jsteube
edb5028a4d
Fix for PR 69122fb554 which wasn't complete
2016-06-30 12:03:12 +02:00
Jens Steube
8fc1306b0b
Fix https://github.com/hashcat/hashcat/issues/395
2016-06-30 08:43:10 +02:00
Moritz Kick
69122fb554
Make linux-build POSIX compatible
...
Also allow it to actually compile on musl-libc systems.
From glibc's termio.h:
/* Compatible <termio.h> for old `struct termio' ioctl interface.
This is obsolete; use the POSIX.1 `struct termios' interface
defined in <termios.h> instead. */
2016-06-29 18:23:08 +02:00
jsteube
6be99f523c
Limit scrypt to 64 threads has no performance impact but makes it easier to handle
2016-06-28 11:47:32 +02:00
jsteube
8702d0e3e1
Workaround memory allocation limit from OpenCL by using multiple buffers for scrypt
2016-06-28 11:03:04 +02:00
jsteube
4ed8977e48
Dynamic tmp_t length for scrypt
2016-06-27 21:28:48 +02:00
Jens Steube
37bf9d65b0
Do our own include-file permission check before an OpenCL runtime fails loading them and erroring out with a confusing error messages like "file not found"
2016-06-27 19:13:56 +02:00
Jens Steube
3370820109
Merge pull request #392 from unix-ninja/master
...
Various refactor changes
2016-06-27 18:51:00 +02:00
Unix-Ninja
2188a01ae9
Refactor OSX preprocessor constants to read 'DARWIN'
2016-06-27 08:13:50 -04:00
jsteube
115c95eb02
Use dedicated variables to control shutdown of parallel threads like keypress and monitoring
2016-06-27 12:13:46 +02:00
Jens Steube
cf9c184cd5
Set exhausted status was too late
2016-06-27 09:25:34 +02:00
Jens Steube
772e2ac33f
Fix benchmark devices_status bug
2016-06-27 08:54:45 +02:00
Jens Steube
ed1863c263
Move macros DGST_R0 - DGST_R3 to host, define dgst_size for opencl kernel from host; both at runtime
2016-06-26 23:39:42 +02:00
jsteube
0e68b2af25
Don't use device_processor_cores for scrypt
2016-06-26 22:59:52 +02:00
jsteube
6191e8bf79
Fix reset to automatic for NvAPI
2016-06-26 20:41:11 +02:00
jsteube
49811cfee1
Set fanspeed to automatic on shutdown
...
Increase retain temperature from 65c to 75c; Because GTX1080 hits its reference boost clocks of 1733mhz at 80c
2016-06-26 20:18:46 +02:00
jsteube
78c1d45b09
Disable use of nvmlDeviceSetComputeMode() and nvmlDeviceSetGpuOperationMode() - no advantage of using it
...
Fix https://github.com/hashcat/oclHashcat/issues/391
2016-06-26 16:59:56 +02:00
jsteube
de1d1da454
Disable some of NVMLs useless throttle reasons
2016-06-26 16:29:51 +02:00
jsteube
a5baa829c5
Move thread_monitor() to outer threads
2016-06-26 16:09:05 +02:00
jsteube
5c338de519
Fix race condition in thread_keypress
2016-06-26 15:55:23 +02:00
jsteube
057909fdfd
Do not overwrite quit status when user forced it with keypress
2016-06-26 11:15:29 +02:00
jsteube
c081d3cf05
Fix a bug related to --limit
2016-06-26 10:01:50 +02:00
jsteube
3a275a95e6
Fix some devices_status handling
2016-06-26 09:47:02 +02:00
jsteube
6ab712d247
Add some small section with basic examples to --help; some device status fixes
2016-06-25 21:56:18 +02:00
Jens Steube
a5a38620d4
Move keyspress detecting thread to outer loop
2016-06-25 12:05:16 +02:00
Jens Steube
e7df9b112a
windows path update
2016-06-25 10:37:24 +02:00
philsmd
ffd90774fc
do not ignore keypress / status changes
2016-06-25 09:55:39 +02:00
Unix-Ninja
fce2363fec
Clean end-of-line output usage (not actually POSIX)
2016-06-24 18:08:02 -04:00
Unix-Ninja
78a819417b
Add -D short flag for --opencl-device-types
2016-06-24 17:05:27 -04:00
Jens Steube
1550b9de07
OSX didn't know about MAX_PATH, had to include limits.h; Use heap not stack memory
2016-06-24 21:40:11 +02:00
Jens Steube
6d2aa559a6
Windows doesn't accept escape characters, falling back to old method for windows
2016-06-24 21:29:23 +02:00
Jens Steube
268b1e3ad0
Remove workaround with multiple include folders; Instead use a naive escape function
2016-06-24 16:57:20 +02:00
philsmd
79d02466af
cosmetic fix: don't show prompt and warning within the same line
2016-06-24 12:55:54 +02:00
Jens Steube
8298cd4926
Drop -m 190, no reason to keep it
2016-06-23 18:20:55 +02:00
Jens Steube
c172da1bd2
Evil bug fixed caused by variable-name collision
2016-06-22 11:58:00 +02:00
Jens Steube
fe18f51406
Fix invalid error CL_INVALID_VALUE caused by a zero length copy when using large hashlists
2016-06-22 11:32:40 +02:00
_NSAKEY
26dada5a19
Capitalization.
