mirror of
https://github.com/AuxXxilium/linux_dsm_epyc7002.git
synced 2024-12-05 12:46:39 +07:00
be0e1e788b
Reduce the number of variables modified by the loop in do_csum() by 1, which seems like a good idea. On Nios II (a RISC CPU with 3-operand instruction set) it reduces the loop from 7 to 6 instructions, including the conditional branch. Signed-off-by: Ian Abbott <abbotti@mev.co.uk> Signed-off-by: David S. Miller <davem@davemloft.net>
201 lines
4.8 KiB
C
201 lines
4.8 KiB
C
/*
|
|
*
|
|
* INET An implementation of the TCP/IP protocol suite for the LINUX
|
|
* operating system. INET is implemented using the BSD Socket
|
|
* interface as the means of communication with the user level.
|
|
*
|
|
* IP/TCP/UDP checksumming routines
|
|
*
|
|
* Authors: Jorge Cwik, <jorge@laser.satlink.net>
|
|
* Arnt Gulbrandsen, <agulbra@nvg.unit.no>
|
|
* Tom May, <ftom@netcom.com>
|
|
* Andreas Schwab, <schwab@issan.informatik.uni-dortmund.de>
|
|
* Lots of code moved from tcp.c and ip.c; see those files
|
|
* for more names.
|
|
*
|
|
* 03/02/96 Jes Sorensen, Andreas Schwab, Roman Hodek:
|
|
* Fixed some nasty bugs, causing some horrible crashes.
|
|
* A: At some points, the sum (%0) was used as
|
|
* length-counter instead of the length counter
|
|
* (%1). Thanks to Roman Hodek for pointing this out.
|
|
* B: GCC seems to mess up if one uses too many
|
|
* data-registers to hold input values and one tries to
|
|
* specify d0 and d1 as scratch registers. Letting gcc
|
|
* choose these registers itself solves the problem.
|
|
*
|
|
* This program is free software; you can redistribute it and/or
|
|
* modify it under the terms of the GNU General Public License
|
|
* as published by the Free Software Foundation; either version
|
|
* 2 of the License, or (at your option) any later version.
|
|
*/
|
|
|
|
/* Revised by Kenneth Albanowski for m68knommu. Basic problem: unaligned access
|
|
kills, so most of the assembly has to go. */
|
|
|
|
#include <linux/module.h>
|
|
#include <net/checksum.h>
|
|
|
|
#include <asm/byteorder.h>
|
|
|
|
#ifndef do_csum
|
|
static inline unsigned short from32to16(unsigned int x)
|
|
{
|
|
/* add up 16-bit and 16-bit for 16+c bit */
|
|
x = (x & 0xffff) + (x >> 16);
|
|
/* add up carry.. */
|
|
x = (x & 0xffff) + (x >> 16);
|
|
return x;
|
|
}
|
|
|
|
static unsigned int do_csum(const unsigned char *buff, int len)
|
|
{
|
|
int odd;
|
|
unsigned int result = 0;
|
|
|
|
if (len <= 0)
|
|
goto out;
|
|
odd = 1 & (unsigned long) buff;
|
|
if (odd) {
|
|
#ifdef __LITTLE_ENDIAN
|
|
result += (*buff << 8);
|
|
#else
|
|
result = *buff;
|
|
#endif
|
|
len--;
|
|
buff++;
|
|
}
|
|
if (len >= 2) {
|
|
if (2 & (unsigned long) buff) {
|
|
result += *(unsigned short *) buff;
|
|
len -= 2;
|
|
buff += 2;
|
|
}
|
|
if (len >= 4) {
|
|
const unsigned char *end = buff + ((unsigned)len & ~3);
|
|
unsigned int carry = 0;
|
|
do {
|
|
unsigned int w = *(unsigned int *) buff;
|
|
buff += 4;
|
|
result += carry;
|
|
result += w;
|
|
carry = (w > result);
|
|
} while (buff < end);
|
|
result += carry;
|
|
result = (result & 0xffff) + (result >> 16);
|
|
}
|
|
if (len & 2) {
|
|
result += *(unsigned short *) buff;
|
|
buff += 2;
|
|
}
|
|
}
|
|
if (len & 1)
|
|
#ifdef __LITTLE_ENDIAN
|
|
result += *buff;
|
|
#else
|
|
result += (*buff << 8);
|
|
#endif
|
|
result = from32to16(result);
|
|
if (odd)
|
|
result = ((result >> 8) & 0xff) | ((result & 0xff) << 8);
|
|
out:
|
|
return result;
|
|
}
|
|
#endif
|
|
|
|
/*
|
|
* This is a version of ip_compute_csum() optimized for IP headers,
|
|
* which always checksum on 4 octet boundaries.
|
|
*/
|
|
__sum16 ip_fast_csum(const void *iph, unsigned int ihl)
|
|
{
|
|
return (__force __sum16)~do_csum(iph, ihl*4);
|
|
}
|
|
EXPORT_SYMBOL(ip_fast_csum);
|
|
|
|
/*
|
|
* computes the checksum of a memory block at buff, length len,
|
|
* and adds in "sum" (32-bit)
|
|
*
|
|
* returns a 32-bit number suitable for feeding into itself
|
|
* or csum_tcpudp_magic
|
|
*
|
|
* this function must be called with even lengths, except
|
|
* for the last fragment, which may be odd
|
|
*
|
|
* it's best to have buff aligned on a 32-bit boundary
|
|
*/
|
|
__wsum csum_partial(const void *buff, int len, __wsum wsum)
|
|
{
|
|
unsigned int sum = (__force unsigned int)wsum;
|
|
unsigned int result = do_csum(buff, len);
|
|
|
|
/* add in old sum, and carry.. */
|
|
result += sum;
|
|
if (sum > result)
|
|
result += 1;
|
|
return (__force __wsum)result;
|
|
}
|
|
EXPORT_SYMBOL(csum_partial);
|
|
|
|
/*
|
|
* this routine is used for miscellaneous IP-like checksums, mainly
|
|
* in icmp.c
|
|
*/
|
|
__sum16 ip_compute_csum(const void *buff, int len)
|
|
{
|
|
return (__force __sum16)~do_csum(buff, len);
|
|
}
|
|
EXPORT_SYMBOL(ip_compute_csum);
|
|
|
|
/*
|
|
* copy from fs while checksumming, otherwise like csum_partial
|
|
*/
|
|
__wsum
|
|
csum_partial_copy_from_user(const void __user *src, void *dst, int len,
|
|
__wsum sum, int *csum_err)
|
|
{
|
|
int missing;
|
|
|
|
missing = __copy_from_user(dst, src, len);
|
|
if (missing) {
|
|
memset(dst + len - missing, 0, missing);
|
|
*csum_err = -EFAULT;
|
|
} else
|
|
*csum_err = 0;
|
|
|
|
return csum_partial(dst, len, sum);
|
|
}
|
|
EXPORT_SYMBOL(csum_partial_copy_from_user);
|
|
|
|
/*
|
|
* copy from ds while checksumming, otherwise like csum_partial
|
|
*/
|
|
__wsum
|
|
csum_partial_copy(const void *src, void *dst, int len, __wsum sum)
|
|
{
|
|
memcpy(dst, src, len);
|
|
return csum_partial(dst, len, sum);
|
|
}
|
|
EXPORT_SYMBOL(csum_partial_copy);
|
|
|
|
#ifndef csum_tcpudp_nofold
|
|
__wsum csum_tcpudp_nofold(__be32 saddr, __be32 daddr,
|
|
unsigned short len,
|
|
unsigned short proto,
|
|
__wsum sum)
|
|
{
|
|
unsigned long long s = (__force u32)sum;
|
|
|
|
s += (__force u32)saddr;
|
|
s += (__force u32)daddr;
|
|
#ifdef __BIG_ENDIAN
|
|
s += proto + len;
|
|
#else
|
|
s += (proto + len) << 8;
|
|
#endif
|
|
s += (s >> 32);
|
|
return (__force __wsum)s;
|
|
}
|
|
EXPORT_SYMBOL(csum_tcpudp_nofold);
|
|
#endif
|