From: Greg Ungerer <gerg@snapgear.com>
To: Geert Uytterhoeven <geert@linux-m68k.org>
Cc: linux-m68k@vger.kernel.org, uclinux-dev@uclinux.org
Subject: Re: [PATCH] m68k: simpler m68k and ColdFire CPU's can use generic csum code
Date: Fri, 25 Nov 2011 09:36:38 +1000 [thread overview]
Message-ID: <4ECED506.6030201@snapgear.com> (raw)
In-Reply-To: <1321597970-23806-1-git-send-email-gerg@snapgear.com>
Hi Geert,
On 18/11/11 16:32, gerg@snapgear.com wrote:
> From: Greg Ungerer<gerg@uclinux.org>
>
> We have two implementations of the IP checksuming code for the m68k arch.
> One uses the more advanced instructions available in 68020 and above
> processors, the other uses the simpler instructions available on the
> original 68000 processors and the modern ColdFire processors.
>
> This simpler code is pretty much the same as the generic lib implementation
> of the IP csum functions. So lets just switch over to using that. That
> means we can completely remove the checksum_no.c file, and only have the
> local fast code used for the more complex 68k CPU family members.
>
> Signed-off-by: Greg Ungerer<gerg@uclinux.org>
> Acked-by: Geert Uytterhoeven<geert@linux-m68k.org>
This is a reworked version of the original patch
(http://www.spinics.net/lists/linux-m68k/msg04271.html), so I probably
should have removed your "Acked-by" here, since you haven't really
acked this version of it at all :-) Sorry about that.
This version goes a bit further than the original, and actually
switches to use the generic csum for the non-mmu and ColdFire cases.
Regards
Greg
> ---
> arch/m68k/Kconfig | 3 +
> arch/m68k/Kconfig.cpu | 2 +
> arch/m68k/include/asm/checksum.h | 31 +---
> arch/m68k/lib/Makefile | 8 +-
> arch/m68k/lib/checksum.c | 425 ++++++++++++++++++++++++++++++++++++++
> arch/m68k/lib/checksum_mm.c | 425 --------------------------------------
> arch/m68k/lib/checksum_no.c | 156 --------------
> 7 files changed, 441 insertions(+), 609 deletions(-)
> create mode 100644 arch/m68k/lib/checksum.c
> delete mode 100644 arch/m68k/lib/checksum_mm.c
> delete mode 100644 arch/m68k/lib/checksum_no.c
>
> diff --git a/arch/m68k/Kconfig b/arch/m68k/Kconfig
> index 2914b03..2fe2d63 100644
> --- a/arch/m68k/Kconfig
> +++ b/arch/m68k/Kconfig
> @@ -40,6 +40,9 @@ config GENERIC_CALIBRATE_DELAY
> config GENERIC_IOMAP
> def_bool MMU
>
> +config GENERIC_CSUM
> + bool
> +
> config TIME_LOW_RES
> bool
> default y
> diff --git a/arch/m68k/Kconfig.cpu b/arch/m68k/Kconfig.cpu
> index 9af9e68..17d37ed 100644
> --- a/arch/m68k/Kconfig.cpu
> +++ b/arch/m68k/Kconfig.cpu
> @@ -3,6 +3,7 @@ comment "Processor Type"
> config M68000
> bool
> select CPU_HAS_NO_BITFIELDS
> + select GENERIC_CSUM
> help
> The Freescale (was Motorola) 68000 CPU is the first generation of
> the well known M68K family of processors. The CPU core as well as
> @@ -23,6 +24,7 @@ config COLDFIRE
> select GENERIC_GPIO
> select ARCH_REQUIRE_GPIOLIB
> select CPU_HAS_NO_BITFIELDS
> + select GENERIC_CSUM
> help
> The Freescale ColdFire family of processors is a modern derivitive
> of the 68000 processor family. They are mainly targeted at embedded
> diff --git a/arch/m68k/include/asm/checksum.h b/arch/m68k/include/asm/checksum.h
> index ec51448..2f88d86 100644
> --- a/arch/m68k/include/asm/checksum.h
> +++ b/arch/m68k/include/asm/checksum.h
> @@ -3,6 +3,10 @@
>
> #include<linux/in6.h>
>
> +#ifdef CONFIG_GENERIC_CSUM
> +#include<asm-generic/checksum.h>
> +#else
> +
> /*
> * computes the checksum of a memory block at buff, length len,
> * and adds in "sum" (32-bit)
> @@ -34,30 +38,6 @@ extern __wsum csum_partial_copy_nocheck(const void *src,
> void *dst, int len,
> __wsum sum);
>
> -
> -#ifdef CONFIG_COLDFIRE
> -
> -/*
> - * The ColdFire cores don't support all the 68k instructions used
> - * in the optimized checksum code below. So it reverts back to using
> - * more standard C coded checksums. The fast checksum code is
> - * significantly larger than the optimized version, so it is not
> - * inlined here.
> - */
> -__sum16 ip_fast_csum(const void *iph, unsigned int ihl);
> -
> -static inline __sum16 csum_fold(__wsum sum)
> -{
> - unsigned int tmp = (__force u32)sum;
> -
> - tmp = (tmp& 0xffff) + (tmp>> 16);
> - tmp = (tmp& 0xffff) + (tmp>> 16);
> -
> - return (__force __sum16)~tmp;
> -}
> -
> -#else
> -
> /*
> * This is a version of ip_fast_csum() optimized for IP headers,
> * which always checksum on 4 octet boundaries.
