Mahmoud Rusty Abdelkader mahmoudimus

Quick Guide For Fixing/Installing Python, PyTorch, CUDA, Triton, Sage Attention and Flash Attention For Local AI Image Generation

With all the new local image & video generation tools coming out I've been seeing a lot of posts and error threads being opened for various issues with cuda/pytorch/sage attantion/triton/flash attention. I was tired of digging links up so I initially made this as a cheat sheet for myself but expanded it with hopes that this will help some of you get your venvs and systems running smoothly.

In This Guide:

Check Installed Python Versions
Set Default Python Version by Changing PATH
Installing VS Build Tools
Check the Currently Active CUDA Version

IDA Microcode notes

References

Ilfak's presentation at Recon 2018
Microcode in pictures
Hex-Rays Microcode API vs. Obfuscating Compiler
Scripts vds10, vds11, vds12 and vds13 from Hex-Rays SDK

Implementation

Usage

In IDAPython,

execfile('<path>/cxxparser.py')
parse_file('<path>/a.cpp',[r'-I<path>\LuaJIT-2.0.5\src', '-D__NT__', '-D__X64__', '-D__EA64__'])
parse_file('<path>/malloc.c',['-target=x86_64-linux-gnu'])

	#!/usr/bin/env python3

	import os
	import platform
	import subprocess
	import time
	from concurrent.futures import ProcessPoolExecutor, as_completed

	import click
	from elftools.elf.elffile import ELFFile

	# (c) Tim Blazytko 2021
	# implementation based on the blog post "Automated Detection of Control-flow Flattening"
	# https://synthesis.to/2021/03/03/flattening_detection.html
	import sys

	from miasm.analysis.binary import Container
	from miasm.analysis.machine import Machine
	from miasm.core.locationdb import LocationDB

	# /etc/nsmb.conf - macOS 11.3 - 2021-04-29
	#------------------------------------------------------------------------------
	# SMB configuration for macOS 11.3 <-> Synology
	#------------------------------------------------------------------------------
	# Additional information:
	# -----------------------
	# https://support.apple.com/de-de/HT211927
	# https://support.apple.com/en-us/HT208209
	# https://apple.stackexchange.com/questions/309016/smb-share-deadlocks-since-high-sierra
	# https://photographylife.com/afp-vs-nfs-vs-smb-performance

	Preparation
	1 - Download lua source in https://www.lua.org/download.html
	2 - Unpack (use 7-zip)
	3 - Create an C:\src\lua
	4 - In C:\src\lua create dirs: include, lib, bin and doc
	5 - Copy 'doc' from lua.tar.gz to C:\src\lua\doc
	6 - Copy all files from lua.tar.gz src, to C:\src\lua\include

	Creating a .lib and .dll
	1 - Create a new blank project in Visual Studio

	-------------------------------
	ODbgScript english plugin by E3
	site : http://odbgscript.sf.net
	-------------------------------

	1. About OllyScript and ODbgScript
	2. Status
	2.1 What's new?
	3. Documentation
	3.1 Language

	/**
	* @file endianness.h
	* @brief Convert Endianness of shorts, longs, long longs, regardless of architecture/OS
	*
	* Defines (without pulling in platform-specific network include headers):
	* bswap16, bswap32, bswap64, ntoh16, hton16, ntoh32 hton32, ntoh64, hton64
	*
	* Should support linux / macos / solaris / windows.
	* Supports GCC (on any platform, including embedded), MSVC2015, and clang,
	* and should support intel, solaris, and ibm compilers as well.

	// system resources physical memory map VM detection trick
	// written by Graham Sutherland (@gsuberland) for Nettitude

	// based on prior work done as part of the al-khaser project
	// https://github.com/LordNoteworthy/al-khaser/

	// ref: https://blog.xpnsec.com/total-meltdown-cve-2018-1038/
	// ref: https://gist.github.com/xpn/3792ec34d712425a5c47caf5677de5fe

	// compile: