diff options
author | Heiko Carstens <heiko.carstens@de.ibm.com> | 2013-09-18 13:45:36 +0400 |
---|---|---|
committer | Martin Schwidefsky <schwidefsky@de.ibm.com> | 2013-10-24 19:16:55 +0400 |
commit | 746479cdcbb131a0645e8cb4a35d6b26126e3e4c (patch) | |
tree | 7cf0fa041804197c14abe21c56bdf27c82274718 /arch/s390/include/asm/bitops.h | |
parent | 8e6a8285668b139460cc6852bfd58fdbd00c7157 (diff) | |
download | linux-746479cdcbb131a0645e8cb4a35d6b26126e3e4c.tar.xz |
s390/bitops: use generic find bit functions / reimplement _left variant
Just like all other architectures we should use out-of-line find bit
operations, since the inline variant bloat the size of the kernel image.
And also like all other architecures we should only supply optimized
variants of the __ffs, ffs, etc. primitives.
Therefore this patch removes the inlined s390 find bit functions and uses
the generic out-of-line variants instead.
The optimization of the primitives follows with the next patch.
With this patch also the functions find_first_bit_left() and
find_next_bit_left() have been reimplemented, since logically, they are
nothing else but a find_first_bit()/find_next_bit() implementation that
use an inverted __fls() instead of __ffs().
Also the restriction that these functions only work on machines which
support the "flogr" instruction is gone now.
This reduces the size of the kernel image (defconfig, -march=z9-109)
by 144,482 bytes.
Alone the size of the function build_sched_domains() gets reduced from
7 KB to 3,5 KB.
We also git rid of unused functions like find_first_bit_le()...
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Diffstat (limited to 'arch/s390/include/asm/bitops.h')
-rw-r--r-- | arch/s390/include/asm/bitops.h | 600 |
1 files changed, 46 insertions, 554 deletions
diff --git a/arch/s390/include/asm/bitops.h b/arch/s390/include/asm/bitops.h index 16df62dde094..e5ca8598c06b 100644 --- a/arch/s390/include/asm/bitops.h +++ b/arch/s390/include/asm/bitops.h @@ -1,10 +1,40 @@ /* - * S390 version - * Copyright IBM Corp. 1999 - * Author(s): Martin Schwidefsky (schwidefsky@de.ibm.com) + * Copyright IBM Corp. 1999,2013 * - * Derived from "include/asm-i386/bitops.h" - * Copyright (C) 1992, Linus Torvalds + * Author(s): Martin Schwidefsky <schwidefsky@de.ibm.com>, + * + * The description below was taken in large parts from the powerpc + * bitops header file: + * Within a word, bits are numbered LSB first. Lot's of places make + * this assumption by directly testing bits with (val & (1<<nr)). + * This can cause confusion for large (> 1 word) bitmaps on a + * big-endian system because, unlike little endian, the number of each + * bit depends on the word size. + * + * The bitop functions are defined to work on unsigned longs, so for an + * s390x system the bits end up numbered: + * |63..............0|127............64|191...........128|255...........196| + * and on s390: + * |31.....0|63....31|95....64|127...96|159..128|191..160|223..192|255..224| + * + * There are a few little-endian macros used mostly for filesystem + * bitmaps, these work on similar bit arrays layouts, but + * byte-oriented: + * |7...0|15...8|23...16|31...24|39...32|47...40|55...48|63...56| + * + * The main difference is that bit 3-5 (64b) or 3-4 (32b) in the bit + * number field needs to be reversed compared to the big-endian bit + * fields. This can be achieved by XOR with 0x38 (64b) or 0x18 (32b). + * + * We also have special functions which work with an MSB0 encoding: + * on an s390x system the bits are numbered: + * |0..............63|64............127|128...........191|192...........255| + * and on s390: + * |0.....31|31....63|64....95|96...127|128..159|160..191|192..223|224..255| + * + * The main difference is that bit 0-63 (64b) or 0-31 (32b) in the bit + * number field needs to be reversed compared to the LSB0 encoded bit + * fields. This can be achieved by XOR with 0x3f (64b) or 0x1f (32b). * */ @@ -18,46 +48,6 @@ #include <linux/typecheck.h> #include <linux/compiler.h> -/* - * 32 bit bitops format: - * bit 0 is the LSB of *addr; bit 31 is the MSB of *addr; - * bit 32 is the LSB of *(addr+4). That combined with the - * big endian byte order on S390 give the following bit - * order in memory: - * 1f 1e 1d 1c 1b 1a 19 18 17 16 15 14 13 12 11 10 \ - * 0f 0e 0d 0c 0b 0a 09 08 07 06 05 04 03 02 01 00 - * after that follows the next long with bit numbers - * 3f 3e 3d 3c 3b 3a 39 38 37 36 35 34 33 32 31 30 - * 2f 2e 2d 2c 2b 2a 29 28 27 26 25 24 23 22 21 20 - * The reason for this bit ordering is the fact that - * in the architecture independent code bits operations - * of the form "flags |= (1 << bitnr)" are used INTERMIXED - * with operation of the form "set_bit(bitnr, flags)". - * - * 64 bit bitops format: - * bit 0 is the LSB of *addr; bit 63 is the MSB of *addr; - * bit 64 is the LSB of *(addr+8). That combined with the - * big endian byte order on S390 give the following bit - * order in memory: - * 3f 3e 3d 3c 3b 3a 39 38 37 36 35 34 33 32 31 30 - * 2f 2e 2d 2c 2b 2a 29 28 27 26 25 24 23 22 21 20 - * 1f 1e 1d 1c 1b 1a 19 18 17 16 15 14 13 12 11 10 - * 0f 0e 0d 0c 0b 0a 09 08 07 06 05 04 03 02 01 00 - * after that follows the next long with bit numbers - * 7f 7e 7d 7c 7b 7a 79 78 77 76 75 74 73 72 71 70 - * 6f 6e 6d 6c 6b 6a 69 68 67 66 65 64 63 62 61 60 - * 5f 5e 5d 5c 5b 5a 59 58 57 56 55 54 53 52 51 50 - * 4f 4e 4d 4c 4b 4a 49 48 47 46 45 44 43 42 41 40 - * The reason for this bit ordering is the fact that - * in the architecture independent code bits operations - * of the form "flags |= (1 << bitnr)" are used INTERMIXED - * with operation of the form "set_bit(bitnr, flags)". - */ - -/* bitmap tables from arch/s390/kernel/bitmap.c */ -extern const char _zb_findmap[]; -extern const char _sb_findmap[]; - #ifndef CONFIG_64BIT #define __BITOPS_OR "or" @@ -310,522 +300,24 @@ static inline int test_bit(unsigned long nr, const volatile unsigned long *ptr) } /* - * Optimized find bit helper functions. - */ - -/** - * __ffz_word_loop - find byte offset of first long != -1UL - * @addr: pointer to array of unsigned long - * @size: size of the array in bits - */ -static inline unsigned long __ffz_word_loop(const unsigned long *addr, - unsigned long size) -{ - typedef struct { long _[__BITOPS_WORDS(size)]; } addrtype; - unsigned long bytes = 0; - - asm volatile( -#ifndef CONFIG_64BIT - " ahi %1,-1\n" - " sra %1,5\n" - " jz 1f\n" - "0: c %2,0(%0,%3)\n" - " jne 1f\n" - " la %0,4(%0)\n" - " brct %1,0b\n" - "1:\n" -#else - " aghi %1,-1\n" - " srag %1,%1,6\n" - " jz 1f\n" - "0: cg %2,0(%0,%3)\n" - " jne 1f\n" - " la %0,8(%0)\n" - " brct %1,0b\n" - "1:\n" -#endif - : "+&a" (bytes), "+&d" (size) - : "d" (-1UL), "a" (addr), "m" (*(addrtype *) addr) - : "cc" ); - return bytes; -} - -/** - * __ffs_word_loop - find byte offset of first long != 0UL - * @addr: pointer to array of unsigned long - * @size: size of the array in bits - */ -static inline unsigned long __ffs_word_loop(const unsigned long *addr, - unsigned long size) -{ - typedef struct { long _[__BITOPS_WORDS(size)]; } addrtype; - unsigned long bytes = 0; - - asm volatile( -#ifndef CONFIG_64BIT - " ahi %1,-1\n" - " sra %1,5\n" - " jz 1f\n" - "0: c %2,0(%0,%3)\n" - " jne 1f\n" - " la %0,4(%0)\n" - " brct %1,0b\n" - "1:\n" -#else - " aghi %1,-1\n" - " srag %1,%1,6\n" - " jz 1f\n" - "0: cg %2,0(%0,%3)\n" - " jne 1f\n" - " la %0,8(%0)\n" - " brct %1,0b\n" - "1:\n" -#endif - : "+&a" (bytes), "+&a" (size) - : "d" (0UL), "a" (addr), "m" (*(addrtype *) addr) - : "cc" ); - return bytes; -} - -/** - * __ffz_word - add number of the first unset bit - * @nr: base value the bit number is added to - * @word: the word that is searched for unset bits - */ -static inline unsigned long __ffz_word(unsigned long nr, unsigned long word) -{ -#ifdef CONFIG_64BIT - if ((word & 0xffffffff) == 0xffffffff) { - word >>= 32; - nr += 32; - } -#endif - if ((word & 0xffff) == 0xffff) { - word >>= 16; - nr += 16; - } - if ((word & 0xff) == 0xff) { - word >>= 8; - nr += 8; - } - return nr + _zb_findmap[(unsigned char) word]; -} - -/** - * __ffs_word - add number of the first set bit - * @nr: base value the bit number is added to - * @word: the word that is searched for set bits - */ -static inline unsigned long __ffs_word(unsigned long nr, unsigned long word) -{ -#ifdef CONFIG_64BIT - if ((word & 0xffffffff) == 0) { - word >>= 32; - nr += 32; - } -#endif - if ((word & 0xffff) == 0) { - word >>= 16; - nr += 16; - } - if ((word & 0xff) == 0) { - word >>= 8; - nr += 8; - } - return nr + _sb_findmap[(unsigned char) word]; -} - - -/** - * __load_ulong_be - load big endian unsigned long - * @p: pointer to array of unsigned long - * @offset: byte offset of source value in the array - */ -static inline unsigned long __load_ulong_be(const unsigned long *p, - unsigned long offset) -{ - p = (unsigned long *)((unsigned long) p + offset); - return *p; -} - -/** - * __load_ulong_le - load little endian unsigned long - * @p: pointer to array of unsigned long - * @offset: byte offset of source value in the array - */ -static inline unsigned long __load_ulong_le(const unsigned long *p, - unsigned long offset) -{ - unsigned long word; - - p = (unsigned long *)((unsigned long) p + offset); -#ifndef CONFIG_64BIT - asm volatile( - " ic %0,%O1(%R1)\n" - " icm %0,2,%O1+1(%R1)\n" - " icm %0,4,%O1+2(%R1)\n" - " icm %0,8,%O1+3(%R1)" - : "=&d" (word) : "Q" (*p) : "cc"); -#else - asm volatile( - " lrvg %0,%1" - : "=d" (word) : "m" (*p) ); -#endif - return word; -} - -/* - * The various find bit functions. - */ - -/* - * ffz - find first zero in word. - * @word: The word to search - * - * Undefined if no zero exists, so code should check against ~0UL first. - */ -static inline unsigned long ffz(unsigned long word) -{ - return __ffz_word(0, word); -} - -/** - * __ffs - find first bit in word. - * @word: The word to search - * - * Undefined if no bit exists, so code should check against 0 first. - */ -static inline unsigned long __ffs (unsigned long word) -{ - return __ffs_word(0, word); -} - -/** - * ffs - find first bit set - * @x: the word to search - * - * This is defined the same way as - * the libc and compiler builtin ffs routines, therefore - * differs in spirit from the above ffz (man ffs). - */ -static inline int ffs(int x) -{ - if (!x) - return 0; - return __ffs_word(1, x); -} - -/** - * find_first_zero_bit - find the first zero bit in a memory region - * @addr: The address to start the search at - * @size: The maximum size to search - * - * Returns the bit-number of the first zero bit, not the number of the byte - * containing a bit. - */ -static inline unsigned long find_first_zero_bit(const unsigned long *addr, - unsigned long size) -{ - unsigned long bytes, bits; - - if (!size) - return 0; - bytes = __ffz_word_loop(addr, size); - bits = __ffz_word(bytes*8, __load_ulong_be(addr, bytes)); - return (bits < size) ? bits : size; -} -#define find_first_zero_bit find_first_zero_bit - -/** - * find_first_bit - find the first set bit in a memory region - * @addr: The address to start the search at - * @size: The maximum size to search - * - * Returns the bit-number of the first set bit, not the number of the byte - * containing a bit. - */ -static inline unsigned long find_first_bit(const unsigned long * addr, - unsigned long size) -{ - unsigned long bytes, bits; - - if (!size) - return 0; - bytes = __ffs_word_loop(addr, size); - bits = __ffs_word(bytes*8, __load_ulong_be(addr, bytes)); - return (bits < size) ? bits : size; -} -#define find_first_bit find_first_bit - -/* - * Big endian variant whichs starts bit counting from left using - * the flogr (find leftmost one) instruction. + * ATTENTION: + * find_first_bit_left() and