yalue’s gists

yalue / view_blocksbysm.py

Created November 5, 2019 21:32

A modified version of view_blocksbysm.py that takes a --min-sm and --max-sm command-line argument.

	# This script reads all JSON result files and uses matplotlib to display a
	# timeline indicating when blocks and threads from multiple jobs were run on
	# GPU, including which SM they ran on. For this to work, all result filenames
	# must end in .json.
	#
	# Usage: python view_blocksbysm.py [results directory (default: ./results)]
	import argparse
	import glob
	import json
	import math

yalue / thread_create_hook.c

Last active June 4, 2021 17:02

Example of hooking pthread_create

	// TO COMPILE:
	// gcc -shared -fPIC -o thread_create_hook.so thread_create_hook.c -ldl
	// TO USE (needs an absolute path):
	// LD_PRELOAD=`pwd`/thread_create_hook.so ./application_to_hook

	#define _GNU_SOURCE
	#include <dlfcn.h>
	#include <pthread.h>
	#include <stdio.h>
	#include <stdlib.h>

yalue / write_to_stdin.c

Last active April 23, 2018 14:11

Note to self: writing to a child process' stdin

	#include <errno.h>
	#include <stdio.h>
	#include <string.h>
	#include <unistd.h>

	// A minimal-ish example of writing data to a child process' stdin.
	int main(int argc, char **argv) {
	char print_buffer[256];
	int stdin_pipe[2];
	int result;

yalue / tester.go

Last active March 26, 2017 12:49

Go question

yalue / printf.cpp

Created October 6, 2016 15:50

How to make using C++ suck less

	void StreamPrintf(std::ostream stream, const char *format, ...) {
	char buffer[1024];
	va_list args;
	memset(buffer, 0, sizeof buffer);
	va_start(args, format);
	vsnprintf(buffer, (sizeof buffer) - 1, format, args);
	va_end(args);
	stream << buffer;
	}

yalue / hello.c

Created December 30, 2015 17:22

yalue / regex_tester.go

Created July 23, 2015 14:51

	package main

	import (
	"fmt"
	"math/rand"
	"regexp"
	"time"
	)

	func getRandomBytes(n uint32) []byte {

yalue / deadwood.rb

Last active September 27, 2025 19:42

Gin rummy optimal meld finder and deadwood calculator in ruby

	#!/usr/bin/ruby
	#
	# If you want to understand how this script works, I'd recommend starting to read it
	# at the "entry point" where it starts validating and parsing the arguments---just
	# look for the lines interacting with ARGV to get the initial list of cards. From
	# there you can follow how it generates the full set of candidate melds (which may
	# overlap) and then how it builds a "tree" where each path to a leaf node defines a
	# non-overlapping set of melds.

	usage = "deadwood.rb:

Nathan O yalue