Moved some assembly optimizations to the target tree

git-svn-id: svn://svn.rockbox.org/rockbox/trunk@10318 a1c6a512-1295-4272-9138-f99709370657

4 files changed, 638 insertions, 0 deletions

diff --git a/firmware/target/sh/memcpy-sh.S b/firmware/target/sh/memcpy-sh.S
new file mode 100755
index 0000000000..0b5e086be9
--- /dev/null
+++ b/firmware/target/sh/memcpy-sh.S
@@ -0,0 +1,217 @@
+/***************************************************************************
+ *             __________               __   ___.
+ *   Open      \______   \ ____   ____ |  | _\_ |__   _______  ___
+ *   Source     |       _//  _ \_/ ___\|  |/ /| __ \ /  _ \  \/  /
+ *   Jukebox    |    |   (  <_> )  \___|    < | \_\ (  <_> > <  <
+ *   Firmware   |____|_  /\____/ \___  >__|_ \|___  /\____/__/\_ \
+ *                     \/            \/     \/    \/            \/
+ * $Id$
+ *
+ * Copyright (C) 2004-2005 by Jens Arnold
+ *
+ * All files in this archive are subject to the GNU General Public License.
+ * See the file COPYING in the source tree root for full license agreement.
+ *
+ * This software is distributed on an "AS IS" basis, WITHOUT WARRANTY OF ANY
+ * KIND, either express or implied.
+ *
+ ****************************************************************************/
+#include "config.h"
+
+    .section    .icode,"ax",@progbits
+
+    .align      2
+    .global     _memcpy
+    .global     ___memcpy_fwd_entry
+    .type       _memcpy,@function
+
+/* Copies <length> bytes of data in memory from <source> to <dest>
+ * This version is optimized for speed
+ *
+ * arguments:
+ *  r4 - destination address
+ *  r5 - source address
+ *  r6 - length
+ *
+ * return value:
+ *  r0 - destination address (like ANSI version)
+ *
+ * register usage:
+ *  r0 - data / scratch
+ *  r1 - 2nd data / scratch
+ *  r2 - scratch
+ *  r3 - first long bound / adjusted end address (only if >= 11 bytes)
+ *  r4 - current dest address
+ *  r5 - current source address
+ *  r6 - source end address
+ *  r7 - stored dest start address
+ *
+ * The instruction order is devised in a way to utilize the pipelining
+ * of the SH1 to the max. The routine also tries to utilize fast page mode.
+ */
+
+_memcpy:
+    mov     r4,r7       /* store dest for returning */
+___memcpy_fwd_entry:
+    add     #-8,r4      /* offset for early increment (max. 2 longs) */
+    mov     #11,r0
+    cmp/hs  r0,r6       /* at least 11 bytes to copy? (ensures 2 aligned longs) */
+    add     r5,r6       /* r6 = source_end */
+    bf      .start_b2   /* no: jump directly to byte loop */
+    
+    mov     #3,r0
+    neg     r5,r3
+    and     r0,r3       /* r3 = (4 - align_offset) % 4 */
+    tst     r3,r3       /* already aligned? */
+    bt      .end_b1     /* yes: skip leading byte loop */
+    
+    add     r5,r3       /* r3 = first source long bound */
+
+    /* leading byte loop: copies 0..3 bytes */
+.loop_b1:
+    mov.b   @r5+,r0     /* load byte & increment source addr */
+    add     #1,r4       /* increment dest addr */
+    mov.b   r0,@(7,r4)  /* store byte */
+    cmp/hi  r5,r3       /* runs r5 up to first long bound */
+    bt      .loop_b1
+    /* now r5 is always at a long boundary */
+    /* -> memory reading is done in longs for all dest alignments */
+
+    /* selector for main copy loop */
+.end_b1:
+    mov     #3,r1
+    and     r4,r1       /* r1 = dest alignment offset */
+    mova    .jmptab,r0
+    mov.b   @(r0,r1),r1 /* select appropriate main loop */
+    add     r0,r1
+    mov     r6,r3       /* move end address to r3 */
+    jmp     @r1         /* and jump to it */
+    add     #-7,r3      /* adjust end addr for main loops doing 2 longs/pass */
+    
+    /** main loops, copying 2 longs per pass to profit from fast page mode **/
+
