arno/hashcat
1
0
Fork 0
mirror of https://github.com/hashcat/hashcat.git synced 2025-07-22 22:48:47 +00:00
Code Issues Releases Wiki Activity
9,950 Commits 1 Branch 27 Tags 304 MiB
d7fb2ffa06
Commit Graph

432 Commits

Author SHA1 Message Date
Jens Steube
c87a87f992 Improvements to SCRYPT autotuning strategy 
General:

The logic for calculating the SCRYPT workload has been moved
from module_extra_buffer_size() to module_extra_tuningdb_block().
Previously, this function just returned values from a static
tuning file. Now, it actually computes tuning values on the fly
based on the device's resources and SCRYPT parameters. This
was always possible, it just wasn't used that way until now.

After running the calculation, the calculated kernel_accel value
is injected into the tuning database as if it had come from a
file. The tmto value is stored internally.

Users can still override kernel-threads, kernel-accel, and
scrypt-tmto via the command line or via tuningdb file.

module_extra_tuningdb_block():

This is now where kernel_accel and tmto are automatically
calculated.

The logic for accel and tmto is now separated and more
flexible. Whether the user is using defaults, tuningdb entries, or
manual command line overrides, the code logic will try to make
smart choices based on what's actually available on the device.

First, it tries to find a kernel_accel value that fits into
available memory. It starts with a base value and simulates
tmto=1 or 2 (which is typical good on GPU).

It also leaves room for other buffers (like pws[], tmps[], etc.).
If the result is close to the actual processor count,
it gets clamped.

This value is then added to the tuning database, so hashcat can pick
it up during startup.

Once that's set, it derives tmto using available memory, thread
count, and the actual SCRYPT parameters.

module_extra_buffer_size():

This function now just returns the size of the SCRYPT B[] buffer,
based on the tmto that was already calculated.

kernel_threads:

Defaults are now set to 32 threads in most cases. On AMD GPUs,
64 threads might give a slight performance bump, but 32 is more
consistent and reliable.

For very memory-heavy algorithms (like Ethereum Wallet), it
scales down the thread count.

Here's a rough reference for other SCRYPT-based modes:

- 64 MiB: 16 threads
- 256 MiB: 4 threads

Tuning files:

All built-in tuningdb entries have been removed, because they
shouldn’t be needed anymore. But you can still add custom entries
if needed. There’s even a commented-out example in the tuningdb
file for mode 22700.

Free memory handling:

Getting the actual amount of free GPU memory is critical for
this to work right. Unfortunately, none of the common GPGPU APIs
give reliable numbers. We now query low-level interfaces like
SYSFS (AMD) and NVML (NVIDIA). Support for those APIs is in
place already, except for ADL, which still needs to be added.

Because of this, hwmon support (which handles those low-level
queries) can no longer be disabled.
 2025-06-09 11:02:34 +02:00
Jens Steube
ed6e967425 Add experimental SCRYPT N-parameter auto-discovery 
Remove existing tuningdb entries due to salsa_r() core
refactor. Update tuningdb engine to prefer file entries,
when available, over automatic discovery.

Improve memory-free detection per device, default
--backend-device-keepfree is now set to 0.

