OpenBSD Journal

Theo and a soldering iron; always ahead of the curve.

Contributed by sean on from the a-sensor-for-everything-and-for-everything-a-sensor dept.

Theo sent a message to misc@ on Sunday that caught my eye. He wrote a new sensor driver: sdtemp(4) (with a very small amount of code). This is a driver for a specific temperature sensor for memory modules and can be probed via an I2C bus.

In the post, Theo notes that he was unable to get his hands on a DIMM with this sensor chip. However, he managed to get his hands on the specific sensor chip itself and soldered it to a DIMM he already had!

It always amazes me how tenacious the developers are to support anything and everything they can get their hands on; and even some things they can't. If you happen to have this particular piece of gear, please send in a dmesg(8) to dmesg@openbsd.org.

Of course this driver is only in -current.

(Comments are closed)


Comments
  1. By Anonymous Coward (128.171.90.200) on

    Wow, thats kinda cool

    Comments
    1. By Anonymous Coward (79.114.19.9) on

      > Wow, thats kinda cool

      Nothing cool here. Just some taste for diversity. It's boring being only in software. :-D

      Comments
      1. By Anonymous Coward (128.171.90.200) on

        > > Wow, thats kinda cool
        >
        > Nothing cool here. Just some taste for diversity. It's boring being only in software. :-D

        That soldering iron is pretty hot

    2. By Motley Fool (MotleyFool) on

      > Wow, thats kinda cool

      hardware hacking is way more cool than software stuff, that's why I have a bunch of Plextors with serial consoles running OpenBSD

  2. By Anonymous Coward (193.25.105.76) on

    This is amazing news ! But PLEASE add support for Attansic Technology L1 network card !

    Comments
    1. By Anonymous Coward (24.113.150.55) on

      > This is amazing news ! But PLEASE add support for Attansic Technology L1 network card !

      Do you have one? Ask Theo which dev you should mail it to, and bug the manufacturer for documentation. They can't add support for anything without hardware, and docs make it much, much easier.

    2. By Chris (LizardKing) on http://www.chriswareham.net/

      > This is amazing news ! But PLEASE add support for Attansic Technology L1 network card !

      I assume you want support for this because it's found in the Asus EEE. If so, you may be able to port the driver from NetBSD, as one was added to -current recently.

      Comments
      1. By Joel Sing (58.6.65.97) jsing@openbsd.org on

        > > This is amazing news ! But PLEASE add support for Attansic Technology
        > > L1 network card !
        >
        > I assume you want support for this because it's found in the Asus EEE.
        > If so, you may be able to port the driver from NetBSD, as one was added
        > to -current recently.

        The ASUS EeePC has an Attansic L2 NIC, support for which is already available in -current. And as per a previous post, if you want support for a specific piece of hardware you need to find a way to get it to a developer, preferably with documentation.

        Comments
        1. By Chris (LizardKing) on http://www.chriswareham.net/

          > > > This is amazing news ! But PLEASE add support for Attansic Technology
          > > > L1 network card !
          > >
          > > I assume you want support for this because it's found in the Asus EEE.
          > > If so, you may be able to port the driver from NetBSD, as one was added
          > > to -current recently.
          >
          > The ASUS EeePC has an Attansic L2 NIC, support for which is already available in -current. And as per a previous post, if you want support for a specific piece of hardware you need to find a way to get it to a developer, preferably with documentation.

          Whoops, my bad. I'd forgotten it's the L*2* that's in the Eee.

          Comments
          1. By Anonymous Coward (193.25.105.76) on

            > > > > This is amazing news ! But PLEASE add support for Attansic Technology
            > > > > L1 network card !
            > > >
            > > > I assume you want support for this because it's found in the Asus EEE.
            > > > If so, you may be able to port the driver from NetBSD, as one was added
            > > > to -current recently.
            > >
            > > The ASUS EeePC has an Attansic L2 NIC, support for which is already available in -current. And as per a previous post, if you want support for a specific piece of hardware you need to find a way to get it to a developer, preferably with documentation.
            >
            > Whoops, my bad. I'd forgotten it's the L*2* that's in the Eee.

            Actually I already emailed the manufacturer to release the docs for the Attansic Technology L1 network card to OpenBSD project and I warned them I will never buy an Attansic product again if they fail to answer my request. The Attansic Technology L1 network chips are now widely used because they are cheap (embedded in Asus P5K-V motherboard in my case). There is currently a linux driver for this chip. Maybe there is someone skilled enough and interested to port this linux driver to OpenBSD.

            Comments
            1. By Anonymous Coward (193.25.105.76) on

              > > > > > This is amazing news ! But PLEASE add support for Attansic Technology
              > > > > > L1 network card !
              > > > >
              > > > > I assume you want support for this because it's found in the Asus EEE.
              > > > > If so, you may be able to port the driver from NetBSD, as one was added
              > > > > to -current recently.
              > > >
              > > > The ASUS EeePC has an Attansic L2 NIC, support for which is already available in -current. And as per a previous post, if you want support for a specific piece of hardware you need to find a way to get it to a developer, preferably with documentation.
              > >
              > > Whoops, my bad. I'd forgotten it's the L*2* that's in the Eee.
              >
              > Actually I already emailed the manufacturer to release the docs for the Attansic Technology L1 network card to OpenBSD project and I warned them I will never buy an Attansic product again if they fail to answer my request. The Attansic Technology L1 network chips are now widely used because they are cheap (embedded in Asus P5K-V motherboard in my case). There is currently a linux driver for this chip. Maybe there is someone skilled enough and interested to port this linux driver to OpenBSD.

              There is also a driver in FreeBSD for this chip http://unix.derkeiler.com/Mailing-Lists/FreeBSD/current/2008-03/msg00224.html

  3. By Andy Elvey (viking) andy.elvey@paradise.net.nz on

    Good stuff!

    Theo should be on "Home Improvement"
    (hammer,hammer,solder,solder... ;) )

    Comments
    1. By Anonymous Coward (84.169.215.130) on

      > Good stuff!
      >
      > Theo should be on "Home Improvement"
      > (hammer,hammer,solder,solder... ;) )
      >
      I don't think so, Theo.

      Comments
      1. By Anonymous Coward (203.97.97.11) on

        > > Good stuff!
        > >
        > > Theo should be on "Home Improvement"
        > > (hammer,hammer,solder,solder... ;) )
        > >
        > I don't think so, Theo.
        >
        Theo armed with a soldering iron - watch out RMS...

    2. By Anonymous Coward (67.170.78.135) on

      > Good stuff!
      >
      > Theo should be on "Home Improvement"
      > (hammer,hammer,solder,solder... ;) )
      >
      >

      MORE POWER! ar, ar, ar

  4. By Anonymous Coward (76.217.209.163) on

    Nothing wrong with wire wrapping, the smell of solder spoils the taste of the bier. Cheers on taking the project to the bleeding edge, some more!

  5. By Pieter Verberne (80.126.42.203) pieterverberne@xs4all.nl on

    So now I can also donate my soldering iron? That's a bit cheaper than a Sun Blade=]

  6. By Jason Meltzer (meltzer) strangeresearch@gmail.com on

    Since no one has asked the pertinent question, 'what interesting things
    might you use the readings off this sensor for?' (besides the usual system
    monitoring) I would like to point everyone to a great paper written by
    Bill Paul, Jacob Appelbaum, and some of Ed Felten's Grad students:

    Lest We Remember: Cold Boot Attacks on Encryption Keys.

    You might just have time to do something when you sense temps
    on your dimms dropping precipitously...

    Comments
    1. By Anonymous Coward (79.65.224.150) on

      > Since no one has asked the pertinent question, 'what interesting things might you use the readings off this sensor for?' (besides the usual system monitoring) I would like to point everyone to a great paper written by Bill Paul, Jacob Appelbaum, and some of Ed Felten's Grad students:
      >
      > Lest We Remember: Cold Boot Attacks on Encryption Keys.
      >
      >
      > You might just have time to do something when you sense temps
      > on your dimms dropping precipitously...
      >

      This argument is getting REALLY old already, if you can't secure your hardware you're fucked regardless, STFU n00b!

      Comments
      1. By Jason Meltzer (meltzer) on

        > This argument is getting REALLY old already, if you can't secure your,
        >hardware you're fucked regardless, STFU n00b!

        Wow, classy and smart! Or is that crude and thoughtless, silly me I always
        get them confused!

        There are absolutely no circumstances where one might have any data that
        they value on say something portable like, ummm, a laptop or something
        accessible by law enforcement or other high capability agents. Right? No, well then.

        There are obvious threat profiles that require you to take into
        consideration loss of physical ownership. Even if we consider them
        edge cases, are you really so silly as to suggest just giving
        up and accepting those risks? Hmmm, well I guess you are. Now, just
        because *you* are doesn't mean that other people aren't, unless I
        missed the memo about you being appointed Lord High Risk Manager of
        Earth.

        Why do anything for that matter? Code and data are interchangeable on
        most of the computer architectures we all use, so I guess we are
        just fundamentally fucked, aren't we?

        All this fucking is making me thirsty, time for more coffee!

        Comments
        1. By Anonymous Coward (79.73.219.81) on

          > This argument is getting REALLY old already, if you can't secure your, hardware you're fucked regardless, STFU n00b!
          >
          > Wow, classy and smart! Or is that crude and thoughtless, silly me I always get them confused!
          >
          > There are absolutely no circumstances where one might have any data that they value on say something portable like, ummm, a laptop or something accessible by law enforcement or other high capability agents. Right? No, well then.
          >
          > There are obvious threat profiles that require you to take into consideration loss of physical ownership. Even if we consider them edge cases, are you really so silly as to suggest just giving up and accepting those risks? Hmmm, well I guess you are. Now, just because *you* are doesn't mean that other people aren't, unless I missed the memo about you being appointed Lord High Risk Manager of Earth.
          >
          > Why do anything for that matter? Code and data are interchangeable on most of the computer architectures we all use, so I guess we are just fundamentally fucked, aren't we?
          >
          > All this fucking is making me thirsty, time for more coffee!

          I see your point now, if you don't want the feds or five o discovering remnants of your home made goat sex pr0n sex tape then you should just wire a remotely activated, motion sensitive, all singing all dancing super tiny bomb to your memory and blow it away.

          If that fails you can grab your goat, tuck it beneath your arm and sprint away as fast as you can.

          Thanks for making an already overly paranoid set of OS users even more anal about temperature sensors!

          Comments
          1. By Anonymous Coward (216.68.197.23) on

            RE: ~'...paranoid..OS...temperature sensors..' < Reminds me of the old computer game, Suspended, where all data is potentially important to what is going on.

            All knowledge about interaction of hardware, OS and sensors, is a great thing.

            Being able to add a sensor to existing *KNOWN* *PROVEN* older hardware is a great thing in a post 911 world.

            Although the memory cooling encryption issues are now being discussed, thats just the tip of the iceberg. How about a radio covert channel that works through a weird temp sensor, would be cool and helpful as well for some environments. Or just perhaps a connection to ground/whatever, that only occurs when system is whatever/hot.

            Being ahead of the curve with the system is a very cool or hot subject indeed.

    2. By Anonymous Coward (128.171.90.200) on

      > Since no one has asked the pertinent question, 'what interesting things might you use the readings off this sensor for?' (besides the usual system monitoring) I would like to point everyone to a great paper written by Bill Paul, Jacob Appelbaum, and some of Ed Felten's Grad students:
      >
      > Lest We Remember: Cold Boot Attacks on Encryption Keys.
      >
      >
      > You might just have time to do something when you sense temps
      > on your dimms dropping precipitously...

      You do not seem to understand the research you are pointing to, a cold boot attack in this case has nothing to do with temperature.

      In your defense, I guess if some ner-do-well were freezing your RAM, you may be aware that the temperature dropped quickly, but then you have to accept you lost the box and there is nothing you can do about it.

      Comments
      1. By Jason Meltzer (meltzer) on

        > >
        > > You might just have time to do something when you sense temps
        > > on your dimms dropping precipitously...
        >
        > You do not seem to understand the research you are pointing to, a
        >cold boot attack in this case has nothing to do with temperature.

        I do understand the paper and I'm perfectly aware of this fact...
        cooling the memory simply increases the reliability of data
        recovery. I was only pointing out an interesting, atypical, use for
        this sensor. FWIW Theo brought up the existence of these sensors in
        a discussion with one of the paper's authors at CanSecWest and no
        one seemed to think it was silly then.

        > In your defense, I guess if some ner-do-well were freezing your RAM,
        >you may be aware that the temperature dropped quickly, but then you
        >have to accept you lost the box and there is nothing you can do about it.

        And I agree to a significant degree, but there is no reason not to
        think about these things. Apparently excepting getting flamed by
        trolls (not you!). Meh.

    3. By Anonymous Coward (79.114.19.83) on

      Too little info presented, I thought I'm watching a Myth Busters episode. So for the moment I don'g buy it. Why the hell do you need the DDRAM refresh then so often ? I'm not an ignorant, but what the heck this is no real experiment, it looks like a home made amusement.

Credits

Copyright © - Daniel Hartmeier. All rights reserved. Articles and comments are copyright their respective authors, submission implies license to publish on this web site. Contents of the archive prior to as well as images and HTML templates were copied from the fabulous original deadly.org with Jose's and Jim's kind permission. This journal runs as CGI with httpd(8) on OpenBSD, the source code is BSD licensed. undeadly \Un*dead"ly\, a. Not subject to death; immortal. [Obs.]