3 ms·

Is anyone getting flashbacks to "The Big Hack"? It feels like maybe Bloomberg knew of something but got the wrong root cause- it seems a lot more likely for so

Is anyone getting flashbacks to "The Big Hack"? It feels like maybe Bloomberg knew of something but got the wrong root cause- it seems a lot more likely for someone to sneak in a slightly-reprogrammed BMC than change board layouts /etc., especially considering just how much control BMC's have. Yikes.

Animats 7y ago
That was my concern. All it would take is a built-in IPMI password. For extra credit, a built-in IPMI password that wasn't listed if you list IPMI passwords. So I search for "IPMI password", and get this.[1] "On modern Supermicro IPMI interfaces the default login/ password is:" Login: ADMIN Password: ADMIN That's convenient. Well-documented, too. Is that enabled by default? [1] https://forums.servethehome.com/index.php?resources/supermicro-ipmi-default-login-and-password.1/ https://forums.servethehome.com/index.php?resources/supermic...
- monocasa 7y ago
  FWIW, the implant they were describing could absolutely change a default password in the bitstream. That'd be really neat because you could swap out the flash and still be screwed.
wanderingjew 7y ago
Bloomberg knew of _something_, but Bloomberg isn't a news company. They're a hardware company. That info was passed on to the journos who went and did their own thing (Telephone, chinese whispers, and a healthy dose of confirmation bias). I was talking to one of the Bloomberg engineers at a bar about six months before the story dropped. There was _something_ and it was _big_, but didn't get any more information than that. The story dropped, and the story was idiotic, but there was still something there. This is probably it.
- monocasa 7y ago
  What was idiotic about the actual story?
  - wanderingjew 7y ago
    That the theory presented (inserting a microcontroller INSIDE a circuit board) is absolutely the hardest way to do things, especially when the article (esp. the article's graphics) heavily implied this chip was implanted right next to the empty pad of an SOT-8 footprint that was used for boot flash for the BMC. You could just put a chip on that empty footprint and hardly anyone would catch it. Or there are the specific quotes from the Bloomberg piece -- "the microchip altered the operating system's core so it could accept modifications" -- is so obtuse that it not only fails to inform the reader, but also sounds like some sort of pseudo-technobabble from someone who doesn't know what they're talking about. Ostensibly, if this story were true, the sources would say something like, "this chip contained a code that overwrote settings, allowing remote access". You know, something that seems like it passed through an editor. Or there is the fact that the source checking on the story amounted to going to one guy, asking a question, having him say, "yeah, that's possible", coming back a few weeks later with another question, having the source say, "yeah, that's possible too", and assuming these possibilities add up to a certainty. The bloomberg big hack story is the worst researched and most poorly presented piece of writing in recent memory. It is idiocy in the classical greek sense -- it is disconnected from the polis and any semblance of reality. The authors are still writing for Bloomberg, too.
    - monocasa 7y ago
      > That the theory presented (inserting a microcontroller INSIDE a circuit board) is absolutely the hardest way to do things, especially when the article (esp. the article's graphics) heavily implied this chip was implanted right next to the empty pad of an SOT-8 footprint that was used for boot flash for the BMC. You could just put a chip on that empty footprint and hardly anyone would catch it. Except if they were looking for it. An implant inside the board would be persistent across swapping out or re-imaging the the flash, _and_ wouldn't benoticable unless you were x-raying the board. And yeah, the graphics weren't perfect, welcome to journalism. > Or there are the specific quotes from the Bloomberg piece -- "the microchip altered the operating system's core so it could accept modifications" -- is so obtuse that it not only fails to inform the reader, but also sounds like some sort of pseudo-technobabble from someone who doesn't know what they're talking about. Ostensibly, if this story were true, the sources would say something like, "this chip contained a code that overwrote settings, allowing remote access". You know, something that seems like it passed through an editor. IDK, it sounds like the implant would override one of the keys built into the image, whether that's an update key or a remote access key. > Or there is the fact that the source checking on the story amounted to going to one guy, asking a question, having him say, "yeah, that's possible", coming back a few weeks later with another question, having the source say, "yeah, that's possible too", and assuming these possibilities add up to a certainty. > The bloomberg big hack story is the worst researched and most poorly presented piece of writing in recent memory. It is idiocy in the classical greek sense -- it is disconnected from the polis and any semblance of reality. The authors are still writing for Bloomberg, too. I still haven't heard a legitimate reason from you what doesn't make sense here other than vague "this doens't make sense, I'm so much more intelligent than them". And I'm an FPGA developer who's personally screwed up SPI flashes via a very similar mechanisms just from bugs.

Animats 7y ago

That was my concern. All it would take is a built-in IPMI password. For extra credit, a built-in IPMI password that wasn't listed if you list IPMI passwords. So I search for "IPMI password", and get this.[1] "On modern Supermicro IPMI interfaces the default login/ password is:" Login: ADMIN Password: ADMIN That's convenient. Well-documented, too. Is that enabled by default? [1] https://forums.servethehome.com/index.php?resources/supermicro-ipmi-default-login-and-password.1/ https://forums.servethehome.com/index.php?resources/supermic...

monocasa 7y ago
FWIW, the implant they were describing could absolutely change a default password in the bitstream. That'd be really neat because you could swap out the flash and still be screwed.

wanderingjew 7y ago

Bloomberg knew of _something_, but Bloomberg isn't a news company. They're a hardware company. That info was passed on to the journos who went and did their own thing (Telephone, chinese whispers, and a healthy dose of confirmation bias). I was talking to one of the Bloomberg engineers at a bar about six months before the story dropped. There was _something_ and it was _big_, but didn't get any more information than that. The story dropped, and the story was idiotic, but there was still something there. This is probably it.

monocasa 7y ago
What was idiotic about the actual story?
- wanderingjew 7y ago
  That the theory presented (inserting a microcontroller INSIDE a circuit board) is absolutely the hardest way to do things, especially when the article (esp. the article's graphics) heavily implied this chip was implanted right next to the empty pad of an SOT-8 footprint that was used for boot flash for the BMC. You could just put a chip on that empty footprint and hardly anyone would catch it. Or there are the specific quotes from the Bloomberg piece -- "the microchip altered the operating system's core so it could accept modifications" -- is so obtuse that it not only fails to inform the reader, but also sounds like some sort of pseudo-technobabble from someone who doesn't know what they're talking about. Ostensibly, if this story were true, the sources would say something like, "this chip contained a code that overwrote settings, allowing remote access". You know, something that seems like it passed through an editor. Or there is the fact that the source checking on the story amounted to going to one guy, asking a question, having him say, "yeah, that's possible", coming back a few weeks later with another question, having the source say, "yeah, that's possible too", and assuming these possibilities add up to a certainty. The bloomberg big hack story is the worst researched and most poorly presented piece of writing in recent memory. It is idiocy in the classical greek sense -- it is disconnected from the polis and any semblance of reality. The authors are still writing for Bloomberg, too.
  - monocasa 7y ago
    > That the theory presented (inserting a microcontroller INSIDE a circuit board) is absolutely the hardest way to do things, especially when the article (esp. the article's graphics) heavily implied this chip was implanted right next to the empty pad of an SOT-8 footprint that was used for boot flash for the BMC. You could just put a chip on that empty footprint and hardly anyone would catch it. Except if they were looking for it. An implant inside the board would be persistent across swapping out or re-imaging the the flash, _and_ wouldn't benoticable unless you were x-raying the board. And yeah, the graphics weren't perfect, welcome to journalism. > Or there are the specific quotes from the Bloomberg piece -- "the microchip altered the operating system's core so it could accept modifications" -- is so obtuse that it not only fails to inform the reader, but also sounds like some sort of pseudo-technobabble from someone who doesn't know what they're talking about. Ostensibly, if this story were true, the sources would say something like, "this chip contained a code that overwrote settings, allowing remote access". You know, something that seems like it passed through an editor. IDK, it sounds like the implant would override one of the keys built into the image, whether that's an update key or a remote access key. > Or there is the fact that the source checking on the story amounted to going to one guy, asking a question, having him say, "yeah, that's possible", coming back a few weeks later with another question, having the source say, "yeah, that's possible too", and assuming these possibilities add up to a certainty. > The bloomberg big hack story is the worst researched and most poorly presented piece of writing in recent memory. It is idiocy in the classical greek sense -- it is disconnected from the polis and any semblance of reality. The authors are still writing for Bloomberg, too. I still haven't heard a legitimate reason from you what doesn't make sense here other than vague "this doens't make sense, I'm so much more intelligent than them". And I'm an FPGA developer who's personally screwed up SPI flashes via a very similar mechanisms just from bugs.