Joel joelagnel

## aref.md

      
              1 file
            
          
              0 forks
            
          
                0 comments
              
            
              0 stars
            
          
                joelagnel
                / aref.md
            
            
              Created
              May 11, 2025 00:43
            
              
                ARef article
              
          
    ARef: The “Abstract Reference” for Rust

I understand references, Rc and Arc. But wtf is ARef and why would I need one?
1. The problem: juggling multiple return types

Imagine you have a function that sometimes:

Returns a borrowed &str (no heap allocation)
Returns an owned String you just built
Returns a shared buffer (e.g. an Rc<String>) without copying


## t4.html
<!DOCTYPE html>
<html>
<head>
<title>RCU Sequence Visualization</title>
</head>
<body>
<div style='margin-top: 20px; margin-bottom: 20px;'>
<h3>Legend:</h3>
<ul>
<li style='background-color: #90EE90'>Green: Grace Period Completed (multiple of 4)</li>

## rcu_seq_done_4bit.html
<!DOCTYPE html>
<html>
<head>
<title>RCU Sequence Visualization</title>
</head>
<body>
<div style='margin-top: 20px; margin-bottom: 20px;'>
<h3>Legend:</h3>
<ul>
<li style='background-color: #90EE90'>Green: Grace Period Completed (multiple of 4)</li>

## RustForCpp.md

      
              1 file
            
          
              0 forks
            
          
                0 comments
              
            
              0 stars
            
          
                joelagnel
                / RustForCpp.md
            
            
              Last active
              December 14, 2024 15:37
            
              
                Rust for C and C++ programmers
              
          
    Rust for C and C++ programmers

Table of Contents


Rust is Explicit about Programmer Intentions
Explicit References When Passing Data
Ownership, Lifetimes and Cloning
Arrays vs. Vectors
Buffer Overflow Protection
Strings and the Copy Trait
Slices as Fat Pointers


## readme
Note, llvm by default does not get on PATH, so should be added.

git remote add scx
git://git.kernel.org/pub/scm/linux/kernel/git/tj/sched_ext.git
git checkout -b scx_611 scx/for-6.11
agi clang-16 llvm-16 lld-16

PATH=/usr/lib/llvm-16/bin/:$PATH LLVM=1 make -j48 x86_64_defconfig

Then run:

## ignore-stdin-for-secs.sh
#!/bin/sh

# Check if at least one numeric arg is passed
if [ $# -eq 0 ]
then
    echo "Usage: $0 <seconds>"
    exit 1
fi

# Read and discard input for 1 second

## busybox-init
#!/bin/sh
# Save this file as "init" in the root of your CPIO archive,
# and chmod +x it. This will chroot into an HDD if one is
# available. Or it will just give a prompt.

mkdir -p /proc /sys /d /pstore /dev
mount -t proc none /proc
mount -t sysfs none /sys
mount -t debugfs none /d
mount -t debugfs none /sys/kernel/debug/

## run-qemu.sh
#!/bin/bash
# Author: Joel Fernandes <joel@joelfernandes.org>
#
# Usage, run-qemu		: Runs with busybox chrooting into a debian qemu.img
# run-qemu --rt			: Set all Qemu vCPU threads as RT and pin them to different CPUs.
# run-qemu --rcu		: Run qemu with rcutorture boot parameters
#
# run-qemu --cpio 		: Runs only with busy-box initramfs (no chrooting)
# run-qemu --debian		: Boot into a debian rootfs (see end of this file for how to build debian images)
# If neither --cpio nor --debian is passed, then it will try to chroot into a qemu.img file.

## hrtimer_trace_parse.py
#!/usr/bin/python3
# Run for tracing:
# With full screen video playing on device, run:
# trace-cmd record -e timer:hrtimer_start -R stacktrace -e hrtimer_cancel -e timer:hrtimer_expire_entry -- sleep 5
# trace-cmd report > trace.txt
# Pass the trace to this script.
# The script will produce a filter_trace.txt which only shows hrtimer events that actually fired and
# did not get canceled.
# The script will also output the top-most frequent stacks that queued a timer.

## mutex_fair_test_mod.c
#include <linux/init.h>
#include <linux/module.h>
#include <linux/kernel.h>
#include <linux/mutex.h>
#include <linux/slab.h>
#include <linux/kthread.h>
#include <linux/delay.h>
#include <linux/moduleparam.h>
#include <linux/list.h>
	<!DOCTYPE html>
	<html>
	<head>
	<title>RCU Sequence Visualization</title>
	</head>
	<body>
	<div style='margin-top: 20px; margin-bottom: 20px;'>
	<h3>Legend:</h3>
	<ul>
	<li style='background-color: #90EE90'>Green: Grace Period Completed (multiple of 4)</li>
	Note, llvm by default does not get on PATH, so should be added.

	git remote add scx
	git://git.kernel.org/pub/scm/linux/kernel/git/tj/sched_ext.git
	git checkout -b scx_611 scx/for-6.11
	agi clang-16 llvm-16 lld-16

	PATH=/usr/lib/llvm-16/bin/:$PATH LLVM=1 make -j48 x86_64_defconfig

	Then run:
	#!/bin/sh

	# Check if at least one numeric arg is passed
	if [ $# -eq 0 ]
	then
	echo "Usage: $0 <seconds>"
	exit 1
	fi

	# Read and discard input for 1 second
	#!/bin/sh
	# Save this file as "init" in the root of your CPIO archive,
	# and chmod +x it. This will chroot into an HDD if one is
	# available. Or it will just give a prompt.

	mkdir -p /proc /sys /d /pstore /dev
	mount -t proc none /proc
	mount -t sysfs none /sys
	mount -t debugfs none /d
	mount -t debugfs none /sys/kernel/debug/
	#!/bin/bash
	# Author: Joel Fernandes <joel@joelfernandes.org>
	#
	# Usage, run-qemu : Runs with busybox chrooting into a debian qemu.img
	# run-qemu --rt : Set all Qemu vCPU threads as RT and pin them to different CPUs.
	# run-qemu --rcu : Run qemu with rcutorture boot parameters
	#
	# run-qemu --cpio : Runs only with busy-box initramfs (no chrooting)
	# run-qemu --debian : Boot into a debian rootfs (see end of this file for how to build debian images)
	# If neither --cpio nor --debian is passed, then it will try to chroot into a qemu.img file.
	#!/usr/bin/python3
	# Run for tracing:
	# With full screen video playing on device, run:
	# trace-cmd record -e timer:hrtimer_start -R stacktrace -e hrtimer_cancel -e timer:hrtimer_expire_entry -- sleep 5
	# trace-cmd report > trace.txt
	# Pass the trace to this script.
	# The script will produce a filter_trace.txt which only shows hrtimer events that actually fired and
	# did not get canceled.
	# The script will also output the top-most frequent stacks that queued a timer.
	#include <linux/init.h>
	#include <linux/module.h>
	#include <linux/kernel.h>
	#include <linux/mutex.h>
	#include <linux/slab.h>
	#include <linux/kthread.h>
	#include <linux/delay.h>
	#include <linux/moduleparam.h>
	#include <linux/list.h>