frank-dittrich
commented
9 years ago Apparently, the i915 driver problems will be fixed with the 4.2 linux kernel.
Closed frank-dittrich closed 8 years ago
frank-dittrich
commented
9 years ago Apparently, the i915 driver problems will be fixed with the 4.2 linux kernel.
magnumripper
commented
9 years ago Here, I am not sure whether the
multiple unsequenced modifications to 'p'warning indicates a real problem:
I guess we should trust it until proven wrong.
The only statements that may trigger that
multiple unsequenced modifications to 'p'warning IMHO areoldoffice_kernel.cl:194: W[i >> 1] = (uint)*p++ | (*p++ << 16U); (...)Can any of the above be a problem?
I guess. I'm not sure what a C99 spec would say about it (I thought it was left-to-right when I wrote that). Please try adding temps that removes these ambiguities! Or maybe we should google for C specs first (or just ask Alexander Cherepanov).
BTW: why is the CPU format using length 4, but the GPU length 5?
That's for benchmarking against cRARk, which has/had it that way. I don't care about that anymore, but it took me a year or so to get on par with cRARk.
@magnum I wasn't sure whether I should create separate issues for every possible problem (failing format self tests etc.)
Perhaps not quite yet. First we should try to find general little fixes that solves many issues (like was done for OSX three years ago - the first results were worse than these)
frank-dittrich
commented
9 years ago Upgrading the Linux kernel from 4.0.8-300.fc22.x86_64 to a vanilla 4.2 rc3 helped to get rid of the OCL_IGNORE_SELF_TEST=1 which was needed to convince beignet not to disable the "non-working" device which didn't pass beignet's self test.
Just upgrading the linux kernel also helped improving the rar-opencl performance (from 364 c/s to 658 c/s).
I guess it is somehow related to the Beignet: warning - disable atomic in L3 feature. which was due to the failing self test with the older kernel.
$ ./john --test --format=rar-opencl --verbosity=5
initUnicode(UNICODE, ASCII/ASCII)
ASCII -> ASCII -> ASCII
Will run 4 OpenMP threads
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
Benchmarking: rar-opencl, RAR3 (length 5) [SHA1 OpenCL AES]... (4xOMP) Options used: -I ./kernels -cl-mad-enable -D__GPU__ -DDEVICE_INFO=34 -DDEV_VER_MAJOR=1 -DDEV_VER_MINOR=1 -D_OPENCL_COMPILER -DPLAINTEXT_LENGTH=22
Calculating best global worksize (GWS); max. 10s total for crypt_all()
Raw speed figures including buffer transfers:
key xfer: 82.733ms*, len xfer: 82.750ms*, init: 2.800us, loop: 16x35.801ms, final: 34.480us, key xfer: 69.059ms*, iv xfer: 111.680us*
gws: 256 317 c/s 83099648 rounds/s 807.510ms per crypt_all()!
key xfer: 132.540ms*, len xfer: 132.549ms*, init: 1.680us, loop: 16x66.152ms, final: 20.160us, key xfer: 132.499ms*, iv xfer: 109.280us*
gws: 512 351 c/s 92012544 rounds/s 1.456s per crypt_all()+
key xfer: 220.683ms*, len xfer: 220.719ms*, init: 1.920us, loop: 16x110.336ms, final: 34.800us, key xfer: 220.567ms*, iv xfer: 130.480us*
gws: 1024 421 c/s 110362624 rounds/s 2.427s per crypt_all()+
key xfer: 396.616ms*, len xfer: 396.585ms*, init: 2.560us, loop: 16x198.144ms, final: 45.040us, key xfer: 396.396ms*, iv xfer: 73.360us*
gws: 2048 469 c/s 122945536 rounds/s 4.360s per crypt_all()+
key xfer: 1.520us, len xfer: 1.360us, init: 3.680us, loop: 16x385.641ms, final: 70.240us, key xfer: 2us, iv xfer: 1.120us
gws: 4096 663 c/s 173801472 rounds/s 6.170s per crypt_all()+
key xfer: 1.525s*, len xfer: 1.525s*, init: 6.800us, loop: 16x762.559ms, final: 129.200us, key xfer: 1.532s*, iv xfer: 7.142ms*
gws: 8192 487 c/s 127664128 rounds/s 16.791s per crypt_all() - too slow
Local worksize (LWS) 64, global worksize (GWS) 4096
DONE
Raw: 658 c/s real, 204800 c/s virtualFor some reason, the encfs-opencl self test takes more than 7 minutes:
$ time ./john --test --format=encfs-opencl --verbosity=5
initUnicode(UNICODE, ASCII/ASCII)
ASCII -> ASCII -> ASCII
Will run 4 OpenMP threads
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
Benchmarking: encfs-opencl, EncFS [PBKDF2-SHA1 OpenCL 4x AES/Blowfish]... (4xOMP) Options used: -I ./kernels -cl-mad-enable -D__GPU__ -DDEVICE_INFO=34 -DDEV_VER_MAJOR=1 -DDEV_VER_MINOR=1 -D_OPENCL_COMPILER -DHASH_LOOPS=753 -DOUTLEN=48 -DPLAINTEXT_LENGTH=64 -DV_WIDTH=4
Calculating best global worksize (GWS); max. 10s total for crypt_all()
Raw speed figures including buffer transfers:
P xfer: 170.097ms*, init: 752.960us, loop: 241x84.120ms, final: 201.840us, res xfer: 168.742ms*
gws: 128 24 c/s 8710800 rounds/s 20.613s per crypt_all()!
P xfer: 174.726ms*, init: 412.080us, loop: 241x86.903ms, final: 229.680us, res xfer: 174.614ms*
gws: 256 48 c/s 17421600 rounds/s 21.294s per crypt_all()+
P xfer: 181.656ms*, init: 538.800us, loop: 241x90.215ms, final: 263.760us, res xfer: 181.576ms*
gws: 512 92 c/s 33391400 rounds/s 22.106s per crypt_all()+
P xfer: 308.815ms*, init: 1.320ms, loop: 241x152.363ms, final: 774us, res xfer: 308.409ms*
gws: 1024 109 c/s 39561550 rounds/s 37.339s per crypt_all() - too slow
Local worksize (LWS) 64, global worksize (GWS) 512
DONE
Speed for cost 1 (iteration count) of 181474 and 181317
Raw: 47.1 c/s real, 5535 c/s virtual
real 7m33.911s
user 0m2.681s
sys 0m4.626s
Sat Jul 25 18:52:44 CEST 2015With a 4.2-rc3 linux kernel, the number of failing opencl formats reduced from 16 to 6:
$ for f in `grep -v "^0 " opencl_errors-4.2.txt |cut -d" " -f 2-`; do time ./john --test --format=$f --verbosity=5; done
initUnicode(UNICODE, ASCII/ASCII)
ASCII -> ASCII -> ASCII
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
Options used: -I ./kernels -cl-mad-enable -D__GPU__ -DDEVICE_INFO=34 -DDEV_VER_MAJOR=1 -DDEV_VER_MINOR=1 -D_OPENCL_COMPILER
ASSERTION FAILED: 0
at file /home/fd/git/beignet/backend/src/backend/gen_context.cpp, function virtual void gbe::GenContext::emitUnaryWithTempInstruction(const gbe::SelectionInstruction&), line 438
Trace/breakpoint trap (core dumped)
real 0m2.114s
user 0m1.414s
sys 0m0.054s
initUnicode(UNICODE, ASCII/ASCII)
ASCII -> ASCII -> ASCII
Will run 4 OpenMP threads
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
Benchmarking: office2013-opencl, MS Office 2013 (100,000 iterations) [SHA512 OpenCL 2x AES]... (4xOMP) Options used: -I ./kernels -cl-mad-enable -D__GPU__ -DDEVICE_INFO=34 -DDEV_VER_MAJOR=1 -DDEV_VER_MINOR=1 -D_OPENCL_COMPILER -DHASH_LOOPS=100 -DUNICODE_LENGTH=96 -DV_WIDTH=2
ASSERTION FAILED: 0
at file /home/fd/git/beignet/backend/src/backend/gen_context.cpp, function virtual void gbe::GenContext::emitUnaryWithTempInstruction(const gbe::SelectionInstruction&), line 438
Trace/breakpoint trap (core dumped)
real 0m3.321s
user 0m3.267s
sys 0m0.041s
initUnicode(UNICODE, ASCII/ASCII)
ASCII -> ASCII -> ASCII
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
Benchmarking: PBKDF2-HMAC-SHA512-opencl, GRUB2 / OS X 10.8+, rounds=10000 [PBKDF2-SHA512 OpenCL]... Options used: -I ./kernels -cl-mad-enable -D__GPU__ -DDEVICE_INFO=34 -DDEV_VER_MAJOR=1 -DDEV_VER_MINOR=1 -D_OPENCL_COMPILER -DHASH_LOOPS=250 -DPLAINTEXT_LENGTH=110 -DMAX_SALT_SIZE=107
ASSERTION FAILED: 0
at file /home/fd/git/beignet/backend/src/backend/gen_context.cpp, function virtual void gbe::GenContext::emitUnaryWithTempInstruction(const gbe::SelectionInstruction&), line 438
Trace/breakpoint trap (core dumped)
real 0m3.161s
user 0m3.103s
sys 0m0.048s
initUnicode(UNICODE, ASCII/ASCII)
ASCII -> ASCII -> ASCII
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
Options used: -I ./kernels -cl-mad-enable -D__GPU__ -DDEVICE_INFO=34 -DDEV_VER_MAJOR=1 -DDEV_VER_MINOR=1 -D_OPENCL_COMPILER
ASSERTION FAILED: 0
at file /home/fd/git/beignet/backend/src/backend/gen_context.cpp, function virtual void gbe::GenContext::emitUnaryWithTempInstruction(const gbe::SelectionInstruction&), line 438
Trace/breakpoint trap (core dumped)
real 0m0.236s
user 0m0.217s
sys 0m0.008s
initUnicode(UNICODE, ASCII/ASCII)
ASCII -> ASCII -> ASCII
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
Benchmarking: wpapsk-opencl, WPA/WPA2 PSK [PBKDF2-SHA1 OpenCL 4x]... Options used: -I ./kernels -cl-mad-enable -D__GPU__ -DDEVICE_INFO=34 -DDEV_VER_MAJOR=1 -DDEV_VER_MINOR=1 -D_OPENCL_COMPILER -DHASH_LOOPS=105 -DITERATIONS=4095 -DPLAINTEXT_LENGTH=63 -DV_WIDTH=4
ASSERTION FAILED: !(ctx->getErrCode() == OUT_OF_RANGE_IF_ENDIF && ctx->getIFENDIFFix())
at file /home/fd/git/beignet/backend/src/backend/gen_program.cpp, function virtual gbe::Kernel* gbe::GenProgram::compileKernel(const gbe::ir::Unit&, const string&, bool), line 196
Trace/breakpoint trap (core dumped)
real 0m15.782s
user 0m13.106s
sys 0m0.146s
initUnicode(UNICODE, ASCII/ASCII)
ASCII -> ASCII -> ASCII
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
Options used: -I ./kernels -cl-mad-enable -D__GPU__ -DDEVICE_INFO=34 -DDEV_VER_MAJOR=1 -DDEV_VER_MINOR=1 -D_OPENCL_COMPILER
ASSERTION FAILED: 0
at file /home/fd/git/beignet/backend/src/backend/gen_context.cpp, function virtual void gbe::GenContext::emitUnaryWithTempInstruction(const gbe::SelectionInstruction&), line 438
Trace/breakpoint trap (core dumped)
real 0m0.243s
user 0m0.225s
sys 0m0.008sTo get an idea why the beignet assertions failed, I added a few fprintf statements:
(master)beignet $ git diff
diff --git a/backend/src/backend/gen_context.cpp b/backend/src/backend/gen_context.cpp
index e16b0a9..183d803 100644
--- a/backend/src/backend/gen_context.cpp
+++ b/backend/src/backend/gen_context.cpp
@@ -435,7 +435,7 @@ namespace gbe
p->MOV(dst, tmp);
} else {
- GBE_ASSERT(0);
+ fprintf(stderr,"src.type %d\n", src.type); GBE_ASSERT(0);
}
}
}
diff --git a/backend/src/backend/gen_program.cpp b/backend/src/backend/gen_program.cpp
index c761a2f..27b979c 100644
--- a/backend/src/backend/gen_program.cpp
+++ b/backend/src/backend/gen_program.cpp
@@ -192,8 +192,10 @@ namespace gbe {
if ( ctx->getErrCode() == OUT_OF_RANGE_IF_ENDIF && !ctx->getIFENDIFFix() ) {
ctx->setIFENDIFFix(true);
codeGen--;
- } else
+ } else {
+ fprintf(stderr, "ctx->getErrCode() %d, ctx->getIFENDIFFix() %d, OUT_OF_RANGE_IF_ENDIF %d\n", ctx->getErrCode(), ctx->getIFENDIFFix(), OUT_OF_RANGE_IF_ENDIF);
GBE_ASSERT(!(ctx->getErrCode() == OUT_OF_RANGE_IF_ENDIF && ctx->getIFENDIFFix()));
+ }
}
GBE_ASSERTM(kernel != NULL, "Fail to compile kernel, may need to increase reserved registers for spilling.");and got:
(bleeding-jumbo)run $ for f in `grep -v "^0 " opencl_errors-4.2.txt |cut -d" " -f 2-`; do ./john --test --format=$f --devices=0; done
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
ctx->getErrCode() 1, ctx->getIFENDIFFix() 0, OUT_OF_RANGE_IF_ENDIF 4
ctx->getErrCode() 1, ctx->getIFENDIFFix() 0, OUT_OF_RANGE_IF_ENDIF 4
ctx->getErrCode() 1, ctx->getIFENDIFFix() 0, OUT_OF_RANGE_IF_ENDIF 4
ctx->getErrCode() 1, ctx->getIFENDIFFix() 0, OUT_OF_RANGE_IF_ENDIF 4
src.type 8
ASSERTION FAILED: 0
at file /home/fd/git/beignet/backend/src/backend/gen_context.cpp, function virtual void gbe::GenContext::emitUnaryWithTempInstruction(const gbe::SelectionInstruction&), line 438
Trace/breakpoint trap (core dumped)
Will run 4 OpenMP threads
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
Benchmarking: office2013-opencl, MS Office 2013 (100,000 iterations) [SHA512 OpenCL 2x AES]... (4xOMP) ctx->getErrCode() 1, ctx->getIFENDIFFix() 0, OUT_OF_RANGE_IF_ENDIF 4
ctx->getErrCode() 1, ctx->getIFENDIFFix() 0, OUT_OF_RANGE_IF_ENDIF 4
src.type 8
ASSERTION FAILED: 0
at file /home/fd/git/beignet/backend/src/backend/gen_context.cpp, function virtual void gbe::GenContext::emitUnaryWithTempInstruction(const gbe::SelectionInstruction&), line 438
Trace/breakpoint trap (core dumped)
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
Benchmarking: PBKDF2-HMAC-SHA512-opencl, GRUB2 / OS X 10.8+, rounds=10000 [PBKDF2-SHA512 OpenCL]... ctx->getErrCode() 1, ctx->getIFENDIFFix() 0, OUT_OF_RANGE_IF_ENDIF 4
ctx->getErrCode() 1, ctx->getIFENDIFFix() 0, OUT_OF_RANGE_IF_ENDIF 4
src.type 8
ASSERTION FAILED: 0
at file /home/fd/git/beignet/backend/src/backend/gen_context.cpp, function virtual void gbe::GenContext::emitUnaryWithTempInstruction(const gbe::SelectionInstruction&), line 438
Trace/breakpoint trap (core dumped)
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
ctx->getErrCode() 1, ctx->getIFENDIFFix() 0, OUT_OF_RANGE_IF_ENDIF 4
src.type 8
ASSERTION FAILED: 0
at file /home/fd/git/beignet/backend/src/backend/gen_context.cpp, function virtual void gbe::GenContext::emitUnaryWithTempInstruction(const gbe::SelectionInstruction&), line 438
Trace/breakpoint trap (core dumped)
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
Benchmarking: wpapsk-opencl, WPA/WPA2 PSK [PBKDF2-SHA1 OpenCL 4x]... ctx->getErrCode() 1, ctx->getIFENDIFFix() 0, OUT_OF_RANGE_IF_ENDIF 4
ctx->getErrCode() 1, ctx->getIFENDIFFix() 0, OUT_OF_RANGE_IF_ENDIF 4
ctx->getErrCode() 4, ctx->getIFENDIFFix() 1, OUT_OF_RANGE_IF_ENDIF 4
ASSERTION FAILED: !(ctx->getErrCode() == OUT_OF_RANGE_IF_ENDIF && ctx->getIFENDIFFix())
at file /home/fd/git/beignet/backend/src/backend/gen_program.cpp, function virtual gbe::Kernel* gbe::GenProgram::compileKernel(const gbe::ir::Unit&, const string&, bool), line 197
Trace/breakpoint trap (core dumped)
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
ctx->getErrCode() 1, ctx->getIFENDIFFix() 0, OUT_OF_RANGE_IF_ENDIF 4
src.type 8
ASSERTION FAILED: 0
at file /home/fd/git/beignet/backend/src/backend/gen_context.cpp, function virtual void gbe::GenContext::emitUnaryWithTempInstruction(const gbe::SelectionInstruction&), line 438
Trace/breakpoint trap (core dumped)Looks like beignet is to blame here.
