Cleaning up some VMs this evening and wanted to mount a RAW disk partition I’d been using with an old VM. Surprised to find that a simple ‘mount’ wouldn’t do the trick. Fortunately pretty easy thanks to the info at the following link – use parted to get the offset and add that as one of the mount options.
https://askubuntu.com/questions/236263/browse-img-without-mounting
December 2, 2017
Mounting VirtualBox RAW disks
November 27, 2016
hostapd in a VM for device testing
While working with my Chamberlain MyQ Garage Door Opener the other day, I ran into issues getting hostapd to work on Kali with the chap ass Netis WP2120 nano USB adapter I had (based on the RTL8192cu chipset/driver). I need to clear off a few browser tabs and eventually see if I can get hostapd working with this adapter on a different box (VM) for testing. A few of the links I worked with unsuccessfully, but provide some useful information.
- https://koharicorp.wordpress.com/2015/03/26/debian-access-point-with-hostapd-and-rtl8188cusrtl8192cu/
- Still threw an error for me, may retry under a different distribution on a different host. I really should set up vagrant.
- https://www.raspberrypi.org/forums/viewtopic.php?t=28589&p=252533
- More of a client config than hostapd
- https://www.psattack.com/articles/20160410/setting-up-a-wireless-access-point-in-kali/
- hostapd configuration in Kali with a friendlier device/chipset
More notes
Installed usbutils for lsusb
Installed hostapd
Installed iw
Installed lshw
Ran through steps at https://adamscheller.com/systems-administration/rtl8192cu-fix-wifi/, minor issue with “linux-headers-generic”, skipped it for now.
Did receive a modprobe error, “could not insert ‘8192cu’: Device or resource busy”. lsmod shows the old rtl8192cu drivers active. Did a reboot. Correct driver modules now listed, though no device listing in ‘iw list’. Does appear in ‘lshw -c network’, and appears as wlan0 in ‘ip link list’.
Started in with hostapd steps from http://sarumait.blogspot.com/2013/11/ubuntu-wireless-hotspot-using-edimax-ew.html and from my previous testing on the Chamberlain page.
Attempted to start hostapd service which seemed to work, but didn’t see my SSID. Ran it interactive debug and got an “invalid/unknown driver ‘8192cu'”.
Instructions at https://github.com/pritambaral/hostapd-rtl871xdrv
Git cloned the above repository and the w1.fi repository listed. Got hostapd to build after installing dependencies (below).
apt-get install apt-file libnl-3-dev libnl-genl-3-dev pkg-config libssl-dev
apt-file not needed, but handy for looking up files that resulted in failures during make.
Ran the patch process noted at https://github.com/pritambaral/hostapd-rtl871xdrv
Ended up having some success after doing a ‘make dkms’ in the rt8192cu directory cloned from https://github.com/dz0ny/rt8192cu and setting the driver in my hostapd.conf to the rtl871xdrv
Pages helpful in getting it to work
- https://github.com/pritambaral/hostapd-rtl871xdrv
- Driver that actually worked and instructions that covered most of what I needed to do
- https://www.raspberrypi.org/forums/viewtopic.php?t=65009&p=478495
- This page was what got me to set the driver rtl871xdrv after the ‘make dkms’ in the rt8192cu directory
- http://askubuntu.com/questions/617973/fatal-error-netlink-genl-genl-h-no-such-file-or-directory
- This was useful in resolving some of the compilation errors I encountered while attempting to build hostapd before patching. I also got some errors during compilation regarding missing pkg-config, which led me to install the pkg-config package.
- http://billauer.co.il/blog/2014/06/linux-realtek-hostapd/
- Referenced several times during the research and build process
Chamberlain MyQ Garage Door Opener Testing
Got a great deal on some Chamberlain MyQ wifi garage door openers recently. As I’m setting them up, I really want to see what kind of conversations they have, so I’ll be making a few notes. As usual, I’ll be working with my VM setup, using Kali in this case.
I have a Netis WF2120 USB wireless adapter that I’ll be using under Kali. The vendor/product code is 0bda:8176. Planning to use Kali as an AP. I installed hostapd on Kali with ‘apt-get install hostapd’. Added a USB filter to pass the Netis through to the guest with ‘vboxmanage usbfilter add 2 –target kali –name NetisWIFI –vendorid 0x0bda –productid 0x8176’. After shutting down and restarting the VM, my wireless adapter was available as wlan0.
Enabled root login for SSH on my Kali box as VRDE… sucked. Slow and very unresponsive.
Created /etc/hostapd/hostapd.conf with approximately the following contents:
ssid=TheSSIDOfYourChoiceWillEndYourLifeTonight
wpa_passphrase=YeahPostThat
ctrl_interface=/var/run/hostapd
interface=wlan0
driver=rtl8192cu
channel=1
wpa=2
wpa_key_mgmt=WPA-PSK
Edited /etc/default/hostapd and configured ‘DAEMON_CONF=”/etc/hostapd/hostapd.conf”‘.
Attempted to start the hostapd service, got “Line 5: invalid/unknown driver ‘rtl8192cu'”.
Tried editing NetworkManager.conf and adding a keyfile configuration to exclude the MAC of the device, followed by restarting /etc/init.d/network-manager, without much success. Yielded same error.
Ran a ‘modprobe -r rtl8192cu’ and attempted to restart the hostapd service with the same error. Tried changing the “driver=” line to ‘driver=rtl871xdrv’ with the same results as well.
Remarked the “driver=” line in hostapd.conf and attempted to restart the service; got a little further, but still did not seem to start.
Went back to the instructions at https://bogeskov.dk/UsbAccessPoint.html. The ‘apt-get install linux-headers’ was not successful on my platform, so ran ‘apt-get install linux-headers-4.6.0-kali1-amd64’ instead, which was already installed in my case. Created and cd’ed to an rtl directory, then ran ‘git clone https://github.com/dz0ny/rt8192cu.git’. Attempt to make drivers failed.
Ran an ‘apt-get update’ and ‘apt-get upgrade’ to see if I could get around the issue. Seem to be issues with the 4.8.x kernel on Kali as well; system fails to boot using that kernel. Have to come back to this one later, will create another post for hostapd with the rtl8192cu and see if I can come up with a better way to work around that problem.
Might be useful to take a look for the referenced article on burp from http://cybergibbons.com/security-2/quick-and-easy-fake-wifi-access-point-in-kali/ once I get hostapd on solid footing on its own (which is getting closer).
Additional Links
- https://iotsecurity.eecs.umich.edu/img/Paper27_CameraReady_SmartThings_Revised_IEEEGen.pdf
- Security Analysis of Emerging Smart Home Applications paper
August 20, 2016
More 434MHz stuff to look at later
Acurite #0899 self emptying rain gauge – at Amazon
Blog post on reversing the signal from Acurite temperature sensors; may be similar for the rain gauges. Also mentions the RTL SDR. Probably wouldn’t be bad to have.
Another blog post regarding reading the Acurite 5n1 station. Possible that part of this will already cover the rain gauge. Also discusses SDR and mentions the rtl_433 tool.
This blog post looks like rtl_433 supports the Acurite 896 decode natively, so if the protocol is similar or the same it may be simple.
RTL Software Defined Radio – at Amazon – seems like it’s probably a decent model, some good reviews. Looks like there’s a little more to it than just “buy this and set up the software”.
Bought a cheap receiver off of eBay from Hong Kong for $7.31 (original listing).
Also, here was an Instructable about making a cheap combination RF/Bluetooth/433 MHz remote using an Arduino.
