compjootery

Dwarf Fortress in a VM

2011-12-24T13:54:00.000Z

My 2Ghz machine is too slow to run DF in a VM so I stopped

Debian net install into a Qemu VM on FreeBSD

NB. Aptitude is an awful program for browsing packages.

Download Dwarf Fortress for Linux & expand

./libs/Dwarf_Fortress: error while loading shared libraries: libSDL-1.2.so.0: cannot open shared object file: No such file or directory

which means

apt-get install libsdl1.2debian

./libs/Dwarf_Fortress: error while loading shared libraries: libgtk-x11-2.0.so.0: cannot open shared object file: No such file or directory

apt-get install libgtk2.0-0

./libs/Dwarf_Fortress: error while loading shared libraries: libSDL_image-1.2.so.0: cannot open shared object file: No such file or directory

apt-get install libsdl-image1.2

./libs/Dwarf_Fortress: error while loading shared libraries: libGLU.so.1: cannot open shared object file: No such file or directory

apt-get install libglu1-mesa

./libs/Dwarf_Fortress: error while loading shared libraries: libSDL_ttf-2.0.so.0: cannot open shared object file: No such file or directory

apt-get install libsdl-ttf2.0-0

and it runs without Sound

apt-get install tightvncserver

then you'll need a window manager etc.

try and pick a simple one, oh look,

My choices but not very good, I had to edit config files to get it to work

apt-get install 9wm

and a terminal

apt-get install rxvt

Pah not so easy. Windows take forever to open, BASH takes ages to start.

So I found it runs much faster (ha)if you copy df_linux to /dev/shm. I imagine it is because reads from the host filesystem require a syscall whereas reading from guest memory is in host user space.

Also use VNC raw and run it over ssh compressed so that the host does the compression work, not the Qemu

Its doing worldgen atm, taking a while. But I reckon its going to be playable via the netbook over vnc.

Standard size worldgen took 2 hours. World size "Smaller", History "Short", No. Civs "Low", No. Sites "Medium", No. Beasts "Medium", Natural Savagery "medium", Mineral Occurence "Sparse" takes about 4 minutes!

I'm a bit concerned about the 8s keypress delay!

Bookmarklet for Vworker

2011-12-02T10:30:00.001Z

You need to be signed in and on your Profile page.

The press this and it gets your rank from the HTML and changes the location to your page in the ranking list with 50 entries per page.

javascript:{window.location = "https://www.vworker.com/RentACoder/DotNet/misc/AllWorkerCompleteList.aspx?intLastRecordOnPage=50&intFirstRecordOnPage=1&txtMaxNumberOfEntriesPerPage=50&intLastRecordInRecordset=43292&cmdGoToPage=" + Math.ceil((/#[0-9]+,[0-9]+/.exec(document.getElementById('TABLE1').innerHTML)[0].replace('#', '').replace(',', '') / 50)).toString() + "&lngMaxNumberOfEntriesPerPage=50"}

Linux From Scratch Part 2

2011-11-19T18:13:00.001Z

Ok, after making a mistake I wrote a script to build the build tools for me. (sorry about the .txt my web server likes octects)I guess this is what Stage1 is in Gentoo.Edit your optionsI've got a Quad core with 8Gb RAM :>export MAKEFLAGS="-j 5"export EXTRACT=/dev/shmThe tools built for me in about 20 mins.My SBU :real 1m2.852suser 1m48.902ssys 0m44.976s

Linux From Scratch Part 1

2011-11-16T19:31:00.001Z

blogger happily deleted this massive postit was a script to download all the sources

Compilin' SSL Apachey with PHP, MySQL client

2011-11-11T20:46:00.001Z

I ended up writing a bunch o'scriptshttp://www.maht0x0r.net/build_lamp.zip

A few scripts for building a small self-contained Apache2 / PHP Installation.

PHP has Mysql + SSL and includes JSON but not XML

Scripts here for fetching the source files by version
Untarring them
and then building them all

Make a directory for putting it all under and then run 'em
e.g.
mkdir -p /opts/LAMPcd /opts/LAMPunzip /tmp/build.zip[vi build/build.vars] # this hard codes the version numbers, that's the point./build/fetch./build/build/opts/LAMP/apache2/bin/apachectl start

By default it checks for a running mysql to find the sock file when creating php.ini
docroot ends up being /opts/LAMP/www/public_html
php path is .:/opts/LAMP/www/php
default port is 801
default user:group is gopher:gopher

also has
rm.src for deleying the source trees
rm.tar for deleting the tarfiles
untar for re-extracting the tarfiles

DKIM signing with Postfix

2011-09-30T16:37:00.000+01:00

Scenario

You use SMTP-AUTH (hopefully over TLS) on your phone to connect to a Postfix server which then sends your email over the public internet. You want to add some authority to the message somehow.

Enter OpenDKIM - look there for ./make install etc.

A DKIM Signature

DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=maht0x0r.net; s=blogger; t=1317392525; 
        bh=AKfekrfl5wnhOUH0Zy1ZDQuS/dTepAwDyGPkx1cTdLk=;
        h=Content-Type:To:Subject:MIME-Version:Content-Transfer-Encoding: From:Date:Message-ID; 
        b=pJL4iPtbmXl2G+58UrJR6SeKbcoWx4sYwqUXitFs/tKvolBnSiy53D3+Mf79pYZ2L   
        5q35MwZcWEx94s4bFb88AYC44TdAhZx9hZuo/NbOvKRdUjDv028XRLJ0mOAgVR0jle  
        bCXqykNnCdduIgyO3xKyEaw+xRhwxPf74g9P2mdY=

Before signing the values of d=maht0x0r.net (domain) s=blogger (section) and the private key chosen are under your control based on the sender's email address.

I've got OpenDKIM running as a service under daemontools which is what I use for long running servers - anyway you need it listening on it's socket

% cat /etc/opendkim.conf # the important bitsCanonicalization        relaxed/simpleKeyTable        file:/etc/postfix/ssl/opendkim/KeyTableSigningTable    refile:/etc/postfix/ssl/opendkim/SigningTableExternalIgnoreList refile:/etc/postfix/ssl/opendkim/trusted-hostsInternalHosts refile:/etc/postfix/ssl/opendkim/trusted-hostsSocket                  inet:8891@localhost

Make some keys

% opendkim-genkey -d maht0x0r.net -s blogger% ls -l --rw------- M 0 root root 887 Sep 30 13:26 blogger.private--rw------- M 0 root root 320 Sep 30 13:26 blogger.txt% cat blogger.private # used to sign outgoing email-----BEGIN RSA PRIVATE KEY-----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-----END RSA PRIVATE KEY-----% cat blogger.txt # used by receivers to check your signateblogger._domainkey IN TXT "v=DKIM1; r=postmaster; g=*; k=rsa;        p=MIGfMA0GCSqGSIb3DQEBAQUAA4GNADCBiQKBgQDEfFH3nxvFnwck2oJCq7JojSPsfAvr93+N        +M5cykwFruPgJXsWEIcNymwKyEABwS7kL78eg1SYggh/8RI1jZiXfLW+PQFZLrwqE9/L0ZzZhjrjYQjKWM182        /Uaj6GWtOhariSLKmR6JYtRPFh/fiOr6YDDAt/puFSLCnUi3Ppj8QIDAQAB" ; ----- DKIM blogger for maht0x0r.net

r=postmaster (DKIM errors to $r@$domain) # something like that :)Then tell opendkim about your keys, here you name them :This is where keyname is used to determine d, s and the private key

% cat /etc/postfix/ssl/opendkim/KeyTable# keyname  d s keyfilemaht	maht0x0r.net:blogger:/etc/postfix/ssl/opendkim/blogger.privatemaht0x0r maht0x0r.net:blogger:/etc/postfix/ssl/opendkim/maht0x0r.net.privatedefault example.com:flex:/etc/postfix/ssl/opendkim/smtp.private

And in here you match email addresses to named keys % cat /etc/postfix/ssl/opendkim/SigningTablemaht@maht0x0r.net maht*@maht0x0r.net maht0x0r* default First match wins in the SigningTable, which took me a while to realise when I had * as the first entry.OpenDKIM by default will only sign mails posted from inside the network. But we can control this

# ExternalIgnoreList refile:/etc/postfix/ssl/opendkim/trusted-hosts# InternalHosts refile:/etc/postfix/ssl/opendkim/trusted-hosts% echo '*' > /etc/postfix/ssl/opendkim/trusted-hosts

That's right, we just sign for anyone but only SASL-Authed clients can be sending so that means only SASL-Auth'd clients get a signature.OK, then you need a DKIM entry in your DNS

% host -t txt blogger._domainkey.maht0x0r.netblogger._domainkey.maht0x0r.net descriptive text "v=DKIM1\; r=maht-dkim\; g=*\; k=rsa\;         p=MIGfMA0GCSqGSIb3DQEBAQUAA4GNADCBiQKBgQDEfFH3nxvFnwck2oJCq7JojSPsfAvr93+N+        M5cykwFruPgJXsWEIcNymwKyEABwS7kL78eg1SYggh/8RI1jZiXfLW+PQFZLrwqE9/L0ZzZhjrjYQjKWM182        /Uaj6GWtOhariSLKmR6JYtRPFh/fiOr6YDDAt/puFSLCnUi3Ppj8QIDAQAB"

I tried to verify it with a few online DKIM testing suites, all failed.This is a header when I send it to google :

Authentication-Results: mx.google.com; spf=pass (google.com: domain of maht@maht0x0r.net          designates x.x.x.x as permitted sender) smtp.mail=maht@maht0x0r.net;          dkim=pass header.i=@maht0x0r.net

I didn't use the keys above, obviously, so my DNS /public key is different, which you can verify :

host -t txt flex._domainkey.maht0x0r.netflex._domainkey.maht0x0r.net descriptive text "v=DKIM1\; r=maht-dkim\; g=*\; k=rsa\;         p=MIGfMA0GCSqGSIb3DQEBAQUAA4GNADCBiQKBgQDeQqlOKodABflhh076+6yueVDiQAm5koXBEg        QG31F04Al01HeyYxM7HllBjvrDfuQei102SyZH298QDDXmk/4Wtb8zh2SZ5xammqSR89HkKmbqFaSHD        2iEAgiEeWGf7nL/lK2NM0QljP5aOUpjOW1SGfRiB1AtipAcKNPLm1NRIQIDAQAB"

Next job is to find out how to sign it manually without using an OpenDKIM process forever listening.Then I can add my own headers *before* signing it.

SSL Encryption in PHP Without HTTPS

2011-09-22T21:07:00.002+01:00

This all started because the EU decided that one can't use cookies without authorisation, so how to hand state around a website without client side storage.

I investigated some things and played with SSL.

It works like this :

plain -> encryption with public key -> cipher

cipher -> decryption with private key -> plain

plain -> encryption with private key -> cipher

cipher -> decryption with public key -> plain

Using this mechanism one can envisage a system where two entities can communicate by first exchanging public keys and then encrypted data that only the recipient can decrypt.

First off we're going to need some keys. These are generated using the openssl command line tools.

% openssl genrsa -out private.pem 1024% openssl rsa -in private.pem -out public.pem -outform PEM -pubout

If you're feeling fruity you can keep them both in the same file, remembering not to send it when you mean to just send the public part !

% cat private.pem public.pem | tee sslkeys.pem-----BEGIN RSA PRIVATE KEY-----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-----END RSA PRIVATE KEY----------BEGIN PUBLIC KEY-----MIGfMA0GCSqGSIb3DQEBAQUAA4GNADCBiQKBgQDnv6zntTRDw9r0O2wJWrzyfyeVO/++9JaXAq6M60uKKNGK/Lhej4rrMYyN4l8arUBWJBHLaAn9VPHymXA9VdoXaPIGu+YS9fhraKUkz3eXcQhxgPXxxAkyzOMsyEbLt7E9rQ9r9cNZVJ7rQBwloY0WHjYsZGBoT2fZI8wzz1+jTwIDAQAB-----END PUBLIC KEY-----

OBVIOUSLY don't use it for anything sensitive !

You can generate pem files that are passworded too, probably a good idea, when I find out how to do it, I'll post an update.The code below would then use openssl_get_privatekey($pem, $passphrase);

On to the PHP bit

All very simple really :

<?phpfunction private_encrypt($pem, $plain) {	openssl_get_privatekey($pem);	openssl_private_encrypt($plain, $cipher, $pem);	return $cipher;}function private_decrypt($pem, $cipher) {	openssl_get_privatekey($pem);	openssl_private_decrypt($cipher, $plain, $pem);	return $plain;}function public_encrypt($pem, $plain) {	openssl_get_publickey($pem); 	openssl_public_encrypt($plain, $cipher, $pem);	return $cipher;}function public_decrypt($pem, $cipher) {	openssl_get_publickey($pem); 	openssl_public_decrypt($cipher, $plain, $pem);	return $plain;}?>

You can use either the combined sslkeys.pem or the appropriate public.pem / private.pem files for $pem

Do not keep the private part or combined file in your webspace

$cipher is binary so if you want to send it over text channels you will need to base64_encode it or something

e.g.

<?phpfunction encode($plain) {	static $pem;	if(!$pem)		$pem = file_get_contents("/etc/ssl/sslkeys.pem");	return base64_encode(public_encrypt($pem, $plain));}function decode($cipher) {	static $pem;	if(!$pem)		$pem = file_get_contents("/etc/ssl/sslkeys.pem");	return private_decrypt($pem, base64_decode($cipher));}?>

the static keyword means that the value of the variable is preserved between function calls

What to use it for

Well as it happens, I'm using it to talk to Paypal. You can sent a custom string from your Buy button and this will be sent back to you later. So I'm sending the buyer an encrypted transaction code, they send it to Paypal and then I get it back from Paypal. Like many things in the world of encryption it's easy to be fooled into thinking things are more secure than they are. But hey, at least no-one knows what I'm sending :)

Another thing that can be done is if I encrypt something with my private key, you can decrypt with my public key. By doing this you know only I (or someone who has stolen my private key :) has sent the message. If I encrypt it with your private key first then no-one else can read it except you.

I could even base64 it, upload it to pastebin, and post the URL - like so http://pastebin.com/kSnWgzHs

Then linking you and I together is much, much harder.

RIPA can lick my balls.

Round and round the garden

2011-08-23T13:04:00.004+01:00

Go was playing up on FreeBSD (or maybe I've learned enough to cure that now, not sure yet) so I decided to get a Qemu image of some Linux up and running.

Debian was the one that finally worked.

Now, I use Plan 9 as my working environment. I run a Qemu image of it on a FreeBSD machine in the Co-lo which I connect to with drawterm.
I need to be able to run commands on the Debian from within the Plan 9, usually via win, so I can plumb the error messages (e.g. file.go:23).

I made a diagram of what I ended up with.

So today I tried to compile some C.
cc1: error: file.c: Value too large for defined data type
A fantastic error message if ever I saw one.
naturally someone else has already had that error and consequently I discovered that GCC can't compile code if it resides on a Samba share because the inode numbers are too big.

The solution is to remap the inode numbers in mount.

So the mount command in the diagram is out of date, it is now

# mount -t smbfs //10.0.2.2/kam /mnt/kam2 \ 

nounix,noserverino,uid=1000,gid=1000,credential=/root/kam.password

Also I had to use the IP address rather than the hostname, even with an entry in /etc/hosts. I probably need a WINS file somewhere, cba with that

Unlimited length path names, how do they work ?!

2011-08-19T11:10:00.006+01:00

On Linux
I'm sending an email, it is reporting an error.
I hit on the idea - I should include the pwd or else relative pathnames might be ambiguous.

"Error in ./config file" is too late for the user to realise they should have included full pathnames.

man 3 getcwd
char *getcwd(char *buf, size_t size);
...
If the current absolute pathname would require a buffer longer than size elements, NULL is returned, and errno is set to ERANGE; an application should check for this error, and allocate a larger buffer if necessary.

Further on :

Note that PATH_MAX need not be a compile-time constant; it may depend on the filesystem and may even be unlimited.

It is also very easily wrong. It is only defined as the maximum for the filesystem directory entry, not the path, it could be returned as ./././././././././././././././././ etc.

So we are limited by SIZE_MAX which is the upper bound on getcwd



char *

pwd() {

    size_t buffsize = PATH_MAX;

    char *buff = malloc(buffsize);

    if(!buff) return NULL;



    while(!getcwd(&buff, buffsize)) {

        if(errno != ERANGE) 

            return NULL; // caller can deal with it



        if(buffsize < (SIZE_MAX / 2)) { // prevent overflow

            buffsize = buffsize * 2;

        } else if(buffsize == SIZE_MAX) {

            fprintf(stderr, "getcwd failed, path too long max %d\n", buffsize);

            return NULL;

        } else {

            buffsize = SIZE_MAX;  // getcwd limit

        }



        buff = realloc(buff, buffsize);

        if(!buff) return NULL; // oom !

    }

    return buff;

}

Unlimited length path names, how do they work ?!

There is another way :
As an extension to the POSIX.1-2001 standard, Linux (libc4, libc5, glibc) getcwd() allocates the buffer dynamically using malloc() if buf is NULL on call. In this case, the allocated buffer has the length size unless size is zero, when buf is allocated as big as necessary.

I wonder how they deal with unlimited length path names, one time I shall look. In the meatime how can one code for UNLIMITED LENGTH PATHS !

UDP Synced MPlayer on Windows - not yet

2011-03-28T14:24:00.004+01:00

Update - udp sync is now part of mplayerNow I *do* have a video wall to build. 28 screens which sounds 7x4 to me.

But they want Windows !! Ok Mplayer for windows exists and there are howtos to compile MPlayer with mingw on Windows.

First ran the automated installer: MinGW from sourceforge

I clicked on Developer version, can see lots of gnu software defiling my system.

First failed instruction is DirectX headers being 404

http://www.mplayerhq.hu/MPlayer/releases/win32-beta/contrib/dx7headers.tgz

Ok, a bit more reading, mplayer shoul djust build on Cygwin so I've switched to that. First I'm trying to get my existing mplayer tree to compile, might work out :)

Had to sdd gcc, make, nasm, libiconv, freetype2. freetype brought lots of friends to my cygwin party.

had to get these http://natsuki.mplayerhq.hu/~reimar/{dinput,ddraw,dsound}.h

but they are included in the source these days apparently

had fun ewith bswap.h, cygwin defines the functiones therein and then everyone #includes libavutil/bswap.h and it treis to refine them. Wouldn't happen if they didn't abuse #IFDEFs.

codec-cfg doesn't work on Windows, I had to copy etc/codecs.conf to Linux and make codecs.conf.h there /home/maht/mplayer/mplayer/codec-cfg codecs.conf > codecs.conf.h
then on Windows just make the Makefile cp it into place rather than run codec-cfg

/usr/local/math.h log2(x) used float_t which confused matters, so I made it use float instead

I got it to build but it won't run

_feinitialise could not be located in the dynamic link library cygwin1.dll

Backing up Venti

2010-08-17T16:26:00.003+01:00

This code cycles through the arenas and if it's not sealed, dump it to dvd formatted iso

#!/usr/local/plan9/bin/rc

path=(/usr/local/plan9/bin $path)
cd /home/venti/dumps

fn arenas {
        wget -O - -q http://127.0.0.1:808/index | awk '
        BEGIN { FS="    | |="}
        /^arena/ { printf "%s", $2}
        $2 == "score" { printf " %s", $3 }
        $2 == "indexed:" { printf "\n" }
        '
}

fn extract {
        venti/rdarena /home/venti/arenas $1 | gpg --homedir /home/maht/.gnupg -e -r maht@maht0x0r.net  > $1/$2
        mkisofs -udf -o $1.$2.iso $1
}

fn process_arenas {
        ifs = '
'
        for(a in `{arenas}) {
                ifs=' '
                i = `{echo -n $a}
                mkdir $i(1)
                if(~ $#i 2) {
                        rm -f $i(1) ^.unsealed.iso
                        if(! test -e $i(1) ^. ^$i(2) ^.iso)
                                extract $i(1) $i(2)

                }
                if not {
                        extract $i(1) unsealed
                }
                rm -rf $i(1)
        }
}

process_arenas

Importing BASH style config files into rc

2010-08-11T17:14:00.005+01:00

So you're got some config in a bash style, something like this

% cat /usr/local/courier/etc/imapd
ADDRESS=0
PORT=143
MAXDAEMONS=80
MAXPERIP=25
PIDFILE=/var/run/imapd.pid
TCPDOPTS="-nodnslookup -noidentlookup"
LOGGEROPTS="-name=imapd"
IMAP_CAPABILITY="IMAP4rev1 UIDPLUS CHILDREN NAMESPACE THREAD=ORDEREDSUBJECT THREAD=REFERENCES SORT QUOTA IDLE"
IMAP_KEYWORDS=1
IMAP_ACL=1
IMAP_CAPABILITY_ORIG="IMAP4rev1 UIDPLUS CHILDREN NAMESPACE THREAD=ORDEREDSUBJECT THREAD=REFERENCES SORT QUOTA AUTH=CR
AM-MD5 AUTH=CRAM-SHA1 AUTH=CRAM-SHA256 IDLE"
IMAP_PROXY=0
IMAP_PROXY_FOREIGN=0
IMAP_IDLE_TIMEOUT=60
IMAP_MAILBOX_SANITY_CHECK=1
IMAP_CAPABILITY_TLS="$IMAP_CAPABILITY AUTH=PLAIN"
IMAP_CAPABILITY_TLS_ORIG="$IMAP_CAPABILITY_ORIG AUTH=PLAIN"
IMAP_DISABLETHREADSORT=0
IMAP_CHECK_ALL_FOLDERS=0
IMAP_OBSOLETE_CLIENT=0
IMAP_UMASK=022
IMAP_ULIMITD=65536
IMAP_USELOCKS=1
IMAP_SHAREDINDEXFILE=/usr/lib/courier-imap/etc/shared/index
IMAP_ENHANCEDIDLE=0
IMAP_TRASHFOLDERNAME=Trash
IMAP_EMPTYTRASH=Trash:7
IMAP_MOVE_EXPUNGE_TO_TRASH=0
SENDMAIL=/usr/sbin/sendmail
HEADERFROM=X-IMAP-Sender
IMAPDSTART=YES
MAILDIRPATH=Maildir

and you want to import it into rc

I give you

fn import_bash {
    . <{ 9 sed -e 's/"//g' -e 's/=/=(''/'  -e 's/(.)$/&'')/' \
               -e 's/''(\$[^ ]+) /\1 ''/' -e 's/''(\$[^ ]+) ''/\1/' \
         | awk ' /OPTS/ && / / { gsub(/ /, "'' ''") } { print } ' }
}

the bit inside the . <{} turns the bash into

ADDRESS=('0')
PORT=('143')
MAXDAEMONS=('80')
MAXPERIP=('25')
PIDFILE=('/var/run/imapd.pid')
TCPDOPTS=('-nodnslookup' '-noidentlookup')
LOGGEROPTS=('-name=imapd')
IMAP_CAPABILITY=('IMAP4rev1 UIDPLUS CHILDREN NAMESPACE THREAD=ORDEREDSUBJECT THREAD=REFERENCES SORT QUOTA IDLE')
IMAP_KEYWORDS=('1')
IMAP_ACL=('1')
IMAP_CAPABILITY_ORIG=('IMAP4rev1 UIDPLUS CHILDREN NAMESPACE THREAD=ORDEREDSUBJECT THREAD=REFERENCES SORT QUOTA AUTH=C
RAM-MD5 AUTH=CRAM-SHA1 AUTH=CRAM-SHA256 IDLE')
IMAP_PROXY=('0')
IMAP_PROXY_FOREIGN=('0')
IMAP_IDLE_TIMEOUT=('60')
IMAP_MAILBOX_SANITY_CHECK=('1')
IMAP_CAPABILITY_TLS=($IMAP_CAPABILITY 'AUTH=PLAIN')
IMAP_CAPABILITY_TLS_ORIG=($IMAP_CAPABILITY_ORIG 'AUTH=PLAIN')
IMAP_DISABLETHREADSORT=('0')
IMAP_CHECK_ALL_FOLDERS=('0')
IMAP_OBSOLETE_CLIENT=('0')
IMAP_UMASK=('022')
IMAP_ULIMITD=('65536')
IMAP_USELOCKS=('1')
IMAP_SHAREDINDEXFILE=('/usr/lib/courier-imap/etc/shared/index')
IMAP_ENHANCEDIDLE=('0')
IMAP_TRASHFOLDERNAME=('Trash')
IMAP_EMPTYTRASH=('Trash:7')
IMAP_MOVE_EXPUNGE_TO_TRASH=('0')
SENDMAIL=('/usr/sbin/sendmail')
HEADERFROM=('X-IMAP-Sender')
IMAPDSTART=('YES')
MAILDIRPATH=('Maildir')

The tricky part was the OPTS, they need to be passed individually not as one string.
Sadly that means keeping an eye on when such things are needed and changing the awk guard appropriately.

UvesaFB

2010-08-02T15:37:00.001+01:00

just a mirror of

http://dev.gentoo.org/~spock/projects/uvesafb/

in case it ever disappears

projects::uvesafb

uvesafb is a generic framebuffer driver for Linux systems and the direct successor of vesafb-tng. Its main features are:

* works on non-x86 systems,
* the Video BIOS code is run in userspace by a helper application,
* can be compiled as a module,
* adjustable refresh rates with VBE 3.0-compliant graphic cards.

uvesafb has been available as a part of the mainline Linux kernel since 2.6.24.
uvesafb::v86d

v86d is the userspace helper that runs x86 code in an emulated environment. uvesafb will not work without v86d. v86d currently supports the x86 and amd64 (x86-64) architectures.
Latest version v86d-0.1.9.tar.bz2
GIT repository v86d.git
uvesafb::installation instructions

uvesafb is included in gentoo-sources >= 2.6.23 and from 2.6.24 onwards also in the mainline kernel. If you happen to be using any of these, you don't have to patch your kernel manually. Otherwise, just download the patch using the link above and run cd /usr/src/linux ; cat uvesafb-0.1-rc3-2.6.23-rc3.patch | patch -p1

The guide below assumes you're using Gentoo Linux. If you happen do be using another distro, replace all emerge steps with unpack ; ./configure ; make ; make install or similar.

1. cd /usr/src/linux
2. configure the kernel; in the "Device Drivers" section select:
Connector - unified userspace <-> kernelspace linker and
Graphics support ---> Support for frame buffer devices ---> Userspace VESA VGA graphics support
(that's CONFIG_CONNECTOR=y and CONFIG_FB_UVESA=y)
3. build the kernel: make
4. remerge klibc: emerge klibc
note: This step is necessary because klibc has to be compiled against a kernel tree that includes uvesafb. You only need to do this once, i.e. you won't have to remerge klibc every time you install a new kernel or a new kernel module.
5. install v86d: emerge v86d
6. reconfigure your kernel; in the "General Setup" section select:
Initial RAM filesystem and RAM disk (initramfs/initrd) support
and use /usr/share/v86d/initramfs in Initramfs source file(s).
(that's CONFIG_INITRAMFS_SOURCE="/usr/share/v86d/initramfs")
7. rebuild the kernel: make, and install it
8. add video=uvesafb:1024x768-32,mtrr:3,ywrap (or similar) to your kernel command line
9. reboot and enjoy uvesafb :)

NOTE: The older versions of klibc seem to have problems with the new unified arches in the kernel (2.6.24). To avoid these issues, use klibc-1.5.8 or newer. Thanks to David Raison for pointing this out.

Network synchronized playback using UDP

2010-07-19T13:16:00.006+01:00

UPDATE 2011 - MPlayer udp-sync is now part of the main trunk.

Someone contacted me on how to do this and I realised I still hadn't written it up.

I tried to recreate how I made the player using svn's checkout by date feature but I failed to find the magic.

Instead here's a link to the patched source code (15M)

And here's the patches on their own

mplayer-udp-patch-17.diff
udp.diff

Seems I don't have the unpatched code. The date of the patch is 17 Feb 2010.

which should checkout with :

svn checkout svn://svn.mplayerhq.hu/mplayer/trunk -r "{2010-02-17}" mplayer

but it failed to build when I tried

G4 PowerBook screen replacement

2010-05-28T20:33:00.002+01:00

screen model B152EW01

Simon turned up with a spares or repairs he's ebayed and we got to work. We couldn't find any instructions so we just took all the screws out!!

The lid is glued to the body which makes it a bit scruffy when you put it back together but all worked out ok, no hassles really except working out what to do. Even though the outer shells looked the same, the two laptops didn't have exactly the same motherboard but the screen luckily had the same connections.

Sadly there were the dreaded vertical stripes when it came on but that might be damage on the MB from when it got smashed. It's still muc better than literally half the screen being lcd juice.

Fuck you jQuery

2010-05-24T10:59:00.002+01:00

> jQuery is a fixed part of RoundCube which allows you to make use of the very comfortable jQuery functions when writing plugin scripts.

Great, now I have to learn TWO frameworks at once.

Here's a tip, esp. for YOU Resig : If Javascript is too hard for you, perhaps you should pick up a broom and get a job you can do.

Mac-Mini ACPI wake-on-lan

2010-05-19T19:29:00.006+01:00

tl;dr : use ACPI, if you want wake-on-lan you can only suspend to ram

# uname -a
Linux mini1 2.6.32-gentoo-r7 #13 SMP Wed May 19 18:11:25 BST 2010 i686 Genuine Intel(R) CPU 1300 @ 1.66GHz GenuineIntel GNU/Linux

.config make.conf

APM

It won't have it :
CONFIG_APM=y
and yet

mini# apmd
No APM support in kernel

ACPI

mini# acpitool -e
... snip ...
   Device       S-state   Status   Sysfs node
  ---------------------------------------
  1. PXS1         S4     disabled   pci:0000:01:00.0
  2. PXS2         S4     disabled   pci:0000:02:00.0

mini# ethtool eth0
Settings for eth0:
... snip ...
        Supports Wake-on: pg
        Wake-on: d

So by default wake-on-lan is disabled (quite right too), thus :
(and *sigh* I have to show you the GPL! - is it really fucking worth it, the notice is longer than the code they gave me) :


mini# cat /etc/init.d/wol
#!/sbin/runscript
# Copyright 1999-2010 Gentoo Foundation
# Distributed under the terms of the GNU General Public License v2

depend() {
        need net
}
 
start() {
        acpitool -w | grep -q enabled || acpitool -W 2
        # -W 2 stupidly toggles enabled/disabled
        ethtool -s eth0 wol g
}
 
restart() {
        start
}

so now acpitool -[Ss] both work, I made a swap for storing disk based and it restores just fine, except mplayer waits until a media change before resuming to play on screen (the sound might work, I have it on mute). That's not good enough for this application, it might be an hour. I read some stuff on't'net earlier about dumping video registers and the like and ended up using sys-power/hibernate-script. It didn't do any better so I went for a quick and dirty : kill `{pgrep mplayer}

Or rather I installed daemontools and made mplayer a service (I use SLiM to auto-login to X but I realise now I could have done it with daemontools (I didn't have it installed at the time). Oh well, I learned about xdm.

Hibernate lets you run scripts on resume so I just have it kill mplayer and unclutter (which removes the cursor).

When the Mac-mini is in suspend-to-disk I can't get it to hear Wake-On-Lan magic packets, I have to use the power button to get it to resume. It works just fine with suspend-to-ram (the power led slowly pulses so you know it's on).

TuxOnIce looks like it might help but I can't be arsed with it no more.

Loeliger's TIL

2010-05-16T17:24:00.003+01:00

I first implemented Loeliger's TIL (ISBN-10: 007038360X) in AVR machine code though as with most FORTH type projects, writing the interpreter is as far as anyone goes. My interest in SIMD has given me an excuse to write it in x86 assembler. Unfortunately I'm going to have to do it on Linux, though it's like choosing which nugget to pick out of the cat litter tray. Alarmingly I read what is effectively the Linux manual. Nearly 400 syscalls. What a diabolical monster!!

Anyway Loeliger. One of my favourite readings.
Here's a template I made for the inner loop. It contains no actual mnemonics because the idea is that to port it to a new platform, one just %defines the required bits, just leaving the platform dictionary to write.

COLON:
        PUSH_I_2_RS             ; PSH I -> RS
        WA_2_I                  ; WA -> I
        JMP_NEXT                ; jmp NEXT
SEMI :
        my_ca                   ; Code Address of SEMI
        POP_RS_2_I              ; POP RS -> I
NEXT:
        DEREF_I_2_WA            ; @I -> WA
        I_2_NEXT_I              ; I = I + 2
RUN:
        DEREF_WA_2_CA           ; @WA -> CA
        WA_2_NEXT_WA            ; WA = WA + 2
        CA_2_PC                 ; CA -> PC ; implicit   JMP CA

I've done the %defines and %macros for x86 now.

work in progress

I've got some nifty macros going there, NASM seems easy to use.

Mac-Mini Gentoo revisited

2010-05-15T19:32:00.002+01:00

I got the Mac-Mini back. This time round I skipped the whole "Install 10.5 and use boot camp to make a partition" and went straight for the Gentoo i386 disk.

mini1 ~ # fdisk /dev/sda

WARNING: GPT (GUID Partition Table) detected on '/dev/sda'! The util fdisk doesn't support GPT. Use GNU Parted.

... snip ...

mini1 ~ # parted /dev/sda
GNU Parted 1.9.0
Using /dev/sda
Welcome to GNU Parted! Type 'help' to view a list of commands.
                                                                
(parted) unit b                                                           
(parted) p                                                                
Model: ATA FUJITSU MHV2060B (scsi)
Disk /dev/sda: 60011642880B
Sector size (logical/physical): 512B/512B
Partition Table: gpt

Number  Start         End           Size          File system  Name                  Flags
 1      20480B        209735679B    209715200B    fat32        EFI system partition  boot
 2      209735680B    1000000511B   790264832B    ext2
 3      1000000512B   10000000511B  9000000000B   ext3
 4      10000000512B  60011625983B  50011625472B  ext3

(parted)

Then I booted using the rEFIt iso disk, it fiddled with the MBR and now it boots.


# cat /etc/fstab
/dev/sda2               /boot           ext2            noauto,noatime  1 2
/dev/sda3               /               ext3            noatime         0 1
/dev/sda4               /mnt/Store      ext3            noatime         0 1
/dev/cdrom              /mnt/cdrom      auto            noauto,ro       0 0

# cat /boot/grub/menu.lst
title Gentoo Linux
root (hd0,1)
kernel /boot/kernel root=/dev/sda3

Wheeee all done, got X, got ALSA, got mplayer

I have to make it do i915 as a module for Xorg, had to add mesa for the dri to work.

But work it does !!

Keep on Clucking with Chickenfoot

2010-05-10T11:58:00.005+01:00

Websites with black backgrounds do my eyes in, so here's a Chickenfoot script to set the background and color for all elements on the current page to Plan9 like calm.


var s = document.getElementsByTagName("*")
for(a in s) {
  s[a].style.background = "#FFD"
  if(s[a].nodeName == "A") { // HREFs
    s[a].style.color = "#00F"
    s[a].style.textDecoration = "none"
  } else {
    s[a].style.color = "#111"
  }
}

And here's a bit of helper code to inspect collections :
(in Chickenfoot the last evaluation is printed on the console)


function inspect(e) {
  var b = ""
  for(k in e) {
    b += k + " = " + e[k] + "\n"
  }
  return b
}

// use it like this :

var c = ""
var s = document.getElementsByTagName("*")
for(a in s)
  c += inspect(s[a])
c

This simple script sat on right click has made my day, and it's only midday!

Mplayer using Nvidia Vdpau

2010-04-30T15:38:00.006+01:00

I did a side by side projection last night. Two 1280x768 projectors pretending to be 1280x1024 (I can probably drive them dot for dot but I didn't have enough tinker time).

Instead of the MatroxDualHeadToGo I used mplayer with Jason Holt's udp-patch.

The playback machine has two Nvidia 8600GTs and I used one for each projector.

The mp4s were only 25Mb each, even so I played them from RAMdisk

# ls -l /dev/shm/LR
total 47164
-rw-r--r-- 1 maht maht 22602673 Apr 29 12:16 L.mp4
-rw-r--r-- 1 maht maht 25579796 Apr 29 12:16 R.mp4

With Vdpau


 PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND
3884 maht      20   0  220m  45m  17m R   24  0.6  32:23.67 mpu
3886 maht      20   0  221m  45m  17m R   23  0.6  31:07.02 mpu
 
# cat /proc/3884/cmdline  | tr '\0' ' '
mpu -vo vdpau -fs -zoom -fixed-vo -udp-master -loop 0 -input file=/dev/null /dev/shm/LR/R.mp4

# cat /proc/3886/cmdline  | tr '\0' ' '
mpu -vo vdpau -framedrop -fs -zoom -fixed-vo -udp-slave -loop -0 -input file=/dev/null /dev/shm/LR/L.mp4

Without Vdpau


DISPLAY=:0.1 mpu -vo xv -framedrop -fs -zoom -fixed-vo -udp-slave -loop -0 -input 'file=/dev/null' /dev/shm/LR/L.mp4 &
DISPLAY=:0.0 mpu -vo xv -fs -zoom -fixed-vo -udp-master -loop 0 -input 'file=/dev/null' /dev/shm/LR/R.mp4

  PID USER      PR  NI  VIRT  RES  SHR S %CPU %MEM    TIME+  COMMAND
 4252 maht      20   0  221m  47m  20m R   31  0.6   0:16.00 mpu
 4253 maht      20   0  221m  47m  20m R   30  0.6   0:14.82 mpu
 3860 root      20   0  114m  40m  26m R   15  0.5   0:16.31 X

So for this low bandwidth mp4 using vdpau saved me 25 %CPU points on the quad core, or even 1 whole CPU if you think about how it will get scheduled with > 2 cores. Looks tasty for the dual core player.

Exhausting FDs in Go-lang

2010-04-28T08:49:00.004+01:00

Looks like the garbage collector is lazy when clearing away unused Fds so always call fd.Close()

Installing Go on Fresh Gentoo

2010-04-21T15:18:00.000+01:00

Gentooed up the Quad Core.
On a fresh install, let's see what needs to be added for Go

# emerge mercurial
# emerge ed
# emerge syslog-ng  # yes folks, Go has a dependency on syslog
# /etc/init.d/syslog-ng start # don't forget like I did

AoE client for Windows

2010-04-21T11:02:00.000+01:00

StarWind's gone AoE client Free

You're supposed to reg

http://www.starwindsoftware.com/

Shame about the *AWFUL* UI

proper belongs in the Hall of Shame

Image Processing with Go

2010-04-20T17:20:00.004+01:00

The usual story of aphid meets programming language is to see what it's like at image processing.

file.go from the Go Tutorial

package main

import (
        "fmt"
        "image"
        png "image/png"
        "./file"
        "os"
)

var errn = 0
func abort(e os.Error) {
        errn++
        if e != nil {
                fmt.Printf("ERR %s\n", e)
                os.Exit(errn)
        }
}

func main() {
        fd, err := file.Open("in.png", 0, 0)
        abort(err)
        var img image.Image
        img, err = png.Decode(fd)
        fd.Close()
        abort(err)
        w, h := img.Width(), img.Height()
        opng := image.NewNRGBA(w, h)
        var c image.Color
        for y := 0; y < h; y++ {
        for x := 0; x < w; x++ {
                c = img.At(x, y)
                opng.Set(x, y, c)
        }
        }

        fd, err = file.Open("create.png", os.O_RDWR | os.O_CREATE, 0644)
        abort(err)
        err = png.Encode(fd, opng)
        abort(err)
        fd.Close()
}


% cd /dev/shm; cp $home/1920x1440.png in.png; time /tmp/8.out
real    0m9.810s
user    0m12.425s
sys     0m0.072s

around 1s per 2000 pixels

Multicore Shell Scripting

2010-04-20T15:00:00.011+01:00

Processing lots of files at once in a process level embarassingly parallel way is v. handy. I've been meaning to get round to doing this shell script spawner that will use a fixed number of processes launching, by default, one process for each CPU core the OS reports.

It reads one line at a time and executes it until N are running then it waits until they have all finished and then spawns N more.

It's the most horrible way of doing it from an I/O point of view, hotspot bonanza!!

So maybe it's best to echo 'sleep 2; gm mogrify -resize 5000x3000 foo.jpg bar.png'
I don't know, we'll see

#!/usr/local/plan9/bin/rc

fn os_cpus { # platform dependent, number of cpus, counting from 1
        # Linux
        grep '^processor' /proc/cpuinfo |
                tail -n 1 |
                sed 's/[^0-9]//g' |
                awk '/./ {print $0 + 1; exit} {print "1"}'
}

fn cpus {
        if(~ $#1 0)
                os_cpus # default
        if not
                echo $1
}

cpuN = `{cpus $1}
while() {
        for(i in `{seq 1 $cpuN}) {
                if(cmd = `{read}) { # brace required :) clever rc parser
                        $cmd &
                }
                if not
                        kill $pid
        }
        wait
}

I just ran this

{ for(f in f*png) echo gm convert $f -interlace None i^$f } | conquer

Worked a treat

Mac-Mini is booting Gentoo Linux

2010-04-18T15:41:00.010+01:00

used these notes
Install 10.5 to the whole disk
run bootcamp from Applications/Utilities and make a Windows partition
install rEFIt
reboot, reboot again
boot gentoo x86, use parted to make root in the fat32 partition bootcamp made
reboot - run rEFIt partition utility to remake MBR
boot gentoo, make file system

parted output

Partition Table: gpt

Number  Start   End     Size    File system  Name                  Flags
 1      20.5kB  210MB   210MB   fat32        EFI System Partition  boot
 2      210MB   23.7GB  23.5GB  hfs+         Untitled
 3      23.7GB  60.0GB  36.3GB  ext2                               boot

......... carry on with install

I used the make.conf from this useful page

mac_mini kernel config

Compiling away .....
come one, compilaroo

grub-install --no-floppy /dev/sda3

whee I now have Gentoo Linux booting on the mac mini

it's currently in VESA 1920x1440x32

added "directfb" to the use flags

emerge mplayer

I'm on 17 of 103 packages .....
80 of 103 ... come on giddy up

mplayer doesn't work using directfb 'initialization error' - maybe return to that later - intel 954gm is the device

works fine in fbdev mode but that doesn't do -fs -zoom properly. -xy 1024:768 works but goes a bit stuttery
fbdev2 doesn't work with -fixed-vo

so emerge xorg-x11

ideally I want to be in 1920x1440x32 and scale everything up to fs, fingers crossed that X11 will win !

.......

fail - X11 couldn't find its driver and I had to give up the MacMini

X -config /root/xorg.conf.new 

X.Org X Server 1.7.6
Release Date: 2010-03-17
X Protocol Version 11, Revision 0
Build Operating System: Linux 2.6.32-gentoo-r7 i686 
Current Operating System: Linux mini1 2.6.32-gentoo-r7 #3 SMP Mon Apr 19 11:31:59 BST 2010 i686
Kernel command line: root=/dev/sda3 vga=0x318 video=uvesafb:1024x768-24@72,mtrr:3,ywrap
Build Date: 19 April 2010  12:47:20PM
 
Current version of pixman: 0.17.2
        Before reporting problems, check http://wiki.x.org
        to make sure that you have the latest version.
Markers: (--) probed, (**) from config file, (==) default setting,
        (++) from command line, (!!) notice, (II) informational,
        (WW) warning, (EE) error, (NI) not implemented, (??) unknown.
(==) Log file: "/var/log/Xorg.0.log", Time: Mon Apr 19 12:54:28 2010
(++) Using config file: "/root/xorg.conf.new"
FATAL: Module i915 not found.
xf86TokenToOptinfo: table is NULL
xf86TokenToOptinfo: table is NULL
error setting MTRR (base = 0x80000000, size = 0x10000000, type = 1) Inappropriate ioctl for device (25)
(EE) AIGLX error: dlopen of /usr/lib/dri/i915_dri.so failed (/usr/lib/dri/i915_dri.so: cannot open shared object file: No such file or directory)
(EE) AIGLX: reverting to software rendering
error setting MTRR (base = 0x80000000, size = 0x10000000, type = 1) Inappropriate ioctl for device (25)

it's right, there's not much in /usr/lib/dri/

maybe that MTRR thing is significant, I need to boot with a LiveCD and find oot, Sabayon Linux 5.2 is rather nice, that looked great (but no shell on the media centre version I booted)

so, until next time ........

MPlayer OSX Revisited - failed

2010-04-13T19:32:00.004+01:00

The MPlayer OSX Extended tells you all you need to know about GUI oriented OSes. You can't launch it from the command line properly you have to tell it what to open with a series of pulleys.

So I'm trying to compile it using these instructions. I already tried using ./configure && make but that failed with _ff_eval2 missing symbol.

For extra points I have to avoid upgrading to 10.5 as there might be no going back.

I'm also a bit hungry, the chip shop is closed. Curse that Rocco and his real life, doesn't he know we researchers of the deep rely on his potato chipping and frying skills.

WTF does subversion need sqlite3 for !. Do they really need to express their data access in SQL ! We're always one step away from being Crystal Reports in the end.

Needed an extra : port install libtool. nah, did a MacPorts upgrade : sudo port -d selfupdate
and then port installed all the pre-requisites again! ages.

Try not to ever do this yourself without a big book to read while it takes forever with no guarantee of success, even if it works I know it won't do what I want. BUt what I will have is working source code, that's golden.

game over, updated subversion source doesn't compile

Mplayer framebuffer

2010-04-13T13:24:00.002+01:00

Mplayer in the framebuffer seems to be best with DirectFB,

mplayer -quiet -vo directfb:buffermode=triple -zoom -fs -aspect 16:9 NSC_TV_ad_30s.avi

DirectFB also gives us buffermode={single, double, triple}
fieldparity={top, bottom, disabled}
layer=N
dfbopts=

stuff that works

-flip -x n -y n -xy n,m (set width, height to n,m) -vf rotate -vf scale -vf mirror

scale resamples the stream before outputting it -xy just hardware scales it

v happy

Why I love AoE

2010-04-12T18:43:00.004+01:00

I'm using a 2GB flash card to boot from but as I build the system it uses /var/tmp as a place to build all the intermediate files.

We can get two wins for the price of one.
First off make a new vblade

daph:/# cd /dev/shm/
daph:/dev/shm# qemu-img create -f raw var_tmp 2G
Formatting 'var_tmp', fmt=raw, size=2097152 kB
daph:/dev/shm# ls -l var_tmp
-rw-r--r-- 1 root root 2147483648 2010-04-12 18:55 var_tmp
daph:/dev/shm# du -h var_tmp 
0       var_tmp

This gives us a maximum size of 2G but until we use it, it's size is actually 0

By using

# vblade 2 3 eth0 /dev/shm/var_tmp
pid 15913: e2.3, 4194304 sectors O_RDWR

run # fdisk /dev/etherd/e2.3
and create 1 primary partition covering the disk (or however you want!) surely you know how
In my case # mke2fs -T small /dev/etherd/e2.3p1
then mount it and importantly bind our memory based /tmp/var in place

mount /dev/etherd/e2.3p1 /mnt/boof
mkdir -p /mnt/boof/var/tmp
mount --bind /mnt/boof/var/tmp /var/tmp
emerge something

and emerge will use the vblade for tmp storage

of course Linux laughs in your face because binds are root level so once it's done, it's done, you can't pop to another window and keep looking in /var/tmp on your local disk, that's gone until you unbind it.

Of course you don't need to be using shm for the vblade you can use disk files or partitions. With a few machines donating RAM you're probably able to saturate your Gbit, with this new machine I'll finally be able to make some measurements.

% time dd if=/dev/zero of=/dev/null count=2000000 bs=512
1024000000 bytes (1.0 GB) copied, 1.22008 s, 839 MB/s 
%  time dd if=/dev/zero of=/mnt/boof/zeros count=2000000 bs=512
1024000000 bytes (1.0 GB) copied, 80.0547 s, 12.8 MB/s
% time dd if=/mnt/boof/zeros of=/dev/null  count=2000000 bs=512
1024000000 bytes (1.0 GB) copied, 1.66719 s, 614 MB/s

This last one is pretty astonishing, I think it needs testing without zeros or something

Framebuffer fun

2010-04-12T16:58:00.006+01:00

Getting the framebuffer splash page going was a fun day.

There's some not so secret sauce - that's the definitive one for me April 2010

% emerge splashutils
% emerge -av splash-themes-gentoo
% splash_geninitramfs --verbose --res 1280x1024 --generate /boot/splash gentoo

it uses /etc/splash/gentoo as the source Dir for the files

title Gentoo Vesa
root (hd0,0)
kernel /boot/vmlinuz root/dev/sda2 vga=0x31b video=nvidiafb:1280x1024-32@72,tty:12 splash=silent,theme:gentoo console=tty1
initrd (hd0,0)/boot/splash

various places claim you need 16bit vga but that's no longer true

make sure the bits in the kernel conf are set

my fb1 config

next stage I guess is making custom ones

CF Write Speed

2010-04-12T12:02:00.005+01:00

I'm writing a whole partition at once and getting an interesting phenomena. The write speed is varying from 4.3-3.3MB/s which strikes me as quite a range 23% on a solid state disk. Ok It's a across a LAN but it's 1Gbit and only me on it.

Final figure was : 1883Mb 3.7MB/s 8m32s

Just to tickle me every time, the damn partition I backed up wasn't cleanly unmounted and I have to e2fsck the darn thing when I restore it thinking the install will be over before I need to do it again, the curse of the positive outlook.

Backing up is a bit quicker


dd if=/dev/sda | ssh 10.0.0.1 'cat > gentoo.sda'
...
2017419264 bytes (2.0 GB) copied, 181.301 s, 11.1MB/s

I did the whole partition this time so I can boot it in Qemu

% time ssh 10.0.0.1 'cat gentoo.sda' > /dev/sda
real 9m1.404s

I'm now a Gentoon

2010-04-11T15:42:00.003+01:00

I spent 5 hours on Friday installing Gentoo to my new playback PC
ASRock K10N78FullHD-hSLI motherboard with 2 GB DDR2 800Mhz RAM and Athlon X2 245 Dual Core 2.9 GHz 2MBL2 65W.

I'm booting it from a 2Gb Compact Flash card in a SATA <> CF device.

boof ~ # cat /etc/make.conf 
ACCEPT_KEYWORDS="amd64"
CHOST="x86_64-pc-linux-gnu"
CFLAGS="-march=athlon64 -O2 -pipe"
CXXFLAGS="${CFLAGS}"
MAKEOPTS="-j3"
GENTOO_MIRRORS="ftp://gentoo.virginmedia.com/sites/gentoo http://gentoo.virginmedia.com/ http://www.mirrorservice.org/sites/www.ibiblio.org/gentoo/ ftp://ftp.mirrorservice.org/sites/www.ibiblio.org/gentoo/ "
SYNC="rsync://rsync.europe.gentoo.org/gentoo-portage"
USE="-hal -dbus -kde -qt3 -qt4 -gnome fbcon -apache2 -aqua -cairo directfb dri gimp -gnustep -handbook -innodb -kontact mmx -mono -mssql -mysql -oci8 openmp -oracle postgres scanner smp sse sse2 ssl vnc x264 -xemacs xvid"
FEATURES="userfetch parallel-fetch"

I dd'd the sda2 partition to use as a vblade but it complains that it has no partition table so I'm going to build it from scratch

read speed from the CF is pretty good

I'm using AOE for /var and /usr/{portage, src}

I've just done ACCEPT_KEYWORDS="~amd64" emerge plan9port
and it's dragging in the whole of X, which could well be a good thing

The vblade I'm using I made with qemu-img create so it uses sparse files, It has ballooned up to 1.1G. I can roll this back as I backed up sda{1,2} before doing it and have already rolled back once when I got caught out by umount.

warning casting double to int : I hope these people know what they are doing. I love Go's attitude : any errors - no build

bah, an error in libXtst and I'm SOL

reset from that, went back to text install pre plan9port, that obviously wants to go all XOrg on me. It locked up and the CF corrupted.

Got the framebuffer going, emerge cdrkit is wodim and friends, dvdauthor brought in imagemagick

Adding venti to an existing Fossil

2010-04-09T16:00:00.002+01:00

Pretty simple but I had to look it up.

Here's where I started

cpu% fossil/conf /dev/sdC0/fossil
fsys main config /dev/sdC0/fossil
fsys main open -V -c 3000

and to change it

cpu% fossil/conf -w /dev/sdC0/fossil <{
echo fsys main config /dev/sdC0/fossil
echo fsys main open -c 3000
}

I just love <{}

Go Slow

2010-03-26T00:40:00.013Z

Cores up to your ears, so stop writing single threaded code! You can even grab multi machine programming, you'll hardly know the difference and there are some things that can really benefit from it. Imagine image editors grabbing spare cycles from the machines on the LAN as required. It's here today and it's called Go.

In this entry I'm going to take someone's existing Go implementation of the smallpt global illumination renderer. I hope to demonstrate the easy multi-threaded nature of CSP based techniques, even spanning machines!

The single threaded Go version gives me :

% time ./smallpt 8
Rendering (8 spp) 100.00
721.50 user 5.08 system 12:06.84 elapsed 99%CPU

Single threaded C++ with -O3

% time ./small-sthread 8
Rendering (8 spp) 100.00%
61.76u 0.27s 62.04r     ./small-sthread 8

OpenMP Multi threaded Code, 2 cores + HT

% time ./small-omp 8
Rendering (8 spp) 100.00%
88.55u 0.90s 24.83r     ./small-omp 8

Oh dear, that's amazingly piss poor. An order of magnitude to find!

OK I added an I/O thread to gather the pixels

time ./smallpt.io 8
Rendering (8 spp) 100.00
849.65 user 54.19 system 14:02.77 elapsed 107%CPU

And not suprisingly I added 50s of Syscalls, some 10% overhead but at least they seem to have ended up on 1 CPU.

Ultimate overload : 1 thread per pixel.


crashed

with a threadpool of 4 on a machine with 2 HT CPUS

time ./smallpt 8
Submitting (8 spp) 100.00
1106.38 user 369.83 system 15:47.15 elapsed 155% CPU

This used more CPU time but the running time was much the same. All the SMP gain is used up by the overhead.

I manually inlined the function calls - getting rid of the Vec methods

% time ./smallpt.un 8
Rendering (8 spp) 100.00
273.82 user 5.62 system 4:40.05 elapsed 99% CPU

shit, now I fixed it it is 4x slower !

One interesting question I have is, do Go routines without a CPU operate in non pre-empting co-op mode like Limbo? So if I spawn my Go routine to do slow I/O will it be interleaved or sit there doing nothing while the heavy CPU work co-ops ?

Go Routined version

time ./smallpt_gr.go 8
1012.35 user 454.21 system 16:21.66 elapsed 149% CPU

Libbed out vector you'll need

My arenas are too big

2010-03-04T10:54:00.003Z

I thought having 4gb arenas would be a good match for dumping to DVD but it's not because you have to read the arenas out in one go, there's no offsetting and although arenas only grow, the daily gpg encrypted version of the unsealed areana has different output depending on the length as well as the contents. Offsite copying is also maximized in this manner.

The ideal solution is to match the arena size to the daily growth and adjust when adding new arenas. I've not added new arenas to my venti yet so maybe there's even a way of automating it.

I'm going to go for 200Mb arenas. Gpg manages some compression so we might get 4 arenas on a CD or 25 on a DVD.

So apart from the occasional massive burst of Gb files, the daily "email & logs" reports will be maxed out at 200Mb instead of the just under 1.2Gb I've got atm. Uploading that offsite is just too much.

UDP broadcast

2010-02-27T15:02:00.007Z

I found the UDP listener here. He calls it a server but in my context it is what it is, a listener. I'm sure Gunnar Gunnarsson won't mind me adapting it.

The listener spits out "IP\tmessage\n". It gets the messages from a the broadcaster.

What happens next is up to you.

Mplayer video wall

2010-02-16T13:02:00.003Z

I found a mplayer syncing patch which will be amazing if it works!

http://persephone.cps.unizar.es/~spd/cmplayer/

If it works I'll look into the code and see how it applies to more recent versions, there's not much, it only patches these files (though the patch file is 5k)
Makefile
mplayer.c
parser-mpcmd.c

It worked fine against rc2

With this configure
./configure --enable-largefiles --enable-menu

but fails when I use checkout by date on the next version
(cave-mplayer source code (v. 0.3) for MPlayer SVN (Fri Jun 13 2008))

With this configure
./configure --enable-menu

% svn checkout -r {2008-06-13} svn://svn.mplayerhq.hu/mplayer/trunk mplayer
% gzip -dc cave-mplayer-0.3.1.tgz | tar  xvf -
% cd mplayer
% patch < cave-mplayer/cave-mplayer.patch

worked a treat

% make
... snip
In file included from ./libavutil/avutil.h:76,
                 from ./libswscale/swscale.h:30,
                 from libvo/vo_yuv4mpeg.c:43:
./libavutil/pixfmt.h:32:32: error: libavutil/avconfig.h: No such file or directory

and I can't fix the bleeder. So I'm going to try newer / older dates randomly !

13 July 2008 also broken
Jan 01 2009 also broken, same shit
Trying Jan 01 2010 to see if  Diego's patch from 19th Jan is the only fix. Yep same shit.
Fuck me these people can't even keep a build working!

I tried to fix it but was falling down a big hole with glass on the sides

rc2 it is



Installing OSX Leopard again
2010-02-14T15:22:00.008Z
Last time I tried to install Leopard I found some version clashes with Final Cut & Compressor - namely PPC compressor doesn't work and one needs Universal Compressor (according to some forums). After the heartache of DVD authoring not working I shall try again.

Step 1.
Install Leopard from DVD

Great start, it's kept the volume name. Scratch01 isn't too great! See you in Calculating... minutes (it said 1 hour, turned out to be 20 mins).

Installed the optional dev tools, X11, gcc et al.

Upon reboot it asked me to fix quartermaster's permissions, it didn't complain on the next reboot

Installed Final Cut Pro Upgrade, including LiveType and Comrpessor

Comrpessor still doesn't work - no suprise there, but now I discover we don't even have iDvd

Downloaded Pacifist from http://www.charlessoft.com/  to look for packages I've missed
DVD Studio doesn't install from the DVD Studio Content disk even though it's universal, FCP refuses because it's a set of Intel disks, agghhh, someone's snaffled our Universal!

It needs DSPGroup, XSKey, SharedComponents, ProCoreA, ProRuntime

And, of course, a valid serial number.


EeePC Repositories
2010-02-07T14:30:00.002Z
So I can keep track between wipes
see also
http://wiki.eeeuser.com/addingxandrosrepos

deb http://update.eeepc.asus.com/1.6 common main
deb http://update.eeepc.asus.com/1.6 p901 main
deb http://update.eeepc.asus.com/1.6 en main

# Original Asus 701 repos, do not change the reference to the 701!
deb http://update.eeepc.asus.com/p701/ p701 main
deb http://update.eeepc.asus.com/p701/en/ p701 main

# Community repos, good for conky. Do not change the reference to the 701
deb http://download.tuxfamily.org/eeepcrepos/ p701 main etch

# Xandros repos
deb http://xnv4.xandros.com/xs2.0/upkg-srv2/ etch main contrib non-free
deb http://www.geekconnection.org/ xandros4 main
deb http://xnv4.xandros.com/4.0/pkg/ xandros4.0-xn main contrib non-free

# XEPC repo, specifically compiled for the eee!
deb http://updates.xepc.org/ p701 main
# XEPC alternative repo, if main repo down
# deb http://download.tuxfamily.org/xepcrepo/ p701 main

# Debian repos now listed as etch and oldstable which are both currently identic
al. 
deb http://www.debian-multimedia.org/ oldstable main 
deb ftp://ftp.uk.debian.org/debian/ etch main contrib non-free


after an apt-get update

apt-get install xandros-archive-keyring

I'm doing an apt-get upgrade, and it's going to take an hour. I might be headed for Wipeout!


compiling drawterm
2010-02-07T14:27:00.002Z
Debian packages required :
% sudo apt-get install build-essential libx11-dev libxt-dev

libx11-dev gets you X11/Xlib.h and friends
libxt-dev gets you X11/IntrinsicP.h and co


Venti a plenti
2010-02-06T20:44:00.003Z
I had to put a few small things in place before it would work.

First off is a script to run, this must live somewhere Inferno can find, I used % cat /usr/maht/backup
bind '#Uc:' /n/local
vacput -a tcp!10.0.0.1!17034 '/n/local/Documents and Settings/maht/My Documents' >> /usr/maht/my.documents.vac

When you run that and try and unvac it on Lunix first time you'll have some fun because 'My Documents' is dr-xr-x-- so it creates the directory and then sigfaults because it can;t write into it!

So I had to run Inferno andchmod 775 '/n/local/Documents and Settings/maht/My Documents/' 
chmod 775 '/n/local/Documents and Settings/maht/My Documents/My Pictures' 
chmod 775 '/n/local/Documents and Settings/maht/My Documents/My Music' 

Next job is to make a Scheduled Task. I did it as Administrator but gave it the credentials of a normal user.

It's effectively a shortcut with the command line "C:\inferno\Nt\386\bin\emu.exe /dis/sh.dis /usr/maht/backup"

I don't know how fast it's copying, hopefully I'll be able to work it out after it's finished running tonight.


Venti on Linux via p9p
2010-01-27T16:51:00.006Z
The plan is to backup arenas to DVD
DVD-Rs store 4707319808 bytes
Four arenas should be enough to start with

(when I posted this mycrotiv shoed me his venti setup script)

% echo '4 * 4707319808 / 1024' | hoc
18387968
% time dd if=/dev/zero of=arenas bs=1024 count=18387968 # stupid linux dd and its =
real    14m0.207s

venti(8) recommends an index size that is 5% of the active data log.

It doesn't describe the difference between the active data log and any other sort so I'll assume 5% of the arena size which in this case is 
% echo '0.05 * 4 * 4707319808 / 1024' | hoc
919398.4
% dd if=/dev/zero of=isect bs=1024 count=919399

Seeing as I'm using it for backup all sorts of shizzle, I'm going to include the bloom filter.

The bloom filter thus has two parameters: nhash (maximum 32) and the total bitmap size (maximum 512MB, 232 bits). nhash × nblock <= 0.7 × b, where nblock is the expected number of blocks stored on the server and b is the bitmap size in bits.
b = 232, nash = 32 
32 × nblock <= 0.7 × 232
nblock <= 0.7 × 232 / 32
nblock <= 5.075
so I guess nblock = 5, anyway says the maximum size is 512Mb
% echo '512 * 1024' | hoc
524288
% dd if=/dev/zero of=bloom bs=1024 count=524288

That's the storage reserved now for formatting

I'm going to round down the actual arena size a bit
% time venti/fmtareanas -a 4705000000 -Z arenas0. /home/venti/arenas # I dd'd in zero
fmtarenas /home/venti/arenas: 5 arenas, 18,828,484,608 bytes storage, 524,288 bytes for index map
real    0m1.312s
% time venti/fmtisect -Z isect0. /home/venti/isect
fmtisect /home/venti/isect: 114,827 buckets of 215 entries, 524,288 bytes for index map
real    0m0.513s
% time venti/fmtbloom -N 32 /home/venti/bloom
fmtbloom: using 512MB, 32 hashes/score, best up to 95,443,717 blocks
real    0m26.742s


Ok that looks like the formatting is done, set up venti.conf
% cat venti.conf
index main
isect /home/venti/isect
arenas /home/venti/arenas
bloom /home/venti/bloom
httpdaddr tcp!127.0.0.1!808
% time /usr/local/plan9/bin/venti/fmtindex venti.conf
fmtindex: 5 arenas, 114,827 index buckets, 18,828,402,688 bytes storage
real    0m0.483s


hmm, ok , lets run this sucker

% /usr/local/plan9/bin/venti/venti
2010/0127 21:12:22 venti: conf...httpd tcp!127.0.0.1!808...init...icache 0 bytes = 1,000 entries; 4 scache
sync...announce tcp!*!venti...serving.
% wget -q -O - http://127.0.0.1:808/storage
index=main
total arenas=5 active=0
total space=18,828,402,688 used=0
clumps=0 compressed clumps=0 data=0 compressed data=0

as usual, despressingly straight forward, I didn't even need a loopback


Restoring Venti from dumped arenas using Qemu
2010-01-08T17:24:00.012Z
I've been using this script to dump my venti arenas from my laptop to RAID.

When I'm on the LAN I run this by hand as bootes

#!/bin/rc
test -e /n/daph/home || 9fs daph
bk = /n/daph/home/bootes/t23/arenas
if(! test -e $bk)
 mkdir -p $bk
if(! test -e $bk)
 status = $mk ^missing && exit
/usr/maht/bin/rc/vbackup /dev/sdC0 $bk

and I kept saying to myself - "I must do a test restore".

Naturally fate intervened and my HD decided that I didn't want to read my isect partiton after all and my hand was forced.

So after a couple of false starts trying to do it with the bootable CD I fired up Qemu with a blank 3G image - my arenas are 1.6Gb - bit of headroom, and installed fossil+venti.

I turned it into a cpu/auth with my script to make logging into it a bit easier

and then as bootes  
% 9fs daph; cd /n/daph/arenas/
% venti/wrarena sdC0.arenas00
% venti/wrarena sdC0.arenas01
% time venti/wrarena sdC0.arenas02
1.67u 46.65s 496.71r  
% ls -l sdC0.arenas02
--rw-rw-rw- M 85 bootes bootes 536870912 Dec 16 06:46 sdC0.arenas02

which seems easy enough, 1Mb sec

boot the laptop from the CDrom, edit plan9.ini to use my qemu is its venti (using -redir tcp:17034::17034 

had to add 
venti=/dev/sdC0/arenas tcp!*!17034
 to qemu's plan9.ini

balls the qemu is sending a reset instead of data - I think there's a problem in there somewhere after a few different tries, fossil/flfmt dies when using my qemu venti as a host too


OK so with the Venti install I added a new disk - recover

My invocation for qemu venti now looks like this :
qemu.exe -L . -m 1024 -hda venti -hdb recover -localtime -M pc -redir tcp:3131::3131  -vnc :8 -snapshot -redir tcp:5356::5356 -redir tcp:17010::17010 -redir tcp:567::567 -redir tcp:17007::17007 -redir tcp:17034::17034

to get my data I did this as bootes


disk/mbr -m /386/mbr /dev/sdC1/data
disk/fdisk -baw /dev/sdC1/data
disk/prep -bw -a fossil /dev/sdC1/plan9
fossil/flfmt -v 22d...ff3 /dev/sdC1/fossil   # 22d...ff3 is the score from fossil/last
fossil/conf -w /dev/sdC1/fossil <<EOF
fsys main config /dev/sdC1/fossil
fsys main open -AWP
fsys main
srv -p ricons
srv rico
EOF
fossil/fossil -f /dev/sdC1/fossil


drawtermed in again as me
mount /srv/rico /n/rico
cd /n/rico/usr/maht
tar cp . > /n/daph/home/maht/plan9_usr_maht.tar


You beauty!!


Acme tag for when editing man pages
2009-12-09T14:34:00.003Z
I've been writing more man pages recently that aren't in the usual Manpath and instead of building a new plumbing rule I made a tag

% cat /acme/bin/Man
#!/bin/rc

if(~ $#* 0) {
    tags=`{cat /mnt/acme/$winid/tag}
    f = $tags(1)
}
if not
    f = $1

nroff -man $f



cd with cd - feature
2009-12-08T13:07:00.003Z
On Lunix cd as an extra feature namely cd -
which takes you to the previous directory

I've got quite into this as I jump about, freebsd had pushcd and popcd which was also fun.

fn cd {
 cdnextlastdir = `{pwd}
 {
  if(~ $#* 0)
   builtin cd
  if not if(~ $1 -)
    builtin cd $cdlastdir
   if not
    builtin cd $1
 } && cdlastdir = $cdnextlastdir
}

I've got it in lib/profile

EXAMPLE


t23% cd
t23% pwd
/usr/maht
t23% cd /
t23% pwd
/
t23% cd -
t23% pwd
/usr/maht
t23% cd ouihouihiuh
Can't cd ouihouihiuh: 'ouihouihiuh' file does not exist
t23% cd -
t23% pwd
/
t23% cd -
t23% pwd
/usr/maht
t23% 



4th-compiler error reporting
2009-12-06T13:43:00.004Z
Compiling; Word 444: Undefined name

is a bogus error message. The only way to match that word number is to compile the code !!
So I edited the compiler :

4th-3.5d3-unix/sources/comp_4th.c:/^Object->ErrNo = M4NONAME;/
s/$/printf("WORD %s", p);/


so it spits out the duff word
Compiling; Word 444: Undefined name
WORD dup4

which I suppose is ok for a start but I can do better than that but let's not get sidetracked

as an aside, programming in Lunix code is like going back in time. I mean FILE* !! gimme a break.


4th-compiler word list
2009-12-04T22:02:00.003Z
comp_4th.c & exec_4th.c 
is where the action happens


4th-compiler opening plan9 devices
2009-12-04T00:25:00.004Z
This didn't work, no surprise.
Refill introduces an integer.
and USE nuked the stack, so I'm investigating that

First is dial
With minimal comments. I stripped my notes out to see how terse it is without them.

With comments

Dial looks ready to test tomorrow so lets hope this post doesn;t last long.


And I also started draw


4th isatty
2009-11-30T18:32:00.002Z
Just encountered my first isatty annoyance. 


% cat t
6 string buff
buff 5 accept . buff count type cr
buff 5 accept . buff count type cr
% echo ABCDEFabcdef | 4th cqx t
5 ABCDEF
-1
% # wtf?
% cat s
s" /fd/0" input open error? abort" no stdin!"
to fd0
fd0 use
6 string buff
buff 5 accept . buff count type cr
buff 5 accept . buff count type cr
% echo ABCDEFabcdef | 4th cqx s
5 ABCDEF
5 abcdef
% # more like it



Acme helpers
2009-11-30T15:49:00.002Z
4t, C, Limbo

Tags for compiling the current window in the appropriate language

4t is 4th, capital 4 was a bit tricky


FreeBSD Qemu Apache 1.3
2009-11-25T13:23:00.003Z
I need to keep a legacy web server running. Got to Linux and there's no Apache 1.3.x in the Debian package manager. I want to replicate what I have on FreeBSD so .....

FreeBSD with ports installed on Qemu [x] easy. - backup your qimage as you go :)

in root's .cshrc 
# setenv PACKAGESITE ftp://ftp.uk.freebsd.org/pub/FreeBSD/ports/packages/All/
# pkg_add -r apache-1.3.41_1
# cat >> /etc/rc.conf
apache_enable="YES"
^D
# echo 127.0.0.1  `hostname` >> /etc/hosts
This doesn't work :
# cd /usr/ports/lang/php4; make install
because strstr is checked using SIGALARM and that locks up QEMU!
# pkg_add -r php4-4.4.9
Doesn't add mod_php4 or these modules :
# pkg_add -r php4-pcre-4.4.9
# pkg_add -r php4-pgsql-4.4.9_1

So a bit of bodging is required
# cd /usr/local/libexec/apache
# scp maht@steponnopets.net:/usr/local/libexec/apache/libphp4.so .

and the other Apache modules I need

# pkg_add -r mod_gzip-1.3.26.1a
# pkg_add -r mod_auth_pgsql-0.9.12_4


et viola, index.php is excecuted

I copied the confs from steponnopets


AVR assembler
2009-11-20T16:02:00.002Z
I've been working on some rc code to spit out the mnemonics used in AVR Studio. My intention was to use RC as a smarter pre-processor than the #define and .macro directives included therein.

http://www.maht0x0r.net/library/computing/avr/til/ is where I'm at.

I've not included a way to map names onto registers yet, which I'll have to do next.


PXE booting QEMU in Debian
2009-09-30T12:09:00.002+01:00
The PXE booting in Debian QEMU is broken

http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=412010

You need the pxe-* files from http://trac.kju-app.org/browser/trunk/qemu/qemu/pc-bios?rev=106

and put them in /usr/share/qemu

That's all well and good but it uses QEMU's DHCP values for the TFTP server etc. so you need *another* qemu running as the TFTP server and that's not what I want.

So, next step is trying gpxe, so it's off to http://rom-o-matic.net/gpxe/ and got the all-drivers version.

Dammit, I can get 9pxeload working but not bootp so I get boot from ether0 but it can't see 10.0.0.9 from 10.0.2.2

bah


markdown to pdf
2009-09-21T15:44:00.004+01:00
pandoc seems interesting. The debian package didn't satisfy its dependencies though.

first up is texlive-latex-base, only 138Mb!

% markdown2pdfrkdown2pdf ./werc.cat-v.org/docs/web_server_setup/lighttpd.md
markdown2pdf: pdfLaTeX failed with error code 1
markdown2pdf: error context:
! LaTeX Error: File `ucs.sty' not found.

! Emergency stop.
 
         
!  ==> Fatal error occurred, no output PDF file produced!
markdown2pdf: Please install the 'unicode' package from CTAN:
  http://www.ctan.org/tex-archive/macros/latex/contrib/unicode/


ffs, thats no help, check the page out, it assumes one knows what's going on.

ok, on to htmldoc

sigh, another program that ignores pipes by default HELLO THIS IS UNIX

markdown.pl ./werc.cat-v.org/docs/web_server_setup/lighttpd.md| htmldoc --book --footer . -f output.pdf -

works good enough, some font embedding issues but not fatal, though to get it in a pipeline you need to give it an output file with .pdf at the end, IDIOTS

#!/usr/local/plan9/bin/rc
markdown.pl $1 | htmldoc --book --footer . -f /tmp/md2pdf.$pid.pdf -
cat /tmp/md2pdf.$pid.pdf
rm -f /tmp/md2pdf.$pid.pdf



DNDR dns caching proxy
2009-09-15T22:18:00.006+01:00
Although that sounds dangerously like a DNS server I just couldn't face bind/named.

Curiously it's not an OpenBSD binary package, I wonder if there's a reason for that.

http://sourceforge.net/projects/dnrd/
untar it then
 ./configure && make && make install
adduser dnrd
mkdir /usr/local/etc/dnrd
mv /usr/local/share/man/man8/dnrd.8 /usr/share/man/man8/dnrd.8

It will proxy DNS from the IP you specifiy on the command line with -s (required) and also serve out entries on a hosts like file : /usr/local/etc/dnrd/master
so for me it's
/usr/local/sbin/dnrd -b -s 85.189.102.5 -s 85.189.39.5
The -b tells it to round robin the -s entries
ooh why haven't I got round to doing it before, it's fabbo, another one to go in the box marked "essential".

Curiously the master file is only used when -d [1-4] is specified (it doesn't daemonize)
Then because loads of debug is spewing out I ended up installing rc-1.6-readline. I really need to get p9p sorted out.

pkg_add ftp://ftp.plig.net/pub/OpenBSD/snapshots/packages/i386/rc-1.6-readline.tgz
 In the meantime at least I can write sane code.


OpenBSD firewall
2009-09-13T10:47:00.005+01:00
Installed 4.6 on my EPIA 5000 fanless & the HD I have in it is too noisy so I want to run it read only.
I've already posted about /etc/pf.conf

I've enabled syslogd on momo to accept udp syslog posts (syslogd -u) which uses UDP 514

Then in the firewall (goo) I've changed /etc/syslog.conf to send all syslog messages to @momo

Easy peasy.

