c/mirror_zfs
1
0
Fork 0
mirror of https://git.proxmox.com/git/mirror_zfs.git synced 2024-11-18 18:31:00 +03:00
Code Issues Packages Projects Releases Wiki Activity
c54ee4c0d3
mirror_zfs/module/zfs/vdev_raidz_math_ssse3.c

2476 lines
118 KiB
C
Raw Normal View History

SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
/*
* CDDL HEADER START
*
* The contents of this file are subject to the terms of the
* Common Development and Distribution License (the "License").
* You may not use this file except in compliance with the License.
*
* You can obtain a copy of the license at usr/src/OPENSOLARIS.LICENSE
* or http://www.opensolaris.org/os/licensing.
* See the License for the specific language governing permissions
* and limitations under the License.
*
* When distributing Covered Code, include this CDDL HEADER in each
* file and include the License file at usr/src/OPENSOLARIS.LICENSE.
* If applicable, add the following below this CDDL HEADER, with the
* fields enclosed by brackets "[]" replaced with your own identifying
* information: Portions Copyright [yyyy] [name of copyright owner]
*
* CDDL HEADER END
*/
/*
* Copyright (C) 2016 Gvozden Nešković. All rights reserved.
*/
#include <sys/isa_defs.h>
#if defined(__x86_64) && defined(HAVE_SSSE3)
#include <sys/types.h>
#include <linux/simd_x86.h>
#define __asm __asm__ __volatile__
#define _REG_CNT(_0, _1, _2, _3, _4, _5, _6, _7, N, ...) N
#define REG_CNT(r...) _REG_CNT(r, 8, 7, 6, 5, 4, 3, 2, 1)
#define VR0_(REG, ...) "xmm"#REG
#define VR1_(_1, REG, ...) "xmm"#REG
#define VR2_(_1, _2, REG, ...) "xmm"#REG
#define VR3_(_1, _2, _3, REG, ...) "xmm"#REG
#define VR4_(_1, _2, _3, _4, REG, ...) "xmm"#REG
#define VR5_(_1, _2, _3, _4, _5, REG, ...) "xmm"#REG
#define VR6_(_1, _2, _3, _4, _5, _6, REG, ...) "xmm"#REG
#define VR7_(_1, _2, _3, _4, _5, _6, _7, REG, ...) "xmm"#REG
#define VR0(r...) VR0_(r)
#define VR1(r...) VR1_(r)
#define VR2(r...) VR2_(r, 1)
#define VR3(r...) VR3_(r, 1, 2)
Fixes and enhancements of SIMD raidz parity - Implementation lock replaced with atomic variable - Trailing whitespace is removed from user specified parameter, to enhance experience when using commands that add newline, e.g. `echo` - raidz_test: remove dependency on `getrusage()` and RUSAGE_THREAD, Issue #4813 - silence `cppcheck` in vdev_raidz, partial solution of Issue #1392 - Minor fixes and cleanups - Enable use of original parity methods in [fastest] configuration. New opaque original ops structure, representing native methods, is added to supported raidz methods. Original parity methods are executed if selected implementation has NULL fn pointer. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Issue #4813 Issue #1392
2016-07-17 20:41:11 +03:00
#define VR4(r...) VR4_(r, 1, 2)
#define VR5(r...) VR5_(r, 1, 2, 3)
#define VR6(r...) VR6_(r, 1, 2, 3, 4)
#define VR7(r...) VR7_(r, 1, 2, 3, 4, 5)
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
#define R_01(REG1, REG2, ...) REG1, REG2
#define _R_23(_0, _1, REG2, REG3, ...) REG2, REG3
#define R_23(REG...) _R_23(REG, 1, 2, 3)
Linux 5.0 compat: ASM_BUG macro The 5.0 kernel defines the macro ASM_BUG. In order to prevent a conflict and build failure rename ASM_BUG to ZFS_ASM_BUG. This is currently only an issue on aarch64 but all instances of ASM_BUG we're renamed to avoid any future conflict on x86_64. Reviewed-by: Tomohiro Kusumi <kusumi.tomohiro@gmail.com> Reviewed-by: Richard Laager <rlaager@wiktel.com> Reviewed-by: Chris Dunlop <chris@onthe.net.au> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #8725 Issue #8545
2019-05-08 20:18:40 +03:00
#define ZFS_ASM_BUG() ASSERT(0)
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
const uint8_t gf_clmul_mod_lt[4*256][16];
#define ELEM_SIZE 16
typedef struct v {
uint8_t b[ELEM_SIZE] __attribute__((aligned(ELEM_SIZE)));
} v_t;
#define XOR_ACC(src, r...) \
{ \
switch (REG_CNT(r)) { \
case 4: \
__asm( \
"pxor 0x00(%[SRC]), %%" VR0(r) "\n" \
"pxor 0x10(%[SRC]), %%" VR1(r) "\n" \
"pxor 0x20(%[SRC]), %%" VR2(r) "\n" \
"pxor 0x30(%[SRC]), %%" VR3(r) "\n" \
: : [SRC] "r" (src)); \
break; \
case 2: \
__asm( \
"pxor 0x00(%[SRC]), %%" VR0(r) "\n" \
"pxor 0x10(%[SRC]), %%" VR1(r) "\n" \
: : [SRC] "r" (src)); \
break; \
default: \
Linux 5.0 compat: ASM_BUG macro The 5.0 kernel defines the macro ASM_BUG. In order to prevent a conflict and build failure rename ASM_BUG to ZFS_ASM_BUG. This is currently only an issue on aarch64 but all instances of ASM_BUG we're renamed to avoid any future conflict on x86_64. Reviewed-by: Tomohiro Kusumi <kusumi.tomohiro@gmail.com> Reviewed-by: Richard Laager <rlaager@wiktel.com> Reviewed-by: Chris Dunlop <chris@onthe.net.au> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #8725 Issue #8545
2019-05-08 20:18:40 +03:00
ZFS_ASM_BUG(); \
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
} \
}
#define XOR(r...) \
{ \
switch (REG_CNT(r)) { \
case 8: \
__asm( \
"pxor %" VR0(r) ", %" VR4(r) "\n" \
"pxor %" VR1(r) ", %" VR5(r) "\n" \
"pxor %" VR2(r) ", %" VR6(r) "\n" \
"pxor %" VR3(r) ", %" VR7(r)); \
break; \
case 4: \
__asm( \
"pxor %" VR0(r) ", %" VR2(r) "\n" \
"pxor %" VR1(r) ", %" VR3(r)); \
break; \
default: \
Linux 5.0 compat: ASM_BUG macro The 5.0 kernel defines the macro ASM_BUG. In order to prevent a conflict and build failure rename ASM_BUG to ZFS_ASM_BUG. This is currently only an issue on aarch64 but all instances of ASM_BUG we're renamed to avoid any future conflict on x86_64. Reviewed-by: Tomohiro Kusumi <kusumi.tomohiro@gmail.com> Reviewed-by: Richard Laager <rlaager@wiktel.com> Reviewed-by: Chris Dunlop <chris@onthe.net.au> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #8725 Issue #8545
2019-05-08 20:18:40 +03:00
ZFS_ASM_BUG(); \
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
} \
}
ABD raidz avx512f support Implement shift based multiplication for 512f. Higher IPC over lookup based methods yields up to 40% better performance on the current hardware. Results on Xeon Phi(TM) CPU 7210: implementation gen_p gen_pq gen_pqr rec_p rec_q rec_r rec_pq rec_pr rec_qr rec_pqr original 142232671 24411492 12948205 283053705 22348167 4215911 9171609 2265548 2378370 1648495 scalar 295711162 49851491 33253815 293198109 88179448 61866752 27941684 25764416 17384442 12138153 sse2 410055998 199642658 117973654 406240463 152688682 121092250 84968180 79291076 47473657 20779719 ssse3 411641595 199669571 117937647 406211024 137638508 117050346 81263322 76120405 46281559 32696722 avx2 616485806 311515332 188595628 605455115 260602390 230554476 148198817 138800254 92273356 62937819 avx512f 832191523 408509425 253599522 810094481 404325734 317590971 218235687 197204920 133101937 94001219 fastest avx512f avx512f avx512f avx512f avx512f avx512f avx512f avx512f avx512f avx512f Signed-off-by: Gvozden Neskovic <neskovic@gmail.com>
2016-11-20 08:01:31 +03:00
#define ZERO(r...) XOR(r, r)
Add parity generation/rebuild using 128-bits NEON for Aarch64 This re-use the framework established for SSE2, SSSE3 and AVX2. However, GCC is using FP registers on Aarch64, so unlike SSE/AVX2 we can't rely on the registers being left alone between ASM statements. So instead, the NEON code uses C variables and GCC extended ASM syntax. Note that since the kernel explicitly disable vector registers, they have to be locally re-enabled explicitly. As we use the variable's number to define the symbolic name, and GCC won't allow duplicate symbolic names, numbers have to be unique. Even when the code is not going to be used (e.g. the case for 4 registers when using the macro with only 2). Only the actually used variables should be declared, otherwise the build will fails in debug mode. This requires the replacement of the XOR(X,X) syntax by a new ZERO(X) macro, which does the same thing but without repeating the argument. And perhaps someday there will be a machine where there is a more efficient way to zero a register than XOR with itself. This affects scalar, SSE2, SSSE3 and AVX2 as they need the new macro. It's possible to write faster implementations (different scheduling, different unrolling, interleaving NEON and scalar, ...) for various cores, but this one has the advantage of fitting in the current state of the code, and thus is likely easier to review/check/merge. The only difference between aarch64-neon and aarch64-neonx2 is that aarch64-neonx2 unroll some functions some more. Reviewed-by: Gvozden Neskovic <neskovic@gmail.com> Reviewed-by: Brian Behlendorf <behlendorf1@llnl.gov> Signed-off-by: Romain Dolbeau <romain.dolbeau@atos.net> Closes #4801
2016-10-03 19:44:00 +03:00
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
#define COPY(r...) \
{ \
switch (REG_CNT(r)) { \
case 8: \
__asm( \
"movdqa %" VR0(r) ", %" VR4(r) "\n" \
"movdqa %" VR1(r) ", %" VR5(r) "\n" \
"movdqa %" VR2(r) ", %" VR6(r) "\n" \
"movdqa %" VR3(r) ", %" VR7(r)); \
break; \
case 4: \
__asm( \
"movdqa %" VR0(r) ", %" VR2(r) "\n" \
"movdqa %" VR1(r) ", %" VR3(r)); \
break; \
default: \
Linux 5.0 compat: ASM_BUG macro The 5.0 kernel defines the macro ASM_BUG. In order to prevent a conflict and build failure rename ASM_BUG to ZFS_ASM_BUG. This is currently only an issue on aarch64 but all instances of ASM_BUG we're renamed to avoid any future conflict on x86_64. Reviewed-by: Tomohiro Kusumi <kusumi.tomohiro@gmail.com> Reviewed-by: Richard Laager <rlaager@wiktel.com> Reviewed-by: Chris Dunlop <chris@onthe.net.au> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #8725 Issue #8545
2019-05-08 20:18:40 +03:00
ZFS_ASM_BUG(); \
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
} \
}
#define LOAD(src, r...) \
{ \
switch (REG_CNT(r)) { \
case 4: \
__asm( \
"movdqa 0x00(%[SRC]), %%" VR0(r) "\n" \
"movdqa 0x10(%[SRC]), %%" VR1(r) "\n" \
"movdqa 0x20(%[SRC]), %%" VR2(r) "\n" \
"movdqa 0x30(%[SRC]), %%" VR3(r) "\n" \
: : [SRC] "r" (src)); \
break; \
case 2: \
__asm( \
"movdqa 0x00(%[SRC]), %%" VR0(r) "\n" \
"movdqa 0x10(%[SRC]), %%" VR1(r) "\n" \
: : [SRC] "r" (src)); \
break; \
default: \
Linux 5.0 compat: ASM_BUG macro The 5.0 kernel defines the macro ASM_BUG. In order to prevent a conflict and build failure rename ASM_BUG to ZFS_ASM_BUG. This is currently only an issue on aarch64 but all instances of ASM_BUG we're renamed to avoid any future conflict on x86_64. Reviewed-by: Tomohiro Kusumi <kusumi.tomohiro@gmail.com> Reviewed-by: Richard Laager <rlaager@wiktel.com> Reviewed-by: Chris Dunlop <chris@onthe.net.au> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #8725 Issue #8545
2019-05-08 20:18:40 +03:00
ZFS_ASM_BUG(); \
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
} \
}
#define STORE(dst, r...) \
{ \
switch (REG_CNT(r)) { \
case 4: \
__asm( \
"movdqa %%" VR0(r)", 0x00(%[DST])\n" \
"movdqa %%" VR1(r)", 0x10(%[DST])\n" \
"movdqa %%" VR2(r)", 0x20(%[DST])\n" \
"movdqa %%" VR3(r)", 0x30(%[DST])\n" \
: : [DST] "r" (dst)); \
break; \
case 2: \
__asm( \
"movdqa %%" VR0(r)", 0x00(%[DST])\n" \
"movdqa %%" VR1(r)", 0x10(%[DST])\n" \
: : [DST] "r" (dst)); \
break; \
default: \
Linux 5.0 compat: ASM_BUG macro The 5.0 kernel defines the macro ASM_BUG. In order to prevent a conflict and build failure rename ASM_BUG to ZFS_ASM_BUG. This is currently only an issue on aarch64 but all instances of ASM_BUG we're renamed to avoid any future conflict on x86_64. Reviewed-by: Tomohiro Kusumi <kusumi.tomohiro@gmail.com> Reviewed-by: Richard Laager <rlaager@wiktel.com> Reviewed-by: Chris Dunlop <chris@onthe.net.au> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #8725 Issue #8545
2019-05-08 20:18:40 +03:00
ZFS_ASM_BUG(); \
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
} \
}
#define MUL2_SETUP() \
{ \
__asm( \
"movd %[mask], %%xmm15\n" \
"pshufd $0x0, %%xmm15, %%xmm15\n" \
: : [mask] "r" (0x1d1d1d1d)); \
}
#define _MUL2_x2(r...) \
{ \
switch (REG_CNT(r)) { \
case 2: \
__asm( \
"pxor %xmm14, %xmm14\n" \
"pxor %xmm13, %xmm13\n" \
"pcmpgtb %" VR0(r)", %xmm14\n" \
"pcmpgtb %" VR1(r)", %xmm13\n" \
"pand %xmm15, %xmm14\n" \
"pand %xmm15, %xmm13\n" \
"paddb %" VR0(r)", %" VR0(r) "\n" \
"paddb %" VR1(r)", %" VR1(r) "\n" \
"pxor %xmm14, %" VR0(r) "\n" \
"pxor %xmm13, %" VR1(r)); \
break; \
default: \
Linux 5.0 compat: ASM_BUG macro The 5.0 kernel defines the macro ASM_BUG. In order to prevent a conflict and build failure rename ASM_BUG to ZFS_ASM_BUG. This is currently only an issue on aarch64 but all instances of ASM_BUG we're renamed to avoid any future conflict on x86_64. Reviewed-by: Tomohiro Kusumi <kusumi.tomohiro@gmail.com> Reviewed-by: Richard Laager <rlaager@wiktel.com> Reviewed-by: Chris Dunlop <chris@onthe.net.au> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #8725 Issue #8545
2019-05-08 20:18:40 +03:00
ZFS_ASM_BUG(); \
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
} \
}
#define MUL2(r...) \
{ \
switch (REG_CNT(r)) { \
case 4: \
_MUL2_x2(R_01(r)); \
_MUL2_x2(R_23(r)); \
break; \
case 2: \
_MUL2_x2(r); \
break; \
default: \
Linux 5.0 compat: ASM_BUG macro The 5.0 kernel defines the macro ASM_BUG. In order to prevent a conflict and build failure rename ASM_BUG to ZFS_ASM_BUG. This is currently only an issue on aarch64 but all instances of ASM_BUG we're renamed to avoid any future conflict on x86_64. Reviewed-by: Tomohiro Kusumi <kusumi.tomohiro@gmail.com> Reviewed-by: Richard Laager <rlaager@wiktel.com> Reviewed-by: Chris Dunlop <chris@onthe.net.au> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #8725 Issue #8545
2019-05-08 20:18:40 +03:00
ZFS_ASM_BUG(); \
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
} \
}
#define MUL4(r...) \
{ \
MUL2(r); \
MUL2(r); \
}
#define _0f "xmm15"
#define _a_save "xmm14"
#define _b_save "xmm13"
#define _lt_mod_a "xmm12"
#define _lt_clmul_a "xmm11"
#define _lt_mod_b "xmm10"
#define _lt_clmul_b "xmm15"
#define _MULx2(c, r...) \
{ \
switch (REG_CNT(r)) { \
case 2: \
__asm( \
/* lts for upper part */ \
"movd %[mask], %%" _0f "\n" \
"pshufd $0x0, %%" _0f ", %%" _0f "\n" \
"movdqa 0x00(%[lt]), %%" _lt_mod_a "\n" \
"movdqa 0x10(%[lt]), %%" _lt_clmul_a "\n" \
/* upper part */ \
"movdqa %%" VR0(r) ", %%" _a_save "\n" \
"movdqa %%" VR1(r) ", %%" _b_save "\n" \
"psraw $0x4, %%" VR0(r) "\n" \
"psraw $0x4, %%" VR1(r) "\n" \
"pand %%" _0f ", %%" _a_save "\n" \
"pand %%" _0f ", %%" _b_save "\n" \
"pand %%" _0f ", %%" VR0(r) "\n" \
"pand %%" _0f ", %%" VR1(r) "\n" \
\
"movdqa %%" _lt_mod_a ", %%" _lt_mod_b "\n" \
"movdqa %%" _lt_clmul_a ", %%" _lt_clmul_b "\n" \
\
"pshufb %%" VR0(r) ",%%" _lt_mod_a "\n" \
"pshufb %%" VR1(r) ",%%" _lt_mod_b "\n" \
