Getting PacketTracer 7.1 to work on Ubuntu 17.10

If you need to use PacketTracer 7.1 and you’re on Ubuntu 17.10 you will need to install a few dependencies.

One annoying thing is that the launcher for pt will hide the library errors from you, unless you try to launch /opt/pt/bin/PacketTracer7 directly.

Installing PacketTracer using the included ‘install’ script and then running the code below should get you up and running.

The easiest way to get PT working is by just installing a few libraries:
sudo apt install libqt5multimedia5 libqt5script5 \ libqt5scripttools5 libqt5webkit5 libqt5xml5

OpenVPN 2.4 and ECDH on OpenBSD 6.1

I recently wanted to test out some of the newer features in OpenVPN 2.4 but I wasn’t able to find all of the needed information in the same place. There are 2 great guides on setting up OpenVPN already but both are a little old and reference some out of date stuff, so here’s my attempt on setting up a more ‘hardened’ OpenVPN server on OpenBSD for road-warriors.

OpenVPN 2.4.1 is currently available for OpenBSD 6.1.

Checklist:
- install openvpn 2.4
- install easy-rsa (currently 3.0.1)

Configure EASY-RSA for our PKI infrastructure

Easy-RSA 3.0’s commands differ from 2.0 and thankfully it’s now easier to use. We need to modify the default config a little to enable the shiny new ECDH support for key exchange.

# cd /usr/local/share/easy-rsa/ && cp vars.example vars

Now edit vars file using your favorite editor. We want to enable Elliptic Curve support in Easy-RSA. Find these 2 variables and change them to the following.

set_var EASYRSA_ALGO ec
set_var EASYRSA_CURVE secp521r1
- feel free to use a different curve, that’s just the one I chose.
It’s not required but you can fill out the “Organizational fields” like EASYRSA_REQ_COUNTRY or EASYRSA_REQ_EMAIL.

Once our vars file is filled out we can setup our PKI. It is highly recommend to run your ‘CA’ on a different machine than your OpenVPN server. You sign your certificates on a different computer and copy them over to your your OpenVPN server and clients.

It’s also best practice to copy the whole easy-rsa directory to somewhere else so your keys aren’t just sitting in /usr/local/share/easy-rsa/ but for this guide, I will reference the default location.

cd to your new easy-rsa location then run: 
# ./easyrsa init-pki
This will purge any left-over configs and rebuild the pki/
#./easyrsa build-ca
It will ask for a password to protect the CA's key, use a strong password!
Then it will ask to enter the Common Name of the CA.
Example: myhost.mydomain.com

Now we need to setup the vpn server’s cert/key pair. This is the cert and key that OpenVPN will run with.

# ./easyrsa build-server-full vpn.mydomain.com nopass
It will prompt for the CA's password, in order to use it to sign the server cert. I use the 'nopass' option because my OpenVPN server
is generally remote and if it reboots, I wouldn't want to have to enter a password just to start OpenVPN

Now need to generate our client(s) cert/key pair(s).

# ./easyrsa build-client-full client1.vpn.mydomain.com
Note: it is recommended to use a password for the client key-file. 
Most clients can cache the key and it provides some security if the key is lost.

Repeat this process for all clients. Once you’ve completed this, you’re done setting up your PKI.

Copying certificates and setting up OpenVPN

Become root now. We need to setup the openvpn config directory and copy some files to it.

# mkdir /etc/openvpn && mkdir /etc/openvpn/private
Now we need to copy the needed certificates and key to the openvpn dirctory
# cp /usr/local/share/examples/openvpn/sample-config-files/server.conf /etc/openvpn/
# cp /path/to/easy-rsa/pki/issued/vpn.mydomain.com.crt /etc/openvpn/
# cp /path/to/easy-rsa/pki/ca.crt /etc/openvpn/
# cp /path/to/easy-rsa/pki/private/vpn.mydomain.com.key /etc/openvpn/private/
# chmod -R 400 /etc/openvpn/private
# cd /etc/openvpn && openvpn --genkey --secret ta.key

What we’ve done is created a configuration directory for openvpn. Within this directory should be the ca.crt file, the vpn server’s crt file and the server.conf file for the openvpn process. Within the private directory is the vpn server’s key file. Then we made sure the private directory is only readable by root. Lastly we generated a tls authentication key, this is optional but it prevents abuse. You can read more about it here.

Configuring OpenVPN through server.conf

In the last steps we copied over the sample configuration file into /etc/openvpn.

Now we need to edit that file, make some changes and reference our certs and keys.

Here is a simplified config file:

port 80
proto udp
dev tun0
ca /etc/openvpn/ca.crt
cert /etc/openvpn/vpn.mydomain.com.crt
key /etc/openvpn/private/vpn.mydomain.com.key
dh none
server 10.8.0.0 255.255.255.0
push "redirect-gateway def1 bypass-dhcp"
keepalive 10 120
tls-auth /etc/openvpn/ta.key 0
cipher AES-256-CBC
compress lz4-v2
push "compress lz4-v2"
max-clients 4
verb 3
explicit-exit-notify 1
mssfix 1440

Firstly you need to reference the ca, crt and key files.
I’ve tested using TCP and performance was horrible, so stick to UDP.
I use port 80 because some cafes will block high ports like 1194
I set ‘dh none’ in order to ‘disable Diffie Hellman key exchange (and use ECDH only)’
I use ‘push redirect…’ line in order to force all client traffic through OpenVPN.
I use tls-auth and the ta.key we generated to mitigate potential abuse.
- tls-crypt is a newer and probably a better alternative to tls-auth but my client (tunnelblick) doesn’t like this new option. Read about the new tls-crypt here.
The last option ‘mssfix’ probably isn’t needed for most people but one place I connect at has MTU problems and this was the magical fix.

Now we need to get the OpenVPN process running.

As root, create and edit the file /etc/hostname.tun0

up
group _openvpn
description "OpenVPN client tunnel"
!/usr/local/sbin/openvpn --daemon --config /etc/openvpn/server.conf  & false

Now the system will automatically bring up the interface on boot and run the openvpn process.

Setting up PF

I’m going to assume you already have a natting pf.conf setup. Here is a minimal pf config to accept the inbound udp packets and allow your vpn clients out to the internet.

pf.conf

# we need to pass in port 80 UDP packets to accept the client requests
pass in on egress inet proto udp from any to (egress) port 80
match out on $ext_if inet from $vpn_net to nat-to ($ext_if)
pass in on tun0 inet

I’m using a table called “internet” that is defined:

table  const {0.0.0.0/0, !10/8, !172.16/12, !192.168/16 }

Edit the file /etc/sysctl.conf and make sure it has the line:

net.inet.ip.forwarding=1

Reboot and enjoy!

CenturyLink Fiber with PRISM in Seattle on OpenBSD

Overview

Most if not all of this information is already out on the internet regarding CenturyLink and connecting your own modem but I’d like to document my own steps as I’ve needed to gather data from a few sources. Almost everyone doing this online uses pfSense examples and not OpenBSD, so this information may be useful to others.

If you have CenturyLink and Prism, it is actually easier to connect your own router/firewall to the modem than without Prism! The TV part is just a minor fix. (and optional really)

Why is that? When you get Prism, CenturyLink disables the PPPoE authentication and uses IPoE which is just, IP over Ethernet, which is another term for what most of use for out home networks.

People have complained that pfSsense and other products cannot push 1gbps through their PPP daemon, so IPoE this eliminates that bottleneck. I don’t have 1Gbps service so I wouldn’t hit this bottleneck in any case

Overview of steps to connect your own firewall to CenturyLink

Write down the MAC address of your CenturyLink provided router’s WAN port, likely the CT2000T, you can find this under the Status page on the router.
- The CL modem would not give a DHCP address to my firewall until I cloned the MAC address of the router. Usually power cycling the modem should remove the learned MAC address of the router from the modem, this isn’t the case for my CL modem.
Change the MAC address of your firewall’s WAN interface to match the one found in step 1.

