[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#618386: gcc-4.5: fails to vectorize trivial loop with aligned operands



Package: gcc-4.5
Version: 4.5.2-6
Severity: wishlist

Attached is a file, ofb.c.  When compiling with -DVECTOR_T, the xor and
copy operation is appropriately vectorized using SSE and SSE2
instructions (these are assumed, since the machine is amd64).  Without
-DVECTOR_T, the operation is not vectorized and is instead replaced by
16 movzbl-xorb-movb operations.  In the original code, vectorization
causes not only a code size decrease, but also a significant speed
increase (87 MiB/s to 96 MiB/s).

It's unclear to me why GCC isn't vectorizing this operation.  All of in,
out, and c->buf are guaranteed to be 16-byte aligned[0].  GCC even
recognizes this when using VECTOR_T because it uses movdqa instead of
the unaligned movdqu.  I don't think I should have to rewrite the code
to use vector types when the operation is clearly trivially
vectorizable.  I want GCC to produce identical code for either case.

I've tried with gcc 4.4, 4.5, 4.6, and snapshot, and all of them have
this bug.  The code is compiled as follows:

  gcc -fPIC -shared -std=c99 -O3 --save-temps -o ofb.so ofb.c

or

  gcc -DVECTOR_T -fPIC -shared -std=c99 -O3 --save-temps -o ofb.so ofb.c

as the case may be.  The original C file is included, as are the two .i
files and the two .s files.  They are marked as either with -DVECTOR_T
or without it.

gcc -v output:

  Using built-in specs.
  COLLECT_GCC=gcc
  COLLECT_LTO_WRAPPER=/usr/lib/gcc/x86_64-linux-gnu/4.5.2/lto-wrapper
  Target: x86_64-linux-gnu
  Configured with: ../src/configure -v --with-pkgversion='Debian 4.5.2-6' --with-bugurl=file:///usr/share/doc/gcc-4.5/README.Bugs --enable-languages=c,c++,fortran,objc,obj-c++ --prefix=/usr --program-suffix=-4.5 --enable-shared --enable-multiarch --enable-linker-build-id --with-system-zlib --libexecdir=/usr/lib --without-included-gettext --enable-threads=posix --with-gxx-include-dir=/usr/include/c++/4.5 --libdir=/usr/lib --enable-nls --enable-clocale=gnu --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-plugin --enable-gold --enable-ld=default --with-plugin-ld=ld.gold --enable-objc-gc --with-arch-32=i586 --with-tune=generic --enable-checking=release --build=x86_64-linux-gnu --host=x86_64-linux-gnu --target=x86_64-linux-gnu
  Thread model: posix
  gcc version 4.5.2 (Debian 4.5.2-6) 

[0] in and out are defined by the API of this function to be 16-byte
aligned, so the cast from a void * to a struct aligned * is guaranteed
to be okay.

-- System Information:
Debian Release: wheezy/sid
  APT prefers unstable
  APT policy: (500, 'unstable'), (1, 'experimental')
Architecture: amd64 (x86_64)

Kernel: Linux 2.6.38-rc7-amd64 (SMP w/2 CPU cores)
Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash

Versions of packages gcc-4.5 depends on:
ii  binutils               2.21.0.20110302-2 The GNU assembler, linker and bina
ii  cpp-4.5                4.5.2-6           The GNU C preprocessor
ii  gcc-4.5-base           4.5.2-6           The GNU Compiler Collection (base 
ii  libc6                  2.11.2-13         Embedded GNU C Library: Shared lib
ii  libcloog-ppl0          0.15.9-3          the Chunky Loop Generator (runtime
ii  libelfg0               0.8.13-1          an ELF object file access library
ii  libgcc1                1:4.6-20110105-1  GCC support library
ii  libgmp10               2:5.0.1+dfsg-6    Multiprecision arithmetic library
ii  libgmpxx4ldbl          2:5.0.1+dfsg-6    Multiprecision arithmetic library 
ii  libgomp1               4.6-20110105-1    GCC OpenMP (GOMP) support library
ii  libmpc2                0.9-2             multiple precision complex floatin
ii  libmpfr4               3.0.0-9           multiple precision floating-point 
ii  libppl-c4              0.11.2-2          Parma Polyhedra Library (C interfa
ii  libppl9                0.11.2-2          Parma Polyhedra Library (runtime l
ii  zlib1g                 1:1.2.3.4.dfsg-3  compression library - runtime

Versions of packages gcc-4.5 recommends:
ii  libc6-dev                     2.11.2-13  Embedded GNU C Library: Developmen

Versions of packages gcc-4.5 suggests:
pn  binutils-gold                 <none>     (no description available)
pn  gcc-4.5-doc                   <none>     (no description available)
pn  gcc-4.5-locales               <none>     (no description available)
ii  gcc-4.5-multilib              4.5.2-6    The GNU C compiler (multilib files
pn  libgcc1-dbg                   <none>     (no description available)
pn  libgomp1-dbg                  <none>     (no description available)
ii  libmudflap0-4.5-dev           4.5.2-6    GCC mudflap support libraries (dev
pn  libmudflap0-dbg               <none>     (no description available)
ii  libppl-c2                     0.10.2-9   Parma Polyhedra Library (C interfa
ii  libppl7                       0.10.2-9   Parma Polyhedra Library (runtime l

-- no debconf information

-- 
brian m. carlson / brian with sandals: Houston, Texas, US
+1 832 623 2791 | http://www.crustytoothpaste.net/~bmc | My opinion only
OpenPGP: RSA v4 4096b: 88AC E9B2 9196 305B A994 7552 F1BA 225C 0223 B187
#include <string.h>
#include <stdint.h>

#define STATIC_ASSERT(e) ((void)sizeof(char[1 - 2*!(e)]))

struct ofb {
	void (*encryptfast)(void *, void *, const void *, size_t);
	uint8_t buf[32] __attribute__((aligned(16)));
	size_t chunks;
};

struct aligned {
	uint8_t data[16] __attribute__((aligned(16)));
};

