cfi info for stack adjust needs to be on the insn doing the adjust.
cfi describing register saves can be anywhere after the save insn but
before the reg is altered. Fewer locations with cfi result in smaller
cfi programs and possibly slightly faster exception handling. Thus
the LR cfi_offset move.
The idea behind ajusting sp after restoring regs is to break a
register dependency chain, in this case not be using r1 immediately
after it is modified.
The missing LR cfi_restore meant that code after the blr,
unaligned_lt_16 and other labels, would have cfi that said LR was at
cfa+16, but that code is reached without LR being saved.
* sysdeps/powerpc/powerpc64/power8/strncpy.S: Move LR cfi.
Adjust stack after restoring regs. Add missing LR cfi_restore.
Reviewed-by: Tulio Magno Quites Machado Filho <tuliom@linux.vnet.ibm.com>
Recent commit 59ba2d2b5421 missed to add __memrchr_power8 in
ifunc list. Also handled discarding unwanted bytes for
unaligned inputs in power8 optimization.
2017-10-05 Rajalakshmi Srinivasaraghavan <raji@linux.vnet.ibm.com>
* sysdeps/powerpc/powerpc64/multiarch/memrchr-ppc64.c: Revert
back to powerpc32 file.
* sysdeps/powerpc/powerpc64/multiarch/memrchr.c
(memrchr): Add __memrchr_power8 to ifunc list.
* sysdeps/powerpc/powerpc64/power8/memrchr.S: Mask
extra bytes for unaligned inputs.
The new generic expf and exp2f code don't need wrappers any more, they
set errno inline, so only use the wrappers on targets that need it.
(If the wrapper is needed, then the top level wrapper code is included,
otherwise empty w_exp*f.c is used to suppress the wrapper.)
A powerpc64 expf implementation includes the expf c code directly which
needed some changes.
* sysdeps/ieee754/flt-32/e_exp2f.c (__exp2f): Define without wrapper.
* sysdeps/ieee754/flt-32/e_expf.c (__expf): Likewise
* sysdeps/ieee754/flt-32/w_exp2f.c: New file.
* sysdeps/ieee754/flt-32/w_expf.c: New file.
* sysdeps/powerpc/powerpc64/fpu/multiarch/e_expf-ppc64.c: Update for
the new expf code.
* sysdeps/powerpc/powerpc64/fpu/multiarch/w_expf.c: New file.
* sysdeps/powerpc/powerpc64/power8/fpu/w_expf.c: New file.
* sysdeps/m68k/m680x0/fpu/w_exp2f.c: New file.
* sysdeps/m68k/m680x0/fpu/w_expf.c: New file.
* sysdeps/i386/fpu/w_exp2f.c: New file.
* sysdeps/i386/fpu/w_expf.c: New file.
* sysdeps/i386/i686/fpu/multiarch/w_expf.c: New file.
* sysdeps/x86_64/fpu/w_expf.c: New file.
Vectorized loops are used for sizes greater than 32B to improve
performance over power7 optimization. This shows as an average
of 25% improvement depending on the position of search
character. The performance is same for shorter strings.
As per the section "3.1.4.2 Alignment Interrupts" of the "POWER8 Processor
User's Manual for the Single-Chip Module", alignment interrupt is reported
for misaligned stores in Caching-inhibited storage. As memset is used in
some drivers for DMA (like xorg), this patch avoids misaligned stores for
sizes less than 8 in memset.
Makes __stpncpy_power8 call __memset_power8 directly rather than via an
IFUNC. Fixes a missing _mcount, and removes some redundant NOPS. The
*_is_local defines are also used in a followup patch.
* sysdeps/powerpc/powerpc64/multiarch/strncpy-power7.S: Define
MEMSET_is_local.
* sysdeps/powerpc/powerpc64/multiarch/strncpy-power8.S: Likewise.
* sysdeps/powerpc/powerpc64/multiarch/stpncpy-power7.S: Likewise.
* sysdeps/powerpc/powerpc64/multiarch/stpncpy-power8.S: Likewise.
Define MEMSET.
* sysdeps/powerpc/powerpc64/multiarch/strstr-power7.S: Define
STRLEN_is_local, STRNLEN_is_local, and STRCHR_is_local.
* sysdeps/powerpc/powerpc64/power7/strstr.S: Likewise. Don't add
nop after local calls.
* sysdeps/powerpc/powerpc64/power7/strncpy.S: Define MEMSET_is_local.
Don't add nop after local call.
* sysdeps/powerpc/powerpc64/power8/strncpy.S: Likewise. Add missing
CALL_MCOUNT.
This implementation is based on the one already used at
sysdeps/powerpc/powerpc64/fpu/multiarch/s_sinf-power8.S.