Next is to mount the filesystems read only which I found here.

change fstab to something like this :
/dev/wd0a / ffs ro,noatime 1 1
swap /tmp mfs rw,noatime,nodev,nosuid,-s=20000 0 0
swap /var mfs rw,noatime,nodev,nosuid,-s=40000 0 0
swap /dev mfs rw,noatime,nosuid,noexec,-s=20000 0 0

do this patch
cp /etc/rc /etc/rc.orig && echo '/fastboot.*XXX
a
mtree -qdef /etc/mtree/4.4BSD.dist -p / -u
cp /root/MAKEDEV /dev; cd /dev; sh MAKEDEV all
.
w
q' | ed /etc/rc



to make changes later 
mount -u -o rw / unless, like me, your CF card reader has a RO jumper !! lucky me
mount -u -o ro / to make it ro again.
I got rid of /usr/ports and ended up with about 500Mb of disk used, so I'm well under the 1Gb target. And I've got stuff on there from where I was putting Mercurial on.

Filesystem     Size    Used   Avail Capacity  Mounted on
/dev/wd0a      158M   32.9M    117M    22%    /
/dev/wd0g      3.5G    3.4M    3.3G     0%    /home
/dev/wd0f      3.5G    490M    2.9G    14%    /usr
mfs:30180      9.3M   37.0K    8.8M     0%    /dev
mfs:17306      9.3M   21.0K    8.8M     0%    /tmp
mfs:28721     18.8M    1.2M   16.7M     7%    /var



mouting a cd in Plan 9
2009-08-31T13:19:00.002+01:00
#!/bin/rc
cdfs
rm -f /srv/cd.d000
9660srv -f /mnt/d000 cd.d000
chmod +r /srv/cd.d000
chmod +w /srv/cd.d000
mount  /srv/cd.d000 /n/cd


Inferno on OpenBSD 4.6
2009-08-31T00:08:00.002+01:00
got mercurial once I got the ports worked out.

ack, all went horribly wrong, no X source, what a todoo


OpenBSD 4.6 packages
2009-08-30T23:52:00.004+01:00
It all started with 
# cd /usr/ports/*/mercurial
# make
===>  mercurial-1.3.1 depends on: python->=2.5,<2.6 - not found
===>  Verifying install for python->=2.5,<2.6 in lang/python/2.5
===>  python-2.5.4p2 uses X11, but /usr/X11R6 not found.
===> Returning to build of mercurial-1.3.1
===>  mercurial-1.3.1 depends on: python->=2.5,<2.6 - not found
===>  Verifying install for python->=2.5,<2.6 in lang/python/2.5
===>  python-2.5.4p2 uses X11, but /usr/X11R6 not found.


er, wtf

Seems the package system has undergone some changes while I've been away.

so the upshot is you need a PKG_PATH set and take binary installs. Ah, the Plan9 way comes to OpenBSD, well done Theo!

The only problem is that you're left to your own devices to find a package store you can trust!

Does it really need so many mirrors. You'd have thought it would find one for you out of the box, perhaps it does, it's not like I've read any documentation.

anyway ratfans, here's mine.

PKG_PATH=ftp://ftp.plig.net/pub/OpenBSD/snapshots/packages/i386


Reverse shell with tcpserver
2009-08-27T13:54:00.003+01:00
I've got an ssh server that is v2 only

on it I run
ssh2srv % tcpserver -v  127.0.0.1 9999 srvrc

srvrc is
#!/usr/local/plan9/bin/rc
/usr/local/plan9/bin/rc <>[2=1]

on a local Lunix box I run
ssh2client % ssh -N -L 192.168.9.2:5999:localhost:9999 ssh2only

then finally on plan9 I run
p9 % srv tcp!192.168.9.2!5999 rc
p9 % con -C /srv/rc

and I get

post...
%

and I have my prompt.


Lunix is your dad.
2009-08-27T12:36:00.001+01:00
"Unix never says please.'" said the cheese.

Some Lunix apps bleat if you make some system files group writeable, some even refuse to run. They email me every week to let me know.

These apps trust neither their ecosystem or their administrators.

Now sit back and wait for "anatomy of my pwnage".


BASH CAN FUCK RIGHT OFF
2009-08-26T15:29:00.001+01:00
When once more one is bitten by a=oh and a = oh being different it is time to stop pretending Bash is a usable product.

That's without thinking abut the runtime penalty of all that shit like history and fucking command line completion that are there to defend you from your badly planned filesystem layout.

A per computer view of the filesystem, shit, you may as well drop to single user.

...

Even hours later my invective hasn't lessened. I've been using it all day. I say using, more like flinging faeces at the plexiglass through which no-one is looking.

Always battling, never a joy.

Emacs or vi is like Beatles or Rolling Stones ?


Plan9 Command Line History
2009-08-25T01:00:00.003+01:00
or rather

"If you would learn to think more clearly Mr. Edison, you wouldn't have to sweat so much"

Command line history rewards sloppy preparation.

fn f { echo do shit; whatis f }; f

losers


Alcatel / Thompson SpeedTouch 330 on OpenBSD
2009-08-24T23:00:00.005+01:00
I've got the green one.

The actual process was a damn sight easier than that blinking instructions suggested.

Get yourself OpenBSD4.6 & install it

Get your source from http://speedtouch.sourceforge.net/
 I used vers 1.3.1 which hasn't seen an update for years which is good.

The Thompson drivers have moved to http://www.thomson.net/GlobalEnglish/Deliver/In-Home-Digital-Distribution/Telco-ISP/dsl-modems-gateways/residential_wired/other_supported_products/thomson_st330/Pages/default.aspx

Though I got mine pre-packaged from here
http://speedtouch.hysplace.co.uk/latest/Kqd6_3.012

I chopped all of the usb startup out of  speedtouch-start :

# Goal: Start the connection

# Ensure to have a working PATH
export PATH="/sbin:/bin:/usr/sbin:/usr/bin:/usr/local/sbin:/usr/local/bin"

# check for root privileges

#
# FIRMWARE_LOADER_FILE
# FIRMWARE_FILE
# PPPD_PEER
#

if [ -f /etc/speedtouch/speedtouch.conf ]; then
  . /etc/speedtouch/speedtouch.conf
fi

the rest was the same

speedtouch.conf is

FIRMWARE_LOADER_FILE="/etc/speedtouch/Kqd6_3.012"
FIRMWARE_FILE="/etc/speedtouch/Kqd6_3.012"
MODEM_RUN_OPTIONS=""
PPPD_PEER=speedtouch-pppoa


/etc/ppp/chap-secrets isn't :>
"alphas@example.dsl" "*" "W00t"

/etc/ppp/ppp.conf
these are supposed to be indented with just the ones with a colon in column zero, like a c label

default:
        ident user-ppp
        set log Phase Chat IPCP CCP tun command
        set ifaddr 10.0.0.1/0 10.0.0.2/0

adsl:
        set authname alphas@example.dsl
        set authkey W00t
        set device !"/usr/local/sbin/pppoa2 -vpi 0 -vci 38 -v 1"  # find your own, this is BT UK on griffin adsl
        accept chap
        set speed sync
        set timeout 0
        set redial 15 10000
        set dial ""
        add default HISADDR
        enable dns

Connection:


/etc/rc.local

/usr/local/bin/speedtouch-start
ppp -nat -ddial adsl


/etc/sysctl.conf
net.inet.ip.forwarding=1        # 1=Permit forwarding (routing) of IPv4 packets
net.inet6.ip6.forwarding=1      # 1=Permit forwarding (routing) of IPv6 packets


Das Brickenn vas Goot
2009-08-24T19:29:00.007+01:00
EPIA as a router, got the OpenBSD 4.5 disk, got it installed once I reset the partitions as the default it gave me for an 8gb can't install the ports on itself!

First thing I have to do is add packet forwarding 

# echo "net.inet.ip.forwarding=1" >>  /etc/sysctl.conf
# echo "net.inet6.ip6.forwarding=1" >>  /etc/sysctl.conf

I guess I should play with IPv6 some time, anyway lets go nattin
/etc/pf.conf

ext_if  = "tun0 "
int_if  = "vr0"
int_net  = "vr0:network"
lan_net = "192.168.9.0/24"

momos = "{ 993 25 22 }"
pigs = "{ 80 222 }"

# nat on $int_if to any -> $(ext) - ppp is natting for me
# even though my IP is fixed, it's still assigned by dhcp

table  const { self }
set skip on lo0
# it doesn't like the next line
# scrub in all
block all
block in quick from urpf-failed
pass in  on $int_if from $lan_net to any
pass out on $int_if from any to $lan_net
pass out on $ext_if proto { tcp udp icmp } all modulate state
pass in on $ext_if proto tcp from any to !  port ssh synproxy state
rdr log on $ext_if proto tcp from any to any port $momos -> 192.168.9.14
pass in on $ext_if proto tcp from any to any port $momos
rdr log on $ext_if proto tcp from any to any port $pigs -> 192.168.9.6
pass in on $ext_if proto tcp from any to any port $pigs



dumping venti arenas
2009-08-23T22:48:00.004+01:00
It's about time I dumped my venti arenas.

Here's the script I'm using to read the arenas.


#!/bin/rc

! test -w /dev/sdC0/arenas && status='not hostowner' && exit

rfork e

echo `{date -n} `{fossil/last $1/fossil} >> vacs

# sort this bit out to cd to the backup dir
9fs momo
cd /n/momo/home/t23/arenas
# in backup dir

fn do{
 b=`{echo $1 | sed -e  's!/dev/!!g' -e  's!/arenas!!g'}
 b=$b(1)
 if(~ $#3 0)
  venti/rdarena $1 $2 > $b.$2
 if not 
  if(! test -e $b.$3)
   venti/rdarena $1 $2 > $b.$2.$3 && rm -f $b
}

fossil/last >> vacs

hget http://127.1:8000/index | 
awk '
/^arena=/ { arena=substr($1, 7); score=""}
/^ arena=/ { dev=$3 }
/^ score=/ { score=substr($1, 7) }
/^ written:/ { print "do " dev " " arena " " score }
' |rc


which is adapted from Russ' restore script.
I added the hostowner check and I' not sure it works so note that

next up I have to work out if the final score is checkable outside venti, i.e it could be the hash of the whole arena.


blogger fs new tool
2009-08-19T21:57:00.001+01:00

% cat /bin/blg
#!/bin/rc

echo -n 'Title: '
title = `{read}
echo -n 'Tags: '
tags = `{read}
echo -n 'Draft: '
draft = `{read}
if(~ $draft Yes) draft = 'Draft: yes
'
{
	echo 'Title: ' $"title
	echo 'Tags: ' $"tags
	echo -n $draft
	cat
} > /n/blogger/new


The slight use advantage of the FS for this is that you can export it once you have authenticated and hand it to a service.

Off the top of my head you could run fossil/last and save it as a draft post on blogger.

It also means you can use blogger as a storage system for your CP


New Post
2009-08-18T23:39:00.001+01:00
Text for new post


write only group blog
2009-08-18T23:37:00.001+01:00
quad% import t23 /srv/ /n/bsrv
quad% mount /n/bsrv/blogger /n/blogger
quad% { echo 'Title: New Post'; echo 'Text for new post' } > /n/blogger/new


bloggerfs in action
2009-08-18T23:18:00.001+01:00
How plan9 people do stuff

Set up your blogger account with factotum (typing your password at the prompt)

% g=maht.lawless@googlemail.com
% auth/factotum -g 'proto=pass server=www.google.com user='$g' !password?'

!Adding key: proto=pass server=www.google.com user=maht.lawless@googlemail.com
password: 
!

% blogger/blogs $g
3062883415508658133	compjootery
% blogger/actionfs $g  3062883415508658133
% { echo 'Title: Date demo'; echo 'Tags: demo, blogger' ; date} > /n/blogger/new
% cat /n/blogger/posts
3931523848251033590 Date demo
% cat /n/blogger/3931523848251033590
Title: Date demo
Tags: blogger, demo
Tue Aug 18 21:01:07 GMT 2009
% { echo 'Title: Date demo'; echo 'Tags: demo, blogger' ; date} > /n/blogger/3931523848251033590
% cat /n/blogger/3931523848251033590
Title: Date demo
Tags: blogger, demo
Tue Aug 18 21:01:25 GMT 2009
%



Fixed some bugs in bloggerfs
2009-08-18T19:42:00.001+01:00
blogger/read | blogger/update now works as it should
This entry is posted with it.


Sounds come to my plan9 terminal
2009-08-18T19:16:00.001+01:00
The Edirol UA-3D works great for dfm 
hget `{hget http://dfm.nu/dfm_1.m3u | grep http} | madplay -
Here's the stats
quad% usb/audio -V
Audio output unit 1
Dev can record from Line connector
Audio input unit 13
Dev can play to USB Streaming
Audio Selector Unit 11
Audio Feature Unit 9 for Playback
	mute control on master channel
	volume control on channel 1
	volume control on channel 2
Interface 1: mono, 16 bits, discrete frequencies are: 32000 44100 48000
Interface 1: stereo, 16 bits, discrete frequencies are: 32000 44100 48000
Interface 2: mono, 16 bits, discrete frequencies are: 32000 44100 48000
Interface 2: stereo, 16 bits, discrete frequencies are: 32000 44100 48000
usb/audio: playback on ep 2
usb/audio: record on ep 2
usb/audio: buttons on ep 1
Setting default play parameters: 44100 Hz, 2 channels at 16 bits
Setting default record parameters: 44100 Hz, 2 channels at 16 bits
Warning, can't configure record for 44100 Hz, configuring for 48000 Hz instead
Playback speed control, value 44100
Playback mute control, value 0
Playback volume control, min -5888, max 2048, step 256, values[1] -2048  [2] -2048  
Playback channels control, min 1, max 2, value 2
Playback resolution control, min 16, max 16, value 16
Record speed control, value 48000
Record channels control, min 1, max 2, value 2
Record resolution control, min 16, max 16, value 16


Here's the stats for the Creative Soundblaster USB
Note the : Warning, can't configure playout for 44100 Hz, configuring for 48000 Hz instead

Audio output unit 1
Dev can record from Microphone
Dev can record from Line connector
Dev can record from S/PDIF
Audio input unit 7
Dev can play to USB Streaming
Audio Selector Unit 8
Audio Feature Unit 9, not known what for
	mute control on master channel
	volume control on channel 1
	volume control on channel 2
Audio Feature Unit 10 for Record
	mute control on master channel
	volume control on master channel
	agc control on master channel
Audio Feature Unit 11 for Record
Second feature unit (10, 11) on record
	mute control on master channel
	volume control on channel 1
	volume control on channel 2
Audio Feature Unit 12 for Record
Second feature unit (11, 12) on record
	mute control on master channel
Audio Feature Unit 13 for Record
Second feature unit (12, 13) on record
	mute control on master channel
	volume control on master channel
Audio Feature Unit 14 for Record
Second feature unit (13, 14) on record
	mute control on master channel
	volume control on channel 1
	volume control on channel 2
Audio Mixer Unit 15
	Mixer 15 for I & O 
	Mixer 15 for I & O 
Interface 1: stereo, 16 bits, discrete frequencies are: 48000
Interface 2: stereo, 16 bits, discrete frequencies are: 48000 44100
usb/audio: playback on ep 1
usb/audio: record on ep 1
Setting default play parameters: 44100 Hz, 2 channels at 16 bits
Warning, can't configure playout for 44100 Hz, configuring for 48000 Hz instead
Setting default record parameters: 44100 Hz, 2 channels at 16 bits
Warning, can't configure record for 44100 Hz, configuring for 48000 Hz instead
Playback speed control, min 48000, max 48000, value 48000
Playback channels control, min 2, max 2, value 2
Playback resolution control, min 16, max 16, value 16
Record speed control, min 48000, max 44100, value 48000
Record mute control, value 1
Record volume control, min 0, max 12240, step 48, values[1] 9168  [2] 9168  
Record agc control
Record channels control, min 2, max 2, value 2
Record resolution control, min 16, max 16, value 16







And acme posts
2009-08-16T19:35:00.001+01:00
I'm going to upload what I have so far and leave the rest to fate.