+    /* long aligned destination (fastest) */
+    .align  2
+.loop_do0:
+    mov.l   @r5+,r1     /* load first long & increment source addr */
+    add     #16,r4      /* increment dest addr & account for decrementing stores */
+    mov.l   @r5+,r0     /* load second long & increment source addr */
+    cmp/hi  r5,r3       /* runs r5 up to last or second last long bound */
+    mov.l   r0,@-r4     /* store second long */
+    mov.l   r1,@-r4     /* store first long; NOT ALIGNED - no speed loss here! */
+    bt      .loop_do0
+           
+    add     #4,r3       /* readjust end address */
+    cmp/hi  r5,r3       /* one long left? */
+    bf      .start_b2   /* no, jump to trailing byte loop */
+    
+    mov.l   @r5+,r0     /* load last long & increment source addr */
+    add     #4,r4       /* increment dest addr */
+    bra     .start_b2   /* jump to trailing byte loop */
+    mov.l   r0,@(4,r4)  /* store last long */
+
+    /* word aligned destination (long + 2) */
+    .align  2
+.loop_do2:
+    mov.l   @r5+,r1     /* load first long & increment source addr */
+    add     #16,r4      /* increment dest addr */
+    mov.l   @r5+,r0     /* load second long & increment source addr */
+    cmp/hi  r5,r3       /* runs r5 up to last or second last long bound */
+    mov.w   r0,@-r4     /* store low word of second long */
+    xtrct   r1,r0       /* extract low word of first long & high word of second long */
+    mov.l   r0,@-r4     /* and store as long */
+    swap.w  r1,r0       /* get high word of first long */
+    mov.w   r0,@-r4     /* and store it */
+    bt      .loop_do2
+
+    add     #4,r3       /* readjust end address */
+    cmp/hi  r5,r3       /* one long left? */
+    bf      .start_b2   /* no, jump to trailing byte loop */
+
+    mov.l   @r5+,r0     /* load last long & increment source addr */
+    add     #4,r4       /* increment dest addr */
+    mov.w   r0,@(6,r4)  /* store low word */
+    shlr16  r0          /* get high word */
+    bra     .start_b2   /* jump to trailing byte loop */
+    mov.w   r0,@(4,r4)  /* and store it */
+
+    /* jumptable for loop selector */
+    .align  2
+.jmptab:
+    .byte   .loop_do0 - .jmptab  /* placed in the middle because the SH1 */
+    .byte   .loop_do1 - .jmptab  /* loads bytes sign-extended. Otherwise */
+    .byte   .loop_do2 - .jmptab  /* the last loop would be out of reach */
+    .byte   .loop_do3 - .jmptab  /* of the offset range. */
+
+    /* byte aligned destination (long + 1) */
+    .align  2
+.loop_do1:
+    mov.l   @r5+,r1     /* load first long & increment source addr */
+    add     #16,r4      /* increment dest addr */
+    mov.l   @r5+,r0     /* load second long & increment source addr */
+    cmp/hi  r5,r3       /* runs r5 up to last or second last long bound */
+    mov.b   r0,@-r4     /* store low byte of second long */
+    shlr8   r0          /* get upper 3 bytes */
+    mov     r1,r2       /* copy first long */
+    shll16  r2          /* move low byte of first long all the way up, .. */
+    shll8   r2
+    or      r2,r0       /* ..combine with the 3 bytes of second long.. */
+    mov.l   r0,@-r4     /* ..and store as long */
+    shlr8   r1          /* get middle 2 bytes */
+    mov.w   r1,@-r4     /* store as word */
+    shlr16  r1          /* get upper byte */
+    mov.b   r1,@-r4     /* and store */
+    bt      .loop_do1
+
+    add     #4,r3       /* readjust end address */
+.last_do13:
+    cmp/hi  r5,r3       /* one long left? */
+    bf      .start_b2   /* no, jump to trailing byte loop */
+
+    mov.l   @r5+,r0     /* load last long & increment source addr */
+    add     #12,r4      /* increment dest addr */
+    mov.b   r0,@-r4     /* store low byte */
+    shlr8   r0          /* get middle 2 bytes */
+    mov.w   r0,@-r4     /* store as word */
+    shlr16  r0          /* get upper byte */
+    mov.b   r0,@-r4     /* and store */
+    bra     .start_b2   /* jump to trailing byte loop */
+    add     #-4,r4      /* readjust destination */
+
+    /* byte aligned destination (long + 3) */
+    .align  2
+.loop_do3:
+    mov.l   @r5+,r1     /* load first long & increment source addr */
+    add     #16,r4      /* increment dest addr */
+    mov.l   @r5+,r0     /* load second long & increment source addr */
+    mov     r1,r2       /* copy first long */
+    mov.b   r0,@-r4     /* store low byte of second long */
+    shlr8   r0          /* get middle 2 bytes */
+    mov.w   r0,@-r4     /* store as word */
+    shlr16  r0          /* get upper byte */
+    shll8   r2          /* move lower 3 bytes of first long one up.. */
+    or      r2,r0       /* ..combine with the 1 byte of second long.. */
+    mov.l   r0,@-r4     /* ..and store as long */
+    shlr16  r1          /* get upper byte of first long.. */
+    shlr8   r1
+    cmp/hi  r5,r3       /* runs r5 up to last or second last long bound */
+    mov.b   r1,@-r4     /* ..and store */
+    bt      .loop_do3
+
+    bra     .last_do13  /* handle last longword: reuse routine for (long + 1) */
+    add     #4,r3       /* readjust end address */
+
+    /* trailing byte loop: copies 0..3 bytes (or all for < 11 in total) */
+    .align  2
+.loop_b2:
+    mov.b   @r5+,r0     /* load byte & increment source addr */
+    add     #1,r4       /* increment dest addr */
+    mov.b   r0,@(7,r4)  /* store byte */
+.start_b2:
+    cmp/hi  r5,r6       /* runs r5 up to end address */
+    bt      .loop_b2
+
+    rts
+    mov     r7,r0       /* return dest start address */
+.end:
+    .size   _memcpy,.end-_memcpy
diff --git a/firmware/target/sh/memmove-sh.S b/firmware/target/sh/memmove-sh.S
new file mode 100755
index 0000000000..9ae9ae5fa2
--- /dev/null
+++ b/firmware/target/sh/memmove-sh.S
@@ -0,0 +1,220 @@
+/***************************************************************************
+ *             __________               __   ___.
+ *   Open      \______   \ ____   ____ |  | _\_ |__   _______  ___
+ *   Source     |       _//  _ \_/ ___\|  |/ /| __ \ /  _ \  \/  /
+ *   Jukebox    |    |   (  <_> )  \___|    < | \_\ (  <_> > <  <
+ *   Firmware   |____|_  /\____/ \___  >__|_ \|___  /\____/__/\_ \
+ *                     \/            \/     \/    \/            \/
+ * $Id$
+ *
+ * Copyright (C) 2006 by Jens Arnold
+ *
+ * All files in this archive are subject to the GNU General Public License.
+ * See the file COPYING in the source tree root for full license agreement.
+ *
+ * This software is distributed on an "AS IS" basis, WITHOUT WARRANTY OF ANY
+ * KIND, either express or implied.
+ *
+ ****************************************************************************/
+#include "config.h"
+
+    .section    .icode,"ax",@progbits
+
+    .align      2
+    .global     _memmove
+    .type       _memmove,@function
+
+/* Moves <length> bytes of data in memory from <source> to <dest>
+ * Regions may overlap.
+ * This version is optimized for speed, and needs the corresponding memcpy
+ * implementation for the forward copy branch.
+ *
+ * arguments:
+ *  r4 - destination address
+ *  r5 - source address
+ *  r6 - length
+ *
+ * return value:
+ *  r0 - destination address (like ANSI version)
+ *
+ * register usage:
+ *  r0 - data / scratch
+ *  r1 - 2nd data / scratch
+ *  r2 - scratch
+ *  r3 - last long bound / adjusted start address (only if >= 11 bytes)
+ *  r4 - current dest address
+ *  r5 - source start address
+ *  r6 - current source address
+ *
+ * The instruction order is devised in a way to utilize the pipelining
+ * of the SH1 to the max. The routine also tries to utilize fast page mode.
+ */
+
+_memmove:
+    cmp/hi  r4,r5       /* source > destination */
+    bf     .backward    /* no: backward copy */
+    mov.l  .memcpy_fwd,r0
+    jmp     @r0
+    mov     r4,r7       /* store dest for returning */
+
+    .align  2
+.memcpy_fwd:
+    .long   ___memcpy_fwd_entry
+
+.backward:
+    add     r6,r4       /* r4 = destination end */
+    mov     #11,r0
+    cmp/hs  r0,r6       /* at least 11 bytes to copy? (ensures 2 aligned longs) */
+    add     #-8,r5      /* adjust for late decrement (max. 2 longs) */
+    add     r5,r6       /* r6 = source end - 8 */
+    bf      .start_b2r  /* no: jump directly to byte loop */
+    
+    mov     #-4,r3      /* r3 = 0xfffffffc */
+    and     r6,r3       /* r3 = last source long bound */
+    cmp/hi  r3,r6       /* already aligned? */
+    bf      .end_b1r    /* yes: skip leading byte loop */
+    
+.loop_b1r:
+    mov.b   @(7,r6),r0  /* load byte */
+    add     #-1,r6      /* decrement source addr */
+    mov.b   r0,@-r4     /* store byte */
+    cmp/hi  r3,r6       /* runs r6 down to last long bound */
+    bt      .loop_b1r
+    
+.end_b1r:
+    mov     #3,r1
+    and     r4,r1       /* r1 = dest alignment offset */
+    mova    .jmptab_r,r0
+    mov.b   @(r0,r1),r1 /* select appropriate main loop.. */
+    add     r0,r1
+    mov     r5,r3       /* copy start adress to r3 */
+    jmp     @r1         /* ..and jump to it */
+    add     #7,r3       /* adjust end addr for main loops doing 2 longs/pass */
+
+    /** main loops, copying 2 longs per pass to profit from fast page mode **/
+
+    /* long aligned destination (fastest) */
+    .align  2
+.loop_do0r:
+    mov.l   @r6,r1      /* load first long */
+    add     #-8,r6      /* decrement source addr */
+    mov.l   @(12,r6),r0 /* load second long */
+    cmp/hi  r3,r6       /* runs r6 down to first or second long bound */
+    mov.l   r0,@-r4     /* store second long */
+    mov.l   r1,@-r4     /* store first long; NOT ALIGNED - no speed loss here! */
+    bt      .loop_do0r
+    
+    add     #-4,r3      /* readjust end address */
+    cmp/hi  r3,r6       /* first long left? */
+    bf      .start_b2r  /* no, jump to trailing byte loop */
+    
+    mov.l   @(4,r6),r0  /* load first long */
+    add     #-4,r6      /* decrement source addr */
+    bra     .start_b2r  /* jump to trailing byte loop */
+    mov.l   r0,@-r4     /* store first long */
+    
+    /* word aligned destination (long + 2) */
+    .align  2
+.loop_do2r:
+    mov.l   @r6,r1      /* load first long */
+    add     #-8,r6      /* decrement source addr */
+    mov.l   @(12,r6),r0 /* load second long */
+    cmp/hi  r3,r6       /* runs r6 down to first or second long bound */
+    mov.w   r0,@-r4     /* store low word of second long */
+    xtrct   r1,r0       /* extract low word of first long & high word of second long */
+    mov.l   r0,@-r4     /* and store as long */
+    shlr16  r1          /* get high word of first long */
+    mov.w   r1,@-r4     /* and store it */
+    bt      .loop_do2r
+
+    add     #-4,r3      /* readjust end address */
+    cmp/hi  r3,r6       /* first long left? */
+    bf      .start_b2r  /* no, jump to trailing byte loop */
+
+    mov.l   @(4,r6),r0  /* load first long & decrement source addr */
+    add     #-4,r6      /* decrement source addr */
+    mov.w   r0,@-r4     /* store low word */
+    shlr16  r0          /* get high word */
+    bra     .start_b2r  /* jump to trailing byte loop */
+    mov.w   r0,@-r4     /* and store it */
+
+    /* jumptable for loop selector */
+    .align  2
+.jmptab_r:
+    .byte   .loop_do0r - .jmptab_r  /* placed in the middle because the SH1 */
+    .byte   .loop_do1r - .jmptab_r  /* loads bytes sign-extended. Otherwise */
+    .byte   .loop_do2r - .jmptab_r  /* the last loop would be out of reach */
+    .byte   .loop_do3r - .jmptab_r  /* of the offset range. */
+    
+    /* byte aligned destination (long + 1) */
+    .align  2
+.loop_do1r:
+    mov.l   @r6,r1      /* load first long */
+    add     #-8,r6      /* decrement source addr */
+    mov.l   @(12,r6),r0 /* load second long */
+    cmp/hi  r3,r6       /* runs r6 down to first or second long bound */
+    mov.b   r0,@-r4     /* store low byte of second long */
+    shlr8   r0          /* get upper 3 bytes */
+    mov     r1,r2       /* copy first long */
+    shll16  r2          /* move low byte of first long all the way up, .. */
+    shll8   r2
+    or      r2,r0       /* ..combine with the 3 bytes of second long.. */
+    mov.l   r0,@-r4     /* ..and store as long */
+    shlr8   r1          /* get middle 2 bytes */
+    mov.w   r1,@-r4     /* store as word */
+    shlr16  r1          /* get upper byte */
+    mov.b   r1,@-r4     /* and store */
+    bt      .loop_do1r
+    
+    add     #-4,r3      /* readjust end address */
+.last_do13r:
+    cmp/hi  r3,r6       /* first long left? */
+    bf      .start_b2r  /* no, jump to trailing byte loop */
+
+    nop                 /* alignment */
+    mov.l   @(4,r6),r0  /* load first long */
+    add     #-4,r6      /* decrement source addr */
+    mov.b   r0,@-r4     /* store low byte */
+    shlr8   r0          /* get middle 2 bytes */
+    mov.w   r0,@-r4     /* store as word */
+    shlr16  r0          /* get upper byte */
+    bra     .start_b2r  /* jump to trailing byte loop */
+    mov.b   r0,@-r4     /* and store */
+
+    /* byte aligned destination (long + 3) */
+    .align  2
+.loop_do3r:
+    mov.l   @r6,r1      /* load first long */
+    add     #-8,r6      /* decrement source addr */
+    mov.l   @(12,r6),r0 /* load second long */
+    mov     r1,r2       /* copy first long */
+    mov.b   r0,@-r4     /* store low byte of second long */
+    shlr8   r0          /* get middle 2 bytes */
+    mov.w   r0,@-r4     /* store as word */
+    shlr16  r0          /* get upper byte */
+    shll8   r2          /* move lower 3 bytes of first long one up.. */
+    or      r2,r0       /* ..combine with the 1 byte of second long.. */
+    mov.l   r0,@-r4     /* ..and store as long */
+    shlr16  r1          /* get upper byte of first long */
+    shlr8   r1
+    cmp/hi  r3,r6       /* runs r6 down to first or second long bound */
+    mov.b   r1,@-r4     /* ..and store */
+    bt      .loop_do3r
+
+    bra     .last_do13r /* handle first longword: reuse routine for (long + 1) */
+    add     #-4,r3      /* readjust end address */
+
+    /* trailing byte loop: copies 0..3 bytes (or all for < 11 in total) */
+    .align  2
+.loop_b2r:
+    mov.b   @(7,r6),r0  /* load byte */
+    add     #-1,r6      /* decrement source addr */
+    mov.b   r0,@-r4     /* store byte */
+.start_b2r:
+    cmp/hi  r5,r6       /* runs r6 down to start address */
+    bt      .loop_b2r
+
+    rts
+    mov     r4,r0       /* return dest start address */
+.end:
+    .size   _memmove,.end-_memmove
diff --git a/firmware/target/sh/memset-sh.S b/firmware/target/sh/memset-sh.S
new file mode 100755
index 0000000000..9b96b93f27
--- /dev/null
+++ b/firmware/target/sh/memset-sh.S
@@ -0,0 +1,107 @@
+/***************************************************************************
+ *             __________               __   ___.