> @@ -97,8 +77,6 @@ static inline __sum16 csum_fold(__wsum sum)
> return (__force __sum16)~sum;
> }
>
> -#endif /* CONFIG_COLDFIRE */
> -
> static inline __wsum
> csum_tcpudp_nofold(__be32 saddr, __be32 daddr, unsigned short len,
> unsigned short proto, __wsum sum)
> @@ -167,4 +145,5 @@ csum_ipv6_magic(const struct in6_addr *saddr, const struct in6_addr *daddr,
> return csum_fold(sum);
> }
>
> +#endif /* CONFIG_GENERIC_CSUM */
> #endif /* _M68K_CHECKSUM_H */
> diff --git a/arch/m68k/lib/Makefile b/arch/m68k/lib/Makefile
> index 1a1bd90..b3b40e4 100644
> --- a/arch/m68k/lib/Makefile
> +++ b/arch/m68k/lib/Makefile
> @@ -7,8 +7,12 @@ lib-y := ashldi3.o ashrdi3.o lshrdi3.o muldi3.o \
> memcpy.o memset.o memmove.o
>
> ifdef CONFIG_MMU
> -lib-y += string.o uaccess.o checksum_mm.o
> +lib-y += string.o uaccess.o
> else
> -lib-y += mulsi3.o divsi3.o udivsi3.o modsi3.o umodsi3.o checksum_no.o
> +lib-y += mulsi3.o divsi3.o udivsi3.o modsi3.o umodsi3.o
> +endif
> +
> +ifndef CONFIG_GENERIC_CSUM
> +lib-y += checksum.o
> endif
>
> diff --git a/arch/m68k/lib/checksum.c b/arch/m68k/lib/checksum.c
> new file mode 100644
> index 0000000..6216f12
> --- /dev/null
> +++ b/arch/m68k/lib/checksum.c
> @@ -0,0 +1,425 @@
> +/*
> + * 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.
> + *
> + * 1998/8/31 Andreas Schwab:
> + * Zero out rest of buffer on exception in
> + * csum_partial_copy_from_user.
> + */
> +
> +#include<linux/module.h>
> +#include<net/checksum.h>
> +
> +/*
> + * computes a partial checksum, e.g. for TCP/UDP fragments
> + */
> +
> +__wsum csum_partial(const void *buff, int len, __wsum sum)
> +{
> + unsigned long tmp1, tmp2;
> + /*
> + * Experiments with ethernet and slip connections show that buff
> + * is aligned on either a 2-byte or 4-byte boundary.
> + */
> + __asm__("movel %2,%3\n\t"
> + "btst #1,%3\n\t" /* Check alignment */
> + "jeq 2f\n\t"
> + "subql #2,%1\n\t" /* buff%4==2: treat first word */
> + "jgt 1f\n\t"
> + "addql #2,%1\n\t" /* len was == 2, treat only rest */
> + "jra 4f\n"
> + "1:\t"
> + "addw %2@+,%0\n\t" /* add first word to sum */
> + "clrl %3\n\t"
> + "addxl %3,%0\n" /* add X bit */
> + "2:\t"
> + /* unrolled loop for the main part: do 8 longs at once */
> + "movel %1,%3\n\t" /* save len in tmp1 */
> + "lsrl #5,%1\n\t" /* len/32 */
> + "jeq 2f\n\t" /* not enough... */
> + "subql #1,%1\n"
> + "1:\t"
> + "movel %2@+,%4\n\t"
> + "addxl %4,%0\n\t"
> + "movel %2@+,%4\n\t"
> + "addxl %4,%0\n\t"
> + "movel %2@+,%4\n\t"
> + "addxl %4,%0\n\t"
> + "movel %2@+,%4\n\t"
> + "addxl %4,%0\n\t"
> + "movel %2@+,%4\n\t"
> + "addxl %4,%0\n\t"
> + "movel %2@+,%4\n\t"
> + "addxl %4,%0\n\t"
> + "movel %2@+,%4\n\t"
> + "addxl %4,%0\n\t"
> + "movel %2@+,%4\n\t"
> + "addxl %4,%0\n\t"
> + "dbra %1,1b\n\t"
> + "clrl %4\n\t"
> + "addxl %4,%0\n\t" /* add X bit */
> + "clrw %1\n\t"
> + "subql #1,%1\n\t"
> + "jcc 1b\n"
> + "2:\t"
> + "movel %3,%1\n\t" /* restore len from tmp1 */
> + "andw #0x1c,%3\n\t" /* number of rest longs */
> + "jeq 4f\n\t"
> + "lsrw #2,%3\n\t"
> + "subqw #1,%3\n"
> + "3:\t"
> + /* loop for rest longs */
> + "movel %2@+,%4\n\t"
> + "addxl %4,%0\n\t"
> + "dbra %3,3b\n\t"
> + "clrl %4\n\t"
> + "addxl %4,%0\n" /* add X bit */
> + "4:\t"
> + /* now check for rest bytes that do not fit into longs */
> + "andw #3,%1\n\t"
> + "jeq 7f\n\t"
> + "clrl %4\n\t" /* clear tmp2 for rest bytes */
> + "subqw #2,%1\n\t"
> + "jlt 5f\n\t"
> + "movew %2@+,%4\n\t" /* have rest>= 2: get word */
> + "swap %4\n\t" /* into bits 16..31 */
> + "tstw %1\n\t" /* another byte? */
> + "jeq 6f\n"
> + "5:\t"
> + "moveb %2@,%4\n\t" /* have odd rest: get byte */
> + "lslw #8,%4\n\t" /* into bits 8..15; 16..31 untouched */
> + "6:\t"
> + "addl %4,%0\n\t" /* now add rest long to sum */
> + "clrl %4\n\t"
> + "addxl %4,%0\n" /* add X bit */
> + "7:\t"
> + : "=d" (sum), "=d" (len), "=a" (buff),
> + "=&d" (tmp1), "=&d" (tmp2)
> + : "0" (sum), "1" (len), "2" (buff)
> + );
> + return(sum);
> +}
> +
> +EXPORT_SYMBOL(csum_partial);
> +
> +
> +/*
> + * copy from user space while checksumming, with exception handling.