In OpenBSD you can set this to change on boot via:

# cat /etc hostname.em0
lladdr c4:ea:1d:01:02:03
up

Create a vlan with the tag 201 and put it on your firewall’s WAN interface.
- By default the CL modem communicates over VLAN 201 with the router. You can get this feature disabled if you talk to a knowledgeable CL technician but why bother.

In OpenBSD this is configured via:

# cat /etc/hostname.vlan201
dhcp vlan 201 vlandev em0 description “Modem Connection”

Now you should be able to bring up the interface and receive an external address from the modem. If you can’t receive an address at this point, then something is wrong.
Your external interface in pf.conf is vlan201 and not em0, (remember tagging), since everything that flows over the external interface (em0), needs the vlan tag.

Step to configure PRISM (IPTV) Setup

You will need install and configure the igmpproxy daemon and allow igmp and udp multicast streams to your CenturyLink STBs.

Install the igmpproxy daemon on OpenBSD ( # pkg_add igmpproxy )
Make sure igmpproxy is added to your pkg_scripts so it starts on reboot.
Configure the igmproxy via /etc/igmpproxy.conf.

Here is a trimmed down example of mine:

# cat /etc/igmpproxy.conf

quickleave
phyint vlan201 upstream ratelimit 0 threshold 1
altnet 67.12.0.0/15
altnet 151.118.0.0/15
phyint em1 downstream ratelimit 0 threshold 1
phyint lo0 disabled
phyint re0 disabled

Since I don’t have a managed switch at home that has the IGMP snooping option, I didn’t want the igmp packets broadcasting over my WiFi APs (connected to re0), so I added a 3rd NIC and the IPTV STB is plugged directly into it. This way the IPTV is confined to just that NIC and not sent to other hosts in my network. Your desired configuration may be different.

Allow the multicast UDP packets and IGMP packets on your ext interface (vlan201 remember) so the IGMPProxy can do it’s thing. Then, since they will be leaving on another interface that points to the STB, you need to allow them out that one as well

pf.conf:

table <iptv> const {151.118.0.0/16, 67.12.0.0/15, 224.0.0.0/4 }

pass quick on $ext_if inet proto udp from <iptv> to any allow-opts
pass quick on $ext_if inet proto igmp allow-opts allow-opts

pass quick on $iptv_if inet proto igmp allow-opts
pass quick on $iptv_if inet proto udp from any to any allow-opts

I found the list of source networks for the table in the 2nd link in my sources below.

Don’t forget to enable multicast forwarding!

# echo “net.inet.ip.mforwarding=1” >> /etc/sysctl.conf

# echo “multicast=YES” >> /etc/rc.conf.local

# reboot (not sure if it’s necessary but the above changes will certainly take effect on reboot)

This isn’t a whole tutorial on setting up a firewall so I’m assuming you have a working pf.conf. Reboot for the multicast forwarding to take effect.

That’s it. Turn on your CenturyLink STB and make sure the TV works!

Great links from which I learned most about this process.

http://kmwoley.com/blog/bypassing-needless-centurylink-wireless-router-on-gigabit-fiber/

http://www.dslreports.com/forum/r30270839-Prism-TV-HOWTO-Use-pfSense-with-CenturyLink-FTTH-and-Prism-TV-in-Seattle

Setting up Piwigo on OpenBSD 5.8

I’ve finally gotten around to setting up Piwigo on OpenBSD and I ran into a few problems that I was able to fix after much googling.

Piwigo 2.7.4 sucks at logging and telling you what it needs/wants so there is a lot of trial and error. Hopefully I can save you some time.

Special considerations. I’ve moved my httpd chroot from the default /var/www into a directory under /mnt which has caused a few headaches.

