_64.S powerpc: New copy_4K_page() 2008-09-15 11:07:42 -07:00
copyuser_64.S [POWERPC] Use mtocrf instruction in asm when CONFIG_POWER4_ONLY=y 2007-04-13 03:55:13 +10:00
crtsavres.S [POWERPC] Fix -Os kernel builds with newer gcc versions 2008-06-16 15:00:54 +10:00
devres.c [POWERPC] devres: Add devm_ioremap_prot() 2008-05-05 16:47:14 +10:00
div64.S powerpc: Fix a corner case in __div64_32 2005-10-20 09:37:02 +10:00
dma-noncoherent.c powerpc: Prevent memory corruption due to cache invalidation of unaligned DMA buffer 2008-07-01 11:28:54 +10:00
feature-fixups-test.S powerpc: Fixup lwsync at runtime 2008-07-03 16:58:10 +10:00
feature-fixups.c powerpc: Use WARN_ON(1) instead of __WARN() 2008-07-22 10:39:34 +10:00
locks.c [POWERPC] Make ppc64_defconfig without CONFIG_PPC_PSERIES build 2007-03-09 15:03:24 +11:00
Makefile powerpc: Remove use of CONFIG_PPC_MERGE 2008-08-04 13:18:17 +10:00
mem_64.S [POWERPC] Use mtocrf instruction in asm when CONFIG_POWER4_ONLY=y 2007-04-13 03:55:13 +10:00
memcpy_64.S [POWERPC] Use mtocrf instruction in asm when CONFIG_POWER4_ONLY=y 2007-04-13 03:55:13 +10:00
rheap.c [POWERPC] Make rheap safe for spinlocks 2008-04-17 09:50:38 -05:00
sstep.c [POWERPC] Emulate more instructions in software 2007-04-24 21:31:57 +10:00
string.S powerpc: Use PPC_LONG and PPC_LONG_ALIGN in lib/string.S 2008-07-22 10:39:35 +10:00
usercopy_64.c powerpc: Rename files to have consistent _32/_64 suffixes 2005-10-10 21:52:43 +10:00