* sysdeps/powerpc/powerpc64/fpu/multiarch/Makefile
[$(subdir) = math] (libm-sysdep_routines): Add s_cosf-power8 and
s_cosf-ppc64.
* sysdeps/powerpc/powerpc64/fpu/multiarch/s_cosf-power8.S: New file.
* sysdeps/powerpc/powerpc64/fpu/multiarch/s_cosf-ppc64.c: Likewise.
* sysdeps/powerpc/powerpc64/fpu/multiarch/s_cosf.c: Likewise.
* sysdeps/powerpc/powerpc64/power8/fpu/s_cosf.S: Likewise.
P7 code is used for <=32B strings and for > 32B vectorized loops are used.
This shows as an average 25% improvement depending on the position of search
character. The performance is same for shorter strings.
Tested on ppc64 and ppc64le.
Clean up the IFUNC implementations for powerpc in order to remove
unneeded macro definitions.
Tested on ppc64le with and without --disable-multi-arch flag.
* sysdeps/powerpc/powerpc64/multiarch/memset-power4.S: Define the
implementation-specific function name and remove unneeded macros
definition.
* sysdeps/powerpc/powerpc64/multiarch/memset-power6.S: Likewise.
* sysdeps/powerpc/powerpc64/multiarch/memset-power7.S: Likewise.
* sysdeps/powerpc/powerpc64/multiarch/memset-power8.S: Likewise.
* sysdeps/powerpc/powerpc64/multiarch/memset-ppc64.S: Likewise.
* sysdeps/powerpc/powerpc64/memset.S: Set a default function name if
not defined and pass as parameter to macros accordingly.
* sysdeps/powerpc/powerpc64/power4/memset.S: Likewise.
* sysdeps/powerpc/powerpc64/power6/memset.S: Likewise.
* sysdeps/powerpc/powerpc64/power7/memset.S: Likewise.
* sysdeps/powerpc/powerpc64/power8/memset.S: Likewise.
Clean up the IFUNC implementations for powerpc in order to remove
unneeded macro definitions.
Tested on ppc64le with and without --disable-multi-arch flag.
* sysdeps/powerpc/powerpc64/multiarch/strcasestr-power8.S: Define the
strcasestr implementation name and remove unneeded macros definition.
* sysdeps/powerpc/powerpc64/multiarch/strstr-power7.S: Define
strstr implementation name and remove unneeded macros definition.
* sysdeps/powerpc/powerpc64/power7/strstr.S: Set a default function
name if not defined and pass as parameter to macros accordingly.
* sysdeps/powerpc/powerpc64/power8/strcasestr.S: Likewise.
Clean up the IFUNC implementations for powerpc in order to remove
unneeded macro definitions.
Tested on ppc64le with and without --disable-multi-arch flag.
* sysdeps/powerpc/powerpc64/multiarch/strchr-power7.S: Define the
implementation-specific function name and remove unneeded macros
definition.
* sysdeps/powerpc/powerpc64/multiarch/strchr-power8.S: Likewise.
* sysdeps/powerpc/powerpc64/multiarch/strchr-ppc64.S: Likewise.
* sysdeps/powerpc/powerpc64/multiarch/strchrnul-power7.S: Likewise.
* sysdeps/powerpc/powerpc64/multiarch/strchrnul-power8.S: Likewise.
* sysdeps/powerpc/powerpc64/multiarch/strrchr-power7.S: Likewise.
* sysdeps/powerpc/powerpc64/power7/strchr.S: Set a default
function name if not defined and pass as parameter to macros
accordingly.
* sysdeps/powerpc/powerpc64/power7/strchrnul.S: Likewise.
* sysdeps/powerpc/powerpc64/power7/strrchr.S: Likewise.
* sysdeps/powerpc/powerpc64/power8/strchr.S: Likewise.
* sysdeps/powerpc/powerpc64/strchr.S: Likewise.
Clean up the IFUNC implementations for powerpc in order to remove
unneeded macro definitions.
Tested on ppc64le with and without --disable-multi-arch flag.
* sysdeps/powerpc/powerpc64/multiarch/strlen-power7.S: Define
the strlen implementation name and remove unneeded macros definition.
* sysdeps/powerpc/powerpc64/multiarch/strlen-power8.S: Likewise.
* sysdeps/powerpc/powerpc64/multiarch/strlen-ppc64.S: Likewise.
* sysdeps/powerpc/powerpc64/multiarch/strnlen-power7.S: Define
the strnlen implementation name and remove unneeded macros definition.
* sysdeps/powerpc/powerpc64/power7/strlen.S: Set a default function
name if not defined and pass as parameter to macros accordingly.
* sysdeps/powerpc/powerpc64/power7/strnlen.S: Likewise.
* sysdeps/powerpc/powerpc64/power8/strlen.S: Likewise.