int ofb_encryptfast(void *ctx, void *outp, const void *inp, size_t len)
{
	struct ofb *c = ctx;
	struct aligned *out = outp;
	const struct aligned *in = inp;
	STATIC_ASSERT(sizeof(*out) == 16);
	STATIC_ASSERT(sizeof(*in) == 16);
	STATIC_ASSERT(sizeof(out->data) == 16);
	STATIC_ASSERT(sizeof(in->data) == 16);
	STATIC_ASSERT(__alignof__(*out) == 16);
	STATIC_ASSERT(__alignof__(*in) == 16);
	STATIC_ASSERT(__alignof__(out->data) == 16);
	STATIC_ASSERT(__alignof__(in->data) == 16);
	STATIC_ASSERT(__alignof__(c->buf) == 16);
	len /= 16;
	for (size_t iters = 0; iters < len; iters++, in++, out++) {
		c->encryptfast(c, c->buf, c->buf, c->chunks);
#ifdef VECTOR_T
		typedef unsigned vector_t __attribute__ ((vector_size (16)));
		vector_t bufv, inv;
		memcpy(&bufv, c->buf, sizeof(vector_t));
		memcpy(&inv, in->data, sizeof(vector_t));
		bufv ^= inv;
		memcpy(out->data, &bufv, sizeof(vector_t));
#else
		for (int i = 0; i < 16; i++)
			out->data[i] = c->buf[i] ^ in->data[i];
#endif
	}
	return 0;
}
# 1 "ofb.c"
# 1 "<built-in>"
# 1 "<command-line>"
# 1 "ofb.c"
# 1 "/usr/include/string.h" 1 3 4
# 26 "/usr/include/string.h" 3 4
# 1 "/usr/include/features.h" 1 3 4
# 313 "/usr/include/features.h" 3 4
# 1 "/usr/include/bits/predefs.h" 1 3 4
# 314 "/usr/include/features.h" 2 3 4
# 346 "/usr/include/features.h" 3 4
# 1 "/usr/include/sys/cdefs.h" 1 3 4
# 353 "/usr/include/sys/cdefs.h" 3 4
# 1 "/usr/include/bits/wordsize.h" 1 3 4
# 354 "/usr/include/sys/cdefs.h" 2 3 4
# 347 "/usr/include/features.h" 2 3 4
# 378 "/usr/include/features.h" 3 4
# 1 "/usr/include/gnu/stubs.h" 1 3 4



# 1 "/usr/include/bits/wordsize.h" 1 3 4
# 5 "/usr/include/gnu/stubs.h" 2 3 4




# 1 "/usr/include/gnu/stubs-64.h" 1 3 4
# 10 "/usr/include/gnu/stubs.h" 2 3 4
# 379 "/usr/include/features.h" 2 3 4
# 27 "/usr/include/string.h" 2 3 4






# 1 "/usr/lib/gcc/x86_64-linux-gnu/4.5.2/include/stddef.h" 1 3 4
# 211 "/usr/lib/gcc/x86_64-linux-gnu/4.5.2/include/stddef.h" 3 4
typedef long unsigned int size_t;
# 34 "/usr/include/string.h" 2 3 4









extern void *memcpy (void *__restrict __dest,
       __const void *__restrict __src, size_t __n)
     __attribute__ ((__nothrow__)) __attribute__ ((__nonnull__ (1, 2)));


extern void *memmove (void *__dest, __const void *__src, size_t __n)
     __attribute__ ((__nothrow__)) __attribute__ ((__nonnull__ (1, 2)));

# 62 "/usr/include/string.h" 3 4


extern void *memset (void *__s, int __c, size_t __n) __attribute__ ((__nothrow__)) __attribute__ ((__nonnull__ (1)));


extern int memcmp (__const void *__s1, __const void *__s2, size_t __n)
     __attribute__ ((__nothrow__)) __attribute__ ((__pure__)) __attribute__ ((__nonnull__ (1, 2)));
# 94 "/usr/include/string.h" 3 4
extern void *memchr (__const void *__s, int __c, size_t __n)
      __attribute__ ((__nothrow__)) __attribute__ ((__pure__)) __attribute__ ((__nonnull__ (1)));


# 125 "/usr/include/string.h" 3 4


extern char *strcpy (char *__restrict __dest, __const char *__restrict __src)
     __attribute__ ((__nothrow__)) __attribute__ ((__nonnull__ (1, 2)));

extern char *strncpy (char *__restrict __dest,
        __const char *__restrict __src, size_t __n)
     __attribute__ ((__nothrow__)) __attribute__ ((__nonnull__ (1, 2)));


extern char *strcat (char *__restrict __dest, __const char *__restrict __src)
     __attribute__ ((__nothrow__)) __attribute__ ((__nonnull__ (1, 2)));

extern char *strncat (char *__restrict __dest, __const char *__restrict __src,
        size_t __n) __attribute__ ((__nothrow__)) __attribute__ ((__nonnull__ (1, 2)));


extern int strcmp (__const char *__s1, __const char *__s2)
     __attribute__ ((__nothrow__)) __attribute__ ((__pure__)) __attribute__ ((__nonnull__ (1, 2)));

extern int strncmp (__const char *__s1, __const char *__s2, size_t __n)
     __attribute__ ((__nothrow__)) __attribute__ ((__pure__)) __attribute__ ((__nonnull__ (1, 2)));


extern int strcoll (__const char *__s1, __const char *__s2)
     __attribute__ ((__nothrow__)) __attribute__ ((__pure__)) __attribute__ ((__nonnull__ (1, 2)));

extern size_t strxfrm (char *__restrict __dest,
         __const char *__restrict __src, size_t __n)
     __attribute__ ((__nothrow__)) __attribute__ ((__nonnull__ (2)));

# 208 "/usr/include/string.h" 3 4

# 233 "/usr/include/string.h" 3 4
extern char *strchr (__const char *__s, int __c)
     __attribute__ ((__nothrow__)) __attribute__ ((__pure__)) __attribute__ ((__nonnull__ (1)));
# 260 "/usr/include/string.h" 3 4
extern char *strrchr (__const char *__s, int __c)
     __attribute__ ((__nothrow__)) __attribute__ ((__pure__)) __attribute__ ((__nonnull__ (1)));


# 279 "/usr/include/string.h" 3 4



extern size_t strcspn (__const char *__s, __const char *__reject)
     __attribute__ ((__nothrow__)) __attribute__ ((__pure__)) __attribute__ ((__nonnull__ (1, 2)));


extern size_t strspn (__const char *__s, __const char *__accept)
     __attribute__ ((__nothrow__)) __attribute__ ((__pure__)) __attribute__ ((__nonnull__ (1, 2)));
# 312 "/usr/include/string.h" 3 4
extern char *strpbrk (__const char *__s, __const char *__accept)
     __attribute__ ((__nothrow__)) __attribute__ ((__pure__)) __attribute__ ((__nonnull__ (1, 2)));
# 340 "/usr/include/string.h" 3 4
extern char *strstr (__const char *__haystack, __const char *__needle)
     __attribute__ ((__nothrow__)) __attribute__ ((__pure__)) __attribute__ ((__nonnull__ (1, 2)));




extern char *strtok (char *__restrict __s, __const char *__restrict __delim)
     __attribute__ ((__nothrow__)) __attribute__ ((__nonnull__ (2)));




extern char *__strtok_r (char *__restrict __s,
    __const char *__restrict __delim,
    char **__restrict __save_ptr)
     __attribute__ ((__nothrow__)) __attribute__ ((__nonnull__ (2, 3)));
# 395 "/usr/include/string.h" 3 4


extern size_t strlen (__const char *__s)
     __attribute__ ((__nothrow__)) __attribute__ ((__pure__)) __attribute__ ((__nonnull__ (1)));