"pshufb %%" VR0(r) ",%%" _lt_clmul_a "\n" \
"pshufb %%" VR1(r) ",%%" _lt_clmul_b "\n" \
\
"pxor %%" _lt_mod_a ",%%" _lt_clmul_a "\n" \
"pxor %%" _lt_mod_b ",%%" _lt_clmul_b "\n" \
"movdqa %%" _lt_clmul_a ",%%" VR0(r) "\n" \
"movdqa %%" _lt_clmul_b ",%%" VR1(r) "\n" \
/* lts for lower part */ \
"movdqa 0x20(%[lt]), %%" _lt_mod_a "\n" \
"movdqa 0x30(%[lt]), %%" _lt_clmul_a "\n" \
"movdqa %%" _lt_mod_a ", %%" _lt_mod_b "\n" \
"movdqa %%" _lt_clmul_a ", %%" _lt_clmul_b "\n" \
/* lower part */ \
"pshufb %%" _a_save ",%%" _lt_mod_a "\n" \
"pshufb %%" _b_save ",%%" _lt_mod_b "\n" \
"pshufb %%" _a_save ",%%" _lt_clmul_a "\n" \
"pshufb %%" _b_save ",%%" _lt_clmul_b "\n" \
\
"pxor %%" _lt_mod_a ",%%" VR0(r) "\n" \
"pxor %%" _lt_mod_b ",%%" VR1(r) "\n" \
"pxor %%" _lt_clmul_a ",%%" VR0(r) "\n" \
"pxor %%" _lt_clmul_b ",%%" VR1(r) "\n" \
: : [mask] "r" (0x0f0f0f0f), \
[lt] "r" (gf_clmul_mod_lt[4*(c)])); \
break; \
default: \
Linux 5.0 compat: ASM_BUG macro The 5.0 kernel defines the macro ASM_BUG. In order to prevent a conflict and build failure rename ASM_BUG to ZFS_ASM_BUG. This is currently only an issue on aarch64 but all instances of ASM_BUG we're renamed to avoid any future conflict on x86_64. Reviewed-by: Tomohiro Kusumi <kusumi.tomohiro@gmail.com> Reviewed-by: Richard Laager <rlaager@wiktel.com> Reviewed-by: Chris Dunlop <chris@onthe.net.au> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #8725 Issue #8545
2019-05-08 20:18:40 +03:00
ZFS_ASM_BUG(); \
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
} \
}
#define MUL(c, r...) \
{ \
switch (REG_CNT(r)) { \
case 4: \
_MULx2(c, R_23(r)); \
_MULx2(c, R_01(r)); \
break; \
case 2: \
_MULx2(c, R_01(r)); \
break; \
default: \
Linux 5.0 compat: ASM_BUG macro The 5.0 kernel defines the macro ASM_BUG. In order to prevent a conflict and build failure rename ASM_BUG to ZFS_ASM_BUG. This is currently only an issue on aarch64 but all instances of ASM_BUG we're renamed to avoid any future conflict on x86_64. Reviewed-by: Tomohiro Kusumi <kusumi.tomohiro@gmail.com> Reviewed-by: Richard Laager <rlaager@wiktel.com> Reviewed-by: Chris Dunlop <chris@onthe.net.au> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #8725 Issue #8545
2019-05-08 20:18:40 +03:00
ZFS_ASM_BUG(); \
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
} \
}
#define raidz_math_begin() kfpu_begin()
#define raidz_math_end() kfpu_end()
ABD Vectorized raidz Enable vectorized raidz code on ABD buffers. The avx512f, avx512bw, neon and aarch64_neonx2 are disabled in this commit. With the exception of avx512bw these implementations are updated for ABD in the subsequent commits. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com>
2016-08-24 16:51:33 +03:00
#define SYN_STRIDE 4
#define ZERO_STRIDE 4
#define ZERO_DEFINE() {}
#define ZERO_D 0, 1, 2, 3
#define COPY_STRIDE 4
#define COPY_DEFINE() {}
#define COPY_D 0, 1, 2, 3
#define ADD_STRIDE 4
#define ADD_DEFINE() {}
#define ADD_D 0, 1, 2, 3
#define MUL_STRIDE 4
#define MUL_DEFINE() {}
#define MUL_D 0, 1, 2, 3
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
#define GEN_P_STRIDE 4
ABD Vectorized raidz Enable vectorized raidz code on ABD buffers. The avx512f, avx512bw, neon and aarch64_neonx2 are disabled in this commit. With the exception of avx512bw these implementations are updated for ABD in the subsequent commits. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com>
2016-08-24 16:51:33 +03:00
#define GEN_P_DEFINE() {}
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
#define GEN_P_P 0, 1, 2, 3
#define GEN_PQ_STRIDE 4
ABD Vectorized raidz Enable vectorized raidz code on ABD buffers. The avx512f, avx512bw, neon and aarch64_neonx2 are disabled in this commit. With the exception of avx512bw these implementations are updated for ABD in the subsequent commits. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com>
2016-08-24 16:51:33 +03:00
#define GEN_PQ_DEFINE() {}
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
#define GEN_PQ_D 0, 1, 2, 3
ABD Vectorized raidz Enable vectorized raidz code on ABD buffers. The avx512f, avx512bw, neon and aarch64_neonx2 are disabled in this commit. With the exception of avx512bw these implementations are updated for ABD in the subsequent commits. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com>
2016-08-24 16:51:33 +03:00
#define GEN_PQ_C 4, 5, 6, 7
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
ABD Vectorized raidz Enable vectorized raidz code on ABD buffers. The avx512f, avx512bw, neon and aarch64_neonx2 are disabled in this commit. With the exception of avx512bw these implementations are updated for ABD in the subsequent commits. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com>
2016-08-24 16:51:33 +03:00
#define GEN_PQR_STRIDE 4
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
#define GEN_PQR_DEFINE() {}
ABD Vectorized raidz Enable vectorized raidz code on ABD buffers. The avx512f, avx512bw, neon and aarch64_neonx2 are disabled in this commit. With the exception of avx512bw these implementations are updated for ABD in the subsequent commits. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com>
2016-08-24 16:51:33 +03:00
#define GEN_PQR_D 0, 1, 2, 3
#define GEN_PQR_C 4, 5, 6, 7
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
ABD Vectorized raidz Enable vectorized raidz code on ABD buffers. The avx512f, avx512bw, neon and aarch64_neonx2 are disabled in this commit. With the exception of avx512bw these implementations are updated for ABD in the subsequent commits. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com>
2016-08-24 16:51:33 +03:00
#define SYN_Q_DEFINE() {}
#define SYN_Q_D 0, 1, 2, 3
#define SYN_Q_X 4, 5, 6, 7
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
ABD Vectorized raidz Enable vectorized raidz code on ABD buffers. The avx512f, avx512bw, neon and aarch64_neonx2 are disabled in this commit. With the exception of avx512bw these implementations are updated for ABD in the subsequent commits. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com>
2016-08-24 16:51:33 +03:00
#define SYN_R_DEFINE() {}
#define SYN_R_D 0, 1, 2, 3
#define SYN_R_X 4, 5, 6, 7
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
ABD Vectorized raidz Enable vectorized raidz code on ABD buffers. The avx512f, avx512bw, neon and aarch64_neonx2 are disabled in this commit. With the exception of avx512bw these implementations are updated for ABD in the subsequent commits. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com>
2016-08-24 16:51:33 +03:00
#define SYN_PQ_DEFINE() {}
#define SYN_PQ_D 0, 1, 2, 3
#define SYN_PQ_X 4, 5, 6, 7
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
#define REC_PQ_STRIDE 2
ABD Vectorized raidz Enable vectorized raidz code on ABD buffers. The avx512f, avx512bw, neon and aarch64_neonx2 are disabled in this commit. With the exception of avx512bw these implementations are updated for ABD in the subsequent commits. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com>
2016-08-24 16:51:33 +03:00
#define REC_PQ_DEFINE() {}
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
#define REC_PQ_X 0, 1
#define REC_PQ_Y 2, 3
ABD Vectorized raidz Enable vectorized raidz code on ABD buffers. The avx512f, avx512bw, neon and aarch64_neonx2 are disabled in this commit. With the exception of avx512bw these implementations are updated for ABD in the subsequent commits. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com>
2016-08-24 16:51:33 +03:00
#define REC_PQ_T 4, 5
#define SYN_PR_DEFINE() {}
#define SYN_PR_D 0, 1, 2, 3
#define SYN_PR_X 4, 5, 6, 7
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
#define REC_PR_STRIDE 2
ABD Vectorized raidz Enable vectorized raidz code on ABD buffers. The avx512f, avx512bw, neon and aarch64_neonx2 are disabled in this commit. With the exception of avx512bw these implementations are updated for ABD in the subsequent commits. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com>
2016-08-24 16:51:33 +03:00
#define REC_PR_DEFINE() {}
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
#define REC_PR_X 0, 1
#define REC_PR_Y 2, 3
ABD Vectorized raidz Enable vectorized raidz code on ABD buffers. The avx512f, avx512bw, neon and aarch64_neonx2 are disabled in this commit. With the exception of avx512bw these implementations are updated for ABD in the subsequent commits. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com>
2016-08-24 16:51:33 +03:00
#define REC_PR_T 4, 5
#define SYN_QR_DEFINE() {}
#define SYN_QR_D 0, 1, 2, 3
#define SYN_QR_X 4, 5, 6, 7
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
#define REC_QR_STRIDE 2
ABD Vectorized raidz Enable vectorized raidz code on ABD buffers. The avx512f, avx512bw, neon and aarch64_neonx2 are disabled in this commit. With the exception of avx512bw these implementations are updated for ABD in the subsequent commits. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com>
2016-08-24 16:51:33 +03:00
#define REC_QR_DEFINE() {}
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
#define REC_QR_X 0, 1
#define REC_QR_Y 2, 3
ABD Vectorized raidz Enable vectorized raidz code on ABD buffers. The avx512f, avx512bw, neon and aarch64_neonx2 are disabled in this commit. With the exception of avx512bw these implementations are updated for ABD in the subsequent commits. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com>
2016-08-24 16:51:33 +03:00
#define REC_QR_T 4, 5
#define SYN_PQR_DEFINE() {}
#define SYN_PQR_D 0, 1, 2, 3
#define SYN_PQR_X 4, 5, 6, 7
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
#define REC_PQR_STRIDE 2
ABD Vectorized raidz Enable vectorized raidz code on ABD buffers. The avx512f, avx512bw, neon and aarch64_neonx2 are disabled in this commit. With the exception of avx512bw these implementations are updated for ABD in the subsequent commits. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com>
2016-08-24 16:51:33 +03:00
#define REC_PQR_DEFINE() {}
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
#define REC_PQR_X 0, 1
#define REC_PQR_Y 2, 3
#define REC_PQR_Z 4, 5
#define REC_PQR_XS 6, 7
#define REC_PQR_YS 8, 9
#include <sys/vdev_raidz_impl.h>
#include "vdev_raidz_math_impl.h"
Add RAID-Z routines for SSE2 instruction set, in x86_64 mode. The patch covers low-end and older x86 CPUs. Parity generation is equivalent to SSSE3 implementation, but reconstruction is somewhat slower. Previous 'sse' implementation is renamed to 'ssse3' to indicate highest instruction set used. Benchmark results: scalar_rec_p 4 720476442 scalar_rec_q 4 187462804 scalar_rec_r 4 138996096 scalar_rec_pq 4 140834951 scalar_rec_pr 4 129332035 scalar_rec_qr 4 81619194 scalar_rec_pqr 4 53376668 sse2_rec_p 4 2427757064 sse2_rec_q 4 747120861 sse2_rec_r 4 499871637 sse2_rec_pq 4 522403710 sse2_rec_pr 4 464632780 sse2_rec_qr 4 319124434 sse2_rec_pqr 4 205794190 ssse3_rec_p 4 2519939444 ssse3_rec_q 4 1003019289 ssse3_rec_r 4 616428767 ssse3_rec_pq 4 706326396 ssse3_rec_pr 4 570493618 ssse3_rec_qr 4 400185250 ssse3_rec_pqr 4 377541245 original_rec_p 4 691658568 original_rec_q 4 195510948 original_rec_r 4 26075538 original_rec_pq 4 103087368 original_rec_pr 4 15767058 original_rec_qr 4 15513175 original_rec_pqr 4 10746357 Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4783
2016-06-28 20:49:53 +03:00
DEFINE_GEN_METHODS(ssse3);
DEFINE_REC_METHODS(ssse3);
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
static boolean_t
Add RAID-Z routines for SSE2 instruction set, in x86_64 mode. The patch covers low-end and older x86 CPUs. Parity generation is equivalent to SSSE3 implementation, but reconstruction is somewhat slower. Previous 'sse' implementation is renamed to 'ssse3' to indicate highest instruction set used. Benchmark results: scalar_rec_p 4 720476442 scalar_rec_q 4 187462804 scalar_rec_r 4 138996096 scalar_rec_pq 4 140834951 scalar_rec_pr 4 129332035 scalar_rec_qr 4 81619194 scalar_rec_pqr 4 53376668 sse2_rec_p 4 2427757064 sse2_rec_q 4 747120861 sse2_rec_r 4 499871637 sse2_rec_pq 4 522403710 sse2_rec_pr 4 464632780 sse2_rec_qr 4 319124434 sse2_rec_pqr 4 205794190 ssse3_rec_p 4 2519939444 ssse3_rec_q 4 1003019289 ssse3_rec_r 4 616428767 ssse3_rec_pq 4 706326396 ssse3_rec_pr 4 570493618 ssse3_rec_qr 4 400185250 ssse3_rec_pqr 4 377541245 original_rec_p 4 691658568 original_rec_q 4 195510948 original_rec_r 4 26075538 original_rec_pq 4 103087368 original_rec_pr 4 15767058 original_rec_qr 4 15513175 original_rec_pqr 4 10746357 Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4783
2016-06-28 20:49:53 +03:00
raidz_will_ssse3_work(void)
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
{
Linux 5.0 compat: SIMD compatibility Restore the SIMD optimization for 4.19.38 LTS, 4.14.120 LTS, and 5.0 and newer kernels. This commit squashes the following commits from master in to a single commit which can be applied to 0.8.2. 10fa2545 - Linux 4.14, 4.19, 5.0+ compat: SIMD save/restore b88ca2ac - Enable SIMD for encryption 095b5412 - Fix CONFIG_X86_DEBUG_FPU build failure e5db3134 - Linux 5.0 compat: SIMD compatibility Reviewed-by: Fabian Grünbichler <f.gruenbichler@proxmox.com> Reviewed-by: Tony Hutter <hutter2@llnl.gov> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> TEST_ZIMPORT_SKIP="yes"
2019-07-12 19:31:20 +03:00
return (kfpu_allowed() && zfs_sse_available() &&
zfs_sse2_available() && zfs_ssse3_available());
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
}
Add RAID-Z routines for SSE2 instruction set, in x86_64 mode. The patch covers low-end and older x86 CPUs. Parity generation is equivalent to SSSE3 implementation, but reconstruction is somewhat slower. Previous 'sse' implementation is renamed to 'ssse3' to indicate highest instruction set used. Benchmark results: scalar_rec_p 4 720476442 scalar_rec_q 4 187462804 scalar_rec_r 4 138996096 scalar_rec_pq 4 140834951 scalar_rec_pr 4 129332035 scalar_rec_qr 4 81619194 scalar_rec_pqr 4 53376668 sse2_rec_p 4 2427757064 sse2_rec_q 4 747120861 sse2_rec_r 4 499871637 sse2_rec_pq 4 522403710 sse2_rec_pr 4 464632780 sse2_rec_qr 4 319124434 sse2_rec_pqr 4 205794190 ssse3_rec_p 4 2519939444 ssse3_rec_q 4 1003019289 ssse3_rec_r 4 616428767 ssse3_rec_pq 4 706326396 ssse3_rec_pr 4 570493618 ssse3_rec_qr 4 400185250 ssse3_rec_pqr 4 377541245 original_rec_p 4 691658568 original_rec_q 4 195510948 original_rec_r 4 26075538 original_rec_pq 4 103087368 original_rec_pr 4 15767058 original_rec_qr 4 15513175 original_rec_pqr 4 10746357 Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4783
2016-06-28 20:49:53 +03:00
const raidz_impl_ops_t vdev_raidz_ssse3_impl = {
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
.init = NULL,
.fini = NULL,
Add RAID-Z routines for SSE2 instruction set, in x86_64 mode. The patch covers low-end and older x86 CPUs. Parity generation is equivalent to SSSE3 implementation, but reconstruction is somewhat slower. Previous 'sse' implementation is renamed to 'ssse3' to indicate highest instruction set used. Benchmark results: scalar_rec_p 4 720476442 scalar_rec_q 4 187462804 scalar_rec_r 4 138996096 scalar_rec_pq 4 140834951 scalar_rec_pr 4 129332035 scalar_rec_qr 4 81619194 scalar_rec_pqr 4 53376668 sse2_rec_p 4 2427757064 sse2_rec_q 4 747120861 sse2_rec_r 4 499871637 sse2_rec_pq 4 522403710 sse2_rec_pr 4 464632780 sse2_rec_qr 4 319124434 sse2_rec_pqr 4 205794190 ssse3_rec_p 4 2519939444 ssse3_rec_q 4 1003019289 ssse3_rec_r 4 616428767 ssse3_rec_pq 4 706326396 ssse3_rec_pr 4 570493618 ssse3_rec_qr 4 400185250 ssse3_rec_pqr 4 377541245 original_rec_p 4 691658568 original_rec_q 4 195510948 original_rec_r 4 26075538 original_rec_pq 4 103087368 original_rec_pr 4 15767058 original_rec_qr 4 15513175 original_rec_pqr 4 10746357 Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4783
2016-06-28 20:49:53 +03:00
.gen = RAIDZ_GEN_METHODS(ssse3),
.rec = RAIDZ_REC_METHODS(ssse3),
.is_supported = &raidz_will_ssse3_work,
.name = "ssse3"
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
};
#endif /* defined(__x86_64) && defined(HAVE_SSSE3) */
ABD raidz avx512f support Implement shift based multiplication for 512f. Higher IPC over lookup based methods yields up to 40% better performance on the current hardware. Results on Xeon Phi(TM) CPU 7210: implementation gen_p gen_pq gen_pqr rec_p rec_q rec_r rec_pq rec_pr rec_qr rec_pqr original 142232671 24411492 12948205 283053705 22348167 4215911 9171609 2265548 2378370 1648495 scalar 295711162 49851491 33253815 293198109 88179448 61866752 27941684 25764416 17384442 12138153 sse2 410055998 199642658 117973654 406240463 152688682 121092250 84968180 79291076 47473657 20779719 ssse3 411641595 199669571 117937647 406211024 137638508 117050346 81263322 76120405 46281559 32696722 avx2 616485806 311515332 188595628 605455115 260602390 230554476 148198817 138800254 92273356 62937819 avx512f 832191523 408509425 253599522 810094481 404325734 317590971 218235687 197204920 133101937 94001219 fastest avx512f avx512f avx512f avx512f avx512f avx512f avx512f avx512f avx512f avx512f Signed-off-by: Gvozden Neskovic <neskovic@gmail.com>
2016-11-20 08:01:31 +03:00
#if defined(__x86_64)
#if defined(HAVE_SSSE3) || defined(HAVE_AVX2) || defined(HAVE_AVX512BW)
codebase style improvements for OpenZFS 6459 port
2017-01-21 00:17:55 +03:00
/* BEGIN CSTYLED */
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
const uint8_t
codebase style improvements for OpenZFS 6459 port
2017-01-21 00:17:55 +03:00
__attribute__((aligned(256))) gf_clmul_mod_lt[4*256][16] =
{
SIMD implementation of vdev_raidz generate and reconstruct routines This is a new implementation of RAIDZ1/2/3 routines using x86_64 scalar, SSE, and AVX2 instruction sets. Included are 3 parity generation routines (P, PQ, and PQR) and 7 reconstruction routines, for all RAIDZ level. On module load, a quick benchmark of supported routines will select the fastest for each operation and they will be used at runtime. Original implementation is still present and can be selected via module parameter. Patch contains: - specialized gen/rec routines for all RAIDZ levels, - new scalar raidz implementation (unrolled), - two x86_64 SIMD implementations (SSE and AVX2 instructions sets), - fastest routines selected on module load (benchmark). - cmd/raidz_test - verify and benchmark all implementations - added raidz_test to the ZFS Test Suite New zfs module parameters: - zfs_vdev_raidz_impl (str): selects the implementation to use. On module load, the parameter will only accept first 3 options, and the other implementations can be set once module is finished loading. Possible values for this option are: "fastest" - use the fastest math available "original" - use the original raidz code "scalar" - new scalar impl "sse" - new SSE impl if available "avx2" - new AVX2 impl if available See contents of `/sys/module/zfs/parameters/zfs_vdev_raidz_impl` to get the list of supported values. If an implementation is not supported on the system, it will not be shown. Currently selected option is enclosed in `[]`. Signed-off-by: Gvozden Neskovic <neskovic@gmail.com> Signed-off-by: Brian Behlendorf <behlendorf1@llnl.gov> Closes #4328
2016-04-25 11:04:31 +03:00
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x10, 0x20, 0x30, 0x40, 0x50, 0x60, 0x70,
0x80, 0x90, 0xa0, 0xb0, 0xc0, 0xd0, 0xe0, 0xf0 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07,
0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0,
0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x02, 0x04, 0x06, 0x08, 0x0a, 0x0c, 0x0e,
0x10, 0x12, 0x14, 0x16, 0x18, 0x1a, 0x1c, 0x1e },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x30, 0x60, 0x50, 0xc0, 0xf0, 0xa0, 0x90,
0x80, 0xb0, 0xe0, 0xd0, 0x40, 0x70, 0x20, 0x10 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x03, 0x06, 0x05, 0x0c, 0x0f, 0x0a, 0x09,
0x18, 0x1b, 0x1e, 0x1d, 0x14, 0x17, 0x12, 0x11 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0,
0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x04, 0x08, 0x0c, 0x10, 0x14, 0x18, 0x1c,
0x20, 0x24, 0x28, 0x2c, 0x30, 0x34, 0x38, 0x3c },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x50, 0xa0, 0xf0, 0x40, 0x10, 0xe0, 0xb0,
0x80, 0xd0, 0x20, 0x70, 0xc0, 0x90, 0x60, 0x30 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x05, 0x0a, 0x0f, 0x14, 0x11, 0x1e, 0x1b,
0x28, 0x2d, 0x22, 0x27, 0x3c, 0x39, 0x36, 0x33 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20,
0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x06, 0x0c, 0x0a, 0x18, 0x1e, 0x14, 0x12,
0x30, 0x36, 0x3c, 0x3a, 0x28, 0x2e, 0x24, 0x22 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x70, 0xe0, 0x90, 0xc0, 0xb0, 0x20, 0x50,
0x80, 0xf0, 0x60, 0x10, 0x40, 0x30, 0xa0, 0xd0 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x07, 0x0e, 0x09, 0x1c, 0x1b, 0x12, 0x15,
0x38, 0x3f, 0x36, 0x31, 0x24, 0x23, 0x2a, 0x2d },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80,
0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x08, 0x10, 0x18, 0x20, 0x28, 0x30, 0x38,
0x40, 0x48, 0x50, 0x58, 0x60, 0x68, 0x70, 0x78 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x90, 0x20, 0xb0, 0x40, 0xd0, 0x60, 0xf0,
0x80, 0x10, 0xa0, 0x30, 0xc0, 0x50, 0xe0, 0x70 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x09, 0x12, 0x1b, 0x24, 0x2d, 0x36, 0x3f,
0x48, 0x41, 0x5a, 0x53, 0x6c, 0x65, 0x7e, 0x77 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60,
0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x0a, 0x14, 0x1e, 0x28, 0x22, 0x3c, 0x36,
0x50, 0x5a, 0x44, 0x4e, 0x78, 0x72, 0x6c, 0x66 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xb0, 0x60, 0xd0, 0xc0, 0x70, 0xa0, 0x10,
0x80, 0x30, 0xe0, 0x50, 0x40, 0xf0, 0x20, 0x90 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x0b, 0x16, 0x1d, 0x2c, 0x27, 0x3a, 0x31,
0x58, 0x53, 0x4e, 0x45, 0x74, 0x7f, 0x62, 0x69 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40,
0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x0c, 0x18, 0x14, 0x30, 0x3c, 0x28, 0x24,
0x60, 0x6c, 0x78, 0x74, 0x50, 0x5c, 0x48, 0x44 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xd0, 0xa0, 0x70, 0x40, 0x90, 0xe0, 0x30,
0x80, 0x50, 0x20, 0xf0, 0xc0, 0x10, 0x60, 0xb0 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x0d, 0x1a, 0x17, 0x34, 0x39, 0x2e, 0x23,
0x68, 0x65, 0x72, 0x7f, 0x5c, 0x51, 0x46, 0x4b },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0,
0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x0e, 0x1c, 0x12, 0x38, 0x36, 0x24, 0x2a,
0x70, 0x7e, 0x6c, 0x62, 0x48, 0x46, 0x54, 0x5a },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xf0, 0xe0, 0x10, 0xc0, 0x30, 0x20, 0xd0,
0x80, 0x70, 0x60, 0x90, 0x40, 0xb0, 0xa0, 0x50 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x0f, 0x1e, 0x11, 0x3c, 0x33, 0x22, 0x2d,
0x78, 0x77, 0x66, 0x69, 0x44, 0x4b, 0x5a, 0x55 },
{ 0x00, 0x1d, 0x3a, 0x27, 0x74, 0x69, 0x4e, 0x53,
0xe8, 0xf5, 0xd2, 0xcf, 0x9c, 0x81, 0xa6, 0xbb },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x10, 0x20, 0x30, 0x40, 0x50, 0x60, 0x70,
0x80, 0x90, 0xa0, 0xb0, 0xc0, 0xd0, 0xe0, 0xf0 },
{ 0x00, 0x1d, 0x3a, 0x27, 0x74, 0x69, 0x4e, 0x53,
0xe8, 0xf5, 0xd2, 0xcf, 0x9c, 0x81, 0xa6, 0xbb },
{ 0x00, 0x10, 0x20, 0x30, 0x40, 0x50, 0x60, 0x70,
0x80, 0x90, 0xa0, 0xb0, 0xc0, 0xd0, 0xe0, 0xf0 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x11, 0x22, 0x33, 0x44, 0x55, 0x66, 0x77,
0x88, 0x99, 0xaa, 0xbb, 0xcc, 0xdd, 0xee, 0xff },
{ 0x00, 0x1d, 0x3a, 0x27, 0x74, 0x69, 0x4e, 0x53,
0xf5, 0xe8, 0xcf, 0xd2, 0x81, 0x9c, 0xbb, 0xa6 },
{ 0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0,
0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x12, 0x24, 0x36, 0x48, 0x5a, 0x6c, 0x7e,
0x90, 0x82, 0xb4, 0xa6, 0xd8, 0xca, 0xfc, 0xee },
{ 0x00, 0x1d, 0x3a, 0x27, 0x74, 0x69, 0x4e, 0x53,
0xf5, 0xe8, 0xcf, 0xd2, 0x81, 0x9c, 0xbb, 0xa6 },
{ 0x00, 0x30, 0x60, 0x50, 0xc0, 0xf0, 0xa0, 0x90,
0x80, 0xb0, 0xe0, 0xd0, 0x40, 0x70, 0x20, 0x10 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x13, 0x26, 0x35, 0x4c, 0x5f, 0x6a, 0x79,
0x98, 0x8b, 0xbe, 0xad, 0xd4, 0xc7, 0xf2, 0xe1 },
{ 0x00, 0x1d, 0x3a, 0x27, 0x69, 0x74, 0x53, 0x4e,
0xd2, 0xcf, 0xe8, 0xf5, 0xbb, 0xa6, 0x81, 0x9c },
{ 0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0,
0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x14, 0x28, 0x3c, 0x50, 0x44, 0x78, 0x6c,
0xa0, 0xb4, 0x88, 0x9c, 0xf0, 0xe4, 0xd8, 0xcc },
{ 0x00, 0x1d, 0x3a, 0x27, 0x69, 0x74, 0x53, 0x4e,
0xd2, 0xcf, 0xe8, 0xf5, 0xbb, 0xa6, 0x81, 0x9c },
{ 0x00, 0x50, 0xa0, 0xf0, 0x40, 0x10, 0xe0, 0xb0,
0x80, 0xd0, 0x20, 0x70, 0xc0, 0x90, 0x60, 0x30 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x15, 0x2a, 0x3f, 0x54, 0x41, 0x7e, 0x6b,
0xa8, 0xbd, 0x82, 0x97, 0xfc, 0xe9, 0xd6, 0xc3 },
{ 0x00, 0x1d, 0x3a, 0x27, 0x69, 0x74, 0x53, 0x4e,
0xcf, 0xd2, 0xf5, 0xe8, 0xa6, 0xbb, 0x9c, 0x81 },
{ 0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20,
0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x16, 0x2c, 0x3a, 0x58, 0x4e, 0x74, 0x62,
0xb0, 0xa6, 0x9c, 0x8a, 0xe8, 0xfe, 0xc4, 0xd2 },
{ 0x00, 0x1d, 0x3a, 0x27, 0x69, 0x74, 0x53, 0x4e,
0xcf, 0xd2, 0xf5, 0xe8, 0xa6, 0xbb, 0x9c, 0x81 },
{ 0x00, 0x70, 0xe0, 0x90, 0xc0, 0xb0, 0x20, 0x50,
0x80, 0xf0, 0x60, 0x10, 0x40, 0x30, 0xa0, 0xd0 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x17, 0x2e, 0x39, 0x5c, 0x4b, 0x72, 0x65,
0xb8, 0xaf, 0x96, 0x81, 0xe4, 0xf3, 0xca, 0xdd },
{ 0x00, 0x1d, 0x27, 0x3a, 0x4e, 0x53, 0x69, 0x74,
0x9c, 0x81, 0xbb, 0xa6, 0xd2, 0xcf, 0xf5, 0xe8 },
{ 0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80,
0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x18, 0x30, 0x28, 0x60, 0x78, 0x50, 0x48,
0xc0, 0xd8, 0xf0, 0xe8, 0xa0, 0xb8, 0x90, 0x88 },
{ 0x00, 0x1d, 0x27, 0x3a, 0x4e, 0x53, 0x69, 0x74,
0x9c, 0x81, 0xbb, 0xa6, 0xd2, 0xcf, 0xf5, 0xe8 },
{ 0x00, 0x90, 0x20, 0xb0, 0x40, 0xd0, 0x60, 0xf0,
0x80, 0x10, 0xa0, 0x30, 0xc0, 0x50, 0xe0, 0x70 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x19, 0x32, 0x2b, 0x64, 0x7d, 0x56, 0x4f,
0xc8, 0xd1, 0xfa, 0xe3, 0xac, 0xb5, 0x9e, 0x87 },
{ 0x00, 0x1d, 0x27, 0x3a, 0x4e, 0x53, 0x69, 0x74,
0x81, 0x9c, 0xa6, 0xbb, 0xcf, 0xd2, 0xe8, 0xf5 },
{ 0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60,
0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x1a, 0x34, 0x2e, 0x68, 0x72, 0x5c, 0x46,
0xd0, 0xca, 0xe4, 0xfe, 0xb8, 0xa2, 0x8c, 0x96 },
{ 0x00, 0x1d, 0x27, 0x3a, 0x4e, 0x53, 0x69, 0x74,
0x81, 0x9c, 0xa6, 0xbb, 0xcf, 0xd2, 0xe8, 0xf5 },
{ 0x00, 0xb0, 0x60, 0xd0, 0xc0, 0x70, 0xa0, 0x10,
0x80, 0x30, 0xe0, 0x50, 0x40, 0xf0, 0x20, 0x90 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x1b, 0x36, 0x2d, 0x6c, 0x77, 0x5a, 0x41,
0xd8, 0xc3, 0xee, 0xf5, 0xb4, 0xaf, 0x82, 0x99 },
{ 0x00, 0x1d, 0x27, 0x3a, 0x53, 0x4e, 0x74, 0x69,
0xa6, 0xbb, 0x81, 0x9c, 0xf5, 0xe8, 0xd2, 0xcf },
{ 0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40,
0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x1c, 0x38, 0x24, 0x70, 0x6c, 0x48, 0x54,
0xe0, 0xfc, 0xd8, 0xc4, 0x90, 0x8c, 0xa8, 0xb4 },
{ 0x00, 0x1d, 0x27, 0x3a, 0x53, 0x4e, 0x74, 0x69,
0xa6, 0xbb, 0x81, 0x9c, 0xf5, 0xe8, 0xd2, 0xcf },
{ 0x00, 0xd0, 0xa0, 0x70, 0x40, 0x90, 0xe0, 0x30,
0x80, 0x50, 0x20, 0xf0, 0xc0, 0x10, 0x60, 0xb0 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x1d, 0x3a, 0x27, 0x74, 0x69, 0x4e, 0x53,
0xe8, 0xf5, 0xd2, 0xcf, 0x9c, 0x81, 0xa6, 0xbb },
{ 0x00, 0x1d, 0x27, 0x3a, 0x53, 0x4e, 0x74, 0x69,
0xbb, 0xa6, 0x9c, 0x81, 0xe8, 0xf5, 0xcf, 0xd2 },
{ 0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0,
0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x1e, 0x3c, 0x22, 0x78, 0x66, 0x44, 0x5a,
0xf0, 0xee, 0xcc, 0xd2, 0x88, 0x96, 0xb4, 0xaa },
{ 0x00, 0x1d, 0x27, 0x3a, 0x53, 0x4e, 0x74, 0x69,
0xbb, 0xa6, 0x9c, 0x81, 0xe8, 0xf5, 0xcf, 0xd2 },
{ 0x00, 0xf0, 0xe0, 0x10, 0xc0, 0x30, 0x20, 0xd0,
0x80, 0x70, 0x60, 0x90, 0x40, 0xb0, 0xa0, 0x50 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x1f, 0x3e, 0x21, 0x7c, 0x63, 0x42, 0x5d,
0xf8, 0xe7, 0xc6, 0xd9, 0x84, 0x9b, 0xba, 0xa5 },
{ 0x00, 0x3a, 0x74, 0x4e, 0xe8, 0xd2, 0x9c, 0xa6,
0xcd, 0xf7, 0xb9, 0x83, 0x25, 0x1f, 0x51, 0x6b },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0,
0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0 },
{ 0x00, 0x3a, 0x74, 0x4e, 0xe8, 0xd2, 0x9c, 0xa6,
0xcd, 0xf7, 0xb9, 0x83, 0x25, 0x1f, 0x51, 0x6b },
{ 0x00, 0x10, 0x20, 0x30, 0x40, 0x50, 0x60, 0x70,
0x80, 0x90, 0xa0, 0xb0, 0xc0, 0xd0, 0xe0, 0xf0 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x21, 0x42, 0x63, 0x84, 0xa5, 0xc6, 0xe7,
0x08, 0x29, 0x4a, 0x6b, 0x8c, 0xad, 0xce, 0xef },
{ 0x00, 0x3a, 0x74, 0x4e, 0xe8, 0xd2, 0x9c, 0xa6,
0xd0, 0xea, 0xa4, 0x9e, 0x38, 0x02, 0x4c, 0x76 },
{ 0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0,
0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x22, 0x44, 0x66, 0x88, 0xaa, 0xcc, 0xee,
0x10, 0x32, 0x54, 0x76, 0x98, 0xba, 0xdc, 0xfe },
{ 0x00, 0x3a, 0x74, 0x4e, 0xe8, 0xd2, 0x9c, 0xa6,
0xd0, 0xea, 0xa4, 0x9e, 0x38, 0x02, 0x4c, 0x76 },
{ 0x00, 0x30, 0x60, 0x50, 0xc0, 0xf0, 0xa0, 0x90,
0x80, 0xb0, 0xe0, 0xd0, 0x40, 0x70, 0x20, 0x10 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x23, 0x46, 0x65, 0x8c, 0xaf, 0xca, 0xe9,