> + */
> +
> +__wsum
> +csum_partial_copy_from_user(const void __user *src, void *dst,
> + int len, __wsum sum, int *csum_err)
> +{
> + /*
> + * GCC doesn't like more than 10 operands for the asm
> + * statements so we have to use tmp2 for the error
> + * code.
> + */
> + unsigned long tmp1, tmp2;
> +
> + __asm__("movel %2,%4\n\t"
> + "btst #1,%4\n\t" /* Check alignment */
> + "jeq 2f\n\t"
> + "subql #2,%1\n\t" /* buff%4==2: treat first word */
> + "jgt 1f\n\t"
> + "addql #2,%1\n\t" /* len was == 2, treat only rest */
> + "jra 4f\n"
> + "1:\n"
> + "10:\t"
> + "movesw %2@+,%4\n\t" /* add first word to sum */
> + "addw %4,%0\n\t"
> + "movew %4,%3@+\n\t"
> + "clrl %4\n\t"
> + "addxl %4,%0\n" /* add X bit */
> + "2:\t"
> + /* unrolled loop for the main part: do 8 longs at once */
> + "movel %1,%4\n\t" /* save len in tmp1 */
> + "lsrl #5,%1\n\t" /* len/32 */
> + "jeq 2f\n\t" /* not enough... */
> + "subql #1,%1\n"
> + "1:\n"
> + "11:\t"
> + "movesl %2@+,%5\n\t"
> + "addxl %5,%0\n\t"
> + "movel %5,%3@+\n\t"
> + "12:\t"
> + "movesl %2@+,%5\n\t"
> + "addxl %5,%0\n\t"
> + "movel %5,%3@+\n\t"
> + "13:\t"
> + "movesl %2@+,%5\n\t"
> + "addxl %5,%0\n\t"
> + "movel %5,%3@+\n\t"
> + "14:\t"
> + "movesl %2@+,%5\n\t"
> + "addxl %5,%0\n\t"
> + "movel %5,%3@+\n\t"
> + "15:\t"
> + "movesl %2@+,%5\n\t"
> + "addxl %5,%0\n\t"
> + "movel %5,%3@+\n\t"
> + "16:\t"
> + "movesl %2@+,%5\n\t"
> + "addxl %5,%0\n\t"
> + "movel %5,%3@+\n\t"
> + "17:\t"
> + "movesl %2@+,%5\n\t"
> + "addxl %5,%0\n\t"
> + "movel %5,%3@+\n\t"
> + "18:\t"
> + "movesl %2@+,%5\n\t"
> + "addxl %5,%0\n\t"
> + "movel %5,%3@+\n\t"
> + "dbra %1,1b\n\t"
> + "clrl %5\n\t"
> + "addxl %5,%0\n\t" /* add X bit */
> + "clrw %1\n\t"
> + "subql #1,%1\n\t"
> + "jcc 1b\n"
> + "2:\t"
> + "movel %4,%1\n\t" /* restore len from tmp1 */
> + "andw #0x1c,%4\n\t" /* number of rest longs */
> + "jeq 4f\n\t"
> + "lsrw #2,%4\n\t"
> + "subqw #1,%4\n"
> + "3:\n"
> + /* loop for rest longs */
> + "19:\t"
> + "movesl %2@+,%5\n\t"
> + "addxl %5,%0\n\t"
> + "movel %5,%3@+\n\t"
> + "dbra %4,3b\n\t"
> + "clrl %5\n\t"
> + "addxl %5,%0\n" /* add X bit */
> + "4:\t"
> + /* now check for rest bytes that do not fit into longs */
> + "andw #3,%1\n\t"
> + "jeq 7f\n\t"
> + "clrl %5\n\t" /* clear tmp2 for rest bytes */
> + "subqw #2,%1\n\t"
> + "jlt 5f\n\t"
> + "20:\t"
> + "movesw %2@+,%5\n\t" /* have rest>= 2: get word */
> + "movew %5,%3@+\n\t"
> + "swap %5\n\t" /* into bits 16..31 */
> + "tstw %1\n\t" /* another byte? */
> + "jeq 6f\n"
> + "5:\n"
> + "21:\t"
> + "movesb %2@,%5\n\t" /* have odd rest: get byte */
> + "moveb %5,%3@+\n\t"
> + "lslw #8,%5\n\t" /* into bits 8..15; 16..31 untouched */
> + "6:\t"
> + "addl %5,%0\n\t" /* now add rest long to sum */
> + "clrl %5\n\t"
> + "addxl %5,%0\n\t" /* add X bit */
> + "7:\t"
> + "clrl %5\n" /* no error - clear return value */
> + "8:\n"
> + ".section .fixup,\"ax\"\n"
> + ".even\n"
> + /* If any exception occurs zero out the rest.