# 409 "/usr/include/string.h" 3 4


extern char *strerror (int __errnum) __attribute__ ((__nothrow__));

# 449 "/usr/include/string.h" 3 4
extern void __bzero (void *__s, size_t __n) __attribute__ ((__nothrow__)) __attribute__ ((__nonnull__ (1)));
# 632 "/usr/include/string.h" 3 4
# 1 "/usr/include/bits/string.h" 1 3 4
# 633 "/usr/include/string.h" 2 3 4


# 1 "/usr/include/bits/string2.h" 1 3 4
# 52 "/usr/include/bits/string2.h" 3 4
# 1 "/usr/include/endian.h" 1 3 4
# 37 "/usr/include/endian.h" 3 4
# 1 "/usr/include/bits/endian.h" 1 3 4
# 38 "/usr/include/endian.h" 2 3 4
# 53 "/usr/include/bits/string2.h" 2 3 4
# 1 "/usr/include/bits/types.h" 1 3 4
# 28 "/usr/include/bits/types.h" 3 4
# 1 "/usr/include/bits/wordsize.h" 1 3 4
# 29 "/usr/include/bits/types.h" 2 3 4


typedef unsigned char __u_char;
typedef unsigned short int __u_short;
typedef unsigned int __u_int;
typedef unsigned long int __u_long;


typedef signed char __int8_t;
typedef unsigned char __uint8_t;
typedef signed short int __int16_t;
typedef unsigned short int __uint16_t;
typedef signed int __int32_t;
typedef unsigned int __uint32_t;

typedef signed long int __int64_t;
typedef unsigned long int __uint64_t;







typedef long int __quad_t;
typedef unsigned long int __u_quad_t;
# 131 "/usr/include/bits/types.h" 3 4
# 1 "/usr/include/bits/typesizes.h" 1 3 4
# 132 "/usr/include/bits/types.h" 2 3 4


typedef unsigned long int __dev_t;
typedef unsigned int __uid_t;
typedef unsigned int __gid_t;
typedef unsigned long int __ino_t;
typedef unsigned long int __ino64_t;
typedef unsigned int __mode_t;
typedef unsigned long int __nlink_t;
typedef long int __off_t;
typedef long int __off64_t;
typedef int __pid_t;
typedef struct { int __val[2]; } __fsid_t;
typedef long int __clock_t;
typedef unsigned long int __rlim_t;
typedef unsigned long int __rlim64_t;
typedef unsigned int __id_t;
typedef long int __time_t;
typedef unsigned int __useconds_t;
typedef long int __suseconds_t;

typedef int __daddr_t;
typedef long int __swblk_t;
typedef int __key_t;


typedef int __clockid_t;


typedef void * __timer_t;


typedef long int __blksize_t;




typedef long int __blkcnt_t;
typedef long int __blkcnt64_t;


typedef unsigned long int __fsblkcnt_t;
typedef unsigned long int __fsblkcnt64_t;


typedef unsigned long int __fsfilcnt_t;
typedef unsigned long int __fsfilcnt64_t;

typedef long int __ssize_t;



typedef __off64_t __loff_t;
typedef __quad_t *__qaddr_t;
typedef char *__caddr_t;


typedef long int __intptr_t;


typedef unsigned int __socklen_t;
# 54 "/usr/include/bits/string2.h" 2 3 4
# 394 "/usr/include/bits/string2.h" 3 4
extern void *__rawmemchr (const void *__s, int __c);
# 969 "/usr/include/bits/string2.h" 3 4
extern __inline __attribute__ ((__gnu_inline__)) size_t __strcspn_c1 (__const char *__s, int __reject);
extern __inline __attribute__ ((__gnu_inline__)) size_t
__strcspn_c1 (__const char *__s, int __reject)
{
  register size_t __result = 0;
  while (__s[__result] != '\0' && __s[__result] != __reject)
    ++__result;
  return __result;
}

extern __inline __attribute__ ((__gnu_inline__)) size_t __strcspn_c2 (__const char *__s, int __reject1,
         int __reject2);
extern __inline __attribute__ ((__gnu_inline__)) size_t
__strcspn_c2 (__const char *__s, int __reject1, int __reject2)
{
  register size_t __result = 0;
  while (__s[__result] != '\0' && __s[__result] != __reject1
  && __s[__result] != __reject2)
    ++__result;
  return __result;
}

extern __inline __attribute__ ((__gnu_inline__)) size_t __strcspn_c3 (__const char *__s, int __reject1,
         int __reject2, int __reject3);
extern __inline __attribute__ ((__gnu_inline__)) size_t
__strcspn_c3 (__const char *__s, int __reject1, int __reject2,
       int __reject3)
{
  register size_t __result = 0;
  while (__s[__result] != '\0' && __s[__result] != __reject1
  && __s[__result] != __reject2 && __s[__result] != __reject3)
    ++__result;
  return __result;
}
# 1045 "/usr/include/bits/string2.h" 3 4
extern __inline __attribute__ ((__gnu_inline__)) size_t __strspn_c1 (__const char *__s, int __accept);
extern __inline __attribute__ ((__gnu_inline__)) size_t
__strspn_c1 (__const char *__s, int __accept)
{
  register size_t __result = 0;

  while (__s[__result] == __accept)
    ++__result;
  return __result;
}

extern __inline __attribute__ ((__gnu_inline__)) size_t __strspn_c2 (__const char *__s, int __accept1,
        int __accept2);
extern __inline __attribute__ ((__gnu_inline__)) size_t
__strspn_c2 (__const char *__s, int __accept1, int __accept2)
{
  register size_t __result = 0;

  while (__s[__result] == __accept1 || __s[__result] == __accept2)
    ++__result;
  return __result;
}

extern __inline __attribute__ ((__gnu_inline__)) size_t __strspn_c3 (__const char *__s, int __accept1,
        int __accept2, int __accept3);
extern __inline __attribute__ ((__gnu_inline__)) size_t
__strspn_c3 (__const char *__s, int __accept1, int __accept2, int __accept3)
{
  register size_t __result = 0;

  while (__s[__result] == __accept1 || __s[__result] == __accept2
  || __s[__result] == __accept3)
    ++__result;
  return __result;
}
# 1121 "/usr/include/bits/string2.h" 3 4
extern __inline __attribute__ ((__gnu_inline__)) char *__strpbrk_c2 (__const char *__s, int __accept1,
         int __accept2);
extern __inline __attribute__ ((__gnu_inline__)) char *
__strpbrk_c2 (__const char *__s, int __accept1, int __accept2)
{

  while (*__s != '\0' && *__s != __accept1 && *__s != __accept2)
    ++__s;
  return *__s == '\0' ? ((void *)0) : (char *) (size_t) __s;
}

