summaryrefslogtreecommitdiffstats
path: root/docs/security
diff options
context:
space:
mode:
authorFrancis Rowe <info@gluglug.org.uk>2014-11-05 20:52:36 (EST)
committer Francis Rowe <info@gluglug.org.uk>2014-11-05 20:52:36 (EST)
commit8b2219bfa2da36e7809588ef723a10483a6e137f (patch)
tree8c4d74b9efbdabc1a2604b8cbd508c006c225241 /docs/security
parent7429bdcdbb4fc51c61897115112468642afeecfc (diff)
downloadlibreboot-8b2219bfa2da36e7809588ef723a10483a6e137f.zip
libreboot-8b2219bfa2da36e7809588ef723a10483a6e137f.tar.gz
libreboot-8b2219bfa2da36e7809588ef723a10483a6e137f.tar.bz2
Documentation: *major* cleanup.
Cleanup was long overdue. Old structure was messy and inefficient.
Diffstat (limited to 'docs/security')
-rw-r--r--docs/security/dock.html154
-rw-r--r--docs/security/index.html40
-rw-r--r--docs/security/t60_security.html440
-rw-r--r--docs/security/x60_security.html301
4 files changed, 935 insertions, 0 deletions
diff --git a/docs/security/dock.html b/docs/security/dock.html
new file mode 100644
index 0000000..2b7fe75
--- /dev/null
+++ b/docs/security/dock.html
@@ -0,0 +1,154 @@
+<!DOCTYPE html>
+<html>
+<head>
+ <meta charset="utf-8">
+ <meta name="viewport" content="width=device-width, initial-scale=1">
+
+ <style type="text/css">
+ @import url('css/main.css');
+ </style>
+
+ <title>Notes about DMA and the docking station (X60/T60)</title>
+</head>
+
+<body>
+ <header>
+ <h1>Notes about DMA and the docking station (X60/T60)</h1>
+ </header>
+
+<pre>
+
+Use case:
+---------
+Usually when people do full disk encryption, it's not really full disk,
+instead they still have a /boot in clear.
+
+So an evil maid attack can still be done, in two passes:
+1) Clone the hdd, Infect the initramfs or the kernel.
+2) Wait for the user to enter its password, recover the password,
+luksOpen the hdd image.
+
+I wanted a real full-disk encryption so I've put grub in flash and I
+have the following: The HDD has a LUKS rootfs(containing /boot) on an
+lvm partition, so no partition is in clear.
+
+So when the computer boots it executes coreboot, then grub as a payload.
+Grub then opens the LUKS partition and loads the kernel and initramfs
+from there.
+
+To prevent hardware level tempering(like reflashing), I used nail
+polish with a lot of gilder, that acts like a seal. Then a high
+resolution picture of it is taken, to be able to tell the difference.
+
+The problem:
+------------
+But then comes the docking port issue: Some LPC pins are exported
+there, such as the CLKRUN and LDRQ#.
+
+LDRQ# is "Encoded DMA/Bus Master Request": "Only needed by
+peripherals that need DMA or bus mastering. Requires an
+individual signal per peripheral. Peripherals may not share
+an LDRQ# signal."
+
+So now DMA access is possible trough the dock connector.
+So I want to be able to turn that off.
+
+If I got it right, the X60 has 2 superio, one is in the dock, and the
+other one is in the laptop, so we have:
+ ________________
+ _________________ | |
+| | | Dock connector:|
+|Dock: NSC pc87982|&lt;--LPC---&gt;D_LPC_DREQ0 |
+|_________________| |_______^________|
+ |
+ |
+ |
+ |
+ ___________________|____
+ | v |
+ | SuperIO: DLDRQ# |
+ | NSC pc87382 LDRQ# |
+ |___________________^____|
+ |
+ |
+ |
+ |
+ ___________________|___
+ | v |
+ | Southbridge: LDRQ0 |
+ | ICH7 |
+ |_______________________|
+
+
+The code:
+---------
+Now if I look at the existing code, there is some superio drivers, like
+pc87382 in src/superio/nsc, the code is very small.
+The only interesting part is the pnp_info pnp_dev_info struct.
+
+Now if I look inside src/mainboard/lenovo/x60 there is some more
+complete dock driver:
+
+Inside dock.c I see some dock_connect and dock_disconnect functions.
+
+Such functions are called during the initialisation (romstage.c) and
+from the x60's SMI handler (smihandler.c).
+
+Questions:
+----------
+1) Would the following be sufficent to prevent DMA access from the
+outside:
+&gt; int dock_connect(void)
+&gt; {
+&gt; int timeout = 1000;
+&gt; + int val;
+&gt; +
+&gt; + if (get_option(&amp;val, &quot;dock&quot;) != CB_SUCCESS)
+&gt; + val = 1;
+&gt; + if (val == 0)
+&gt; + return 0;
+&gt; [...]
+&gt; }
+>
+&gt; void dock_disconnect(void) {
+&gt; + if (dock_present())
+&gt; + return;
+&gt; [...]
+&gt; }
+2) Would an nvram option be ok for that? Should a Kconfig option be
+added too?
+
+&gt; config DOCK_AUTODETECT
+&gt; bool "Autodetect"
+&gt; help
+&gt; The dock is autodetected. If unsure select this option.