Old brute-force OpenCL behavior can be restored using
--backend-device-keepfree 100.
 2025-06-08 07:32:32 +02:00
Gabriele Gristina
12f1fe56aa OpenCL Backend: added workaround to set device_available_memory from CUDA/HIP alias device 2025-06-07 20:22:01 +02:00
Gabriele Gristina
378abf6957 Metal Backend: added workaround to set the true Processor value in Metal devices on Apple Intel 2025-06-07 17:52:34 +02:00
Jens Steube
d60658102b Added option --backend-devices-keepfree to configure X percentage of device memory available to keep free 2025-06-04 10:13:29 +02:00
Jens Steube
bf110ee10f
Merge branch 'master' into asahi-fix 2025-06-03 14:59:04 +02:00
Jens Steube
b02b1b5033 - Add code to recognize Microsofts OpenCL D3D12 platform 
- Skip memory-free detection on MS OpenCL platform to avoid crashes
- Improve salt usage of 70100/70200, use decoder/kernels from 8900
- Add REPLACE bridge type support (eg. BRIDGE_TYPE_REPLACE_LOOP)
- Switch 70000, 70100 and 70200 to BRIDGE_TYPE_REPLACE_LOOP
- Add synchronization barriers on d2h copy when using bridges
- Improve speed status display updates when using bridges
- Set AMD_DIRECT_DISPATCH=0 to reduce CPU burning loop on AMD backends
- Set benchmark/selftest hash on 70100/70200 to 16:8:1
 2025-06-02 06:59:36 +02:00
Jens Steube
22c25b3ea1 Fix cross-compiler target in makefiles 
Do not show timeout patch warnings in virtualized mode
 2025-06-01 06:39:55 +02:00
Jens Steube
ceb5ff5641 The Assimilation Bridge (Framework) 2025-05-29 15:38:13 +02:00
Jens Steube
f9822fe569
Merge pull request #4185 from matrix/deep_comp_aux_workaround 
workaround for issue #3914
 2025-05-25 08:14:23 +02:00
Gabriele Gristina
55f53ba076 assigned -H to --hash-info 2025-05-11 17:16:03 +02:00
Jens Steube
b0c7765e88 - Backport code from hipDllPath to hiprtcDllPath 
- Fix some compiler warning
- Remove some unused code
- We will add changes to docs/ and user-notifications on versioning later
 2025-05-11 14:08:04 +02:00
Jens Steube
d10ffbb461 Change amdhip64.dll installation path detection strategy 2025-05-10 12:11:40 +00:00
Gabriele Gristina
6568c5988b fix #3914 2025-04-18 14:17:21 +02:00
Code Curiously
5b26392adb Fix RAM usage bug for Linux Apple Silicon #4125 2024-12-01 19:20:26 +00:00
wizardsd
0ba76629c0 Fixed a host buffer overflow bug when copying rules from host to device 2024-11-07 12:47:36 +03:00
Mathias
d756a6617c
Update backend.c 
Fixes hiprtcCompileProgram(): HIPRTC_ERROR_COMPILATION on AMD 6900XT.

```bash
hiprtcCompileProgram(): HIPRTC_ERROR_COMPILATION

ld.lld: error: undefined hidden symbol: __ockl_get_group_id
>>> referenced by /home/mathias/.local/share/hashcat/comgr-69ec34/input/shared_kernel.o:(gpu_decompress)
>>> referenced by /home/mathias/.local/share/hashcat/comgr-69ec34/input/shared_kernel.o:(gpu_decompress)
>>> referenced by /home/mathias/.local/share/hashcat/comgr-69ec34/input/shared_kernel.o:(gpu_memset)
>>> referenced 7 more times

ld.lld: error: undefined hidden symbol: __ockl_get_local_size
>>> referenced by /home/mathias/.local/share/hashcat/comgr-69ec34/input/shared_kernel.o:(gpu_decompress)
>>> referenced by /home/mathias/.local/share/hashcat/comgr-69ec34/input/shared_kernel.o:(gpu_decompress)
>>> referenced by /home/mathias/.local/share/hashcat/comgr-69ec34/input/shared_kernel.o:(gpu_memset)
>>> referenced 7 more times

ld.lld: error: undefined hidden symbol: __ockl_get_local_id
>>> referenced by /home/mathias/.local/share/hashcat/comgr-69ec34/input/shared_kernel.o:(gpu_decompress)
>>> referenced by /home/mathias/.local/share/hashcat/comgr-69ec34/input/shared_kernel.o:(gpu_decompress)
>>> referenced by /home/mathias/.local/share/hashcat/comgr-69ec34/input/shared_kernel.o:(gpu_memset)
>>> referenced 7 more times

* Device #1: Kernel /usr/local/share/hashcat/OpenCL/shared.cl build failed.

* Device #1: Kernel /usr/local/share/hashcat/OpenCL/shared.cl build failed.
```
 2024-10-29 15:35:10 +01:00
fse-a
f8c0899670 Increased-virtual-backend-limit 
Increased the virtual backend limit.
 2024-01-25 10:27:38 +01:00
Sergey Popov
c7923fe8ba Fix RAM usage for Intel iGPUs 2023-11-24 23:03:00 +02:00
jsteube
c59b4bffd5 Add note on how to get CUDA running on WSL2 2023-09-27 08:14:51 +00:00
jsteube
0d0a07012c Do not show HIPRTC notice if HIP is not installed. In that case we can assume the system doesn't have an AMD GPU 2023-09-26 08:52:46 +00:00
Rosen Penev
ae07d65f34 clang-tidy: remove useless casts 
Now that const was fixed, pointless casts can be removed.

Signed-off-by: Rosen Penev <rosenp@gmail.com>
 2023-08-20 21:10:34 -07:00
Rosen Penev
795674c939 fix some const-qual warnings 
Signed-off-by: Rosen Penev <rosenp@gmail.com>
 2023-08-20 20:55:54 -07:00
Rosen Penev
a59e030d7f
fix MinGW compilation 
HC_API_CALL needs to be placed not just on the declaration.

Signed-off-by: Rosen Penev <rosenp@gmail.com>
 2023-08-11 23:16:27 -07:00
jsteube
96b05b2630 AMD Driver: Updated requirements for AMD Windows drivers to "AMD Adrenalin Edition" (23.7.2 or later) and "AMD HIP SDK" (23.Q3 or later) 2023-08-05 15:58:22 +00:00
justpretending
b2f14f2f5d Fix some typos 2023-07-27 23:11:55 +07:00
jsteube
ab932815ea Fixed out-of-boundary read in autotune.c when a fast hash defines a kernel_loops_min value that is higher than the actual number of amplifiers provided by the user 2023-07-04 09:40:39 +00:00
Gabriele Gristina
ef81520158 Metal Backend: disable Metal devices only if at least one OpenCL device is active 2023-06-14 21:11:42 +02:00
Jens Steube
32517211a5
Merge pull request #3760 from matrix/check_invalid_kern_type 
do not allow negative kern_type
 2023-06-14 03:25:35 +02:00
Gabriele Gristina
eb276f12f6 do not allow negative kern_type 2023-06-07 21:06:57 +02:00
Gabriele Gristina
737989b7cf Metal Backend: allow use of devices with Metal if runtime version is >= 200 2023-06-07 01:17:30 +02:00
Gabriele Gristina
36606addac Help: show supported hash-modes only with -hh 2023-05-01 17:38:42 +02:00
jsteube
6785a32e7a Added new feature (-Y) that creates N virtual instances for each device in your system at the cost of N times the device memory consumption 2023-04-16 12:04:34 +00:00
Jens Steube
b1ca2ca539 Only try to allocate memory on a opencl device if it actually has memory 2023-04-11 10:17:32 +00:00
jsteube
6ad36db9ba Add some detailed information/warnings about dropped HIPRTC symbols in the current AMD Adrenalin driver 2023-03-29 20:16:36 +00:00
Dávid Bolvanský
29a461cb51
Respect quiet flag when printing Metal API warning 
Fixes https://github.com/hashcat/hashcat/issues/3624
 2023-03-28 00:05:46 +02:00
Royce Williams
ebe8741000 link to driver FAQ on CUDA failure 2023-03-07 12:27:30 -09:00
beschio
69a945ea75
fix Metal API advice 2023-02-04 22:47:10 +01:00
jsteube
3390159bb2 AMD Driver: Updated requirements for AMD Windows drivers to "AMD Adrenalin Edition" (Adrenalin 22.5.1 exactly) 2023-01-04 08:36:37 +00:00
jsteube
c6a7826d12 Do not show metal disabled warning in backend information mode 2023-01-03 13:59:59 +00:00
jsteube
9651d3025c Fix error in backend.c if compiling on macOS 2022-12-31 16:04:35 +00:00
jsteube
b7925f1149 Add warning about automatically disabled Metal devices 2022-12-29 19:37:56 +00:00
jsteube
b228816067 Fix some missing newlines 2022-12-28 18:28:30 +00:00
Gabriele Gristina
15d74b7c1c Disable Metal < 200, commented unused and deprecated code 2022-12-08 18:59:34 +01:00
Gabriele Gristina
0f63294b8a Allow using Metal < 300 with --force 2022-12-08 17:37:50 +01:00
Jens Steube
a7dbd73612 Add Radeon Pro W5700X to more scrypt based hash-modes and fix some metal issues 2022-11-26 13:32:33 +01:00
Jens Steube
47b49d81de Prepare fixes to enable scrypt cracking on macOS 13 2022-11-24 22:46:03 +01:00
Jens Steube
18fcf28d64 Apple Driver: Automatically enable GPU support on Apple OpenCL instead of CPU support 
Apple Driver: Updated requirements to use Apple OpenCL API to macOS 13.0
 2022-11-20 14:27:16 +01:00
jsteube
9a52ee6d01 Fix error message in backend.c and add missing new hash-mode entries in changes.txt 2022-11-18 09:47:17 +00:00
Jens Steube
66b22fa644 Add support for Metal > 300 and reject support for older version 2022-11-16 14:26:54 +01:00
Jens Steube
8f0d69ea9c Always assume 8 compute unit for metal GPUs 2022-11-09 16:09:57 +01:00
Jens Steube
6d5d9a266f AMD Driver: Updated requirements for AMD Windows drivers to "AMD Adrenalin Edition" (Adrenalin 22.5.1 or later) 2022-11-05 12:56:36 +01:00
Jens Steube
b8dd806944 Revert accidental removal of check to backend_devices_cnt 2022-11-01 19:39:03 +01:00
Jens Steube
eaf3de164b Remove option "-flegacy-pass-manager" if HIP backend is used. Support was removed from LLVM 15. There's no alternative option available. 
Fixes #3469
 2022-11-01 19:36:03 +01:00
pelle
f456973928 Added error message describing workaround when detecting more than 64 backend devices. 2022-10-27 12:09:37 +02:00
Jens Steube
5aaf8e2c42 AMD Driver: Updated requirements for AMD Linux drivers to "AMDGPU" (21.50 or later) and "ROCm" (5.0 or later) 
AMD Driver: Updated requirements for AMD Windows drivers to "AMD Adrenalin Edition" (Adrenalin 22.5.1 exactly)
 2022-08-30 10:12:26 +02:00
Jens Steube
78ee811fa9 Use -flegacy-pass-manager for -m 501, -m 1600 and -m 6300 on AMD HIP for massive speed improvement 2022-08-12 16:20:14 +02:00
Jens Steube
3d898c5daf AMD Driver: Updated requirements for AMD Linux drivers to ROCm 5.2.1 or later due to new HIP interface 2022-08-10 19:19:34 +02:00
Jens Steube
a841af000f Fix updated HIP version handling 2022-04-22 13:28:33 +02:00
Jens Steube
4791cca0aa AMD Driver: Updated requirements for AMD Linux drivers to ROCm 5.1.1 or later due to new HIP interface 
AMD Driver: Updated requirements for AMD Windows drivers to Adrenalin 22.3.1 or later due to new HIP interface
 2022-04-19 13:02:51 +02:00
Jens Steube
9a8f3ebcab Update HIP version check on Linux and Windows 2022-04-18 18:07:40 +02:00
Jens Steube
c916ae821d Fix HIP include folder usage on Windows 2022-04-16 18:32:21 +02:00
philsmd
a24d723d66
minor code style fixes 2022-03-31 19:19:16 +02:00
philsmd
b51f536ac5
sanity check for u32 cast for combs_cnt, bfs_cnt 2022-03-13 14:41:34 +01:00
philsmd
5d9d3c7eb7
sanity check for u32 cast for combs_cnt, bfs_cnt 2022-03-13 14:25:09 +01:00
philsmd
cd33a8a200
fixes #2560: mention "main driver" within the error message 2022-03-09 18:15:54 +01:00
Will Crozier
edf7365cda stdout mode: transfer indexes and compressed pw buffer from device in blocks, reducing overhead 2022-02-25 16:35:55 +00:00
Jens Steube
e9cb796528
Merge pull request #3186 from willcrozi/opencl-include-quoting 
Improve handling of whitespace in kernel compiler include-path
 2022-02-24 17:50:13 +01:00
Will Crozier
919ad5a1fd warn/skip for AMD devices on Linux when OpenCL/HIP include-path has spaces 2022-02-20 20:14:57 +00:00
Giulio Garzia
c49238e9cd OpenCL include-path workaround for Windows with AMD GPU 2022-02-20 17:55:16 +00:00
Gabriele Gristina
3156b487b7 Avoid quoting OpenCL include path when it's not required to 2022-02-20 17:55:16 +00:00
Gabriele Gristina
f890cd8134 Improved management of systems with multiple OpenCL platforms and show empty OpenCL platforms only in backend information mode 2022-02-19 21:17:56 +01:00
Gabriele Gristina
24de156ce8 Fixed backend active devices checks 2022-02-19 17:21:27 +01:00
Gabriele Gristina
fb12de4be6 fix style 2022-02-16 05:40:34 +01:00
Gabriele Gristina
7eb66e9936 Hardware Monitor: Add support for GPU device utilization readings using iokit on Apple Silicon (OpenCL and Metal) 2022-02-15 22:20:03 +01:00
Gabriele Gristina
3064c63c71 User Options: Change --backend-info/-I option type, from bool to int 2022-02-13 12:33:11 +01:00
Gabriele Gristina
b3d3b31c3e Metal: added support for vectors up to 4 2022-02-10 21:53:08 +01:00
Jens Steube
be75e4b4ea Rename STR() to M2S() to avoid future collisions and move from kernel source to command line parameter 2022-02-07 09:57:08 +01:00
Gabriele Gristina
7ac879f1e4 Fixed unused variable warning on Windows 2022-02-06 02:04:27 +01:00
Gabriele Gristina
829d49c8ba resync src/backend.c 2022-02-06 01:17:19 +01:00
Gabriele Gristina
490702fcfa Backends: added Metal host-code 2022-02-05 22:48:16 +01:00
Jens Steube
8293964097 Fix coding convention 2022-02-04 10:28:45 +01:00
Gabriele Gristina
cd363b32f6
Merge branch 'master' into metal_prepare_kernelIncludes_v2 2022-01-28 19:56:26 +01:00
Gabriele Gristina
01a28f80f7 Updated handling of POCL's known bugs 2022-01-22 22:40:18 +01:00
Gabriele Gristina
a1ced24564 Fixed bug on benchmark engine, add some unstable warnings, updated negative status code 2022-01-22 12:10:09 +01:00
Gabriele Gristina
2e4a136758 Refactored standard kernel includes in order to support Apple Metal runtime, updated backend, test units and status code 2022-01-18 22:52:14 +01:00
Gabriele Gristina
7650894e02 fixed bug in benchmark engine, updated negative status code 2022-01-11 04:58:30 +01:00
Jens Steube
dfd316c653
Merge pull request #3103 from matrix/backend_session_update_mp_rl 
Removed hc_clSetKernelArg() call from backend_session_update_mp_rl()
 2022-01-09 13:22:00 +01:00
Jens Steube
7a9a1b37d0
Merge pull request #3104 from matrix/backend_session_update_mp 
Removed hc_clSetKernelArg() call from backend_session_update_mp()
 2022-01-09 13:21:52 +01:00
Jens Steube
56ef2b4bde
Merge pull request #3102 from matrix/backend_cuda_restyle 
CUDA Backend: moved functions to ext_cuda.c/ext_nvrtc.c and includes to ext_cuda.h/ext_nvrtc.h
 2022-01-06 11:29:56 +01:00
Jens Steube
045ca5cb7a Fixed method how OPTS_TYPE_AUX* kernels are called in association mode, for instance WPA/WPA2 kernels 2022-01-05 10:46:00 +01:00
Jens Steube
668d2179cd Kernels: Refactored standard kernel declaration to use a structure holding u32/u64 attributes to reduce the number of attributes 2022-01-04 18:07:18 +01:00
Gabriele Gristina
994083eaf5 Removed hc_clSetKernelArg() call from backend_session_update_mp() 2022-01-03 18:28:19 +01:00
Gabriele Gristina
0f0cf1fe08 Removed hc_clSetKernelArg() call from backend_session_update_mp_rl() 2022-01-03 17:16:27 +01:00
Gabriele Gristina
f8ceb8785e CUDA Backend: moved functions to ext_cuda.c/ext_nvrtc.c and includes to ext_cuda.h/ext_nvrtc.h 2022-01-03 16:29:15 +01:00
Gabriele Gristina
78c7ee2af6 HIP Backend: moved functions to ext_hip.c/ext_hiprtc.c and includes to ext_hip.h/ext_hiprtc.h 2022-01-02 19:12:41 +01:00
Gabriele Gristina
26b6054cab OpenCL Backend: moved functions to ext_OpenCL.c and includes to ext_OpenCL.h 2021-12-28 21:55:05 +01:00
Gabriele Gristina
861e644057 OpenCL Backend: added workaround to make optimized kernels work on Apple Silicon 2021-12-24 17:30:49 +01:00
Jens Steube
df6e5480ca Print module_extra_tuningdb_block undefined compute device warning only on GPU 2021-12-23 19:23:35 +01:00
Gabriele Gristina
3fd6dac523 Set default device-type to GPU with Apple M1 2021-12-14 00:37:17 +01:00