* sysdeps/powerpc/powerpc64/strlen.S: Likewise.
Clean up the IFUNC implementations for powerpc in order to remove
unneeded macro definitions.
Tested on ppc64le with and without --disable-multi-arch flag.
* sysdeps/powerpc/powerpc64/multiarch/strcasecmp_l-power7.S: Define
the implementation-specific function name and remove unneeded
macros definition.
* sysdeps/powerpc/powerpc64/multiarch/strcmp-power7.S: Likewise.
* sysdeps/powerpc/powerpc64/multiarch/strcmp-power8.S Likewise.
* sysdeps/powerpc/powerpc64/multiarch/strcmp-power9.S: Likewise.
* sysdeps/powerpc/powerpc64/multiarch/strcmp-ppc64.S: Likewise.
* sysdeps/powerpc/powerpc64/multiarch/strncmp-power4.S: Likewise.
* sysdeps/powerpc/powerpc64/multiarch/strncmp-power7.S: Likewise.
* sysdeps/powerpc/powerpc64/multiarch/strncmp-power8.S: Likewise.
* sysdeps/powerpc/powerpc64/multiarch/strncmp-power9.S: Likewise.
* sysdeps/powerpc/powerpc64/multiarch/strncmp-ppc64.S: Likewise.
* sysdeps/powerpc/powerpc64/power4/strncmp.S: Set a default function
name if not defined and pass as parameter to macros accordingly.
* sysdeps/powerpc/powerpc64/power7/strcmp.S: Likewise.
* sysdeps/powerpc/powerpc64/power7/strncmp.S: Likewise.
* sysdeps/powerpc/powerpc64/power8/strcmp.S: Likewise.
* sysdeps/powerpc/powerpc64/power8/strncmp.S: Likewise.
* sysdeps/powerpc/powerpc64/power9/strcmp.S: Likewise.
* sysdeps/powerpc/powerpc64/power9/strncmp.S: Likewise.
* sysdeps/powerpc/powerpc64/strcmp.S: Likewise.
* sysdeps/powerpc/powerpc64/strncmp.S: Likewise.
Clean up the IFUNC implementations for powerpc in order to remove
unneeded macro definitions.
Tested on ppc64le with and without --disable-multi-arch flag.
* sysdeps/powerpc/powerpc64/multiarch/stpcpy-power8.S: Define the
implementation-specific function name and remove unneeded macros
definition.
* sysdeps/powerpc/powerpc64/multiarch/stpncpy-power7.S: Likewise.
* sysdeps/powerpc/powerpc64/multiarch/stpncpy-power8.S: Likewise.
* sysdeps/powerpc/powerpc64/multiarch/strcpy-power8.S: Likewise.
* sysdeps/powerpc/powerpc64/multiarch/strncpy-power7.S: Likewise.
* sysdeps/powerpc/powerpc64/multiarch/strncpy-power8.S: Likewise.
* sysdeps/powerpc/powerpc64/power7/strncpy.S: Set a default
function name if not defined.
* sysdeps/powerpc/powerpc64/power8/strcpy.S: Likewise.
* sysdeps/powerpc/powerpc64/power8/strncpy.S: Likewise.
Added strnlen POWER8 otimized for long strings. It delivers
same performance as POWER7 implementation for short strings.
This takes advantage of reasonably performing unaligned loads
and bit permutes to check the first 1-16 bytes until
quadword aligned, then checks in 64 bytes strides until unsafe,
then 16 bytes, truncating the count if need be.
Likewise, the POWER7 code is recycled for less than 32 bytes strings.
Tested on ppc64 and ppc64le.
* sysdeps/powerpc/powerpc64/multiarch/Makefile
(sysdep_routines): Add strnlen-power8.
* sysdeps/powerpc/powerpc64/multiarch/ifunc-impl-list.c
(strnlen): Add __strnlen_power8 to list of strnlen functions.
* sysdeps/powerpc/powerpc64/multiarch/strnlen-power8.S:
New file.
* sysdeps/powerpc/powerpc64/multiarch/strnlen.c
(__strnlen): Add __strnlen_power8 to ifunc list.
* sysdeps/powerpc/powerpc64/power8/strnlen.S: New file.
For strings >16B and <32B existing algorithm takes more time than default
implementation when strings are placed closed to end of page. This is due
to byte by byte access for handling page cross. This is improved by
following >32B code path where the address is adjusted to aligned memory
before doing load doubleword operation instead of loading bytes.
Tested on powerpc64 and powerpc64le.
The P7 code is used for <=32B strings and for > 32B vectorized loops are used.
This shows as an average 25% improvement depending on the position of search
character. The performance is same for shorter strings.
Tested on ppc64 and ppc64le.
If the input values are unaligned and if there are null characters in the
memory before the starting address of the input values, strcasecmp
gives incorrect return code. Fixed it by adding mask the bits that
are not part of the string.
