~ [ source navigation ] ~ [ diff markup ] ~ [ identifier search ] ~

TOMOYO Linux Cross Reference
Linux/arch/riscv/lib/memset.S

Version: ~ [ linux-6.12-rc7 ] ~ [ linux-6.11.7 ] ~ [ linux-6.10.14 ] ~ [ linux-6.9.12 ] ~ [ linux-6.8.12 ] ~ [ linux-6.7.12 ] ~ [ linux-6.6.60 ] ~ [ linux-6.5.13 ] ~ [ linux-6.4.16 ] ~ [ linux-6.3.13 ] ~ [ linux-6.2.16 ] ~ [ linux-6.1.116 ] ~ [ linux-6.0.19 ] ~ [ linux-5.19.17 ] ~ [ linux-5.18.19 ] ~ [ linux-5.17.15 ] ~ [ linux-5.16.20 ] ~ [ linux-5.15.171 ] ~ [ linux-5.14.21 ] ~ [ linux-5.13.19 ] ~ [ linux-5.12.19 ] ~ [ linux-5.11.22 ] ~ [ linux-5.10.229 ] ~ [ linux-5.9.16 ] ~ [ linux-5.8.18 ] ~ [ linux-5.7.19 ] ~ [ linux-5.6.19 ] ~ [ linux-5.5.19 ] ~ [ linux-5.4.285 ] ~ [ linux-5.3.18 ] ~ [ linux-5.2.21 ] ~ [ linux-5.1.21 ] ~ [ linux-5.0.21 ] ~ [ linux-4.20.17 ] ~ [ linux-4.19.323 ] ~ [ linux-4.18.20 ] ~ [ linux-4.17.19 ] ~ [ linux-4.16.18 ] ~ [ linux-4.15.18 ] ~ [ linux-4.14.336 ] ~ [ linux-4.13.16 ] ~ [ linux-4.12.14 ] ~ [ linux-4.11.12 ] ~ [ linux-4.10.17 ] ~ [ linux-4.9.337 ] ~ [ linux-4.4.302 ] ~ [ linux-3.10.108 ] ~ [ linux-2.6.32.71 ] ~ [ linux-2.6.0 ] ~ [ linux-2.4.37.11 ] ~ [ unix-v6-master ] ~ [ ccs-tools-1.8.12 ] ~ [ policy-sample ] ~
Architecture: ~ [ i386 ] ~ [ alpha ] ~ [ m68k ] ~ [ mips ] ~ [ ppc ] ~ [ sparc ] ~ [ sparc64 ] ~

Diff markup

Differences between /arch/riscv/lib/memset.S (Version linux-6.12-rc7) and /arch/alpha/lib/memset.S (Version linux-5.19.17)