2016-06-21 06:11:00 -05:00
_NSAKEY
c713196d8b
Another trivial log message edit.
2016-06-21 06:09:27 -05:00
_NSAKEY
2504b59eea
Trivial log message touch-ups.
2016-06-21 06:05:07 -05:00
jsteube
39042d39e8
Fix a race condition in output file handling
2016-06-21 12:25:07 +02:00
Jens Steube
5a2f0bb5be
Remove benchmark info; Better align status display if more than 9 devices
2016-06-21 10:39:23 +02:00
radix
5d41b93514
Update wording on some warning messages.
2016-06-19 13:42:22 -05:00
dropdead
30aebe81fa
Cleaned log output
2016-06-19 16:36:41 +02:00
jsteube
6b2b11fdc0
Increase --stdout output performance
2016-06-19 15:58:52 +02:00
jsteube
4444c17ec6
Speed up startup time for --stdout by using an empty kernel
2016-06-19 14:33:50 +02:00
jsteube
07acd52ed4
Some optical change on startup
2016-06-19 12:55:03 +02:00
jsteube
d57ac912cc
Fix https://github.com/hashcat/oclHashcat/issues/374
...
Fix https://github.com/hashcat/oclHashcat/issues/375
2016-06-19 10:08:55 +02:00
jsteube
92a111be74
Fixed some macros
2016-06-18 21:56:19 +02:00
jsteube
e8aefb412c
Reenable --gpu-temp-retain (65c by default), now that we have fan set support on Linux and Windows
2016-06-18 21:29:24 +02:00
jsteube
2bd49e6720
Added --gpu-temp-retain support for windows through nvapi
2016-06-18 18:21:40 +02:00
jsteube
6a2ec331ef
Check permission to set fan speed before actually doing it, otherwise X11 becomes mad and drops the connection
2016-06-18 14:42:57 +02:00
jsteube
e43be7aba3
Do not try to set fanspeed if connection to X11 failed
2016-06-18 11:26:51 +02:00
jsteube
9a970c0955
Added support for XNVCTRL on Linux to add support for --gpu-temp-retain for NVidia GPU
2016-06-18 10:59:58 +02:00
jsteube
28505c350b
Add commandline support to override cpu spinning damper for nvidia runtime
2016-06-16 07:46:24 +02:00
Jens Steube
fd3813f279
save_hash() always works with a file, therefore no branch required
2016-06-15 12:17:18 +02:00
Jens Steube
c61798434e
Merge branch 'master' of https://github.com/hashcat/oclHashcat
2016-06-15 12:10:38 +02:00
Jens Steube
90181e565a
This patch should remove a bottleneck causing slow writing performance of hash files which can be annoying when using --remove in combination with large hashlists
2016-06-15 12:10:09 +02:00
jsteube
5542e46022
Added option --stdout to print candidates instead of trying to crack a hash
2016-06-14 21:49:33 +02:00
jsteube
eda36a9877
Make sure to not overflow
2016-06-14 10:55:03 +02:00
Jens Steube
4f0709b3c9
Another slow-hash improvement
2016-06-14 10:18:42 +02:00
jsteube
d275772b9f
Fix for 100% CPU burning loop workaround for slow hashes
2016-06-14 09:17:03 +02:00
jsteube
8abc42e346
Simple, but effective workaround to solve the 100% CPU burning loop bug in Nvidias OpenCL runtime, works Linux and Windows, 0% cpu (alsmost), 100% performance in comparison to CPU burning loop variant. Replaces old previous try with overwritten clock_gettime() function