+>
+&gt; config DOCK_DISABLED
+&gt; bool "Disabled"
+&gt; help
+&gt; The dock is always disabled.
+>
+&gt; config DOCK_NVRAM_ENABLE
+&gt; bool "Nvram"
+&gt; help
+&gt; The dock autodetection is tried only if it is also enabled
+&gt; trough nvram.
+
+</pre>
+
+<hr/>
+
+ <p>
+ Copyright &copy; 2014 Francis Rowe &lt;info@gluglug.org.uk&gt;<br/>
+ This document is released under the Creative Commons Attribution-ShareAlike 4.0 International Public License and all future versions.
+ A copy of the license can be found at <a href="../license.txt">../license.txt</a>.
+ </p>
+
+ <p>
+ This document is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See <a href="../license.txt">../license.txt</a> for more information.
+ </p>
+
+</body>
+</html>
diff --git a/docs/security/index.html b/docs/security/index.html
new file mode 100644
index 0000000..64008b8
--- /dev/null
+++ b/docs/security/index.html
@@ -0,0 +1,40 @@
+<!DOCTYPE html>
+<html>
+<head>
+ <meta charset="utf-8">
+ <meta name="viewport" content="width=device-width, initial-scale=1">
+
+ <style type="text/css">
+ @import url('../css/main.css');
+ </style>
+
+ <title>Security topics</title>
+</head>
+
+<body>
+
+ <h1 id="pagetop">Security topics</h1>
+ <p>
+ Or <a href="../index.html">Back to main index</a>.
+ </p>
+ <ul>
+ <li><a href="x60_security.html">ThinkPad X60/X60S: hardware security</a></li>
+ <li><a href="t60_security.html">ThinkPad T60: hardware security</a></li>
+ </ul>
+
+<hr/>
+
+ <p>
+ Copyright &copy; 2014 Francis Rowe &lt;info@gluglug.org.uk&gt;<br/>
+ This document is released under the Creative Commons Attribution-ShareAlike 4.0 International Public License and all future versions.
+ A copy of the license can be found at <a href="../license.txt">../license.txt</a>.
+ </p>
+
+ <p>
+ This document is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See <a href="../license.txt">../license.txt</a> for more information.
+ </p>
+
+</body>
+</html>
diff --git a/docs/security/t60_security.html b/docs/security/t60_security.html
new file mode 100644
index 0000000..eb3db03
--- /dev/null
+++ b/docs/security/t60_security.html
@@ -0,0 +1,440 @@
+<!DOCTYPE html>
+<html>
+<head>
+ <meta charset="utf-8">
+ <meta name="viewport" content="width=device-width, initial-scale=1">
+
+ <style type="text/css">
+ @import url('../css/main.css');
+ </style>
+
+ <title>Libreboot documentation: Security on the ThinkPad T60</title>
+</head>
+
+<body>
+
+ <header>
+ <h1>Security on the ThinkPad T60</h1>
+ <aside>Hardware modifications to enhance security on the ThinkPad T60. This tutorial is <b>incomplete</b> at the time of writing.</aside>
+ </header>
+
+ <p>Or go <a href="index.html">back to main index</a></p>
+
+ <h2>Table of Contents</h2>
+ <ul>
+ <li><a href="#hardware_requirements">Hardware Requirements</a></li>
+ <li><a href="#software_requirements">Software Requirements</a></li>
+ <li><a href="#procedure">The procedure</a></li>
+ </ul>
+
+ <h1 id="hardware_requirements">Hardware requirements</h1>
+ <ul>
+ <li>A T60</li>
+ <li>screwdriver</li>
+ <li>(in a later version of this tutorial: soldering iron and scalpel)</li>
+ </ul>
+
+ <h1 id="software_requirements">Software requirements</h1>
+ <ul>
+ <li>none (at least in the scope of the article as-is)</li>
+ <li>You probably want to encrypt your GNU/Linux install using LUKS</li>
+ </ul>
+
+ <h1>
+ Rationale
+ </h1>
+ <p>
+ Most people think of security on the software side: the hardware is important aswell.
+ Hardware security is useful in particular to journalists (or activists in a given movement) who need absolute privacy in their work.
+ It is also generally useful to all those that believe security and privacy are inalienable rights.
+ Security starts with the hardware; crypto and network security come later.
+ </p>
+ <p>
+ Paradoxically, going this far to increase your security also makes you a bigger target.
+ At the same time, it protects you in the case that someone does attack your machine.
+ This paradox only exists while few people take adequate steps to protect yourself: it is your <b>duty</b>
+ to protect yourself, not only for your benefit but to make strong security <i>normal</i> so
+ that those who do need protection (and claim it) are a smaller target against the masses.
+ </p>
+ <p>
+ Even if there are levels of security beyond your ability (technically, financially and so on)
+ doing at least <i>something</i> (what you are able to do) is extremely important.
+ If you use the internet and your computer without protection, attacking you is cheap (some say it is
+ only a few US cents). If everyone (majority of people) use strong security by default,
+ it makes attacks more costly and time consuming; in effect, making them disappear.
+ </p>
+ <p>
+ This tutorial deals with reducing the number of devices that have direct memory access that
+ could communicate with inputs/outputs that could be used to remotely
+ command the machine (or leak data).