This implementation is based on the one already used at
sysdeps/x86_64/fpu/e_expf.S.
This implementation improves the performance by ~14% on average in synthetic
benchmarks at the cost of decreasing accuracy to 1 ULP.
This implementation utilizes vectors to improve performance
compared to current byte by byte implementation for POWER7.
The performance improvement is upto 4x. This patch is tested
on powerpc64 and powerpc64le.
The file sysdeps/powerpc/sysdeps.h defines aliases for condition register
operands. E.g.: 'cr7' means condition register 7. On the one hand, this
increases readability, as it makes it easier for readers to know whether the
operand is a condition register, a general purpose register or an immediate.
On the other hand, this permits that condition registers be written as if they
were general purpose, and vice-versa, thus reducing the readability of the
code.
This commit removes some of these unintentional misuses.
The changes have no effect on the final code. Checked with objdump.
Call __memset_power8 to pad, with zeros, the remaining bytes in the
dest string on __strncpy_power8 and __stpncpy_power8. This improves
performance when n is larger than the input string, giving ~30% gain for
larger strings without impacting much shorter strings.
This patch optimizes strcasestr function for power >= 8 systems. The average
improvement of this optimization is ~40% and compares 16 bytes at a time
using vector instructions. This patch is tested on powerpc64 and powerpc64le.
This utilizes vectors and bitmasks. For small needle, large
haystack, the performance improvement is upto 8x. For short
strings (0-4B), the cost of computing the bitmask dominates,
and is a tad slower.
The file sysdeps/powerpc/sysdeps.h defines aliases for register operands,
which add the letter 'r' as a prefix to a register name. E.g.: register 20
can be written as 'r20', instead of '20'. On the one hand, this increases
readability, as it makes it easier for readers to know whether the operand is a
register or an immediate. On the other hand, this permits that immediate
operands be written as if they were registers, and vice-versa, thus reducing
the readability of the code.
This commit removes some of these unintentional misuses.
This commit also increases readability of the code by adding the prefix 'cr' to
some uses of the control register.
Both changes have no effect on the final code. Checked with objdump.
* sysdeps/powerpc/powerpc64/power8/strncpy.S: Remove or add register
prefix from operands.
GLIBC memset optimization for POWER8 uses the '.machine power8'
directive, which is only supported officially on binutils 2.24+. This
causes a build failure on older binutils.
Since the requirement of .machine power8 is to correctly assembly the
'mtvsrd' instruction and it is already handled by the MTVSRD_V1_R4
macro, there is no really needed of using it.
The patch replaces the power8 with power7 for .machine directive.
It fixes BZ#17869.
This patch adds an optimized POWER8 strncmp. The implementation focus
on speeding up unaligned cases follwing the ideas of power8 strcmp.
The algorithm first check the initial 16 bytes, then align the first
function source and uses unaligned loads on second argument only.
Aditional checks for page boundaries are done for unaligned cases
(where sources alignment are different).
This patch adds an optimized POWER8 strcmp using unaligned accesses.
The algorithm first check the initial 16 bytes, then align the first
function source and uses unaligned loads on second argument only.
Aditional checks for page boundaries are done for unaligned cases
This patch adds an optimized POWER8 st{r,p}ncpy using unaligned accesses.
It shows 10%-80% improvement over the optimized POWER7 one that uses
only aligned accesses, specially on unaligned inputs.
The algorithm first read and check 16 bytes (if inputs do not cross a 4K
page size). The it realign source to 16-bytes and issue a 16 bytes read
and compare loop to speedup null byte checks for large strings. Also,
different from POWER7 optimization, the null pad is done inline in the
implementation using possible unaligned accesses, instead of realying on
a memset call. Special case is added for page cross reads.
This patch adds an optimized POWER8 strcpy using unaligned accesses.
For strings up to 16 bytes the implementation first calculate the
string size, like strlen, and issues a memcpy. For larger strings,
source is first aligned to 16 bytes and then tested over a loop that
reads 16 bytes am combine the cmpb results for speedup. Special case is
added for page cross reads.
It shows 30%-60% improvement over the optimized POWER7 one that uses
only aligned accesses.
This patch adds an optimized memset implementation for POWER8. For
sizes from 0 to 255 bytes, a word/doubleword algorithm similar to
POWER7 optimized one is used.
For size higher than 255 two strategies are used:
1. If the constant is different than 0, the memory is written with
altivec vector instruction;
2. If constant is 0, dbcz instructions are used. The loop is unrolled
to clear 512 byte at time.
Using vector instructions increases throughput considerable, with a
double performance for sizes larger than 1024. The dcbz loops unrolls
also shows performance improvement, by doubling throughput for sizes
larger than 8192 bytes.