find_next_bit_left() use MSB0 encoding. */ -static inline unsigned long __flo_word(unsigned long nr, unsigned long val) -{ - register unsigned long bit asm("2") = val; - register unsigned long out asm("3"); - - asm volatile ( - " .insn rre,0xb9830000,%[bit],%[bit]\n" - : [bit] "+d" (bit), [out] "=d" (out) : : "cc"); - return nr + bit; -} - -/* - * 64 bit special left bitops format: - * order in memory: - * 00 01 02 03 04 05 06 07 08 09 0a 0b 0c 0d 0e 0f - * 10 11 12 13 14 15 16 17 18 19 1a 1b 1c 1d 1e 1f - * 20 21 22 23 24 25 26 27 28 29 2a 2b 2c 2d 2e 2f - * 30 31 32 33 34 35 36 37 38 39 3a 3b 3c 3d 3e 3f - * after that follows the next long with bit numbers - * 40 41 42 43 44 45 46 47 48 49 4a 4b 4c 4d 4e 4f - * 50 51 52 53 54 55 56 57 58 59 5a 5b 5c 5d 5e 5f - * 60 61 62 63 64 65 66 67 68 69 6a 6b 6c 6d 6e 6f - * 70 71 72 73 74 75 76 77 78 79 7a 7b 7c 7d 7e 7f - * The reason for this bit ordering is the fact that - * the hardware sets bits in a bitmap starting at bit 0 - * and we don't want to scan the bitmap from the 'wrong - * end'. - */ -static inline unsigned long find_first_bit_left(const unsigned long *addr, - unsigned long size) -{ - unsigned long bytes, bits; - - if (!size) - return 0; - bytes = __ffs_word_loop(addr, size); - bits = __flo_word(bytes * 8, __load_ulong_be(addr, bytes)); - return (bits < size) ? bits : size; -} +unsigned long find_first_bit_left(const unsigned long *addr, unsigned long size); +unsigned long find_next_bit_left(const unsigned long *addr, unsigned long size, + unsigned long offset); -static inline int find_next_bit_left(const unsigned long *addr, - unsigned long size, - unsigned long offset) -{ - const unsigned long *p; - unsigned long bit, set; - - if (offset >= size) - return size; - bit = offset & (BITS_PER_LONG - 1); - offset -= bit; - size -= offset; - p = addr + offset / BITS_PER_LONG; - if (bit) { - set = __flo_word(0, *p & (~0UL >> bit)); - if (set >= size) - return size + offset; - if (set < BITS_PER_LONG) - return set + offset; - offset += BITS_PER_LONG; - size -= BITS_PER_LONG; - p++; - } - return offset + find_first_bit_left(p, size); -} - -#define for_each_set_bit_left(bit, addr, size) \ - for ((bit) = find_first_bit_left((addr), (size)); \ - (bit) < (size); \ - (bit) = find_next_bit_left((addr), (size), (bit) + 1)) - -/* same as for_each_set_bit() but use bit as value to start with */ -#define for_each_set_bit_left_cont(bit, addr, size) \ - for ((bit) = find_next_bit_left((addr), (size), (bit)); \ - (bit) < (size); \ - (bit) = find_next_bit_left((addr), (size), (bit) + 1)) - -/** - * find_next_zero_bit - find the first zero bit in a memory region - * @addr: The address to base the search on - * @offset: The bitnumber to start searching at - * @size: The maximum size to search - */ -static inline int find_next_zero_bit (const unsigned long * addr, - unsigned long size, - unsigned long offset) -{ - const unsigned long *p; - unsigned long bit, set; - - if (offset >= size) - return size; - bit = offset & (BITS_PER_LONG - 1); - offset -= bit; - size -= offset; - p = addr + offset / BITS_PER_LONG; - if (bit) { - /* - * __ffz_word returns BITS_PER_LONG - * if no zero bit is present in the word. - */ - set = __ffz_word(bit, *p >> bit); - if (set >= size) - return size + offset; - if (set < BITS_PER_LONG) - return set + offset; - offset += BITS_PER_LONG; - size -= BITS_PER_LONG; - p++; - } - return offset + find_first_zero_bit(p, size); -} -#define find_next_zero_bit find_next_zero_bit - -/** - * find_next_bit - find the first set bit in a memory region - * @addr: The address to base the search on - * @offset: The bitnumber to start searching at - * @size: The maximum size to search - */ -static inline int find_next_bit (const unsigned long * addr, - unsigned long size, - unsigned long offset) -{ - const unsigned long *p; - unsigned long bit, set; - - if (offset >= size) - return