+ *   Open      \______   \ ____   ____ |  | _\_ |__   _______  ___
+ *   Source     |       _//  _ \_/ ___\|  |/ /| __ \ /  _ \  \/  /
+ *   Jukebox    |    |   (  <_> )  \___|    < | \_\ (  <_> > <  <
+ *   Firmware   |____|_  /\____/ \___  >__|_ \|___  /\____/__/\_ \
+ *                     \/            \/     \/    \/            \/
+ * $Id$
+ *
+ * Copyright (C) 2004 by Jens Arnold
+ *
+ * All files in this archive are subject to the GNU General Public License.
+ * See the file COPYING in the source tree root for full license agreement.
+ *
+ * This software is distributed on an "AS IS" basis, WITHOUT WARRANTY OF ANY
+ * KIND, either express or implied.
+ *
+ ****************************************************************************/
+#include "config.h"
+
+    .section    .icode,"ax",@progbits
+
+    .align      2
+    .global     _memset
+    .type       _memset,@function
+
+/* Fills a memory region with specified byte value
+ * This version is optimized for speed
+ *
+ * arguments:
+ *  r4 - start address
+ *  r5 - data
+ *  r6 - length
+ *
+ * return value:
+ *  r0 - start address (like ANSI version)
+ *
+ * register usage:
+ *  r0 - temporary
+ *  r1 - start address +11 for main loop
+ *  r4 - start address
+ *  r5 - data (spread to all 4 bytes when using long stores)
+ *  r6 - current address (runs down from end to start)
+ *
+ * The instruction order below is devised in a way to utilize the pipelining
+ * of the SH1 to the max. The routine fills memory from end to start in
+ * order to utilize the auto-decrementing store instructions.
+ */
+
+_memset:
+    neg     r4,r0
+    and     #3,r0       /* r0 = (4 - align_offset) % 4 */
+    add     #4,r0
+    cmp/hs  r0,r6       /* at least one aligned longword to fill? */
+    add     r4,r6       /* r6 = end_address */
+    bf      .no_longs   /* no, jump directly to byte loop */
+
+    extu.b  r5,r5       /* start: spread data to all 4 bytes */
+    swap.b  r5,r0
+    or      r0,r5       /* data now in 2 lower bytes of r5 */
+    swap.w  r5,r0
+    or      r0,r5       /* data now in all 4 bytes of r5 */
+    
+    mov     r6,r0
+    tst     #3,r0       /* r0 already long aligned? */
+    bt      .end_b1     /* yes: skip loop */
+
+    /* leading byte loop: sets 0..3 bytes */
+.loop_b1:
+    mov.b   r5,@-r0     /* store byte */
+    tst     #3,r0       /* r0 long aligned? */
+    bf      .loop_b1    /* runs r0 down until long aligned */
+    
+    mov     r0,r6       /* r6 = last long bound */
+    nop                 /* keep alignment */
+
+.end_b1:
+    mov     r4,r1       /* r1 = start_address... */
+    add     #11,r1      /* ... + 11, combined for rounding and offset */
+    xor     r1,r0
+    tst     #4,r0       /* bit 2 tells whether an even or odd number of */
+    bf      .loop_odd   /* longwords to set */
+
+    /* main loop: set 2 longs per pass */
+.loop_2l:
+    mov.l   r5,@-r6     /* store first long */
+.loop_odd:
+    cmp/hi  r1,r6       /* runs r6 down to first long bound */
+    mov.l   r5,@-r6     /* store second long */
+    bt      .loop_2l
+
+.no_longs:
+    cmp/hi  r4,r6       /* any bytes left? */
+    bf      .end_b2     /* no: skip loop */
+
+    /* trailing byte loop */
+.loop_b2:
+    mov.b   r5,@-r6     /* store byte */
+    cmp/hi  r4,r6       /* runs r6 down to the start address */
+    bt      .loop_b2
+
+.end_b2:
+    rts
+    mov     r4,r0       /* return start address */
+
+.end:
+    .size   _memset,.end-_memset
diff --git a/firmware/target/sh/strlen-sh.S b/firmware/target/sh/strlen-sh.S
new file mode 100755
index 0000000000..34837605ac
--- /dev/null
+++ b/firmware/target/sh/strlen-sh.S
@@ -0,0 +1,94 @@
+/***************************************************************************
+ *             __________               __   ___.