  1 /* SPDX-License-Identifier: GPL-2.0-only */    !!   1 /* SPDX-License-Identifier: GPL-2.0 */
  2 /*                                                  2 /*
  3  * Copyright (C) 2013 Regents of the Universit !!   3  * linux/arch/alpha/lib/memset.S
                                                   >>   4  *
                                                   >>   5  * This is an efficient (and small) implementation of the C library "memset()"
                                                   >>   6  * function for the alpha.
                                                   >>   7  *
                                                   >>   8  *      (C) Copyright 1996 Linus Torvalds
                                                   >>   9  *
                                                   >>  10  * This routine is "moral-ware": you are free to use it any way you wish, and
                                                   >>  11  * the only obligation I put on you is a moral one: if you make any improvements
                                                   >>  12  * to the routine, please send me your improvements for me to use similarly.
                                                   >>  13  *
                                                   >>  14  * The scheduling comments are according to the EV5 documentation (and done by 
                                                   >>  15  * hand, so they might well be incorrect, please do tell me about it..)
  4  */                                                16  */
  5                                                !!  17 #include <asm/export.h>
  6                                                !!  18         .set noat
  7 #include <linux/linkage.h>                     !!  19         .set noreorder
  8 #include <asm/asm.h>                           !!  20 .text
  9                                                !!  21         .globl memset
 10 /* void *memset(void *, int, size_t) */        !!  22         .globl __memset
 11 SYM_FUNC_START(__memset)                       !!  23         .globl ___memset
 12         move t0, a0  /* Preserve return value  !!  24         .globl __memset16
 13                                                !!  25         .globl __constant_c_memset
 14         /* Defer to byte-oriented fill for sma !!  26 
 15         sltiu a3, a2, 16                       !!  27         .ent ___memset
 16         bnez a3, 4f                            !!  28 .align 5
 17                                                !!  29 ___memset:
 18         /*                                     !!  30         .frame $30,0,$26,0
 19          * Round to nearest XLEN-aligned addre !!  31         .prologue 0
 20          * greater than or equal to start addr !!  32 
 21          */                                    !!  33         and $17,255,$1          /* E1 */
 22         addi a3, t0, SZREG-1                   !!  34         insbl $17,1,$17         /* .. E0 */
 23         andi a3, a3, ~(SZREG-1)                !!  35         bis $17,$1,$17          /* E0 (p-c latency, next cycle) */
 24         beq a3, t0, 2f  /* Skip if already ali !!  36         sll $17,16,$1           /* E1 (p-c latency, next cycle) */
 25         /* Handle initial misalignment */      !!  37 
 26         sub a4, a3, t0                         !!  38         bis $17,$1,$17          /* E0 (p-c latency, next cycle) */
 27 1:                                             !!  39         sll $17,32,$1           /* E1 (p-c latency, next cycle) */
 28         sb a1, 0(t0)                           !!  40         bis $17,$1,$17          /* E0 (p-c latency, next cycle) */
 29         addi t0, t0, 1                         !!  41         ldq_u $31,0($30)        /* .. E1 */
 30         bltu t0, a3, 1b                        !!  42 
 31         sub a2, a2, a4  /* Update count */     !!  43 .align 5
 32                                                !!  44 __constant_c_memset:
 33 2: /* Duff's device with 32 XLEN stores per it !!  45         addq $18,$16,$6         /* E0 */
 34         /* Broadcast value into all bytes */   !!  46         bis $16,$16,$0          /* .. E1 */
 35         andi a1, a1, 0xff                      !!  47         xor $16,$6,$1           /* E0 */
 36         slli a3, a1, 8                         !!  48         ble $18,end             /* .. E1 */
 37         or a1, a3, a1                          !!  49 
 38         slli a3, a1, 16                        !!  50         bic $1,7,$1             /* E0 */
 39         or a1, a3, a1                          !!  51         beq $1,within_one_quad  /* .. E1 (note EV5 zero-latency forwarding) */
 40 #ifdef CONFIG_64BIT                            !!  52         and $16,7,$3            /* E0 */
 41         slli a3, a1, 32                        !!  53         beq $3,aligned          /* .. E1 (note EV5 zero-latency forwarding) */
 42         or a1, a3, a1                          !!  54 
 43 #endif                                         !!  55         ldq_u $4,0($16)         /* E0 */
 44                                                !!  56         bis $16,$16,$5          /* .. E1 */
 45         /* Calculate end address */            !!  57         insql $17,$16,$2        /* E0 */
 46         andi a4, a2, ~(SZREG-1)                !!  58         subq $3,8,$3            /* .. E1 */
 47         add a3, t0, a4                         !!  59 
 48                                                !!  60         addq $18,$3,$18         /* E0   $18 is new count ($3 is negative) */
 49         andi a4, a4, 31*SZREG  /* Calculate re !!  61         mskql $4,$16,$4         /* .. E1 (and possible load stall) */
 50         beqz a4, 3f            /* Shortcut if  !!  62         subq $16,$3,$16         /* E0   $16 is new aligned destination */
 51         neg a4, a4                             !!  63         bis $2,$4,$1            /* .. E1 */
 52         addi a4, a4, 32*SZREG  /* Calculate in !!  64 
 53                                                !!  65         bis $31,$31,$31         /* E0 */
 54         /* Adjust start address with offset */ !!  66         ldq_u $31,0($30)        /* .. E1 */