2016-06-13 23:50:49 +02:00
jsteube
217f9e6004
Do not try to use clock_gettime() on OSX
2016-06-12 18:07:26 +02:00
jsteube
7ef80fb12b
Do not automatically enable powertune in benchmark, but allow the user to enable it manually
2016-06-12 17:51:32 +02:00
jsteube
9174767847
Inform user about how to verify reported cracking speed
2016-06-12 14:39:42 +02:00
jsteube
94e9fe7836
No need to check for cracked hashes in benchmark mode
2016-06-12 14:25:41 +02:00
jsteube
b2ff9922c9
Pure black magic to workaround 100% CPU usage bug in NVidias OpenCL runtime on Linux
2016-06-12 12:26:47 +02:00
Jens Steube
eb88171bdd
Fix precompute bug in PeopleSoft PS_TOKEN
2016-06-11 23:29:20 +02:00
Jens Steube
3022abf956
some axcrypt fixes
2016-06-11 23:08:33 +02:00
Jens Steube
0d38489c9d
Fix typo
2016-06-11 20:16:52 +02:00
jsteube
aa78cf2c74
Fix sed call in Makefile
2016-06-11 11:39:49 +02:00
jsteube
f600b9dd59
Get rid of cp in Makefile install target
2016-06-11 10:52:00 +02:00
philsmd
eb08afa543
makefile: indentation fix (replace tabs with spaces) + comments about cygwin builds
2016-06-11 08:08:22 +02:00
Jens Steube
f30629b21a
Merge pull request #369 from hlein/DESTDIR
...
Added DESTDIR support
2016-06-10 23:54:46 +02:00
Jens Steube
10b086a31d
Merge pull request #364 from neheb/patch-2
...
Add cygwin support
2016-06-10 23:51:04 +02:00
Mangix
37ff7fbc11
Small fix for < 10 Windows
2016-06-10 14:14:22 -07:00
Hank Leininger
8254e2e3c4
Make sure the INSTALL_FOLDER path exists.
...
This is only likely to come up when using DESTDIR to point to an
alternate directory tree, or if the user has pointed at some
exotic PREFIX for some reason, but should be harmless in the
normal case.
2016-06-10 15:29:10 -04:00
Hank Leininger
8a2bfff917
Added DESTDIR support.
...
This is a common thing for distros or build environments that want
to install to a sandbox location as part of testing or package-building.
It should be a harmless/no-op change when make is called without
DESTDIR defined.
2016-06-10 15:20:53 -04:00
Hank Leininger
ecb7ba6f3e
Removed obsolete entries from install: target.
...
These files have been renamed+moved under OpenCL, and are already
installed from there elsewhere in the install: target.
2016-06-10 15:10:50 -04:00
jsteube
1c695b1185
Fixed DES kernel_loops only in case of BF
2016-06-10 19:03:29 +02:00
jsteube
b9eefd9b20
Added new hash-mode 13800 = Windows 8+ phone PIN/Password
2016-06-10 09:46:41 +02:00
jsteube
b875cb51ac
Use CRT_glob.o to enable commandline globbing on windows binaries
2016-06-07 12:39:41 +02:00
jsteube
106e781a2f
Fix some workload balancing issue
2016-06-06 21:39:11 +02:00
neheb
a8bebb4a55
Add cygwin support
...
Doesn't actually build but "make win64" or "make win32" still works.
Actually, why is the operating system even checked in the makefile?
2016-06-05 13:22:38 -07:00
jsteube
9d2c24fc0d
Fix memory allocation for clBuildProgram() retry version
2016-06-04 18:20:32 +02:00
Jens Steube
4075577340
Do not show hardware manager in final status display in case status is EXHAUSTED, CRACKED, ABORTED OR QUIT -- it could be confusing for users to see an utilization of 0% at this point
2016-06-04 11:21:21 +02:00
jsteube
829289b109
Fix calling conventions for external libraries (mingw doesn't handle it alone); Fix first print of PROMPT after autotune;
2016-06-04 00:54:28 +02:00
jsteube
79c44186fc
Finally added: Temperature limit check on NVidia + Windows to notice User about performance drop. This is a NvAPI undocumented feature. Thanks to @philsmd for information gathering, an outstanding job
2016-06-03 23:10:43 +02:00
jsteube
2e706edc18
Do not show *throttle* in status display if the reason is unknown
2016-06-03 16:26:05 +02:00
jsteube
08934db58c
Add NvAPI_GPU_GetPerfPoliciesInfo() and NvAPI_GPU_GetPerfPoliciesStatus() for later use
2016-06-03 16:08:32 +02:00
jsteube
a3e6690a9c
Remove union in hm_attrs_t; Remove code related to threshold_slowdown which will be replaced with NvGetPerfPoliciesInfo()
2016-06-03 15:37:53 +02:00
jsteube
02dd1f20ab
Fix of invalid use of free()
2016-06-03 14:18:56 +02:00
jsteube
e9960709b5
Makefile fix
2016-06-03 13:05:18 +02:00
jsteube
c970ad0f83
Fix dos-windows buffer size to maximum, users won't complain if they can scroll back
2016-06-03 13:03:38 +02:00
Jens Steube
0ffa9f4b82
Do not show warnings message in case --force is used
2016-06-03 11:37:54 +02:00
jsteube
040bbd416a
More nvapi preparations and display skip reason in benchmark
2016-06-03 11:33:59 +02:00
jsteube
34bcbb3091
Prepare for reintegration of nvapi for checking thermal limit
2016-06-02 14:20:51 +02:00
jsteube
3c6296801a
Prepare for reintegration of nvapi for checking thermal limit
2016-06-02 14:20:15 +02:00
Jens Steube
7ba1322d7f
Redesign some of the kernel_power features like too small wordlist detection etc, now based on autotuning results
2016-06-02 12:32:24 +02:00
Jens Steube
725e3677cb
Only load ADL or NVML if we really need it
2016-06-02 11:10:36 +02:00
jsteube
04dfe6e89e
Fix -m 12500 by limiting max. length to 20; Limit fake rule copy in autotune() to it's max size
2016-06-01 22:56:33 +02:00
jsteube
d329451cc1
Here it is: --powertune-enable for nvidia -- Works on both linux and windows
2016-06-01 19:01:44 +02:00
jsteube
a23c0c4716
Added a ton of new NVML stuff
2016-06-01 00:57:57 +02:00
jsteube
d89a8a68d6
Finally got rid of NvAPI on Windows, replace with NVML
2016-05-31 22:48:58 +02:00
jsteube
a381cb651d
Fix autotune in case of rule-based attack is used
2016-05-31 13:21:41 +02:00
jsteube
9a548298ee
Fix some OSX compiler errors
2016-05-30 10:29:18 +02:00
jsteube
19e36694da
Added support for --gpu-temp-retain for nvidia devices on windows
...
Disabled retain support by default, you can reactive it using --gpu-temp-retain
Completely get rid of HAVE_ADL, HAVE_NVML and HAVE_NVAPI in sources
2016-05-30 00:05:46 +02:00
jsteube
8ffbeac954
Add NvAPI_GPU_GetPerfDecreaseInfo() -- but i doesn't work, so deactive it
2016-05-29 20:51:52 +02:00
jsteube
63de7cdca8
Add NvAPI support for PCIE Lanes
2016-05-29 17:43:25 +02:00
jsteube
3a49227748
Only show Hardware-Monitor information we have, do not show N/A any longer
2016-05-29 17:25:52 +02:00
jsteube
72384c1fc3
Add NVML support for PCIE Lanes
2016-05-29 16:54:07 +02:00
Jens Steube
53a4e0cbb7
Add PCI-E Lanes to status display, if available
2016-05-29 15:54:51 +02:00
Jens Steube
3e6ae89297
Add ADL_Overdrive6_TargetTemperatureData_Get() which for some reason does not work
2016-05-29 15:34:28 +02:00
jsteube
3e4badd9b4
Send a notice to user in case the drivers temperature threshold for slowdown is reached and a performance drop is expected due to throttling
2016-05-29 00:59:24 +02:00
jsteube
f91dc3ac80
Fix osx compilation
2016-05-28 22:13:46 +02:00
jsteube
1b05d09327
Remove coreclock frequency returned back from OpenCL runtime (always wrong since GTX5xx) - we now have a better reading in status display
...
Only adjust screen width on windows, not height
2016-05-28 18:05:04 +02:00
jsteube
3227ef167e
Added NvAPI support for querying current engine clock and current memory clock
2016-05-28 17:44:09 +02:00
jsteube
36def60bfa
Added NVML support for querying current engine clock and current memory clock
2016-05-28 16:49:23 +02:00
Jens Steube
e97fa06a7a
Added current engine clock and current memory clock to the status display (ADL only atm)
...
Automatically enable AMD powertune in benchmark-mode
2016-05-28 16:32:45 +02:00
jsteube
75a6e432db
Turns out default gpu retain temp of 80c is a bit too high, latest nvidia driver sets it to 70c, so we move with it
2016-05-28 16:01:26 +02:00
jsteube
bfeacfde75
Automatically increase console size on windows if it's too small
2016-05-28 14:46:54 +02:00
radix
083c90a0ab
Update macros to match option --machine-readable
2016-05-27 04:26:39 -05:00
radix
2b2a0c585e
Change --status-automate to --machine-readble. The former didn't really illustrate the point of the command though the functionallity is great to have. Updated ASCII tables around options output.
2016-05-27 03:50:52 -05:00
radix
bb332eaa71
Merge remote-tracking branch 'upstream/master'
2016-05-26 21:38:34 -05:00
radix
ed72953c8b
Change format of option headers to something less eye cancerous.
...
Fix a typo with --status-automate.
2016-05-26 21:34:46 -05:00
jsteube
911c8ae1db
Release kernel_memset when finished
2016-05-26 17:08:00 +02:00
jsteube
9f821aa3d2
Set maximum password length for SAP CODVN B (BCODE) hashes
2016-05-26 16:55:30 +02:00
jsteube
2dd8156d24
Introduce a true memset kernel, currently operates on 16 byte per item
2016-05-26 16:45:52 +02:00
jsteube
fd7bc2736d
More vendor related changes
2016-05-26 14:56:42 +02:00
jsteube
f5f9073f45
Better distinguish between platform vendor and device vendor
2016-05-26 14:28:05 +02:00
jsteube
be91174794
Detect if the user has both an Intel and and AMD OpenCL runtime and not filtered any of them then skip the AMD CPU support automatically
2016-05-26 14:01:42 +02:00
Jens Steube
964a901f2a
Allow numbers not power of 2 in autotune
2016-05-26 12:40:12 +02:00
Jens Steube
ca8bb56404
Make it a real balancing loop
2016-05-26 00:51:11 +02:00
jsteube
167d763795
Workaround some AMD device_maxmem_alloc / device_global_mem problem
2016-05-26 00:07:08 +02:00
jsteube
250dbde2a2
Add more include-dir variants to also make Intel OpenCL SDK on windows happy
2016-05-25 23:47:02 +02:00
Jens Steube
6ed35891f6
Trying to find a workaround for Issue https://github.com/hashcat/oclHashcat/issues/335
2016-05-25 23:15:53 +02:00
Jens Steube
2899f53a15
Move files from include/ to OpenCL/ if they are used within kernels
...
Rename includes in OpenCL so that it's easier to recognize them as such
2016-05-25 23:04:26 +02:00
Jens Steube
083c8ed515
Reduce use of mux_display to a minimum
2016-05-25 11:35:17 +02:00
Jens Steube
e1a9c435cc
Merge branch 'master' of https://github.com/hashcat/oclHashcat
2016-05-24 19:31:14 +02:00
Jens Steube
57e1191307
Fix bug in -a 1 mode: If left wordlist has less entries than right wordlist then 0x80 was not added
2016-05-24 19:30:55 +02:00
jsteube
35d7e67354
Fix some typos
2016-05-24 12:10:43 +02:00
jsteube
6b0a13d998
Reduce helptext width from 134 to 118; Reduce -w 1 target to 2ms
2016-05-23 21:34:50 +02:00
jsteube
d4cc8e3701
Adjust target workload profile timings slightly
2016-05-23 01:28:11 +02:00
jsteube
3a42ab3577
Fix changes
2016-05-22 22:25:15 +02:00
jsteube
73fd95ac01
Redesigned hashcat --help menu
2016-05-22 21:46:06 +02:00
jsteube
c901935843
Update autotune engine
2016-05-21 15:49:09 +02:00
jsteube
63f5ecf9d7
Mark -n and -u as outdated function and restrict them to --force
2016-05-21 12:26:44 +02:00
jsteube
4ed418f504
Fix some indents
2016-05-21 12:16:09 +02:00
jsteube
bb513afe13
Fix some indents
2016-05-21 12:13:37 +02:00
Jens Steube
c6c865e32e
Some final fixes for d_return_buf refactorization; Initial kernels vor veracrypts SHA256 KDF (not working)
2016-05-21 00:39:22 +02:00
