#qi-hardware IRC log for Tuesday, 2012-12-18

qi-bot[commit] Werner Almesberger: ioscript/ioscript: allow positions to be redefined (master) http://qi-hw.com/p/ben-blinkenlights/b4530ff01:45
qi-bot[commit] Werner Almesberger: swuart/: split general UART functions from application (master) http://qi-hw.com/p/ben-blinkenlights/2e749f901:45
qi-bot[commit] Werner Almesberger: swuart/: merge library part into libubb (master) http://qi-hw.com/p/ben-blinkenlights/1c51cc601:45
qi-bot[commit] Werner Almesberger: swuart/: convert "hammering the peer" demo into a primitive interactive chat (master) http://qi-hw.com/p/ben-blinkenlights/519ebe601:45
qi-bot[commit] Werner Almesberger: libubb/swuart.c (swuart_open): start TX high, not low (master) http://qi-hw.com/p/ben-blinkenlights/842766b01:45
kyakwpwrak: are you thinking about a particular use case for the SW UART?06:00
kyakis it possible to use it for serial console?06:00
kyaki mean, use UBB for serial console then06:00
wpwrakkyak: (use case) yes, i want t use it to in-circuit program NXP LPC111x series MCUs. they use a surprisingly text-heavy protocol and it all goes over a UART.10:45
wpwrakit wouldn't work so well for a serial console (either end of it) since you always get small breaks where the receiver is not polling10:46
wpwrakfor a driver that doesn't have this problem, things would have to move into the kernel, with an interrupt handler signaling the reception of a start bit10:47
hellekinwpwrak: did you have a look at http://www.kosagi.com/w/index.php?title=Novena_Main_Page ?11:13
wpwrakheard of it, yes. dunno. laptops didn't strike me as a big openness problem so far.11:15
wpwraki.e., all the equipment i have does what it has to do without nasty closed source drivers or messy binary firmware. maybe there is binary firmware somewhere (though i don't think so) but it's handled transparently enough that i don't have to worry about it.11:16
wpwrakplus, a laptop is easily extended via USB, so - without any more specific requirements - there's not really a pressing need for having special built-in hardware features (such as expansion headers, fpgas, etc.)11:17
wpwrakplus, a laptop is a complex system, so this is a lot of work and it's expensive to make. not as bad as a smartphone, but substantial11:18
wpwrakah, and my laptops have of course all intel graphics. no money wasted on nvidia or ati here. if they want to keep their products a secret, then they can keep their products ;-)11:20
kristianpaulwpwrak: how old is your laptop?11:54
lindi-wpwrak: http://lindi.iki.fi/lindi/firmware/free-firmware.txt has two laptop related entries11:55
kristianpaulthats a point need more openess, well you hack the pack but..11:58
wpwrakkristianpaul: one is some 4-5 years old (a fujitsu u1010), the other maybe 2 (an asus netbook)11:58
kristianpaulwpwrak: so you believe the history of restricted boot etc?11:59
kristianpaulor drm inside the processors..12:00
kristianpauli would partially agree that laptops still fair open as the restrictions seems to be imposed somwhere else (tablets. for example)12:00
lindi-I'd very much like to have a TPM to hold the hdd encryption keys of my laptop but I have a cheap model that doesn't support it12:01
lindi-now it is quite trivial to bruteforce against my passphrase12:02
kristianpaulI do like that for portable harddrives12:02
wpwrakkristianpaul: my systems all predate restricted boot :)12:03
lindi-I also think that RMS is doing more harm than good inventing these odd alternative terms :)12:04
kristianpaulokay lets wait some more years 12:04
kristianpaullindi-: he good point :)12:04
wpwrakkristianpaul: and i don't know of any DRM in my cpus either. of course, i stick with intel there.12:05
lindi-wpwrak: I guess it depends on what you exactly count as DRM. do you include the SENTER instruction in your CPU?12:08
wpwrakhmm, don't even know what it is :)12:16
lindi-wpwrak: it's a way to build dynamic root of trust: e.g. to make it possible for a userland process to operate on encryption keys even when rest of the system is compromised12:21
lindi-wpwrak: SENTER instruction is like a normal branch except that it 1) disables other cores and interrupts 2) calculates a hash of the target code (up to 64KB) 3) tells the TPM that "we are now executing code that matches hash X and there is nothing that can interfere"12:22
lindi-wpwrak: the TPM can then be told things like "please create a new RSA key and let only blocks with hashes X, Y and Z operate using it"12:23
lindi-wpwrak: when you are done you execute the SEXIT instruction12:24
wpwraknice. but as far as i know, there's no TPM in my machines. and if there's anything like that, it's dormant. i.e., i had no troubles installing linux.12:26
lindi-yeah you need several primitives to make anything useful with SENTER12:27
lindi-wpwrak: if you want to hear more about the positive uses of these technologies I suggest reading about http://qubes-os.org/trac12:36
wpwrakyeah,even the most vile technology can have good uses :)12:42
wpwraki do wonder whether it's worth the effort to worry about things at such a low level. i mean, apart from academic interest and the tin foil hat faction.12:42
lindi-wpwrak: well I'm just interested in the technology and want to know what is possible12:43
--- Wed Dec 19 201200:00

Generated by irclog2html.py 2.9.2 by Marius Gedminas - find it at mg.pov.lt!