0x18, 0x3b, 0x5e, 0x7d, 0x94, 0xb7, 0xd2, 0xf1 },
{ 0x00, 0x3a, 0x74, 0x4e, 0xf5, 0xcf, 0x81, 0xbb,
0xf7, 0xcd, 0x83, 0xb9, 0x02, 0x38, 0x76, 0x4c },
{ 0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0,
0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x24, 0x48, 0x6c, 0x90, 0xb4, 0xd8, 0xfc,
0x20, 0x04, 0x68, 0x4c, 0xb0, 0x94, 0xf8, 0xdc },
{ 0x00, 0x3a, 0x74, 0x4e, 0xf5, 0xcf, 0x81, 0xbb,
0xf7, 0xcd, 0x83, 0xb9, 0x02, 0x38, 0x76, 0x4c },
{ 0x00, 0x50, 0xa0, 0xf0, 0x40, 0x10, 0xe0, 0xb0,
0x80, 0xd0, 0x20, 0x70, 0xc0, 0x90, 0x60, 0x30 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x25, 0x4a, 0x6f, 0x94, 0xb1, 0xde, 0xfb,
0x28, 0x0d, 0x62, 0x47, 0xbc, 0x99, 0xf6, 0xd3 },
{ 0x00, 0x3a, 0x74, 0x4e, 0xf5, 0xcf, 0x81, 0xbb,
0xea, 0xd0, 0x9e, 0xa4, 0x1f, 0x25, 0x6b, 0x51 },
{ 0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20,
0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x26, 0x4c, 0x6a, 0x98, 0xbe, 0xd4, 0xf2,
0x30, 0x16, 0x7c, 0x5a, 0xa8, 0x8e, 0xe4, 0xc2 },
{ 0x00, 0x3a, 0x74, 0x4e, 0xf5, 0xcf, 0x81, 0xbb,
0xea, 0xd0, 0x9e, 0xa4, 0x1f, 0x25, 0x6b, 0x51 },
{ 0x00, 0x70, 0xe0, 0x90, 0xc0, 0xb0, 0x20, 0x50,
0x80, 0xf0, 0x60, 0x10, 0x40, 0x30, 0xa0, 0xd0 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x27, 0x4e, 0x69, 0x9c, 0xbb, 0xd2, 0xf5,
0x38, 0x1f, 0x76, 0x51, 0xa4, 0x83, 0xea, 0xcd },
{ 0x00, 0x3a, 0x69, 0x53, 0xd2, 0xe8, 0xbb, 0x81,
0xb9, 0x83, 0xd0, 0xea, 0x6b, 0x51, 0x02, 0x38 },
{ 0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80,
0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x28, 0x50, 0x78, 0xa0, 0x88, 0xf0, 0xd8,
0x40, 0x68, 0x10, 0x38, 0xe0, 0xc8, 0xb0, 0x98 },
{ 0x00, 0x3a, 0x69, 0x53, 0xd2, 0xe8, 0xbb, 0x81,
0xb9, 0x83, 0xd0, 0xea, 0x6b, 0x51, 0x02, 0x38 },
{ 0x00, 0x90, 0x20, 0xb0, 0x40, 0xd0, 0x60, 0xf0,
0x80, 0x10, 0xa0, 0x30, 0xc0, 0x50, 0xe0, 0x70 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x29, 0x52, 0x7b, 0xa4, 0x8d, 0xf6, 0xdf,
0x48, 0x61, 0x1a, 0x33, 0xec, 0xc5, 0xbe, 0x97 },
{ 0x00, 0x3a, 0x69, 0x53, 0xd2, 0xe8, 0xbb, 0x81,
0xa4, 0x9e, 0xcd, 0xf7, 0x76, 0x4c, 0x1f, 0x25 },
{ 0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60,
0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x2a, 0x54, 0x7e, 0xa8, 0x82, 0xfc, 0xd6,
0x50, 0x7a, 0x04, 0x2e, 0xf8, 0xd2, 0xac, 0x86 },
{ 0x00, 0x3a, 0x69, 0x53, 0xd2, 0xe8, 0xbb, 0x81,
0xa4, 0x9e, 0xcd, 0xf7, 0x76, 0x4c, 0x1f, 0x25 },
{ 0x00, 0xb0, 0x60, 0xd0, 0xc0, 0x70, 0xa0, 0x10,
0x80, 0x30, 0xe0, 0x50, 0x40, 0xf0, 0x20, 0x90 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x2b, 0x56, 0x7d, 0xac, 0x87, 0xfa, 0xd1,
0x58, 0x73, 0x0e, 0x25, 0xf4, 0xdf, 0xa2, 0x89 },
{ 0x00, 0x3a, 0x69, 0x53, 0xcf, 0xf5, 0xa6, 0x9c,
0x83, 0xb9, 0xea, 0xd0, 0x4c, 0x76, 0x25, 0x1f },
{ 0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40,
0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x2c, 0x58, 0x74, 0xb0, 0x9c, 0xe8, 0xc4,
0x60, 0x4c, 0x38, 0x14, 0xd0, 0xfc, 0x88, 0xa4 },
{ 0x00, 0x3a, 0x69, 0x53, 0xcf, 0xf5, 0xa6, 0x9c,
0x83, 0xb9, 0xea, 0xd0, 0x4c, 0x76, 0x25, 0x1f },
{ 0x00, 0xd0, 0xa0, 0x70, 0x40, 0x90, 0xe0, 0x30,
0x80, 0x50, 0x20, 0xf0, 0xc0, 0x10, 0x60, 0xb0 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x2d, 0x5a, 0x77, 0xb4, 0x99, 0xee, 0xc3,
0x68, 0x45, 0x32, 0x1f, 0xdc, 0xf1, 0x86, 0xab },
{ 0x00, 0x3a, 0x69, 0x53, 0xcf, 0xf5, 0xa6, 0x9c,
0x9e, 0xa4, 0xf7, 0xcd, 0x51, 0x6b, 0x38, 0x02 },
{ 0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0,
0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x2e, 0x5c, 0x72, 0xb8, 0x96, 0xe4, 0xca,
0x70, 0x5e, 0x2c, 0x02, 0xc8, 0xe6, 0x94, 0xba },
{ 0x00, 0x3a, 0x69, 0x53, 0xcf, 0xf5, 0xa6, 0x9c,
0x9e, 0xa4, 0xf7, 0xcd, 0x51, 0x6b, 0x38, 0x02 },
{ 0x00, 0xf0, 0xe0, 0x10, 0xc0, 0x30, 0x20, 0xd0,
0x80, 0x70, 0x60, 0x90, 0x40, 0xb0, 0xa0, 0x50 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x2f, 0x5e, 0x71, 0xbc, 0x93, 0xe2, 0xcd,
0x78, 0x57, 0x26, 0x09, 0xc4, 0xeb, 0x9a, 0xb5 },
{ 0x00, 0x27, 0x4e, 0x69, 0x9c, 0xbb, 0xd2, 0xf5,
0x25, 0x02, 0x6b, 0x4c, 0xb9, 0x9e, 0xf7, 0xd0 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x30, 0x60, 0x50, 0xc0, 0xf0, 0xa0, 0x90,
0x80, 0xb0, 0xe0, 0xd0, 0x40, 0x70, 0x20, 0x10 },
{ 0x00, 0x27, 0x4e, 0x69, 0x9c, 0xbb, 0xd2, 0xf5,
0x25, 0x02, 0x6b, 0x4c, 0xb9, 0x9e, 0xf7, 0xd0 },
{ 0x00, 0x10, 0x20, 0x30, 0x40, 0x50, 0x60, 0x70,
0x80, 0x90, 0xa0, 0xb0, 0xc0, 0xd0, 0xe0, 0xf0 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x31, 0x62, 0x53, 0xc4, 0xf5, 0xa6, 0x97,
0x88, 0xb9, 0xea, 0xdb, 0x4c, 0x7d, 0x2e, 0x1f },
{ 0x00, 0x27, 0x4e, 0x69, 0x9c, 0xbb, 0xd2, 0xf5,
0x38, 0x1f, 0x76, 0x51, 0xa4, 0x83, 0xea, 0xcd },
{ 0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0,
0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x32, 0x64, 0x56, 0xc8, 0xfa, 0xac, 0x9e,
0x90, 0xa2, 0xf4, 0xc6, 0x58, 0x6a, 0x3c, 0x0e },
{ 0x00, 0x27, 0x4e, 0x69, 0x9c, 0xbb, 0xd2, 0xf5,
0x38, 0x1f, 0x76, 0x51, 0xa4, 0x83, 0xea, 0xcd },
{ 0x00, 0x30, 0x60, 0x50, 0xc0, 0xf0, 0xa0, 0x90,
0x80, 0xb0, 0xe0, 0xd0, 0x40, 0x70, 0x20, 0x10 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x33, 0x66, 0x55, 0xcc, 0xff, 0xaa, 0x99,
0x98, 0xab, 0xfe, 0xcd, 0x54, 0x67, 0x32, 0x01 },
{ 0x00, 0x27, 0x4e, 0x69, 0x81, 0xa6, 0xcf, 0xe8,
0x1f, 0x38, 0x51, 0x76, 0x9e, 0xb9, 0xd0, 0xf7 },
{ 0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0,
0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x34, 0x68, 0x5c, 0xd0, 0xe4, 0xb8, 0x8c,
0xa0, 0x94, 0xc8, 0xfc, 0x70, 0x44, 0x18, 0x2c },
{ 0x00, 0x27, 0x4e, 0x69, 0x81, 0xa6, 0xcf, 0xe8,
0x1f, 0x38, 0x51, 0x76, 0x9e, 0xb9, 0xd0, 0xf7 },
{ 0x00, 0x50, 0xa0, 0xf0, 0x40, 0x10, 0xe0, 0xb0,
0x80, 0xd0, 0x20, 0x70, 0xc0, 0x90, 0x60, 0x30 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x35, 0x6a, 0x5f, 0xd4, 0xe1, 0xbe, 0x8b,
0xa8, 0x9d, 0xc2, 0xf7, 0x7c, 0x49, 0x16, 0x23 },
{ 0x00, 0x27, 0x4e, 0x69, 0x81, 0xa6, 0xcf, 0xe8,
0x02, 0x25, 0x4c, 0x6b, 0x83, 0xa4, 0xcd, 0xea },
{ 0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20,
0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x36, 0x6c, 0x5a, 0xd8, 0xee, 0xb4, 0x82,
0xb0, 0x86, 0xdc, 0xea, 0x68, 0x5e, 0x04, 0x32 },
{ 0x00, 0x27, 0x4e, 0x69, 0x81, 0xa6, 0xcf, 0xe8,
0x02, 0x25, 0x4c, 0x6b, 0x83, 0xa4, 0xcd, 0xea },
{ 0x00, 0x70, 0xe0, 0x90, 0xc0, 0xb0, 0x20, 0x50,
0x80, 0xf0, 0x60, 0x10, 0x40, 0x30, 0xa0, 0xd0 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x37, 0x6e, 0x59, 0xdc, 0xeb, 0xb2, 0x85,
0xb8, 0x8f, 0xd6, 0xe1, 0x64, 0x53, 0x0a, 0x3d },
{ 0x00, 0x27, 0x53, 0x74, 0xa6, 0x81, 0xf5, 0xd2,
0x51, 0x76, 0x02, 0x25, 0xf7, 0xd0, 0xa4, 0x83 },
{ 0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80,
0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x38, 0x70, 0x48, 0xe0, 0xd8, 0x90, 0xa8,
0xc0, 0xf8, 0xb0, 0x88, 0x20, 0x18, 0x50, 0x68 },
{ 0x00, 0x27, 0x53, 0x74, 0xa6, 0x81, 0xf5, 0xd2,
0x51, 0x76, 0x02, 0x25, 0xf7, 0xd0, 0xa4, 0x83 },
{ 0x00, 0x90, 0x20, 0xb0, 0x40, 0xd0, 0x60, 0xf0,
0x80, 0x10, 0xa0, 0x30, 0xc0, 0x50, 0xe0, 0x70 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x39, 0x72, 0x4b, 0xe4, 0xdd, 0x96, 0xaf,
0xc8, 0xf1, 0xba, 0x83, 0x2c, 0x15, 0x5e, 0x67 },
{ 0x00, 0x27, 0x53, 0x74, 0xa6, 0x81, 0xf5, 0xd2,
0x4c, 0x6b, 0x1f, 0x38, 0xea, 0xcd, 0xb9, 0x9e },
{ 0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60,
0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x3a, 0x74, 0x4e, 0xe8, 0xd2, 0x9c, 0xa6,
0xd0, 0xea, 0xa4, 0x9e, 0x38, 0x02, 0x4c, 0x76 },
{ 0x00, 0x27, 0x53, 0x74, 0xa6, 0x81, 0xf5, 0xd2,
0x4c, 0x6b, 0x1f, 0x38, 0xea, 0xcd, 0xb9, 0x9e },
{ 0x00, 0xb0, 0x60, 0xd0, 0xc0, 0x70, 0xa0, 0x10,
0x80, 0x30, 0xe0, 0x50, 0x40, 0xf0, 0x20, 0x90 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x3b, 0x76, 0x4d, 0xec, 0xd7, 0x9a, 0xa1,
0xd8, 0xe3, 0xae, 0x95, 0x34, 0x0f, 0x42, 0x79 },
{ 0x00, 0x27, 0x53, 0x74, 0xbb, 0x9c, 0xe8, 0xcf,
0x6b, 0x4c, 0x38, 0x1f, 0xd0, 0xf7, 0x83, 0xa4 },
{ 0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40,
0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x3c, 0x78, 0x44, 0xf0, 0xcc, 0x88, 0xb4,
0xe0, 0xdc, 0x98, 0xa4, 0x10, 0x2c, 0x68, 0x54 },
{ 0x00, 0x27, 0x53, 0x74, 0xbb, 0x9c, 0xe8, 0xcf,
0x6b, 0x4c, 0x38, 0x1f, 0xd0, 0xf7, 0x83, 0xa4 },
{ 0x00, 0xd0, 0xa0, 0x70, 0x40, 0x90, 0xe0, 0x30,
0x80, 0x50, 0x20, 0xf0, 0xc0, 0x10, 0x60, 0xb0 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x3d, 0x7a, 0x47, 0xf4, 0xc9, 0x8e, 0xb3,
0xe8, 0xd5, 0x92, 0xaf, 0x1c, 0x21, 0x66, 0x5b },
{ 0x00, 0x27, 0x53, 0x74, 0xbb, 0x9c, 0xe8, 0xcf,
0x76, 0x51, 0x25, 0x02, 0xcd, 0xea, 0x9e, 0xb9 },
{ 0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0,
0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x3e, 0x7c, 0x42, 0xf8, 0xc6, 0x84, 0xba,
0xf0, 0xce, 0x8c, 0xb2, 0x08, 0x36, 0x74, 0x4a },
{ 0x00, 0x27, 0x53, 0x74, 0xbb, 0x9c, 0xe8, 0xcf,
0x76, 0x51, 0x25, 0x02, 0xcd, 0xea, 0x9e, 0xb9 },
{ 0x00, 0xf0, 0xe0, 0x10, 0xc0, 0x30, 0x20, 0xd0,
0x80, 0x70, 0x60, 0x90, 0x40, 0xb0, 0xa0, 0x50 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d, 0x1d },
{ 0x00, 0x3f, 0x7e, 0x41, 0xfc, 0xc3, 0x82, 0xbd,
0xf8, 0xc7, 0x86, 0xb9, 0x04, 0x3b, 0x7a, 0x45 },
{ 0x00, 0x74, 0xe8, 0x9c, 0xcd, 0xb9, 0x25, 0x51,
0x87, 0xf3, 0x6f, 0x1b, 0x4a, 0x3e, 0xa2, 0xd6 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0,
0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0 },
{ 0x00, 0x74, 0xe8, 0x9c, 0xcd, 0xb9, 0x25, 0x51,
0x87, 0xf3, 0x6f, 0x1b, 0x4a, 0x3e, 0xa2, 0xd6 },
{ 0x00, 0x10, 0x20, 0x30, 0x40, 0x50, 0x60, 0x70,
0x80, 0x90, 0xa0, 0xb0, 0xc0, 0xd0, 0xe0, 0xf0 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x41, 0x82, 0xc3, 0x04, 0x45, 0x86, 0xc7,
0x08, 0x49, 0x8a, 0xcb, 0x0c, 0x4d, 0x8e, 0xcf },
{ 0x00, 0x74, 0xe8, 0x9c, 0xcd, 0xb9, 0x25, 0x51,
0x9a, 0xee, 0x72, 0x06, 0x57, 0x23, 0xbf, 0xcb },
{ 0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0,
0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x42, 0x84, 0xc6, 0x08, 0x4a, 0x8c, 0xce,
0x10, 0x52, 0x94, 0xd6, 0x18, 0x5a, 0x9c, 0xde },
{ 0x00, 0x74, 0xe8, 0x9c, 0xcd, 0xb9, 0x25, 0x51,
0x9a, 0xee, 0x72, 0x06, 0x57, 0x23, 0xbf, 0xcb },
{ 0x00, 0x30, 0x60, 0x50, 0xc0, 0xf0, 0xa0, 0x90,
0x80, 0xb0, 0xe0, 0xd0, 0x40, 0x70, 0x20, 0x10 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x43, 0x86, 0xc5, 0x0c, 0x4f, 0x8a, 0xc9,
0x18, 0x5b, 0x9e, 0xdd, 0x14, 0x57, 0x92, 0xd1 },
{ 0x00, 0x74, 0xe8, 0x9c, 0xd0, 0xa4, 0x38, 0x4c,
0xbd, 0xc9, 0x55, 0x21, 0x6d, 0x19, 0x85, 0xf1 },
{ 0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0,
0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x44, 0x88, 0xcc, 0x10, 0x54, 0x98, 0xdc,
0x20, 0x64, 0xa8, 0xec, 0x30, 0x74, 0xb8, 0xfc },
{ 0x00, 0x74, 0xe8, 0x9c, 0xd0, 0xa4, 0x38, 0x4c,
0xbd, 0xc9, 0x55, 0x21, 0x6d, 0x19, 0x85, 0xf1 },
{ 0x00, 0x50, 0xa0, 0xf0, 0x40, 0x10, 0xe0, 0xb0,
0x80, 0xd0, 0x20, 0x70, 0xc0, 0x90, 0x60, 0x30 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x45, 0x8a, 0xcf, 0x14, 0x51, 0x9e, 0xdb,
0x28, 0x6d, 0xa2, 0xe7, 0x3c, 0x79, 0xb6, 0xf3 },
{ 0x00, 0x74, 0xe8, 0x9c, 0xd0, 0xa4, 0x38, 0x4c,
0xa0, 0xd4, 0x48, 0x3c, 0x70, 0x04, 0x98, 0xec },
{ 0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20,
0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x46, 0x8c, 0xca, 0x18, 0x5e, 0x94, 0xd2,
0x30, 0x76, 0xbc, 0xfa, 0x28, 0x6e, 0xa4, 0xe2 },
{ 0x00, 0x74, 0xe8, 0x9c, 0xd0, 0xa4, 0x38, 0x4c,
0xa0, 0xd4, 0x48, 0x3c, 0x70, 0x04, 0x98, 0xec },
{ 0x00, 0x70, 0xe0, 0x90, 0xc0, 0xb0, 0x20, 0x50,
0x80, 0xf0, 0x60, 0x10, 0x40, 0x30, 0xa0, 0xd0 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x47, 0x8e, 0xc9, 0x1c, 0x5b, 0x92, 0xd5,
0x38, 0x7f, 0xb6, 0xf1, 0x24, 0x63, 0xaa, 0xed },
{ 0x00, 0x74, 0xf5, 0x81, 0xf7, 0x83, 0x02, 0x76,
0xf3, 0x87, 0x06, 0x72, 0x04, 0x70, 0xf1, 0x85 },
{ 0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80,
0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x48, 0x90, 0xd8, 0x20, 0x68, 0xb0, 0xf8,
0x40, 0x08, 0xd0, 0x98, 0x60, 0x28, 0xf0, 0xb8 },
{ 0x00, 0x74, 0xf5, 0x81, 0xf7, 0x83, 0x02, 0x76,
0xf3, 0x87, 0x06, 0x72, 0x04, 0x70, 0xf1, 0x85 },
{ 0x00, 0x90, 0x20, 0xb0, 0x40, 0xd0, 0x60, 0xf0,
0x80, 0x10, 0xa0, 0x30, 0xc0, 0x50, 0xe0, 0x70 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x49, 0x92, 0xdb, 0x24, 0x6d, 0xb6, 0xff,
0x48, 0x01, 0xda, 0x93, 0x6c, 0x25, 0xfe, 0xb7 },
{ 0x00, 0x74, 0xf5, 0x81, 0xf7, 0x83, 0x02, 0x76,
0xee, 0x9a, 0x1b, 0x6f, 0x19, 0x6d, 0xec, 0x98 },
{ 0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60,
0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x4a, 0x94, 0xde, 0x28, 0x62, 0xbc, 0xf6,
0x50, 0x1a, 0xc4, 0x8e, 0x78, 0x32, 0xec, 0xa6 },
{ 0x00, 0x74, 0xf5, 0x81, 0xf7, 0x83, 0x02, 0x76,
0xee, 0x9a, 0x1b, 0x6f, 0x19, 0x6d, 0xec, 0x98 },
{ 0x00, 0xb0, 0x60, 0xd0, 0xc0, 0x70, 0xa0, 0x10,
0x80, 0x30, 0xe0, 0x50, 0x40, 0xf0, 0x20, 0x90 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x4b, 0x96, 0xdd, 0x2c, 0x67, 0xba, 0xf1,
0x58, 0x13, 0xce, 0x85, 0x74, 0x3f, 0xe2, 0xa9 },
{ 0x00, 0x74, 0xf5, 0x81, 0xea, 0x9e, 0x1f, 0x6b,
0xc9, 0xbd, 0x3c, 0x48, 0x23, 0x57, 0xd6, 0xa2 },
{ 0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40,
0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x4c, 0x98, 0xd4, 0x30, 0x7c, 0xa8, 0xe4,
0x60, 0x2c, 0xf8, 0xb4, 0x50, 0x1c, 0xc8, 0x84 },
{ 0x00, 0x74, 0xf5, 0x81, 0xea, 0x9e, 0x1f, 0x6b,
0xc9, 0xbd, 0x3c, 0x48, 0x23, 0x57, 0xd6, 0xa2 },
{ 0x00, 0xd0, 0xa0, 0x70, 0x40, 0x90, 0xe0, 0x30,
0x80, 0x50, 0x20, 0xf0, 0xc0, 0x10, 0x60, 0xb0 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x4d, 0x9a, 0xd7, 0x34, 0x79, 0xae, 0xe3,
0x68, 0x25, 0xf2, 0xbf, 0x5c, 0x11, 0xc6, 0x8b },
{ 0x00, 0x74, 0xf5, 0x81, 0xea, 0x9e, 0x1f, 0x6b,
0xd4, 0xa0, 0x21, 0x55, 0x3e, 0x4a, 0xcb, 0xbf },
{ 0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0,
0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x4e, 0x9c, 0xd2, 0x38, 0x76, 0xa4, 0xea,
0x70, 0x3e, 0xec, 0xa2, 0x48, 0x06, 0xd4, 0x9a },
{ 0x00, 0x74, 0xf5, 0x81, 0xea, 0x9e, 0x1f, 0x6b,
0xd4, 0xa0, 0x21, 0x55, 0x3e, 0x4a, 0xcb, 0xbf },
{ 0x00, 0xf0, 0xe0, 0x10, 0xc0, 0x30, 0x20, 0xd0,
0x80, 0x70, 0x60, 0x90, 0x40, 0xb0, 0xa0, 0x50 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x4f, 0x9e, 0xd1, 0x3c, 0x73, 0xa2, 0xed,
0x78, 0x37, 0xe6, 0xa9, 0x44, 0x0b, 0xda, 0x95 },
{ 0x00, 0x69, 0xd2, 0xbb, 0xb9, 0xd0, 0x6b, 0x02,
0x6f, 0x06, 0xbd, 0xd4, 0xd6, 0xbf, 0x04, 0x6d },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x50, 0xa0, 0xf0, 0x40, 0x10, 0xe0, 0xb0,
0x80, 0xd0, 0x20, 0x70, 0xc0, 0x90, 0x60, 0x30 },
{ 0x00, 0x69, 0xd2, 0xbb, 0xb9, 0xd0, 0x6b, 0x02,