extern __inline __attribute__ ((__gnu_inline__)) char *__strpbrk_c3 (__const char *__s, int __accept1,
         int __accept2, int __accept3);
extern __inline __attribute__ ((__gnu_inline__)) char *
__strpbrk_c3 (__const char *__s, int __accept1, int __accept2,
       int __accept3)
{

  while (*__s != '\0' && *__s != __accept1 && *__s != __accept2
  && *__s != __accept3)
    ++__s;
  return *__s == '\0' ? ((void *)0) : (char *) (size_t) __s;
}
# 1172 "/usr/include/bits/string2.h" 3 4
extern __inline __attribute__ ((__gnu_inline__)) char *__strtok_r_1c (char *__s, char __sep, char **__nextp);
extern __inline __attribute__ ((__gnu_inline__)) char *
__strtok_r_1c (char *__s, char __sep, char **__nextp)
{
  char *__result;
  if (__s == ((void *)0))
    __s = *__nextp;
  while (*__s == __sep)
    ++__s;
  __result = ((void *)0);
  if (*__s != '\0')
    {
      __result = __s++;
      while (*__s != '\0')
 if (*__s++ == __sep)
   {
     __s[-1] = '\0';
     break;
   }
    }
  *__nextp = __s;
  return __result;
}
# 1204 "/usr/include/bits/string2.h" 3 4
extern char *__strsep_g (char **__stringp, __const char *__delim);
# 1222 "/usr/include/bits/string2.h" 3 4
extern __inline __attribute__ ((__gnu_inline__)) char *__strsep_1c (char **__s, char __reject);
extern __inline __attribute__ ((__gnu_inline__)) char *
__strsep_1c (char **__s, char __reject)
{
  register char *__retval = *__s;
  if (__retval != ((void *)0) && (*__s = (__extension__ (__builtin_constant_p (__reject) && !__builtin_constant_p (__retval) && (__reject) == '\0' ? (char *) __rawmemchr (__retval, __reject) : __builtin_strchr (__retval, __reject)))) != ((void *)0))
    *(*__s)++ = '\0';
  return __retval;
}

extern __inline __attribute__ ((__gnu_inline__)) char *__strsep_2c (char **__s, char __reject1, char __reject2);
extern __inline __attribute__ ((__gnu_inline__)) char *
__strsep_2c (char **__s, char __reject1, char __reject2)
{
  register char *__retval = *__s;
  if (__retval != ((void *)0))
    {
      register char *__cp = __retval;
      while (1)
 {
   if (*__cp == '\0')
     {
       __cp = ((void *)0);
   break;
     }
   if (*__cp == __reject1 || *__cp == __reject2)
     {
       *__cp++ = '\0';
       break;
     }
   ++__cp;
 }
      *__s = __cp;
    }
  return __retval;
}

extern __inline __attribute__ ((__gnu_inline__)) char *__strsep_3c (char **__s, char __reject1, char __reject2,
       char __reject3);
extern __inline __attribute__ ((__gnu_inline__)) char *
__strsep_3c (char **__s, char __reject1, char __reject2, char __reject3)
{
  register char *__retval = *__s;
  if (__retval != ((void *)0))
    {
      register char *__cp = __retval;
      while (1)
 {
   if (*__cp == '\0')
     {
       __cp = ((void *)0);
   break;
     }
   if (*__cp == __reject1 || *__cp == __reject2 || *__cp == __reject3)
     {
       *__cp++ = '\0';
       break;
     }
   ++__cp;
 }
      *__s = __cp;
    }
  return __retval;
}
# 636 "/usr/include/string.h" 2 3 4
# 644 "/usr/include/string.h" 3 4

# 2 "ofb.c" 2
# 1 "/usr/lib/gcc/x86_64-linux-gnu/4.5.2/include/stdint.h" 1 3 4


# 1 "/usr/include/stdint.h" 1 3 4
# 27 "/usr/include/stdint.h" 3 4
# 1 "/usr/include/bits/wchar.h" 1 3 4
# 28 "/usr/include/stdint.h" 2 3 4
# 1 "/usr/include/bits/wordsize.h" 1 3 4
# 29 "/usr/include/stdint.h" 2 3 4
# 37 "/usr/include/stdint.h" 3 4
typedef signed char int8_t;
typedef short int int16_t;
typedef int int32_t;

typedef long int int64_t;







typedef unsigned char uint8_t;
typedef unsigned short int uint16_t;

typedef unsigned int uint32_t;



typedef unsigned long int uint64_t;
# 66 "/usr/include/stdint.h" 3 4
typedef signed char int_least8_t;
typedef short int int_least16_t;
typedef int int_least32_t;

typedef long int int_least64_t;






typedef unsigned char uint_least8_t;
typedef unsigned short int uint_least16_t;
typedef unsigned int uint_least32_t;

typedef unsigned long int uint_least64_t;
# 91 "/usr/include/stdint.h" 3 4
typedef signed char int_fast8_t;

typedef long int int_fast16_t;
typedef long int int_fast32_t;
typedef long int int_fast64_t;
# 104 "/usr/include/stdint.h" 3 4
typedef unsigned char uint_fast8_t;

typedef unsigned long int uint_fast16_t;
typedef unsigned long int uint_fast32_t;
typedef unsigned long int uint_fast64_t;
# 120 "/usr/include/stdint.h" 3 4
typedef long int intptr_t;


typedef unsigned long int uintptr_t;
# 135 "/usr/include/stdint.h" 3 4
typedef long int intmax_t;
typedef unsigned long int uintmax_t;
# 4 "/usr/lib/gcc/x86_64-linux-gnu/4.5.2/include/stdint.h" 2 3 4
# 3 "ofb.c" 2



struct ofb {
 void (*encryptfast)(void *, void *, const void *, size_t);
 uint8_t buf[32] __attribute__((aligned(16)));
 size_t chunks;
};

struct aligned {
 uint8_t data[16] __attribute__((aligned(16)));
};

int ofb_encryptfast(void *ctx, void *outp, const void *inp, size_t len)
{
 struct ofb *c = ctx;
 struct aligned *out = outp;
 const struct aligned *in = inp;
 ((void)sizeof(char[1 - 2*!(sizeof(*out) == 16)]));
 ((void)sizeof(char[1 - 2*!(sizeof(*in) == 16)]));
 ((void)sizeof(char[1 - 2*!(sizeof(out->data) == 16)]));
 ((void)sizeof(char[1 - 2*!(sizeof(in->data) == 16)]));
 ((void)sizeof(char[1 - 2*!(__alignof__(*out) == 16)]));
 ((void)sizeof(char[1 - 2*!(__alignof__(*in) == 16)]));
 ((void)sizeof(char[1 - 2*!(__alignof__(out->data) == 16)]));
 ((void)sizeof(char[1 - 2*!(__alignof__(in->data) == 16)]));
 ((void)sizeof(char[1 - 2*!(__alignof__(c->buf) == 16)]));
 len /= 16;
 for (size_t iters = 0; iters < len; iters++, in++, out++) {
  c->encryptfast(c, c->buf, c->buf, c->chunks);

  typedef unsigned vector_t __attribute__ ((vector_size (16)));
  vector_t bufv, inv;
  memcpy(&bufv, c->buf, sizeof(vector_t));
  memcpy(&inv, in->data, sizeof(vector_t));
  bufv ^= inv;
  memcpy(out->data, &bufv, sizeof(vector_t));