Date demo
2009-08-16T17:57:00.001+01:00
Tue Aug 18 22:29:27 GMT 2009


post from bloggerfs
2009-08-16T10:45:00.001+01:00
A new darn post

Now with editing in Acme



Hello from plan9 python
2009-08-16T09:07:00.001+01:00
bloggerfs incoming


HG Unwells?
2009-08-14T15:33:00.005+01:00
(while making this and feeding back, the packages are being changed so its worth checking

--rw-rw-r-- M 149 bichued sys 5510112 Oct 17 2008 python-2.5.1-ape.tgz
quad% md5sum python-2.5.1-ape.tgz
ac73a8b4fcb38b966e81fe8ad3aa50e1 python-2.5.1-ape.tgz
% 
)

ok, somehow bichued is involved, that dastardly hacker

http://plan9.bell-labs.com/sources/contrib/bichued/python/

He's got hg, python the darn lot going on in there

The excitement is killing me as it copies. I wonder if I'll ever write copy-with-progress, everything is just waiting for it.

Oh I added python to linuxemu already, ver 2.3.5, handy little box I suppose. 

The instructions for it are in /n/sources/contrib/bichued/python-2.5.1-ape.tgz
cat Readme.Plan9

got some contribs to install (ooh first time I have fgb/contrib tools, I am honoured)

the order in the list of contribs is wrong
lcms depends on tiff
and png depends on z
tiff & xml depend on z

openssl is a monster


k now I need to patch ape
ramfs
bind -ac /tmp /n/sources/patch/ape-chmod-dirbit
patch/apply ape-chmod-dirbit
bind -ac /tmp /n/sources/patch/applied/ape-ftruncate
patch/apply applied/ape-ftruncate

I got some perm errors on the last one but the update seems to have worked

The mkfiles need adjustment for python-2.5.1-ape, I got away with this :
% cd python-2.5.1-ape
% grep 'CFLAGS=' `{walk | grep mkfile}
Extra/mkfile:CFLAGS=-c -I.. -I../Include -DT$objtype -DNDEBUG -D_SUSV2_SOURCE
Modules/mkfile:CFLAGS=-c -I. -I.. -I../Include -DT$objtype -DNDEBUG -DPLAN9 -D_SUSV2_SOURCE 
Objects/mkfile:CFLAGS=-c -I.. -I../Include -DT$objtype -DPy_BUILD_CORE -DNDEBUG
Parser/mkfile:CFLAGS=-c -I.. -I../Include -DT$objtype -DPy_BUILD_CORE -DNDEBUG
Python/mkfile:CFLAGS=-c -I.. -I../Include -DT$objtype -DPy_BUILD_CORE -DNDEBUG
mkfile:CFLAGS=-FVwN
mkfile:CFLAGS=-c -I. -IInclude -DT$objtype -DPy_BUILD_CORE -DNDEBUG -D_SUSV2_SOURCE
% mk install
...
% python
Python 2.5.1 (r251:54863, Aug 15 2009, 21:45:20) [C] on plan9
Type "help", "copyright", "credits" or "license" for more information.
>>>

now for hg

bind python-2.5.1-ape /n/python
; bind -a /n /tmp
; cd hg-1.0.2
; mk install

; hg -v
Mercurial Distributed SCM (version 1.0.2)

Adventures in Linuxemu

2009-08-13T23:14:00.007+01:00

Installation is exactly as cinap says

But the apt sources have moved
deb http://archive.debian.org/debian sarge main

Now apt-update worked

apt-get install host
got me a working host

apt-get install python
would have got me python2.3 but I cancelled it

Upgrading to etch is not possible because etch requires the "dswap" instruction

adding privoxy to test sockets

apt-get -y install privoxy
it's also installing adduser & cron

adduser is failing to lock /etc/passwd so wont add new users

trying apt-get -y install ssh
it asks you if you want to disable ssh 1 buy i can only say yes!

worse than that, it ignores the -y

So I went for proftpd and it's hauled in the army - 6M expanding to 25M - no questions asked!

Same option problem so I mashed the keys and it worked but I don't know what I chose, inetd or standalone

dammit, that wants user ftp, didnt think of that

should have tried nc instead

anyway gone for vnc4server which drags in X !

can't open display

Net booting Plan9

2009-08-12T18:38:00.002+01:00

This *should* be easy, eventually it was!

NVidia Bootagent doesn't like it but IBM Boot Agent v.1.2.2 is a worker

/cfg/pxf/ is being picked up
it needed a ether= ip= in /lib/ndb/local

after working out the pxe file I ended up with :

ether0=igbe
monitor=vesa
vgasize=800x600x8
mouseport=ps2intellimouse
user=maht
bootfile=ether0!/386/9pc
bootargs=tcp!t23
nobootprompt=tcp!t23
*nomp=1 # vesa graphics boo hoo

I added secstore to the server and added myself as a user

I fiddled around with lib/profile and added factotum
now you get prompted for your secstore password twice, profile then imports your home directory using your auth and you can write to it. Means you have to shuffle your profile round a little bit, mine's now


fn bins {
bind -a $home/bin/rc /bin
bind -a $home/bin/$cputype /bin
bind -c $home/tmp /tmp
bind -a /usr/inferno/Plan9/386/bin/ /bin
}

font = /lib/font/bit/lucidasans/unicode.8.font
prompt=($sysname^'% ' ' ')
fn $sysname%{ $* }
fn acme { /bin/acme -f /lib/font/bit/lucidasans/unicode.7.font $* }
timezone=`{cat /adm/timezone/local}
switch($service){
case terminal
 auth/factotum
 import t23 / /n/t23
 bind -c /n/t23/usr/$user /usr/$user
 bind -c /n/t23/usr/$user/tmp /tmp
 bind -c /n/t23/usr/inferno /usr/inferno
 bins
 plumber
 upas/fs
 echo -n accelerated > '#m/mousectl'
 echo -n 'res 3' > '#m/mousectl'
 exec rio
case cpu
 bins
 if (test -e /mnt/term/mnt/wsys) {
  # rio already running
  wsys = /mnt/term^`{cat /mnt/term/env/wsys}
  bind -a /mnt/term/mnt/wsys /dev
  echo -n $sysname > /dev/label
 }
 bind /mnt/term/dev/cons /dev/cons
 bind /mnt/term/dev/consctl /dev/consctl
 bind -a /mnt/term/dev /dev
 upas/fs
 news
 if (! test -e /mnt/term/mnt/wsys) {
  # cpu call from drawterm
  plumber
  auth/factotum
  exec rio
 }
case con
 bins
 news
}

Updated compiling Inferno

2009-07-29T11:31:00.002+01:00

The moved to Mercurial is not without its problems - empty directories, missing build targets, but at least it is more transparent these days. I think people overlook what a tricky task it has been choosing and maintaining a distribution system that works for everybody.

Anyway I edited my old entry so that there weren't two

Serialising multiple writers in the shell

2009-06-20T18:26:00.012+01:00

A few years ago I came across wait_on for monitoring file changes on FreeBSD (from the kqueue) in the shell. Linux has inotify-tools for using inotify the same way. Using these tools we can build a fifo daemon with multiple writers and one reader. I'm not sure if Plan 9 already has a way to do this in userland, maybe even by using snoopy.

Here's the Linux version I use.


% cat /usr/local/bin/manage_queue
#!/usr/local/plan9/bin/rc

# we assume that / will never appear in a filename and thus is a safe delimeter
DELIMETER=/

fn waiter {        
        inotifywait  -q $3  --format '%f'^$DELIMETER -e $2 $1
}

fn watch {
        if(~ $1 -1) # do it once
                waiter $*(2-)
        if not # do it forever
                waiter $1 $2 -m
}

fn process {
        read_until $DELIMETER | sed 's/.$//' | $*
        /usr/local/plan9/bin/dd -ibs 1 -count 1 >[2] /dev/null
}

# start the processing cmd in the queue directory
if(~ $1 -1) { # do it once
        cd $2
        watch -1 $2 $3 | process $*(4-)
}
if not { # ad infinitum
        cd $1
        watch $1 $2 | while() { process $*(3-) }
}

and you invoke it thus :
su -c 'manage_queue /tmp/uploaded_images/ moved_to make_thumbnails' uploadprocessor
although I use daemontools so that is the contents of /etc/service/uploads/run

My other usual option is close_write instead of moved_to, it depends how your files get there and are processed once they arrive.

The -1 option is for doing a spot of debugging, it does one file and then exits

The bit of script missing from there is read_until



% cat /usr/local/bin/read_until
#!/usr/local/plan9/bin/rc

ifs=()

while() {
        c=`{/usr/local/plan9/bin/dd -ibs 1 -count 1 >[2] /dev/null}
        if(~ $#c 0)
                echo -n ' '
        if not
                switch($c) {
                case $1
                        exit
                case *
                        echo -n $c
                }
        }
}

I had to make that because newline is a valid filename character thus defeating various line reading commands.

Now you've got the thing going, you'll need a processor like make_thumbnails, which I've made up here as an example (gm is graphicsmagick which kept the older api when ImageMagick changed):


#!/usr/local/plan9/bin/rc
ifs=()
fin=`{cat}
gm convert -resize 100x100 /tmp/uploads/$fin /var/www/thumbs/$fin.jpeg
gm convert -resize 350x500 /tmp/uploads/$fin /var/www/mids/$fin.jpeg

So, using this mechanism the user uploads the image via a web browser, when the CGI is satisfied it can do 'mv /uploadir/$up /tmp/uploads/$up' and carry on sending html to the user, the thumbnail will get made when it gets to the front of the queue.

The other issues this can solve is ring fencing users from each other and just communicating via files and filenames (of which the above is an example but also internal users). www can write to /tmp/downloads but not /var/www/*. By passing the uploaded image through gm it gets (hopefully) sanitized, by running gm as its own user that mitigates attacks against gm, one can use the OSes limiters to let gm only have so much cpu etc.

By serialising like this it also protects against a DoS uploading loads of massive images and trying to thumbnail them in parallel. You've still got the problem of processing them but at least only one process will be pegged at 100% not one per upload.

Though I've not done it yet, you could also use it as a marshall which does some load balancing (perhaps using Xcpu), or just forked threads.

--

Strip invalid characters UTF-8 from a bytestream

2009-05-16T12:18:00.002+01:00

Found this at http://snippets.dzone.com/posts/show/6194


iconv -c -f UTF-8 -t UTF-8 < invalid > valid

VNC serving

2009-04-10T18:04:00.004+01:00

if you want passwordless vnc, use Debian package vnc4server

vncserver -SecurityTypes None

Build your own Inferno

2009-04-10T16:44:00.019+01:00

first you'll need mercurial

apt-get install mercurial
then get the code

cd /usr/local; hg clone https://inferno-os.googlecode.com/hg/ inferno-os

edit mkconfig
ROOT=/usr/local/inferno-os
SYSHOST=Linux
OBJTYPE=386

edit makemk.sh
ROOT=/usr/local/inferno-os
SYSTARG=Linux

Atm. you have to make some empty directories because hg doesn't like those !

mkdir /usr/local/inferno-os/Linux/386/lib/

% sh makemk.sh

[ n.b.

(11:42:04) mjl-: mahjadin: i think that on linux/amd64, you'll need a mk nuke inbetween makemk.sh and mk install, makemk.sh didn't use -m32 last time i checked, so mk install will try to link all -m32 object files against the few amd64 libs that makemk.sh created.
(11:42:11) mjl-: but i'll have to check that.
...
(11:51:08) mjl-: this is the error:
(11:51:21) mjl-: //usr/bin/ld: warning: i386:x86-64 architecture of input file `/home/mjl/inf.test/Linux/386/lib/libbio.a(bgetc.o)' is incompatible with i386 output

]

Then add the inferno bin to your path for mk and later for other tools such as iyacc

PATH=$PATH:/usr/local/inferno-os/Linux/386/bin
mk install

bah, failed with something different this time

(cd Linux; mk install)
cc -c -m32 asm-386.S
mk: no recipe to make 'devtab.o'
mk: echo "(cd $SYSTARG; ... : exit status=exit(1)
mk: for j in ... : exit status=exit(1)

ok I solved that one
touch /usr/local/inferno-os/emu/port/devtab.c

on to the next

touch /usr/local/inferno-os/dis/lib/oldauth.dis

finally mk install will work

Though you have to pretend to yourself none of those compiler warnings matter as they fly past.

Other directories to make

mkdir OpenBSD/arm
mkdir `{hget 'http://inferno-os.googlecode.com/issues/attachment?aid=-8986372491888485913&name=empties'}
chmod 555 mnt/* n/client/* n/*
chmod 755 n/client
chmod 555 prof

you'll get an error 'cos we made it earlier

mkdir: ./Linux/386/lib already exists

ohoh fonts

can't auto-download cdinstall.iso any more, have to guess at the release date (unless this bit of p9p script works)

cd /tmp
dist=`{hget http://www.vitanuova.com/inferno/downloads.html | grep '/dist/4e/' | sed -e 's/^[^"]+"//' -e 's/".*//'}
hget $dist | gunzip | tar x inferno/fonts; @{cd inferno/fonts && tar cp .} | @{cd /usr/local/inferno-os/fonts/&& tar xT} ; rm -rf inferno

chmod 777 /usr/local/inferno-os/tmp

For plan9 is is virtually the same, not all the missing bits are missing, you'll work it out.