0x6f, 0x06, 0xbd, 0xd4, 0xd6, 0xbf, 0x04, 0x6d },
{ 0x00, 0x10, 0x20, 0x30, 0x40, 0x50, 0x60, 0x70,
0x80, 0x90, 0xa0, 0xb0, 0xc0, 0xd0, 0xe0, 0xf0 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x51, 0xa2, 0xf3, 0x44, 0x15, 0xe6, 0xb7,
0x88, 0xd9, 0x2a, 0x7b, 0xcc, 0x9d, 0x6e, 0x3f },
{ 0x00, 0x69, 0xd2, 0xbb, 0xb9, 0xd0, 0x6b, 0x02,
0x72, 0x1b, 0xa0, 0xc9, 0xcb, 0xa2, 0x19, 0x70 },
{ 0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0,
0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x52, 0xa4, 0xf6, 0x48, 0x1a, 0xec, 0xbe,
0x90, 0xc2, 0x34, 0x66, 0xd8, 0x8a, 0x7c, 0x2e },
{ 0x00, 0x69, 0xd2, 0xbb, 0xb9, 0xd0, 0x6b, 0x02,
0x72, 0x1b, 0xa0, 0xc9, 0xcb, 0xa2, 0x19, 0x70 },
{ 0x00, 0x30, 0x60, 0x50, 0xc0, 0xf0, 0xa0, 0x90,
0x80, 0xb0, 0xe0, 0xd0, 0x40, 0x70, 0x20, 0x10 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x53, 0xa6, 0xf5, 0x4c, 0x1f, 0xea, 0xb9,
0x98, 0xcb, 0x3e, 0x6d, 0xd4, 0x87, 0x72, 0x21 },
{ 0x00, 0x69, 0xd2, 0xbb, 0xa4, 0xcd, 0x76, 0x1f,
0x55, 0x3c, 0x87, 0xee, 0xf1, 0x98, 0x23, 0x4a },
{ 0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0,
0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x54, 0xa8, 0xfc, 0x50, 0x04, 0xf8, 0xac,
0xa0, 0xf4, 0x08, 0x5c, 0xf0, 0xa4, 0x58, 0x0c },
{ 0x00, 0x69, 0xd2, 0xbb, 0xa4, 0xcd, 0x76, 0x1f,
0x55, 0x3c, 0x87, 0xee, 0xf1, 0x98, 0x23, 0x4a },
{ 0x00, 0x50, 0xa0, 0xf0, 0x40, 0x10, 0xe0, 0xb0,
0x80, 0xd0, 0x20, 0x70, 0xc0, 0x90, 0x60, 0x30 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x55, 0xaa, 0xff, 0x54, 0x01, 0xfe, 0xab,
0xa8, 0xfd, 0x02, 0x57, 0xfc, 0xa9, 0x56, 0x03 },
{ 0x00, 0x69, 0xd2, 0xbb, 0xa4, 0xcd, 0x76, 0x1f,
0x48, 0x21, 0x9a, 0xf3, 0xec, 0x85, 0x3e, 0x57 },
{ 0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20,
0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x56, 0xac, 0xfa, 0x58, 0x0e, 0xf4, 0xa2,
0xb0, 0xe6, 0x1c, 0x4a, 0xe8, 0xbe, 0x44, 0x12 },
{ 0x00, 0x69, 0xd2, 0xbb, 0xa4, 0xcd, 0x76, 0x1f,
0x48, 0x21, 0x9a, 0xf3, 0xec, 0x85, 0x3e, 0x57 },
{ 0x00, 0x70, 0xe0, 0x90, 0xc0, 0xb0, 0x20, 0x50,
0x80, 0xf0, 0x60, 0x10, 0x40, 0x30, 0xa0, 0xd0 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x57, 0xae, 0xf9, 0x5c, 0x0b, 0xf2, 0xa5,
0xb8, 0xef, 0x16, 0x41, 0xe4, 0xb3, 0x4a, 0x1d },
{ 0x00, 0x69, 0xcf, 0xa6, 0x83, 0xea, 0x4c, 0x25,
0x1b, 0x72, 0xd4, 0xbd, 0x98, 0xf1, 0x57, 0x3e },
{ 0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80,
0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x58, 0xb0, 0xe8, 0x60, 0x38, 0xd0, 0x88,
0xc0, 0x98, 0x70, 0x28, 0xa0, 0xf8, 0x10, 0x48 },
{ 0x00, 0x69, 0xcf, 0xa6, 0x83, 0xea, 0x4c, 0x25,
0x1b, 0x72, 0xd4, 0xbd, 0x98, 0xf1, 0x57, 0x3e },
{ 0x00, 0x90, 0x20, 0xb0, 0x40, 0xd0, 0x60, 0xf0,
0x80, 0x10, 0xa0, 0x30, 0xc0, 0x50, 0xe0, 0x70 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x59, 0xb2, 0xeb, 0x64, 0x3d, 0xd6, 0x8f,
0xc8, 0x91, 0x7a, 0x23, 0xac, 0xf5, 0x1e, 0x47 },
{ 0x00, 0x69, 0xcf, 0xa6, 0x83, 0xea, 0x4c, 0x25,
0x06, 0x6f, 0xc9, 0xa0, 0x85, 0xec, 0x4a, 0x23 },
{ 0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60,
0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x5a, 0xb4, 0xee, 0x68, 0x32, 0xdc, 0x86,
0xd0, 0x8a, 0x64, 0x3e, 0xb8, 0xe2, 0x0c, 0x56 },
{ 0x00, 0x69, 0xcf, 0xa6, 0x83, 0xea, 0x4c, 0x25,
0x06, 0x6f, 0xc9, 0xa0, 0x85, 0xec, 0x4a, 0x23 },
{ 0x00, 0xb0, 0x60, 0xd0, 0xc0, 0x70, 0xa0, 0x10,
0x80, 0x30, 0xe0, 0x50, 0x40, 0xf0, 0x20, 0x90 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x5b, 0xb6, 0xed, 0x6c, 0x37, 0xda, 0x81,
0xd8, 0x83, 0x6e, 0x35, 0xb4, 0xef, 0x02, 0x59 },
{ 0x00, 0x69, 0xcf, 0xa6, 0x9e, 0xf7, 0x51, 0x38,
0x21, 0x48, 0xee, 0x87, 0xbf, 0xd6, 0x70, 0x19 },
{ 0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40,
0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x5c, 0xb8, 0xe4, 0x70, 0x2c, 0xc8, 0x94,
0xe0, 0xbc, 0x58, 0x04, 0x90, 0xcc, 0x28, 0x74 },
{ 0x00, 0x69, 0xcf, 0xa6, 0x9e, 0xf7, 0x51, 0x38,
0x21, 0x48, 0xee, 0x87, 0xbf, 0xd6, 0x70, 0x19 },
{ 0x00, 0xd0, 0xa0, 0x70, 0x40, 0x90, 0xe0, 0x30,
0x80, 0x50, 0x20, 0xf0, 0xc0, 0x10, 0x60, 0xb0 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x5d, 0xba, 0xe7, 0x74, 0x29, 0xce, 0x93,
0xe8, 0xb5, 0x52, 0x0f, 0x9c, 0xc1, 0x26, 0x7b },
{ 0x00, 0x69, 0xcf, 0xa6, 0x9e, 0xf7, 0x51, 0x38,
0x3c, 0x55, 0xf3, 0x9a, 0xa2, 0xcb, 0x6d, 0x04 },
{ 0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0,
0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x5e, 0xbc, 0xe2, 0x78, 0x26, 0xc4, 0x9a,
0xf0, 0xae, 0x4c, 0x12, 0x88, 0xd6, 0x34, 0x6a },
{ 0x00, 0x69, 0xcf, 0xa6, 0x9e, 0xf7, 0x51, 0x38,
0x3c, 0x55, 0xf3, 0x9a, 0xa2, 0xcb, 0x6d, 0x04 },
{ 0x00, 0xf0, 0xe0, 0x10, 0xc0, 0x30, 0x20, 0xd0,
0x80, 0x70, 0x60, 0x90, 0x40, 0xb0, 0xa0, 0x50 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x3a, 0x3a, 0x3a, 0x3a, 0x27, 0x27, 0x27, 0x27 },
{ 0x00, 0x5f, 0xbe, 0xe1, 0x7c, 0x23, 0xc2, 0x9d,
0xf8, 0xa7, 0x46, 0x19, 0x84, 0xdb, 0x3a, 0x65 },
{ 0x00, 0x4e, 0x9c, 0xd2, 0x25, 0x6b, 0xb9, 0xf7,
0x4a, 0x04, 0xd6, 0x98, 0x6f, 0x21, 0xf3, 0xbd },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20,
0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20 },
{ 0x00, 0x4e, 0x9c, 0xd2, 0x25, 0x6b, 0xb9, 0xf7,
0x4a, 0x04, 0xd6, 0x98, 0x6f, 0x21, 0xf3, 0xbd },
{ 0x00, 0x10, 0x20, 0x30, 0x40, 0x50, 0x60, 0x70,
0x80, 0x90, 0xa0, 0xb0, 0xc0, 0xd0, 0xe0, 0xf0 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x61, 0xc2, 0xa3, 0x84, 0xe5, 0x46, 0x27,
0x08, 0x69, 0xca, 0xab, 0x8c, 0xed, 0x4e, 0x2f },
{ 0x00, 0x4e, 0x9c, 0xd2, 0x25, 0x6b, 0xb9, 0xf7,
0x57, 0x19, 0xcb, 0x85, 0x72, 0x3c, 0xee, 0xa0 },
{ 0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0,
0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x62, 0xc4, 0xa6, 0x88, 0xea, 0x4c, 0x2e,
0x10, 0x72, 0xd4, 0xb6, 0x98, 0xfa, 0x5c, 0x3e },
{ 0x00, 0x4e, 0x9c, 0xd2, 0x25, 0x6b, 0xb9, 0xf7,
0x57, 0x19, 0xcb, 0x85, 0x72, 0x3c, 0xee, 0xa0 },
{ 0x00, 0x30, 0x60, 0x50, 0xc0, 0xf0, 0xa0, 0x90,
0x80, 0xb0, 0xe0, 0xd0, 0x40, 0x70, 0x20, 0x10 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x63, 0xc6, 0xa5, 0x8c, 0xef, 0x4a, 0x29,
0x18, 0x7b, 0xde, 0xbd, 0x94, 0xf7, 0x52, 0x31 },
{ 0x00, 0x4e, 0x9c, 0xd2, 0x38, 0x76, 0xa4, 0xea,
0x70, 0x3e, 0xec, 0xa2, 0x48, 0x06, 0xd4, 0x9a },
{ 0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0,
0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x64, 0xc8, 0xac, 0x90, 0xf4, 0x58, 0x3c,
0x20, 0x44, 0xe8, 0x8c, 0xb0, 0xd4, 0x78, 0x1c },
{ 0x00, 0x4e, 0x9c, 0xd2, 0x38, 0x76, 0xa4, 0xea,
0x70, 0x3e, 0xec, 0xa2, 0x48, 0x06, 0xd4, 0x9a },
{ 0x00, 0x50, 0xa0, 0xf0, 0x40, 0x10, 0xe0, 0xb0,
0x80, 0xd0, 0x20, 0x70, 0xc0, 0x90, 0x60, 0x30 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x65, 0xca, 0xaf, 0x94, 0xf1, 0x5e, 0x3b,
0x28, 0x4d, 0xe2, 0x87, 0xbc, 0xd9, 0x76, 0x13 },
{ 0x00, 0x4e, 0x9c, 0xd2, 0x38, 0x76, 0xa4, 0xea,
0x6d, 0x23, 0xf1, 0xbf, 0x55, 0x1b, 0xc9, 0x87 },
{ 0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20,
0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x66, 0xcc, 0xaa, 0x98, 0xfe, 0x54, 0x32,
0x30, 0x56, 0xfc, 0x9a, 0xa8, 0xce, 0x64, 0x02 },
{ 0x00, 0x4e, 0x9c, 0xd2, 0x38, 0x76, 0xa4, 0xea,
0x6d, 0x23, 0xf1, 0xbf, 0x55, 0x1b, 0xc9, 0x87 },
{ 0x00, 0x70, 0xe0, 0x90, 0xc0, 0xb0, 0x20, 0x50,
0x80, 0xf0, 0x60, 0x10, 0x40, 0x30, 0xa0, 0xd0 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x67, 0xce, 0xa9, 0x9c, 0xfb, 0x52, 0x35,
0x38, 0x5f, 0xf6, 0x91, 0xa4, 0xc3, 0x6a, 0x0d },
{ 0x00, 0x4e, 0x81, 0xcf, 0x1f, 0x51, 0x9e, 0xd0,
0x3e, 0x70, 0xbf, 0xf1, 0x21, 0x6f, 0xa0, 0xee },
{ 0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80,
0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x68, 0xd0, 0xb8, 0xa0, 0xc8, 0x70, 0x18,
0x40, 0x28, 0x90, 0xf8, 0xe0, 0x88, 0x30, 0x58 },
{ 0x00, 0x4e, 0x81, 0xcf, 0x1f, 0x51, 0x9e, 0xd0,
0x3e, 0x70, 0xbf, 0xf1, 0x21, 0x6f, 0xa0, 0xee },
{ 0x00, 0x90, 0x20, 0xb0, 0x40, 0xd0, 0x60, 0xf0,
0x80, 0x10, 0xa0, 0x30, 0xc0, 0x50, 0xe0, 0x70 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x69, 0xd2, 0xbb, 0xa4, 0xcd, 0x76, 0x1f,
0x48, 0x21, 0x9a, 0xf3, 0xec, 0x85, 0x3e, 0x57 },
{ 0x00, 0x4e, 0x81, 0xcf, 0x1f, 0x51, 0x9e, 0xd0,
0x23, 0x6d, 0xa2, 0xec, 0x3c, 0x72, 0xbd, 0xf3 },
{ 0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60,
0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x6a, 0xd4, 0xbe, 0xa8, 0xc2, 0x7c, 0x16,
0x50, 0x3a, 0x84, 0xee, 0xf8, 0x92, 0x2c, 0x46 },
{ 0x00, 0x4e, 0x81, 0xcf, 0x1f, 0x51, 0x9e, 0xd0,
0x23, 0x6d, 0xa2, 0xec, 0x3c, 0x72, 0xbd, 0xf3 },
{ 0x00, 0xb0, 0x60, 0xd0, 0xc0, 0x70, 0xa0, 0x10,
0x80, 0x30, 0xe0, 0x50, 0x40, 0xf0, 0x20, 0x90 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x6b, 0xd6, 0xbd, 0xac, 0xc7, 0x7a, 0x11,
0x58, 0x33, 0x8e, 0xe5, 0xf4, 0x9f, 0x22, 0x49 },
{ 0x00, 0x4e, 0x81, 0xcf, 0x02, 0x4c, 0x83, 0xcd,
0x04, 0x4a, 0x85, 0xcb, 0x06, 0x48, 0x87, 0xc9 },
{ 0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40,
0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x6c, 0xd8, 0xb4, 0xb0, 0xdc, 0x68, 0x04,
0x60, 0x0c, 0xb8, 0xd4, 0xd0, 0xbc, 0x08, 0x64 },
{ 0x00, 0x4e, 0x81, 0xcf, 0x02, 0x4c, 0x83, 0xcd,
0x04, 0x4a, 0x85, 0xcb, 0x06, 0x48, 0x87, 0xc9 },
{ 0x00, 0xd0, 0xa0, 0x70, 0x40, 0x90, 0xe0, 0x30,
0x80, 0x50, 0x20, 0xf0, 0xc0, 0x10, 0x60, 0xb0 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x6d, 0xda, 0xb7, 0xb4, 0xd9, 0x6e, 0x03,
0x68, 0x05, 0xb2, 0xdf, 0xdc, 0xb1, 0x06, 0x6b },
{ 0x00, 0x4e, 0x81, 0xcf, 0x02, 0x4c, 0x83, 0xcd,
0x19, 0x57, 0x98, 0xd6, 0x1b, 0x55, 0x9a, 0xd4 },
{ 0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0,
0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x6e, 0xdc, 0xb2, 0xb8, 0xd6, 0x64, 0x0a,
0x70, 0x1e, 0xac, 0xc2, 0xc8, 0xa6, 0x14, 0x7a },
{ 0x00, 0x4e, 0x81, 0xcf, 0x02, 0x4c, 0x83, 0xcd,
0x19, 0x57, 0x98, 0xd6, 0x1b, 0x55, 0x9a, 0xd4 },
{ 0x00, 0xf0, 0xe0, 0x10, 0xc0, 0x30, 0x20, 0xd0,
0x80, 0x70, 0x60, 0x90, 0x40, 0xb0, 0xa0, 0x50 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x6f, 0xde, 0xb1, 0xbc, 0xd3, 0x62, 0x0d,
0x78, 0x17, 0xa6, 0xc9, 0xc4, 0xab, 0x1a, 0x75 },
{ 0x00, 0x53, 0xa6, 0xf5, 0x51, 0x02, 0xf7, 0xa4,
0xa2, 0xf1, 0x04, 0x57, 0xf3, 0xa0, 0x55, 0x06 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x70, 0xe0, 0x90, 0xc0, 0xb0, 0x20, 0x50,
0x80, 0xf0, 0x60, 0x10, 0x40, 0x30, 0xa0, 0xd0 },
{ 0x00, 0x53, 0xa6, 0xf5, 0x51, 0x02, 0xf7, 0xa4,
0xa2, 0xf1, 0x04, 0x57, 0xf3, 0xa0, 0x55, 0x06 },
{ 0x00, 0x10, 0x20, 0x30, 0x40, 0x50, 0x60, 0x70,
0x80, 0x90, 0xa0, 0xb0, 0xc0, 0xd0, 0xe0, 0xf0 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x71, 0xe2, 0x93, 0xc4, 0xb5, 0x26, 0x57,
0x88, 0xf9, 0x6a, 0x1b, 0x4c, 0x3d, 0xae, 0xdf },
{ 0x00, 0x53, 0xa6, 0xf5, 0x51, 0x02, 0xf7, 0xa4,
0xbf, 0xec, 0x19, 0x4a, 0xee, 0xbd, 0x48, 0x1b },
{ 0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0,
0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x72, 0xe4, 0x96, 0xc8, 0xba, 0x2c, 0x5e,
0x90, 0xe2, 0x74, 0x06, 0x58, 0x2a, 0xbc, 0xce },
{ 0x00, 0x53, 0xa6, 0xf5, 0x51, 0x02, 0xf7, 0xa4,
0xbf, 0xec, 0x19, 0x4a, 0xee, 0xbd, 0x48, 0x1b },
{ 0x00, 0x30, 0x60, 0x50, 0xc0, 0xf0, 0xa0, 0x90,
0x80, 0xb0, 0xe0, 0xd0, 0x40, 0x70, 0x20, 0x10 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x73, 0xe6, 0x95, 0xcc, 0xbf, 0x2a, 0x59,
0x98, 0xeb, 0x7e, 0x0d, 0x54, 0x27, 0xb2, 0xc1 },
{ 0x00, 0x53, 0xa6, 0xf5, 0x4c, 0x1f, 0xea, 0xb9,
0x98, 0xcb, 0x3e, 0x6d, 0xd4, 0x87, 0x72, 0x21 },
{ 0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0,
0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x74, 0xe8, 0x9c, 0xd0, 0xa4, 0x38, 0x4c,
0xa0, 0xd4, 0x48, 0x3c, 0x70, 0x04, 0x98, 0xec },
{ 0x00, 0x53, 0xa6, 0xf5, 0x4c, 0x1f, 0xea, 0xb9,
0x98, 0xcb, 0x3e, 0x6d, 0xd4, 0x87, 0x72, 0x21 },
{ 0x00, 0x50, 0xa0, 0xf0, 0x40, 0x10, 0xe0, 0xb0,
0x80, 0xd0, 0x20, 0x70, 0xc0, 0x90, 0x60, 0x30 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x75, 0xea, 0x9f, 0xd4, 0xa1, 0x3e, 0x4b,
0xa8, 0xdd, 0x42, 0x37, 0x7c, 0x09, 0x96, 0xe3 },
{ 0x00, 0x53, 0xa6, 0xf5, 0x4c, 0x1f, 0xea, 0xb9,
0x85, 0xd6, 0x23, 0x70, 0xc9, 0x9a, 0x6f, 0x3c },
{ 0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20,
0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x76, 0xec, 0x9a, 0xd8, 0xae, 0x34, 0x42,
0xb0, 0xc6, 0x5c, 0x2a, 0x68, 0x1e, 0x84, 0xf2 },
{ 0x00, 0x53, 0xa6, 0xf5, 0x4c, 0x1f, 0xea, 0xb9,
0x85, 0xd6, 0x23, 0x70, 0xc9, 0x9a, 0x6f, 0x3c },
{ 0x00, 0x70, 0xe0, 0x90, 0xc0, 0xb0, 0x20, 0x50,
0x80, 0xf0, 0x60, 0x10, 0x40, 0x30, 0xa0, 0xd0 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x77, 0xee, 0x99, 0xdc, 0xab, 0x32, 0x45,
0xb8, 0xcf, 0x56, 0x21, 0x64, 0x13, 0x8a, 0xfd },
{ 0x00, 0x53, 0xbb, 0xe8, 0x6b, 0x38, 0xd0, 0x83,
0xd6, 0x85, 0x6d, 0x3e, 0xbd, 0xee, 0x06, 0x55 },
{ 0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80,
0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x78, 0xf0, 0x88, 0xe0, 0x98, 0x10, 0x68,
0xc0, 0xb8, 0x30, 0x48, 0x20, 0x58, 0xd0, 0xa8 },
{ 0x00, 0x53, 0xbb, 0xe8, 0x6b, 0x38, 0xd0, 0x83,
0xd6, 0x85, 0x6d, 0x3e, 0xbd, 0xee, 0x06, 0x55 },
{ 0x00, 0x90, 0x20, 0xb0, 0x40, 0xd0, 0x60, 0xf0,
0x80, 0x10, 0xa0, 0x30, 0xc0, 0x50, 0xe0, 0x70 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x79, 0xf2, 0x8b, 0xe4, 0x9d, 0x16, 0x6f,
0xc8, 0xb1, 0x3a, 0x43, 0x2c, 0x55, 0xde, 0xa7 },
{ 0x00, 0x53, 0xbb, 0xe8, 0x6b, 0x38, 0xd0, 0x83,
0xcb, 0x98, 0x70, 0x23, 0xa0, 0xf3, 0x1b, 0x48 },
{ 0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60,
0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x7a, 0xf4, 0x8e, 0xe8, 0x92, 0x1c, 0x66,
0xd0, 0xaa, 0x24, 0x5e, 0x38, 0x42, 0xcc, 0xb6 },
{ 0x00, 0x53, 0xbb, 0xe8, 0x6b, 0x38, 0xd0, 0x83,
0xcb, 0x98, 0x70, 0x23, 0xa0, 0xf3, 0x1b, 0x48 },
{ 0x00, 0xb0, 0x60, 0xd0, 0xc0, 0x70, 0xa0, 0x10,
0x80, 0x30, 0xe0, 0x50, 0x40, 0xf0, 0x20, 0x90 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x7b, 0xf6, 0x8d, 0xec, 0x97, 0x1a, 0x61,
0xd8, 0xa3, 0x2e, 0x55, 0x34, 0x4f, 0xc2, 0xb9 },
{ 0x00, 0x53, 0xbb, 0xe8, 0x76, 0x25, 0xcd, 0x9e,
0xec, 0xbf, 0x57, 0x04, 0x9a, 0xc9, 0x21, 0x72 },
{ 0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40,
0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x7c, 0xf8, 0x84, 0xf0, 0x8c, 0x08, 0x74,
0xe0, 0x9c, 0x18, 0x64, 0x10, 0x6c, 0xe8, 0x94 },
{ 0x00, 0x53, 0xbb, 0xe8, 0x76, 0x25, 0xcd, 0x9e,
0xec, 0xbf, 0x57, 0x04, 0x9a, 0xc9, 0x21, 0x72 },
{ 0x00, 0xd0, 0xa0, 0x70, 0x40, 0x90, 0xe0, 0x30,
0x80, 0x50, 0x20, 0xf0, 0xc0, 0x10, 0x60, 0xb0 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x7d, 0xfa, 0x87, 0xf4, 0x89, 0x0e, 0x73,
0xe8, 0x95, 0x12, 0x6f, 0x1c, 0x61, 0xe6, 0x9b },
{ 0x00, 0x53, 0xbb, 0xe8, 0x76, 0x25, 0xcd, 0x9e,