 }
 return 0;
}
	.file	"ofb.c"
	.text
	.p2align 4,,15
.globl ofb_encryptfast
	.type	ofb_encryptfast, @function
ofb_encryptfast:
.LFB12:
	.cfi_startproc
	pushq	%r15
	.cfi_def_cfa_offset 16
	movq	%rcx, %r15
	.cfi_offset 15, -16
	shrq	$4, %r15
	pushq	%r14
	.cfi_def_cfa_offset 24
	pushq	%r13
	.cfi_def_cfa_offset 32
	pushq	%r12
	.cfi_def_cfa_offset 40
	pushq	%rbp
	.cfi_def_cfa_offset 48
	pushq	%rbx
	.cfi_def_cfa_offset 56
	movq	%rdi, %rbx
	.cfi_offset 3, -56
	.cfi_offset 6, -48
	.cfi_offset 12, -40
	.cfi_offset 13, -32
	.cfi_offset 14, -24
	subq	$8, %rsp
	.cfi_def_cfa_offset 64
	testq	%r15, %r15
	je	.L2
	leaq	16(%rdi), %rbp
	movq	%rsi, %r14
	movq	%rdx, %r13
	xorl	%r12d, %r12d
	.p2align 4,,10
	.p2align 3
.L3:
	movq	48(%rbx), %rcx
	movq	%rbp, %rdx
	movq	%rbp, %rsi
	movq	%rbx, %rdi
	addq	$1, %r12
	call	*(%rbx)
	movdqa	0(%rbp), %xmm0
	pxor	0(%r13), %xmm0
	addq	$16, %r13
	movdqa	%xmm0, (%r14)
	addq	$16, %r14
	cmpq	%r12, %r15
	ja	.L3
.L2:
	addq	$8, %rsp
	.cfi_def_cfa_offset 56
	xorl	%eax, %eax
	popq	%rbx
	.cfi_def_cfa_offset 48
	popq	%rbp
	.cfi_def_cfa_offset 40
	popq	%r12
	.cfi_def_cfa_offset 32
	popq	%r13
	.cfi_def_cfa_offset 24
	popq	%r14
	.cfi_def_cfa_offset 16
	popq	%r15
	.cfi_def_cfa_offset 8
	ret
	.cfi_endproc
.LFE12:
	.size	ofb_encryptfast, .-ofb_encryptfast
	.ident	"GCC: (Debian 4.5.2-6) 4.5.2"
	.section	.note.GNU-stack,"",@progbits
# 1 "ofb.c"
# 1 "<built-in>"
# 1 "<command-line>"
# 1 "ofb.c"
# 1 "/usr/include/string.h" 1 3 4
# 26 "/usr/include/string.h" 3 4
# 1 "/usr/include/features.h" 1 3 4
# 313 "/usr/include/features.h" 3 4
# 1 "/usr/include/bits/predefs.h" 1 3 4
# 314 "/usr/include/features.h" 2 3 4
# 346 "/usr/include/features.h" 3 4
# 1 "/usr/include/sys/cdefs.h" 1 3 4
# 353 "/usr/include/sys/cdefs.h" 3 4
# 1 "/usr/include/bits/wordsize.h" 1 3 4
# 354 "/usr/include/sys/cdefs.h" 2 3 4
# 347 "/usr/include/features.h" 2 3 4
# 378 "/usr/include/features.h" 3 4
# 1 "/usr/include/gnu/stubs.h" 1 3 4



# 1 "/usr/include/bits/wordsize.h" 1 3 4
# 5 "/usr/include/gnu/stubs.h" 2 3 4




# 1 "/usr/include/gnu/stubs-64.h" 1 3 4
# 10 "/usr/include/gnu/stubs.h" 2 3 4
# 379 "/usr/include/features.h" 2 3 4
# 27 "/usr/include/string.h" 2 3 4






# 1 "/usr/lib/gcc/x86_64-linux-gnu/4.5.2/include/stddef.h" 1 3 4
# 211 "/usr/lib/gcc/x86_64-linux-gnu/4.5.2/include/stddef.h" 3 4
typedef long unsigned int size_t;
# 34 "/usr/include/string.h" 2 3 4









extern void *memcpy (void *__restrict __dest,
       __const void *__restrict __src, size_t __n)
     __attribute__ ((__nothrow__)) __attribute__ ((__nonnull__ (1, 2)));


extern void *memmove (void *__dest, __const void *__src, size_t __n)
     __attribute__ ((__nothrow__)) __attribute__ ((__nonnull__ (1, 2)));

# 62 "/usr/include/string.h" 3 4


extern void *memset (void *__s, int __c, size_t __n) __attribute__ ((__nothrow__)) __attribute__ ((__nonnull__ (1)));


extern int memcmp (__const void *__s1, __const void *__s2, size_t __n)
     __attribute__ ((__nothrow__)) __attribute__ ((__pure__)) __attribute__ ((__nonnull__ (1, 2)));
# 94 "/usr/include/string.h" 3 4
extern void *memchr (__const void *__s, int __c, size_t __n)
      __attribute__ ((__nothrow__)) __attribute__ ((__pure__)) __attribute__ ((__nonnull__ (1)));


# 125 "/usr/include/string.h" 3 4


extern char *strcpy (char *__restrict __dest, __const char *__restrict __src)
     __attribute__ ((__nothrow__)) __attribute__ ((__nonnull__ (1, 2)));

extern char *strncpy (char *__restrict __dest,
        __const char *__restrict __src, size_t __n)
     __attribute__ ((__nothrow__)) __attribute__ ((__nonnull__ (1, 2)));


extern char *strcat (char *__restrict __dest, __const char *__restrict __src)
     __attribute__ ((__nothrow__)) __attribute__ ((__nonnull__ (1, 2)));

extern char *strncat (char *__restrict __dest, __const char *__restrict __src,
        size_t __n) __attribute__ ((__nothrow__)) __attribute__ ((__nonnull__ (1, 2)));


extern int strcmp (__const char *__s1, __const char *__s2)
     __attribute__ ((__nothrow__)) __attribute__ ((__pure__)) __attribute__ ((__nonnull__ (1, 2)));

extern int strncmp (__const char *__s1, __const char *__s2, size_t __n)
     __attribute__ ((__nothrow__)) __attribute__ ((__pure__)) __attribute__ ((__nonnull__ (1, 2)));


extern int strcoll (__const char *__s1, __const char *__s2)
     __attribute__ ((__nothrow__)) __attribute__ ((__pure__)) __attribute__ ((__nonnull__ (1, 2)));

extern size_t strxfrm (char *__restrict __dest,
         __const char *__restrict __src, size_t __n)
     __attribute__ ((__nothrow__)) __attribute__ ((__nonnull__ (2)));