> + Similarities with the code above are intentional :-) */
> + "90:\t"
> + "clrw %3@+\n\t"
> + "movel %1,%4\n\t"
> + "lsrl #5,%1\n\t"
> + "jeq 1f\n\t"
> + "subql #1,%1\n"
> + "91:\t"
> + "clrl %3@+\n"
> + "92:\t"
> + "clrl %3@+\n"
> + "93:\t"
> + "clrl %3@+\n"
> + "94:\t"
> + "clrl %3@+\n"
> + "95:\t"
> + "clrl %3@+\n"
> + "96:\t"
> + "clrl %3@+\n"
> + "97:\t"
> + "clrl %3@+\n"
> + "98:\t"
> + "clrl %3@+\n\t"
> + "dbra %1,91b\n\t"
> + "clrw %1\n\t"
> + "subql #1,%1\n\t"
> + "jcc 91b\n"
> + "1:\t"
> + "movel %4,%1\n\t"
> + "andw #0x1c,%4\n\t"
> + "jeq 1f\n\t"
> + "lsrw #2,%4\n\t"
> + "subqw #1,%4\n"
> + "99:\t"
> + "clrl %3@+\n\t"
> + "dbra %4,99b\n\t"
> + "1:\t"
> + "andw #3,%1\n\t"
> + "jeq 9f\n"
> + "100:\t"
> + "clrw %3@+\n\t"
> + "tstw %1\n\t"
> + "jeq 9f\n"
> + "101:\t"
> + "clrb %3@+\n"
> + "9:\t"
> +#define STR(X) STR1(X)
> +#define STR1(X) #X
> + "moveq #-" STR(EFAULT) ",%5\n\t"
> + "jra 8b\n"
> + ".previous\n"
> + ".section __ex_table,\"a\"\n"
> + ".long 10b,90b\n"
> + ".long 11b,91b\n"
> + ".long 12b,92b\n"
> + ".long 13b,93b\n"
> + ".long 14b,94b\n"
> + ".long 15b,95b\n"
> + ".long 16b,96b\n"
> + ".long 17b,97b\n"
> + ".long 18b,98b\n"
> + ".long 19b,99b\n"
> + ".long 20b,100b\n"
> + ".long 21b,101b\n"
> + ".previous"
> + : "=d" (sum), "=d" (len), "=a" (src), "=a" (dst),
> + "=&d" (tmp1), "=d" (tmp2)
> + : "0" (sum), "1" (len), "2" (src), "3" (dst)
> + );
> +
> + *csum_err = tmp2;
> +
> + return(sum);
> +}
> +
> +EXPORT_SYMBOL(csum_partial_copy_from_user);
> +
> +
> +/*
> + * copy from kernel space while checksumming, otherwise like csum_partial
> + */
> +
> +__wsum
> +csum_partial_copy_nocheck(const void *src, void *dst, int len, __wsum sum)
> +{
> + unsigned long tmp1, tmp2;
> + __asm__("movel %2,%4\n\t"
> + "btst #1,%4\n\t" /* Check alignment */
> + "jeq 2f\n\t"
> + "subql #2,%1\n\t" /* buff%4==2: treat first word */
> + "jgt 1f\n\t"
> + "addql #2,%1\n\t" /* len was == 2, treat only rest */
> + "jra 4f\n"
> + "1:\t"
> + "movew %2@+,%4\n\t" /* add first word to sum */
> + "addw %4,%0\n\t"
> + "movew %4,%3@+\n\t"
> + "clrl %4\n\t"
> + "addxl %4,%0\n" /* add X bit */
> + "2:\t"
> + /* unrolled loop for the main part: do 8 longs at once */
> + "movel %1,%4\n\t" /* save len in tmp1 */
> + "lsrl #5,%1\n\t" /* len/32 */
> + "jeq 2f\n\t" /* not enough... */
> + "subql #1,%1\n"
> + "1:\t"
> + "movel %2@+,%5\n\t"
> + "addxl %5,%0\n\t"
> + "movel %5,%3@+\n\t"
> + "movel %2@+,%5\n\t"
> + "addxl %5,%0\n\t"
> + "movel %5,%3@+\n\t"
> + "movel %2@+,%5\n\t"
> + "addxl %5,%0\n\t"
> + "movel %5,%3@+\n\t"
> + "movel %2@+,%5\n\t"
> + "addxl %5,%0\n\t"
> + "movel %5,%3@+\n\t"
> + "movel %2@+,%5\n\t"
> + "addxl %5,%0\n\t"
> + "movel %5,%3@+\n\t"
> + "movel %2@+,%5\n\t"
> + "addxl %5,%0\n\t"
> + "movel %5,%3@+\n\t"
> + "movel %2@+,%5\n\t"
> + "addxl %5,%0\n\t"
> + "movel %5,%3@+\n\t"
> + "movel %2@+,%5\n\t"
> + "addxl %5,%0\n\t"
> + "movel %5,%3@+\n\t"
> + "dbra %1,1b\n\t"
> + "clrl %5\n\t"
> + "addxl %5,%0\n\t" /* add X bit */
> + "clrw %1\n\t"
> + "subql #1,%1\n\t"
> + "jcc 1b\n"
> + "2:\t"
> + "movel %4,%1\n\t" /* restore len from tmp1 */
> + "andw #0x1c,%4\n\t" /* number of rest longs */
> + "jeq 4f\n\t"
> + "lsrw #2,%4\n\t"
> + "subqw #1,%4\n"
> + "3:\t"
> + /* loop for rest longs */
> + "movel %2@+,%5\n\t"
> + "addxl %5,%0\n\t"
> + "movel %5,%3@+\n\t"
> + "dbra %4,3b\n\t"
> + "clrl %5\n\t"
> + "addxl %5,%0\n" /* add X bit */
> + "4:\t"
> + /* now check for rest bytes that do not fit into longs */
> + "andw #3,%1\n\t"
> + "jeq 7f\n\t"
> + "clrl %5\n\t" /* clear tmp2 for rest bytes */
> + "subqw #2,%1\n\t"
> + "jlt 5f\n\t"
> + "movew %2@+,%5\n\t" /* have rest>= 2: get word */