0xf1, 0xa2, 0x4a, 0x19, 0x87, 0xd4, 0x3c, 0x6f },
{ 0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0,
0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x7e, 0xfc, 0x82, 0xf8, 0x86, 0x04, 0x7a,
0xf0, 0x8e, 0x0c, 0x72, 0x08, 0x76, 0xf4, 0x8a },
{ 0x00, 0x53, 0xbb, 0xe8, 0x76, 0x25, 0xcd, 0x9e,
0xf1, 0xa2, 0x4a, 0x19, 0x87, 0xd4, 0x3c, 0x6f },
{ 0x00, 0xf0, 0xe0, 0x10, 0xc0, 0x30, 0x20, 0xd0,
0x80, 0x70, 0x60, 0x90, 0x40, 0xb0, 0xa0, 0x50 },
{ 0x00, 0x00, 0x00, 0x00, 0x1d, 0x1d, 0x1d, 0x1d,
0x27, 0x27, 0x27, 0x27, 0x3a, 0x3a, 0x3a, 0x3a },
{ 0x00, 0x7f, 0xfe, 0x81, 0xfc, 0x83, 0x02, 0x7d,
0xf8, 0x87, 0x06, 0x79, 0x04, 0x7b, 0xfa, 0x85 },
{ 0x00, 0xe8, 0xcd, 0x25, 0x87, 0x6f, 0x4a, 0xa2,
0x13, 0xfb, 0xde, 0x36, 0x94, 0x7c, 0x59, 0xb1 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80,
0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80 },
{ 0x00, 0xe8, 0xcd, 0x25, 0x87, 0x6f, 0x4a, 0xa2,
0x13, 0xfb, 0xde, 0x36, 0x94, 0x7c, 0x59, 0xb1 },
{ 0x00, 0x10, 0x20, 0x30, 0x40, 0x50, 0x60, 0x70,
0x80, 0x90, 0xa0, 0xb0, 0xc0, 0xd0, 0xe0, 0xf0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x81, 0x02, 0x83, 0x04, 0x85, 0x06, 0x87,
0x08, 0x89, 0x0a, 0x8b, 0x0c, 0x8d, 0x0e, 0x8f },
{ 0x00, 0xe8, 0xcd, 0x25, 0x87, 0x6f, 0x4a, 0xa2,
0x0e, 0xe6, 0xc3, 0x2b, 0x89, 0x61, 0x44, 0xac },
{ 0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0,
0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x82, 0x04, 0x86, 0x08, 0x8a, 0x0c, 0x8e,
0x10, 0x92, 0x14, 0x96, 0x18, 0x9a, 0x1c, 0x9e },
{ 0x00, 0xe8, 0xcd, 0x25, 0x87, 0x6f, 0x4a, 0xa2,
0x0e, 0xe6, 0xc3, 0x2b, 0x89, 0x61, 0x44, 0xac },
{ 0x00, 0x30, 0x60, 0x50, 0xc0, 0xf0, 0xa0, 0x90,
0x80, 0xb0, 0xe0, 0xd0, 0x40, 0x70, 0x20, 0x10 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x83, 0x06, 0x85, 0x0c, 0x8f, 0x0a, 0x89,
0x18, 0x9b, 0x1e, 0x9d, 0x14, 0x97, 0x12, 0x91 },
{ 0x00, 0xe8, 0xcd, 0x25, 0x9a, 0x72, 0x57, 0xbf,
0x29, 0xc1, 0xe4, 0x0c, 0xb3, 0x5b, 0x7e, 0x96 },
{ 0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0,
0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x84, 0x08, 0x8c, 0x10, 0x94, 0x18, 0x9c,
0x20, 0xa4, 0x28, 0xac, 0x30, 0xb4, 0x38, 0xbc },
{ 0x00, 0xe8, 0xcd, 0x25, 0x9a, 0x72, 0x57, 0xbf,
0x29, 0xc1, 0xe4, 0x0c, 0xb3, 0x5b, 0x7e, 0x96 },
{ 0x00, 0x50, 0xa0, 0xf0, 0x40, 0x10, 0xe0, 0xb0,
0x80, 0xd0, 0x20, 0x70, 0xc0, 0x90, 0x60, 0x30 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x85, 0x0a, 0x8f, 0x14, 0x91, 0x1e, 0x9b,
0x28, 0xad, 0x22, 0xa7, 0x3c, 0xb9, 0x36, 0xb3 },
{ 0x00, 0xe8, 0xcd, 0x25, 0x9a, 0x72, 0x57, 0xbf,
0x34, 0xdc, 0xf9, 0x11, 0xae, 0x46, 0x63, 0x8b },
{ 0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20,
0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x86, 0x0c, 0x8a, 0x18, 0x9e, 0x14, 0x92,
0x30, 0xb6, 0x3c, 0xba, 0x28, 0xae, 0x24, 0xa2 },
{ 0x00, 0xe8, 0xcd, 0x25, 0x9a, 0x72, 0x57, 0xbf,
0x34, 0xdc, 0xf9, 0x11, 0xae, 0x46, 0x63, 0x8b },
{ 0x00, 0x70, 0xe0, 0x90, 0xc0, 0xb0, 0x20, 0x50,
0x80, 0xf0, 0x60, 0x10, 0x40, 0x30, 0xa0, 0xd0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x87, 0x0e, 0x89, 0x1c, 0x9b, 0x12, 0x95,
0x38, 0xbf, 0x36, 0xb1, 0x24, 0xa3, 0x2a, 0xad },
{ 0x00, 0xe8, 0xd0, 0x38, 0xbd, 0x55, 0x6d, 0x85,
0x67, 0x8f, 0xb7, 0x5f, 0xda, 0x32, 0x0a, 0xe2 },
{ 0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80,
0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x88, 0x10, 0x98, 0x20, 0xa8, 0x30, 0xb8,
0x40, 0xc8, 0x50, 0xd8, 0x60, 0xe8, 0x70, 0xf8 },
{ 0x00, 0xe8, 0xd0, 0x38, 0xbd, 0x55, 0x6d, 0x85,
0x67, 0x8f, 0xb7, 0x5f, 0xda, 0x32, 0x0a, 0xe2 },
{ 0x00, 0x90, 0x20, 0xb0, 0x40, 0xd0, 0x60, 0xf0,
0x80, 0x10, 0xa0, 0x30, 0xc0, 0x50, 0xe0, 0x70 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x89, 0x12, 0x9b, 0x24, 0xad, 0x36, 0xbf,
0x48, 0xc1, 0x5a, 0xd3, 0x6c, 0xe5, 0x7e, 0xf7 },
{ 0x00, 0xe8, 0xd0, 0x38, 0xbd, 0x55, 0x6d, 0x85,
0x7a, 0x92, 0xaa, 0x42, 0xc7, 0x2f, 0x17, 0xff },
{ 0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60,
0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x8a, 0x14, 0x9e, 0x28, 0xa2, 0x3c, 0xb6,
0x50, 0xda, 0x44, 0xce, 0x78, 0xf2, 0x6c, 0xe6 },
{ 0x00, 0xe8, 0xd0, 0x38, 0xbd, 0x55, 0x6d, 0x85,
0x7a, 0x92, 0xaa, 0x42, 0xc7, 0x2f, 0x17, 0xff },
{ 0x00, 0xb0, 0x60, 0xd0, 0xc0, 0x70, 0xa0, 0x10,
0x80, 0x30, 0xe0, 0x50, 0x40, 0xf0, 0x20, 0x90 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x8b, 0x16, 0x9d, 0x2c, 0xa7, 0x3a, 0xb1,
0x58, 0xd3, 0x4e, 0xc5, 0x74, 0xff, 0x62, 0xe9 },
{ 0x00, 0xe8, 0xd0, 0x38, 0xa0, 0x48, 0x70, 0x98,
0x5d, 0xb5, 0x8d, 0x65, 0xfd, 0x15, 0x2d, 0xc5 },
{ 0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40,
0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x8c, 0x18, 0x94, 0x30, 0xbc, 0x28, 0xa4,
0x60, 0xec, 0x78, 0xf4, 0x50, 0xdc, 0x48, 0xc4 },
{ 0x00, 0xe8, 0xd0, 0x38, 0xa0, 0x48, 0x70, 0x98,
0x5d, 0xb5, 0x8d, 0x65, 0xfd, 0x15, 0x2d, 0xc5 },
{ 0x00, 0xd0, 0xa0, 0x70, 0x40, 0x90, 0xe0, 0x30,
0x80, 0x50, 0x20, 0xf0, 0xc0, 0x10, 0x60, 0xb0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x8d, 0x1a, 0x97, 0x34, 0xb9, 0x2e, 0xa3,
0x68, 0xe5, 0x72, 0xff, 0x5c, 0xd1, 0x46, 0xcb },
{ 0x00, 0xe8, 0xd0, 0x38, 0xa0, 0x48, 0x70, 0x98,
0x40, 0xa8, 0x90, 0x78, 0xe0, 0x08, 0x30, 0xd8 },
{ 0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0,
0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x8e, 0x1c, 0x92, 0x38, 0xb6, 0x24, 0xaa,
0x70, 0xfe, 0x6c, 0xe2, 0x48, 0xc6, 0x54, 0xda },
{ 0x00, 0xe8, 0xd0, 0x38, 0xa0, 0x48, 0x70, 0x98,
0x40, 0xa8, 0x90, 0x78, 0xe0, 0x08, 0x30, 0xd8 },
{ 0x00, 0xf0, 0xe0, 0x10, 0xc0, 0x30, 0x20, 0xd0,
0x80, 0x70, 0x60, 0x90, 0x40, 0xb0, 0xa0, 0x50 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x8f, 0x1e, 0x91, 0x3c, 0xb3, 0x22, 0xad,
0x78, 0xf7, 0x66, 0xe9, 0x44, 0xcb, 0x5a, 0xd5 },
{ 0x00, 0xf5, 0xf7, 0x02, 0xf3, 0x06, 0x04, 0xf1,
0xfb, 0x0e, 0x0c, 0xf9, 0x08, 0xfd, 0xff, 0x0a },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x90, 0x20, 0xb0, 0x40, 0xd0, 0x60, 0xf0,
0x80, 0x10, 0xa0, 0x30, 0xc0, 0x50, 0xe0, 0x70 },
{ 0x00, 0xf5, 0xf7, 0x02, 0xf3, 0x06, 0x04, 0xf1,
0xfb, 0x0e, 0x0c, 0xf9, 0x08, 0xfd, 0xff, 0x0a },
{ 0x00, 0x10, 0x20, 0x30, 0x40, 0x50, 0x60, 0x70,
0x80, 0x90, 0xa0, 0xb0, 0xc0, 0xd0, 0xe0, 0xf0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x91, 0x22, 0xb3, 0x44, 0xd5, 0x66, 0xf7,
0x88, 0x19, 0xaa, 0x3b, 0xcc, 0x5d, 0xee, 0x7f },
{ 0x00, 0xf5, 0xf7, 0x02, 0xf3, 0x06, 0x04, 0xf1,
0xe6, 0x13, 0x11, 0xe4, 0x15, 0xe0, 0xe2, 0x17 },
{ 0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0,
0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x92, 0x24, 0xb6, 0x48, 0xda, 0x6c, 0xfe,
0x90, 0x02, 0xb4, 0x26, 0xd8, 0x4a, 0xfc, 0x6e },
{ 0x00, 0xf5, 0xf7, 0x02, 0xf3, 0x06, 0x04, 0xf1,
0xe6, 0x13, 0x11, 0xe4, 0x15, 0xe0, 0xe2, 0x17 },
{ 0x00, 0x30, 0x60, 0x50, 0xc0, 0xf0, 0xa0, 0x90,
0x80, 0xb0, 0xe0, 0xd0, 0x40, 0x70, 0x20, 0x10 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x93, 0x26, 0xb5, 0x4c, 0xdf, 0x6a, 0xf9,
0x98, 0x0b, 0xbe, 0x2d, 0xd4, 0x47, 0xf2, 0x61 },
{ 0x00, 0xf5, 0xf7, 0x02, 0xee, 0x1b, 0x19, 0xec,
0xc1, 0x34, 0x36, 0xc3, 0x2f, 0xda, 0xd8, 0x2d },
{ 0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0,
0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x94, 0x28, 0xbc, 0x50, 0xc4, 0x78, 0xec,
0xa0, 0x34, 0x88, 0x1c, 0xf0, 0x64, 0xd8, 0x4c },
{ 0x00, 0xf5, 0xf7, 0x02, 0xee, 0x1b, 0x19, 0xec,
0xc1, 0x34, 0x36, 0xc3, 0x2f, 0xda, 0xd8, 0x2d },
{ 0x00, 0x50, 0xa0, 0xf0, 0x40, 0x10, 0xe0, 0xb0,
0x80, 0xd0, 0x20, 0x70, 0xc0, 0x90, 0x60, 0x30 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x95, 0x2a, 0xbf, 0x54, 0xc1, 0x7e, 0xeb,
0xa8, 0x3d, 0x82, 0x17, 0xfc, 0x69, 0xd6, 0x43 },
{ 0x00, 0xf5, 0xf7, 0x02, 0xee, 0x1b, 0x19, 0xec,
0xdc, 0x29, 0x2b, 0xde, 0x32, 0xc7, 0xc5, 0x30 },
{ 0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20,
0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x96, 0x2c, 0xba, 0x58, 0xce, 0x74, 0xe2,
0xb0, 0x26, 0x9c, 0x0a, 0xe8, 0x7e, 0xc4, 0x52 },
{ 0x00, 0xf5, 0xf7, 0x02, 0xee, 0x1b, 0x19, 0xec,
0xdc, 0x29, 0x2b, 0xde, 0x32, 0xc7, 0xc5, 0x30 },
{ 0x00, 0x70, 0xe0, 0x90, 0xc0, 0xb0, 0x20, 0x50,
0x80, 0xf0, 0x60, 0x10, 0x40, 0x30, 0xa0, 0xd0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x97, 0x2e, 0xb9, 0x5c, 0xcb, 0x72, 0xe5,
0xb8, 0x2f, 0x96, 0x01, 0xe4, 0x73, 0xca, 0x5d },
{ 0x00, 0xf5, 0xea, 0x1f, 0xc9, 0x3c, 0x23, 0xd6,
0x8f, 0x7a, 0x65, 0x90, 0x46, 0xb3, 0xac, 0x59 },
{ 0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80,
0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x98, 0x30, 0xa8, 0x60, 0xf8, 0x50, 0xc8,
0xc0, 0x58, 0xf0, 0x68, 0xa0, 0x38, 0x90, 0x08 },
{ 0x00, 0xf5, 0xea, 0x1f, 0xc9, 0x3c, 0x23, 0xd6,
0x8f, 0x7a, 0x65, 0x90, 0x46, 0xb3, 0xac, 0x59 },
{ 0x00, 0x90, 0x20, 0xb0, 0x40, 0xd0, 0x60, 0xf0,
0x80, 0x10, 0xa0, 0x30, 0xc0, 0x50, 0xe0, 0x70 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x99, 0x32, 0xab, 0x64, 0xfd, 0x56, 0xcf,
0xc8, 0x51, 0xfa, 0x63, 0xac, 0x35, 0x9e, 0x07 },
{ 0x00, 0xf5, 0xea, 0x1f, 0xc9, 0x3c, 0x23, 0xd6,
0x92, 0x67, 0x78, 0x8d, 0x5b, 0xae, 0xb1, 0x44 },
{ 0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60,
0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x9a, 0x34, 0xae, 0x68, 0xf2, 0x5c, 0xc6,
0xd0, 0x4a, 0xe4, 0x7e, 0xb8, 0x22, 0x8c, 0x16 },
{ 0x00, 0xf5, 0xea, 0x1f, 0xc9, 0x3c, 0x23, 0xd6,
0x92, 0x67, 0x78, 0x8d, 0x5b, 0xae, 0xb1, 0x44 },
{ 0x00, 0xb0, 0x60, 0xd0, 0xc0, 0x70, 0xa0, 0x10,
0x80, 0x30, 0xe0, 0x50, 0x40, 0xf0, 0x20, 0x90 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x9b, 0x36, 0xad, 0x6c, 0xf7, 0x5a, 0xc1,
0xd8, 0x43, 0xee, 0x75, 0xb4, 0x2f, 0x82, 0x19 },
{ 0x00, 0xf5, 0xea, 0x1f, 0xd4, 0x21, 0x3e, 0xcb,
0xb5, 0x40, 0x5f, 0xaa, 0x61, 0x94, 0x8b, 0x7e },
{ 0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40,
0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x9c, 0x38, 0xa4, 0x70, 0xec, 0x48, 0xd4,
0xe0, 0x7c, 0xd8, 0x44, 0x90, 0x0c, 0xa8, 0x34 },
{ 0x00, 0xf5, 0xea, 0x1f, 0xd4, 0x21, 0x3e, 0xcb,
0xb5, 0x40, 0x5f, 0xaa, 0x61, 0x94, 0x8b, 0x7e },
{ 0x00, 0xd0, 0xa0, 0x70, 0x40, 0x90, 0xe0, 0x30,
0x80, 0x50, 0x20, 0xf0, 0xc0, 0x10, 0x60, 0xb0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x9d, 0x3a, 0xa7, 0x74, 0xe9, 0x4e, 0xd3,
0xe8, 0x75, 0xd2, 0x4f, 0x9c, 0x01, 0xa6, 0x3b },
{ 0x00, 0xf5, 0xea, 0x1f, 0xd4, 0x21, 0x3e, 0xcb,
0xa8, 0x5d, 0x42, 0xb7, 0x7c, 0x89, 0x96, 0x63 },
{ 0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0,
0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x9e, 0x3c, 0xa2, 0x78, 0xe6, 0x44, 0xda,
0xf0, 0x6e, 0xcc, 0x52, 0x88, 0x16, 0xb4, 0x2a },
{ 0x00, 0xf5, 0xea, 0x1f, 0xd4, 0x21, 0x3e, 0xcb,
0xa8, 0x5d, 0x42, 0xb7, 0x7c, 0x89, 0x96, 0x63 },
{ 0x00, 0xf0, 0xe0, 0x10, 0xc0, 0x30, 0x20, 0xd0,
0x80, 0x70, 0x60, 0x90, 0x40, 0xb0, 0xa0, 0x50 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x74, 0x74, 0x69, 0x69, 0x4e, 0x4e, 0x53, 0x53 },
{ 0x00, 0x9f, 0x3e, 0xa1, 0x7c, 0xe3, 0x42, 0xdd,
0xf8, 0x67, 0xc6, 0x59, 0x84, 0x1b, 0xba, 0x25 },
{ 0x00, 0xd2, 0xb9, 0x6b, 0x6f, 0xbd, 0xd6, 0x04,
0xde, 0x0c, 0x67, 0xb5, 0xb1, 0x63, 0x08, 0xda },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60,
0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60 },
{ 0x00, 0xd2, 0xb9, 0x6b, 0x6f, 0xbd, 0xd6, 0x04,
0xde, 0x0c, 0x67, 0xb5, 0xb1, 0x63, 0x08, 0xda },
{ 0x00, 0x10, 0x20, 0x30, 0x40, 0x50, 0x60, 0x70,
0x80, 0x90, 0xa0, 0xb0, 0xc0, 0xd0, 0xe0, 0xf0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xa1, 0x42, 0xe3, 0x84, 0x25, 0xc6, 0x67,
0x08, 0xa9, 0x4a, 0xeb, 0x8c, 0x2d, 0xce, 0x6f },
{ 0x00, 0xd2, 0xb9, 0x6b, 0x6f, 0xbd, 0xd6, 0x04,
0xc3, 0x11, 0x7a, 0xa8, 0xac, 0x7e, 0x15, 0xc7 },
{ 0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0,
0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xa2, 0x44, 0xe6, 0x88, 0x2a, 0xcc, 0x6e,
0x10, 0xb2, 0x54, 0xf6, 0x98, 0x3a, 0xdc, 0x7e },
{ 0x00, 0xd2, 0xb9, 0x6b, 0x6f, 0xbd, 0xd6, 0x04,
0xc3, 0x11, 0x7a, 0xa8, 0xac, 0x7e, 0x15, 0xc7 },
{ 0x00, 0x30, 0x60, 0x50, 0xc0, 0xf0, 0xa0, 0x90,
0x80, 0xb0, 0xe0, 0xd0, 0x40, 0x70, 0x20, 0x10 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xa3, 0x46, 0xe5, 0x8c, 0x2f, 0xca, 0x69,
0x18, 0xbb, 0x5e, 0xfd, 0x94, 0x37, 0xd2, 0x71 },
{ 0x00, 0xd2, 0xb9, 0x6b, 0x72, 0xa0, 0xcb, 0x19,
0xe4, 0x36, 0x5d, 0x8f, 0x96, 0x44, 0x2f, 0xfd },
{ 0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0,
0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xa4, 0x48, 0xec, 0x90, 0x34, 0xd8, 0x7c,
0x20, 0x84, 0x68, 0xcc, 0xb0, 0x14, 0xf8, 0x5c },
{ 0x00, 0xd2, 0xb9, 0x6b, 0x72, 0xa0, 0xcb, 0x19,
0xe4, 0x36, 0x5d, 0x8f, 0x96, 0x44, 0x2f, 0xfd },
{ 0x00, 0x50, 0xa0, 0xf0, 0x40, 0x10, 0xe0, 0xb0,
0x80, 0xd0, 0x20, 0x70, 0xc0, 0x90, 0x60, 0x30 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xa5, 0x4a, 0xef, 0x94, 0x31, 0xde, 0x7b,
0x28, 0x8d, 0x62, 0xc7, 0xbc, 0x19, 0xf6, 0x53 },
{ 0x00, 0xd2, 0xb9, 0x6b, 0x72, 0xa0, 0xcb, 0x19,
0xf9, 0x2b, 0x40, 0x92, 0x8b, 0x59, 0x32, 0xe0 },
{ 0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20,
0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xa6, 0x4c, 0xea, 0x98, 0x3e, 0xd4, 0x72,
0x30, 0x96, 0x7c, 0xda, 0xa8, 0x0e, 0xe4, 0x42 },
{ 0x00, 0xd2, 0xb9, 0x6b, 0x72, 0xa0, 0xcb, 0x19,
0xf9, 0x2b, 0x40, 0x92, 0x8b, 0x59, 0x32, 0xe0 },
{ 0x00, 0x70, 0xe0, 0x90, 0xc0, 0xb0, 0x20, 0x50,
0x80, 0xf0, 0x60, 0x10, 0x40, 0x30, 0xa0, 0xd0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xa7, 0x4e, 0xe9, 0x9c, 0x3b, 0xd2, 0x75,
0x38, 0x9f, 0x76, 0xd1, 0xa4, 0x03, 0xea, 0x4d },
{ 0x00, 0xd2, 0xa4, 0x76, 0x55, 0x87, 0xf1, 0x23,
0xaa, 0x78, 0x0e, 0xdc, 0xff, 0x2d, 0x5b, 0x89 },
{ 0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80,
0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xa8, 0x50, 0xf8, 0xa0, 0x08, 0xf0, 0x58,
0x40, 0xe8, 0x10, 0xb8, 0xe0, 0x48, 0xb0, 0x18 },
{ 0x00, 0xd2, 0xa4, 0x76, 0x55, 0x87, 0xf1, 0x23,
0xaa, 0x78, 0x0e, 0xdc, 0xff, 0x2d, 0x5b, 0x89 },
{ 0x00, 0x90, 0x20, 0xb0, 0x40, 0xd0, 0x60, 0xf0,
0x80, 0x10, 0xa0, 0x30, 0xc0, 0x50, 0xe0, 0x70 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xa9, 0x52, 0xfb, 0xa4, 0x0d, 0xf6, 0x5f,
0x48, 0xe1, 0x1a, 0xb3, 0xec, 0x45, 0xbe, 0x17 },
{ 0x00, 0xd2, 0xa4, 0x76, 0x55, 0x87, 0xf1, 0x23,
0xb7, 0x65, 0x13, 0xc1, 0xe2, 0x30, 0x46, 0x94 },
{ 0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60,
0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xaa, 0x54, 0xfe, 0xa8, 0x02, 0xfc, 0x56,
0x50, 0xfa, 0x04, 0xae, 0xf8, 0x52, 0xac, 0x06 },
{ 0x00, 0xd2, 0xa4, 0x76, 0x55, 0x87, 0xf1, 0x23,
0xb7, 0x65, 0x13, 0xc1, 0xe2, 0x30, 0x46, 0x94 },