frank-dittrich
commented
9 years ago BTW: not only is the GPU faster than the CPU for rar (660 c/s vs. 400 c/s), I can even run them in parallel, and after GPU auto-tuning is finished, both run at about the same speed they would run when started separately, i.e., I get a combined speed of 1060 c/s for my Intel(R) Core(TM) i5-4570 CPU @ 3.20GHz
frank-dittrich
commented
9 years ago These are the rar-opencl format tests of ./jtrts.pl:
form=rar-opencl guesses: 130 0:00:02:21 DONE [PASSED]
.pot CHK:rar-opencl guesses: 130 0:00:01:36 DONE [PASSED] (130 val-pwd)
form=rar-opencl guesses: 216 -show= 216 0:00:05:24 DONE : Expected count(s) (297) [!!!FAILED1!!!]
.pot CHK:rar-opencl guesses: 216 0:00:02:40 DONE [PASSED] (216 val-pwd)Several other opencl formats pass self test (will report in detail later, probably next week).
frank-dittrich
commented
9 years ago This is the jtrts result:
$ (for f in `../run/john --list=formats --format=opencl|tr "," " "`; do echo -n "$f "; date; ./jtrts.pl -q $f -passthru="--skip-self-test"; done; date ) 2> /dev/null |grep -v "Failed to release test userptr object"
Sun Jul 26 00:38:36 CEST 2015
sha1crypt-opencl Sun Jul 26 00:38:36 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
All tests passed without error. Performed 1 tests. Time used was 50 seconds
oldoffice-opencl Sun Jul 26 00:39:26 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
NO tests were performed. Time used was 8 seconds
PBKDF2-HMAC-MD4-opencl Sun Jul 26 00:39:34 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
NO tests were performed. Time used was 6 seconds
PBKDF2-HMAC-MD5-opencl Sun Jul 26 00:39:40 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
NO tests were performed. Time used was 7 seconds
PBKDF2-HMAC-SHA1-opencl Sun Jul 26 00:39:47 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
All tests passed without error. Performed 3 tests. Time used was 733 seconds
rar-opencl Sun Jul 26 00:52:00 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
form=rar-opencl guesses: 216 -show= 216 0:00:05:24 DONE : Expected count(s) (297) [!!!FAILED1!!!]
Some tests had Errors. Performed 2 tests. 1 errors
Time used was 734 seconds
RAR5-opencl Sun Jul 26 01:04:14 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
NO tests were performed. Time used was 8 seconds
lotus5-opencl Sun Jul 26 01:04:22 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
All tests passed without error. Performed 1 tests. Time used was 8 seconds
agilekeychain-opencl Sun Jul 26 01:04:30 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
NO tests were performed. Time used was 6 seconds
bcrypt-opencl Sun Jul 26 01:04:36 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
form=Bcrypt-opencl guesses: 300 -show= 300 0:00:00:02 DONE : Expected count(s) (1500) [!!!FAILED1!!!]
form=Bcrypt_broken-opencl guesses: 300 -show= 300 0:00:00:02 DONE : Expected count(s) (1500) [!!!FAILED1!!!]
Some tests had Errors. Performed 2 tests. 2 errors
Time used was 26 seconds
blockchain-opencl Sun Jul 26 01:05:02 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
NO tests were performed. Time used was 8 seconds
md5crypt-opencl Sun Jul 26 01:05:10 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
All tests passed without error. Performed 1 tests. Time used was 40 seconds
sha256crypt-opencl Sun Jul 26 01:05:50 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
All tests passed without error. Performed 3 tests. Time used was 253 seconds
sha512crypt-opencl Sun Jul 26 01:10:03 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
form=sha512crypt-opencl guesses: 0 -show= 0 unk unk : Expected count(s) (1389)(1500) [!!!FAILED2!!! exited, return code 133]
.pot CHK:sha512crypt-opencl guesses: 0 -show= unk unk : Expected count(s) (1389)(1500) [!!!FAILED4!!!] (0 val-pwd 1 inval-pwd)
form=sha512crypt-opencl-new guesses: 0 -show= 0 unk unk : Expected count(s) (760) [!!!FAILED2!!! exited, return code 133]
.pot CHK:sha512crypt-opencl-new guesses: 0 -show= unk unk : Expected count(s) (760) [!!!FAILED4!!!] (0 val-pwd 1 inval-pwd)
form=sha512crypt-opencl-xsalt guesses: 0 -show= 0 unk unk : Expected count(s) (1168) [!!!FAILED2!!! exited, return code 133]
.pot CHK:sha512crypt-opencl-xsalt guesses: 0 -show= unk unk : Expected count(s) (1168) [!!!FAILED4!!!] (0 val-pwd 1 inval-pwd)
Some tests had Errors. Performed 3 tests. 3 errors 3 errors reprocessing the .POT files 3 runs had non-clean exit
Time used was 21 seconds
descrypt-opencl Sun Jul 26 01:10:24 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
All tests passed without error. Performed 1 tests. Time used was 35 seconds
dmg-opencl Sun Jul 26 01:10:59 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
NO tests were performed. Time used was 8 seconds
encfs-opencl Sun Jul 26 01:11:07 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
NO tests were performed. Time used was 6 seconds
gpg-opencl Sun Jul 26 01:11:13 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
NO tests were performed. Time used was 7 seconds
keychain-opencl Sun Jul 26 01:11:20 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
All tests passed without error. Performed 1 tests. Time used was 128 seconds
keyring-opencl Sun Jul 26 01:13:28 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
NO tests were performed. Time used was 9 seconds
krb5pa-md5-opencl Sun Jul 26 01:13:37 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
form=krb5pa-md5-opencl guesses: 0 -show= 0 0:00:00:21 DONE : Expected count(s) (1500) [!!!FAILED1!!!]
.pot CHK:krb5pa-md5-opencl guesses: 0 -show= 0 unk unk : Expected count(s) (1500) [!!!FAILED4!!!] (0 val-pwd 1 inval-pwd)
Some tests had Errors. Performed 1 tests. 1 errors 1 errors reprocessing the .POT files
Time used was 28 seconds
krb5pa-sha1-opencl Sun Jul 26 01:14:05 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
NO tests were performed. Time used was 9 seconds
LM-opencl Sun Jul 26 01:14:14 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
NO tests were performed. Time used was 6 seconds
mscash-opencl Sun Jul 26 01:14:20 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
form=mscash-opencl guesses: 0 -show= 0 unk unk : Expected count(s) (1500) [!!!FAILED2!!! exited, return code 1]
.pot CHK:mscash-opencl guesses: 0 -show= 0 unk unk : Expected count(s) (1500) [!!!FAILED4!!!] (0 val-pwd 1 inval-pwd)
form=mscash-opencl-utf8 guesses: 0 -show= 0 unk unk : Expected count(s) (1500) [!!!FAILED2!!! exited, return code 1]
.pot CHK:mscash-opencl-utf8 guesses: 0 -show= 0 unk unk : Expected count(s) (1500) [!!!FAILED4!!!] (0 val-pwd 1 inval-pwd)
Some tests had Errors. Performed 2 tests. 2 errors 2 errors reprocessing the .POT files 2 runs had non-clean exit
Time used was 7 seconds
mscash2-opencl Sun Jul 26 01:14:27 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
form=mscash2-opencl guesses: 1050 -show=1050 0:00:00:05 DONE : Expected count(s) (1410)(1500) [!!!FAILED1!!!]
Some tests had Errors. Performed 2 tests. 1 errors
Time used was 34 seconds
mysql-sha1-opencl Sun Jul 26 01:15:01 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
All tests passed without error. Performed 1 tests. Time used was 10 seconds
ssha-opencl Sun Jul 26 01:15:11 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
All tests passed without error. Performed 1 tests. Time used was 11 seconds
nt-opencl Sun Jul 26 01:15:22 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
form=nt-opencl guesses: 0 -show= 0 unk unk : Expected count(s) (1478) [!!!FAILED2!!! exited, return code 1]
.pot CHK:nt-opencl guesses: 0 -show= 0 unk unk : Expected count(s) (1478) [!!!FAILED4!!!] (0 val-pwd 1 inval-pwd)
Some tests had Errors. Performed 1 tests. 1 errors 1 errors reprocessing the .POT files 1 runs had non-clean exit
Time used was 9 seconds
ntlmv2-opencl Sun Jul 26 01:15:31 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
form=ntlmv2-opencl guesses: 0 -show= 0 0:00:00:06 DONE : Expected count(s) (1452)(1500) [!!!FAILED1!!!]
.pot CHK:ntlmv2-opencl guesses: 0 -show= 0 unk unk : Expected count(s) (1452)(1500) [!!!FAILED4!!!] (0 val-pwd 1 inval-pwd)
form=netntlmv2-opencl-utf8 guesses: 0 -show= 0 0:00:00:02 DONE : Expected count(s) (1442)(1500) [!!!FAILED1!!!]
.pot CHK:netntlmv2-opencl-utf8 guesses: 0 -show= 0 unk unk : Expected count(s) (1442)(1500) [!!!FAILED4!!!] (0 val-pwd 1 inval-pwd)
.pot CHK:netntlmv2-opencl-koi8r guesses: 0 -show= 0 0:00:00:01 DONE : Expected count(s) (1429)(1500) [!!!FAILED4!!!] (0 val-pwd 0 inval-pwd)
Some tests had Errors. Performed 3 tests. 2 errors 3 errors reprocessing the .POT files
Time used was 25 seconds
o5logon-opencl Sun Jul 26 01:15:56 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
.pot CHK:o5logon-opencl guesses: 1499 -show=1499 0:00:00:01 DONE : Expected count(s) (1500) [!!!FAILED4!!!] (1498 val-pwd 1 inval-pwd)
Some tests had Errors. Performed 1 tests. 1 errors reprocessing the .POT files
Time used was 13 seconds
ODF-opencl Sun Jul 26 01:16:09 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
All tests passed without error. Performed 2 tests. Time used was 18 seconds
ODF-AES-opencl Sun Jul 26 01:16:27 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
All tests passed without error. Performed 1 tests. Time used was 15 seconds
office2007-opencl Sun Jul 26 01:16:42 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
form=office_2007-opencl guesses: 1190 -show=1190 0:00:00:12 DONE : Expected count(s) (1500) [!!!FAILED1!!!]