 55         sub t0, t0, a4                         !!  67         stq_u $1,0($5)          /* E0 */
 56                                                !!  68         bis $31,$31,$31         /* .. E1 */
 57         /* Jump into loop body */              !!  69 
 58         /* Assumes 32-bit instruction lengths  !!  70 .align 4
 59         la a5, 3f                              !!  71 aligned:
 60 #ifdef CONFIG_64BIT                            !!  72         sra $18,3,$3            /* E0 */
 61         srli a4, a4, 1                         !!  73         and $18,7,$18           /* .. E1 */
 62 #endif                                         !!  74         bis $16,$16,$5          /* E0 */
 63         add a5, a5, a4                         !!  75         beq $3,no_quad          /* .. E1 */
 64         jr a5                                  !!  76 
 65 3:                                             !!  77 .align 3
 66         REG_S a1,        0(t0)                 !!  78 loop:
 67         REG_S a1,    SZREG(t0)                 !!  79         stq $17,0($5)           /* E0 */
 68         REG_S a1,  2*SZREG(t0)                 !!  80         subq $3,1,$3            /* .. E1 */
 69         REG_S a1,  3*SZREG(t0)                 !!  81         addq $5,8,$5            /* E0 */
 70         REG_S a1,  4*SZREG(t0)                 !!  82         bne $3,loop             /* .. E1 */
 71         REG_S a1,  5*SZREG(t0)                 !!  83 
 72         REG_S a1,  6*SZREG(t0)                 !!  84 no_quad:
 73         REG_S a1,  7*SZREG(t0)                 !!  85         bis $31,$31,$31         /* E0 */
 74         REG_S a1,  8*SZREG(t0)                 !!  86         beq $18,end             /* .. E1 */
 75         REG_S a1,  9*SZREG(t0)                 !!  87         ldq $7,0($5)            /* E0 */
 76         REG_S a1, 10*SZREG(t0)                 !!  88         mskqh $7,$6,$2          /* .. E1 (and load stall) */
 77         REG_S a1, 11*SZREG(t0)                 !!  89 
 78         REG_S a1, 12*SZREG(t0)                 !!  90         insqh $17,$6,$4         /* E0 */
 79         REG_S a1, 13*SZREG(t0)                 !!  91         bis $2,$4,$1            /* .. E1 */
 80         REG_S a1, 14*SZREG(t0)                 !!  92         stq $1,0($5)            /* E0 */
 81         REG_S a1, 15*SZREG(t0)                 !!  93         ret $31,($26),1         /* .. E1 */
 82         REG_S a1, 16*SZREG(t0)                 !!  94 
 83         REG_S a1, 17*SZREG(t0)                 !!  95 .align 3
 84         REG_S a1, 18*SZREG(t0)                 !!  96 within_one_quad:
 85         REG_S a1, 19*SZREG(t0)                 !!  97         ldq_u $1,0($16)         /* E0 */
 86         REG_S a1, 20*SZREG(t0)                 !!  98         insql $17,$16,$2        /* E1 */
 87         REG_S a1, 21*SZREG(t0)                 !!  99         mskql $1,$16,$4         /* E0 (after load stall) */
 88         REG_S a1, 22*SZREG(t0)                 !! 100         bis $2,$4,$2            /* E0 */
 89         REG_S a1, 23*SZREG(t0)                 !! 101 
 90         REG_S a1, 24*SZREG(t0)                 !! 102         mskql $2,$6,$4          /* E0 */
 91         REG_S a1, 25*SZREG(t0)                 !! 103         mskqh $1,$6,$2          /* .. E1 */
 92         REG_S a1, 26*SZREG(t0)                 !! 104         bis $2,$4,$1            /* E0 */
 93         REG_S a1, 27*SZREG(t0)                 !! 105         stq_u $1,0($16)         /* E0 */
 94         REG_S a1, 28*SZREG(t0)                 !! 106 
 95         REG_S a1, 29*SZREG(t0)                 !! 107 end:
 96         REG_S a1, 30*SZREG(t0)                 !! 108         ret $31,($26),1         /* E1 */
 97         REG_S a1, 31*SZREG(t0)                 !! 109         .end ___memset
 98         addi t0, t0, 32*SZREG                  !! 110 EXPORT_SYMBOL(___memset)
 99         bltu t0, a3, 3b                        !! 111 EXPORT_SYMBOL(__constant_c_memset)
100         andi a2, a2, SZREG-1  /* Update count  !! 112 
101                                                !! 113         .align 5
102 4:                                             !! 114         .ent __memset16
103         /* Handle trailing misalignment */     !! 115 __memset16:
104         beqz a2, 6f                            !! 116         .prologue 0
105         add a3, t0, a2                         !! 117 
106 5:                                             !! 118         inswl $17,0,$1          /* E0 */
107         sb a1, 0(t0)                           !! 119         inswl $17,2,$2          /* E0 */
108         addi t0, t0, 1                         !! 120         inswl $17,4,$3          /* E0 */
109         bltu t0, a3, 5b                        !! 121         or $1,$2,$1             /* .. E1 */
110 6:                                             !! 122         inswl $17,6,$4          /* E0 */
111         ret                                    !! 123         or $1,$3,$1             /* .. E1 */
112 SYM_FUNC_END(__memset)                         !! 124         or $1,$4,$17            /* E0 */
113 SYM_FUNC_ALIAS_WEAK(memset, __memset)          !! 125         br __constant_c_memset  /* .. E1 */
114 SYM_FUNC_ALIAS(__pi_memset, __memset)          !! 126 
115 SYM_FUNC_ALIAS(__pi___memset, __memset)        !! 127         .end __memset16
                                                   >> 128 EXPORT_SYMBOL(__memset16)
                                                   >> 129 
                                                   >> 130 memset = ___memset
                                                   >> 131 __memset = ___memset
                                                   >> 132         EXPORT_SYMBOL(memset)
                                                   >> 133         EXPORT_SYMBOL(__memset)
                                                      

~ [ source navigation ] ~ [ diff markup ] ~ [ identifier search ] ~

kernel.org | git.kernel.org | LWN.net | Project Home | SVN repository | Mail admin

Linux® is a registered trademark of Linus Torvalds in the United States and other countries.
TOMOYO® is a registered trademark of NTT DATA CORPORATION.

sflogo.php