+ </p>
+
+ <h1 id="procedure">Disassembly</h1>
+
+ <p>
+ Remove those screws and remove the HDD:<br/>
+ <img src="../images/t60_dev/0001.JPG" alt="" /> <img src="../images/t60_dev/0002.JPG" alt="" />
+ </p>
+
+ <p>
+ Lift off the palm rest:<br/>
+ <img src="../images/t60_dev/0003.JPG" alt="" />
+ </p>
+
+ <p>
+ Lift up the keyboard, pull it back a bit, flip it over like that and then disconnect it from the board:<br/>
+ <img src="../images/t60_dev/0004.JPG" alt="" /> <img src="../images/t60_dev/0005.JPG" alt="" /> <img src="../images/t60_dev/0006.JPG" alt="" />
+ </p>
+
+ <p>
+ Gently wedge both sides loose:<br/>
+ <img src="../images/t60_dev/0007.JPG" alt="" /> <img src="../images/t60_dev/0008.JPG" alt="" />
+ </p>
+
+ <p>
+ Remove that cable from the position:<br/>
+ <img src="../images/t60_dev/0009.JPG" alt="" /> <img src="../images/t60_dev/0010.JPG" alt="" />
+ </p>
+
+ <p>
+ Now remove that bezel. Remove wifi, nvram battery and speaker connector (also remove 56k modem, on the left of wifi):<br/>
+ <img src="../images/t60_dev/0011.JPG" alt="" /><br/>
+ Reason: has direct (and very fast) memory access, and could (theoretically) leak data over a side-channel.<br/>
+ <b>Wifi:</b> The ath5k/ath9k cards might not have firmware at all. They might safe but could have
+ access to the computer's RAM trough DMA. If people have an intel
+ card(most T60's come with Intel wifi by default, until you change it),then that card runs
+ a non-free firwamre and has access to the computer's RAM trough DMA! So
+ it's risk-level is very high.
+ </p>
+
+ <p>
+ Remove those screws:<br/>
+ <img src="../images/t60_dev/0012.JPG" alt="" />
+ </p>
+
+ <p>
+ Disconnect the power jack:<br/>
+ <img src="../images/t60_dev/0013.JPG" alt="" />
+ </p>
+
+ <p>
+ Remove nvram battery (we will put it back later):<br/>
+ <img src="../images/t60_dev/0014.JPG" alt="" />
+ </p>
+
+ <p>
+ Disconnect cable (for 56k modem) and disconnect the other cable:<br/>
+ <img src="../images/t60_dev/0015.JPG" alt="" /> <img src="../images/t60_dev/0016.JPG" alt="" />
+ </p>
+
+ <p>
+ Disconnect speaker cable:<br/>
+ <img src="../images/t60_dev/0017.JPG" alt="" />
+ </p>
+
+ <p>
+ Disconnect the other end of the 56k modem cable:<br/>
+ <img src="../images/t60_dev/0018.JPG" alt="" />
+ </p>
+
+ <p>
+ Make sure you removed it:<br/>
+ <img src="../images/t60_dev/0019.JPG" alt="" />
+ </p>
+
+ <p>
+ Unscrew those:<br/>
+ <img src="../images/t60_dev/0020.JPG" alt="" />
+ </p>
+
+ <p>
+ Make sure you removed those:<br/>
+ <img src="../images/t60_dev/0021.JPG" alt="" />
+ </p>
+
+ <p>
+ Disconnect LCD cable from board:<br/>
+ <img src="../images/t60_dev/0022.JPG" alt="" />
+ </p>
+
+ <p>
+ Remove those screws then remove the LCD assembly:<br/>
+ <img src="../images/t60_dev/0023.JPG" alt="" /> <img src="../images/t60_dev/0024.JPG" alt="" /> <img src="../images/t60_dev/0025.JPG" alt="" />
+ </p>
+
+ <p>
+ Once again, make sure you removed those:<br/>
+ <img src="../images/t60_dev/0026.JPG" alt="" />
+ </p>
+
+ <p>
+ Remove the shielding containing the motherboard, then flip it over. Remove these screws, placing them on a steady
+ surface in the same layout as they were in before you removed them. Also, you should mark each screw hole after removing the
+ screw (a permanent marker pen will do), this is so that you have a point of reference when re-assembling the machine:<br/>
+ <img src="../images/t60_dev/0027.JPG" alt="" /> <img src="../images/t60_dev/0028.JPG" alt="" /> <img src="../images/t60_dev/0029.JPG" alt="" />
+ <img src="../images/t60_dev/0031.JPG" alt="" /> <img src="../images/t60_dev/0032.JPG" alt="" /> <img src="../images/t60_dev/0033.JPG" alt="" />
+ </p>
+
+ <p>
+ Remove microphone (soldering iron not needed. Just wedge it out gently):<br/>
+ <img src="../images/t60_dev/0039.JPG" alt="" /><br/>
+ <b>Rationale:</b><br/>
+ Another reason to remove the microphone: If your computer gets<a href="#ref1">[1]</a> compromised, it can
+ record what you say, and use it to receive data from nearby devices if
+ they're compromised too. Also, we do not know what the built-in microcode (in the CPU) is doing; it could theoretically
+ be programmed to accept remote commands from some speaker somewhere (remote security hole). <b>In other words,
+ the machine could already be compromised from the factory.</b>
+ </p>
+
+ <p>
+ Remove infrared:<br/>
+ <img src="../images/t60_dev/0040.JPG" alt="" /> <img src="../images/t60_dev/0042.JPG" alt="" />
+ </p>
+
+ <p>
+ Remove cardbus (it's in a socket, no need to disable. Just remove the port itself):<br/>
+ <img src="../images/t60_dev/0041.JPG" alt="" /><br/>
+ <b>Rationale:</b><br/>
+ It has direct memory access and can be used to extract sensitive details (such as LUKS keys). See
+ 'GoodBIOS' video linked at the end (speaker is Peter Stuge, a coreboot hacker). The video covers X60
+ but the same topics apply to T60.
+ </p>
+
+ <p>
+ Before re-installing the upper chassis, remove the speaker:<br/>
+ <img src="../images/t60_dev/0043.JPG" alt="" /> <img src="../images/t60_dev/0044.JPG" alt="" /><br/>
+ Reason: combined with the microphone issue, this could be used to leak data.<br/>
+ If your computer gets<a href="#ref1">[1]</a> compromised, it can be used to
+ transmit data to nearby compromised devices. It's unknown if it can be
+ turned into a microphone<a href="#ref2">[2]</a>.<br/>
+ Replacement: headphones/speakers (line-out) or external DAC (USB).
+ </p>
+
+ <p>
+ Remove the wwan:<br/>
+ <img src="../images/t60_dev/0045.JPG" alt="" /><br/>
+ <b>Wwan (3d modem):</b> They run proprietary software and have access to the
+ computer's RAM! So it's like AMT but over the GSM network which is
+ probably even worse.<br/>
+ Replacement: external USB wifi dongle. (or USB wwan/3g dongle; note, this has all the same privacy issues as mobile phones. wwan not recommended).
+ </p>
+
+ <p>
+ This is where the simcard connector is soldered. See notes above about wwan. Remove simcard by removing battery
+ and then it's accessible (so, remember to do this when you re-assemble. or you could do it now?)<br/>
+ <img src="../images/t60_dev/0046.JPG" alt="" />
+ </p>
+
+ <p>
+ Put those screws back:<br/>
+ <img src="../images/t60_dev/0047.JPG" alt="" />
+ </p>
+
+ <p>
+ Put it back into lower chassis:<br/>
+ <img src="../images/t60_dev/0048.JPG" alt="" />
+ </p>
+
+ <p>
+ Attach LCD and insert screws (also, attach the lcd cable to the board):<br/>
+ <img src="../images/t60_dev/0049.JPG" alt="" />
+ </p>
+
+ <p>
+ Insert those screws:<br/>
+ <img src="../images/t60_dev/0050.JPG" alt="" />
+ </p>
+
+ <p>
+ On the CPU (and there is another chip south-east to it, sorry forgot to take pic)
+ clean off the old thermal paste (rubbing a1ocheal (misspelling intentional. halal internet)) and apply new (Artic Silver 5 is good, others are good too)
+ you should also clean the heatsink the same way<br/>
+ <img src="../images/t60_dev/0051.JPG" alt="" />
+ </p>
+
+ <p>
+ Attach the heatsink and install the screws (also, make sure to install the AC jack as highlighted):<br/>
+ <img src="../images/t60_dev/0052.JPG" alt="" />
+ </p>
+
+ <p>
+ Reinstall that upper bezel:<br/>
+ <img src="../images/t60_dev/0053.JPG" alt="" />
+ </p>
+
+ <p>
+ Do that:<br/>
+ <img src="../images/t60_dev/0054.JPG" alt="" /> <img src="../images/t60_dev/0055.JPG" alt="" />
+ </p>
+
+ <p>
+ Attach keyboard and install nvram battery:<br/>
+ <img src="../images/t60_dev/0056.JPG" alt="" /> <img src="../images/t60_dev/0057.JPG" alt="" />
+ </p>
+
+ <p>
+ Place keyboard and (sorry, forgot to take pics) reinstall the palmrest and insert screws on the underside:<br/>
+ <img src="../images/t60_dev/0058.JPG" alt="" />
+ </p>
+
+ <p>
+ Remove those covers and unscrew:<br/>
+ <img src="../images/t60_dev/0059.JPG" alt="" /> <img src="../images/t60_dev/0060.JPG" alt="" /> <img src="../images/t60_dev/0061.JPG" alt="" />
+ </p>
+
+ <p>
+ Gently pry off the front bezel (sorry, forgot to take pics).
+ </p>
+
+ <p>
+ Remove bluetooth module:<br/>
+ <img src="../images/t60_dev/0062.JPG" alt="" /> <img src="../images/t60_dev/0063.JPG" alt="" />
+ </p>
+
+ <p>
+ Re-attach the front bezel and re-insert the screws (sorry, forgot to take pics).
+ </p>
+
+ <p>
+ It lives!<br/>
+ <img src="../images/t60_dev/0071.JPG" alt="" /> <img src="../images/t60_dev/0072.JPG" alt="" /> <img src="../images/t60_dev/0073.JPG" alt="" />
+ </p>
+
+ <p>
+ Always stress test ('stress -c 2' and xsensors. below 90C is ok) when replacing cpu paste/heatsink:<br/>
+ <img src="../images/t60_dev/0074.JPG" alt="" />
+ </p>
+
+ <h2>
+ Not covered yet:
+ </h2>
+ <ul>
+ <li>Disable flashing the ethernet firmware</li>
+ <li>Disable SPI flash writes (can be re-enabled by unsoldering two parts)</li>
+ <li>Disable use of xrandr/edid on external monitor (cut 2 pins on VGA)</li>
+ <li>Disable docking station (might be possible to do it in software, in coreboot upstream as a Kconfig option)</li>
+ </ul>
+ <p>
+ Go to <a href="http://media.ccc.de/browse/congress/2013/30C3_-_5529_-_en_-_saal_2_-_201312271830_-_hardening_hardware_and_choosing_a_goodbios_-_peter_stuge.html">http://media.ccc.de/browse/congress/2013/30C3_-_5529_-_en_-_saal_2_-_201312271830_-_hardening_hardware_and_choosing_a_goodbios_-_peter_stuge.html</a>
+ or directly to the video: <a href="http://mirror.netcologne.de/CCC/congress/2013/webm/30c3-5529-en-Hardening_hardware_and_choosing_a_goodBIOS_webm.webm">http://mirror.netcologne.de/CCC/congress/2013/webm/30c3-5529-en-Hardening_hardware_and_choosing_a_goodBIOS_webm.webm</a>.
+ </p>
+ <p>
+ A lot of this tutorial is based on that video. Look towards the second half of the video to see how to do the above.
+ </p>
+
+ <h2>
+ Also not covered yet:
+ </h2>
+ <ul>
+ <li>
+ Intrusion detection: randomized seal on screws<br/>
+ Just put nail polish with lot of glider on the important screws, take
+ some good pictures. Keep the pictueres and make sure of their integrity.
+ Compare the nail polish with the pictures before powering on the laptop.
+ </li>
+ <li>
+ Tips about preventing/mitigating risk of cold boot attack.
+ <ul>
+ <li>soldered RAM?</li>
+ <li>wipe all RAM at boot/power-off/power-on? (patch in coreboot upstream?)</li>
+ <li>ask gnutoo about fallback patches (counts number of boots)</li>
+ </ul>
+ </li>
+ <li>
+ General tips/advice and web links showing how to detect physical intrusions.
+ </li>
+ <li>
+ For example: <a href="http://cs.tau.ac.il/~tromer/acoustic/">http://cs.tau.ac.il/~tromer/acoustic/</a>
+ </li>
+ <li>
+ https://gitorious.org/gnutoo-for-coreboot/grub-assemble/source/a61f636797777a742f65f4c9c58032aa6a9b23c3:
+ </li>
+ </ul>
+
+ <h1>
+ Extra notes
+ </h1>
+ <p>
+ EC: Cannot be removed but can be mitigated: it contains non-free
+ non-loadable code, but it has no access to the computer's RAM.
+ It has access to the on-switch of the wifi, bluetooth, modem and some
+ other power management features. The issue is that it has access to the
+ keyboard, however if the software security howto <b>(not yet written)</b> is followed correctly,
+ it won't be able to leak data to a local attacker. It has no network
+ access but it may still be able to leak data remotely, but that
+ requires someone to be nearby to recover the data with the help of an
+ SDR and some directional antennas<a href="#ref3">[3]</a>.
+ </p>
+ <p>
+ <a href="http://www.coreboot.org/Intel_82573_Ethernet_controller">Intel 82573 Ethernet controller</a>
+ on the X60 seems safe, according to Denis.
+ </p>
+
+ <h2>
+ Risk level
+ </h2>
+ <ul>
+ <li>Modem (3g/wwan): highest</li>
+ <li>Intel wifi: Near highest</li>
+ <li>Atheros PCI wifi: unknown, but lower than intel wifi.</li>
+ <li>Microphone: only problematic if the computer gets compromised.</li>
+ <li>Speakers: only problematic if the computer gets compromised.</li>
+ <li>EC: can be mitigated if following the guide on software security.</li>
+ </ul>
+
+ <h1>
+ Further reading material (software security)
+ </h1>
+ <ul>
+ <li><a href="../gnulinux/encrypted_trisquel.html">Installing Trisquel GNU/Linux with full disk encryption (including /boot)</a></li>
+ <li><a href="../gnulinux/encrypted_parabola.html">Installing Parabola GNU/Linux with full disk encryption (including /boot)</a></li>
+ <li><a href="dock.html">Notes about DMA access and the docking station</a></li>
+ </ul>
+
+ <h1>
+ References
+ </h1>
+ <h2 id="ref1">[1] physical access</h2>
+ <p>
+ Explain that black hats, TAO, and so on might use a 0day to get in,
+ and explain that in this case it mitigates what the attacker can do.
+ Also the TAO do some evaluation before launching an attack: they take
+ the probability of beeing caught into account, along with the kind of
+ target. A 0day costs a lot of money, I heard that it was from 100000$
+ to 400000$, some other websites had prices 10 times lower but that
+ but it was probably a typo. So if people increase their security it
+ makes it more risky and more costly to attack people.
+ </p>
+ <h2 id="ref2">[2] microphone</h2>
+ <p>
+ It's possible to turn headphones into a microphone, you could try
+ yourself, however they don't record loud at all. Also intel cards have
+ the capability to change a connector's function, for instance the
+ microphone jack can now become a headphone plug, that's called
+ retasking. There is some support for it in GNU/Linux but it's not very
+ well known.
+ </p>
+ <h2 id="ref3">[3] Video (CCC)</h2>
+ <p>
+ 30c3-5356-en-Firmware_Fat_Camp_webm.webm from the 30th CCC. While
+ their demo is experimental(their hardware also got damaged during the
+ transport), the spies probably already have that since a long time.
+ <a href="http://berlin.ftp.media.ccc.de/congress/2013/webm/30c3-5356-en-Firmware_Fat_Camp_webm.webm">http://berlin.ftp.media.ccc.de/congress/2013/webm/30c3-5356-en-Firmware_Fat_Camp_webm.webm</a>
+ </p>
+
+<hr/>
+
+ <p>
+ Copyright &copy; 2014 Francis Rowe &lt;info@gluglug.org.uk&gt;<br/>
+ This document is released under the Creative Commons Attribution-ShareAlike 4.0 International Public License and all future versions.
+ A copy of the license can be found at <a href="../license.txt">../license.txt</a>.
+ </p>
+
+ <p>
+ This document is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See <a href="../license.txt">../license.txt</a> for more information.
+ </p>
+
+</body>
+</html>
diff --git a/docs/security/x60_security.html b/docs/security/x60_security.html
new file mode 100644
index 0000000..33ccb6d
--- /dev/null
+++ b/docs/security/x60_security.html
@@ -0,0 +1,301 @@
+<!DOCTYPE html>
+<html>
+<head>
+ <meta charset="utf-8">
+ <meta name="viewport" content="width=device-width, initial-scale=1">
+
+ <style type="text/css">
+ @import url('../css/main.css');
+ </style>
+
+ <title>Libreboot documentation: Security on the ThinkPad X60</title>
+</head>
+
+<body>
+
+ <header>
+ <h1>Security on the ThinkPad X60</h1>
+ <aside>Hardware modifications to enhance security on the ThinkPad X60. This tutorial is <b>incomplete</b> at the time of writing.</aside>
+ </header>
+
+ <p>Or go <a href="index.html">back to main index</a></p>
+
+ <h2>Table of Contents</h2>
+ <ul>
+ <li><a href="#hardware_requirements">Hardware Requirements</a></li>
+ <li><a href="#software_requirements">Software Requirements</a></li>
+ <li><a href="#procedure">The procedure</a></li>
+ </ul>
+
+ <h1 id="hardware_requirements">Hardware requirements</h1>
+ <ul>
+ <li>An X60</li>
+ <li>screwdriver</li>
+ <li>(in a later version of this tutorial: soldering iron and scalpel)</li>
+ </ul>
+
+ <h1 id="software_requirements">Software requirements</h1>
+ <ul>
+ <li>none (at least in the scope of the article as-is)</li>
+ <li>You probably want to encrypt your GNU/Linux install using LUKS</li>
+ </ul>
+
+ <h1>
+ Rationale
+ </h1>
+ <p>
+ Most people think of security on the software side: the hardware is important aswell.
+ Hardware security is useful in particular to journalists (or activists in a given movement) who need absolute privacy in their work.
+ It is also generally useful to all those that believe security and privacy are inalienable rights.
+ Security starts with the hardware; crypto and network security come later.
+ </p>
+ <p>
+ Paradoxically, going this far to increase your security also makes you a bigger target.
+ At the same time, it protects you in the case that someone does attack your machine.
+ This paradox only exists while few people take adequate steps to protect yourself: it is your <b>duty</b>
+ to protect yourself, not only for your benefit but to make strong security <i>normal</i> so
+ that those who do need protection (and claim it) are a smaller target against the masses.
+ </p>
+ <p>
+ Even if there are levels of security beyond your ability (technically, financially and so on)
+ doing at least <i>something</i> (what you are able to do) is extremely important.
+ If you use the internet and your computer without protection, attacking you is cheap (some say it is
+ only a few US cents). If everyone (majority of people) use strong security by default,
+ it makes attacks more costly and time consuming; in effect, making them disappear.
+ </p>
+ <p>
+ This tutorial deals with reducing the number of devices that have direct memory access that
+ could communicate with inputs/outputs that could be used to remotely
+ command the machine (or leak data).
+ </p>
+
+ <h1 id="procedure">Disassembly</h1>
+
+ <p>
+ Firstly remove the bluetooth (if your X60 has this):<br/>
+ The marked screws are underneath those stickers (marked in those 3 locations at the bottom of the LCD assembly):<br/>
+ <img src="../images/x60_security/0000_bluetooth0.jpg" alt="" /><br/>
+ Now gently pry off the bottom part of the front bezel, and the bluetooth module is on the left (easily removable):<br/>
+ <img src="../images/x60_security/0000_bluetooth.jpg" alt="" /><br/>
+ </p>
+
+ <p>
+ If your model was WWAN, remove the simcard (check anyway):<br/>
+ Uncover those 2 screws at the bottom:<br/>
+ <img src="../images/x60_security/0000_simcard0.jpg" alt="" /><br/>
+ SIM card (not present in the picture) is in the marked location:<br/>
+ <img src="../images/x60_security/0000_simcard1.jpg" alt="" /><br/>
+ Replacement: USB dongle.
+ </p>
+
+ <p>
+ Now get into the motherboard.
+ </p>
+
+ <p>
+ Remove those screws:<br/>
+ <img src="../images/x60_security/0000.jpg" alt="" />
+ </p>
+ <p>
+ Push the keyboard forward (carefully):<br/>
+ <img src="../images/x60_security/0001.jpg" alt="" />
+ </p>
+ <p>
+ Lift the keyboard up and disconnect it from the board:<br/>
+ <img src="../images/x60_security/0002.jpg" alt="" />
+ </p>
+ <p>
+ Grab the right-hand side of the chassis and force it off (gently) and pry up the rest of the chassis:<br/>
+ <img src="../images/x60_security/0003.jpg" alt="" />
+ </p>
+ <p>
+ You should now have this:<br/>
+ <img src="../images/x60_security/0004.jpg" alt="" />
+ </p>
+
+ <p>
+ The following is a summary of what you will remove (already done to this machine):<br/>
+ <img src="../images/x60_security/0001_overview.jpg" alt="" /><br/>
+ Note: the blue lines represent antenna cables and modem cables. You don't need to remove these, but you can if you want
+ (to make it tidier after removing other parts). I removed the antenna wires, the modem jack, the modem cable and
+ also (on another model) a device inside the part where the wwan antenna goes (wasn't sure what it was, but I knew it wasn't needed). <b>This is optional</b>
+ </p>
+
+ <p>
+ Remove the microphone (can desolder it, but you can also easily pull it off with you hands). Already removed here:<br/>
+ <img src="../images/x60_security/0001_microphone.jpg" alt="" /><br/>
+ <b>Rationale:</b><br/>
+ Another reason to remove the microphone: If your computer gets<a href="#ref1">[1]</a> compromised, it can
+ record what you say, and use it to receive data from nearby devices if
+ they're compromised too. Also, we do not know what the built-in microcode (in the CPU) is doing; it could theoretically
+ be programmed to accept remote commands from some speaker somewhere (remote security hole). <b>In other words,
+ the machine could already be compromised from the factory.</b>
+ </p>
+
+ <p>
+ Remove the modem:<br/>
+ <img src="../images/x60_security/0001_modem.jpg" alt="" /><br/>
+ (useless, obsolete device)
+ </p>
+
+ <p>
+ Remove the speaker:<br/>
+ <img src="../images/x60_security/0001_speaker.jpg" alt="" /><br/>
+ Reason: combined with the microphone issue, this could be used to leak data.<br/>
+ If your computer gets<a href="#ref1">[1]</a> compromised, it can be used to
+ transmit data to nearby compromised devices. It's unknown if it can be
+ turned into a microphone<a href="#ref2">[2]</a>.<br/>
+ Replacement: headphones/speakers (line-out) or external DAC (USB).
+ </p>
+
+ <p>
+ Remove the wlan (also remove wwan if you have it):<br/>
+ <img src="../images/x60_security/0001_wlan_wwan.jpg" alt="" /><br/>
+ Reason: has direct (and very fast) memory access, and could (theoretically) leak data over a side-channel.<br/>
+ <b>Wifi:</b> The ath5k/ath9k cards might not have firmware at all. They might safe but could have
+ access to the computer's RAM trough DMA. If people have an intel
+ card(most X60's come with Intel wifi by default, until you change it),then that card runs
+ a non-free firwamre and has access to the computer's RAM trough DMA! So
+ it's risk-level is very high.<br/>
+ <b>Wwan (3d modem):</b> They run proprietary software and have access to the
+ computer's RAM! So it's like AMT but over the GSM network which is
+ probably even worse.<br/>
+ Replacement: external USB wifi dongle. (or USB wwan/3g dongle; note, this has all the same privacy issues as mobile phones. wwan not recommended).
+ </p>
+
+ <h2>
+ Not covered yet:
+ </h2>
+ <ul>
+ <li>Disable cardbus (has fast/direct memory access)</li>
+ <li>Disable firewire (has fast/direct memory access)</li>
+ <li>Disable flashing the ethernet firmware</li>
+ <li>Disable SPI flash writes (can be re-enabled by unsoldering two parts)</li>
+ <li>Disable use of xrandr/edid on external monitor (cut 2 pins on VGA)</li>
+ <li>Disable docking station (might be possible to do it in software, in coreboot upstream as a Kconfig option)</li>
+ </ul>
+ <p>
+ Go to <a href="http://media.ccc.de/browse/congress/2013/30C3_-_5529_-_en_-_saal_2_-_201312271830_-_hardening_hardware_and_choosing_a_goodbios_-_peter_stuge.html">http://media.ccc.de/browse/congress/2013/30C3_-_5529_-_en_-_saal_2_-_201312271830_-_hardening_hardware_and_choosing_a_goodbios_-_peter_stuge.html</a>
+ or directly to the video: <a href="http://mirror.netcologne.de/CCC/congress/2013/webm/30c3-5529-en-Hardening_hardware_and_choosing_a_goodBIOS_webm.webm">http://mirror.netcologne.de/CCC/congress/2013/webm/30c3-5529-en-Hardening_hardware_and_choosing_a_goodBIOS_webm.webm</a>.
+ </p>
+ <p>
+ A lot of this tutorial is based on that video. Look towards the second half of the video to see how to do the above.
+ </p>
+
+ <h2>
+ Also not covered yet:
+ </h2>
+ <ul>
+ <li>
+ Intrusion detection: randomized seal on screws<br/>
+ Just put nail polish with lot of glider on the important screws, take
+ some good pictures. Keep the pictueres and make sure of their integrity.
+ Compare the nail polish with the pictures before powering on the laptop.
+ </li>
+ <li>
+ Tips about preventing/mitigating risk of cold boot attack.
+ <ul>
+ <li>soldered RAM?</li>
+ <li>seal RAM door shut (possibly modified lower chassis) so that system has to be disassembled (which has to go through the nail polish)</li>
+ <li>wipe all RAM at boot/power-off/power-on? (patch in coreboot upstream?)</li>
+ <li>ask gnutoo about fallback patches (counts number of boots)</li>
+ </ul>
+ </li>
+ <li>
+ General tips/advice and web links showing how to detect physical intrusions.
+ </li>
+ <li>
+ For example: <a href="http://cs.tau.ac.il/~tromer/acoustic/">http://cs.tau.ac.il/~tromer/acoustic/</a>
+ </li>
+ <li>
+ https://gitorious.org/gnutoo-for-coreboot/grub-assemble/source/a61f636797777a742f65f4c9c58032aa6a9b23c3:
+ </li>
+ </ul>
+
+ <h1>
+ Extra notes
+ </h1>
+ <p>
+ EC: Cannot be removed but can be mitigated: it contains non-free
+ non-loadable code, but it has no access to the computer's RAM.
+ It has access to the on-switch of the wifi, bluetooth, modem and some
+ other power management features. The issue is that it has access to the
+ keyboard, however if the software security howto <b>(not yet written)</b> is followed correctly,
+ it won't be able to leak data to a local attacker. It has no network
+ access but it may still be able to leak data remotely, but that
+ requires someone to be nearby to recover the data with the help of an
+ SDR and some directional antennas<a href="#ref3">[3]</a>.
+ </p>
+ <p>
+ <a href="http://www.coreboot.org/Intel_82573_Ethernet_controller">Intel 82573 Ethernet controller</a>
+ on the X60 seems safe, according to Denis.
+ </p>
+
+ <h2>
+ Risk level
+ </h2>
+ <ul>
+ <li>Modem (3g/wwan): highest</li>
+ <li>Intel wifi: Near highest</li>
+ <li>Atheros PCI wifi: unknown, but lower than intel wifi.</li>
+ <li>Microphone: only problematic if the computer gets compromised.</li>
+ <li>Speakers: only problematic if the computer gets compromised.</li>
+ <li>EC: can be mitigated if following the guide on software security.</li>
+ </ul>
+
+ <h1>
+ Further reading material (software security)
+ </h1>
+ <ul>
+ <li><a href="../gnulinux/encrypted_trisquel.html">Installing Trisquel GNU/Linux with full disk encryption (including /boot)</a></li>
+ <li><a href="../gnulinux/encrypted_parabola.html">Installing Parabola GNU/Linux with full disk encryption (including /boot)</a></li>
+ <li><a href="dock.html">Notes about DMA access and the docking station</a></li>
+ </ul>
+
+ <h1>
+ References
+ </h1>
+ <h2 id="ref1">[1] physical access</h2>
+ <p>
+ Explain that black hats, TAO, and so on might use a 0day to get in,
+ and explain that in this case it mitigates what the attacker can do.
+ Also the TAO do some evaluation before launching an attack: they take
+ the probability of beeing caught into account, along with the kind of
+ target. A 0day costs a lot of money, I heard that it was from 100000$
+ to 400000$, some other websites had prices 10 times lower but that
+ but it was probably a typo. So if people increase their security it
+ makes it more risky and more costly to attack people.
+ </p>
+ <h2 id="ref2">[2] microphone</h2>
+ <p>
+ It's possible to turn headphones into a microphone, you could try
+ yourself, however they don't record loud at all. Also intel cards have
+ the capability to change a connector's function, for instance the
+ microphone jack can now become a headphone plug, that's called
+ retasking. There is some support for it in GNU/Linux but it's not very
+ well known.
+ </p>
+ <h2 id="ref3">[3] Video (CCC)</h2>
+ <p>
+ 30c3-5356-en-Firmware_Fat_Camp_webm.webm from the 30th CCC. While
+ their demo is experimental(their hardware also got damaged during the
+ transport), the spies probably already have that since a long time.
+ <a href="http://berlin.ftp.media.ccc.de/congress/2013/webm/30c3-5356-en-Firmware_Fat_Camp_webm.webm">http://berlin.ftp.media.ccc.de/congress/2013/webm/30c3-5356-en-Firmware_Fat_Camp_webm.webm</a>
+ </p>
+
+<hr/>
+
+ <p>
+ Copyright &copy; 2014 Francis Rowe &lt;info@gluglug.org.uk&gt;<br/>
+ This document is released under the Creative Commons Attribution-ShareAlike 4.0 International Public License and all future versions.
+ A copy of the license can be found at <a href="../license.txt">../license.txt</a>.
+ </p>
+
+ <p>
+ This document is distributed in the hope that it will be useful,
+ but WITHOUT ANY WARRANTY; without even the implied warranty of
+ MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See <a href="../license.txt">../license.txt</a> for more information.
+ </p>
+
+</body>
+</html>