Mysql. You need a mysql.sock visible to httpd inside it’s chroot, if you’ve moved it away from /var/xxx then you need to update /etc/my.conf and modify the socket to be created in say /mnt/www/run/mysql/mysql.sock
- A lot of the options out there just have you create symlinks but they never seem to work for http when it’s inside a chroot.
- You need to install php-mysql and follow the instructions to enable it or else everything will just fail silently for piwigo when it tries to use a database connection
- On the installation page the Database HOST field will be something like this: localhost:/run/mysql/mysql.sock
- UPDATE: do not put localhost:/path/…. Piwigo will die a painful death thinking this is a network and port combination. Instead make sure it looks more like:
  - $conf[‘db_host’] = ‘/run/mysql/mysql.sock’;
- If you’re importing an existing piwigo installation then you need to edit: /path/to/piwigo/local/config/database.inc.php to update the db_host variable
PHP If you want to be able to install plugins via the admin GUI then you need to install php-curl manually and enable it, or Piwigo will fail with a generic connection error.
- To enable hostname resolution create /path/to/www/etc
- and then copy your /etc/resolv.conf to the chroot’ed /etc folder
- edit /etc/php-5.6.ini (or whatever version you have) and enable turn on: allow_url_fopen
  - You can also edit upload_max_filesize and max_file_uploads if you think you’ll be uploading pictures bigger than 2mb or more than 20 at a time
Thumbnails not working?
- You also need to install php-gd or no pictures will be able to generate thumbnails. This shows up as just triangular warning signs with exclamation points on everything

Well I hope this helps someone. Httpd seems to be a capable replacement for apache/nginx even when running a program like Piwigo which wasn’t developed with it in mind.

Mounting an encrypted OpenBSD partition from the ramdisk

There a several great guides online which show you how-to create an OpenBSD system with an encrypted partition.

However one thing that isn’t called out anywhere that I can find: How do you manually mount it in the case that you don’t want to boot straight into the encrypted boot partition?

Let’s say you need to boot the system using bsd.rd to do something to the system.

Ok booting using bsd.rd is easy, plop it in the / partition, at boottime say boot bsd.rd and then enter the encryption password to boot the system. However in my setup, once this boots, I still can’t mount the encrypted filesystem.

Running ‘bioctl softraid0’ will show that my crypto raid drive is on sd1 however trying to mount sd1,it will fail and say it doesn’t exist.

All you need to do is make the proper device node. If ‘bioctl softraid0’ reports the cryptodrive as sd1, then run this:

chmod +x /dev/MAKEDEV

/dev/MAKEDEV sd1

All of a sudden the system has access to the drive, you can see it with

disklabel -h sd1

I’m not sure if this is an oversight in the softraid0 program or if this is by design so something won’t get it’s toes stepped on.

It seems a little nuts to me the system could say, hey we have something running on sd1…except we won’t create the device node so you can’t access it.

==========

This is all mentioned in http://www.openbsd.org/faq/faq14.html#softraid

however the section on using MAKEDEV would only apply (in my head) to when you’re first creating a raid/crypt setup, not after it’s creation. I didn’t want to overwrite my data so I was paranoid.

Triple boot Windows 8.1 Linux and OpenBSD on a laptop

I want to document this somewhere because searching around on the Internet only made me think that I was going to need to use some arcane spells or install some special programs to get this working. I did not really have to do anything special in the end, which was a relief.

I have a work laptop and I’ve wanted to install OpenBSD on it for a bit but I still need the ability to run a few other programs that I can only use in Linux and then sometimes Windows (Vmware mainly) I’ll keep windows around for any unknowns.

So in a nutshell I have a lenovo laptop with a 256GB SSD. All I had to do was Install Windows 8.1, I told the installer to create a 100GB partition. It also created a 2nd smaller boot partition. So sda1 and sda2 are taken up.

Windows boots and updates just fine.

