Minki/linux
1
0
Fork 1
mirror of https://github.com/torvalds/linux.git synced 2024-12-11 21:52:04 +00:00
Code Issues Packages Projects Releases Wiki Activity

ARM: 8785/1: use compiler built-ins for ffs and fls

Browse Source 
On ARMv5 and above, it is beneficial to use compiler built-ins such as
__builtin_ffs() and __builtin_ctzl() to implement ffs(), __ffs(), fls()
and __fls(). The compiler does inline the clz instruction and even the
rbit instruction when available, or provide a constant value when
possible. On ARMv4 the compiler calls out to helper functions for those
built-ins so it is best to keep the open coded versions in that case.

Signed-off-by: Nicolas Pitre <nico@linaro.org>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
This commit is contained in:
Nicolas Pitre 2018-07-25 04:38:54 +01:00 committed by Russell King
parent cbfc5619e0
commit 001a30c4d0
1 changed files with 9 additions and 83 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

92
arch/arm/include/asm/bitops.h
View File

@ -215,7 +215,6 @@ extern int _find_next_bit_be(const unsigned long *p, int size, int offset);
#if __LINUX_ARM_ARCH__ < 5
#include <asm-generic/bitops/ffz.h>
#include <asm-generic/bitops/__fls.h>
#include <asm-generic/bitops/__ffs.h>
#include <asm-generic/bitops/fls.h>
@ -223,93 +222,20 @@ extern int _find_next_bit_be(const unsigned long *p, int size, int offset);
#else
static inline int constant_fls(int x)
{
int r = 32;
if (!x)
return 0;
if (!(x & 0xffff0000u)) {
x <<= 16;
r -= 16;
}
if (!(x & 0xff000000u)) {
x <<= 8;
r -= 8;
}
if (!(x & 0xf0000000u)) {
x <<= 4;
r -= 4;
}
if (!(x & 0xc0000000u)) {
x <<= 2;
r -= 2;
}
if (!(x & 0x80000000u)) {
x <<= 1;
r -= 1;
}
return r;
}
/*
* On ARMv5 and above those functions can be implemented around the
* clz instruction for much better code efficiency. __clz returns
* the number of leading zeros, zero input will return 32, and
* 0x80000000 will return 0.
* On ARMv5 and above, the gcc built-ins may rely on the clz instruction
* and produce optimal inlined code in all cases. On ARMv7 it is even
* better by also using the rbit instruction.
*/
static inline unsigned int __clz(unsigned int x)
{
unsigned int ret;
asm("clz\t%0, %1" : "=r" (ret) : "r" (x));
return ret;
}
/*
* fls() returns zero if the input is zero, otherwise returns the bit
* position of the last set bit, where the LSB is 1 and MSB is 32.
*/
static inline int fls(int x)
{
if (__builtin_constant_p(x))
return constant_fls(x);
return 32 - __clz(x);
}
/*
* __fls() returns the bit position of the last bit set, where the
* LSB is 0 and MSB is 31. Zero input is undefined.
*/
static inline unsigned long __fls(unsigned long x)
{
return fls(x) - 1;
}
/*
* ffs() returns zero if the input was zero, otherwise returns the bit
* position of the first set bit, where the LSB is 1 and MSB is 32.
*/
static inline int ffs(int x)
{
return fls(x & -x);
}
/*
* __ffs() returns the bit position of the first bit set, where the
* LSB is 0 and MSB is 31. Zero input is undefined.
*/
static inline unsigned long __ffs(unsigned long x)
{
return ffs(x) - 1;
}
#define ffz(x) __ffs( ~(x) )
#include <asm-generic/bitops/builtin-__fls.h>
#include <asm-generic/bitops/builtin-__ffs.h>
#include <asm-generic/bitops/builtin-fls.h>
#include <asm-generic/bitops/builtin-ffs.h>
#endif
#include <asm-generic/bitops/ffz.h>
#include <asm-generic/bitops/fls64.h>
#include <asm-generic/bitops/sched.h>