Yuriy Nazarov
Love machine learning
Author Archives: Yuriy Nazarov
Link assembly and c
01.06.2013 – 12:53
I using nasm to generate object code from assembly source. 3 application required to compile executable:
– nasm
– gcc from gcc-
– ld from binutils package
*package names in ubuntu distribution
To install this packages run following command:
|
1 |
sudo apt-get install nasm binutils gcc-4.7-base |
Files contents:
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 |
global func section .text func: push rbp mov rbp, rsp mov rax, 1 ; sys_write mov rdi, 1 ; stdout mov rsi, message ; message address mov rdx, length ; message string length syscall pop rbp ret section .data message: db 'Hello, world!',0x0A ; message and newline length: equ $-message ; NASM definition pseudo-instruction |
|
1 2 3 4 5 6 |
void func(); //may be extern void func(); is more true way int main(){ func(); return 0; } |
|
1 2 3 |
nasm -f elf64 func.asm gcc -c use_asm.c gcc use_asm.o func.o -o use_asm |
And finaly:
|
1 2 |
$ ./use_asm Hello, world! |
Precision of calculation with floating point numbers
25.05.2013 – 14:26
I found some curiosity issues during finding optimal tool for compilation our program. Results of computations be different for different tools and different optimisation flags.
For investigating this issue I write simple code(full sources on github):
|
1 2 3 4 5 |
int i; long double res; for(i=0;i<100000000;i++){ res += cos(i); } |
In this small test program I can’t reproduce different results for different flags. But 5 different results for so similar programs is enough to demonstrate discovered effect.
| Source code and tool | Result | Computation time, s |
| Assembler | 1.53436944477410652787 | 3.81 |
| gcc long | 1.53436944477462278158 | 4.4 |
| gcc double | 1.53436944477389380914 | 4.4 |
| icc long | 1.53436944477397574360 | 6.88 |
| icc double | 1.53436944477334602510 | 1.50 |
p.s. also I found that manually written assembly code 15% faster than code on C in spite of using “-O2” optimisation flag.
Remote SVN backup with svnsync
07.09.2012 – 01:50
Create local svn repository with name “my_local_repo”
|
1 |
$ svnadmin create my_local_repo |
prepare local repo
|
1 2 |
$ echo "#!/bin/sh" > hooks/pre-revprop-change $ chmod 755 hooks/pre-revprop-change |
initialize local copy
|
1 |
$ svnsync init file:///path/to/my_local_repo http://from.server/from_repo |
first run must be interactive to input and save your login and password
|
1 |
svnsync sync file:///path/to/my_local_repo |
write script to run it in cron
backup_svn_script.sh:
|
1 2 3 |
#!/bin/bash svnsync --non-interactive sync file:///path/to/my_local_repo tar -C /path/to/ -czf /path/to/backup/$(date +"svn_%Y_%m_%d_%H_%M_%S").tar.gz my_local_repo |
it produce files like “svn_2012_09_07_01_34_01.tar.gz” in your backup directory
Now you can add task to cron. Execute:
|
1 |
$ crontab -e |
and add following line to run backup at 4:00 (4 a.m.) every day
|
1 |
0 4 * * * /path/to/backup_svn_script.sh |
This method is not disk space efficient, but very simple.
Sources:
using svnsync
installing subversion server in ubuntu
Web server on bash
04.09.2012 – 00:17
web server on bash, nc, awk and sed
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 |
#!/bin/sh rm /tmp/f mkfifo /tmp/f #runloop while true do nc -l 8080 < /tmp/f | ( #read request read line #parse request args=$(echo -n "$line" | awk '{print $2}' | sed s~'?'~' '~ | awk '{print $2}') page=$(echo -n "$line" | awk '{print $2}' | sed s~'?'~' '~ | awk '{print $1}' | sed s~'/'~''~) user=$(echo -n "$args" | sed s~','~"\n"~ | sed s~'='~' '~ | awk '{if($1 == "user") print $2}') #some logic if [ "$page" != 'index.sh' ] then conttype="application/octet-stream" len=$(stat -c%s $page) echo "HTTP/1.0 200 OK" echo "Server: MadnessBASH" echo "Content-Length: $len" echo "Content-Type: $conttype" echo "" cat $page else conttype="text/html" if [ -z "$user" ] then data=`cat login.htm` else data=`cat index.htm` data=$(echo -n "$data" | sed s~_user_~"$user"~) fi len=`echo $data | wc -c` echo "HTTP/1.0 200 OK" echo "Server: MadnessBASH" echo "Content-Length: $len" echo "Content-Type: $conttype" echo "" echo $data fi ) > /tmp/f done |
Remote access without “white” ip
03.09.2012 – 00:19
How to get access to shell on the remote computer that don’t have “white”(“real” or “public”) IP without any additional software in linux?
Quick and dirty solution:
on server side:
|
1 2 3 4 |
mkfifo /tmp/p0.0 mkfifo /tmp/p0.1 screen sh -c 'cat < /tmp/p0.1 & cat > /tmp/p0.0' #now you can detach this from current console and attach it back when you want it |
on client side:
|
1 2 |
mkfifo /tmp/localpipe ssh -p 22 my.host.name 'cat < /tmp/p0.0 & cat > /tmp/p0.1' < /tmp/localpipe | bash > /tmp/localpipe |
pipe ‘/tmp/localpipe’ used to redirect bash output back to ssh
note: use exit command to stop redirection
