jxu

These are simplified examples I came up with to go along with William Woodruff's excellent blog post on x86-64 addressing modes.

All the examples are compiled with GCC -O3 or -Os, which is indicative of what the compiler thinks the best option is in practice.

Base + Index

Use case: Indexing into a byte array (whose address is not fixed)

char f(char buf[], int index)

jxu

#include <vector>
#include <cstdint>
#include <iostream>
#include <cassert>
#include <bit>


struct popcnt_array
{
    uint32_t len; // length of array in 64-bit words

jxu

library(tidyverse)

# find duplicates by col
df %>% filter(n() > 1, .by = "col")

# drop duplicates by col
df %>% distinct(col)

# summarize per group by
mtcars %>% 

jxu

#include <vector>
#include <iostream>

struct fenwick_tree {
    size_t len; // 0-based len
    std::vector<long long> t; // 1-based tree, indexes [1:len]

    fenwick_tree(std::vector<long long> const &a) 
    {
        len = a.size();

jxu

x86-32 cdecl

Arguments pushed on stack, right-to-left.

Caller-saved: eax, ecx, edx. All others are callee-saved.

Return value in eax.

General purpose registers:

1. EAX (Accumulator)

jxu

Kipling
Stirling
Viking
bidding
biding
biffing
biking
bilking
billing
binding

jxu

# Horizontal bar chart, ordered by value
ggplot(df, aes(x = reorder(name, value), y = value)) + 
  geom_bar(stat = "identity") + 
  labs(title = "Title", x = "Name", y = "Value") +
  scale_y_continuous(expand = c(0,0)) + 
  coord_flip() + 
  theme_light()

jxu

import pandas as pd

### Data import ###

# https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.to_csv.html 
df = pd.read_csv("file.csv")  


### Data types ###

jxu

grep -Pi '^(?=.*u)[bluntam]{4,}$' /usr/share/dict/words

jxu

-- Syntax order
-- https://learn.microsoft.com/en-us/sql/t-sql/queries/select-transact-sql

WITH cte AS (cte_query)
SELECT cols
FROM table 
WHERE cond
GROUP BY col
HAVING cond
WINDOW window_expr