Then I installed Arch linux. Since I’m most familiar with fdisk on Linux, I created a third partition /dev/sda3 for OpenBSD, I set the partition type to A6 and then I created the last partition. Since it had to be an extended partition I created it, then I created a logical partition inside the extended partition for linux to live in. So linux would be installed to sda5

you will get an error about the filesystem containing a ufs1 filesystem and it throws and error. BUT don’t worry, it actually did install grub successfully. I’m not sure
why it seems to say it didn’t
grub-install –target=i386-pc /dev/sda
a reboot will prove it when you get the grub menu.

Then I found this site which shows how to add a section to GRUB2 to boot openbsd
http://www.kariliq.nl/openbsd/grub2.html

Once you’ve followed that, you can run this:

grub-mkconfig -o /boot/grub/grub.cfg

which will write the grub menu with sections for Linux, OpenBSD and Windows.

Reboot and install BSD

When the installer starts, since you already set a partition type to A6, it should give you the option to select ‘O’ which uses the OpenBSD space.
This way OpenBSD will set itself up inside /dev/sda3

I didn’t have to do anything special and once the system finished, I was able to reboot and I still got the GRUB menu which offered all 3 OSs.
I’ve booted into all of them and let Windows update itself, nothing has messed with the grub and I can happily triple boot.

Disk /dev/sda: 238.5 GiB, 256060514304 bytes, 500118192 sectors
Units: sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disklabel type: dos
Disk identifier: 0xab36ea15

Here is an example of the partition table when looking at it from fdisk on linux.

Device     Boot     Start       End   Sectors   Size Id Type
/dev/sda1 *         2048    718847    716800   350M 7 HPFS/NTFS/exFAT
/dev/sda2          718848 213397503 212678656 101.4G 7 HPFS/NTFS/exFAT
/dev/sda3       213397504 381169663 167772160    80G a6 OpenBSD
/dev/sda4       381169664 500118191 118948528 56.7G 5 Extended
/dev/sda5       381171712 500118191 118946480 56.7G 83 Linux

this document isn’t well thought out however I thought I’d better just publish it before I completely forget about it

To Gentoo or not to Gentoo

Ever since I installed Gentoo back in 2005 I’ve been hooked. Of course I’ve tried various other distributions, Ubuntu and many of it’s variants (I’m writing from Lubuntu currently) Fedora (I don’t think I’ve eve let it stay on a system for very long) and many many years ago before Gentoo I’ve had Mandrake, Red Hat, Slackware, Suse (open) and a few others.

There seems to exist a cycle of installing Gentoo, having it around for awhile, something breaks and I feel like I don’t have the time so I’ll try out the latest Ubuntu…a few weeks later I’ll reinstall Gentoo.

Now let me clarify on point, I’m referring to Gentoo as a desktop. I have 2 servers which have always had Gentoo installed on them and probably always will.

I realized that my main problem with Gentoo is usually graphics related in someway. Either my graphics break and I need to do a bunch of troubleshooting, finally getting something working or it has to do with endless configuration.

I think the documentation on the Gentoo-wiki has gone downhill for sometime. There just doesn’t seem to be enough manpower to update items. I’ve tried in vain to get NetworkManager integrated into the system so that it can connect before I login but alas, it never works. I would love to contribute more to Gentoo, help with documentation and such but I’m not quite done with school yet and my willpower to write wiki documents usually doesn’t last very long.

I’ve also realized half the fun of a Gentoo system is that installation procedure for me, I love reinstalling from scratch and having a squeaky clean system. If anyone is reading this you’ve probably realized that I’m beyond mortal help 🙂 as most people loath installing Gentoo.

Currently my main desktop still has Gentoo on it but I’ve installed Lubuntu on my netbook. I was just tired of fucking with freedesktop.org to view my own partitions or trying to get Network Manager working in fluxbox (and saving the configuration files, automatically connecting and such) or having LibreOffice and Skype randomly crash for no reason.

I might put Gentoo back on my netbook someday but for now Lubuntu is fine. I had first tried Xubuntu but after updating the system, turning off some unneeded boot services, the boot time was 1 minute and 20 seconds. Lubuntu boots in about 50 seconds (my netbook has a harddrive not a SSD) so it is something of a booting dog.

One last complaint about Gentoo is the speed of portage. When I first booted my working Gentoo box a few months ago, I asked it to emerge -pv something, within 1 second i had my detailed results. After installing a bunch of packages and adding packages.mask, unmask,use my performance was now more like a minute later I’d get my results. I understand that I’ve added many more dependencies for the system to check but a fucking minute? It reminds me of going back to 2005 when searching the portage tree with -s was rediculous and everyone installed some sort of program that would cache all of the data to allow for quick searching (yes the devs have fixed this sometime back in 2007 or 2008 I think)

Alright enough ranting. Arch Linux. I have yet to try it but I’m intrigued by reading through the installation document. It could be a very interesting distribution, and I’ll probably try it out sometime after this weekend.

Who knows the future? I very well could end up becoming involved in Gentoo a lot more, fixing problems and documenting on the wiki. I’m almost done with my double degree, I’ve put my 2 weeks in a my job and in just 7 weeks I’ll be going over to Germany for a month to work on my Deutsch. When I get back I’ll be a graduate without a job (by choice). 🙂

eeePC 1000 funtoo

Gentoo using Funtoo on the eee-pc 1000

First of some history : I bought the eee-1000 (not H or HA) for my wife’s birthday, she didn’t like the pre-installed xandros so we blew it off and put ubuntu-eee on it. Boot times were slow and she didn’t like it too much. She asked if I’d help her put Gentoo on it.

Lets get down to it. The specs.Newegg described it as “ASUS Eee PC 1000 40G – Fine Ebony NetBook Intel Atom 10.0″ Wide SVGA 1GB Memory 40GB SSD ”

Don’t be fooled though, it has an 8 GB SSD and a 32GB Flash drive.

Specs:

Intel atom processor N270 2.6GHz

Wireless card Ralink RA2860 Works great**

see network section, needs third party drivers

Integrated SD card works great

1GB of ram, expandable to 2 GB, works fine

Screen resolution of 1024×600, works fine with no extra configuration needed in xorg.conf

SSD hard drive and flash drives works fine

I have / on the SSD drive and /home on the Flash drive

make sure to add elevator=noop to your kernel boot parameters in your grub.conf

Webcam

in progress

bluetooth

in progress

SD cardslot

works fine

Installation

Funtoo?
- My choice to use the funtoo image is because I don’t want to compile so much when I install. The funtoo pentium4 image is ACCEPT_KEYWORDS=”~x86″ so I get a prebuilt system that’s the latest for my needs, emerging packages on the eee-pc does take sometime. I do however use a standard portage snapshot.
Suspend? Hibernate? Not for me.
- I’ve never cared much for suspend or hibernate as they ususally cause random problems and since I’m using funtoo by default, I’m using openrc which helps decrease boot times. I also use fluxbox with wbar for an app launcher and conky for some system monitoring, my boot times from cold boot are 10s, which is quick enough for me.

First grab yourself a bootcd from http://www.sysresccd.org/Main_Page

At this time the gentoo minimal boot cd dosen’t detect enough of the hardware for it to be usefull so that’s why I’m using system rescue cd.

Since the eee-pc doesn’t have a cdrom drive you will need to create a bootable usb stick. Directions are on the site, I won’t discuss them here.

So now we have a bootable usbstick, check the bios for some options, use the right key to force boot the usbstick.

Before you boot up you might want to head over to funtoo.org and grab the latest ~pentium4 stage 3 build and put it on the usbstick. Also you can grab a portage snapshot from somewhere perhaps here . This step isn’t absolutely necessary but it easier now than using the text webbrowser on the bootcd.

I had some trouble getting my usbstick to boot so I disabled both hard drives from the boot list in the BIOS and that helped a lot, just re-enable them later.

Hard disks
- I simply partitioned the whole SSD as /dev/sda1. Then I formatted it as ext3. There is a lot out there about journals and wearing down the ssd but its not worth the risk, I think, to not use a journaling fs.
- I partitioned the whole flashdrive as sdb1 and I formatted it as ext3. Then I mount it as /home Later in my make.conf I use /home/distfiles as the location for downloaded programs cache for gentoo

continue as your normally would with the install following the gentoo handbook. When it is time to reboot don’t forget to set your root password.

One problem I have run into (and it may be fixed by now) is that since we are using funtoo and it has openrc by default, it seems to fail to setup the /dev properly so before you exit your chroot do the following rc-update add devfs sysinit && rc-update add udev sysinit

Network

wireless

NOTE: As of the linux kernel 2.6.29-rc1 the driver for this wireless card is included in the staging drivers section, I haven’t tested it yet

There is no opensource driver right now for this card, the card has a good driver from ralink however. Go to this link and grab the latest driver its the link starting with “RT2860PCI/mPCI/PCIe…”Read the readme if you need to but for me, all i needed to do was to edit os/linux/config.mk

and set
'HAS_WPA_SUPPLICANT=y' and 'HAS_NATIVE_WPA_SUPPLICANT_SUPPORT=y'

since i run network manager, if you don’t run it then don’t bother.
Now go into the main dir of the drive and do “make” and “make install”. then reboot, modprobe rt2860sta and see if iwconfig lists the link ra0
Network manager can do the rest from there, as I use nm-applet in the system tray to configure the wireless.

wired-network

To get the rj45 port working all you need to do in your kernel config for 2.6.28 is:

set experimental drivers to yes

General setup —>

(Y) Prompt for development and/or incomplete code/drivers

Then

-> Device Drivers
│ -> Network device support (NETDEVICES [=y])
│ -> Ethernet (1000 Mbit) (NETDEV_1000 [=y])

Yes to Atheros L1E

Config files

Here are some of the config files I have created on this funtoo laptop. I welcome feedback.

make.conf
FEATURES="parallel-fetch distlocks sandbox userpriv usersandbox" ACCEPT_KEYWORDS="~x86" CHOST="i686-pc-linux-gnu" CFLAGS="-Os -march=prescott -pipe -fomit-frame-pointer" MAKEOPTS="-j2" GENTOO_MIRRORS="http://gentoo.osuosl.org/ "

SYNC=”rsync://rsync.namerica.gentoo.org/gentoo-portage”
USE=”gtk -gnome -kde -qt3 -qt3support -sqlite -qt4 alsa dvd -cdr wifi ipv6 -berkdb -cups \
X networkmanager samba wireless bz2 xml lzo unicode laptop dri drm rar \
tiff gif png jpeg win32codecs mpeg mp3 mp4 divx libcaca intel opengl \
svg dbus hal -esd -eds -ldap -fam mmx sse v4l2 sse2 ssse3 pni xvid xvmc”
VIDEO_CARDS=”intel vesa”
AUDIO_CARDS=”intel-hd”
LINGUAS=”en_US en”
DISTDIR=”/home/distfiles”
PORTDIR_OVERLAY=”/usr/local/portage”

A note about my use flags: I am using fluxbox and therefore don’t need kde,qt or gnome, gtk by default. Your intended setup may differ

fstab

# NOTE: If your BOOT partition is ReiserFS, add the notail option to opts.
/dev/sdb1 /home ext3 noatime 0 0
/dev/sda1 / ext3 noatime 0 1
/dev/sda5 none swap sw 0 0
#/dev/cdrom /mnt/cdrom auto noauto,ro 0 0
#/dev/fd0 /mnt/floppy auto noauto 0 0
shm /dev/shm tmpfs nodev,nosuid,noexec 0 0

Done

now reboot and enojy your system. I’ll add some software titles that I use later, such as eee-control. Please give me feedback, I intend this article as a starting point and it should change overtime as more gets tested and optimized.

Hardware/software problems