> + "movew %5,%3@+\n\t"
> + "swap %5\n\t" /* into bits 16..31 */
> + "tstw %1\n\t" /* another byte? */
> + "jeq 6f\n"
> + "5:\t"
> + "moveb %2@,%5\n\t" /* have odd rest: get byte */
> + "moveb %5,%3@+\n\t"
> + "lslw #8,%5\n" /* into bits 8..15; 16..31 untouched */
> + "6:\t"
> + "addl %5,%0\n\t" /* now add rest long to sum */
> + "clrl %5\n\t"
> + "addxl %5,%0\n" /* add X bit */
> + "7:\t"
> + : "=d" (sum), "=d" (len), "=a" (src), "=a" (dst),
> + "=&d" (tmp1), "=&d" (tmp2)
> + : "0" (sum), "1" (len), "2" (src), "3" (dst)
> + );
> + return(sum);
> +}
> +EXPORT_SYMBOL(csum_partial_copy_nocheck);
> diff --git a/arch/m68k/lib/checksum_mm.c b/arch/m68k/lib/checksum_mm.c
> deleted file mode 100644
> index 6216f12..0000000
> --- a/arch/m68k/lib/checksum_mm.c
> +++ /dev/null
> @@ -1,425 +0,0 @@
> -/*
> - * 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.
> - *
> - * 1998/8/31 Andreas Schwab:
> - * Zero out rest of buffer on exception in
> - * csum_partial_copy_from_user.
> - */
> -
> -#include<linux/module.h>
> -#include<net/checksum.h>
> -
> -/*
> - * computes a partial checksum, e.g. for TCP/UDP fragments
> - */
> -
> -__wsum csum_partial(const void *buff, int len, __wsum sum)
> -{
> - unsigned long tmp1, tmp2;
> - /*
> - * Experiments with ethernet and slip connections show that buff
> - * is aligned on either a 2-byte or 4-byte boundary.
> - */
> - __asm__("movel %2,%3\n\t"
> - "btst #1,%3\n\t" /* Check alignment */
> - "jeq 2f\n\t"
> - "subql #2,%1\n\t" /* buff%4==2: treat first word */
> - "jgt 1f\n\t"
> - "addql #2,%1\n\t" /* len was == 2, treat only rest */
> - "jra 4f\n"
> - "1:\t"
> - "addw %2@+,%0\n\t" /* add first word to sum */
> - "clrl %3\n\t"
> - "addxl %3,%0\n" /* add X bit */
> - "2:\t"
> - /* unrolled loop for the main part: do 8 longs at once */
> - "movel %1,%3\n\t" /* save len in tmp1 */
> - "lsrl #5,%1\n\t" /* len/32 */
> - "jeq 2f\n\t" /* not enough... */
> - "subql #1,%1\n"
> - "1:\t"
> - "movel %2@+,%4\n\t"
> - "addxl %4,%0\n\t"
> - "movel %2@+,%4\n\t"
> - "addxl %4,%0\n\t"
> - "movel %2@+,%4\n\t"
> - "addxl %4,%0\n\t"
> - "movel %2@+,%4\n\t"
> - "addxl %4,%0\n\t"
> - "movel %2@+,%4\n\t"
> - "addxl %4,%0\n\t"
> - "movel %2@+,%4\n\t"
> - "addxl %4,%0\n\t"
> - "movel %2@+,%4\n\t"
> - "addxl %4,%0\n\t"
> - "movel %2@+,%4\n\t"
> - "addxl %4,%0\n\t"
> - "dbra %1,1b\n\t"
> - "clrl %4\n\t"
> - "addxl %4,%0\n\t" /* add X bit */
> - "clrw %1\n\t"
> - "subql #1,%1\n\t"
> - "jcc 1b\n"
> - "2:\t"
> - "movel %3,%1\n\t" /* restore len from tmp1 */
> - "andw #0x1c,%3\n\t" /* number of rest longs */
> - "jeq 4f\n\t"
> - "lsrw #2,%3\n\t"
> - "subqw #1,%3\n"
> - "3:\t"
> - /* loop for rest longs */
> - "movel %2@+,%4\n\t"
> - "addxl %4,%0\n\t"
> - "dbra %3,3b\n\t"
> - "clrl %4\n\t"
> - "addxl %4,%0\n" /* add X bit */
> - "4:\t"
> - /* now check for rest bytes that do not fit into longs */
> - "andw #3,%1\n\t"
> - "jeq 7f\n\t"
> - "clrl %4\n\t" /* clear tmp2 for rest bytes */
> - "subqw #2,%1\n\t"
> - "jlt 5f\n\t"
> - "movew %2@+,%4\n\t" /* have rest>= 2: get word */
> - "swap %4\n\t" /* into bits 16..31 */
> - "tstw %1\n\t" /* another byte? */
> - "jeq 6f\n"
> - "5:\t"
> - "moveb %2@,%4\n\t" /* have odd rest: get byte */
> - "lslw #8,%4\n\t" /* into bits 8..15; 16..31 untouched */
> - "6:\t"
> - "addl %4,%0\n\t" /* now add rest long to sum */
> - "clrl %4\n\t"
> - "addxl %4,%0\n" /* add X bit */
> - "7:\t"
> - : "=d" (sum), "=d" (len), "=a" (buff),
> - "=&d" (tmp1), "=&d" (tmp2)
> - : "0" (sum), "1" (len), "2" (buff)
> - );
> - return(sum);
> -}
> -
> -EXPORT_SYMBOL(csum_partial);
> -
> -
> -/*
> - * copy from user space while checksumming, with exception handling.