{ 0x00, 0xb0, 0x60, 0xd0, 0xc0, 0x70, 0xa0, 0x10,
0x80, 0x30, 0xe0, 0x50, 0x40, 0xf0, 0x20, 0x90 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xab, 0x56, 0xfd, 0xac, 0x07, 0xfa, 0x51,
0x58, 0xf3, 0x0e, 0xa5, 0xf4, 0x5f, 0xa2, 0x09 },
{ 0x00, 0xd2, 0xa4, 0x76, 0x48, 0x9a, 0xec, 0x3e,
0x90, 0x42, 0x34, 0xe6, 0xd8, 0x0a, 0x7c, 0xae },
{ 0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40,
0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xac, 0x58, 0xf4, 0xb0, 0x1c, 0xe8, 0x44,
0x60, 0xcc, 0x38, 0x94, 0xd0, 0x7c, 0x88, 0x24 },
{ 0x00, 0xd2, 0xa4, 0x76, 0x48, 0x9a, 0xec, 0x3e,
0x90, 0x42, 0x34, 0xe6, 0xd8, 0x0a, 0x7c, 0xae },
{ 0x00, 0xd0, 0xa0, 0x70, 0x40, 0x90, 0xe0, 0x30,
0x80, 0x50, 0x20, 0xf0, 0xc0, 0x10, 0x60, 0xb0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xad, 0x5a, 0xf7, 0xb4, 0x19, 0xee, 0x43,
0x68, 0xc5, 0x32, 0x9f, 0xdc, 0x71, 0x86, 0x2b },
{ 0x00, 0xd2, 0xa4, 0x76, 0x48, 0x9a, 0xec, 0x3e,
0x8d, 0x5f, 0x29, 0xfb, 0xc5, 0x17, 0x61, 0xb3 },
{ 0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0,
0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xae, 0x5c, 0xf2, 0xb8, 0x16, 0xe4, 0x4a,
0x70, 0xde, 0x2c, 0x82, 0xc8, 0x66, 0x94, 0x3a },
{ 0x00, 0xd2, 0xa4, 0x76, 0x48, 0x9a, 0xec, 0x3e,
0x8d, 0x5f, 0x29, 0xfb, 0xc5, 0x17, 0x61, 0xb3 },
{ 0x00, 0xf0, 0xe0, 0x10, 0xc0, 0x30, 0x20, 0xd0,
0x80, 0x70, 0x60, 0x90, 0x40, 0xb0, 0xa0, 0x50 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xaf, 0x5e, 0xf1, 0xbc, 0x13, 0xe2, 0x4d,
0x78, 0xd7, 0x26, 0x89, 0xc4, 0x6b, 0x9a, 0x35 },
{ 0x00, 0xcf, 0x83, 0x4c, 0x1b, 0xd4, 0x98, 0x57,
0x36, 0xf9, 0xb5, 0x7a, 0x2d, 0xe2, 0xae, 0x61 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xb0, 0x60, 0xd0, 0xc0, 0x70, 0xa0, 0x10,
0x80, 0x30, 0xe0, 0x50, 0x40, 0xf0, 0x20, 0x90 },
{ 0x00, 0xcf, 0x83, 0x4c, 0x1b, 0xd4, 0x98, 0x57,
0x36, 0xf9, 0xb5, 0x7a, 0x2d, 0xe2, 0xae, 0x61 },
{ 0x00, 0x10, 0x20, 0x30, 0x40, 0x50, 0x60, 0x70,
0x80, 0x90, 0xa0, 0xb0, 0xc0, 0xd0, 0xe0, 0xf0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xb1, 0x62, 0xd3, 0xc4, 0x75, 0xa6, 0x17,
0x88, 0x39, 0xea, 0x5b, 0x4c, 0xfd, 0x2e, 0x9f },
{ 0x00, 0xcf, 0x83, 0x4c, 0x1b, 0xd4, 0x98, 0x57,
0x2b, 0xe4, 0xa8, 0x67, 0x30, 0xff, 0xb3, 0x7c },
{ 0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0,
0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xb2, 0x64, 0xd6, 0xc8, 0x7a, 0xac, 0x1e,
0x90, 0x22, 0xf4, 0x46, 0x58, 0xea, 0x3c, 0x8e },
{ 0x00, 0xcf, 0x83, 0x4c, 0x1b, 0xd4, 0x98, 0x57,
0x2b, 0xe4, 0xa8, 0x67, 0x30, 0xff, 0xb3, 0x7c },
{ 0x00, 0x30, 0x60, 0x50, 0xc0, 0xf0, 0xa0, 0x90,
0x80, 0xb0, 0xe0, 0xd0, 0x40, 0x70, 0x20, 0x10 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xb3, 0x66, 0xd5, 0xcc, 0x7f, 0xaa, 0x19,
0x98, 0x2b, 0xfe, 0x4d, 0x54, 0xe7, 0x32, 0x81 },
{ 0x00, 0xcf, 0x83, 0x4c, 0x06, 0xc9, 0x85, 0x4a,
0x0c, 0xc3, 0x8f, 0x40, 0x0a, 0xc5, 0x89, 0x46 },
{ 0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0,
0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xb4, 0x68, 0xdc, 0xd0, 0x64, 0xb8, 0x0c,
0xa0, 0x14, 0xc8, 0x7c, 0x70, 0xc4, 0x18, 0xac },
{ 0x00, 0xcf, 0x83, 0x4c, 0x06, 0xc9, 0x85, 0x4a,
0x0c, 0xc3, 0x8f, 0x40, 0x0a, 0xc5, 0x89, 0x46 },
{ 0x00, 0x50, 0xa0, 0xf0, 0x40, 0x10, 0xe0, 0xb0,
0x80, 0xd0, 0x20, 0x70, 0xc0, 0x90, 0x60, 0x30 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xb5, 0x6a, 0xdf, 0xd4, 0x61, 0xbe, 0x0b,
0xa8, 0x1d, 0xc2, 0x77, 0x7c, 0xc9, 0x16, 0xa3 },
{ 0x00, 0xcf, 0x83, 0x4c, 0x06, 0xc9, 0x85, 0x4a,
0x11, 0xde, 0x92, 0x5d, 0x17, 0xd8, 0x94, 0x5b },
{ 0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20,
0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xb6, 0x6c, 0xda, 0xd8, 0x6e, 0xb4, 0x02,
0xb0, 0x06, 0xdc, 0x6a, 0x68, 0xde, 0x04, 0xb2 },
{ 0x00, 0xcf, 0x83, 0x4c, 0x06, 0xc9, 0x85, 0x4a,
0x11, 0xde, 0x92, 0x5d, 0x17, 0xd8, 0x94, 0x5b },
{ 0x00, 0x70, 0xe0, 0x90, 0xc0, 0xb0, 0x20, 0x50,
0x80, 0xf0, 0x60, 0x10, 0x40, 0x30, 0xa0, 0xd0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xb7, 0x6e, 0xd9, 0xdc, 0x6b, 0xb2, 0x05,
0xb8, 0x0f, 0xd6, 0x61, 0x64, 0xd3, 0x0a, 0xbd },
{ 0x00, 0xcf, 0x9e, 0x51, 0x21, 0xee, 0xbf, 0x70,
0x42, 0x8d, 0xdc, 0x13, 0x63, 0xac, 0xfd, 0x32 },
{ 0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80,
0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xb8, 0x70, 0xc8, 0xe0, 0x58, 0x90, 0x28,
0xc0, 0x78, 0xb0, 0x08, 0x20, 0x98, 0x50, 0xe8 },
{ 0x00, 0xcf, 0x9e, 0x51, 0x21, 0xee, 0xbf, 0x70,
0x42, 0x8d, 0xdc, 0x13, 0x63, 0xac, 0xfd, 0x32 },
{ 0x00, 0x90, 0x20, 0xb0, 0x40, 0xd0, 0x60, 0xf0,
0x80, 0x10, 0xa0, 0x30, 0xc0, 0x50, 0xe0, 0x70 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xb9, 0x72, 0xcb, 0xe4, 0x5d, 0x96, 0x2f,
0xc8, 0x71, 0xba, 0x03, 0x2c, 0x95, 0x5e, 0xe7 },
{ 0x00, 0xcf, 0x9e, 0x51, 0x21, 0xee, 0xbf, 0x70,
0x5f, 0x90, 0xc1, 0x0e, 0x7e, 0xb1, 0xe0, 0x2f },
{ 0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60,
0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xba, 0x74, 0xce, 0xe8, 0x52, 0x9c, 0x26,
0xd0, 0x6a, 0xa4, 0x1e, 0x38, 0x82, 0x4c, 0xf6 },
{ 0x00, 0xcf, 0x9e, 0x51, 0x21, 0xee, 0xbf, 0x70,
0x5f, 0x90, 0xc1, 0x0e, 0x7e, 0xb1, 0xe0, 0x2f },
{ 0x00, 0xb0, 0x60, 0xd0, 0xc0, 0x70, 0xa0, 0x10,
0x80, 0x30, 0xe0, 0x50, 0x40, 0xf0, 0x20, 0x90 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xbb, 0x76, 0xcd, 0xec, 0x57, 0x9a, 0x21,
0xd8, 0x63, 0xae, 0x15, 0x34, 0x8f, 0x42, 0xf9 },
{ 0x00, 0xcf, 0x9e, 0x51, 0x3c, 0xf3, 0xa2, 0x6d,
0x78, 0xb7, 0xe6, 0x29, 0x44, 0x8b, 0xda, 0x15 },
{ 0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40,
0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xbc, 0x78, 0xc4, 0xf0, 0x4c, 0x88, 0x34,
0xe0, 0x5c, 0x98, 0x24, 0x10, 0xac, 0x68, 0xd4 },
{ 0x00, 0xcf, 0x9e, 0x51, 0x3c, 0xf3, 0xa2, 0x6d,
0x78, 0xb7, 0xe6, 0x29, 0x44, 0x8b, 0xda, 0x15 },
{ 0x00, 0xd0, 0xa0, 0x70, 0x40, 0x90, 0xe0, 0x30,
0x80, 0x50, 0x20, 0xf0, 0xc0, 0x10, 0x60, 0xb0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xbd, 0x7a, 0xc7, 0xf4, 0x49, 0x8e, 0x33,
0xe8, 0x55, 0x92, 0x2f, 0x1c, 0xa1, 0x66, 0xdb },
{ 0x00, 0xcf, 0x9e, 0x51, 0x3c, 0xf3, 0xa2, 0x6d,
0x65, 0xaa, 0xfb, 0x34, 0x59, 0x96, 0xc7, 0x08 },
{ 0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0,
0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xbe, 0x7c, 0xc2, 0xf8, 0x46, 0x84, 0x3a,
0xf0, 0x4e, 0x8c, 0x32, 0x08, 0xb6, 0x74, 0xca },
{ 0x00, 0xcf, 0x9e, 0x51, 0x3c, 0xf3, 0xa2, 0x6d,
0x65, 0xaa, 0xfb, 0x34, 0x59, 0x96, 0xc7, 0x08 },
{ 0x00, 0xf0, 0xe0, 0x10, 0xc0, 0x30, 0x20, 0xd0,
0x80, 0x70, 0x60, 0x90, 0x40, 0xb0, 0xa0, 0x50 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x3a, 0x3a, 0x27, 0x27,
0x69, 0x69, 0x74, 0x74, 0x53, 0x53, 0x4e, 0x4e },
{ 0x00, 0xbf, 0x7e, 0xc1, 0xfc, 0x43, 0x82, 0x3d,
0xf8, 0x47, 0x86, 0x39, 0x04, 0xbb, 0x7a, 0xc5 },
{ 0x00, 0x9c, 0x25, 0xb9, 0x4a, 0xd6, 0x6f, 0xf3,
0x94, 0x08, 0xb1, 0x2d, 0xde, 0x42, 0xfb, 0x67 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40,
0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40 },
{ 0x00, 0x9c, 0x25, 0xb9, 0x4a, 0xd6, 0x6f, 0xf3,
0x94, 0x08, 0xb1, 0x2d, 0xde, 0x42, 0xfb, 0x67 },
{ 0x00, 0x10, 0x20, 0x30, 0x40, 0x50, 0x60, 0x70,
0x80, 0x90, 0xa0, 0xb0, 0xc0, 0xd0, 0xe0, 0xf0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xc1, 0x82, 0x43, 0x04, 0xc5, 0x86, 0x47,
0x08, 0xc9, 0x8a, 0x4b, 0x0c, 0xcd, 0x8e, 0x4f },
{ 0x00, 0x9c, 0x25, 0xb9, 0x4a, 0xd6, 0x6f, 0xf3,
0x89, 0x15, 0xac, 0x30, 0xc3, 0x5f, 0xe6, 0x7a },
{ 0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0,
0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xc2, 0x84, 0x46, 0x08, 0xca, 0x8c, 0x4e,
0x10, 0xd2, 0x94, 0x56, 0x18, 0xda, 0x9c, 0x5e },
{ 0x00, 0x9c, 0x25, 0xb9, 0x4a, 0xd6, 0x6f, 0xf3,
0x89, 0x15, 0xac, 0x30, 0xc3, 0x5f, 0xe6, 0x7a },
{ 0x00, 0x30, 0x60, 0x50, 0xc0, 0xf0, 0xa0, 0x90,
0x80, 0xb0, 0xe0, 0xd0, 0x40, 0x70, 0x20, 0x10 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xc3, 0x86, 0x45, 0x0c, 0xcf, 0x8a, 0x49,
0x18, 0xdb, 0x9e, 0x5d, 0x14, 0xd7, 0x92, 0x51 },
{ 0x00, 0x9c, 0x25, 0xb9, 0x57, 0xcb, 0x72, 0xee,
0xae, 0x32, 0x8b, 0x17, 0xf9, 0x65, 0xdc, 0x40 },
{ 0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0,
0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xc4, 0x88, 0x4c, 0x10, 0xd4, 0x98, 0x5c,
0x20, 0xe4, 0xa8, 0x6c, 0x30, 0xf4, 0xb8, 0x7c },
{ 0x00, 0x9c, 0x25, 0xb9, 0x57, 0xcb, 0x72, 0xee,
0xae, 0x32, 0x8b, 0x17, 0xf9, 0x65, 0xdc, 0x40 },
{ 0x00, 0x50, 0xa0, 0xf0, 0x40, 0x10, 0xe0, 0xb0,
0x80, 0xd0, 0x20, 0x70, 0xc0, 0x90, 0x60, 0x30 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xc5, 0x8a, 0x4f, 0x14, 0xd1, 0x9e, 0x5b,
0x28, 0xed, 0xa2, 0x67, 0x3c, 0xf9, 0xb6, 0x73 },
{ 0x00, 0x9c, 0x25, 0xb9, 0x57, 0xcb, 0x72, 0xee,
0xb3, 0x2f, 0x96, 0x0a, 0xe4, 0x78, 0xc1, 0x5d },
{ 0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20,
0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xc6, 0x8c, 0x4a, 0x18, 0xde, 0x94, 0x52,
0x30, 0xf6, 0xbc, 0x7a, 0x28, 0xee, 0xa4, 0x62 },
{ 0x00, 0x9c, 0x25, 0xb9, 0x57, 0xcb, 0x72, 0xee,
0xb3, 0x2f, 0x96, 0x0a, 0xe4, 0x78, 0xc1, 0x5d },
{ 0x00, 0x70, 0xe0, 0x90, 0xc0, 0xb0, 0x20, 0x50,
0x80, 0xf0, 0x60, 0x10, 0x40, 0x30, 0xa0, 0xd0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xc7, 0x8e, 0x49, 0x1c, 0xdb, 0x92, 0x55,
0x38, 0xff, 0xb6, 0x71, 0x24, 0xe3, 0xaa, 0x6d },
{ 0x00, 0x9c, 0x38, 0xa4, 0x70, 0xec, 0x48, 0xd4,
0xe0, 0x7c, 0xd8, 0x44, 0x90, 0x0c, 0xa8, 0x34 },
{ 0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80,
0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xc8, 0x90, 0x58, 0x20, 0xe8, 0xb0, 0x78,
0x40, 0x88, 0xd0, 0x18, 0x60, 0xa8, 0xf0, 0x38 },
{ 0x00, 0x9c, 0x38, 0xa4, 0x70, 0xec, 0x48, 0xd4,
0xe0, 0x7c, 0xd8, 0x44, 0x90, 0x0c, 0xa8, 0x34 },
{ 0x00, 0x90, 0x20, 0xb0, 0x40, 0xd0, 0x60, 0xf0,
0x80, 0x10, 0xa0, 0x30, 0xc0, 0x50, 0xe0, 0x70 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xc9, 0x92, 0x5b, 0x24, 0xed, 0xb6, 0x7f,
0x48, 0x81, 0xda, 0x13, 0x6c, 0xa5, 0xfe, 0x37 },
{ 0x00, 0x9c, 0x38, 0xa4, 0x70, 0xec, 0x48, 0xd4,
0xfd, 0x61, 0xc5, 0x59, 0x8d, 0x11, 0xb5, 0x29 },
{ 0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60,
0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xca, 0x94, 0x5e, 0x28, 0xe2, 0xbc, 0x76,
0x50, 0x9a, 0xc4, 0x0e, 0x78, 0xb2, 0xec, 0x26 },
{ 0x00, 0x9c, 0x38, 0xa4, 0x70, 0xec, 0x48, 0xd4,
0xfd, 0x61, 0xc5, 0x59, 0x8d, 0x11, 0xb5, 0x29 },
{ 0x00, 0xb0, 0x60, 0xd0, 0xc0, 0x70, 0xa0, 0x10,
0x80, 0x30, 0xe0, 0x50, 0x40, 0xf0, 0x20, 0x90 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xcb, 0x96, 0x5d, 0x2c, 0xe7, 0xba, 0x71,
0x58, 0x93, 0xce, 0x05, 0x74, 0xbf, 0xe2, 0x29 },
{ 0x00, 0x9c, 0x38, 0xa4, 0x6d, 0xf1, 0x55, 0xc9,
0xda, 0x46, 0xe2, 0x7e, 0xb7, 0x2b, 0x8f, 0x13 },
{ 0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40,
0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xcc, 0x98, 0x54, 0x30, 0xfc, 0xa8, 0x64,
0x60, 0xac, 0xf8, 0x34, 0x50, 0x9c, 0xc8, 0x04 },
{ 0x00, 0x9c, 0x38, 0xa4, 0x6d, 0xf1, 0x55, 0xc9,
0xda, 0x46, 0xe2, 0x7e, 0xb7, 0x2b, 0x8f, 0x13 },
{ 0x00, 0xd0, 0xa0, 0x70, 0x40, 0x90, 0xe0, 0x30,
0x80, 0x50, 0x20, 0xf0, 0xc0, 0x10, 0x60, 0xb0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xcd, 0x9a, 0x57, 0x34, 0xf9, 0xae, 0x63,
0x68, 0xa5, 0xf2, 0x3f, 0x5c, 0x91, 0xc6, 0x0b },
{ 0x00, 0x9c, 0x38, 0xa4, 0x6d, 0xf1, 0x55, 0xc9,
0xc7, 0x5b, 0xff, 0x63, 0xaa, 0x36, 0x92, 0x0e },
{ 0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0,
0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xce, 0x9c, 0x52, 0x38, 0xf6, 0xa4, 0x6a,
0x70, 0xbe, 0xec, 0x22, 0x48, 0x86, 0xd4, 0x1a },
{ 0x00, 0x9c, 0x38, 0xa4, 0x6d, 0xf1, 0x55, 0xc9,
0xc7, 0x5b, 0xff, 0x63, 0xaa, 0x36, 0x92, 0x0e },
{ 0x00, 0xf0, 0xe0, 0x10, 0xc0, 0x30, 0x20, 0xd0,
0x80, 0x70, 0x60, 0x90, 0x40, 0xb0, 0xa0, 0x50 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xcf, 0x9e, 0x51, 0x3c, 0xf3, 0xa2, 0x6d,
0x78, 0xb7, 0xe6, 0x29, 0x44, 0x8b, 0xda, 0x15 },
{ 0x00, 0x81, 0x1f, 0x9e, 0x3e, 0xbf, 0x21, 0xa0,
0x7c, 0xfd, 0x63, 0xe2, 0x42, 0xc3, 0x5d, 0xdc },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xd0, 0xa0, 0x70, 0x40, 0x90, 0xe0, 0x30,
0x80, 0x50, 0x20, 0xf0, 0xc0, 0x10, 0x60, 0xb0 },
{ 0x00, 0x81, 0x1f, 0x9e, 0x3e, 0xbf, 0x21, 0xa0,
0x7c, 0xfd, 0x63, 0xe2, 0x42, 0xc3, 0x5d, 0xdc },
{ 0x00, 0x10, 0x20, 0x30, 0x40, 0x50, 0x60, 0x70,
0x80, 0x90, 0xa0, 0xb0, 0xc0, 0xd0, 0xe0, 0xf0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xd1, 0xa2, 0x73, 0x44, 0x95, 0xe6, 0x37,
0x88, 0x59, 0x2a, 0xfb, 0xcc, 0x1d, 0x6e, 0xbf },
{ 0x00, 0x81, 0x1f, 0x9e, 0x3e, 0xbf, 0x21, 0xa0,
0x61, 0xe0, 0x7e, 0xff, 0x5f, 0xde, 0x40, 0xc1 },
{ 0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0,
0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xd2, 0xa4, 0x76, 0x48, 0x9a, 0xec, 0x3e,
0x90, 0x42, 0x34, 0xe6, 0xd8, 0x0a, 0x7c, 0xae },
{ 0x00, 0x81, 0x1f, 0x9e, 0x3e, 0xbf, 0x21, 0xa0,
0x61, 0xe0, 0x7e, 0xff, 0x5f, 0xde, 0x40, 0xc1 },
{ 0x00, 0x30, 0x60, 0x50, 0xc0, 0xf0, 0xa0, 0x90,
0x80, 0xb0, 0xe0, 0xd0, 0x40, 0x70, 0x20, 0x10 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xd3, 0xa6, 0x75, 0x4c, 0x9f, 0xea, 0x39,
0x98, 0x4b, 0x3e, 0xed, 0xd4, 0x07, 0x72, 0xa1 },
{ 0x00, 0x81, 0x1f, 0x9e, 0x23, 0xa2, 0x3c, 0xbd,
0x46, 0xc7, 0x59, 0xd8, 0x65, 0xe4, 0x7a, 0xfb },
{ 0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0,
0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xd4, 0xa8, 0x7c, 0x50, 0x84, 0xf8, 0x2c,
0xa0, 0x74, 0x08, 0xdc, 0xf0, 0x24, 0x58, 0x8c },
{ 0x00, 0x81, 0x1f, 0x9e, 0x23, 0xa2, 0x3c, 0xbd,
0x46, 0xc7, 0x59, 0xd8, 0x65, 0xe4, 0x7a, 0xfb },
{ 0x00, 0x50, 0xa0, 0xf0, 0x40, 0x10, 0xe0, 0xb0,
0x80, 0xd0, 0x20, 0x70, 0xc0, 0x90, 0x60, 0x30 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xd5, 0xaa, 0x7f, 0x54, 0x81, 0xfe, 0x2b,
0xa8, 0x7d, 0x02, 0xd7, 0xfc, 0x29, 0x56, 0x83 },
{ 0x00, 0x81, 0x1f, 0x9e, 0x23, 0xa2, 0x3c, 0xbd,
0x5b, 0xda, 0x44, 0xc5, 0x78, 0xf9, 0x67, 0xe6 },
{ 0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20,
0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xd6, 0xac, 0x7a, 0x58, 0x8e, 0xf4, 0x22,
0xb0, 0x66, 0x1c, 0xca, 0xe8, 0x3e, 0x44, 0x92 },
{ 0x00, 0x81, 0x1f, 0x9e, 0x23, 0xa2, 0x3c, 0xbd,
0x5b, 0xda, 0x44, 0xc5, 0x78, 0xf9, 0x67, 0xe6 },
{ 0x00, 0x70, 0xe0, 0x90, 0xc0, 0xb0, 0x20, 0x50,
0x80, 0xf0, 0x60, 0x10, 0x40, 0x30, 0xa0, 0xd0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xd7, 0xae, 0x79, 0x5c, 0x8b, 0xf2, 0x25,
0xb8, 0x6f, 0x16, 0xc1, 0xe4, 0x33, 0x4a, 0x9d },
{ 0x00, 0x81, 0x02, 0x83, 0x04, 0x85, 0x06, 0x87,
0x08, 0x89, 0x0a, 0x8b, 0x0c, 0x8d, 0x0e, 0x8f },