As a new user you'll need to do this (probably as root):
cd /usr/local/inferno-os/
mkdir usr/$user
mkdir usr/$user/lib
cp usr/inferno/lib/plumbing usr/$user/lib
touch usr/$user/lib/keyring
chown -R $user usr/$user

more trig

2009-03-24T11:01:00.002Z

for a half circle

for a point X, Y

if X < w / 2

x = ( w / 2 ) - X

a = atan(y / x)
R = sqrt(x * x + h * h)
scale = h / R

Projection trig

2009-03-23T23:57:00.003Z

the throw ratio (R) is defined as distance to screen (D) divided by the width of the image (W)

or rather R = D / W
As we normally have R, are constrained by D and want to know W rewrite that as

W = D / R

if we assume D = 1 then W = 1 / R
for atan we only use half the triangle for the angle from straight :
thus
a = atan (W / 2) = atan ( (1 / R) * (1 / 2) ) = atan ( 1 / 2R )
ergo the angle of the cone A = 2 * a = 2 * atan ( 1 / 2R )
for a Ratio of 0.8, A = 2 * atan( 1 / 1.6) = 64.011 degrees

mpeg2enc & dvdauthoring on debian

2009-03-23T13:16:00.006Z

basically you have to install
http://debian-multimedia.org/

apt-get install mjpegtools mkisofs dvdauthor ffmpeg

2mpeg


#!/usr/local/plan9/bin/rc 

f = 8 # for dvdauthor otherwise go with 3
br = 7500
a = 3
q = 5

while(~ $1 -*) {
 switch($1) {
 case -b
  br = $2
  shift 2
 case -f
  f = $2
  shift 2
 case -a
  a = $2
  shift 2 
  # 1 - 1  - 1:1 display
  # 2 - 2  - 4:3 display
    #  3 - 3  - 16:9 display
   #       4 - 4  - 2.21:1 display
 case -i
  deinterlace = '-deinterlace'
  shift
 case -q
  q = $2
  shift 2
 }
}

paff = $1
shift
fname = `{basename $paff}

2mp2 $paff &
2m2v -q $q -a $a -f $f -b $br $paff $deinterlace &
wait

mplex -f 8 -o $fname.mpeg $fname.m2v $fname.mp2 && rm $fname.m2v $fname.mp2

2mp2


#!/usr/local/plan9/bin/rc

mplayer -vc null -vo null -ao 'pcm:fast:file=' ^$1.wav $1
mp2enc  < $1.wav -o `{basename $1} ^.mp2 && rm $1.wav

2wav


#!/usr/local/plan9/bin/rc

aid = ''
while(~ $1 -*) {
 switch($1) {
 case -aid
  aid = '-aid ' $2
  shift 2
 }
}

mplayer  $aid -vc null -vo null -ao 'pcm:fast:file=/dev/fd/1' -

make some mpegs : 2mpeg 1.mov
2mpeg 2.mov
so you'll end up with 1.mov.mpeg 2.mov.mpeg
ffmpeg can deduce your aspect if you don't want to specify it
then make an xml (dvd.xml) file thus :


<dvdauthor dest="dvd">
<vmgm />
<titleset>
 <titles>
  <video format="pal" aspect="4:3" />
  <audio format="pcm" lang="EN" />
  <subpicture lang="EN" />
  <pgc>
   <vob file="1.mov.mpeg" />
   <vob file="2.mov.mpeg" />
  </pgc>
 </titles>
</titleset>
</dvdauthor>

then


dvdauthor -x dvd.xml
mkisofs -dvd-video -udf -o dvd.iso dvd
wodim dev=/dev/cdrw1 -dao dvd.iso

done

Mplayer OSX - working

2009-03-17T12:03:00.006Z

Anothy_x reminded me about MacPorts

yay, mplayer working

except .... getting it to run at boot. putting it in login_items you have to associate it with terminal and then the mouse being in the top left means the mplayer menu bar is across the top.

yay, done it. The power of Java moving the mouse off screen

import java.awt.AWTException;
import java.awt.Robot;
import java.awt.event.KeyEvent;

public class RM {    
    public static void main(String[] args) {        
        try {            
            Robot robot = new Robot();
            robot.delay(5000);
            robot.mouseMove(100, 10000);            
        } catch (AWTException e) {
            e.printStackTrace();
        }
    }
}

I presume the class file has to be in the classpath, Looks like that includes .


#!/bin/bash
cd /Users/broadway/bin
java RM RM.class
mplayer -playlist /Users/broadway/media/playlist

Yay, remove the bullet and sew up the wound; I'm out of this crazy hospital.

Mplayer OSX

2009-03-16T18:42:00.004Z

Adapted from http://jonsimpson.co.uk/log/2005/mplayer_cvs_osx_tiger
FAIL

shoot me

sshd on MacMini / OSX 10.4.11

2009-03-16T11:06:00.004Z

On the machine I got control of there were no ssh keys

nothing unusual going on, keys are in /etc instead of /etc/ssh

ssh-keygen -t rsa1 -q -f /etc/ssh_host_key
ssh-keygen -t rsa -q -f /etc/ssh_host_rsa_key
ssh-keygen -t dsa -q -f /etc/ssh_host_dsa_key

/usr/sbin/sshd

getting it to startup on boot - Preferences - Network - Remote Access [x]

I suspect that it would have generated the proper keys if I'd done that first !

Building p9p from hg

2009-03-11T17:56:00.005Z

FAIL

I spend more time setting up the environment than doing the damn work when I get there!
There was once a golden age where I could say to someone "I need a machine with Apache and Postgresql and CGI enabled for a domain foo.com" and they would email me when it was working. Happy days, but nothing to blog about!

Ok so now I have a plan9port from Mercurial and it doesn't have some cmds built like awk and others I've not even looked at.

First off is I can't build the awk because I don't have mk and I can't build mk because I don't have mk!

Luckily VitaNuova have their source tree available and for the moment the source tree is this file but it is referenced by date.

To make it work you have to unpack it in /usr and then I couldn't build it anyway. Aha but mk is in /usr/inferno/Linux/386/bin/mk and runs so I copied it to /usr/local/bin/mk

cd /usr/local/plan9/src/cmd/awk
which now fails with


warning: skipping missing include file: /src/mkhdr: No such file or directory
warning: skipping missing include file: /src/mkone: No such file or directory

because of <$PLAN9/src/mkhdr
even if $PLAN9 is set so the solution is PLAN9=$PLAN9 mk
but more joy


9 yacc -S -d -v awkgram.y
/usr/local/plan9/bin/9: line 17: exec: yacc: not found
mk: 9 yacc -S ...  : exit status=exit(127)

again
cp /usr/inferno/Linux/386/bin/yacc /usr/local/bin/

PLAN9=$PLAN9 mk
9 yacc -S -d -v awkgram.y

conflicts: 42 shift/reduce, 83 reduce/reduce
9c  re.c
9c  lex.c
9c  main.c
9c  parse.c
9c  maketab.c
9c  tran.c
9c  lib.c
9c  run.c
9c  y.tab.c
9l -o o.maketab maketab.o
/usr/bin/ld: cannot find -l9
collect2: ld returned 1 exit status
mk: 9l -o o.maketab maketab.o  : exit status=exit(1)

time to download http://swtch.com/plan9port/plan9port.tgz ! and

apt-get install xorg-dev
Need to get 14.7MB of archives.
After this operation, 47.5MB of additional disk space will be used.
Do you want to continue [Y/n]? y

Working the Werc

2009-03-09T12:58:00.006Z

I'm giving Werc a try out, I hope the fella's done a good job.

First shizzle I need is an hg client do get a checkout, I notice there's a lot of checkins so it's better this way I reckon. Plus I'll be a contributor before long.

Hg only needs 16.8Mb of files !! hgsvn is the package name - hope it's the right one.

slap:/var/www# rmdir werc
slap:/var/www# hg clone http://gsoc.cat-v.org/hg/werc/

ok that worked out

lighttpd.conf

need to add "mod_cgi", to the modules section or run : lighty-enable-mod cgi

$HTTP["host"] =~ ".*feedideas.$" {
    index-file.names = ( )
    server.error-handler-404 = "/werc.rc"
    alias.url       += ( "/werc.rc" => "/var/www/werc/bin/werc.rc" ) # Here use the path to werc.rc in your system.
    cgi.assign += ( ".rc" => "")
    server.dir-listing = "disable"
}

Then

mkdir /var/www/werc/sites/feedideas
echo HelloWorld > /var/www/werc/sites/feedideas/index.html
/etc/init.d/lighttpd restart

Then you can begin the task of skinning your website so it doesn't look like the paedophile cat-v ;)

Apparently werc utilizes markdown but when I tried it, I got literals.

Not totally happy with the directory structure but been working with Uriel on moving a few functions around so you can overload your own.

O, that way madness lies.

2009-03-03T23:38:00.008Z

Thanks Skip for the title.

#include <u.h>
#include <libc.h>
#include <draw.h>
#include <cursor.h>

Image *col[2];

void
main(void) {
 Point zero;
 zero.x = 0;
 zero.y = 0;

 newwindow("-r 0 0 700 300");
 if(initdraw(nil, nil, "tri") < 0){
  fprint(2, "stats: initdraw failed: %r\n");
  exits("initdraw");
 }

/* code in the boxes goes here */
 col[0] = allocimage(display, Rect(0,0,1,1), RGB24, 1, setalpha(0xFF0000FF, 0xFF));
 col[1] = allocimage(display, Rect(0,0,1,1), RGB24, 1, setalpha(0x00FF00FF, 0xFF));
 draw(screen, screen->r, col[0], nil, zero);
 draw(screen, Rect(0,0,350,150), col[1], nil, zero);
/* to here */
     
 flushimage(display, 1);
 sleep(10000);
 closedisplay(display);
}

This is the image with no alpha

As expected, apply alpha to the red or the green and the intensity diminishes. No blending.

But with RGB24 you can't do Porter-Duff operations. But when you apply alpha to RGBA32 images something unexpected happens :

I can't work out how that transformation even occurs - The green turns to fd007f, the red to 7f80ff. Where does the blue come from?

If I drawop them too I get :

Green = bd40ff
Red = 7f80ff

I'm afraid it defies my comprehension!

serving 9p from debian

2009-03-02T15:25:00.005Z

Despite a tried and tested u9fs working, I went the npfs route.

No docs.

I prefer tcpserver over inetd so # apt-get install ucspi-tcp

naively I tried slap:/home/maht/spfs# tcpserver -v 10.0.0.1 564 fs/ufs

mount failed : Hangup

back to u9fs, cba pissing about
can't find a download for u9fs at sourceforge - though I can find the cvs browser so I uploaded it to my webspace

once you've built u9fs with make


cp u9fs /usr/local/bin/u9fs-bin
touch /usr/local/bin/u9fs
ed /usr/local/bin/u9fs
a
#!/bin/sh

u9fs-bin -a none
.
w
q

tcpserver 10.0.0.1 564 u9fs

and away you go
(you'll need your plan9 users as users on Debian and you trust the names)

so you could use -a p9any and use /etc/u9fs.key
I'll check it works in a bit when my 4Gb has finished copying - nope failed for me - giving up


time cp linux /n/slap/home/glenda
0.88u 105.31s 2176.35r cp linux /n/slap/home/glenda

which is 36.26 mins

Installing Debian 5 to AoE

2009-03-02T13:27:00.004Z

before you copy this procedure, the initrd.aoe didn't work out, I've got to inspect it an compare it against the one that worked for etch

haha found aoe.ko on the install disk
on the vblade server aptitude vblade

 
dd if=/dev/zero of=lenny bs=1048576  count=4096
vblade 3 1 eth0 lenny
dd if=/dev/zero of=home bs=1024 count=204800
vblade 3 2 eth0 home
dd if=/dev/zero of=swap bs=1024 count=204800
vblade 3 3 eth0 swap

run the installer until you can get a shell (alt-f2)
which for me was "configure the network"


udpkg --unpack /cdrom/pool/main/l/linux-2.6/linux-image-2.6.26-1-486_2.6.26-13_i386.deb
insmod /lib/modules/2.6.26-1-486/kernel/drivers/block/aoe/aoe.ko

cd /dev
ln -s /dev/etherd/e3.1 hda
ln -s /dev/etherd/e3.2 hdb
ln -s /dev/etherd/e3.3 hdd

The installer will moan that it can't install GRUB and give you the choice to skip along, do that.
Eventually you get "Installation is complete" and a
choice.

Now it's hack time. Crack open the initrd.img the installer just made


chroot /target
mkdir /tmp/initrd && cd /tmp/initrd
gunzip < /initrd.img | cpio -i --make-directories

I got the scripts from http://www.etherboot.org/wiki/sanboot/debian_and_ubuntu


apt-get install aoetools
cp /sbin/aoe-discover /tmp/initrd/sbin/
cd /etc/initramfs-tools/
wget http://www.proweb.co.uk/~matt/aoe/hooks -O hooks/aoetools
wget http://www.proweb.co.uk/~matt/aoe/scripts -O scripts/aoetools
chmod 755 hooks/aoetools scripts/aoetools
cp scripts/aoetools /tmp/initrd/scripts/

Then make the initrd.img again


cd /tmp/initrd
find . | cpio -o -H newc | gzip -9 > /boot/initrd.aoe

Then get it off the machine and on to the TFTP server.
scp /boot/initrd.aoe /boot/vmlinuz-2.6.6.26-1.686 TFTP-server:/var/lib/tftpboot/

Now to taste the pudding

NFS stole my afternoon

2009-02-27T10:55:00.009Z

NFS is awful for using as part of the boot sequence. It's half ok once the thing is up, but mounting /usr and /var via NFS at boot was a fail.

An afternoon wasted, thanks

yay for AoE

Creating a vblade for Debian Lenny

2009-02-26T17:14:00.005Z

Getting the right aoe.ko to install straight onto an aoe target is (probably) a lot of fannying
And I think this is the starting point. At least now I know how to run Qemu better so I don't need to install onto a real PC

Download the netinst iso from debian - debian-500-i386-netinst.iso

make a qcow disk
slap:/home/qemus# qemu-img create -f qcow lenny 4G
Formatting 'lenny', fmt=qcow, size=4194304 kB

I discovered you can specify a full IP for the vnc so you can expose it on the lan :)

slap:/home/qemus# qemu -hda lenny -cdrom debian-500-i386-netinst.iso -boot d -vnc 10.0.0.1:3

vnc into it, install as normal
slightly concerned one sets the hostname at install. I must find out how to change that later.
I'm going to split the partitions up so I can serve / /usr /home /tmp swap and /var from different targets. I'm thinking it *might* be possible to mount them ro so multiple comps can read the same vblade
4 hours later and it's still installing

All installed, now to shift /usr /var /home to NFS targets

Qemu installation

2009-02-26T14:03:00.007Z

I aptituded
Setting up kqemu-common (1.3.0~pre11-8) and qemu (0.9.1-10)

and it added a load of extras, vde2 included

I had to add kqemu-source
It depended gcc et al and kernel headers

cd /usr/src

tar -jxvf kqemu-source.tar.bz2
cd modules/kqemu
aptitude linux-headers-2.6.26-686 and dependents
./configure
make
make install
insmod kqemu.ko

echo keqmu >> /etc/modules

add tap0 to /etc/network/interfaces

auto lo eth0 eth1 tap0

iface tap0 inet static
        address 192.168.254.254
        netmask 255.255.255.0
        network 192.168.254.0
        broadcast 192.168.254.255

echo tun >> /etc/modules

maht@slap:~/storm_bin$ cat /etc/service/w2k/run

#!/usr/local/plan9/bin/rc

vdeq qemu -boot c -m 256 -hda /home/qemus/w2k -localtime -k en-us -net 'nic,macaddr=00:99:99:99:02:01,vlan=0' -net 'vde,vlan=0' -vnc 10.0.0.1:21 -win2k-hack -usbdevice tablet -snapshot

so, all done

SSH config - getting protocol 1 to work

2009-02-25T10:54:00.005Z

slap:/etc/ssh# ed /etc/ssh/sshd_config
/^Proto
s/^.*/Protocol 2,1
/dsa
a
# HostKey for protocol version 1
HostKey /etc/ssh/ssh_host_key
.
w
q

slap:/etc/ssh# ssh-keygen -t rsa1
Generating public/private rsa1 key pair.
Enter file in which to save the key (/root/.ssh/identity): /etc/ssh/ssh_host_key
Enter passphrase (empty for no passphrase):
Enter same passphrase again:
Your identification has been saved in /etc/ssh/ssh_host_key.
Your public key has been saved in /etc/ssh/ssh_host_key.pub.

slap:/etc/ssh# kill `ps axw | grep sbin/sshd | grep -v grep | awk ' { print $1} '` && /usr/sbin/sshd