# 208 "/usr/include/string.h" 3 4

# 233 "/usr/include/string.h" 3 4
extern char *strchr (__const char *__s, int __c)
     __attribute__ ((__nothrow__)) __attribute__ ((__pure__)) __attribute__ ((__nonnull__ (1)));
# 260 "/usr/include/string.h" 3 4
extern char *strrchr (__const char *__s, int __c)
     __attribute__ ((__nothrow__)) __attribute__ ((__pure__)) __attribute__ ((__nonnull__ (1)));


# 279 "/usr/include/string.h" 3 4



extern size_t strcspn (__const char *__s, __const char *__reject)
     __attribute__ ((__nothrow__)) __attribute__ ((__pure__)) __attribute__ ((__nonnull__ (1, 2)));


extern size_t strspn (__const char *__s, __const char *__accept)
     __attribute__ ((__nothrow__)) __attribute__ ((__pure__)) __attribute__ ((__nonnull__ (1, 2)));
# 312 "/usr/include/string.h" 3 4
extern char *strpbrk (__const char *__s, __const char *__accept)
     __attribute__ ((__nothrow__)) __attribute__ ((__pure__)) __attribute__ ((__nonnull__ (1, 2)));
# 340 "/usr/include/string.h" 3 4
extern char *strstr (__const char *__haystack, __const char *__needle)
     __attribute__ ((__nothrow__)) __attribute__ ((__pure__)) __attribute__ ((__nonnull__ (1, 2)));




extern char *strtok (char *__restrict __s, __const char *__restrict __delim)
     __attribute__ ((__nothrow__)) __attribute__ ((__nonnull__ (2)));




extern char *__strtok_r (char *__restrict __s,
    __const char *__restrict __delim,
    char **__restrict __save_ptr)
     __attribute__ ((__nothrow__)) __attribute__ ((__nonnull__ (2, 3)));
# 395 "/usr/include/string.h" 3 4


extern size_t strlen (__const char *__s)
     __attribute__ ((__nothrow__)) __attribute__ ((__pure__)) __attribute__ ((__nonnull__ (1)));

# 409 "/usr/include/string.h" 3 4


extern char *strerror (int __errnum) __attribute__ ((__nothrow__));

# 449 "/usr/include/string.h" 3 4
extern void __bzero (void *__s, size_t __n) __attribute__ ((__nothrow__)) __attribute__ ((__nonnull__ (1)));
# 632 "/usr/include/string.h" 3 4
# 1 "/usr/include/bits/string.h" 1 3 4
# 633 "/usr/include/string.h" 2 3 4


# 1 "/usr/include/bits/string2.h" 1 3 4
# 52 "/usr/include/bits/string2.h" 3 4
# 1 "/usr/include/endian.h" 1 3 4
# 37 "/usr/include/endian.h" 3 4
# 1 "/usr/include/bits/endian.h" 1 3 4
# 38 "/usr/include/endian.h" 2 3 4
# 53 "/usr/include/bits/string2.h" 2 3 4
# 1 "/usr/include/bits/types.h" 1 3 4
# 28 "/usr/include/bits/types.h" 3 4
# 1 "/usr/include/bits/wordsize.h" 1 3 4
# 29 "/usr/include/bits/types.h" 2 3 4


typedef unsigned char __u_char;
typedef unsigned short int __u_short;
typedef unsigned int __u_int;
typedef unsigned long int __u_long;


typedef signed char __int8_t;
typedef unsigned char __uint8_t;
typedef signed short int __int16_t;
typedef unsigned short int __uint16_t;
typedef signed int __int32_t;
typedef unsigned int __uint32_t;

typedef signed long int __int64_t;
typedef unsigned long int __uint64_t;







typedef long int __quad_t;
typedef unsigned long int __u_quad_t;
# 131 "/usr/include/bits/types.h" 3 4
# 1 "/usr/include/bits/typesizes.h" 1 3 4
# 132 "/usr/include/bits/types.h" 2 3 4


typedef unsigned long int __dev_t;
typedef unsigned int __uid_t;
typedef unsigned int __gid_t;
typedef unsigned long int __ino_t;
typedef unsigned long int __ino64_t;
typedef unsigned int __mode_t;
typedef unsigned long int __nlink_t;
typedef long int __off_t;
typedef long int __off64_t;
typedef int __pid_t;
typedef struct { int __val[2]; } __fsid_t;
typedef long int __clock_t;
typedef unsigned long int __rlim_t;
typedef unsigned long int __rlim64_t;
typedef unsigned int __id_t;
typedef long int __time_t;
typedef unsigned int __useconds_t;
typedef long int __suseconds_t;

typedef int __daddr_t;
typedef long int __swblk_t;
typedef int __key_t;


typedef int __clockid_t;


typedef void * __timer_t;


typedef long int __blksize_t;




typedef long int __blkcnt_t;
typedef long int __blkcnt64_t;


typedef unsigned long int __fsblkcnt_t;
typedef unsigned long int __fsblkcnt64_t;


typedef unsigned long int __fsfilcnt_t;
typedef unsigned long int __fsfilcnt64_t;

typedef long int __ssize_t;



typedef __off64_t __loff_t;
typedef __quad_t *__qaddr_t;
typedef char *__caddr_t;


typedef long int __intptr_t;


typedef unsigned int __socklen_t;
# 54 "/usr/include/bits/string2.h" 2 3 4
# 394 "/usr/include/bits/string2.h" 3 4
extern void *__rawmemchr (const void *__s, int __c);
# 969 "/usr/include/bits/string2.h" 3 4
extern __inline __attribute__ ((__gnu_inline__)) size_t __strcspn_c1 (__const char *__s, int __reject);
extern __inline __attribute__ ((__gnu_inline__)) size_t
__strcspn_c1 (__const char *__s, int __reject)
{
  register size_t __result = 0;
  while (__s[__result] != '\0' && __s[__result] != __reject)
    ++__result;
  return __result;
}

extern __inline __attribute__ ((__gnu_inline__)) size_t __strcspn_c2 (__const char *__s, int __reject1,
         int __reject2);
extern __inline __attribute__ ((__gnu_inline__)) size_t
__strcspn_c2 (__const char *__s, int __reject1, int __reject2)
{
  register size_t __result = 0;
  while (__s[__result] != '\0' && __s[__result] != __reject1
  && __s[__result] != __reject2)
    ++__result;
  return __result;
}