Some tests had Errors. Performed 1 tests. 1 errors
Time used was 35 seconds
office2010-opencl Sun Jul 26 01:17:17 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
form=office_2010-opencl guesses: 32 -show= 32 0:00:01:55 DONE : Expected count(s) (48) [!!!FAILED1!!!]
.pot CHK:office_2010-opencl guesses: 32 -show= 32 0:00:01:17 DONE : Expected count(s) (48) [!!!FAILED4!!!] (30 val-pwd 2 inval-pwd)
Some tests had Errors. Performed 1 tests. 1 errors 1 errors reprocessing the .POT files
Time used was 205 seconds
office2013-opencl Sun Jul 26 01:20:42 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
form=office_2013-opencl guesses: 0 -show= 0 unk unk : Expected count(s) (48) [!!!FAILED2!!! exited, return code 133]
.pot CHK:office_2013-opencl guesses: 0 -show= 0 unk unk : Expected count(s) (48) [!!!FAILED4!!!] (0 val-pwd 1 inval-pwd)
Some tests had Errors. Performed 1 tests. 1 errors 1 errors reprocessing the .POT files 1 runs had non-clean exit
Time used was 12 seconds
PBKDF2-HMAC-SHA256-opencl Sun Jul 26 01:20:54 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
All tests passed without error. Performed 2 tests. Time used was 64 seconds
PBKDF2-HMAC-SHA512-opencl Sun Jul 26 01:21:58 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
form=pbkdf2-hmac-sha512-opencl guesses: 0 -show= 0 unk unk : Expected count(s) (1500) [!!!FAILED2!!! exited, return code 133]
.pot CHK:pbkdf2-hmac-sha512-openc guesses: 0 -show= 0 unk unk : Expected count(s) (1500) [!!!FAILED4!!!] (0 val-pwd 1 inval-pwd)
Some tests had Errors. Performed 1 tests. 1 errors 1 errors reprocessing the .POT files 1 runs had non-clean exit
Time used was 12 seconds
phpass-opencl Sun Jul 26 01:22:10 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
All tests passed without error. Performed 1 tests. Time used was 194 seconds
pwsafe-opencl Sun Jul 26 01:25:24 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
All tests passed without error. Performed 1 tests. Time used was 69 seconds
RAKP-opencl Sun Jul 26 01:26:33 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
All tests passed without error. Performed 1 tests. Time used was 55 seconds
Raw-MD4-opencl Sun Jul 26 01:27:28 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
form=raw-md4-opencl guesses: 0 -show= 172 unk unk : Expected count(s) (1500) [!!!FAILED2!!! exited, return code 1]
.pot CHK:raw-md4-opencl guesses: 0 unk unk [pass, but exited, return code 139]
Some tests had Errors. Performed 1 tests. 1 errors 2 runs had non-clean exit
Time used was 10 seconds
Raw-MD5-opencl Sun Jul 26 01:27:38 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
form=raw-md5-opencl guesses: 0 -show= 172 unk unk : Expected count(s) (1500) [!!!FAILED2!!! exited, return code 1]
.pot CHK:raw-md5-opencl guesses: 0 -show= 64 unk unk : Expected count(s) (1500) [!!!FAILED4!!!] (64 val-pwd 1 inval-pwd)
Some tests had Errors. Performed 1 tests. 1 errors 1 errors reprocessing the .POT files 1 runs had non-clean exit
Time used was 9 seconds
Raw-SHA1-opencl Sun Jul 26 01:27:47 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
form=raw-sha1-opencl guesses: 0 -show= 172 unk unk : Expected count(s) (1500) [!!!FAILED2!!! exited, return code 1]
.pot CHK:raw-sha1-opencl guesses: 0 -show= 64 unk unk : Expected count(s) (1500) [!!!FAILED4!!!] (64 val-pwd 1 inval-pwd)
Some tests had Errors. Performed 1 tests. 1 errors 1 errors reprocessing the .POT files 1 runs had non-clean exit
Time used was 8 seconds
Raw-SHA256-opencl Sun Jul 26 01:27:55 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
All tests passed without error. Performed 1 tests. Time used was 11 seconds
Raw-SHA512-opencl Sun Jul 26 01:28:06 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
form=raw-sha512-opencl guesses: 0 -show= 0 unk unk : Expected count(s) (1500) [!!!FAILED2!!! exited, return code 133]
.pot CHK:raw-sha512-opencl guesses: 0 -show= unk unk : Expected count(s) (1500) [!!!FAILED4!!!] (0 val-pwd 1 inval-pwd)
Some tests had Errors. Performed 1 tests. 1 errors 1 errors reprocessing the .POT files 1 runs had non-clean exit
Time used was 7 seconds
7z-opencl Sun Jul 26 01:28:13 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
NO tests were performed. Time used was 6 seconds
strip-opencl Sun Jul 26 01:28:19 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
All tests passed without error. Performed 1 tests. Time used was 476 seconds
sxc-opencl Sun Jul 26 01:36:15 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
NO tests were performed. Time used was 8 seconds
wpapsk-opencl Sun Jul 26 01:36:23 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
form=wpapsk-opencl guesses: 0 -show= 0 unk unk : Expected count(s) (500) [!!!FAILED2!!! exited, return code 133]
.pot CHK:wpapsk-opencl guesses: 0 -show= 0 unk unk : Expected count(s) (500) [!!!FAILED4!!!] (0 val-pwd 1 inval-pwd)
Some tests had Errors. Performed 1 tests. 1 errors 1 errors reprocessing the .POT files 1 runs had non-clean exit
Time used was 23 seconds
XSHA512-opencl Sun Jul 26 01:36:46 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
form=xsha512-opencl guesses: 0 -show= 0 unk unk : Expected count(s) (1500) [!!!FAILED2!!! exited, return code 133]
.pot CHK:xsha512-opencl guesses: 0 -show= unk unk : Expected count(s) (1500) [!!!FAILED4!!!] (0 val-pwd 1 inval-pwd)
Some tests had Errors. Performed 1 tests. 1 errors 1 errors reprocessing the .POT files 1 runs had non-clean exit
Time used was 7 seconds
zip-opencl Sun Jul 26 01:36:53 CEST 2015
-------------------------------------------------------------------------------
- JtR-TestSuite (jtrts). Version 1.13, Dec 21, 2014. By, Jim Fougeron & others
- Testing: John the Ripper 1.8.0.6-jumbo-1-531-g1b0031f OMP [linux-gnu 64-bit AVX2-ac]
--------------------------------------------------------------------------------
All tests passed without error. Performed 4 tests. Time used was 720 seconds
Sun Jul 26 01:48:53 CEST 2015And these are the --test results:
$ (for f in `../run/john --list=formats --format=opencl|tr "," " "`; do echo -n "$f "; date; ./john --test --format=$f | tee test-1-$f.txt; done ; date ) 2> /dev/null |grep -v "Failed to release test user"
sha1crypt-opencl Sun Jul 26 00:20:48 CEST 2015
Benchmarking: sha1crypt-opencl, (NetBSD) [PBKDF1-SHA1 OpenCL 4x]... DONE
Speed for cost 1 (iteration count) of 64000 and 40000
Raw: 469 c/s real, 51200 c/s virtual
oldoffice-opencl Sun Jul 26 00:21:28 CEST 2015
Benchmarking: oldoffice-opencl, MS Office <= 2003 [MD5/SHA1 RC4 OpenCL]... DONE
Speed for cost 1 (hash type) of 1 and 0
Many salts: 1542K c/s real, 45875K c/s virtual
Only one salt: 1497K c/s real, 31457K c/s virtual
PBKDF2-HMAC-MD4-opencl Sun Jul 26 00:21:33 CEST 2015
Benchmarking: PBKDF2-HMAC-MD4-opencl [PBKDF2-MD4 OpenCL 4x]... DONE
Speed for cost 1 (iterations) of 1000
Raw: 113975 c/s real, 13107K c/s virtual
PBKDF2-HMAC-MD5-opencl Sun Jul 26 00:21:51 CEST 2015
Benchmarking: PBKDF2-HMAC-MD5-opencl [PBKDF2-MD5 OpenCL 4x]... DONE
Speed for cost 1 (iterations) of 1000
Raw: 74898 c/s real, 26214K c/s virtual
PBKDF2-HMAC-SHA1-opencl Sun Jul 26 00:22:09 CEST 2015
Benchmarking: PBKDF2-HMAC-SHA1-opencl [PBKDF2-SHA1 OpenCL 4x]... DONE
Speed for cost 1 (iterations) of 1000
Raw: 29520 c/s real, 13107K c/s virtual
rar-opencl Sun Jul 26 00:22:37 CEST 2015
Benchmarking: rar-opencl, RAR3 (length 5) [SHA1 OpenCL AES]... (4xOMP) DONE
Raw: 661 c/s real, 136533 c/s virtual
RAR5-opencl Sun Jul 26 00:23:07 CEST 2015
Benchmarking: RAR5-opencl [PBKDF2-SHA256 OpenCL]... DONE
Speed for cost 1 (iteration count) of 32768
Raw: 408 c/s real, 136533 c/s virtual
lotus5-opencl Sun Jul 26 00:23:36 CEST 2015
Benchmarking: lotus5-opencl, Lotus Notes/Domino 5 [OpenCL]... DONE
Raw: 865569 c/s real, 11468K c/s virtual
agilekeychain-opencl Sun Jul 26 00:23:40 CEST 2015
Benchmarking: agilekeychain-opencl, 1Password Agile Keychain [PBKDF2-SHA1 OpenCL AES]... (4xOMP) DONE
Speed for cost 1 (iteration count) of 1000
Raw: 31207 c/s real, 409600 c/s virtual
bcrypt-opencl Sun Jul 26 00:23:46 CEST 2015
Benchmarking: bcrypt-opencl ("$2a$05", 32 iterations) [Blowfish OpenCL]... DONE
Speed for cost 1 (iteration count) of 32
Raw: 322 c/s real, 12800 c/s virtual
blockchain-opencl Sun Jul 26 00:23:51 CEST 2015
Benchmarking: blockchain-opencl, blockchain My Wallet [PBKDF2-SHA1 OpenCL AES]... (4xOMP) DONE
Raw: 674635 c/s real, 395475 c/s virtual
md5crypt-opencl Sun Jul 26 00:23:56 CEST 2015
Benchmarking: md5crypt-opencl, crypt(3) $1$ [MD5 OpenCL]... DONE
Raw: 109226 c/s real, 11468K c/s virtual
sha256crypt-opencl Sun Jul 26 00:23:59 CEST 2015
Benchmarking: sha256crypt-opencl, crypt(3) $5$ (rounds=5000) [SHA256 OpenCL]... DONE
Speed for cost 1 (iteration count) of 5000
Raw: 687 c/s real, 34133 c/s virtual
sha512crypt-opencl Sun Jul 26 00:24:14 CEST 2015
descrypt-opencl Sun Jul 26 00:24:16 CEST 2015
Benchmarking: descrypt-opencl, traditional crypt(3) [DES OpenCL]... DONE
Many salts: 1055K c/s real, 157286K c/s virtual
Only one salt: 1038K c/s real, 104857K c/s virtual
dmg-opencl Sun Jul 26 00:24:21 CEST 2015
Benchmarking: dmg-opencl, Apple DMG [PBKDF2-SHA1 OpenCL 3DES/AES]... (4xOMP) DONE
Speed for cost 1 (iteration count) of 1000
Raw: 11592 c/s real, 12934 c/s virtual
encfs-opencl Sun Jul 26 00:24:27 CEST 2015
Benchmarking: encfs-opencl, EncFS [PBKDF2-SHA1 OpenCL 4x AES/Blowfish]... (4xOMP) DONE
Speed for cost 1 (iteration count) of 181474 and 181317
Raw: 47.2 c/s real, 5851 c/s virtual
gpg-opencl Sun Jul 26 00:32:01 CEST 2015
Benchmarking: gpg-opencl, OpenPGP / GnuPG Secret Key [SHA1 OpenCL]... (4xOMP) DONE
Speed for cost 1 (s2k-count) of 65536, cost 2 (hash algorithm [1:MD5 2:SHA1 3:RIPEMD160 8:SHA256 9:SHA384 10:SHA512 11:SHA224]) of 2, cost 3 (cipher algorithm [1:IDEA 2:3DES 3:CAST5 4:Blowfish 7:AES128 8:AES192 9:AES256]) of 3
Raw: 43497 c/s real, 132843 c/s virtual
keychain-opencl Sun Jul 26 00:32:11 CEST 2015
Benchmarking: keychain-opencl, Mac OS X Keychain [PBKDF2-SHA1 OpenCL 3DES]... (4xOMP) DONE
Raw: 14894 c/s real, 102400 c/s virtual
keyring-opencl Sun Jul 26 00:32:16 CEST 2015
Benchmarking: keyring-opencl, GNOME Keyring [SHA256 OpenCL AES]... (4xOMP) DONE
Speed for cost 1 (iteration count) of 3221 and 2439
Raw: 3757 c/s real, 27306 c/s virtual