Additional Links
- http://www.instructables.com/id/Arduino-Frequency-Detection/
- Frequency detection on Arduino
- http://www.instructables.com/id/RF-315433-MHz-Transmitter-receiver-Module-and-Ardu/
- 434 MHz transmitter/receiver on Arduino
- http://www.princetronics.com/how-to-read-433-mhz-codes-w-raspberry-pi-433-mhz-receiver/
- 433 MHz receiver on Raspberry Pi
- Did this 10-Oct-2018 with an ESP8266-12E, and it worked well in receiving from a Spigen 433MHz doorbell transmitter. Only minor change was configuring “mySwitch.enableReceive(4)” instead of 0 for use with pin marked D2 on the ESP. Range was only a few feet with the receiver I was using, though I wonder a bit if using the “ReceiveDemo_advanced” would improve things at all. Need to play with this more
- https://hackaday.io/project/63004-esp433-wifi-to-433mhz-bridge
- 433 MHz with an ESP8266… not complete though.
- https://www.instructables.com/id/Using-an-ESP8266-to-Control-Mains-Sockets-Using-43/
- Another partial guide on the ESP, would need to determine how to wire up the receiver for the ReceiveDemo, but otherwise this could be pretty useful for supported protocols.
June 13, 2016
On Networks N150R Routers
Picked up a couple of these from Newegg for free plus a couple of bucks shipping after rebate. Figured that in combination with a couple of relays on the GPIOs, I could probably use them for switches or work with the serial port for a sensor connection. Just a few notes on them.
OpenWRT 15.05 Chaos Calmer patched to expose the GPIOs: https://micu.webserwer.pl/pmwiki/pmwiki.php?n=OpenWRT.Downloads
Forum thread on not flashing direct to 15.05.1; I went to 14.07, then to 15.05. https://forum.openwrt.org/viewtopic.php?id=51576
OpenWRT Wiki page with device info including GPIOs https://wiki.openwrt.org/toh/netgear/wnr612v2
Controlling the GPIOs
From here, the simple way to control the GPIOs natively is with the “brightness” value for the LEDs. By example for the LAN1 LED (others are in the same path):
root@OpenWrt:~# echo 0 > /sys/class/leds/netgear\:green\:lan1/brightness
root@OpenWrt:~# echo 1 > /sys/class/leds/netgear\:green\:lan1/brightness
Update 9/18/2016
Finally got the LAN2 LED peeled off of the board, had a heck of a time with it. After doing so, I discovered that I only get ~2.5v across the pins when the GPIO is active… not enough to drive my 3.3v relays. I assume they are probably using a voltage divider here. Measuring the resistance between the positive and negative ends of the LED without power applied, I get about 1580 ohms one direction, and 670 in the other. There appears to be a ~960 ohm resistor at CA111, but I haven’t figured out exactly how this ties in.
Update 10/9/2016
Spent some time with this again over the past couple of days. I removed the resistor at CA111, but apparently it is not a divider. May need to spend some time to see if I can make something work with a 2N7000 MOSFET or something like that, but I’m pretty novice so that will take some effort.
I did test out my 433MHz ASK receiver on the TTL serial port, but without much luck. This OpenWRT page was helpful in getting stty installed and setting the baud rate on the serial port,
‘stty -F /dev/ttyS0 9600 clocal cread cs8 -cstopb -parenb’ to set the config, and
‘stty -F /dev/ttyS0 -a’ to verify the config
but still didn’t get any output. I’ll have to give it a shot with my Raspberry Pi just to see if I can get anything out of there.
April 30, 2016
More Pogoplug E02 uBoot Work
I’ve written a few times about configuring Pogoplug devices to work with uBoot and Linux. Recently I’ve run into some problems with one of my Pogoplug E02 devices, where it would no longer boot to the attached USB drive, and didn’t seem to be booting properly to the native firmware either. Finally got around to looking at that a little bit in the past couple of days.
Since I wasn’t able to SSH to the device, I had to open it up and connect to the internal serial console. I had already purchased a USB to TTL serial cable adapter based on the PC-PL2303HX chipset (similar to these), so I opened up the Pogoplug and modified a CD-ROM audio cable I had laying around to connect between the TTL adapter cable and the serial port on the board (opening and serial connection documentation here).
Once I was able to get consoled in, I determined that my USB drive was only being detected properly on a soft boot. So pulling the power cable was doing me no good as far as getting it to boot, but if I issued a reset or reboot from the failed boot environment, it would detect the USB device correctly and boot up. While the reset command was functional, it was not practical, as it would have required a perpetual serial connection and for me to issue a reset every time it lost power before it would come up properly. I want it to reload to normal running state automatically.
I played around with the firmware environment variables for a bit before going ahead with an upgrade of my uBoot environment (I was on a 2011 version). I followed most of the instructions including flashing the default uBoot environment variables. I was still having some issues with the USB drive not being detected from a cold boot, but modifying the “scan_disk” environment variable appeared to provide a resolution. I changed the “scan_usb” portion of the command from “usb start” to “usb start; sleep 3; usb reset”, and thereafter I was able to successfully boot from both a cold start and a warm start. The full variable after changes was as follows:
scan_disk=echo running scan_disk ...; scan_done=0; setenv scan_usb "usb start; sleep 3; usb reset"; setenv scan_ide "ide reset"; setenv scan_mmc "mmc rescan"; for dev in $devices; do if test $scan_done -eq 0; then echo Scan device $dev; run scan_$dev; for disknum in $disks; do if test $scan_done -eq 0; then echo device $dev $disknum:1; if ext2load $dev $disknum:1 0x800000 /boot/uImage 1; then scan_done=1; echo Found bootable drive on $dev $disknum; setenv device $disknum:1; setenv bootdev $dev; fi; fi; done; fi; done
Finally, I made some modifications to allow the default Pogo OS to load in the event a drive with a “rootfs” label was not present. I added a “pogo_bootcmd” environment variable and modified the “bootcmd_exec” as in the following. This appears to be allowing the device to boot the defacto Pogo OS in the event the rootfs drive is not present, though I still had difficulties SSH’ing in with my Pogo portal configured account.
pogo_bootcmd=if fsload uboot-original-mtd0.kwb; then go 0x800200; fi bootcmd_exec=mw 0x800000 0 1; run load_uimage; if run load_initrd; then if run load_dtb; then bootm 0x800000 0x1100000 0x1c00000; else bootm 0x800000 0x1100000; fi; else if run load_dtb; then bootm 0x800000 - 0x1c00000; else if bootm 0x800000; then echo Booting; else run pogo_bootcmd; fi; fi; fi
March 31, 2015
VM System With VGA Pass-Through
Hardware hangup note for Ctrl+C to get past.
Debain Wheezy network installation freezes after trying to Detect Hardware
byu/SpaceMagnet indebian
Configured LVM environment with /home, /usr/, and /var on spinning platter drive.
Configured an IP reservation on my DHCP server, though I may change this to a true static address at some point.
Instructions for linux-headers and virtualbox setup: https://wiki.debian.org/VirtualBox
Downloaded the extension pack with wget http://download.virtualbox.org/virtualbox/4.1.36/Oracle_VM_VirtualBox_Extension_Pack-4.1.36-97684.vbox-extpack (NOTE: you will need to download the appropriate version for your installed VirtualBox version. Run vboxmanage -v and compare the output version to the versions available at https://www.virtualbox.org/wiki/Downloads; it apears you will need to match at least Major.Minor version numbers) and installed using instructions at https://www.virtualbox.org/manual/ch08.html#vboxmanage-extpack.
Created a separate logical volume on my SSD drive. Dedicated it to the Win10 guest (see https://wiki.archlinux.org/index.php/VirtualBox#Create_a_VM_configuration_to_boot_from_the_physical_drive).
Created a VM from the CLI. Had to actually go into the BIOS and enable AMD-V on my Gigabyte 78LMT-USB3 as it was disabled. Could not get the version of VirtualBox included with the Debian distribution to give anything except the “Your PC needs to restart” 0x0000005D error, even when trying to force the NX bit using info for setting CMPXCHG16B as found at http://www.makeuseof.com/tag/try-windows-10-youre-running-linux/. Ultimately ended up loosely following instructions from http://www.admin-magazine.com/Articles/Server-Virtualization-with-VirtualBox but substituting “wheezy contrib” for “lucid contrib” early in the instructions, and substituting 4.3 for 4.1, then running a “modifyvm” and setting ostype to Windows81_64. Once this was done, I was able to start the install process. Yay.
Had to modify the e1000g0 adapter config from http://www.perkin.org.uk/posts/create-virtualbox-vm-from-the-command-line.html to instead be “eth0” since I’m on a linux box (also had to remove it before re-adding, got an error when even attempting to start the machine).
Mouse pointer was batty, like it was scaled. Ended up downloading the appropriate VirtualBox guest additions to the host machine, mounting them in the DVD drive, selecting the AMD64 file, bringing up properties and setting to compatibility mode Windows 8 (can’t seem to locate where I found this), and was able to install. Pointer synced, system is on the web.
Second Steps – Debian LXC Guest for KDE
Debian 7.8
#From https://wiki.debian.org/LXC - loosely apt-get install lxc apt-get install bridge-utils libvirt-bin debootstrap # Failed with: # Failed to fetch http://security.debian.org/pool/updates/main/libx/libxml2/libxml2-utils_2.8.0+dfsg1-7+wheezy3_amd64.deb 404 Not Found [IP: 128.31.0.63 80] # E: Unable to fetch some archives, maybe run apt-get update or try with --fix-missing? apt-get apt-get install bridge-utils libvirt-bin debootstrap # Successful this time echo -e "cgroup\t/sys/fs/cgroup\tcgroup\tdefaults\t0\t0" >>/etc/fstab mount /sys/fs/cgroup mount # Confirmed that cgroup is mounted lxc-checkconfig # All show enabled # Check out http://www.turnkeylinux.org/blog/announcing-turnkey-lxc, maybe add # Turnkey template for some other functionality on this box # Switched tracks and decided to try out Docker # From http://docs.docker.com/installation/debian/#debian-wheezystable-7x-64-bit echo "deb http://http.debian.net/debian wheezy-backports main" >> /etc/apt/sources.list apt-get update apt-get install -t wheezy-backports linux-image-amd64 reboot # SSH'ed back into machine, verify new kernel uname -a wget https://get.docker.com/ chmod +x index.html gpasswd -a my-account docker service docker restart # Service apparently did not restart, so had to start it again manually service docker start # Now as non-root # Determined the following from http://www.syncano.com/getting-started-docker/ docker pull ubuntu:14.10 docker images # Lists the newly pulled image; attempted to run it. docker run -t ubuntu:14.10 # At this point, continuously encountered an error similar to the following: # FATA[0034] Error response from daemon: Cannot start container (container UUID here): [8] System error: write /sys/fs/cgroup/docker/(container UUID here)/cgroup.procs: no space left on device # There is definitely space left on the device, I do not appear to be lacking inodes as others seem to have encountered similarly, and permissions look OK. # Started to install strace to troubleshoot further... but I need to get an accounting system back online. # Switched to a VirtualBox KDE environment plan. I guess I can always kill it later :) lvm lvcreate -L 10g -n kmm wget http://peppermintos.com/iso/Peppermint-5-20140623-amd64.iso vboxmanage createvm --name kmm --register vboxmanage modifyvm kmm --memory 1024 --acpi on --boot1 dvd --nic1 bridged --bridgeadapter1 eth0 VBoxManage internalcommands createrawvmdk -filename ~/VirtualBox\ VMs/kmm/kmmSSD.vmdk -rawdisk /dev/amdfx/kmm vboxmanage storagectl kmm --name "SATA Controller" --add sata vboxmanage storageattach kmm --storagectl "SATA Controller" --port 0 --device 0 --type hdd --medium ~/VirtualBox\ VMs/kmm/kmmSSD.vmdk vboxmanage storageattach kmm --storagectl "SATA Controller" --port 1 --device 0 --type dvddrive --medium ./Peppermint-5-20140623-amd64.iso vboxmanage modifyvm kmm --ostype Ubuntu_64 --vrde on --vrdeport 22888 vboxmanage startvm kmm --type headless # Able to RDP in and started Peppermint install process.