size; - bit = offset & (BITS_PER_LONG - 1); - offset -= bit; - size -= offset; - p = addr + offset / BITS_PER_LONG; - if (bit) { - /* - * __ffs_word returns BITS_PER_LONG - * if no one bit is present in the word. - */ - set = __ffs_word(0, *p & (~0UL << bit)); - if (set >= size) - return size + offset; - if (set < BITS_PER_LONG) - return set + offset; - offset += BITS_PER_LONG; - size -= BITS_PER_LONG; - p++; - } - return offset + find_first_bit(p, size); -} -#define find_next_bit find_next_bit - -/* - * Every architecture must define this function. It's the fastest - * way of searching a 140-bit bitmap where the first 100 bits are - * unlikely to be set. It's guaranteed that at least one of the 140 - * bits is cleared. - */ -static inline int sched_find_first_bit(unsigned long *b) -{ - return find_first_bit(b, 140); -} - -#include <asm-generic/bitops/fls.h> +#include <asm-generic/bitops/__ffs.h> +#include <asm-generic/bitops/ffs.h> #include <asm-generic/bitops/__fls.h> +#include <asm-generic/bitops/fls.h> #include <asm-generic/bitops/fls64.h> - +#include <asm-generic/bitops/ffz.h> +#include <asm-generic/bitops/find.h> #include <asm-generic/bitops/hweight.h> #include <asm-generic/bitops/lock.h> - -/* - * ATTENTION: intel byte ordering convention for ext2 and minix !! - * bit 0 is the LSB of addr; bit 31 is the MSB of addr; - * bit 32 is the LSB of (addr+4). - * That combined with the little endian byte order of Intel gives the - * following bit order in memory: - * 07 06 05 04 03 02 01 00 15 14 13 12 11 10 09 08 \ - * 23 22 21 20 19 18 17 16 31 30 29 28 27 26 25 24 - */ - -static inline int find_first_zero_bit_le(void *vaddr, unsigned int size) -{ - unsigned long bytes, bits; - - if (!size) - return 0; - bytes = __ffz_word_loop(vaddr, size); - bits = __ffz_word(bytes*8, __load_ulong_le(vaddr, bytes)); - return (bits < size) ? bits : size; -} -#define find_first_zero_bit_le find_first_zero_bit_le - -static inline int find_next_zero_bit_le(void *vaddr, unsigned long size, - unsigned long offset) -{ - unsigned long *addr = vaddr, *p; - unsigned long bit, set; - - if (offset >= size) - return size; - bit = offset & (BITS_PER_LONG - 1); - offset -= bit; - size -= offset; - p = addr + offset / BITS_PER_LONG; - if (bit) { - /* - * s390 version of ffz returns BITS_PER_LONG - * if no zero bit is present in the word. - */ - set = __ffz_word(bit, __load_ulong_le(p, 0) >> bit); - if (set >= size) - return size + offset; - if (set < BITS_PER_LONG) - return set + offset; - offset += BITS_PER_LONG; - size -= BITS_PER_LONG; - p++; - } - return offset + find_first_zero_bit_le(p, size); -} -#define find_next_zero_bit_le find_next_zero_bit_le - -static inline unsigned long find_first_bit_le(void *vaddr, unsigned long size) -{ - unsigned long bytes, bits; - - if (!size) - return 0; - bytes = __ffs_word_loop(vaddr, size); - bits = __ffs_word(bytes*8, __load_ulong_le(vaddr, bytes)); - return (bits < size) ? bits : size; -} -#define find_first_bit_le find_first_bit_le - -static inline int find_next_bit_le(void *vaddr, unsigned long size, - unsigned long offset) -{ - unsigned long *addr = vaddr, *p; - unsigned long bit, set; - - if (offset >= size) - return size; - bit = offset & (BITS_PER_LONG - 1); - offset -= bit; - size -= offset; - p = addr + offset / BITS_PER_LONG; - if (bit) { - /* - * s390 version of ffz returns BITS_PER_LONG - * if no zero bit is present in the word. - */ - set = __ffs_word(0, __load_ulong_le(p, 0) & (~0UL << bit)); - if (set >= size) - return size + offset; - if (set < BITS_PER_LONG) - return set + offset; - offset += BITS_PER_LONG; - size -= BITS_PER_LONG; - p++; - } - return offset + find_first_bit_le(p, size); -} -#define find_next_bit_le find_next_bit_le - +#include <asm-generic/bitops/sched.h> #include <asm-generic/bitops/le.h> - #include <asm-generic/bitops/ext2-atomic-setbit.h> #endif /* _S390_BITOPS_H */ |