+ *   Open      \______   \ ____   ____ |  | _\_ |__   _______  ___
+ *   Source     |       _//  _ \_/ ___\|  |/ /| __ \ /  _ \  \/  /
+ *   Jukebox    |    |   (  <_> )  \___|    < | \_\ (  <_> > <  <
+ *   Firmware   |____|_  /\____/ \___  >__|_ \|___  /\____/__/\_ \
+ *                     \/            \/     \/    \/            \/
+ * $Id$
+ *
+ * Copyright (C) 2005 by Jens Arnold
+ *
+ * All files in this archive are subject to the GNU General Public License.
+ * See the file COPYING in the source tree root for full license agreement.
+ *
+ * This software is distributed on an "AS IS" basis, WITHOUT WARRANTY OF ANY
+ * KIND, either express or implied.
+ *
+ ****************************************************************************/
+#include "config.h"
+
+    .section    .icode,"ax",@progbits
+
+    .align      2
+    .global     _strlen
+    .type       _strlen,@function
+
+/* Works out the length of a string
+ * This version is optimized for speed
+ *
+ * arguments:
+ *  r4 - start address
+ *
+ * return value:
+ *  r0 - string length
+ *
+ * register usage:
+ *  r0 - current address
+ *  r1 - current value (byte/long)
+ *  r2 - mask for alignment / zero (for cmp/str)
+ *  r4 - start address
+ *
+ */
+
+_strlen:
+    mov     r4,r0       /* r0 = start address */
+    tst     #3,r0       /* long aligned? */
+    bt      .start_l    /* yes, jump directly to the longword loop */
+
+    /* not long aligned: check the first 3 bytes */
+    mov.b   @r0+,r1     /* fetch first byte */
+    tst     r1,r1       /* byte == 0 ? */
+    bt      .hitzero    /* yes, string end found */
+    mov.b   @r0+,r1     /* fetch second byte */
+    mov     #3,r2       /* prepare mask: r2 = 0..00000011b */
+    tst     r1,r1       /* byte == 0 ? */
+    bt      .hitzero    /* yes, string end found */
+    mov.b   @r0+,r1     /* fetch third byte */
+    not     r2,r2       /* prepare mask: r2 = 1..11111100b */
+    tst     r1,r1       /* byte == 0 ? */
+    bt      .hitzero    /* yes, string end found */
+
+    /* not yet found, fall through into longword loop */
+    and     r2,r0       /* align down to long bound */
+
+    /* main loop: check longwords */
+.start_l:
+    mov     #0,r2       /* zero longword for cmp/str */
+.loop_l:
+    mov.l   @r0+,r1     /* fetch long word */
+    cmp/str r1,r2       /* any zero byte within? */
+    bf      .loop_l     /* no, loop */
+    add     #-4,r0      /* set address back to start of this longword */
+
+    /* the last longword contains the string end: figure out the byte */
+    mov.b   @r0+,r1     /* fetch first byte */    
+    tst     r1,r1       /* byte == 0 ? */
+    bt      .hitzero    /* yes, string end found */
+    mov.b   @r0+,r1     /* fetch second byte */
+    tst     r1,r1       /* byte == 0 ? */
+    bt      .hitzero    /* yes, string end found */
+    mov.b   @r0+,r1     /* fetch third byte */
+    tst     r1,r1       /* byte == 0 ? */
+    bt      .hitzero    /* yes, string end found */
+    rts                 /* must be the fourth byte */
+    sub     r4,r0       /* len = string_end - string_start */
+
+.hitzero:
+    add     #-1,r0      /* undo address increment */
+    rts
+    sub     r4,r0       /* len = string_end - string_start */
+
+.end:
+    .size   _strlen,.end-_strlen
+