extern __inline __attribute__ ((__gnu_inline__)) size_t __strcspn_c3 (__const char *__s, int __reject1,
         int __reject2, int __reject3);
extern __inline __attribute__ ((__gnu_inline__)) size_t
__strcspn_c3 (__const char *__s, int __reject1, int __reject2,
       int __reject3)
{
  register size_t __result = 0;
  while (__s[__result] != '\0' && __s[__result] != __reject1
  && __s[__result] != __reject2 && __s[__result] != __reject3)
    ++__result;
  return __result;
}
# 1045 "/usr/include/bits/string2.h" 3 4
extern __inline __attribute__ ((__gnu_inline__)) size_t __strspn_c1 (__const char *__s, int __accept);
extern __inline __attribute__ ((__gnu_inline__)) size_t
__strspn_c1 (__const char *__s, int __accept)
{
  register size_t __result = 0;

  while (__s[__result] == __accept)
    ++__result;
  return __result;
}

extern __inline __attribute__ ((__gnu_inline__)) size_t __strspn_c2 (__const char *__s, int __accept1,
        int __accept2);
extern __inline __attribute__ ((__gnu_inline__)) size_t
__strspn_c2 (__const char *__s, int __accept1, int __accept2)
{
  register size_t __result = 0;

  while (__s[__result] == __accept1 || __s[__result] == __accept2)
    ++__result;
  return __result;
}

extern __inline __attribute__ ((__gnu_inline__)) size_t __strspn_c3 (__const char *__s, int __accept1,
        int __accept2, int __accept3);
extern __inline __attribute__ ((__gnu_inline__)) size_t
__strspn_c3 (__const char *__s, int __accept1, int __accept2, int __accept3)
{
  register size_t __result = 0;

  while (__s[__result] == __accept1 || __s[__result] == __accept2
  || __s[__result] == __accept3)
    ++__result;
  return __result;
}
# 1121 "/usr/include/bits/string2.h" 3 4
extern __inline __attribute__ ((__gnu_inline__)) char *__strpbrk_c2 (__const char *__s, int __accept1,
         int __accept2);
extern __inline __attribute__ ((__gnu_inline__)) char *
__strpbrk_c2 (__const char *__s, int __accept1, int __accept2)
{

  while (*__s != '\0' && *__s != __accept1 && *__s != __accept2)
    ++__s;
  return *__s == '\0' ? ((void *)0) : (char *) (size_t) __s;
}

extern __inline __attribute__ ((__gnu_inline__)) char *__strpbrk_c3 (__const char *__s, int __accept1,
         int __accept2, int __accept3);
extern __inline __attribute__ ((__gnu_inline__)) char *
__strpbrk_c3 (__const char *__s, int __accept1, int __accept2,
       int __accept3)
{

  while (*__s != '\0' && *__s != __accept1 && *__s != __accept2
  && *__s != __accept3)
    ++__s;
  return *__s == '\0' ? ((void *)0) : (char *) (size_t) __s;
}
# 1172 "/usr/include/bits/string2.h" 3 4
extern __inline __attribute__ ((__gnu_inline__)) char *__strtok_r_1c (char *__s, char __sep, char **__nextp);
extern __inline __attribute__ ((__gnu_inline__)) char *
__strtok_r_1c (char *__s, char __sep, char **__nextp)
{
  char *__result;
  if (__s == ((void *)0))
    __s = *__nextp;
  while (*__s == __sep)
    ++__s;
  __result = ((void *)0);
  if (*__s != '\0')
    {
      __result = __s++;
      while (*__s != '\0')
 if (*__s++ == __sep)
   {
     __s[-1] = '\0';
     break;
   }
    }
  *__nextp = __s;
  return __result;
}
# 1204 "/usr/include/bits/string2.h" 3 4
extern char *__strsep_g (char **__stringp, __const char *__delim);
# 1222 "/usr/include/bits/string2.h" 3 4
extern __inline __attribute__ ((__gnu_inline__)) char *__strsep_1c (char **__s, char __reject);
extern __inline __attribute__ ((__gnu_inline__)) char *
__strsep_1c (char **__s, char __reject)
{
  register char *__retval = *__s;
  if (__retval != ((void *)0) && (*__s = (__extension__ (__builtin_constant_p (__reject) && !__builtin_constant_p (__retval) && (__reject) == '\0' ? (char *) __rawmemchr (__retval, __reject) : __builtin_strchr (__retval, __reject)))) != ((void *)0))
    *(*__s)++ = '\0';
  return __retval;
}

extern __inline __attribute__ ((__gnu_inline__)) char *__strsep_2c (char **__s, char __reject1, char __reject2);
extern __inline __attribute__ ((__gnu_inline__)) char *
__strsep_2c (char **__s, char __reject1, char __reject2)
{
  register char *__retval = *__s;
  if (__retval != ((void *)0))
    {
      register char *__cp = __retval;
      while (1)
 {
   if (*__cp == '\0')
     {
       __cp = ((void *)0);
   break;
     }
   if (*__cp == __reject1 || *__cp == __reject2)
     {
       *__cp++ = '\0';
       break;
     }
   ++__cp;
 }
      *__s = __cp;
    }
  return __retval;
}

extern __inline __attribute__ ((__gnu_inline__)) char *__strsep_3c (char **__s, char __reject1, char __reject2,
       char __reject3);
extern __inline __attribute__ ((__gnu_inline__)) char *
__strsep_3c (char **__s, char __reject1, char __reject2, char __reject3)
{
  register char *__retval = *__s;
  if (__retval != ((void *)0))
    {
      register char *__cp = __retval;
      while (1)
 {
   if (*__cp == '\0')
     {
       __cp = ((void *)0);
   break;
     }
   if (*__cp == __reject1 || *__cp == __reject2 || *__cp == __reject3)
     {
       *__cp++ = '\0';
       break;
     }
   ++__cp;
 }
      *__s = __cp;
    }
  return __retval;
}
# 636 "/usr/include/string.h" 2 3 4
# 644 "/usr/include/string.h" 3 4