krb5pa-md5-opencl Sun Jul 26 00:32:22 CEST 2015
Benchmarking: krb5pa-md5-opencl, Kerberos 5 AS-REQ Pre-Auth etype 23 [MD4 HMAC-MD5 RC4 OpenCL]... DONE
Many salts: 2746K c/s real, 72089K c/s virtual
Only one salt: 2427K c/s real, 23831K c/s virtual
krb5pa-sha1-opencl Sun Jul 26 00:32:25 CEST 2015
Benchmarking: krb5pa-sha1-opencl, Kerberos 5 AS-REQ Pre-Auth etype 17/18 [PBKDF2-SHA1 OpenCL 4x]... (4xOMP) DONE
Raw: 3296 c/s real, 126030 c/s virtual
LM-opencl Sun Jul 26 00:32:46 CEST 2015
Benchmarking: LM-opencl [LM BS OpenCL(inefficient)]... DONE
Raw: 13865K c/s real, 167772K c/s virtual
mscash-opencl Sun Jul 26 00:32:53 CEST 2015
Benchmarking: mscash-opencl, M$ Cache Hash [MD4 OpenCL]... DONE
Raw: 7557K c/s real, 419430K c/s virtual
mscash2-opencl Sun Jul 26 00:32:55 CEST 2015
Benchmarking: mscash2-opencl, MS Cache Hash 2 (DCC2) [PBKDF2-SHA1 OpenCL]... DONE
Raw: 3068 c/s real, 3079 c/s virtual
mysql-sha1-opencl Sun Jul 26 00:33:23 CEST 2015
Benchmarking: mysql-sha1-opencl, MySQL 4.1+ [SHA1 OpenCL (inefficient, development use only)]... DONE
Raw: 5041K c/s real, 47662K c/s virtual
ssha-opencl Sun Jul 26 00:33:25 CEST 2015
Benchmarking: ssha-opencl, Netscape LDAP {SSHA} [SHA1 OpenCL (inefficient, development use mostly)]... DONE
Many salts: 9460K c/s real, 139810K c/s virtual
Only one salt: 8388K c/s real, 48395K c/s virtual
nt-opencl Sun Jul 26 00:33:31 CEST 2015
Benchmarking: nt-opencl, NT [MD4 OpenCL]... DONE
Raw: 10847K c/s real, 89877K c/s virtual
ntlmv2-opencl Sun Jul 26 00:33:34 CEST 2015
Benchmarking: ntlmv2-opencl, NTLMv2 C/R [MD4 HMAC-MD5 OpenCL 4x]... DONE
Many salts: 14979K c/s real, 196608K c/s virtual
Only one salt: 9799K c/s real, 49932K c/s virtual
o5logon-opencl Sun Jul 26 00:33:39 CEST 2015
Benchmarking: o5logon-opencl, Oracle O5LOGON protocol [SHA1 OpenCL AES 32/64]... DONE
Raw: 2746K c/s real, 4179K c/s virtual
ODF-opencl Sun Jul 26 00:33:42 CEST 2015
Benchmarking: ODF-opencl [SHA1 OpenCL Blowfish]... (4xOMP) DONE
Raw: 18618 c/s real, 12800 c/s virtual
ODF-AES-opencl Sun Jul 26 00:33:48 CEST 2015
Benchmarking: ODF-AES-opencl [SHA256 OpenCL AES]... (4xOMP) DONE
Speed for cost 1 (iteration count) of 1
Raw: 12047 c/s real, 21557 c/s virtual
office2007-opencl Sun Jul 26 00:33:53 CEST 2015
Benchmarking: office2007-opencl, MS Office 2007 (50,000 iterations) [SHA1 OpenCL 4x AES]... (4xOMP) DONE
Raw: 1110 c/s real, 81920 c/s virtual
office2010-opencl Sun Jul 26 00:34:38 CEST 2015
Benchmarking: office2010-opencl, MS Office 2010 (100,000 iterations) [SHA1 OpenCL 4x AES]... (4xOMP) DONE
Speed for cost 1 (iteration count) of 100000
Raw: 519 c/s real, 24094 c/s virtual
office2013-opencl Sun Jul 26 00:35:27 CEST 2015
Benchmarking: office2013-opencl, MS Office 2013 (100,000 iterations) [SHA512 OpenCL 2x AES]... (4xOMP) PBKDF2-HMAC-SHA256-opencl Sun Jul 26 00:35:32 CEST 2015
Benchmarking: PBKDF2-HMAC-SHA256-opencl, rounds=12000 [PBKDF2-SHA256 OpenCL]... DONE
Speed for cost 1 (iteration count) of 12000
Raw: 1116 c/s real, 409600 c/s virtual
PBKDF2-HMAC-SHA512-opencl Sun Jul 26 00:36:03 CEST 2015
Benchmarking: PBKDF2-HMAC-SHA512-opencl, GRUB2 / OS X 10.8+, rounds=10000 [PBKDF2-SHA512 OpenCL]... phpass-opencl Sun Jul 26 00:36:07 CEST 2015
Benchmarking: phpass-opencl ($P$9 lengths 0 to 15) [MD5 OpenCL]... DONE
Raw: 56219 c/s real, 5734K c/s virtual
pwsafe-opencl Sun Jul 26 00:36:10 CEST 2015
Benchmarking: pwsafe-opencl, Password Safe [SHA256 OpenCL]... DONE
Speed for cost 1 (iteration count) of 2048
Raw: 13212 c/s real, 819200 c/s virtual
RAKP-opencl Sun Jul 26 00:36:17 CEST 2015
Benchmarking: RAKP-opencl, IPMI 2.0 RAKP (RMCP+) [HMAC-SHA1 OpenCL 4x]... DONE
Many salts: 10586K c/s real, 110100K c/s virtual
Only one salt: 8824K c/s real, 38751K c/s virtual
Raw-MD4-opencl Sun Jul 26 00:36:42 CEST 2015
Benchmarking: Raw-MD4-opencl [MD4 OpenCL]... DONE
Raw: 4802K c/s real, 78643K c/s virtual
Raw-MD5-opencl Sun Jul 26 00:36:44 CEST 2015
Benchmarking: Raw-MD5-opencl [MD5 OpenCL]... DONE
Raw: 4660K c/s real, 62914K c/s virtual
Raw-SHA1-opencl Sun Jul 26 00:36:46 CEST 2015
Benchmarking: Raw-SHA1-opencl [SHA1 OpenCL]... DONE
Raw: 4481K c/s real, 65536K c/s virtual
Raw-SHA256-opencl Sun Jul 26 00:36:47 CEST 2015
Benchmarking: Raw-SHA256-opencl [SHA256 OpenCL]... DONE
Raw: 5969K c/s real, 54811K c/s virtual
Raw-SHA512-opencl Sun Jul 26 00:36:51 CEST 2015
7z-opencl Sun Jul 26 00:36:51 CEST 2015
Benchmarking: 7z-opencl, 7-Zip (512K iterations) [SHA256 OPENCL AES]... (4xOMP) DONE
Speed for cost 1 (iteration count) of 524288
Raw: 76.2 c/s real, 8533 c/s virtual
strip-opencl Sun Jul 26 00:38:02 CEST 2015
Benchmarking: strip-opencl, STRIP Password Manager [PBKDF2-SHA1 OpenCL]... (4xOMP) DONE
Raw: 3531 c/s real, 81920 c/s virtual
sxc-opencl Sun Jul 26 00:38:08 CEST 2015
Benchmarking: sxc-opencl, StarOffice .sxc [PBKDF2-SHA1 OpenCL Blowfish]... (4xOMP) DONE
Speed for cost 1 (iteration count) of 0
Raw: 21140 c/s real, 18004 c/s virtual
wpapsk-opencl Sun Jul 26 00:38:14 CEST 2015
Benchmarking: wpapsk-opencl, WPA/WPA2 PSK [PBKDF2-SHA1 OpenCL 4x]... XSHA512-opencl Sun Jul 26 00:38:30 CEST 2015
zip-opencl Sun Jul 26 00:38:30 CEST 2015
Benchmarking: zip-opencl, ZIP [PBKDF2-SHA1 OpenCL AES]... (4xOMP) DONE
Raw: 7656 c/s real, 273066 c/s virtual
Sun Jul 26 00:38:36 CEST 2015I fixed oldoffice. Maybe it was actually proper code but better safe than sorry.
frank-dittrich
commented
9 years ago A copy of magnum's link from #1515 https://gist.github.com/spiralray/cae0bc235509e495fec1
Just in case someone wants to test this on ubuntu...
frank-dittrich
commented
9 years ago BTW, I reported two beignet bugs, let's we what happens. https://bugs.freedesktop.org/show_bug.cgi?id=91524 https://bugs.freedesktop.org/show_bug.cgi?id=91525
frank-dittrich
commented
9 years ago One of these beignet bugs has recently been fixed; https://bugs.freedesktop.org/show_bug.cgi?id=91524
But for WPAPSK, the GPU can't compete...
(bleeding-jumbo)run $ ../run/john --test=20 --format=wpapsk-opencl --verbosity=5
initUnicode(UNICODE, ASCII/ASCII)
ASCII -> ASCII -> ASCII
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
Benchmarking: wpapsk-opencl, WPA/WPA2 PSK [PBKDF2-SHA1 OpenCL 4x]... Options used: -I ../run/kernels -cl-mad-enable -D__GPU__ -DDEVICE_INFO=34 -DDEV_VER_MAJOR=1 -DDEV_VER_MINOR=2 -D_OPENCL_COMPILER -DHASH_LOOPS=105 -DITERATIONS=4095 -DPLAINTEXT_LENGTH=63 -DV_WIDTH=4
Calculating best global worksize (GWS); max. 10s total for crypt_all()
Raw speed figures including buffer transfers:
xfer: 64.517ms*, init: 677.920us, loop: 78x20.535ms, pass2: 362.800us, final: 1.050ms, xfer: 62.358ms*
gws: 512 1183 c/s 19379906 rounds/s 1.730s per crypt_all()!
xfer: 66.064ms*, init: 1.298ms, loop: 78x20.818ms, pass2: 665.600us, final: 1.084ms, xfer: 64.368ms*
gws: 1024 2330 c/s 38170060 rounds/s 1.757s per crypt_all()+
xfer: 1.040us, init: 2.650ms, loop: 78x38.103ms, pass2: 1.302ms, final: 2.075ms, xfer: 118.578ms*
gws: 2048 2645 c/s 43330390 rounds/s 3.097s per crypt_all()+
xfer: 2us, init: 5.187ms, loop: 78x68.788ms, pass2: 2.546ms, final: 3.807ms, xfer: 1.280us
gws: 4096 3046 c/s 49899572 rounds/s 5.377s per crypt_all()+
xfer: 404.082ms*, init: 10.301ms, loop: 78x126.007ms, pass2: 4.866ms, final: 6.951ms, xfer: 389.816ms*
gws: 8192 3078 c/s 50423796 rounds/s 10.645s per crypt_all() - too slow
Local worksize (LWS) 512, global worksize (GWS) 4096
DONE
Raw: 2958 c/s real, 192752 c/s virtual
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
(bleeding-jumbo)run $ ../run/john --test=20 --format=wpapsk --verbosity=5
initUnicode(UNICODE, ASCII/ASCII)
ASCII -> ASCII -> ASCII
Will run 4 OpenMP threads
Benchmarking: wpapsk, WPA/WPA2 PSK [PBKDF2-SHA1 256/256 AVX2 8x]... (4xOMP) DONE
Raw: 9982 c/s real, 2512 c/s virtualNative vector widths: char 8, short 8, int 4, long 2
This is not what the Intel driver says (all 1). Perhaps a beigenet bug. Or maybe they are not the same model? Here's your i5-4570 output with warnings redacted:
$ OCL_IGNORE_SELF_TEST=1 ./john --list=opencl-devices --verbosity=5
Platform #0 name: Intel Gen OCL Driver
Platform version: OpenCL 1.2 beignet 1.1 (git-834d0ae)
Platform extensions: cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_khr_byte_addressable_store cl_khr_spir cl_khr_icd
Device #0 (0) name: Intel(R) HD Graphics Haswell GT2 Desktop
Device vendor: Intel
Device type: GPU (LE)
Device version: OpenCL 1.2 beignet 1.1 (git-834d0ae)
Driver version: 1.1
Native vector widths: char 8, short 8, int 4, long 2
Preferred vector width: char 16, short 8, int 4, long 2
Global Memory: 2.0 GB
Device extensions: cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_khr_byte_addressable_store cl_khr_spir cl_khr_icd
Global Memory Cache: 8.0 KB
Local Memory: 64.0 KB (Global)
Max memory alloc. size: 1024.0 MB
Max clock (MHz): 1000
Profiling timer res.: 80 ns
Max Work Group Size: 512
Parallel compute cores: 20Here's well's i7-4770K w/ Intel driver
Platform #1 name: Intel(R) OpenCL, version: OpenCL 1.2
Platform extensions: cl_khr_3d_image_writes cl_khr_byte_addressable_store cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_icd cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_khr_spir
Device #0 (3) name: Intel(R) HD Graphics
Device vendor: Intel(R) Corporation
Device type: GPU (LE)
Device version: OpenCL 1.2
Driver version: 16.4.2.1.39163
Native vector widths: char 1, short 1, int 1, long 1
Preferred vector width: char 1, short 1, int 1, long 1
Global Memory: 1.0 GB
Device extensions: cl_intel_accelerator cl_intel_advanced_motion_estimation cl_intel_motion_estimation cl_intel_subgroups cl_intel_va_api_media_sharing cl_khr_3d_image_writes cl_khr_byte_addressable_store cl_khr_global_int32_base_atomics cl_khr_global_int32_extended_atomics cl_khr_icd cl_khr_local_int32_base_atomics cl_khr_local_int32_extended_atomics cl_khr_spir
Global Memory Cache: 256.2 KB
Local Memory: 64.0 KB (Local)
Max memory alloc. size: 407.4 MB
Max clock (MHz): 1250
Profiling timer res.: 80 ns
Max Work Group Size: 512
Parallel compute cores: 20With the latest beignet commit http://cgit.freedesktop.org/beignet/commit/?id=18a52ffc966027a3004b85c7c03c9416e1a84c3a the bug https://bugs.freedesktop.org/show_bug.cgi?id=91525 seems to be fixed. I don't get ASSERTION FAILED anymore.
However, this is what happens now:
$ time ./john --test --format=office2013-opencl
Will run 4 OpenMP threads
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
Benchmarking: office2013-opencl, MS Office 2013 (100,000 iterations) [SHA512 OpenCL 2x AES]... (4xOMP)
total:1 hash_loops 100 rounds 100004 looptime 0
drm_intel_gem_bo_context_exec() failed: Input/output error
drm_intel_gem_bo_context_exec() failed: Input/output error
drm_intel_gem_bo_context_exec() failed: Input/output error
OpenCL CL_INVALID_EVENT error in common-opencl.c:1357 - WaitForEvents failed
real 0m33.325s
user 0m17.857s
sys 0m0.074s$ time ./john --test --format=office2013-opencl --verbosity=5
initUnicode(UNICODE, ASCII/ASCII)
ASCII -> ASCII -> ASCII
Will run 4 OpenMP threads
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
Benchmarking: office2013-opencl, MS Office 2013 (100,000 iterations) [SHA512 OpenCL 2x AES]... (4xOMP) Options used: -I ./kernels -cl-mad-enable -D__GPU__ -DDEVICE_INFO=34 -DDEV_VER_MAJOR=1 -DDEV_VER_MINOR=2 -D_OPENCL_COMPILER -DHASH_LOOPS=100 -DUNICODE_LENGTH=96 -DV_WIDTH=2
Calculating best GWS for LWS=32; max. 5s total for crypt_all()
Raw speed figures including buffer transfers:
xfer: 5.385s*, xfer: 5.385s*, init: 1.274ms, loop: 1000x0ns, final: 6.219ms, xfer: 6.328ms*
total:1 hash_loops 100 rounds 100004 looptime 0
gws: 128 23 c/s 2300092 rounds/s 10.785s per crypt_all()!
drm_intel_gem_bo_context_exec() failed: Input/output error
drm_intel_gem_bo_context_exec() failed: Input/output error
OpenCL CL_OUT_OF_RESOURCES error in opencl_office2013_fmt_plug.c:328 - failed in clEnqueueNDRangeKernel
(error occurred)Calculating best LWS for GWS=128
drm_intel_gem_bo_context_exec() failed: Input/output error
drm_intel_gem_bo_context_exec() failed: Input/output error
OpenCL CL_OUT_OF_RESOURCES error in opencl_office2013_fmt_plug.c:320 - failed in clEnqueueNDRangeKernel
drm_intel_gem_bo_context_exec() failed: Input/output error
drm_intel_gem_bo_context_exec() failed: Input/output error
OpenCL CL_OUT_OF_RESOURCES error in opencl_office2013_fmt_plug.c:320 - failed in clEnqueueNDRangeKernel
OpenCL CL_INVALID_EVENT error in common-opencl.c:1357 - WaitForEvents failed
real 0m33.834s
user 0m17.863s
sys 0m0.079sNot sure whether this is a bleeding-jumbo error, a beignet error, a kernel driver error, or a combination of errors. Any thoughts?
I'll try to upgrade to Linux kernel version 4.3.0-0.rc1 and see whether the problem disappears.
magnumripper
commented
9 years ago I see nothing that points to a bug in Jumbo. Whether it's beignet or kernel I have no idea.
frank-dittrich
commented
9 years ago What about this then?
$ ./john --list=format-all-details --format=descrypt-opencl
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Format label descrypt-opencl
Disabled in configuration file no
Min. password length in bytes 0
Max. password length in bytes 8
Min. keys per crypt 32
Max. keys per crypt 32
Flags
Case sensitive yes
Truncates at (our) max. length yes
Supports 8-bit characters no
Converts 8859-1 to UTF-16/UCS-2 no
Honours --encoding=NAME no
False positives possible no
Uses a bitslice implementation yes
The split() method unifies case no
A $dynamic$ format no
A dynamic sized salt no
Parallelized with OpenMP no
Number of test vectors 6
Algorithm name DES OpenCL
Format name traditional crypt(3)
Benchmark comment
Benchmark length 0
Binary size 8
Salt size 4
Tunable cost parameters
Example ciphertext CCNf8Sbh3HDfQ
OpenCL CL_INVALID_PROGRAM error in opencl_DES_bs_b_plug.c:104 - Error releasing Program
(bleeding-jumbo)run $ echo $?
1 $ ./john --list=format-details --format=descrypt-opencl
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
descrypt-opencl 8 32 32 00210001 6 DES OpenCL traditional crypt(3) 0 8 4 0 CCNf8Sbh3HDfQ
OpenCL CL_INVALID_PROGRAM error in opencl_DES_bs_b_plug.c:104 - Error releasing Program$ ./john --list=format-details --format=opencl
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
sha1crypt-opencl 64 1 1 00000003 4 PBKDF1-SHA1 OpenCL 4x (NetBSD) -1001 20 64 iteration count 0 $sha1$64000$wnUR8T1U$vt1TFQ50tBMFgkflAFAOer2CwdYZ
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
oldoffice-opencl 57 1 1 0002000f 5 MD5/SHA1 RC4 OpenCL MS Office <= 2003 -1000 0 72 hash type 0 $oldoffice$1*de17a7f3c3ff03a39937ba9666d6e952*2374d5b6ce7449f57c9f252f9f9b53d2*e60e1185f7aecedba262f869c0236f81
[...]
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
md5crypt-opencl 15 1 1 00000003 61 MD5 OpenCL crypt(3) $1$ -1 16 9 0 $1$Btiy90iG$bGn4vzF3g1rIVGZ5odGIp/
sha256crypt-opencl 24 1 1 00000003 7 SHA256 OpenCL crypt(3) $5$ (rounds=5000) -1 32 28 iteration count 0 $5$LKO/Ute40T3FNF95$U0prpBQd4PloSGU0pnpM4z9wKn4vZ1.jsrzQfPqxph9
sha512crypt-opencl 23 1 1 00000003 6 SHA512 OpenCL crypt(3) $6$ (rounds=5000) -1 64 32 iteration count 0 $6$LKO/Ute40T3FNF95$6S/6T2YuOIHY0N3XpLKABJ3soYcXD9mB7uVbtEZDj/LNscVhZoZ9DEH.sBciDrMsHOWOoASbNLTypH/5X26gN0
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
descrypt-opencl 8 32 32 00210001 6 DES OpenCL traditional crypt(3) 0 8 4 0 CCNf8Sbh3HDfQ
Segmentation fault (core dumped)Backtrace of that segfault:
Program received signal SIGSEGV, Segmentation fault.
0x00000000000000ff in ?? ()
(gdb) bt
#0 0x00000000000000ff in ?? ()
#1 0x00007ffff70dc9c7 in clReleaseProgram () from /lib64/libOpenCL.so.1
#2 0x00000000005a6ca3 in clean_all_buffers () at opencl_DES_bs_b_plug.c:104
#3 0x00000000006c4b87 in fmt_done (format=format@entry=0xb0d300 <fmt_opencl_DES>) at formats.c:109
#4 0x00000000006f47fc in listconf_parse_late () at listconf.c:570
#5 0x00000000006ce07d in john_init (name=0x7fffffffe314 "john", argc=3, argv=0x7fffffffdfd8) at john.c:1439
#6 0x00000000006ced1f in main (argc=3, argv=0x7fffffffdfd8) at john.c:1867For this one
$ ./john --test --format=descrypt-opencl --verbosity=5
initUnicode(UNICODE, ASCII/ASCII)
ASCII -> ASCII -> ASCII
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Failed to release test userptr object! (9) i915 kernel driver may not be sane!
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
Benchmarking: descrypt-opencl, traditional crypt(3) [DES OpenCL]... Options used: -I ./kernels -cl-mad-enable -D__GPU__ -DDEVICE_INFO=34 -DDEV_VER_MAJOR=1 -DDEV_VER_MINOR=2 -D_OPENCL_COMPILER
ASSERTION FAILED: isScalarType(type)
at file /home/fd/git/beignet/backend/src/llvm/llvm_gen_backend.cpp, function gbe::ir::Type gbe::getType(gbe::ir::Context&, const llvm::Type*), line 196
Trace/breakpoint trap (core dumped)I filed a new beignet bug: https://bugs.freedesktop.org/show_bug.cgi?id=92081
magnumripper
commented
9 years ago I opened #1778 for the problem with new DEScrypt-opencl and --list (it's unrelated to beignet).
frank-dittrich
commented
9 years ago Beignet bug https://bugs.freedesktop.org/show_bug.cgi?id=92081 (fixed earlier) reappeared for a different opencl format.
frank-dittrich
commented
8 years ago Finally I found the time to re-test https://bugs.freedesktop.org/show_bug.cgi?id=91525 (GPU hang when testing --format=office2013-opencl on my Haswell with latest beignet.
Running that test after echo -n 0 > /sys/module/i915/parameters/enable_hangcheck to disable the hang check didn't help.
I kept john running for about half an hour, the system didn't respond, so I has to use the power button to reboot.
That's why I doubt it is john's fault.
But let's see what beignet developers think.
Even with GWS=1 LWS=1 I get the hang.
$ GWS=1 LWS=1 ./john --test=0 --format=office2013-opencl --verbosity=5
initUnicode(UNICODE, ASCII/ASCII)
ASCII -> ASCII -> ASCII
Will run 4 OpenMP threads
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
Testing: office2013-opencl, MS Office 2013 (100,000 iterations) [SHA512 OpenCL 2x AES]... (4xOMP) Options used: -I ./kernels -cl-mad-enable -D__GPU__ -DDEVICE_INFO=34 -DSIZEOF_SIZE_T=8 -DDEV_VER_MAJOR=1 -DDEV_VER_MINOR=2 -D_OPENCL_COMPILER -DHASH_LOOPS=100 -DUNICODE_LENGTH=96 -DV_WIDTH=2 $JOHN/kernels/office2013_kernel.cl
Local worksize (LWS) 1, global worksize (GWS) 1
drm_intel_gem_bo_context_exec() failed: Input/output error
drm_intel_gem_bo_context_exec() failed: Input/output error
OpenCL CL_OUT_OF_RESOURCES error in opencl_office2013_fmt_plug.c:323 - failed in clEnqueueNDRangeKernelThis is the corresponting dmesg output:
[ 149.302616] [drm] stuck on render ring
[ 149.303861] [drm] GPU HANG: ecode 7:0:0x85ddfffc, in john [2110], reason: Ring hung, action: reset
[ 149.303865] [drm] GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace.
[ 149.303867] [drm] Please file a _new_ bug report on bugs.freedesktop.org against DRI -> DRM/Intel
[ 149.303870] [drm] drm/i915 developers can then reassign to the right component if it's not a kernel issue.
[ 149.303872] [drm] The gpu crash dump is required to analyze gpu hangs, so please always attach it.
[ 149.303874] [drm] GPU crash dump saved to /sys/class/drm/card0/error
[ 149.306033] drm/i915: Resetting chip after gpu hang
[ 155.301738] [drm] stuck on render ring
[ 155.302955] [drm] GPU HANG: ecode 7:0:0x85ddfffc, in john [2110], reason: Ring hung, action: reset
[ 155.305137] drm/i915: Resetting chip after gpu hangThe contents of /sys/class/drm/card0/error has been attached to the beignet bug report, see https://bugs.freedesktop.org/show_bug.cgi?id=91525#c11.
Repeated tests of that formats just result in these dmesg lines:
[ 520.187119] [drm] stuck on render ring
[ 520.188347] [drm] GPU HANG: ecode 7:0:0x85ddfffc, in john [2239], reason: Ring hung, action: reset
[ 520.190142] drm/i915: Resetting chip after gpu hang
[ 526.179121] [drm] stuck on render ring
[ 526.180357] [drm] GPU HANG: ecode 7:0:0x85ddfffc, in john [2239], reason: Ring hung, action: reset
[ 526.182550] drm/i915: Resetting chip after gpu hang
[ 3991.026791] [drm] stuck on render ring
[ 3991.027969] [drm] GPU HANG: ecode 7:0:0x85ddfffc, in john [2733], reason: Ring hung, action: reset
[ 3991.030133] drm/i915: Resetting chip after gpu hang
[ 3997.024799] [drm] stuck on render ring
[ 3997.026080] [drm] GPU HANG: ecode 7:0:0xf3cffffe, in john [2733], reason: Ring hung, action: reset
[ 3997.027766] drm/i915: Resetting chip after gpu hangBut with linux kernel 4.3, the
Failed to release test userptr object! (9) i915 kernel driver may not be sane!lines finally disappeared.
frank-dittrich
commented
8 years ago Next week I'll most likely have time to re-test. At least for one of the problems there should be a commit fixing the problem.
frank-dittrich
commented
8 years ago Status update:
With latest John the Ripper bleeding-jumbo commit 8d4470ff9f2357fc10c8e5769dbb164eb5118f40. latest beignet commit
commit 032b606f8c5baa53e52b1f55c4f7c0bafdd6ff37
Author: Junyan He <junyan.he@linux.intel.com>
Date: Mon Dec 14 14:51:26 2015 +0800
Backend: Fix a memory leak for structurizer.
In structurizer, the useless instruction is just be
erased from block. The iintrusive_list::erase() just
unlink the instruction, but not free its resource.
We should use remove() to deallocate the instruction
object.
Signed-off-by: Junyan He <junyan.he@linux.intel.com>
Reviewed-by: Yang Rong <rong.r.yang@intel.com>and a newer Linux kernel (4.4.0-0.rc6.git0.1.vanilla.knurd.1.fc22.x86_64), I get these results:
https://bugs.freedesktop.org/show_bug.cgi?id=91524
./john --test=0 --format=PBKDF2-HMAC-SHA512-opencl --verbosity=5
initUnicode(UNICODE, ASCII/ASCII)
ASCII -> ASCII -> ASCII
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
Testing: PBKDF2-HMAC-SHA512-opencl, GRUB2 / OS X 10.8+, rounds=1000 [PBKDF2-SHA512 OpenCL]... Loaded 11 hashes with 10 different salts to test db from test vectors
Options used: -I ./kernels -cl-mad-enable -D__GPU__ -DDEVICE_INFO=34 -DSIZEOF_SIZE_T=8 -DDEV_VER_MAJOR=1 -DDEV_VER_MINOR=2 -D_OPENCL_COMPILER -DHASH_LOOPS=250 -DPLAINTEXT_LENGTH=110 -DMAX_SALT_SIZE=107 $JOHN/kernels/pbkdf2_hmac_sha512_kernel.cl
ASSERTION FAILED: !(ctx->getErrCode() == OUT_OF_RANGE_IF_ENDIF && ctx->getIFENDIFFix())
at file /home/fd/git/beignet/backend/src/backend/gen_program.cpp, function virtual gbe::Kernel* gbe::GenProgram::compileKernel(const gbe::ir::Unit&, const string&, bool, int), line 206
Trace/breakpoint trap (core dumped)https://bugs.freedesktop.org/show_bug.cgi?id=91525
Sometimes I still get
GWS=1 LWS=1 ./john --test=0 --format=office2013-opencl --verbosity=5
initUnicode(UNICODE, ASCII/ASCII)
ASCII -> ASCII -> ASCII
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
Testing: office2013-opencl, MS Office 2013 (100,000 iterations) [SHA512 OpenCL 2x AES]... Loaded 5 hashes with 5 different salts to test db from test vectors
Options used: -I ./kernels -cl-mad-enable -D__GPU__ -DDEVICE_INFO=34 -DSIZEOF_SIZE_T=8 -DDEV_VER_MAJOR=1 -DDEV_VER_MINOR=2 -D_OPENCL_COMPILER -DHASH_LOOPS=100 -DUNICODE_LENGTH=96 -DV_WIDTH=2 $JOHN/kernels/office2013_kernel.cl
Local worksize (LWS) 1, global worksize (GWS) 1
drm_intel_gem_bo_context_exec() failed: Input/output error
OpenCL CL_OUT_OF_RESOURCES error in opencl_office2013_fmt_plug.c:323 - failed in clEnqueueNDRangeKerneland dmesg shows
[ 2236.597358] [drm] stuck on render ring
[ 2236.598523] [drm] GPU HANG: ecode 7:0:0x85ddfffc, in john [2894], reason: Ring hung, action: reset
[ 2236.598527] [drm] GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace.
[ 2236.598529] [drm] Please file a _new_ bug report on bugs.freedesktop.org against DRI -> DRM/Intel
[ 2236.598531] [drm] drm/i915 developers can then reassign to the right component if it's not a kernel issue.
[ 2236.598533] [drm] The gpu crash dump is required to analyze gpu hangs, so please always attach it.
[ 2236.598536] [drm] GPU crash dump saved to /sys/class/drm/card0/error
[ 2236.600685] drm/i915: Resetting chip after gpu hang
[ 2242.597091] [drm] stuck on render ring
[ 2242.597661] [drm] GPU HANG: ecode 7:0:0x85ddfffc, in john [2894], reason: Ring hung, action: reset
[ 2242.599757] drm/i915: Resetting chip after gpu hangI attached the contents of /sys/class/drm/card0/error to the beignet bug.
But sometimes I don't run into this problem, instead I see
$ GWS=1 LWS=1 ./john --test=0 --format=office2013-opencl --verbosity=5
initUnicode(UNICODE, ASCII/ASCII)
ASCII -> ASCII -> ASCII
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
Testing: office2013-opencl, MS Office 2013 (100,000 iterations) [SHA512 OpenCL 2x AES]... Loaded 5 hashes with 5 different salts to test db from test vectors
Options used: -I ./kernels -cl-mad-enable -D__GPU__ -DDEVICE_INFO=34 -DSIZEOF_SIZE_T=8 -DDEV_VER_MAJOR=1 -DDEV_VER_MINOR=2 -D_OPENCL_COMPILER -DHASH_LOOPS=100 -DUNICODE_LENGTH=96 -DV_WIDTH=2 $JOHN/kernels/office2013_kernel.cl
Local worksize (LWS) 1, global worksize (GWS) 1
FAILED (cmp_all(1))and when running that format with --skip-self-tests against the self test hashes, the correct passwords don't crack any hash.
https://bugs.freedesktop.org/show_bug.cgi?id=92081
$ ./john --test --format=descrypt-opencl --verbosity=5
initUnicode(UNICODE, ASCII/ASCII)
ASCII -> ASCII -> ASCII
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
Benchmarking: descrypt-opencl, traditional crypt(3) [DES OpenCL]... Loaded 6 hashes with 4 different salts to test db from test vectors
Options used: -I ./kernels -cl-mad-enable -D__GPU__ -DDEVICE_INFO=34 -DSIZEOF_SIZE_T=8 -DDEV_VER_MAJOR=1 -DDEV_VER_MINOR=2 -D_OPENCL_COMPILER $JOHN/kernels/DES_bs_hash_checking_kernel.cl
ASSERTION FAILED: it != allocatedBlocks.end()
at file /home/fd/git/beignet/backend/src/backend/context.cpp, function void gbe::SimpleAllocator::splitBlock(int32_t, int32_t), line 293
Trace/breakpoint trap (core dumped)https://bugs.freedesktop.org/show_bug.cgi?id=91525
looks like john could be the culprit, causing kernel hangs.
Evidence: Try to crack a single hash:
$ ./john hash.office2013-opencl --wordlist=pw.office2013-opencl --skip-self-tests --verbosity=5 --format=office2013-opencl --pot=office2013-opencl.pot
initUnicode(UNICODE, UTF-8/ISO-8859-1)
UTF-8 -> UTF-8 -> UTF-8
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
Using default input encoding: UTF-8
Loaded 1 password hash (office2013-opencl, MS Office 2013 [SHA512 OpenCL 2x AES])
Cost 1 (iteration count) is 100000 for all loaded hashes
Loaded 5 hashes with 5 different salts to test db from test vectors
Options used: -I ./kernels -cl-mad-enable -D__GPU__ -DDEVICE_INFO=34 -DSIZEOF_SIZE_T=8 -DDEV_VER_MAJOR=1 -DDEV_VER_MINOR=2 -D_OPENCL_COMPILER -DHASH_LOOPS=100 -DUNICODE_LENGTH=96 -DV_WIDTH=2 $JOHN/kernels/office2013_kernel.cl
Calculating best GWS for LWS=8; max. 5s total for crypt_all()
Raw speed figures including buffer transfers:
xfer: 1.680us, xfer: 1.120us, init: 1.384ms, loop: 1000x4.293s, final: 6.285ms, xfer: 1.120us
gws: 128 0c/s 0 rounds/s 4293.777s per crypt_all()
drm_intel_gem_bo_context_exec() failed: Input/output error
OpenCL CL_OUT_OF_RESOURCES error in opencl_office2013_fmt_plug.c:328 - failed in clEnqueueNDRangeKernel
(error occurred)Calculating best LWS for GWS=8
drm_intel_gem_bo_context_exec() failed: Input/output error
OpenCL CL_OUT_OF_RESOURCES error in opencl_office2013_fmt_plug.c:320 - failed in clEnqueueNDRangeKernel
drm_intel_gem_bo_context_exec() failed: Input/output error
OpenCL CL_OUT_OF_RESOURCES error in opencl_office2013_fmt_plug.c:320 - failed in clEnqueueNDRangeKernel
OpenCL CL_INVALID_EVENT error in common-opencl.c:1406 - WaitForEvents failedLook at
Raw speed figures including buffer transfers:
xfer: 1.680us, xfer: 1.120us, init: 1.384ms, loop: 1000x4.293s, final: 6.285ms, xfer: 1.120us
gws: 128 0c/s 0 rounds/s 4293.777s per crypt_all()4293.777s per crypt_all() really looks bad.
Lets try LWS=1:
$ LWS=1 ./john hash.office2013-opencl --wordlist=pw.office2013-opencl --skip-self-tests --verbosity=5 --format=office2013-opencl --pot=office2013-opencl.pot
initUnicode(UNICODE, UTF-8/ISO-8859-1)
UTF-8 -> UTF-8 -> UTF-8
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
Using default input encoding: UTF-8
Loaded 1 password hash (office2013-opencl, MS Office 2013 [SHA512 OpenCL 2x AES])
Cost 1 (iteration count) is 100000 for all loaded hashes
Loaded 5 hashes with 5 different salts to test db from test vectors
Options used: -I ./kernels -cl-mad-enable -D__GPU__ -DDEVICE_INFO=34 -DSIZEOF_SIZE_T=8 -DDEV_VER_MAJOR=1 -DDEV_VER_MINOR=2 -D_OPENCL_COMPILER -DHASH_LOOPS=100 -DUNICODE_LENGTH=96 -DV_WIDTH=2 $JOHN/kernels/office2013_kernel.cl
Calculating best GWS for LWS=1; max. 10s total for crypt_all()
Raw speed figures including buffer transfers:
xfer: 13.398ms*, xfer: 13.405ms*, init: 2.327ms, loop: 1000x3.343ms, final: 6.401ms, xfer: 6.562ms*
gws: 20 11c/s 1100044 rounds/s 3.385s per crypt_all()!
xfer: 22.188ms*, xfer: 22.191ms*, init: 2.771ms, loop: 1000x6.444ms, final: 12.701ms, xfer: 12.809ms*
gws: 40 12c/s 1200048 rounds/s 6.517s per crypt_all()+
xfer: 33.229ms*, xfer: 33.230ms*, init: 4.456ms, loop: 1000x9.377ms, final: 18.996ms, xfer: 19.241ms*
gws: 80 16c/s 1600064 rounds/s 9.486s per crypt_all()+
xfer: 55.998ms*, xfer: 55.994ms*, init: 8.632ms, loop: 1000x15.485ms, final: 31.531ms, xfer: 31.723ms*
gws: 160 20c/s 2000080 rounds/s 15.669s per crypt_all() - too slow
Local worksize (LWS) 1, global worksize (GWS) 80
ERROR: SessionFileProtect enabled in john.conf, and ./john.rec existsOops
$ rm john.rec
$ LWS=1 ./john hash.office2013-opencl --wordlist=pw.office2013-opencl --skip-self-tests --verbosity=5 --format=office2013-opencl --pot=office2013-opencl.pot
initUnicode(UNICODE, UTF-8/ISO-8859-1)
UTF-8 -> UTF-8 -> UTF-8
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
Using default input encoding: UTF-8
Loaded 1 password hash (office2013-opencl, MS Office 2013 [SHA512 OpenCL 2x AES])
Cost 1 (iteration count) is 100000 for all loaded hashes
Loaded 5 hashes with 5 different salts to test db from test vectors
Options used: -I ./kernels -cl-mad-enable -D__GPU__ -DDEVICE_INFO=34 -DSIZEOF_SIZE_T=8 -DDEV_VER_MAJOR=1 -DDEV_VER_MINOR=2 -D_OPENCL_COMPILER -DHASH_LOOPS=100 -DUNICODE_LENGTH=96 -DV_WIDTH=2 $JOHN/kernels/office2013_kernel.cl
Calculating best GWS for LWS=1; max. 10s total for crypt_all()
Raw speed figures including buffer transfers:
xfer: 1.600us, xfer: 1.040us, init: 2.445ms, loop: 1000x4.292s, final: 6.303ms, xfer: 1.040us
gws: 20 0c/s 0 rounds/s 4292.715s per crypt_all()
drm_intel_gem_bo_context_exec() failed: Input/output error
OpenCL CL_OUT_OF_RESOURCES error in opencl_office2013_fmt_plug.c:328 - failed in clEnqueueNDRangeKernel
(error occurred)Local worksize (LWS) 1, global worksize (GWS) 20
Press 'q' or Ctrl-C to abort, almost any other key for status
drm_intel_gem_bo_context_exec() failed: Input/output error
OpenCL CL_OUT_OF_RESOURCES error in opencl_office2013_fmt_plug.c:320 - failed in clEnqueueNDRangeKernelGWS=1:
$ rm john.rec
$ GWS=1 ./john hash.office2013-opencl --wordlist=pw.office2013-opencl --skip-self-tests --verbosity=5 --format=office2013-opencl --pot=office2013-opencl.pot
initUnicode(UNICODE, UTF-8/ISO-8859-1)
UTF-8 -> UTF-8 -> UTF-8
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
Using default input encoding: UTF-8
Loaded 1 password hash (office2013-opencl, MS Office 2013 [SHA512 OpenCL 2x AES])
Cost 1 (iteration count) is 100000 for all loaded hashes
Loaded 5 hashes with 5 different salts to test db from test vectors
Options used: -I ./kernels -cl-mad-enable -D__GPU__ -DDEVICE_INFO=34 -DSIZEOF_SIZE_T=8 -DDEV_VER_MAJOR=1 -DDEV_VER_MINOR=2 -D_OPENCL_COMPILER -DHASH_LOOPS=100 -DUNICODE_LENGTH=96 -DV_WIDTH=2 $JOHN/kernels/office2013_kernel.cl
Calculating best GWS for LWS=8; max. 5s total for crypt_all()
Raw speed figures including buffer transfers:
xfer: 12.762ms*, xfer: 12.763ms*, init: 1.373ms, loop: 1000x3.594ms, final: 6.501ms, xfer: 6.671ms*
gws: 128 70c/s 7000280 rounds/s 3.635s per crypt_all()!
xfer: 12.492ms*, xfer: 12.485ms*, init: 2.275ms, loop: 1000x3.414ms, final: 6.501ms, xfer: 6.611ms*
gws: 256 148c/s 14800592 rounds/s 3.454s per crypt_all()!
xfer: 1.440us, xfer: 25.674ms*, init: 4.196ms, loop: 1000x6.666ms, final: 12.887ms, xfer: 14.659ms*
gws: 512 152c/s 15200608 rounds/s 6.724s per crypt_all() - too slow
Calculating best LWS for GWS=256
Testing LWS=8 GWS=256 ... 16.893ms+
Testing LWS=16 GWS=256 ... 17.542ms
Testing LWS=32 GWS=256 ... 18.119ms
Testing LWS=64 GWS=256 ... 17.855ms
Testing LWS=128 GWS=256 ... 17.521ms
Testing LWS=256 GWS=256 ... 17.902ms
Calculating best GWS for LWS=8; max. 10s total for crypt_all()
Raw speed figures including buffer transfers:
xfer: 12.319ms*, xfer: 12.324ms*, init: 2.278ms, loop: 1000x3.502ms, final: 6.274ms, xfer: 6.371ms*
gws: 160 90c/s 9000360 rounds/s 3.542s per crypt_all()!
xfer: 22.604ms*, xfer: 22.608ms*, init: 3.058ms, loop: 1000x6.693ms, final: 12.610ms, xfer: 12.772ms*
gws: 320 94c/s 9400376 rounds/s 6.767s per crypt_all()+
xfer: 39.586ms*, xfer: 39.555ms*, init: 5.431ms, loop: 1000x10.037ms, final: 19.238ms, xfer: 16.160us
gws: 640 126c/s 12600504 rounds/s 10.141s per crypt_all() - too slow
Local worksize (LWS) 8, global worksize (GWS) 320
Press 'q' or Ctrl-C to abort, almost any other key for status
0g 0:00:00:02 DONE (2015-12-23 01:59) 0g/s 0.7117p/s 0.7117c/s 0.7117C/s openwall..password
Session completed$ GWS=8 LWS=8 ./john --test --verbosity=5 --format=office2013-opencl
initUnicode(UNICODE, ASCII/ASCII)
ASCII -> ASCII -> ASCII
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
Benchmarking: office2013-opencl, MS Office 2013 (100,000 iterations) [SHA512 OpenCL 2x AES]... Loaded 5 hashes with 5 different salts to test db from test vectors
Options used: -I ./kernels -cl-mad-enable -D__GPU__ -DDEVICE_INFO=34 -DSIZEOF_SIZE_T=8 -DDEV_VER_MAJOR=1 -DDEV_VER_MINOR=2 -D_OPENCL_COMPILER -DHASH_LOOPS=100 -DUNICODE_LENGTH=96 -DV_WIDTH=2 $JOHN/kernels/office2013_kernel.cl
Local worksize (LWS) 8, global worksize (GWS) 8
drm_intel_gem_bo_context_exec() failed: Input/output error
OpenCL CL_OUT_OF_RESOURCES error in opencl_office2013_fmt_plug.c:323 - failed in clEnqueueNDRangeKernelStill consuming too much time?
$ $ time GWS=2 LWS=2 ./john --test=0 --verbosity=5 --format=office2013-opencl
initUnicode(UNICODE, ASCII/ASCII)
ASCII -> ASCII -> ASCII
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
Testing: office2013-opencl, MS Office 2013 (100,000 iterations) [SHA512 OpenCL 2x AES]... Loaded 5 hashes with 5 different salts to test db from test vectors
Options used: -I ./kernels -cl-mad-enable -D__GPU__ -DDEVICE_INFO=34 -DSIZEOF_SIZE_T=8 -DDEV_VER_MAJOR=1 -DDEV_VER_MINOR=2 -D_OPENCL_COMPILER -DHASH_LOOPS=100 -DUNICODE_LENGTH=96 -DV_WIDTH=2 $JOHN/kernels/office2013_kernel.cl
Local worksize (LWS) 2, global worksize (GWS) 2
FAILED (cmp_all(1))
real 0m20.240s
user 0m16.818s
sys 0m1.081sAny idea how I should try to debug the failing self test?
And why is GWS=1 LWS=1 failing?
$ time GWS=1 LWS=1 ./john --test=0 --verbosity=5 --format=office2013-opencl
initUnicode(UNICODE, ASCII/ASCII)
ASCII -> ASCII -> ASCII
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
Testing: office2013-opencl, MS Office 2013 (100,000 iterations) [SHA512 OpenCL 2x AES]... Loaded 5 hashes with 5 different salts to test db from test vectors
Options used: -I ./kernels -cl-mad-enable -D__GPU__ -DDEVICE_INFO=34 -DSIZEOF_SIZE_T=8 -DDEV_VER_MAJOR=1 -DDEV_VER_MINOR=2 -D_OPENCL_COMPILER -DHASH_LOOPS=100 -DUNICODE_LENGTH=96 -DV_WIDTH=2 $JOHN/kernels/office2013_kernel.cl
Local worksize (LWS) 1, global worksize (GWS) 1
drm_intel_gem_bo_context_exec() failed: Input/output error
OpenCL CL_OUT_OF_RESOURCES error in opencl_office2013_fmt_plug.c:323 - failed in clEnqueueNDRangeKernel
real 0m29.673s
user 0m17.569s
sys 0m0.118sNow that https://bugs.freedesktop.org/show_bug.cgi?id=91525 looks like there could be a john the ripper bug (hanging the GPU) and a second one (failing self test), should I create one or two new bugs, and keep issue 1571 for https://bugs.freedesktop.org/show_bug.cgi?id=91524 and https://bugs.freedesktop.org/show_bug.cgi?id=92081 instead?
frank-dittrich
commented
8 years ago $ ./john --list=build-info
Version: 1.8.0.6-jumbo-1-1790-g8d4470f+
Build: linux-gnu 64-bit AVX2-ac
SIMD: AVX2, interleaving: MD4:3 MD5:3 SHA1:1 SHA256:1 SHA512:1
$JOHN is ./
Format interface version: 13
Max. number of reported tunable costs: 3
Rec file version: REC4
Charset file version: CHR3
CHARSET_MIN: 1 (0x01)
CHARSET_MAX: 255 (0xff)
CHARSET_LENGTH: 24
SALT_HASH_SIZE: 1048576
Max. Markov mode level: 400
Max. Markov mode password length: 30
gcc version: 5.3.1
GNU libc version: 2.21 (loaded: 2.21)
OpenCL headers version: 1.2
Crypto library: OpenSSL
OpenSSL library version: 0100010bf
OpenSSL 1.0.1k-fips 8 Jan 2015
GMP library version: 6.0.0
Regex library version: 1.2 (loaded: 1.2.3)
File locking: fcntl()
fseek(): fseek
ftell(): ftell
fopen(): fopen
memmem(): System's
(bleeding-jumbo)run $ git diff
diff --git a/src/bt.c b/src/bt.c
index 8a658c1..0194c46 100644
--- a/src/bt.c
+++ b/src/bt.c
@@ -623,7 +623,7 @@ unsigned int create_perfect_hash_table(int htype, void *loaded_hashes_ptr,
if (sigaction(SIGALRM, &new_action, NULL) < 0)
bt_error("Error setting new signal handler.");
- if (setitimer(ITIMER_REAL, NULL, &old_it) < 0)
+ if (getitimer(ITIMER_REAL, &old_it) < 0)
bt_error("Error retriving timer info.");
inc_ht = 0.005;Just to get an idea about the performance, here a --test (without =0) and without LWS and GWS:
$ ./john --test --format=office2013-opencl --verbosity=5
initUnicode(UNICODE, ASCII/ASCII)
ASCII -> ASCII -> ASCII
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
Benchmarking: office2013-opencl, MS Office 2013 (100,000 iterations) [SHA512 OpenCL 2x AES]... Loaded 5 hashes with 5 different salts to test db from test vectors
Options used: -I ./kernels -cl-mad-enable -D__GPU__ -DDEVICE_INFO=34 -DSIZEOF_SIZE_T=8 -DDEV_VER_MAJOR=1 -DDEV_VER_MINOR=2 -D_OPENCL_COMPILER -DHASH_LOOPS=100 -DUNICODE_LENGTH=96 -DV_WIDTH=2 $JOHN/kernels/office2013_kernel.cl
Calculating best GWS for LWS=8; max. 5s total for crypt_all()
Raw speed figures including buffer transfers:
xfer: 12.875ms*, xfer: 12.878ms*, init: 1.454ms, loop: 1000x3.607ms, final: 6.533ms, xfer: 6.692ms*
gws: 128 70c/s 7000280 rounds/s 3.648s per crypt_all()!
xfer: 12.066ms*, xfer: 12.057ms*, init: 2.157ms, loop: 1000x3.325ms, final: 6.345ms, xfer: 6.440ms*
gws: 256 152c/s 15200608 rounds/s 3.364s per crypt_all()!
xfer: 25.251ms*, xfer: 25.253ms*, init: 4.019ms, loop: 1000x6.549ms, final: 12.789ms, xfer: 14.532ms*
gws: 512 154c/s 15400616 rounds/s 6.631s per crypt_all() - too slow
Calculating best LWS for GWS=256
Testing LWS=8 GWS=256 ... 16.895ms+
Testing LWS=16 GWS=256 ... 17.606ms
Testing LWS=32 GWS=256 ... 17.716ms
Testing LWS=64 GWS=256 ... 17.535ms
Testing LWS=128 GWS=256 ... 17.200ms
Testing LWS=256 GWS=256 ... 16.827ms+
Calculating best GWS for LWS=256; max. 10s total for crypt_all()
Raw speed figures including buffer transfers:
xfer: 76.066ms*, xfer: 75.797ms*, init: 8.015ms, loop: 1000x20.641ms, final: 41.258ms, xfer: 46.837ms*
gws: 5120 490c/s 49001960 rounds/s 20.890s per crypt_all()!
xfer: 144.864ms*, xfer: 144.340ms*, init: 15.394ms, loop: 1000x41.793ms, final: 81.045ms, xfer: 85.929ms*
gws: 10240 484c/s 48401936 rounds/s 42.266s per crypt_all() - too slow
Local worksize (LWS) 256, global worksize (GWS) 5120
FAILED (cmp_all(1))
What in these walls of text suggests to you that the bug is in JtR?
frank-dittrich
commented
8 years ago with latest bleeding-jumbo commit ca11872eaf094b0dbe90ba3f74fae5366d2b3125, Linux kernel 4.4.0-0.rc6.git1.1.vanilla.knurd.1.fc22.x86_64 and latest beignet commit f74980864a41daa5d644a033a16132768ce89296, all 53 opencl formats pass self test.
(bleeding-jumbo)run $ ./john --test=10 --format=opencl --verbosity=3
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
Benchmarking: sha1crypt-opencl, (NetBSD) [PBKDF1-SHA1 OpenCL 4x]... DONE
Speed for cost 1 (iteration count) of 64000 and 40000
Raw: 117 c/s real, 18285 c/s virtual
Benchmarking: oldoffice-opencl, MS Office <= 2003 [MD5/SHA1 RC4 OpenCL]... DONE
Speed for cost 1 (hash type) of 1 and 0
Warning: "Many salts" test limited: 96/256
Many salts: 1566K c/s real, 120989K c/s virtual
Only one salt: 1473K c/s real, 23405K c/s virtual
Benchmarking: PBKDF2-HMAC-MD4-opencl [PBKDF2-MD4 OpenCL 4x]... DONE
Speed for cost 1 (iterations) of 1000
Raw: 108503 c/s real, 16384K c/s virtual
Benchmarking: PBKDF2-HMAC-MD5-opencl [PBKDF2-MD5 OpenCL 4x]... DONE
Speed for cost 1 (iterations) of 1000
Raw: 73251 c/s real, 32768K c/s virtual
Benchmarking: PBKDF2-HMAC-SHA1-opencl [PBKDF2-SHA1 OpenCL 4x]... DONE
Speed for cost 1 (iterations) of 1000
Raw: 30453 c/s real, 16384K c/s virtual
Benchmarking: rar-opencl, RAR3 (length 5) [SHA1 OpenCL AES]... DONE
Raw: 605 c/s real, 106666 c/s virtual
Benchmarking: RAR5-opencl [PBKDF2-SHA256 OpenCL]... DONE
Speed for cost 1 (iteration count) of 32768
Raw: 396 c/s real, 128000 c/s virtual
Benchmarking: truecrypt-opencl, TrueCrypt AES256_XTS [RIPEMD160 OpenCL]... DONE
Raw: 1547 c/s real, 277333 c/s virtual
Benchmarking: lotus5-opencl, Lotus Notes/Domino 5 [OpenCL]... DONE
Raw: 1032K c/s real, 1224K c/s virtual
Benchmarking: agilekeychain-opencl, 1Password Agile Keychain [PBKDF2-SHA1 OpenCL AES]... DONE
Speed for cost 1 (iteration count) of 1000
Raw: 31177 c/s real, 840205 c/s virtual
Benchmarking: bcrypt-opencl ("$2a$05", 32 iterations) [Blowfish OpenCL]... DONE
Speed for cost 1 (iteration count) of 32
Raw: 141 c/s real, 13381 c/s virtual
Benchmarking: blockchain-opencl, blockchain My Wallet [PBKDF2-SHA1 OpenCL AES]... DONE
Raw: 473526 c/s real, 809752 c/s virtual
Benchmarking: md5crypt-opencl, crypt(3) $1$ [MD5 OpenCL]... DONE
Raw: 117491 c/s real, 29696K c/s virtual
Benchmarking: sha256crypt-opencl, crypt(3) $5$ (rounds=5000) [SHA256 OpenCL]... DONE
Speed for cost 1 (iteration count) of 5000
Raw: 860 c/s real, 60235 c/s virtual
Benchmarking: sha512crypt-opencl, crypt(3) $6$ (rounds=5000) [SHA512 OpenCL]... DONE
Speed for cost 1 (iteration count) of 5000
Raw: 156 c/s real, 28444 c/s virtual
Benchmarking: descrypt-opencl, traditional crypt(3) [DES OpenCL]... DONE
Warning: "Many salts" test limited: 17/256
Many salts: 1694K c/s real, 594193K c/s virtual
Only one salt: 1640K c/s real, 50840K c/s virtual
Benchmarking: dmg-opencl, Apple DMG [PBKDF2-SHA1 OpenCL 3DES/AES]... DONE
Speed for cost 1 (iteration count) of 1000
Raw: 7937 c/s real, 15814 c/s virtual
Benchmarking: encfs-opencl, EncFS [PBKDF2-SHA1 OpenCL 4x AES/Blowfish]... DONE
Speed for cost 1 (iteration count) of 181474 and 181317
Raw: 64.0 c/s real, 7314 c/s virtual
Benchmarking: gpg-opencl, OpenPGP / GnuPG Secret Key [SHA1 OpenCL]... DONE
Speed for cost 1 (s2k-count) of 65536, cost 2 (hash algorithm [1:MD5 2:SHA1 3:RIPEMD160 8:SHA256 9:SHA384 10:SHA512 11:SHA224]) of 2, cost 3 (cipher algorithm [1:IDEA 2:3DES 3:CAST5 4:Blowfish 7:AES128 8:AES192 9:AES256]) of 3
Raw: 40715 c/s real, 315076 c/s virtual
Benchmarking: iwork-opencl, Apple iWork '09 / '13 / '14 [PBKDF2-SHA1 OpenCL 4x AES]... DONE
Speed for cost 1 (iteration count) of 100000
Raw: 60.1 c/s real, 7111 c/s virtual
Benchmarking: keychain-opencl, Mac OS X Keychain [PBKDF2-SHA1 OpenCL 3DES]... DONE
Raw: 15029 c/s real, 216338 c/s virtual
Benchmarking: keyring-opencl, GNOME Keyring [SHA256 OpenCL AES]... DONE
Speed for cost 1 (iteration count) of 3221 and 2439
Raw: 3704 c/s real, 195368 c/s virtual
Benchmarking: krb5pa-md5-opencl, Kerberos 5 AS-REQ Pre-Auth etype 23 [MD4 HMAC-MD5 RC4 OpenCL]... DONE
Warning: "Many salts" test limited: 82/256
Many salts: 2684K c/s real, 92654K c/s virtual
Only one salt: 2317K c/s real, 29449K c/s virtual
Benchmarking: krb5pa-sha1-opencl, Kerberos 5 AS-REQ Pre-Auth etype 17/18 [PBKDF2-SHA1 OpenCL 4x]... DONE
Raw: 2243 c/s real, 93090 c/s virtual
Benchmarking: LM-opencl [DES BS OpenCL]...
DONE
Raw: 14085K c/s real, 100663K c/s virtual
Benchmarking: mscash-opencl, M$ Cache Hash [MD4 OpenCL]... DONE
Many salts: 5417K c/s real, 216924K c/s virtual
Only one salt: 4517K c/s real, 24311K c/s virtual
Benchmarking: mscash2-opencl, MS Cache Hash 2 (DCC2) [PBKDF2-SHA1 OpenCL]...
DONE
Raw: 3114 c/s real, 3114 c/s virtual
Benchmarking: mysql-sha1-opencl, MySQL 4.1+ [SHA1 OpenCL (inefficient, development use only)]... DONE
Raw: 4059K c/s real, 22573K c/s virtual
Benchmarking: Raw-SHA512-ng-opencl [SHA512 OpenCL]... DONE
Raw: 1683K c/s real, 28398K c/s virtual
Benchmarking: XSHA512-ng-opencl, Mac OS X 10.7 salted [SHA512 OpenCL]... DONE
Warning: "Many salts" test limited: 28/256
Many salts: 1816K c/s real, 305834K c/s virtual
Only one salt: 1731K c/s real, 32172K c/s virtual
Benchmarking: nt-opencl, NT [MD4 OpenCL]... DONE
Raw: 2670K c/s real, 31243K c/s virtual
Benchmarking: ntlmv2-opencl, NTLMv2 C/R [MD4 HMAC-MD5 OpenCL 4x]... DONE
Many salts: 15435K c/s real, 386662K c/s virtual
Only one salt: 4271K c/s real, 30229K c/s virtual
Benchmarking: o5logon-opencl, Oracle O5LOGON protocol [SHA1 OpenCL AES 32/64]... DONE
Raw: 2396K c/s real, 3662K c/s virtual
Benchmarking: ODF-opencl [SHA1 OpenCL Blowfish]... DONE
Raw: 13373 c/s real, 23774 c/s virtual
Benchmarking: ODF-AES-opencl [SHA256 OpenCL AES]... DONE
Speed for cost 1 (iteration count) of 1
Raw: 11409 c/s real, 46900 c/s virtual
Benchmarking: office2007-opencl, MS Office 2007 (50,000 iterations) [SHA1 OpenCL 4x AES]... DONE
Raw: 1159 c/s real, 107789 c/s virtual
Benchmarking: office2010-opencl, MS Office 2010 (100,000 iterations) [SHA1 OpenCL 4x AES]... DONE
Speed for cost 1 (iteration count) of 100000
Raw: 350 c/s real, 18285 c/s virtual
Benchmarking: PBKDF2-HMAC-SHA256-opencl, rounds=1000 [PBKDF2-SHA256 OpenCL]... DONE
Speed for cost 1 (iteration count) of 1000
Raw: 13550 c/s real, 4778K c/s virtual
Benchmarking: phpass-opencl ($P$9 lengths 0 to 15) [MD5 OpenCL]... DONE
Raw: 36753 c/s real, 5266K c/s virtual
Benchmarking: pwsafe-opencl, Password Safe [SHA256 OpenCL]... DONE
Speed for cost 1 (iteration count) of 2048
Raw: 12928 c/s real, 275063 c/s virtual
Benchmarking: RAKP-opencl, IPMI 2.0 RAKP (RMCP+) [HMAC-SHA1 OpenCL 4x]... DONE
Warning: "Many salts" test limited: 178/256
Many salts: 2904K c/s real, 112167K c/s virtual
Only one salt: 2657K c/s real, 24958K c/s virtual
Benchmarking: Raw-MD4-opencl [MD4 OpenCL]...
DONE
Raw: 4364K c/s real, 35805K c/s virtual
Benchmarking: Raw-MD5-opencl [MD5 OpenCL]...
DONE
Raw: 4473K c/s real, 51826K c/s virtual
Benchmarking: Raw-SHA1-opencl [SHA1 OpenCL]...
DONE
Raw: 4088K c/s real, 35246K c/s virtual
Benchmarking: Raw-SHA256-opencl [SHA256 OpenCL]... DONE
Raw: 4779K c/s real, 28476K c/s virtual
Benchmarking: Raw-SHA512-opencl [SHA512 OpenCL (inefficient, development use mostly)]... DONE
Raw: 1427K c/s real, 14869K c/s virtual
Benchmarking: salted-sha1-opencl [SHA1 OpenCL]...
DONE
Warning: "Many salts" test limited: 58/256
Many salts: 6051K c/s real, 1216M c/s virtual
Only one salt: 5227K c/s real, 39125K c/s virtual
Benchmarking: 7z-opencl, 7-Zip (512K iterations) [SHA256 OPENCL AES]... DONE
Speed for cost 1 (iteration count) of 524288
Warning: "Many salts" test limited: 1/256
Many salts: 125 c/s real, 42666 c/s virtual
Only one salt: 125 c/s real, 42666 c/s virtual
Benchmarking: strip-opencl, STRIP Password Manager [PBKDF2-SHA1 OpenCL]... DONE
Raw: 3559 c/s real, 325818 c/s virtual
Benchmarking: sxc-opencl, StarOffice .sxc [PBKDF2-SHA1 OpenCL Blowfish]... DONE
Speed for cost 1 (iteration count) of 0
Raw: 13038 c/s real, 23942 c/s virtual
Benchmarking: wpapsk-opencl, WPA/WPA2 PSK [PBKDF2-SHA1 OpenCL 4x]... DONE
Raw: 2268 c/s real, 180705 c/s virtual
Benchmarking: XSHA512-opencl, Mac OS X 10.7+ [SHA512 OpenCL (efficient at "many salts" only)]... DONE
Many salts: 1497K c/s real, 34863K c/s virtual
Only one salt: 1477K c/s real, 25941K c/s virtual
Benchmarking: zip-opencl, ZIP [PBKDF2-SHA1 OpenCL AES]... DONE
Raw: 10315 c/s real, 5632K c/s virtual
All 53 formats passed self-tests!Cool! Thanks for your efforts with this.
frank-dittrich
commented
8 years ago Unfortunately, meanwhile I managed to reproduce the GPU hang again.
I wanted to close those beignet bugs and, after restarting my system, I ran each individual command which caused one of those bugs in the past.
Some bugs seem to have been fixed, but some still remain. May be I was just lucky with those tests where "All 53 formats passed self-tests!"
Before doing a full --test=10 run for --format=opencl, I tested each individual opencl format with --test=0, redirecting sdtout and stderr into files, and copying dmesg output into files. No problems there (at least with a --disable-openmp build). May be testing all those formats in a certain sequence somehow initialized the GPU state or memory in a certain way, and running certain format tests after shutting down and booting the system just triggers those bugs. It also looks like after a GPU hang, the GPU isn't reset correctly, because afterwards I get
(bleeding-jumbo)run $ ./john --test=0 --format=office2013-opencl --verbosity=3
Will run 4 OpenMP threads
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
Testing: office2013-opencl, MS Office 2013 (100,000 iterations) [SHA512 OpenCL 2x AES]... (4xOMP) FAILED (cmp_all(1))instead of the GPU hang
(bleeding-jumbo)run $ ./john --test=0 --format=office2013-opencl --verbosity=3
Will run 4 OpenMP threads
Device 0: Intel(R) HD Graphics Haswell GT2 Desktop
Testing: office2013-opencl, MS Office 2013 (100,000 iterations) [SHA512 OpenCL 2x AES]... (4xOMP) drm_intel_gem_bo_context_exec() failed: Input/output error
OpenCL CL_OUT_OF_RESOURCES error in opencl_office2013_fmt_plug.c:323 - failed in clEnqueueNDRangeKernelInstead of reporting this as a beignet bug, I now picked DRI as the product and DRM/Intel as the component, as suggested by the dmesg output
[ 1791.185004] [drm] stuck on render ring
[ 1791.186261] [drm] GPU HANG: ecode 7:0:0x85ddfffc, in john [2277], reason: Ring hung, action: reset
[ 1791.186265] [drm] GPU hangs can indicate a bug anywhere in the entire gfx stack, including userspace.
[ 1791.186268] [drm] Please file a _new_ bug report on bugs.freedesktop.org against DRI -> DRM/Intel
[ 1791.186270] [drm] drm/i915 developers can then reassign to the right component if it's not a kernel issue.
[ 1791.186273] [drm] The gpu crash dump is required to analyze gpu hangs, so please always attach it.
[ 1791.186275] [drm] GPU crash dump saved to /sys/class/drm/card0/error
[ 1791.188446] drm/i915: Resetting chip after gpu hanghttps://bugs.freedesktop.org/show_bug.cgi?id=93579
Let's just keep this issue closed.
I'll open new issues depending on the response to my freedesktop.org bug reports.
BTW some of the opencl format self tests seem to write '\n' to stderr (look at the additional new lines in the --test --verbosity=3 output.
magnumripper
commented
8 years ago BTW some of the opencl format self tests seem to write '\n' to stderr (look at the additional new lines in the --test --verbosity=3 output.
You had to filter out the "spinning wheel" from the output, right? I suspect it's just something with that.
frank-dittrich
commented
8 years ago Now I know why the complete --test --format=opencl run didn't fail. I forgot to check my local config. for disabled formats
office2013-opencl = Y
PBKDF2-HMAC-SHA512-opencl = YBut I think it is still better to keep this issue closed. I'll open new issues for the remaining problems.
During my experiments to get the Haswell GPU working, I probably found some beignet / libdrm / i915 driver issues and may be some bleeding-jumbo issues.
On my 64bit Fedora 22 with Haswell CPU Intel(R) Core(TM) i5-4570 CPU @ 3.20GHz, after installing beignet (I tried both the fedora package and the latest http://anongit.freedesktop.org/git/beignet.git commit):
Here, I am not sure whether the
multiple unsequenced modifications to 'p'warning indicates a real problem:The only statements that may trigger that
multiple unsequenced modifications to 'p'warning IMHO areCan any of the above be a problem?
I hope that these warnings
will disappear after a future libdrm or linux kernel upgrade and that the
problem will be fixed in future beignet versions.
May be when I get rid of the
disable atomic in L3 featuredue to the failed self test and when I disable the debug CFLAGS from my beignet build, the performance will improve...For now, I just use OCL_IGNORE_SELF_TEST=1, otherwise the Haswell GPU will be skipped:
Most opencl formats pass self test, with c/s rate between 10% and 90% of the CPU (4*OMP). For rar, opencl gets closest:
BTW: why is the CPU format using length 4, but the GPU length 5?
And these are all the failing self tests:
The common warnings
which appeared for every single test have been filtered using
grep -v.@magnum I wasn't sure whether I should create separate issues for every possible problem (failing format self tests etc.)
May be beignet, libdrm or the i915 driver are to blame for many of these. If you think I should have created multiple issues, I suggest that we keep this one for the
multiple unsequenced modifications to 'p'warning in oldoffice-opencl (and may be the failing oldoffice-opencl self test), and I'll create new issues for the other failing self tests.