arno/hashcat
1
0
Fork 0
mirror of https://github.com/hashcat/hashcat.git synced 2024-12-04 13:58:22 +00:00
Code Issues Releases Wiki Activity
08a35772a9
hashcat/src/Makefile

207 lines
7.0 KiB
Makefile
Raw Normal View History

Initial commit
2015-12-04 14:47:52 +00:00
##
## Author......: Jens Steube <jens.steube@gmail.com>
## License.....: MIT
##
##
Revert "Remove NPROCS from Makefile, make is able to automatically detect the optimal number of parallel threads" Problem here is; This process requires alot of memory. If host memory is too small it will segfault somewhere inside AMD's OpenCL runtime library. Therefore it's better to stick to NPROCS as part of a workaround. This reverts commit e5adccbf3804d58cb4beb8f06797c0ba179043c6.
2015-12-06 12:09:17 +00:00
## Detect number of processors
Initial commit
2015-12-04 14:47:52 +00:00
##
Revert "Remove NPROCS from Makefile, make is able to automatically detect the optimal number of parallel threads" Problem here is; This process requires alot of memory. If host memory is too small it will segfault somewhere inside AMD's OpenCL runtime library. Therefore it's better to stick to NPROCS as part of a workaround. This reverts commit e5adccbf3804d58cb4beb8f06797c0ba179043c6.
2015-12-06 12:09:17 +00:00
NPROCS := $(shell grep -c ^processor /proc/cpuinfo)
Initial commit
2015-12-04 14:47:52 +00:00
OS := $(shell uname)
##
## Makefile flags
##
Revert "Remove NPROCS from Makefile, make is able to automatically detect the optimal number of parallel threads" Problem here is; This process requires alot of memory. If host memory is too small it will segfault somewhere inside AMD's OpenCL runtime library. Therefore it's better to stick to NPROCS as part of a workaround. This reverts commit e5adccbf3804d58cb4beb8f06797c0ba179043c6.
2015-12-06 12:09:17 +00:00
MAKEFLAGS += -l -j $(NPROCS) -rR --no-print-directory
Initial commit
2015-12-04 14:47:52 +00:00
ifneq ($(findstring clean,$(MAKECMDGOALS)),)
MAKEFLAGS += -j 1
endif
##
Prepare reordering or files to help integration into linux distributions as discussed in https://github.com/hashcat/oclHashcat/issues/20 TODOS: - Let oclHashcat actually use the new paths - Find a better way for native compilation - Replace /bin/cp with /usr/bin/install where it has to copy files recursive
2015-12-13 11:21:36 +00:00
## Installation paths (Linux only)
##
INSTALL_FOLDER ?= /opt/test/usr/bin
SHARED_FOLDER ?= /opt/test/usr/share/oclHashcat
DOCUMENT_FOLDER ?= /opt/test/usr/share/doc/oclHashcat
#INSTALL_FOLDER ?= /usr/bin
#SHARED_FOLDER ?= /usr/share/oclHashcat
#DOCUMENT_FOLDER ?= /usr/share/doc/oclHashcat
##
## Library paths
Initial commit
2015-12-04 14:47:52 +00:00
##
issue #16: use just "deps" instead of "hashcat-deps" as directory name
2015-12-09 14:55:06 +00:00
CUDA := deps/cuda-7.5
CUDALIBPATH32 := deps/NVIDIA-Linux-x86_64-352.21/32
CUDALIBPATH64 := deps/NVIDIA-Linux-x86_64-352.21
Initial commit
2015-12-04 14:47:52 +00:00
- Dropped all vector code since new GPU's are all scalar, makes the code much easier - Some performance on low-end GPU may drop because of that, but only for a few hash-modes - Dropped scalar code (aka warp) since we do not have any vector datatypes anymore - Renamed C++ overloading functions memcat32_9 -> memcat_c32_w4x4_a3x4 - Still need to fix kernels to new function names, needs to be done manually - Temperature Management needs to be rewritten partially because of conflicting datatypes names - Added code to create different codepaths for NV on AMD in runtime in host (see data.vendor_id) - Added code to create different codepaths for NV on AMD in runtime in kernels (see IS_NV and IS_AMD) - First tests working for -m 0, for example - Great performance increases in general for NV so far - Tested amp_* and markov_* kernel - Migrated special NV optimizations for rule processor
2015-12-15 11:04:22 +00:00
OCL := $(CUDA)
OCLLIBPATH32 := $(CUDALIBPATH32)
OCLLIBPATH64 := $(CUDALIBPATH64)
issue #16: use just "deps" instead of "hashcat-deps" as directory name
2015-12-09 14:55:06 +00:00
ADL := deps/adl-sdk
Initial commit
2015-12-04 14:47:52 +00:00
issue #16: use just "deps" instead of "hashcat-deps" as directory name
2015-12-09 14:55:06 +00:00
GDK := deps/nvidia-gdk
NVAPI := deps/R352-developer
Initial commit
2015-12-04 14:47:52 +00:00
NVMLINCPATH := $(GDK)/usr/include/nvidia/gdk/
NVMLLIBPATH32 := $(CUDALIBPATH32)
NVMLLIBPATH64 := $(CUDALIBPATH64)
CLCOMPILE_PATH := tools/clcompile
RULES_OPTIMIZE_PATH := tools/rules_optimize
##
## Compiler paths
##
CC_LINUX_32 := gcc
CC_LINUX_64 := gcc
CC_WIN_32 := i686-w64-mingw32-gcc
CC_WIN_64 := x86_64-w64-mingw32-gcc
DLL_WIN_32 := i686-w64-mingw32-dlltool
DLL_WIN_64 := x86_64-w64-mingw32-dlltool
CLCOMPILE := $(CLCOMPILE_PATH)/clcompile.bin
BIN := .
Prepare reordering or files to help integration into linux distributions as discussed in https://github.com/hashcat/oclHashcat/issues/20 TODOS: - Let oclHashcat actually use the new paths - Find a better way for native compilation - Replace /bin/cp with /usr/bin/install where it has to copy files recursive
2015-12-13 11:21:36 +00:00
INSTALL := install
CP := cp
Initial commit
2015-12-04 14:47:52 +00:00
##
## Misc stuff
##
NOW := $(shell perl -e 'print time')
##
## Compiler flags
##
CFLAGS := -O2 -s -pipe -W -Wall -Iinclude/ -std=c99
#CFLAGS := -g -pipe -W -Wall -Iinclude/ -std=c99
CFLAGS_32 := -m32
CFLAGS_64 := -m64
CFLAGS_LINUX := -D_POSIX -DLINUX
CFLAGS_WIN := -D_WIN -DWIN -D__MSVCRT__ -D__USE_MINGW_ANSI_STDIO=1
- Dropped all vector code since new GPU's are all scalar, makes the code much easier - Some performance on low-end GPU may drop because of that, but only for a few hash-modes - Dropped scalar code (aka warp) since we do not have any vector datatypes anymore - Renamed C++ overloading functions memcat32_9 -> memcat_c32_w4x4_a3x4 - Still need to fix kernels to new function names, needs to be done manually - Temperature Management needs to be rewritten partially because of conflicting datatypes names - Added code to create different codepaths for NV on AMD in runtime in host (see data.vendor_id) - Added code to create different codepaths for NV on AMD in runtime in kernels (see IS_NV and IS_AMD) - First tests working for -m 0, for example - Great performance increases in general for NV so far - Tested amp_* and markov_* kernel - Migrated special NV optimizations for rule processor
2015-12-15 11:04:22 +00:00
CFLAGS_LINUX += -I$(OCL)/include/ -I$(ADL)/include/ -I$(GDK)/usr/include/nvidia/gdk/
CFLAGS_WIN += -I$(OCL)/include/ -I$(ADL)/include/ -I$(NVAPI)/
Initial commit
2015-12-04 14:47:52 +00:00
change in library loading: libnvidia-ml now does load dynamically only on NVidia systems
2015-12-15 19:34:07 +00:00
LFLAGS_LINUX := -lpthread -lOpenCL -ldl
- Dropped all vector code since new GPU's are all scalar, makes the code much easier - Some performance on low-end GPU may drop because of that, but only for a few hash-modes - Dropped scalar code (aka warp) since we do not have any vector datatypes anymore - Renamed C++ overloading functions memcat32_9 -> memcat_c32_w4x4_a3x4 - Still need to fix kernels to new function names, needs to be done manually - Temperature Management needs to be rewritten partially because of conflicting datatypes names - Added code to create different codepaths for NV on AMD in runtime in host (see data.vendor_id) - Added code to create different codepaths for NV on AMD in runtime in kernels (see IS_NV and IS_AMD) - First tests working for -m 0, for example - Great performance increases in general for NV so far - Tested amp_* and markov_* kernel - Migrated special NV optimizations for rule processor
2015-12-15 11:04:22 +00:00
LFLAGS_WIN := -lpsapi -L./lib
Initial commit
2015-12-04 14:47:52 +00:00
##
## Targets: Global
##
- Dropped all vector code since new GPU's are all scalar, makes the code much easier - Some performance on low-end GPU may drop because of that, but only for a few hash-modes - Dropped scalar code (aka warp) since we do not have any vector datatypes anymore - Renamed C++ overloading functions memcat32_9 -> memcat_c32_w4x4_a3x4 - Still need to fix kernels to new function names, needs to be done manually - Temperature Management needs to be rewritten partially because of conflicting datatypes names - Added code to create different codepaths for NV on AMD in runtime in host (see data.vendor_id) - Added code to create different codepaths for NV on AMD in runtime in kernels (see IS_NV and IS_AMD) - First tests working for -m 0, for example - Great performance increases in general for NV so far - Tested amp_* and markov_* kernel - Migrated special NV optimizations for rule processor
2015-12-15 11:04:22 +00:00
all: binaries_all
Initial commit
2015-12-04 14:47:52 +00:00
binaries_all: linux32 linux64 win32 win64 rules_optimize
release: binaries
linux: linux32 linux64 rules_optimize
windows: win32 win64 rules_optimize
clean: clean_bin
- Dropped all vector code since new GPU's are all scalar, makes the code much easier - Some performance on low-end GPU may drop because of that, but only for a few hash-modes - Dropped scalar code (aka warp) since we do not have any vector datatypes anymore - Renamed C++ overloading functions memcat32_9 -> memcat_c32_w4x4_a3x4 - Still need to fix kernels to new function names, needs to be done manually - Temperature Management needs to be rewritten partially because of conflicting datatypes names - Added code to create different codepaths for NV on AMD in runtime in host (see data.vendor_id) - Added code to create different codepaths for NV on AMD in runtime in kernels (see IS_NV and IS_AMD) - First tests working for -m 0, for example - Great performance increases in general for NV so far - Tested amp_* and markov_* kernel - Migrated special NV optimizations for rule processor
2015-12-15 11:04:22 +00:00
clean_all: clean_bin clean_clcompile
Initial commit
2015-12-04 14:47:52 +00:00
clean_bin:
rm -f obj/*.o lib/*.a $(BIN)/*.bin $(BIN)/*.exe $(BIN)/*.app *.restore *.out *.pot *.dictstat *.log
rm -rf *Hashcat.induct
rm -rf *Hashcat.outfiles
$(MAKE) -C $(RULES_OPTIMIZE_PATH) clean
clean_clcompile:
rm -f ${CLCOMPILE_PATH}/clcompile.bin
${CLCOMPILE}: ${CLCOMPILE_PATH}/clcompile.c
- Dropped all vector code since new GPU's are all scalar, makes the code much easier - Some performance on low-end GPU may drop because of that, but only for a few hash-modes - Dropped scalar code (aka warp) since we do not have any vector datatypes anymore - Renamed C++ overloading functions memcat32_9 -> memcat_c32_w4x4_a3x4 - Still need to fix kernels to new function names, needs to be done manually - Temperature Management needs to be rewritten partially because of conflicting datatypes names - Added code to create different codepaths for NV on AMD in runtime in host (see data.vendor_id) - Added code to create different codepaths for NV on AMD in runtime in kernels (see IS_NV and IS_AMD) - First tests working for -m 0, for example - Great performance increases in general for NV so far - Tested amp_* and markov_* kernel - Migrated special NV optimizations for rule processor
2015-12-15 11:04:22 +00:00
$(CC_LINUX_64) $(CFLAGS) -o $@ -I$(CFLAGS_LINUX) $< -L$(OCLLIBPATH64) $(LFLAGS_LINUX)
Initial commit
2015-12-04 14:47:52 +00:00
rules_optimize:
$(MAKE) -C $(RULES_OPTIMIZE_PATH)
- Dropped all vector code since new GPU's are all scalar, makes the code much easier - Some performance on low-end GPU may drop because of that, but only for a few hash-modes - Dropped scalar code (aka warp) since we do not have any vector datatypes anymore - Renamed C++ overloading functions memcat32_9 -> memcat_c32_w4x4_a3x4 - Still need to fix kernels to new function names, needs to be done manually - Temperature Management needs to be rewritten partially because of conflicting datatypes names - Added code to create different codepaths for NV on AMD in runtime in host (see data.vendor_id) - Added code to create different codepaths for NV on AMD in runtime in kernels (see IS_NV and IS_AMD) - First tests working for -m 0, for example - Great performance increases in general for NV so far - Tested amp_* and markov_* kernel - Migrated special NV optimizations for rule processor
2015-12-15 11:04:22 +00:00
linux32: oclHashcat32.bin
linux64: oclHashcat64.bin
Initial commit
2015-12-04 14:47:52 +00:00
- Dropped all vector code since new GPU's are all scalar, makes the code much easier - Some performance on low-end GPU may drop because of that, but only for a few hash-modes - Dropped scalar code (aka warp) since we do not have any vector datatypes anymore - Renamed C++ overloading functions memcat32_9 -> memcat_c32_w4x4_a3x4 - Still need to fix kernels to new function names, needs to be done manually - Temperature Management needs to be rewritten partially because of conflicting datatypes names - Added code to create different codepaths for NV on AMD in runtime in host (see data.vendor_id) - Added code to create different codepaths for NV on AMD in runtime in kernels (see IS_NV and IS_AMD) - First tests working for -m 0, for example - Great performance increases in general for NV so far - Tested amp_* and markov_* kernel - Migrated special NV optimizations for rule processor
2015-12-15 11:04:22 +00:00
win32: oclHashcat32.exe
win64: oclHashcat64.exe
Initial commit
2015-12-04 14:47:52 +00:00
Prepare reordering or files to help integration into linux distributions as discussed in https://github.com/hashcat/oclHashcat/issues/20 TODOS: - Let oclHashcat actually use the new paths - Find a better way for native compilation - Replace /bin/cp with /usr/bin/install where it has to copy files recursive
2015-12-13 11:21:36 +00:00
##
## Targets: Linux install
## We need some better native compilation logic here, fixing it to 64 bit for now and AMD kernels
## How to make /usr/bin/install doing recursive??
##
install: linux64 amd_all
$(INSTALL) -m 755 -d $(DOCUMENT_FOLDER)
$(CP) -a docs/* $(DOCUMENT_FOLDER)/
$(INSTALL) -m 755 -d $(DOCUMENT_FOLDER)/extra
$(CP) -a extra/* $(DOCUMENT_FOLDER)/extra/
$(INSTALL) -m 755 -d $(SHARED_FOLDER)
$(INSTALL) -m 755 -d $(SHARED_FOLDER)/charsets
$(CP) -a charsets/* $(SHARED_FOLDER)/charsets/
$(INSTALL) -m 755 -d $(SHARED_FOLDER)/masks
$(CP) -a masks/* $(SHARED_FOLDER)/masks/
$(INSTALL) -m 755 -d $(SHARED_FOLDER)/rules
$(CP) -a rules/* $(SHARED_FOLDER)/rules/
$(INSTALL) -m 755 -T hashcat.hcstat $(SHARED_FOLDER)/hashcat.hcstat
$(INSTALL) -m 755 -s -T oclHashcat64.bin $(INSTALL_FOLDER)/oclHashcat
uninstall:
$(RM) -f $(INSTALL_FOLDER)/oclHashcat
$(RM) -rf $(SHARED_FOLDER)
$(RM) -rf $(DOCUMENT_FOLDER)
Initial commit
2015-12-04 14:47:52 +00:00
##
## Targets: Libraries
##
lib/libOpenCL.a:
${DLL_WIN_32} -A -k -l lib/libOpenCL.a -d lib/OpenCL.def
lib/libOpenCL64.a:
${DLL_WIN_64} -A -k -l lib/libOpenCL64.a -d lib/OpenCL64.def
##
## oclHashcat
##
obj/%.oclHashcat.LINUX.32.o: src/%.c
- Dropped all vector code since new GPU's are all scalar, makes the code much easier - Some performance on low-end GPU may drop because of that, but only for a few hash-modes - Dropped scalar code (aka warp) since we do not have any vector datatypes anymore - Renamed C++ overloading functions memcat32_9 -> memcat_c32_w4x4_a3x4 - Still need to fix kernels to new function names, needs to be done manually - Temperature Management needs to be rewritten partially because of conflicting datatypes names - Added code to create different codepaths for NV on AMD in runtime in host (see data.vendor_id) - Added code to create different codepaths for NV on AMD in runtime in kernels (see IS_NV and IS_AMD) - First tests working for -m 0, for example - Great performance increases in general for NV so far - Tested amp_* and markov_* kernel - Migrated special NV optimizations for rule processor
2015-12-15 11:04:22 +00:00
$(CC_LINUX_32) $(CFLAGS) $(CFLAGS_LINUX) $(CFLAGS_32) $(CFLAGS_LINUX) -c -o $@ $<
Initial commit
2015-12-04 14:47:52 +00:00
obj/%.oclHashcat.LINUX.64.o: src/%.c
- Dropped all vector code since new GPU's are all scalar, makes the code much easier - Some performance on low-end GPU may drop because of that, but only for a few hash-modes - Dropped scalar code (aka warp) since we do not have any vector datatypes anymore - Renamed C++ overloading functions memcat32_9 -> memcat_c32_w4x4_a3x4 - Still need to fix kernels to new function names, needs to be done manually - Temperature Management needs to be rewritten partially because of conflicting datatypes names - Added code to create different codepaths for NV on AMD in runtime in host (see data.vendor_id) - Added code to create different codepaths for NV on AMD in runtime in kernels (see IS_NV and IS_AMD) - First tests working for -m 0, for example - Great performance increases in general for NV so far - Tested amp_* and markov_* kernel - Migrated special NV optimizations for rule processor
2015-12-15 11:04:22 +00:00
$(CC_LINUX_64) $(CFLAGS) $(CFLAGS_LINUX) $(CFLAGS_64) $(CFLAGS_LINUX) -c -o $@ $<
Initial commit
2015-12-04 14:47:52 +00:00
obj/%.oclHashcat.WIN.32.o: src/%.c
- Dropped all vector code since new GPU's are all scalar, makes the code much easier - Some performance on low-end GPU may drop because of that, but only for a few hash-modes - Dropped scalar code (aka warp) since we do not have any vector datatypes anymore - Renamed C++ overloading functions memcat32_9 -> memcat_c32_w4x4_a3x4 - Still need to fix kernels to new function names, needs to be done manually - Temperature Management needs to be rewritten partially because of conflicting datatypes names - Added code to create different codepaths for NV on AMD in runtime in host (see data.vendor_id) - Added code to create different codepaths for NV on AMD in runtime in kernels (see IS_NV and IS_AMD) - First tests working for -m 0, for example - Great performance increases in general for NV so far - Tested amp_* and markov_* kernel - Migrated special NV optimizations for rule processor
2015-12-15 11:04:22 +00:00
$(CC_WIN_32) $(CFLAGS) $(CFLAGS_WIN) $(CFLAGS_32) $(CFLAGS_WIN) -c -o $@ $<
Initial commit
2015-12-04 14:47:52 +00:00
obj/%.oclHashcat.WIN.64.o: src/%.c
- Dropped all vector code since new GPU's are all scalar, makes the code much easier - Some performance on low-end GPU may drop because of that, but only for a few hash-modes - Dropped scalar code (aka warp) since we do not have any vector datatypes anymore - Renamed C++ overloading functions memcat32_9 -> memcat_c32_w4x4_a3x4 - Still need to fix kernels to new function names, needs to be done manually - Temperature Management needs to be rewritten partially because of conflicting datatypes names - Added code to create different codepaths for NV on AMD in runtime in host (see data.vendor_id) - Added code to create different codepaths for NV on AMD in runtime in kernels (see IS_NV and IS_AMD) - First tests working for -m 0, for example - Great performance increases in general for NV so far - Tested amp_* and markov_* kernel - Migrated special NV optimizations for rule processor
2015-12-15 11:04:22 +00:00
$(CC_WIN_64) $(CFLAGS) $(CFLAGS_WIN) $(CFLAGS_64) $(CFLAGS_WIN) -c -o $@ $<
Initial commit
2015-12-04 14:47:52 +00:00
- Dropped all vector code since new GPU's are all scalar, makes the code much easier - Some performance on low-end GPU may drop because of that, but only for a few hash-modes - Dropped scalar code (aka warp) since we do not have any vector datatypes anymore - Renamed C++ overloading functions memcat32_9 -> memcat_c32_w4x4_a3x4 - Still need to fix kernels to new function names, needs to be done manually - Temperature Management needs to be rewritten partially because of conflicting datatypes names - Added code to create different codepaths for NV on AMD in runtime in host (see data.vendor_id) - Added code to create different codepaths for NV on AMD in runtime in kernels (see IS_NV and IS_AMD) - First tests working for -m 0, for example - Great performance increases in general for NV so far - Tested amp_* and markov_* kernel - Migrated special NV optimizations for rule processor
2015-12-15 11:04:22 +00:00
oclHashcat32.bin: src/oclHashcat.c obj/ext_OpenCL.oclHashcat.LINUX.32.o obj/ext_nvml.oclHashcat.LINUX.32.o obj/ext_ADL.oclHashcat.LINUX.32.o obj/shared.oclHashcat.LINUX.32.o obj/rp_gpu_on_cpu.oclHashcat.LINUX.32.o
$(CC_LINUX_32) $(CFLAGS) $(CFLAGS_LINUX) $(CFLAGS_32) $(CFLAGS_LINUX) -o $@ -DCOMPTIME=$(NOW) $^ -L$(OCLLIBPATH32) $(LFLAGS_LINUX) -L$(NVMLLIBPATH32) -DINSTALL_FOLDER=\"$(INSTALL_FOLDER)\" -DSHARED_FOLDER=\"$(SHARED_FOLDER)\" -DDOCUMENT_FOLDER=\"$(DOCUMENT_FOLDER)\"
Initial commit
2015-12-04 14:47:52 +00:00
- Dropped all vector code since new GPU's are all scalar, makes the code much easier - Some performance on low-end GPU may drop because of that, but only for a few hash-modes - Dropped scalar code (aka warp) since we do not have any vector datatypes anymore - Renamed C++ overloading functions memcat32_9 -> memcat_c32_w4x4_a3x4 - Still need to fix kernels to new function names, needs to be done manually - Temperature Management needs to be rewritten partially because of conflicting datatypes names - Added code to create different codepaths for NV on AMD in runtime in host (see data.vendor_id) - Added code to create different codepaths for NV on AMD in runtime in kernels (see IS_NV and IS_AMD) - First tests working for -m 0, for example - Great performance increases in general for NV so far - Tested amp_* and markov_* kernel - Migrated special NV optimizations for rule processor
2015-12-15 11:04:22 +00:00
oclHashcat64.bin: src/oclHashcat.c obj/ext_OpenCL.oclHashcat.LINUX.64.o obj/ext_nvml.oclHashcat.LINUX.64.o obj/ext_ADL.oclHashcat.LINUX.64.o obj/shared.oclHashcat.LINUX.64.o obj/rp_gpu_on_cpu.oclHashcat.LINUX.64.o
$(CC_LINUX_64) $(CFLAGS) $(CFLAGS_LINUX) $(CFLAGS_64) $(CFLAGS_LINUX) -o $@ -DCOMPTIME=$(NOW) $^ -L$(OCLLIBPATH64) $(LFLAGS_LINUX) -L$(NVMLLIBPATH64) -DINSTALL_FOLDER=\"$(INSTALL_FOLDER)\" -DSHARED_FOLDER=\"$(SHARED_FOLDER)\" -DDOCUMENT_FOLDER=\"$(DOCUMENT_FOLDER)\"
Initial commit
2015-12-04 14:47:52 +00:00
- Dropped all vector code since new GPU's are all scalar, makes the code much easier - Some performance on low-end GPU may drop because of that, but only for a few hash-modes - Dropped scalar code (aka warp) since we do not have any vector datatypes anymore - Renamed C++ overloading functions memcat32_9 -> memcat_c32_w4x4_a3x4 - Still need to fix kernels to new function names, needs to be done manually - Temperature Management needs to be rewritten partially because of conflicting datatypes names - Added code to create different codepaths for NV on AMD in runtime in host (see data.vendor_id) - Added code to create different codepaths for NV on AMD in runtime in kernels (see IS_NV and IS_AMD) - First tests working for -m 0, for example - Great performance increases in general for NV so far - Tested amp_* and markov_* kernel - Migrated special NV optimizations for rule processor
2015-12-15 11:04:22 +00:00
oclHashcat32.exe: src/oclHashcat.c obj/ext_OpenCL.oclHashcat.WIN.32.o obj/ext_nvapi.oclHashcat.WIN.32.o obj/ext_ADL.oclHashcat.WIN.32.o obj/shared.oclHashcat.WIN.32.o obj/rp_gpu_on_cpu.oclHashcat.WIN.32.o lib/libOpenCL.a /usr/i686-w64-mingw32/lib/CRT_glob.o
$(CC_WIN_32) $(CFLAGS) $(CFLAGS_WIN) $(CFLAGS_32) $(CFLAGS_WIN) -o $@ -DCOMPTIME=$(NOW) $^ -L$(OCLLIBPATH32) $(LFLAGS_WIN) -static-libgcc $(NVAPI)/x86/nvapi.lib
Initial commit
2015-12-04 14:47:52 +00:00
- Dropped all vector code since new GPU's are all scalar, makes the code much easier - Some performance on low-end GPU may drop because of that, but only for a few hash-modes - Dropped scalar code (aka warp) since we do not have any vector datatypes anymore - Renamed C++ overloading functions memcat32_9 -> memcat_c32_w4x4_a3x4 - Still need to fix kernels to new function names, needs to be done manually - Temperature Management needs to be rewritten partially because of conflicting datatypes names - Added code to create different codepaths for NV on AMD in runtime in host (see data.vendor_id) - Added code to create different codepaths for NV on AMD in runtime in kernels (see IS_NV and IS_AMD) - First tests working for -m 0, for example - Great performance increases in general for NV so far - Tested amp_* and markov_* kernel - Migrated special NV optimizations for rule processor
2015-12-15 11:04:22 +00:00
oclHashcat64.exe: src/oclHashcat.c obj/ext_OpenCL.oclHashcat.WIN.64.o obj/ext_nvapi.oclHashcat.WIN.64.o obj/ext_ADL.oclHashcat.WIN.64.o obj/shared.oclHashcat.WIN.64.o obj/rp_gpu_on_cpu.oclHashcat.WIN.64.o lib/libOpenCL64.a /usr/x86_64-w64-mingw32/lib/CRT_glob.o
$(CC_WIN_64) $(CFLAGS) $(CFLAGS_WIN) $(CFLAGS_64) $(CFLAGS_WIN) -o $@ -DCOMPTIME=$(NOW) $^ -L$(OCLLIBPATH64) $(LFLAGS_WIN) -static-libgcc $(NVAPI)/amd64/nvapi64.lib
Reference in New Issue Copy Permalink