# 2 "ofb.c" 2
# 1 "/usr/lib/gcc/x86_64-linux-gnu/4.5.2/include/stdint.h" 1 3 4


# 1 "/usr/include/stdint.h" 1 3 4
# 27 "/usr/include/stdint.h" 3 4
# 1 "/usr/include/bits/wchar.h" 1 3 4
# 28 "/usr/include/stdint.h" 2 3 4
# 1 "/usr/include/bits/wordsize.h" 1 3 4
# 29 "/usr/include/stdint.h" 2 3 4
# 37 "/usr/include/stdint.h" 3 4
typedef signed char int8_t;
typedef short int int16_t;
typedef int int32_t;

typedef long int int64_t;







typedef unsigned char uint8_t;
typedef unsigned short int uint16_t;

typedef unsigned int uint32_t;



typedef unsigned long int uint64_t;
# 66 "/usr/include/stdint.h" 3 4
typedef signed char int_least8_t;
typedef short int int_least16_t;
typedef int int_least32_t;

typedef long int int_least64_t;






typedef unsigned char uint_least8_t;
typedef unsigned short int uint_least16_t;
typedef unsigned int uint_least32_t;

typedef unsigned long int uint_least64_t;
# 91 "/usr/include/stdint.h" 3 4
typedef signed char int_fast8_t;

typedef long int int_fast16_t;
typedef long int int_fast32_t;
typedef long int int_fast64_t;
# 104 "/usr/include/stdint.h" 3 4
typedef unsigned char uint_fast8_t;

typedef unsigned long int uint_fast16_t;
typedef unsigned long int uint_fast32_t;
typedef unsigned long int uint_fast64_t;
# 120 "/usr/include/stdint.h" 3 4
typedef long int intptr_t;


typedef unsigned long int uintptr_t;
# 135 "/usr/include/stdint.h" 3 4
typedef long int intmax_t;
typedef unsigned long int uintmax_t;
# 4 "/usr/lib/gcc/x86_64-linux-gnu/4.5.2/include/stdint.h" 2 3 4
# 3 "ofb.c" 2



struct ofb {
 void (*encryptfast)(void *, void *, const void *, size_t);
 uint8_t buf[32] __attribute__((aligned(16)));
 size_t chunks;
};

struct aligned {
 uint8_t data[16] __attribute__((aligned(16)));
};

int ofb_encryptfast(void *ctx, void *outp, const void *inp, size_t len)
{
 struct ofb *c = ctx;
 struct aligned *out = outp;
 const struct aligned *in = inp;
 ((void)sizeof(char[1 - 2*!(sizeof(*out) == 16)]));
 ((void)sizeof(char[1 - 2*!(sizeof(*in) == 16)]));
 ((void)sizeof(char[1 - 2*!(sizeof(out->data) == 16)]));
 ((void)sizeof(char[1 - 2*!(sizeof(in->data) == 16)]));
 ((void)sizeof(char[1 - 2*!(__alignof__(*out) == 16)]));
 ((void)sizeof(char[1 - 2*!(__alignof__(*in) == 16)]));
 ((void)sizeof(char[1 - 2*!(__alignof__(out->data) == 16)]));
 ((void)sizeof(char[1 - 2*!(__alignof__(in->data) == 16)]));
 ((void)sizeof(char[1 - 2*!(__alignof__(c->buf) == 16)]));
 len /= 16;
 for (size_t iters = 0; iters < len; iters++, in++, out++) {
  c->encryptfast(c, c->buf, c->buf, c->chunks);
# 41 "ofb.c"
  for (int i = 0; i < 16; i++)
   out->data[i] = c->buf[i] ^ in->data[i];

 }
 return 0;
}
	.file	"ofb.c"
	.text
	.p2align 4,,15
.globl ofb_encryptfast
	.type	ofb_encryptfast, @function
ofb_encryptfast:
.LFB12:
	.cfi_startproc
	pushq	%r15
	.cfi_def_cfa_offset 16
	movq	%rcx, %r15
	.cfi_offset 15, -16
	shrq	$4, %r15
	pushq	%r14
	.cfi_def_cfa_offset 24
	pushq	%r13
	.cfi_def_cfa_offset 32
	pushq	%r12
	.cfi_def_cfa_offset 40
	pushq	%rbp
	.cfi_def_cfa_offset 48
	pushq	%rbx
	.cfi_def_cfa_offset 56
	movq	%rdi, %rbx
	.cfi_offset 3, -56
	.cfi_offset 6, -48
	.cfi_offset 12, -40
	.cfi_offset 13, -32
	.cfi_offset 14, -24
	subq	$8, %rsp
	.cfi_def_cfa_offset 64
	testq	%r15, %r15
	je	.L2
	leaq	16(%rdi), %r14
	movq	%rsi, %r12
	movq	%rdx, %rbp
	xorl	%r13d, %r13d
	.p2align 4,,10
	.p2align 3
.L3:
	movq	48(%rbx), %rcx
	movq	%r14, %rdx
	movq	%r14, %rsi
	movq	%rbx, %rdi
	addq	$1, %r13
	call	*(%rbx)
	movzbl	0(%rbp), %eax
	xorb	16(%rbx), %al
	movb	%al, (%r12)
	movzbl	1(%rbp), %eax
	xorb	17(%rbx), %al
	movb	%al, 1(%r12)
	movzbl	2(%rbp), %eax
	xorb	18(%rbx), %al
	movb	%al, 2(%r12)
	movzbl	3(%rbp), %eax
	xorb	19(%rbx), %al
	movb	%al, 3(%r12)
	movzbl	4(%rbp), %eax
	xorb	20(%rbx), %al
	movb	%al, 4(%r12)
	movzbl	5(%rbp), %eax
	xorb	21(%rbx), %al
	movb	%al, 5(%r12)
	movzbl	6(%rbp), %eax
	xorb	22(%rbx), %al
	movb	%al, 6(%r12)
	movzbl	7(%rbp), %eax
	xorb	23(%rbx), %al
	movb	%al, 7(%r12)
	movzbl	8(%rbp), %eax
	xorb	24(%rbx), %al
	movb	%al, 8(%r12)
	movzbl	9(%rbp), %eax
	xorb	25(%rbx), %al
	movb	%al, 9(%r12)
	movzbl	10(%rbp), %eax
	xorb	26(%rbx), %al
	movb	%al, 10(%r12)
	movzbl	11(%rbp), %eax
	xorb	27(%rbx), %al
	movb	%al, 11(%r12)
	movzbl	12(%rbp), %eax
	xorb	28(%rbx), %al
	movb	%al, 12(%r12)
	movzbl	13(%rbp), %eax
	xorb	29(%rbx), %al
	movb	%al, 13(%r12)
	movzbl	14(%rbp), %eax
	xorb	30(%rbx), %al
	movb	%al, 14(%r12)
	movzbl	15(%rbp), %eax
	addq	$16, %rbp
	xorb	31(%rbx), %al
	movb	%al, 15(%r12)
	addq	$16, %r12
	cmpq	%r13, %r15
	ja	.L3
.L2:
	addq	$8, %rsp
	.cfi_def_cfa_offset 56
	xorl	%eax, %eax
	popq	%rbx
	.cfi_def_cfa_offset 48
	popq	%rbp
	.cfi_def_cfa_offset 40
	popq	%r12
	.cfi_def_cfa_offset 32
	popq	%r13
	.cfi_def_cfa_offset 24
	popq	%r14
	.cfi_def_cfa_offset 16
	popq	%r15
	.cfi_def_cfa_offset 8
	ret
	.cfi_endproc
.LFE12:
	.size	ofb_encryptfast, .-ofb_encryptfast
	.ident	"GCC: (Debian 4.5.2-6) 4.5.2"
	.section	.note.GNU-stack,"",@progbits

Attachment: signature.asc
Description: Digital signature


Reply to: