linux-mips
[Top] [All Lists]

[PATCH 2/3] Optimize flow of csum_partial

To: linux-mips@linux-mips.org
Subject: [PATCH 2/3] Optimize flow of csum_partial
From: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
Date: Fri, 08 Dec 2006 01:04:45 +0900 (JST)
Cc: ralf@linux-mips.org
Original-recipient: rfc822;linux-mips@linux-mips.org
Sender: linux-mips-bounce@linux-mips.org
Delete dead codes at end of the function and move small_csumcopy
there.  This makes some labels (maybe_end_cruft, small_memcpy,
end_bytes, out) needless and eliminates some branches.

Signed-off-by: Atsushi Nemoto <anemo@mba.ocn.ne.jp>
---
diff --git a/arch/mips/lib/csum_partial.S b/arch/mips/lib/csum_partial.S
index 3bffdbb..b04475d 100644
--- a/arch/mips/lib/csum_partial.S
+++ b/arch/mips/lib/csum_partial.S
@@ -65,64 +65,6 @@ #define sum v0
 
        .text
        .set    noreorder
-
-/* unknown src alignment and < 8 bytes to go  */
-small_csumcpy:
-       move    a1, t2
-
-       andi    t0, a1, 4
-       beqz    t0, 1f
-        andi   t0, a1, 2
-
-       /* Still a full word to go  */
-       ulw     t1, (src)
-       PTR_ADDIU       src, 4
-       ADDC(sum, t1)
-
-1:     move    t1, zero
-       beqz    t0, 1f
-        andi   t0, a1, 1
-
-       /* Still a halfword to go  */
-       ulhu    t1, (src)
-       PTR_ADDIU       src, 2
-
-1:     beqz    t0, 1f
-        sll    t1, t1, 16
-
-       lbu     t2, (src)
-        nop
-
-#ifdef __MIPSEB__
-       sll     t2, t2, 8
-#endif
-       or      t1, t2
-
-1:     ADDC(sum, t1)
-
-       /* fold checksum */
-       sll     v1, sum, 16
-       addu    sum, v1
-       sltu    v1, sum, v1
-       srl     sum, sum, 16
-       addu    sum, v1
-
-       /* odd buffer alignment? */
-       beqz    t7, 1f
-        nop
-       sll     v1, sum, 8
-       srl     sum, sum, 8
-       or      sum, v1
-       andi    sum, 0xffff
-1:
-       .set    reorder
-       /* Add the passed partial csum.  */
-       ADDC(sum, a2)
-       jr      ra
-       .set    noreorder
-
-/* ------------------------------------------------------------------------- */
-
        .align  5
 LEAF(csum_partial)
        move    sum, zero
@@ -132,8 +74,7 @@ LEAF(csum_partial)
        bnez    t8, small_csumcpy               /* < 8 bytes to copy */
         move   t2, a1
 
-       beqz    a1, out
-        andi   t7, src, 0x1                    /* odd buffer? */
+       andi    t7, src, 0x1                    /* odd buffer? */
 
 hword_align:
        beqz    t7, word_align
@@ -232,8 +173,9 @@ move_32bytes:
        PTR_ADDU        src, src, 0x20
 
 do_end_words:
-       beqz    t8, maybe_end_cruft
-        LONG_SRL       t8, t8, 0x2
+       beqz    t8, small_csumcpy
+        andi   t2, a1, 0x3
+       LONG_SRL        t8, t8, 0x2
 
 end_words:
        lw      t0, (src)
@@ -242,21 +184,58 @@ end_words:
        bnez    t8, end_words
         PTR_ADDU       src, src, 0x4
 
-maybe_end_cruft:
-       andi    t2, a1, 0x3
+/* unknown src alignment and < 8 bytes to go  */
+small_csumcpy:
+       move    a1, t2
 
-small_memcpy:
- j small_csumcpy; move a1, t2          /* XXX ??? */
-       beqz    t2, out
-        move   a1, t2
+       andi    t0, a1, 4
+       beqz    t0, 1f
+        andi   t0, a1, 2
 
-end_bytes:
-       lb      t0, (src)
-       LONG_SUBU       a1, a1, 0x1
-       bnez    a2, end_bytes
-        PTR_ADDU       src, src, 0x1
+       /* Still a full word to go  */
+       ulw     t1, (src)
+       PTR_ADDIU       src, 4
+       ADDC(sum, t1)
+
+1:     move    t1, zero
+       beqz    t0, 1f
+        andi   t0, a1, 1
+
+       /* Still a halfword to go  */
+       ulhu    t1, (src)
+       PTR_ADDIU       src, 2
+
+1:     beqz    t0, 1f
+        sll    t1, t1, 16
+
+       lbu     t2, (src)
+        nop
+
+#ifdef __MIPSEB__
+       sll     t2, t2, 8
+#endif
+       or      t1, t2
+
+1:     ADDC(sum, t1)
 
-out:
+       /* fold checksum */
+       sll     v1, sum, 16
+       addu    sum, v1
+       sltu    v1, sum, v1
+       srl     sum, sum, 16
+       addu    sum, v1
+
+       /* odd buffer alignment? */
+       beqz    t7, 1f
+        nop
+       sll     v1, sum, 8
+       srl     sum, sum, 8
+       or      sum, v1
+       andi    sum, 0xffff
+1:
+       .set    reorder
+       /* Add the passed partial csum.  */
+       ADDC(sum, a2)
        jr      ra
-        move   v0, sum
+       .set    noreorder
        END(csum_partial)

<Prev in Thread] Current Thread [Next in Thread>
  • [PATCH 2/3] Optimize flow of csum_partial, Atsushi Nemoto <=