Arduino and other embedded boards

Posted on by
Reply

My list of embedded boards, I hope to keep it updated as I learn more about embedded systems options:

Very interesting interview with Monty, MySQL inventor, in FLOSS weekly

Posted on by
Reply

Don’t miss the latest episode of FLOSS weekly podcast with Monty the original author of MySQL, to learn why you should consider Maria DB instead of MySQL in your next project, some highlights from the interview:

  • He is Finnish not Swedish!! yet another Finnish changing the world!!
  • He has 2 daughters My, and Maria, and one son Max … this where “My”SQL, “Maria”DB, “Max”DB names come from.
  • Even after selling MySQL for 1 Billion dollar to Sun, he still lives in the same neighborhood.
  • Why he created MariaDB, how he feel about MySQL under Oracle
  • Still coding, at least 4 hours a day!!

Thanks Randal and Simon for great interview, and thank you Monty for great product.

Bash auto completion heaven

Posted on by
Reply

bash completion extends bash’s standard completion behavior to achieve complex command lines with just a few keystrokes. This project was conceived to produce programmable completion routines for the most common Linux/UNIX commands, reducing the amount of typing sysadmins and programmers need to do on a daily basis.

Just install:
sudo aptitude install bash-completion

Loop Unwinding Experiment

Posted on by
Reply

Few months ago I read an interesting post by Mike Haertel the original author of GNU grep titled “why GNU grep is fast“, one of the reason given is loop unrolling.
Few weeks ago I came a cross the same post which made me more interested on Loop Unrolling, so I decided to experiment with it, I implemented simple function to search for an item inside a list of items, and return as soon as it finds it, soo here is my result:

First Middle Not in Array
Normal Loop 0.387 1.495 2.605
Unrolled Loop 0.397 1.07 1.701
Improvement -2.52 39.72 53.15
O2 Normal Loop 0.045 0.279 0.373
O2 Unrolled Loop 0.045 0.217 0.285
Improvement 0.00 28.57 30.88

The result is taken using the “time” command, “O2″ means compiling the program with optimization level 2.

#include <stdio.h>

inline int item_match_fast(int item, int * items, int items_size)
{
    int i;
    int repeat, left;

    repeat = items_size / 8;
    left = items_size % 8;

    i = 0;
    while (repeat-- > 0) {
        if (item==items[i]) { return 1; }
        if (item==items[i+1]) { return 1; }
        if (item==items[i+2]) { return 1; }
        if (item==items[i+3]) { return 1; }
        if (item==items[i+4]) { return 1; }
        if (item==items[i+5]) { return 1; }
        if (item==items[i+6]) { return 1; }
        if (item==items[i+7]) { return 1; }
        i += 8;
    }

    switch(left) {
        case 7: if (item==items[i+7]) { return 1; }
        case 6: if (item==items[i+6]) { return 1; }
        case 5: if (item==items[i+5]) { return 1; }
        case 4: if (item==items[i+4]) { return 1; }
        case 3: if (item==items[i+3]) { return 1; }
        case 2: if (item==items[i+2]) { return 1; }
        case 1: if (item==items[i+1]) { return 1; }
        case 0: ; /* don nothing */
    }

    return 0;
}

inline int item_match(int item, int * items, int items_size)
{
    int i;

    /* check items array */
    for (i=0; i<items_size ; i++) {
        if(item==items[i]) {
            return 1;
        }
    }
    return 0;
}

#define REPEAT 10000000
#define ITEM 7070

int main(int argc, char ** argv)
{
    int i;
    int items[] = { 1001, 1002, 1003, 1004, 1005, 1006, 1007, 1008, 1009, 1010,
                    1011, 1012, 1013, 1014, 1015, 1016, 1017, 1018, 1019, 1020 };
    int items_size = 20;

    if (argc==2) {
        printf("Loop unrolled\n");
        for (i=0;i<REPEAT; i++) {
            item_match_fast(ITEM, items, items_size);
        }

    } else {
        printf("Normal Loop\n");
        for (i=0;i<REPEAT; i++) {
            item_match(ITEM, items, items_size);
        }
    }

    return 0;
}