> - */
> -
> -__wsum
> -csum_partial_copy_from_user(const void __user *src, void *dst,
> - int len, __wsum sum, int *csum_err)
> -{
> - /*
> - * GCC doesn't like more than 10 operands for the asm
> - * statements so we have to use tmp2 for the error
> - * code.
> - */
> - unsigned long tmp1, tmp2;
> -
> - __asm__("movel %2,%4\n\t"
> - "btst #1,%4\n\t" /* Check alignment */
> - "jeq 2f\n\t"
> - "subql #2,%1\n\t" /* buff%4==2: treat first word */
> - "jgt 1f\n\t"
> - "addql #2,%1\n\t" /* len was == 2, treat only rest */
> - "jra 4f\n"
> - "1:\n"
> - "10:\t"
> - "movesw %2@+,%4\n\t" /* add first word to sum */
> - "addw %4,%0\n\t"
> - "movew %4,%3@+\n\t"
> - "clrl %4\n\t"
> - "addxl %4,%0\n" /* add X bit */
> - "2:\t"
> - /* unrolled loop for the main part: do 8 longs at once */
> - "movel %1,%4\n\t" /* save len in tmp1 */
> - "lsrl #5,%1\n\t" /* len/32 */
> - "jeq 2f\n\t" /* not enough... */
> - "subql #1,%1\n"
> - "1:\n"
> - "11:\t"
> - "movesl %2@+,%5\n\t"
> - "addxl %5,%0\n\t"
> - "movel %5,%3@+\n\t"
> - "12:\t"
> - "movesl %2@+,%5\n\t"
> - "addxl %5,%0\n\t"
> - "movel %5,%3@+\n\t"
> - "13:\t"
> - "movesl %2@+,%5\n\t"
> - "addxl %5,%0\n\t"
> - "movel %5,%3@+\n\t"
> - "14:\t"
> - "movesl %2@+,%5\n\t"
> - "addxl %5,%0\n\t"
> - "movel %5,%3@+\n\t"
> - "15:\t"
> - "movesl %2@+,%5\n\t"
> - "addxl %5,%0\n\t"
> - "movel %5,%3@+\n\t"
> - "16:\t"
> - "movesl %2@+,%5\n\t"
> - "addxl %5,%0\n\t"
> - "movel %5,%3@+\n\t"
> - "17:\t"
> - "movesl %2@+,%5\n\t"
> - "addxl %5,%0\n\t"
> - "movel %5,%3@+\n\t"
> - "18:\t"
> - "movesl %2@+,%5\n\t"
> - "addxl %5,%0\n\t"
> - "movel %5,%3@+\n\t"
> - "dbra %1,1b\n\t"
> - "clrl %5\n\t"
> - "addxl %5,%0\n\t" /* add X bit */
> - "clrw %1\n\t"
> - "subql #1,%1\n\t"
> - "jcc 1b\n"
> - "2:\t"
> - "movel %4,%1\n\t" /* restore len from tmp1 */
> - "andw #0x1c,%4\n\t" /* number of rest longs */
> - "jeq 4f\n\t"
> - "lsrw #2,%4\n\t"
> - "subqw #1,%4\n"
> - "3:\n"
> - /* loop for rest longs */
> - "19:\t"
> - "movesl %2@+,%5\n\t"
> - "addxl %5,%0\n\t"
> - "movel %5,%3@+\n\t"
> - "dbra %4,3b\n\t"
> - "clrl %5\n\t"
> - "addxl %5,%0\n" /* add X bit */
> - "4:\t"
> - /* now check for rest bytes that do not fit into longs */
> - "andw #3,%1\n\t"
> - "jeq 7f\n\t"
> - "clrl %5\n\t" /* clear tmp2 for rest bytes */
> - "subqw #2,%1\n\t"
> - "jlt 5f\n\t"
> - "20:\t"
> - "movesw %2@+,%5\n\t" /* have rest>= 2: get word */
> - "movew %5,%3@+\n\t"
> - "swap %5\n\t" /* into bits 16..31 */
> - "tstw %1\n\t" /* another byte? */
> - "jeq 6f\n"
> - "5:\n"
> - "21:\t"
> - "movesb %2@,%5\n\t" /* have odd rest: get byte */
> - "moveb %5,%3@+\n\t"
> - "lslw #8,%5\n\t" /* into bits 8..15; 16..31 untouched */
> - "6:\t"
> - "addl %5,%0\n\t" /* now add rest long to sum */
> - "clrl %5\n\t"
> - "addxl %5,%0\n\t" /* add X bit */
> - "7:\t"
> - "clrl %5\n" /* no error - clear return value */
> - "8:\n"
> - ".section .fixup,\"ax\"\n"
> - ".even\n"
> - /* If any exception occurs zero out the rest.
> - Similarities with the code above are intentional :-) */
> - "90:\t"
> - "clrw %3@+\n\t"
> - "movel %1,%4\n\t"
> - "lsrl #5,%1\n\t"
> - "jeq 1f\n\t"
> - "subql #1,%1\n"
> - "91:\t"
> - "clrl %3@+\n"
> - "92:\t"
> - "clrl %3@+\n"
> - "93:\t"
> - "clrl %3@+\n"
> - "94:\t"
> - "clrl %3@+\n"
> - "95:\t"
> - "clrl %3@+\n"
> - "96:\t"
> - "clrl %3@+\n"
> - "97:\t"
> - "clrl %3@+\n"
> - "98:\t"
> - "clrl %3@+\n\t"
> - "dbra %1,91b\n\t"
> - "clrw %1\n\t"
> - "subql #1,%1\n\t"
> - "jcc 91b\n"
> - "1:\t"
> - "movel %4,%1\n\t"
> - "andw #0x1c,%4\n\t"
> - "jeq 1f\n\t"
> - "lsrw #2,%4\n\t"
> - "subqw #1,%4\n"
> - "99:\t"
> - "clrl %3@+\n\t"
> - "dbra %4,99b\n\t"
> - "1:\t"
> - "andw #3,%1\n\t"
> - "jeq 9f\n"
> - "100:\t"
> - "clrw %3@+\n\t"
> - "tstw %1\n\t"
> - "jeq 9f\n"
> - "101:\t"
> - "clrb %3@+\n"
> - "9:\t"
> -#define STR(X) STR1(X)
> -#define STR1(X) #X
> - "moveq #-" STR(EFAULT) ",%5\n\t"
> - "jra 8b\n"
> - ".previous\n"
> - ".section __ex_table,\"a\"\n"
> - ".long 10b,90b\n"
> - ".long 11b,91b\n"
> - ".long 12b,92b\n"
> - ".long 13b,93b\n"
> - ".long 14b,94b\n"
> - ".long 15b,95b\n"
> - ".long 16b,96b\n"
> - ".long 17b,97b\n"
> - ".long 18b,98b\n"
> - ".long 19b,99b\n"
> - ".long 20b,100b\n"
> - ".long 21b,101b\n"
> - ".previous"
> - : "=d" (sum), "=d" (len), "=a" (src), "=a" (dst),
> - "=&d" (tmp1), "=d" (tmp2)
> - : "0" (sum), "1" (len), "2" (src), "3" (dst)
> - );
> -
> - *csum_err = tmp2;
> -
> - return(sum);
> -}
> -
> -EXPORT_SYMBOL(csum_partial_copy_from_user);
> -
> -
> -/*
> - * copy from kernel space while checksumming, otherwise like csum_partial
> - */
> -
> -__wsum
> -csum_partial_copy_nocheck(const void *src, void *dst, int len, __wsum sum)
> -{
> - unsigned long tmp1, tmp2;
> - __asm__("movel %2,%4\n\t"
> - "btst #1,%4\n\t" /* Check alignment */
> - "jeq 2f\n\t"
> - "subql #2,%1\n\t" /* buff%4==2: treat first word */
> - "jgt 1f\n\t"
> - "addql #2,%1\n\t" /* len was == 2, treat only rest */
> - "jra 4f\n"
> - "1:\t"
> - "movew %2@+,%4\n\t" /* add first word to sum */
> - "addw %4,%0\n\t"
> - "movew %4,%3@+\n\t"
> - "clrl %4\n\t"
> - "addxl %4,%0\n" /* add X bit */
> - "2:\t"
> - /* unrolled loop for the main part: do 8 longs at once */
> - "movel %1,%4\n\t" /* save len in tmp1 */
> - "lsrl #5,%1\n\t" /* len/32 */
> - "jeq 2f\n\t" /* not enough... */
> - "subql #1,%1\n"
> - "1:\t"
> - "movel %2@+,%5\n\t"
> - "addxl %5,%0\n\t"
> - "movel %5,%3@+\n\t"
> - "movel %2@+,%5\n\t"
> - "addxl %5,%0\n\t"
> - "movel %5,%3@+\n\t"
> - "movel %2@+,%5\n\t"
> - "addxl %5,%0\n\t"
> - "movel %5,%3@+\n\t"
> - "movel %2@+,%5\n\t"
> - "addxl %5,%0\n\t"
> - "movel %5,%3@+\n\t"
> - "movel %2@+,%5\n\t"
> - "addxl %5,%0\n\t"
> - "movel %5,%3@+\n\t"
> - "movel %2@+,%5\n\t"
> - "addxl %5,%0\n\t"
> - "movel %5,%3@+\n\t"
> - "movel %2@+,%5\n\t"
> - "addxl %5,%0\n\t"
> - "movel %5,%3@+\n\t"
> - "movel %2@+,%5\n\t"
> - "addxl %5,%0\n\t"
> - "movel %5,%3@+\n\t"
> - "dbra %1,1b\n\t"
> - "clrl %5\n\t"
> - "addxl %5,%0\n\t" /* add X bit */
> - "clrw %1\n\t"
> - "subql #1,%1\n\t"
> - "jcc 1b\n"
> - "2:\t"
> - "movel %4,%1\n\t" /* restore len from tmp1 */
> - "andw #0x1c,%4\n\t" /* number of rest longs */
> - "jeq 4f\n\t"
> - "lsrw #2,%4\n\t"
> - "subqw #1,%4\n"
> - "3:\t"
> - /* loop for rest longs */
> - "movel %2@+,%5\n\t"
> - "addxl %5,%0\n\t"
> - "movel %5,%3@+\n\t"
> - "dbra %4,3b\n\t"
> - "clrl %5\n\t"
> - "addxl %5,%0\n" /* add X bit */
> - "4:\t"
> - /* now check for rest bytes that do not fit into longs */
> - "andw #3,%1\n\t"
> - "jeq 7f\n\t"
> - "clrl %5\n\t" /* clear tmp2 for rest bytes */
> - "subqw #2,%1\n\t"
> - "jlt 5f\n\t"
> - "movew %2@+,%5\n\t" /* have rest>= 2: get word */