{ 0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80,
0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xd8, 0xb0, 0x68, 0x60, 0xb8, 0xd0, 0x08,
0xc0, 0x18, 0x70, 0xa8, 0xa0, 0x78, 0x10, 0xc8 },
{ 0x00, 0x81, 0x02, 0x83, 0x04, 0x85, 0x06, 0x87,
0x08, 0x89, 0x0a, 0x8b, 0x0c, 0x8d, 0x0e, 0x8f },
{ 0x00, 0x90, 0x20, 0xb0, 0x40, 0xd0, 0x60, 0xf0,
0x80, 0x10, 0xa0, 0x30, 0xc0, 0x50, 0xe0, 0x70 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xd9, 0xb2, 0x6b, 0x64, 0xbd, 0xd6, 0x0f,
0xc8, 0x11, 0x7a, 0xa3, 0xac, 0x75, 0x1e, 0xc7 },
{ 0x00, 0x81, 0x02, 0x83, 0x04, 0x85, 0x06, 0x87,
0x15, 0x94, 0x17, 0x96, 0x11, 0x90, 0x13, 0x92 },
{ 0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60,
0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xda, 0xb4, 0x6e, 0x68, 0xb2, 0xdc, 0x06,
0xd0, 0x0a, 0x64, 0xbe, 0xb8, 0x62, 0x0c, 0xd6 },
{ 0x00, 0x81, 0x02, 0x83, 0x04, 0x85, 0x06, 0x87,
0x15, 0x94, 0x17, 0x96, 0x11, 0x90, 0x13, 0x92 },
{ 0x00, 0xb0, 0x60, 0xd0, 0xc0, 0x70, 0xa0, 0x10,
0x80, 0x30, 0xe0, 0x50, 0x40, 0xf0, 0x20, 0x90 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xdb, 0xb6, 0x6d, 0x6c, 0xb7, 0xda, 0x01,
0xd8, 0x03, 0x6e, 0xb5, 0xb4, 0x6f, 0x02, 0xd9 },
{ 0x00, 0x81, 0x02, 0x83, 0x19, 0x98, 0x1b, 0x9a,
0x32, 0xb3, 0x30, 0xb1, 0x2b, 0xaa, 0x29, 0xa8 },
{ 0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40,
0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xdc, 0xb8, 0x64, 0x70, 0xac, 0xc8, 0x14,
0xe0, 0x3c, 0x58, 0x84, 0x90, 0x4c, 0x28, 0xf4 },
{ 0x00, 0x81, 0x02, 0x83, 0x19, 0x98, 0x1b, 0x9a,
0x32, 0xb3, 0x30, 0xb1, 0x2b, 0xaa, 0x29, 0xa8 },
{ 0x00, 0xd0, 0xa0, 0x70, 0x40, 0x90, 0xe0, 0x30,
0x80, 0x50, 0x20, 0xf0, 0xc0, 0x10, 0x60, 0xb0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xdd, 0xba, 0x67, 0x74, 0xa9, 0xce, 0x13,
0xe8, 0x35, 0x52, 0x8f, 0x9c, 0x41, 0x26, 0xfb },
{ 0x00, 0x81, 0x02, 0x83, 0x19, 0x98, 0x1b, 0x9a,
0x2f, 0xae, 0x2d, 0xac, 0x36, 0xb7, 0x34, 0xb5 },
{ 0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0,
0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xde, 0xbc, 0x62, 0x78, 0xa6, 0xc4, 0x1a,
0xf0, 0x2e, 0x4c, 0x92, 0x88, 0x56, 0x34, 0xea },
{ 0x00, 0x81, 0x02, 0x83, 0x19, 0x98, 0x1b, 0x9a,
0x2f, 0xae, 0x2d, 0xac, 0x36, 0xb7, 0x34, 0xb5 },
{ 0x00, 0xf0, 0xe0, 0x10, 0xc0, 0x30, 0x20, 0xd0,
0x80, 0x70, 0x60, 0x90, 0x40, 0xb0, 0xa0, 0x50 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x4e, 0x4e, 0x53, 0x53, 0x69, 0x69, 0x74, 0x74 },
{ 0x00, 0xdf, 0xbe, 0x61, 0x7c, 0xa3, 0xc2, 0x1d,
0xf8, 0x27, 0x46, 0x99, 0x84, 0x5b, 0x3a, 0xe5 },
{ 0x00, 0xa6, 0x51, 0xf7, 0xa2, 0x04, 0xf3, 0x55,
0x59, 0xff, 0x08, 0xae, 0xfb, 0x5d, 0xaa, 0x0c },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0,
0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0 },
{ 0x00, 0xa6, 0x51, 0xf7, 0xa2, 0x04, 0xf3, 0x55,
0x59, 0xff, 0x08, 0xae, 0xfb, 0x5d, 0xaa, 0x0c },
{ 0x00, 0x10, 0x20, 0x30, 0x40, 0x50, 0x60, 0x70,
0x80, 0x90, 0xa0, 0xb0, 0xc0, 0xd0, 0xe0, 0xf0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xe1, 0xc2, 0x23, 0x84, 0x65, 0x46, 0xa7,
0x08, 0xe9, 0xca, 0x2b, 0x8c, 0x6d, 0x4e, 0xaf },
{ 0x00, 0xa6, 0x51, 0xf7, 0xa2, 0x04, 0xf3, 0x55,
0x44, 0xe2, 0x15, 0xb3, 0xe6, 0x40, 0xb7, 0x11 },
{ 0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0,
0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xe2, 0xc4, 0x26, 0x88, 0x6a, 0x4c, 0xae,
0x10, 0xf2, 0xd4, 0x36, 0x98, 0x7a, 0x5c, 0xbe },
{ 0x00, 0xa6, 0x51, 0xf7, 0xa2, 0x04, 0xf3, 0x55,
0x44, 0xe2, 0x15, 0xb3, 0xe6, 0x40, 0xb7, 0x11 },
{ 0x00, 0x30, 0x60, 0x50, 0xc0, 0xf0, 0xa0, 0x90,
0x80, 0xb0, 0xe0, 0xd0, 0x40, 0x70, 0x20, 0x10 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xe3, 0xc6, 0x25, 0x8c, 0x6f, 0x4a, 0xa9,
0x18, 0xfb, 0xde, 0x3d, 0x94, 0x77, 0x52, 0xb1 },
{ 0x00, 0xa6, 0x51, 0xf7, 0xbf, 0x19, 0xee, 0x48,
0x63, 0xc5, 0x32, 0x94, 0xdc, 0x7a, 0x8d, 0x2b },
{ 0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0,
0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xe4, 0xc8, 0x2c, 0x90, 0x74, 0x58, 0xbc,
0x20, 0xc4, 0xe8, 0x0c, 0xb0, 0x54, 0x78, 0x9c },
{ 0x00, 0xa6, 0x51, 0xf7, 0xbf, 0x19, 0xee, 0x48,
0x63, 0xc5, 0x32, 0x94, 0xdc, 0x7a, 0x8d, 0x2b },
{ 0x00, 0x50, 0xa0, 0xf0, 0x40, 0x10, 0xe0, 0xb0,
0x80, 0xd0, 0x20, 0x70, 0xc0, 0x90, 0x60, 0x30 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xe5, 0xca, 0x2f, 0x94, 0x71, 0x5e, 0xbb,
0x28, 0xcd, 0xe2, 0x07, 0xbc, 0x59, 0x76, 0x93 },
{ 0x00, 0xa6, 0x51, 0xf7, 0xbf, 0x19, 0xee, 0x48,
0x7e, 0xd8, 0x2f, 0x89, 0xc1, 0x67, 0x90, 0x36 },
{ 0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20,
0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xe6, 0xcc, 0x2a, 0x98, 0x7e, 0x54, 0xb2,
0x30, 0xd6, 0xfc, 0x1a, 0xa8, 0x4e, 0x64, 0x82 },
{ 0x00, 0xa6, 0x51, 0xf7, 0xbf, 0x19, 0xee, 0x48,
0x7e, 0xd8, 0x2f, 0x89, 0xc1, 0x67, 0x90, 0x36 },
{ 0x00, 0x70, 0xe0, 0x90, 0xc0, 0xb0, 0x20, 0x50,
0x80, 0xf0, 0x60, 0x10, 0x40, 0x30, 0xa0, 0xd0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xe7, 0xce, 0x29, 0x9c, 0x7b, 0x52, 0xb5,
0x38, 0xdf, 0xf6, 0x11, 0xa4, 0x43, 0x6a, 0x8d },
{ 0x00, 0xa6, 0x4c, 0xea, 0x98, 0x3e, 0xd4, 0x72,
0x2d, 0x8b, 0x61, 0xc7, 0xb5, 0x13, 0xf9, 0x5f },
{ 0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80,
0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xe8, 0xd0, 0x38, 0xa0, 0x48, 0x70, 0x98,
0x40, 0xa8, 0x90, 0x78, 0xe0, 0x08, 0x30, 0xd8 },
{ 0x00, 0xa6, 0x4c, 0xea, 0x98, 0x3e, 0xd4, 0x72,
0x2d, 0x8b, 0x61, 0xc7, 0xb5, 0x13, 0xf9, 0x5f },
{ 0x00, 0x90, 0x20, 0xb0, 0x40, 0xd0, 0x60, 0xf0,
0x80, 0x10, 0xa0, 0x30, 0xc0, 0x50, 0xe0, 0x70 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xe9, 0xd2, 0x3b, 0xa4, 0x4d, 0x76, 0x9f,
0x48, 0xa1, 0x9a, 0x73, 0xec, 0x05, 0x3e, 0xd7 },
{ 0x00, 0xa6, 0x4c, 0xea, 0x98, 0x3e, 0xd4, 0x72,
0x30, 0x96, 0x7c, 0xda, 0xa8, 0x0e, 0xe4, 0x42 },
{ 0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60,
0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xea, 0xd4, 0x3e, 0xa8, 0x42, 0x7c, 0x96,
0x50, 0xba, 0x84, 0x6e, 0xf8, 0x12, 0x2c, 0xc6 },
{ 0x00, 0xa6, 0x4c, 0xea, 0x98, 0x3e, 0xd4, 0x72,
0x30, 0x96, 0x7c, 0xda, 0xa8, 0x0e, 0xe4, 0x42 },
{ 0x00, 0xb0, 0x60, 0xd0, 0xc0, 0x70, 0xa0, 0x10,
0x80, 0x30, 0xe0, 0x50, 0x40, 0xf0, 0x20, 0x90 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xeb, 0xd6, 0x3d, 0xac, 0x47, 0x7a, 0x91,
0x58, 0xb3, 0x8e, 0x65, 0xf4, 0x1f, 0x22, 0xc9 },
{ 0x00, 0xa6, 0x4c, 0xea, 0x85, 0x23, 0xc9, 0x6f,
0x17, 0xb1, 0x5b, 0xfd, 0x92, 0x34, 0xde, 0x78 },
{ 0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40,
0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xec, 0xd8, 0x34, 0xb0, 0x5c, 0x68, 0x84,
0x60, 0x8c, 0xb8, 0x54, 0xd0, 0x3c, 0x08, 0xe4 },
{ 0x00, 0xa6, 0x4c, 0xea, 0x85, 0x23, 0xc9, 0x6f,
0x17, 0xb1, 0x5b, 0xfd, 0x92, 0x34, 0xde, 0x78 },
{ 0x00, 0xd0, 0xa0, 0x70, 0x40, 0x90, 0xe0, 0x30,
0x80, 0x50, 0x20, 0xf0, 0xc0, 0x10, 0x60, 0xb0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xed, 0xda, 0x37, 0xb4, 0x59, 0x6e, 0x83,
0x68, 0x85, 0xb2, 0x5f, 0xdc, 0x31, 0x06, 0xeb },
{ 0x00, 0xa6, 0x4c, 0xea, 0x85, 0x23, 0xc9, 0x6f,
0x0a, 0xac, 0x46, 0xe0, 0x8f, 0x29, 0xc3, 0x65 },
{ 0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0,
0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xee, 0xdc, 0x32, 0xb8, 0x56, 0x64, 0x8a,
0x70, 0x9e, 0xac, 0x42, 0xc8, 0x26, 0x14, 0xfa },
{ 0x00, 0xa6, 0x4c, 0xea, 0x85, 0x23, 0xc9, 0x6f,
0x0a, 0xac, 0x46, 0xe0, 0x8f, 0x29, 0xc3, 0x65 },
{ 0x00, 0xf0, 0xe0, 0x10, 0xc0, 0x30, 0x20, 0xd0,
0x80, 0x70, 0x60, 0x90, 0x40, 0xb0, 0xa0, 0x50 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xef, 0xde, 0x31, 0xbc, 0x53, 0x62, 0x8d,
0x78, 0x97, 0xa6, 0x49, 0xc4, 0x2b, 0x1a, 0xf5 },
{ 0x00, 0xbb, 0x6b, 0xd0, 0xd6, 0x6d, 0xbd, 0x06,
0xb1, 0x0a, 0xda, 0x61, 0x67, 0xdc, 0x0c, 0xb7 },
{ 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,
0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xf0, 0xe0, 0x10, 0xc0, 0x30, 0x20, 0xd0,
0x80, 0x70, 0x60, 0x90, 0x40, 0xb0, 0xa0, 0x50 },
{ 0x00, 0xbb, 0x6b, 0xd0, 0xd6, 0x6d, 0xbd, 0x06,
0xb1, 0x0a, 0xda, 0x61, 0x67, 0xdc, 0x0c, 0xb7 },
{ 0x00, 0x10, 0x20, 0x30, 0x40, 0x50, 0x60, 0x70,
0x80, 0x90, 0xa0, 0xb0, 0xc0, 0xd0, 0xe0, 0xf0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xf1, 0xe2, 0x13, 0xc4, 0x35, 0x26, 0xd7,
0x88, 0x79, 0x6a, 0x9b, 0x4c, 0xbd, 0xae, 0x5f },
{ 0x00, 0xbb, 0x6b, 0xd0, 0xd6, 0x6d, 0xbd, 0x06,
0xac, 0x17, 0xc7, 0x7c, 0x7a, 0xc1, 0x11, 0xaa },
{ 0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0,
0x00, 0x20, 0x40, 0x60, 0x80, 0xa0, 0xc0, 0xe0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xf2, 0xe4, 0x16, 0xc8, 0x3a, 0x2c, 0xde,
0x90, 0x62, 0x74, 0x86, 0x58, 0xaa, 0xbc, 0x4e },
{ 0x00, 0xbb, 0x6b, 0xd0, 0xd6, 0x6d, 0xbd, 0x06,
0xac, 0x17, 0xc7, 0x7c, 0x7a, 0xc1, 0x11, 0xaa },
{ 0x00, 0x30, 0x60, 0x50, 0xc0, 0xf0, 0xa0, 0x90,
0x80, 0xb0, 0xe0, 0xd0, 0x40, 0x70, 0x20, 0x10 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xf3, 0xe6, 0x15, 0xcc, 0x3f, 0x2a, 0xd9,
0x98, 0x6b, 0x7e, 0x8d, 0x54, 0xa7, 0xb2, 0x41 },
{ 0x00, 0xbb, 0x6b, 0xd0, 0xcb, 0x70, 0xa0, 0x1b,
0x8b, 0x30, 0xe0, 0x5b, 0x40, 0xfb, 0x2b, 0x90 },
{ 0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0,
0x00, 0x40, 0x80, 0xc0, 0x00, 0x40, 0x80, 0xc0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xf4, 0xe8, 0x1c, 0xd0, 0x24, 0x38, 0xcc,
0xa0, 0x54, 0x48, 0xbc, 0x70, 0x84, 0x98, 0x6c },
{ 0x00, 0xbb, 0x6b, 0xd0, 0xcb, 0x70, 0xa0, 0x1b,
0x8b, 0x30, 0xe0, 0x5b, 0x40, 0xfb, 0x2b, 0x90 },
{ 0x00, 0x50, 0xa0, 0xf0, 0x40, 0x10, 0xe0, 0xb0,
0x80, 0xd0, 0x20, 0x70, 0xc0, 0x90, 0x60, 0x30 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xf5, 0xea, 0x1f, 0xd4, 0x21, 0x3e, 0xcb,
0xa8, 0x5d, 0x42, 0xb7, 0x7c, 0x89, 0x96, 0x63 },
{ 0x00, 0xbb, 0x6b, 0xd0, 0xcb, 0x70, 0xa0, 0x1b,
0x96, 0x2d, 0xfd, 0x46, 0x5d, 0xe6, 0x36, 0x8d },
{ 0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20,
0x00, 0x60, 0xc0, 0xa0, 0x80, 0xe0, 0x40, 0x20 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xf6, 0xec, 0x1a, 0xd8, 0x2e, 0x34, 0xc2,
0xb0, 0x46, 0x5c, 0xaa, 0x68, 0x9e, 0x84, 0x72 },
{ 0x00, 0xbb, 0x6b, 0xd0, 0xcb, 0x70, 0xa0, 0x1b,
0x96, 0x2d, 0xfd, 0x46, 0x5d, 0xe6, 0x36, 0x8d },
{ 0x00, 0x70, 0xe0, 0x90, 0xc0, 0xb0, 0x20, 0x50,
0x80, 0xf0, 0x60, 0x10, 0x40, 0x30, 0xa0, 0xd0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xf7, 0xee, 0x19, 0xdc, 0x2b, 0x32, 0xc5,
0xb8, 0x4f, 0x56, 0xa1, 0x64, 0x93, 0x8a, 0x7d },
{ 0x00, 0xbb, 0x76, 0xcd, 0xec, 0x57, 0x9a, 0x21,
0xc5, 0x7e, 0xb3, 0x08, 0x29, 0x92, 0x5f, 0xe4 },
{ 0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80,
0x00, 0x80, 0x00, 0x80, 0x00, 0x80, 0x00, 0x80 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xf8, 0xf0, 0x08, 0xe0, 0x18, 0x10, 0xe8,
0xc0, 0x38, 0x30, 0xc8, 0x20, 0xd8, 0xd0, 0x28 },
{ 0x00, 0xbb, 0x76, 0xcd, 0xec, 0x57, 0x9a, 0x21,
0xc5, 0x7e, 0xb3, 0x08, 0x29, 0x92, 0x5f, 0xe4 },
{ 0x00, 0x90, 0x20, 0xb0, 0x40, 0xd0, 0x60, 0xf0,
0x80, 0x10, 0xa0, 0x30, 0xc0, 0x50, 0xe0, 0x70 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xf9, 0xf2, 0x0b, 0xe4, 0x1d, 0x16, 0xef,
0xc8, 0x31, 0x3a, 0xc3, 0x2c, 0xd5, 0xde, 0x27 },
{ 0x00, 0xbb, 0x76, 0xcd, 0xec, 0x57, 0x9a, 0x21,
0xd8, 0x63, 0xae, 0x15, 0x34, 0x8f, 0x42, 0xf9 },
{ 0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60,
0x00, 0xa0, 0x40, 0xe0, 0x80, 0x20, 0xc0, 0x60 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xfa, 0xf4, 0x0e, 0xe8, 0x12, 0x1c, 0xe6,
0xd0, 0x2a, 0x24, 0xde, 0x38, 0xc2, 0xcc, 0x36 },
{ 0x00, 0xbb, 0x76, 0xcd, 0xec, 0x57, 0x9a, 0x21,
0xd8, 0x63, 0xae, 0x15, 0x34, 0x8f, 0x42, 0xf9 },
{ 0x00, 0xb0, 0x60, 0xd0, 0xc0, 0x70, 0xa0, 0x10,
0x80, 0x30, 0xe0, 0x50, 0x40, 0xf0, 0x20, 0x90 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xfb, 0xf6, 0x0d, 0xec, 0x17, 0x1a, 0xe1,
0xd8, 0x23, 0x2e, 0xd5, 0x34, 0xcf, 0xc2, 0x39 },
{ 0x00, 0xbb, 0x76, 0xcd, 0xf1, 0x4a, 0x87, 0x3c,
0xff, 0x44, 0x89, 0x32, 0x0e, 0xb5, 0x78, 0xc3 },
{ 0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40,
0x00, 0xc0, 0x80, 0x40, 0x00, 0xc0, 0x80, 0x40 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xfc, 0xf8, 0x04, 0xf0, 0x0c, 0x08, 0xf4,
0xe0, 0x1c, 0x18, 0xe4, 0x10, 0xec, 0xe8, 0x14 },
{ 0x00, 0xbb, 0x76, 0xcd, 0xf1, 0x4a, 0x87, 0x3c,
0xff, 0x44, 0x89, 0x32, 0x0e, 0xb5, 0x78, 0xc3 },
{ 0x00, 0xd0, 0xa0, 0x70, 0x40, 0x90, 0xe0, 0x30,
0x80, 0x50, 0x20, 0xf0, 0xc0, 0x10, 0x60, 0xb0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xfd, 0xfa, 0x07, 0xf4, 0x09, 0x0e, 0xf3,
0xe8, 0x15, 0x12, 0xef, 0x1c, 0xe1, 0xe6, 0x1b },
{ 0x00, 0xbb, 0x76, 0xcd, 0xf1, 0x4a, 0x87, 0x3c,
0xe2, 0x59, 0x94, 0x2f, 0x13, 0xa8, 0x65, 0xde },
{ 0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0,
0x00, 0xe0, 0xc0, 0x20, 0x80, 0x60, 0x40, 0xa0 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xfe, 0xfc, 0x02, 0xf8, 0x06, 0x04, 0xfa,
0xf0, 0x0e, 0x0c, 0xf2, 0x08, 0xf6, 0xf4, 0x0a },
{ 0x00, 0xbb, 0x76, 0xcd, 0xf1, 0x4a, 0x87, 0x3c,
0xe2, 0x59, 0x94, 0x2f, 0x13, 0xa8, 0x65, 0xde },
{ 0x00, 0xf0, 0xe0, 0x10, 0xc0, 0x30, 0x20, 0xd0,
0x80, 0x70, 0x60, 0x90, 0x40, 0xb0, 0xa0, 0x50 },
{ 0x00, 0x00, 0x1d, 0x1d, 0x27, 0x27, 0x3a, 0x3a,
0x53, 0x53, 0x4e, 0x4e, 0x74, 0x74, 0x69, 0x69 },
{ 0x00, 0xff, 0xfe, 0x01, 0xfc, 0x03, 0x02, 0xfd,
0xf8, 0x07, 0x06, 0xf9, 0x04, 0xfb, 0xfa, 0x05 }
};
codebase style improvements for OpenZFS 6459 port
2017-01-21 00:17:55 +03:00
/* END CSTYLED */
ABD raidz avx512f support Implement shift based multiplication for 512f. Higher IPC over lookup based methods yields up to 40% better performance on the current hardware. Results on Xeon Phi(TM) CPU 7210: implementation gen_p gen_pq gen_pqr rec_p rec_q rec_r rec_pq rec_pr rec_qr rec_pqr original 142232671 24411492 12948205 283053705 22348167 4215911 9171609 2265548 2378370 1648495 scalar 295711162 49851491 33253815 293198109 88179448 61866752 27941684 25764416 17384442 12138153 sse2 410055998 199642658 117973654 406240463 152688682 121092250 84968180 79291076 47473657 20779719 ssse3 411641595 199669571 117937647 406211024 137638508 117050346 81263322 76120405 46281559 32696722 avx2 616485806 311515332 188595628 605455115 260602390 230554476 148198817 138800254 92273356 62937819 avx512f 832191523 408509425 253599522 810094481 404325734 317590971 218235687 197204920 133101937 94001219 fastest avx512f avx512f avx512f avx512f avx512f avx512f avx512f avx512f avx512f avx512f Signed-off-by: Gvozden Neskovic <neskovic@gmail.com>
2016-11-20 08:01:31 +03:00
#endif /* defined(HAVE_SSSE3) || defined(HAVE_AVX2) || defined(HAVE_AVX512BW) */
#endif /* defined(__x86_64) */
Reference in New Issue Copy Permalink