3 files changed, 82 insertions, 45 deletions

diff --git a/arch/ia64/lib/checksum.c b/arch/ia64/lib/checksum.c
index beb11721d9f5..4411d9baeb21 100644
--- a/arch/ia64/lib/checksum.c
+++ b/arch/ia64/lib/checksum.c
@@ -33,32 +33,32 @@ from64to16 (unsigned long x)
  * computes the checksum of the TCP/UDP pseudo-header
  * returns a 16-bit checksum, already complemented.
  */
-unsigned short int
-csum_tcpudp_magic (unsigned long saddr, unsigned long daddr, unsigned short len,
-                   unsigned short proto, unsigned int sum)
+__sum16
+csum_tcpudp_magic (__be32 saddr, __be32 daddr, unsigned short len,
+                   unsigned short proto, __wsum sum)
 {
-        return ~from64to16(saddr + daddr + sum + ((unsigned long) ntohs(len) << 16) +
-                           ((unsigned long) proto << 8));
+        return (__force __sum16)~from64to16(
+                (__force u64)saddr + (__force u64)daddr +
+                (__force u64)sum + ((len + proto) << 8));
 }
 
 EXPORT_SYMBOL(csum_tcpudp_magic);
 
-unsigned int
-csum_tcpudp_nofold (unsigned long saddr, unsigned long daddr, unsigned short len,
-                    unsigned short proto, unsigned int sum)
+__wsum
+csum_tcpudp_nofold (__be32 saddr, __be32 daddr, unsigned short len,
+                    unsigned short proto, __wsum sum)
 {
         unsigned long result;
 
-        result = (saddr + daddr + sum +
-                  ((unsigned long) ntohs(len) << 16) +
-                  ((unsigned long) proto << 8));
+        result = (__force u64)saddr + (__force u64)daddr +
+                 (__force u64)sum + ((len + proto) << 8);
 
         /* Fold down to 32-bits so we don't lose in the typedef-less network stack.  */
         /* 64 to 33 */
         result = (result & 0xffffffff) + (result >> 32);
         /* 33 to 32 */
         result = (result & 0xffffffff) + (result >> 32);
-        return result;
+        return (__force __wsum)result;
 }
 
 extern unsigned long do_csum (const unsigned char *, long);
@@ -75,16 +75,15 @@ extern unsigned long do_csum (const unsigned char *, long);
  *
  * it's best to have buff aligned on a 32-bit boundary
  */
-unsigned int
-csum_partial (const unsigned char * buff, int len, unsigned int sum)
+__wsum csum_partial(const void *buff, int len, __wsum sum)
 {
-        unsigned long result = do_csum(buff, len);
+        u64 result = do_csum(buff, len);
 
         /* add in old sum, and carry.. */
-        result += sum;
+        result += (__force u32)sum;
         /* 32+c bits -> 32 bits */
         result = (result & 0xffffffff) + (result >> 32);
-        return result;
+        return (__force __wsum)result;
 }
 
 EXPORT_SYMBOL(csum_partial);
@@ -93,10 +92,9 @@ EXPORT_SYMBOL(csum_partial);
  * this routine is used for miscellaneous IP-like checksums, mainly
  * in icmp.c
  */
-unsigned short
-ip_compute_csum (unsigned char * buff, int len)
+__sum16 ip_compute_csum (const void *buff, int len)
 {
-        return ~do_csum(buff,len);
+        return (__force __sum16)~do_csum(buff,len);
 }
 
 EXPORT_SYMBOL(ip_compute_csum);
diff --git a/arch/ia64/lib/csum_partial_copy.c b/arch/ia64/lib/csum_partial_copy.c
index 36866e8a5d2b..503dfe6d1450 100644
--- a/arch/ia64/lib/csum_partial_copy.c
+++ b/arch/ia64/lib/csum_partial_copy.c
@@ -104,9 +104,9 @@ out:
  */
 extern unsigned long do_csum(const unsigned char *, long);
 
-static unsigned int
-do_csum_partial_copy_from_user (const unsigned char __user *src, unsigned char *dst,
-                                int len, unsigned int psum, int *errp)
+__wsum
+csum_partial_copy_from_user(const void __user *src, void *dst,
+                                int len, __wsum psum, int *errp)
 {
         unsigned long result;
 
@@ -122,30 +122,17 @@ do_csum_partial_copy_from_user (const unsigned char __user *src, unsigned char *
         result = do_csum(dst, len);
 
         /* add in old sum, and carry.. */
-        result += psum;
+        result += (__force u32)psum;
         /* 32+c bits -> 32 bits */
         result = (result & 0xffffffff) + (result >> 32);
-        return result;
-}
-
-unsigned int
-csum_partial_copy_from_user (const unsigned char __user *src, unsigned char *dst,
-                             int len, unsigned int sum, int *errp)
-{
-        if (!access_ok(VERIFY_READ, src, len)) {
-                *errp = -EFAULT;
-                memset(dst, 0, len);
-                return sum;
-        }
-
-        return do_csum_partial_copy_from_user(src, dst, len, sum, errp);
+        return (__force __wsum)result;
 }
 
-unsigned int
-csum_partial_copy_nocheck(const unsigned char __user *src, unsigned char *dst,
-                          int len, unsigned int sum)
+__wsum
+csum_partial_copy_nocheck(const void *src, void *dst, int len, __wsum sum)
 {
-        return do_csum_partial_copy_from_user(src, dst, len, sum, NULL);
+        return csum_partial_copy_from_user((__force const void __user *)src,
+                                           dst, len, sum, NULL);
 }
 
 EXPORT_SYMBOL(csum_partial_copy_nocheck);
diff --git a/arch/ia64/lib/ip_fast_csum.S b/arch/ia64/lib/ip_fast_csum.S
index 19674ca2acfc..1f86aeb2c948 100644
--- a/arch/ia64/lib/ip_fast_csum.S
+++ b/arch/ia64/lib/ip_fast_csum.S
@@ -8,8 +8,8 @@
  *      in0: address of buffer to checksum (char *)
  *      in1: length of the buffer (int)
  *
- * Copyright (C) 2002 Intel Corp.
- * Copyright (C) 2002 Ken Chen <kenneth.w.chen@intel.com>
+ * Copyright (C) 2002, 2006 Intel Corp.
+ * Copyright (C) 2002, 2006 Ken Chen <kenneth.w.chen@intel.com>
  */
 
 #include <asm/asmmacro.h>
@@ -25,6 +25,9 @@
 
 #define in0     r32
 #define in1     r33
+#define in2     r34
+#define in3     r35
+#define in4     r36
 #define ret0    r8
 
 GLOBAL_ENTRY(ip_fast_csum)
@@ -65,8 +68,9 @@ GLOBAL_ENTRY(ip_fast_csum)
         zxt2    r20=r20
         ;;
         add     r20=ret0,r20
+        mov     r9=0xffff
         ;;
-        andcm   ret0=-1,r20
+        andcm   ret0=r9,r20
         .restore sp             // reset frame state
         br.ret.sptk.many b0
         ;;
@@ -88,3 +92,51 @@ GLOBAL_ENTRY(ip_fast_csum)
         mov     b0=r34
         br.ret.sptk.many b0
 END(ip_fast_csum)
+
+GLOBAL_ENTRY(csum_ipv6_magic)
+        ld4     r20=[in0],4
+        ld4     r21=[in1],4
+        dep     r15=in3,in2,32,16
+        ;;
+        ld4     r22=[in0],4
+        ld4     r23=[in1],4
+        mux1    r15=r15,@rev
+        ;;
+        ld4     r24=[in0],4
+        ld4     r25=[in1],4
+        shr.u   r15=r15,16
+        add     r16=r20,r21
+        add     r17=r22,r23
+        ;;
+        ld4     r26=[in0],4
+        ld4     r27=[in1],4
+        add     r18=r24,r25
+        add     r8=r16,r17
+        ;;
+        add     r19=r26,r27
+        add     r8=r8,r18
+        ;;
+        add     r8=r8,r19
+        add     r15=r15,in4
+        ;;
+        add     r8=r8,r15
+        ;;
+        shr.u   r10=r8,32       // now fold sum into short
+        zxt4    r11=r8
+        ;;
+        add     r8=r10,r11
+        ;;
+        shr.u   r10=r8,16       // yeah, keep it rolling
+        zxt2    r11=r8
+        ;;
+        add     r8=r10,r11
+        ;;
+        shr.u   r10=r8,16       // three times lucky
+        zxt2    r11=r8
+        ;;
+        add     r8=r10,r11
+        mov     r9=0xffff
+        ;;
+        andcm   r8=r9,r8
+        br.ret.sptk.many b0
+END(csum_ipv6_magic)