> - "movew %5,%3@+\n\t"
> - "swap %5\n\t" /* into bits 16..31 */
> - "tstw %1\n\t" /* another byte? */
> - "jeq 6f\n"
> - "5:\t"
> - "moveb %2@,%5\n\t" /* have odd rest: get byte */
> - "moveb %5,%3@+\n\t"
> - "lslw #8,%5\n" /* into bits 8..15; 16..31 untouched */
> - "6:\t"
> - "addl %5,%0\n\t" /* now add rest long to sum */
> - "clrl %5\n\t"
> - "addxl %5,%0\n" /* add X bit */
> - "7:\t"
> - : "=d" (sum), "=d" (len), "=a" (src), "=a" (dst),
> - "=&d" (tmp1), "=&d" (tmp2)
> - : "0" (sum), "1" (len), "2" (src), "3" (dst)
> - );
> - return(sum);
> -}
> -EXPORT_SYMBOL(csum_partial_copy_nocheck);
> diff --git a/arch/m68k/lib/checksum_no.c b/arch/m68k/lib/checksum_no.c
> deleted file mode 100644
> index e4c6354..0000000
> --- a/arch/m68k/lib/checksum_no.c
> +++ /dev/null
> @@ -1,156 +0,0 @@
> -/*
> - * 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>
> -
> -static inline unsigned short from32to16(unsigned long 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 long do_csum(const unsigned char * buff, int len)
> -{
> - int odd, count;
> - unsigned long result = 0;
> -
> - if (len<= 0)
> - goto out;
> - odd = 1& (unsigned long) buff;
> - if (odd) {
> - result = *buff;
> - len--;
> - buff++;
> - }
> - count = len>> 1; /* nr of 16-bit words.. */
> - if (count) {
> - if (2& (unsigned long) buff) {
> - result += *(unsigned short *) buff;
> - count--;
> - len -= 2;
> - buff += 2;
> - }
> - count>>= 1; /* nr of 32-bit words.. */
> - if (count) {
> - unsigned long carry = 0;
> - do {
> - unsigned long w = *(unsigned long *) buff;
> - count--;
> - buff += 4;
> - result += carry;
> - result += w;
> - carry = (w> result);
> - } while (count);
> - result += carry;
> - result = (result& 0xffff) + (result>> 16);
> - }
> - if (len& 2) {
> - result += *(unsigned short *) buff;
> - buff += 2;
> - }
> - }
> - if (len& 1)
> - result += (*buff<< 8);
> - result = from32to16(result);
> - if (odd)
> - result = ((result>> 8)& 0xff) | ((result& 0xff)<< 8);
> -out:
> - return result;
> -}
> -
> -#ifdef CONFIG_COLDFIRE
> -/*
> - * 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);
> -#endif
> -
> -/*
> - * 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 sum)
> -{
> - unsigned int result = do_csum(buff, len);
> -
> - /* add in old sum, and carry.. */
> - result += (__force u32)sum;
> - if ((__force u32)sum> result)
> - result += 1;
> - return (__force __wsum)result;
> -}
> -
> -EXPORT_SYMBOL(csum_partial);
> -
> -/*
> - * 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)
> -{
> - if (csum_err) *csum_err = 0;
> - memcpy(dst, (__force const void *)src, len);
> - 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_nocheck(const void *src, void *dst, int len, __wsum sum)
> -{
> - memcpy(dst, src, len);
> - return csum_partial(dst, len, sum);
> -}
> -EXPORT_SYMBOL(csum_partial_copy_nocheck);
--
------------------------------------------------------------------------
Greg Ungerer -- Principal Engineer EMAIL: gerg@snapgear.com
SnapGear Group, McAfee PHONE: +61 7 3435 2888
8 Gardner Close FAX: +61 7 3217 5323
Milton, QLD, 4064, Australia WEB: http://www.SnapGear.com
prev parent reply other threads:[~2011-11-24 23:34 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-11-18 6:32 [PATCH] m68k: simpler m68k and ColdFire CPU's can use generic csum code gerg
2011-11-24 23:36 ` Greg Ungerer [this message]
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=4ECED506.6030201@snapgear.com \
--to=gerg@snapgear.com \
--cc=geert@linux-m68k.org \
--cc=linux-m68k@vger.kernel.org \
--cc=uclinux-dev@uclinux.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox