2012-06-17 00:32 http://arstechnica.com/information-technology/2012/06/amd-to-add-arm-processors-to-boost-chip-security/ 2012-06-17 00:32 one of the rare times I see a situation and see a nanonote as an immediately useful solution 2012-06-17 00:36 Ayla has quit [Quit: Lost terminal] 2012-06-17 00:36 you want to add a nanonote to the cpu ? 2012-06-17 00:36 naw 2012-06-17 00:37 Ayla has joined #qi-hardware 2012-06-17 00:37 just recalled a new Cory Doctorow's book 2012-06-17 00:37 they used a small and presumably somewhat trusted computer to do a key signing party 2012-06-17 00:38 (they) the book is 1984, but in 2014. you got the idea. I'm too sleepy now to explain it properly anyway. 2012-06-17 00:39 (small computer) ah, i see. yes, the ben is quite good at being small ;-) 2012-06-17 00:39 while I'm not absolutely sure that NN's CPU does not contain backdoors (I haven't examined the actual silicon), that would have a really really low probability 2012-06-17 00:39 and it definitely does not have backdoors as features 2012-06-17 00:40 the next-NN with M1 SoC should fix this problem 2012-06-17 00:40 CPU backdoors would also be kinda tricky 2012-06-17 00:40 yes 2012-06-17 00:40 fpgas could still have "backdoors" ;-) 2012-06-17 00:40 besides that, NN does not have any wireless ifaces 2012-06-17 00:40 which makes it quite perfectly safe for the task 2012-06-17 00:40 mine do, sometimes :) 2012-06-17 00:41 well, a cpu backdoor would require extensible software cooperation 2012-06-17 00:41 so I don't think it's even a theoretically realistic scenario 2012-06-17 00:41 Ayla has quit [Ping timeout: 245 seconds] 2012-06-17 00:41 after all, if you're _that_ paranoid, you can do RSA with a calculator 2012-06-17 00:42 or a pen and a piece of paper 2012-06-17 00:42 definitely no backdoors there. 2012-06-17 00:42 someone whispering numbers to make you miscalculate 2012-06-17 00:42 (fpgas) yeah, I won't blindly trust a fab to not alter the netlist. but I would evaluate the complexity of such a task, and again, it is not realistic at all 2012-06-17 00:43 not significantly more realistic than (/me drops a pen on the floor) that pen tunneling through the floor. 2012-06-17 00:43 Ayla has joined #qi-hardware 2012-06-17 00:43 it's possible according to quantum mechanics, just not very probable :D 2012-06-17 00:43 besides which, netlists were OCR'd with a microscope and a polarizer, and if you have the sources 2012-06-17 00:44 (see: visual6502.org) 2012-06-17 00:44 it's all verifiable, through hard 2012-06-17 00:44 but military supposedly does that verification, and so could any interested party 2012-06-17 00:45 wpwrak: are you perfectly sure that your C compiler doesn't have a, say, self-reproducible `login' backdoor which it inserts in `login' and `cc' itself, but which does not appear in the sources? 2012-06-17 00:45 gcc is built with gcc for ages 2012-06-17 00:45 have it always had digital signatures? were they enforced? 2012-06-17 00:46 * kristianpaul remenber gnu's ftp was compromised time ago 2012-06-17 00:47 I often think that when working with security, you just need to go to the extreme. not to do something practical, but just to evaluate the risk 2012-06-17 00:47 what once seemed impossible is easy now 2012-06-17 00:47 e.g. trusted computing 2012-06-17 00:47 people would laugh on this idea in '85 2012-06-17 00:48 (or should I write "trusted" each time?..) 2012-06-17 00:48 [that is, quoted.] 2012-06-17 00:48 my paranoia knows limits :) 2012-06-17 00:48 wpwrak: would you consider someone placing a supplementary core as a backdoor in your CPU 20 years ago? 2012-06-17 00:49 or, for current CPUs, there is SMM 2012-06-17 00:49 which is technically an undetectable backdoor 2012-06-17 00:51 it's not paranoia and I'm not going to throw away my Galaxy SII because it has a trusted computing module in the CPU (maybe because of the blobs; they're so badly written that they can be broken not because of malice but simply of stupidity) 2012-06-17 00:51 so what it is? 2012-06-17 00:51 an evaluation of possibilites 2012-06-17 00:52 modern CPUs are incredibly complex (they're more like SoCs already, even in PCs), and they get more complex each year 2012-06-17 00:52 it's anything but hard to hide a backdoor in a device like this 2012-06-17 00:52 yup 2012-06-17 00:53 for example, anyone familiar with FPGAs could confirm that it's easy to make a special long command stream that 2012-06-17 00:53 will instantly throw you to ring0 2012-06-17 00:53 or just instantly smm 2012-06-17 00:53 enough, you're pwned. 2012-06-17 00:53 you can even hide it in microcode if you want 2012-06-17 00:54 besides which, AMD openly states that they put a backdoor in the CPU. 2012-06-17 00:54 what else do you fucking need?! 2012-06-17 00:57 ... and silence was an answer to him :) 2012-06-17 00:57 whatever, I'm gonna get some sleep. 2012-06-17 00:57 5AM here 2012-06-17 00:57 I was about to said that (sleep) :_) 2012-06-17 01:36 Textmode has quit [Ping timeout: 255 seconds] 2012-06-17 01:47 wej has quit [Ping timeout: 248 seconds] 2012-06-17 01:50 Textmode has joined #qi-hardware 2012-06-17 01:52 wej has joined #qi-hardware 2012-06-17 02:31 wej has quit [Ping timeout: 272 seconds] 2012-06-17 02:36 wej has joined #qi-hardware 2012-06-17 02:41 Textmode has quit [Ping timeout: 255 seconds] 2012-06-17 02:47 Ayla has quit [Quit: dodo] 2012-06-17 02:47 rz2k has quit [Ping timeout: 245 seconds] 2012-06-17 02:52 rzk has joined #qi-hardware 2012-06-17 02:55 Textmode has joined #qi-hardware 2012-06-17 03:04 dandon has joined #qi-hardware 2012-06-17 03:28 Textmode has quit [Ping timeout: 255 seconds] 2012-06-17 03:57 my first practical goal for fpgatools will be an inverter :-) 2012-06-17 03:57 I think that's the shortest path to something that runs 2012-06-17 03:58 so I will take the smallest spartan-6 (xc6slx4), buy a bunch of them (like 10-20), then make the most minimal board possible to just power the chip and expose jtag 2012-06-17 03:59 then use fpgatools to program the chip and get the inverter to work 2012-06-17 03:59 oh I guess I need 2 pads for the inverter itself as well 2012-06-17 03:59 Textmode has joined #qi-hardware 2012-06-17 04:00 inside the chip, the inverter will start at one pad, pass through an I/OLOGIC and some switch boxes before coming out negated on the other pad 2012-06-17 04:00 that's the plan :-) 2012-06-17 04:04 you can develop this on milkymist ... 2012-06-17 04:14 Textmode has quit [Ping timeout: 255 seconds] 2012-06-17 04:21 * DocScrutinizer05 reads acklog and smiles 2012-06-17 04:21 back* 2012-06-17 04:27 Textmode has joined #qi-hardware 2012-06-17 04:45 compcube has quit [Quit: Leaving] 2012-06-17 04:48 re 2012-06-17 04:52 Textmode has quit [Ping timeout: 256 seconds] 2012-06-17 05:08 whitequark: got a link to the AMD CPU backdoor thing? 2012-06-17 05:11 jekhor has joined #qi-hardware 2012-06-17 05:12 emeb has joined #qi-hardware 2012-06-17 05:16 Textmode has joined #qi-hardware 2012-06-17 05:56 emeb has quit [Ping timeout: 246 seconds] 2012-06-17 06:11 emeb has joined #qi-hardware 2012-06-17 06:12 jekhor has quit [Ping timeout: 246 seconds] 2012-06-17 06:20 emeb has left #qi-hardware [#qi-hardware] 2012-06-17 06:39 Textmode has quit [Ping timeout: 255 seconds] 2012-06-17 06:39 wpwrak: no that would miss the point, right now 2012-06-17 06:39 I am working in the bitstream, and already dealing with 2+ million bits 2012-06-17 06:39 I don't want to deal with 10 million instead 2012-06-17 06:39 plus a slx4-based inverter allows me to go through the pcb-making as well 2012-06-17 06:40 but I'm coming at it from the side of the fpgatools, where I do need small incremental steps to work through the many many different 'things' that are on any spartan-6 chip, even the smallest 2012-06-17 06:40 of course after the inverter I will do a few more gates, include slices, more switch boxes, and eventually work my way up to larger chips like a slx45 2012-06-17 06:40 but one by one, that's the point: inverter on slx4 :-) 2012-06-17 06:41 the inverter doesn't even need a slice because it uses an invert mux right in the io block 2012-06-17 06:41 good for me 2012-06-17 07:17 larsc has joined #qi-hardware 2012-06-17 08:18 DocScrutinizer: so Nokia is joining the mobile patent wars, nice 2012-06-17 08:48 wpwrak: this wasn't a good time? huh? 2012-06-17 08:52 lekernel: they were working on other changes that would have interfered 2012-06-17 08:53 wolfspraul: hmm, so the difference between the FPGA in M1 and the FPGA you want to play with would merely be size but not structural variety ? in that case, can't you just ignore the extra 8 Mbits, just like you'll already ignore some ~2 Mbits your inverter doesn't need ? 2012-06-17 08:54 wolfspraul: (pcb-making) you should really pick something simpler to get started. don't worry, even simple things can get quite hard when you do them for the first time :) 2012-06-17 08:56 lekernel_ has joined #qi-hardware 2012-06-17 08:56 lekernel has quit [Ping timeout: 246 seconds] 2012-06-17 08:57 lekernel_ is now known as lekernel 2012-06-17 09:01 kyak has quit [] 2012-06-17 09:15 DocScrutinizer05: yeah, I remember our discussion about paranoia, yes :) it'd be wrong to say that it didn't make me think 2012-06-17 09:15 pabs3: http://arstechnica.com/information-technology/2012/06/amd-to-add-arm-processors-to-boost-chip-security/ 2012-06-17 09:33 phirsch has quit [Ping timeout: 245 seconds] 2012-06-17 09:34 DocScrutinizer05: just found an interesting feature in SGS2 BP 2012-06-17 09:34 if you'll send 0xDEADDEAD to the BP bootloader instead of data length, then, instead of writing, it'll read entire modem RAM and send it to host 2012-06-17 09:34 yes, it can write too, and reboot 2012-06-17 09:35 and it doesn't exactly verify any signatures 2012-06-17 09:35 well, theoretically it does, but looks like someone fucked it up 2012-06-17 09:35 whitequark: djbclark gave them to me 2012-06-17 09:35 looks like I'm lucky on baseband modules with interesting features 2012-06-17 09:38 interesting discovery 2012-06-17 09:39 SGS2 looks like that everywhere 2012-06-17 09:39 it's more of devboard than a phone: there are *some* locks, but they're not very difficult to bypass 2012-06-17 09:40 ... and I tried asking a Korean dev about a feature of a PMIC which was under NDA 2012-06-17 09:40 and got a helpful reply 2012-06-17 09:41 maybe osmocombb folks would be interested? 2012-06-17 09:41 the BP is based on xgold262 2012-06-17 09:42 er, 626 2012-06-17 09:43 phirsch has joined #qi-hardware 2012-06-17 09:44 just build your own fucking BP 2012-06-17 09:44 it's really frustrating to see how many hours are spent reverse engineering proprietary stuff that becomes obsolete in 1 yr 2012-06-17 09:45 what is a bp? 2012-06-17 09:45 baseband processor 2012-06-17 09:46 cpu in mobile phones which talks to the GSM network 2012-06-17 09:50 ah ok 2012-06-17 09:54 *hmm*. i need a C function that returns a pointer to a function just like itself. do i need more sleep/caffeine or is void * really the best i can do in terms of type safety ? 2012-06-17 10:03 hmmm 2012-06-17 10:03 interesting 2012-06-17 10:04 nice, isn't it ? an infinitely recursive type declaration :) 2012-06-17 10:04 yeah, C isn't haskell 2012-06-17 10:05 so, only void* 2012-06-17 10:06 and of course, typedef won't accept the same name at two places 2012-06-17 10:07 rejon_ has quit [Remote host closed the connection] 2012-06-17 10:19 rejon has joined #qi-hardware 2012-06-17 10:20 yeah 2012-06-17 10:20 that's why I said it's not haskell 2012-06-17 10:30 kyak has joined #qi-hardware 2012-06-17 10:30 kyak has quit [Changing host] 2012-06-17 10:30 kyak has joined #qi-hardware 2012-06-17 10:32 Aylax has joined #qi-hardware 2012-06-17 10:53 wpwrak: http://c-faq.com/decl/recurfuncp.html 2012-06-17 10:54 the struct trick looks good to me 2012-06-17 10:55 if you can afford a struct. :) 2012-06-17 10:56 oh, that's pretty nice. thanks ! 2012-06-17 11:43 phirsch has quit [Excess Flood] 2012-06-17 11:46 phirsch has joined #qi-hardware 2012-06-17 12:04 Aylax has quit [Ping timeout: 240 seconds] 2012-06-17 12:16 Aylax has joined #qi-hardware 2012-06-17 12:44 What do you think, about GPL.... 2012-06-17 12:44 There is a GPL software... 2012-06-17 12:44 but I can download it only if I identify myself. 2012-06-17 12:45 DocScrutinizer has quit [Disconnected by services] 2012-06-17 12:45 DocScrutinizer has joined #qi-hardware 2012-06-17 12:45 Does GPL say anything about the right to (more or less) private access to the sources? 2012-06-17 12:46 DocScrutinizer06 has joined #qi-hardware 2012-06-17 12:46 - A government website lets me download a GPL criptography software they developed, but only if I identify myself and explain why I download that. 2012-06-17 12:46 DocScrutinizer05 has quit [Ping timeout: 246 seconds] 2012-06-17 12:48 viric: wpwrak: this again convinced me I don't know shit about c and better resort to beating up grannies on the street and robbing their handbags 2012-06-17 12:48 :) 2012-06-17 12:49 or simply refuse coding in anything but assembler 2012-06-17 12:57 DocScrutinizer06 is now known as DocScrutinizer05 2012-06-17 12:59 viric: opencores does a similar thing :) 2012-06-17 13:00 ahh 2012-06-17 13:03 Ayla has joined #qi-hardware 2012-06-17 13:03 Aylax has quit [Quit: Bye] 2012-06-17 13:08 Ayla has quit [Ping timeout: 252 seconds] 2012-06-17 13:09 Ayla has joined #qi-hardware 2012-06-17 13:11 Ayla has quit [Client Quit] 2012-06-17 13:13 viric: on the first access, they can make you jump as many hoops they want. make you agree to deliver a million euro, your soul, your firstborn, droit du seigneur, whatever. 2012-06-17 13:14 viric: redoi 2012-06-17 13:14 sorry 2012-06-17 13:14 viric: but then you're FREE to spread as many copies as you want, to whomever you want :) 2012-06-17 13:14 viric: redistribution is not mandatory at least you use that software i understand 2012-06-17 13:15 DocScrutinizer: C is an amazing language. it grows with you. the better your skills get, the more you understand its perfection. 2012-06-17 13:15 yup, (grows part) 2012-06-17 13:15 the better you understand C the more you wish it never got invented ;-) 2012-06-17 13:16 DocScrutinizer: you prefer junk languages like C++ ? :) 2012-06-17 13:16 eeeeeeeeeeeeEEEEEEEEEEEeeeeeeeeeeeeeek 2012-06-17 13:16 I'm one of those Wirth softies ;-P 2012-06-17 13:17 aha, a quiche-eater :) 2012-06-17 13:17 indeed 2012-06-17 13:17 luckily, i was able to shake that bad habit a good while ago 2012-06-17 13:18 strict typechecks - heaven! 2012-06-17 13:18 kristianpaul: I've to use that software to send my taxes report 2012-06-17 13:18 C has strict type checking 2012-06-17 13:18 BWAHAHAHA 2012-06-17 13:18 Yes, I like the C strict type checking 2012-06-17 13:18 C? perfection? lol. 2012-06-17 13:19 typedef is not defining a new type, though. 2012-06-17 13:19 some of it in the form of warnings, but you get your diagnostic when you need it 2012-06-17 13:19 viric: and u already go it as object form? 2012-06-17 13:19 bool foo() {...; return 345} 2012-06-17 13:19 kristianpaul: sure. a java applet. 2012-06-17 13:19 kristianpaul: I even run it. But it fails for me :) 2012-06-17 13:19 hmm 2012-06-17 13:19 we had that less than one week ago, in some popular shite I can't recall 2012-06-17 13:19 especially for e.g. string manipulations, which keeps infosec people fed 2012-06-17 13:20 lekernel: my most illuminating experience with C was when i wrote a language that incorporated much of C (for a scriptable debugger). there, i learned to appreciate many of the more subtle points of that language 2012-06-17 13:20 talking of string manipulations, I recently discovered stpcpy 2012-06-17 13:20 now I recall: it was that funny vulnerability of MySQL 2012-06-17 13:20 i fed with binary related operations, usually perl is more ready in such cases 2012-06-17 13:20 there's also strfry() 2012-06-17 13:21 that made every one out of 256 auth tries succeed no matter which credentials 2012-06-17 13:22 and you tell me there's something like strict typechecks in C 2012-06-17 13:22 the C library is sucky as well. how do you move/rename a folder? system("mv ...") 2012-06-17 13:23 is not that more a posix problem (folder) ? 2012-06-17 13:23 and of course what you give to system() has to come from the aforementioned string manipulation mess 2012-06-17 13:23 how does 'mv' move the folder? 2012-06-17 13:23 lekernel: rename() will do just fine for renaming 2012-06-17 13:24 wpwrak: no, if you cross mount points it doesn't work. 2012-06-17 13:24 lekernel: that's not a rename :) 2012-06-17 13:25 DocScrutinizer: booleans are a perversion you quiche-eaters added. you deserve the consequences :) 2012-06-17 13:26 lekernel: well, some OS have different operations for move and rename 2012-06-17 13:26 for example. 2012-06-17 13:26 well, a particularly nice type we quiche-eaters added are sets, which are defined as bit fields internally 2012-06-17 13:27 basically a uint1 2012-06-17 13:27 rename() renames a file, moving it between directories if required. 2012-06-17 13:27 I didn't know it. 2012-06-17 13:27 C89. 2012-06-17 13:28 oh and there's also memcpy vs. memmove 2012-06-17 13:28 of course 2012-06-17 13:29 that allow operations of different performance 2012-06-17 13:29 now if you think that sort of thing doesn't waste developer time, look at this: https://bugzilla.redhat.com/show_bug.cgi?id=638477 2012-06-17 13:30 well, lemme guess. memmove remapping if possible? 2012-06-17 13:30 viric: wrong. the test for the direction of copy at the beginning of memmove() is unnoticeable. 2012-06-17 13:30 ah you mean it could check if it overlaps or not? 2012-06-17 13:31 it could simply tweak the mmu table 2012-06-17 13:31 DocScrutinizer05: no it's much simpler than that. copying memory. only if you use memcpy() you must make sure your two regions don't overlap. 2012-06-17 13:31 hm I think you miss what's memmove about :) 2012-06-17 13:31 which e.g. NeXTStep did per definitionem for inter-process messages 2012-06-17 13:32 language flamewars, always so much fun 2012-06-17 13:32 as if, your brain was available to pollute it with such details 2012-06-17 13:32 larsc: :) 2012-06-17 13:32 larsc: hey sorry. what was your comment about 'ast generator' about the other day? 2012-06-17 13:32 Let's see if C11 addresses any of that ;) 2012-06-17 13:32 I didn't get it 2012-06-17 13:33 (that was in #milkymist, about migen) 2012-06-17 13:33 my question comes down to "what is the difference between an 'ast generator' and a 'proper language'?" 2012-06-17 13:35 lekernel: well, seems a straightforward pretty natural approach. I know if my two pointers point to unrelated objects or if I just want to shift a mem-area by a few bytes "in place" 2012-06-17 13:35 DocScrutinizer05: but that could be checked 2012-06-17 13:36 what for? 2012-06-17 13:36 AIUI it *gets* checked, in memmove 2012-06-17 13:37 only have one function: memcpy 2012-06-17 13:37 I mean, it's not a mandatory prerequisite that mem overlaps to use memmove on it 2012-06-17 13:37 if the regions overlap, run memmove. If not, run memcpy. 2012-06-17 13:37 wolfspraul: well with a ast generator you basically write in another language, a metalanguage if you want to say so. which can be very powerful. 2012-06-17 13:38 Ayla has joined #qi-hardware 2012-06-17 13:38 I meant the opposite. if overlap, memcpy. If not, memmove. 2012-06-17 13:38 if you however know your areas never can overlap (a very usual case), why use a function with a useless check? 2012-06-17 13:38 because memmove can be implemented faster. 2012-06-17 13:38 amh 2012-06-17 13:38 no 2012-06-17 13:38 the opposite. 2012-06-17 13:38 ;) 2012-06-17 13:39 viric: really you can gain 10 nanoseconds or so on a modern machine. is it really worth all the wasted hours on that bugzilla report? no. 2012-06-17 13:39 lekernel: I agree I agree 2012-06-17 13:41 DocScrutinizer05: because developers make mistakes. 2012-06-17 13:41 lekernel: 10ns??? >> The memory areas may overlap: copying takes place as though the bytes in src are first copied into a temporary array that does not overlap src or dest, and the bytes are then copied from the temporary array to dest. 2012-06-17 13:42 seems to include a malloc 2012-06-17 13:42 larsc has quit [Ping timeout: 244 seconds] 2012-06-17 13:42 yes, 10ns. in all implementations I've seen it's only about the direction of copy. 2012-06-17 13:42 no, there's no malloc 2012-06-17 13:43 I'd expect a tinmy bit longer than 10ns for this to complete 2012-06-17 13:43 just a test at the beginning to determine the copy direction 2012-06-17 13:44 now if that's true then this manpage is pretty buggy and fuckedup 2012-06-17 13:44 the subtlety is in the "as though" ... 2012-06-17 13:44 as the result of the operation may differ vastly in certain situations 2012-06-17 13:44 e.g. memove IO memmapped area 2012-06-17 13:45 there a complete mem mapping swap for each switch between read and write a byte or word could take pretty looooong 2012-06-17 13:46 I don't think you're supposed to use the regular libc memmove on weird memory-mapped I/O... it doesn't even guarantee the size/alignment of accesses 2012-06-17 13:46 hehe, true 2012-06-17 13:46 but manpages are supposed to be accurate 2012-06-17 13:47 larsc (in absence) ok, thanks. I still don't get it :-) 2012-06-17 13:47 but that's ok 2012-06-17 13:48 "language" for me is a set of rules, grammar, syntax, vocabulary/keywords, etc. 2012-06-17 13:48 it's used to express something, that someone (or some program) can understand, or interpret in some way 2012-06-17 13:48 wolfspraul: if you write vhdl or verilog, you have a single language that expresses the logic more or less directly 2012-06-17 13:48 I'm just not familiar with the term 'ast generator' 2012-06-17 13:49 lekernel: btw the rationale of "because developers make mistakes" clearly suggests to abandon C all together and rather use a proper lang like pascal or modula, which comes with runtime checks for all those more commonly done mistakes, like array-index out of bounds, etc 2012-06-17 13:49 pascal? no again ;-) 2012-06-17 13:50 with the migen "ast generator", you are writing python (the "metalanguage") that manipulates fragments of verilog-style logic 2012-06-17 13:51 maybe I should look at migen more and then I would get it 2012-06-17 13:51 wolfspraul: does it make sense? 2012-06-17 13:51 no, doesn't 2012-06-17 13:51 but no problem, don't worry 2012-06-17 13:51 interstingly googling for 'ast generator' also yields very little 2012-06-17 13:53 DocScrutinizer05: ...and as you can see I'm writing python and moving milkymist software to lua those days :) 2012-06-17 13:55 wolfspraul: verilog or vhdl is lekernels' assembler, and let's say he's writing a compiler in python, for some his language to that assembler. 2012-06-17 13:55 or a python lib that emits that assembler (vhdl/verilog). something like this? 2012-06-17 13:55 assembler? 2012-06-17 13:55 thats the net list 2012-06-17 13:55 metaphoric 2012-06-17 13:55 :-) 2012-06-17 13:55 analogy to software 2012-06-17 13:55 hmm 2012-06-17 13:55 this is why I love 'ast generator' 2012-06-17 13:56 I think this is not a widely understood term at all, given how hard it is to even google for a definition 2012-06-17 13:56 So an ast generator is not cooking chicken? 2012-06-17 13:56 lekernel: btw differences between the supposed operation of memmove and the way it's probably done in real world also may arise easily in context of multitasking 2012-06-17 13:56 I did find a 82 page 2008 paper from a german university 2012-06-17 13:56 I don't need an 'ast generator' 2012-06-17 13:56 :-) 2012-06-17 13:57 DocScrutinizer05: totally. but the C language has the concept of a traditional single-core CPU built in (through pointers), so it's not appropriate here. 2012-06-17 13:57 when a concurrent process looks at first byte of dest to determine if origin is already free to rewrite it 2012-06-17 13:58 sure you usually solve those issues with mutex etc 2012-06-17 13:58 Ayla has quit [Quit: brb] 2012-06-17 13:58 or define memmove section to be atomic 2012-06-17 14:00 Ayla has joined #qi-hardware 2012-06-17 14:03 ~wiki ast generator 2012-06-17 14:03 I couldn't find a matching article in wikipedia, look for yerselves: http://en.wikipedia.org/wiki/Special:Search?search=ast+generator&go=Go 2012-06-17 14:03 thought as much 2012-06-17 14:05 http://en.wikipedia.org/wiki/Abstract_syntax_tree ?? 2012-06-17 14:08 or http://en.wikipedia.org/wiki/Andrew_S._Tanenbaum ? ;-P 2012-06-17 14:10 1st one ;) 2012-06-17 14:12 if it's about Abstract Syntax Tree then I wonder why anybody would need a generator for that, since my approach always been I think of code in AST in my head, and then convert it to any arbitrary lang and syntax. That's what I always called "I don't mind which languge to use - I can program" - until somebody pointed me to http://c-faq.com/decl/recurfuncp.html :-S 2012-06-17 14:13 .s/I can program/ I know to design programs/ 2012-06-17 14:14 ast generator is an excuse in the absence of floss synthesizers that could allow those language extensions for our topic i think 2012-06-17 14:17 mh I think it'd need a book covering state machines clear design and implementation (for software, in C for example). 2012-06-17 14:18 C conding style? :-) 2012-06-17 14:18 because it's one of those things I start with a big function and some 'if/else' clauses.. then a switch... and soon some mess with out of band information... 2012-06-17 14:18 C and C++ approaches to state machines would be nice 2012-06-17 14:19 well, you can imagine that every program is some state machine, but with the state spread in multiple variables :) 2012-06-17 14:31 btw http://c-faq.com/ptrs/funccall.html kinda reassured me I understand a tiny bit of C at least 2012-06-17 14:31 on ASTs 2012-06-17 14:32 As for function pointers... I like to declare function types, and instead of pointers to functions. 2012-06-17 14:32 typedef int fptr(char x); 2012-06-17 14:32 then I use: 2012-06-17 14:32 larsc has joined #qi-hardware 2012-06-17 14:32 fptr *x = myfunction; x('z'); 2012-06-17 14:33 even in this context, I assume, AST is not some free-form structure resembling a free-form algorithm. If you have verilog, you can resemble it with AST form, which allows you to manipulate the code very conveniently (compared to e.g. string functions or regexen) 2012-06-17 14:33 basically AST is an incredibly simple storage structure 2012-06-17 14:33 viric: that's sth I feel familiar with 2012-06-17 14:33 nested s-exps resemble an AST 2012-06-17 14:33 I hate the "typedef int (*fptr)(char x)" kind of parenthesis. :) 2012-06-17 14:33 but these later looks much more spread. 2012-06-17 14:33 C is not a programming language 2012-06-17 14:34 DocScrutinizer05: I disagree with many things that Paul Graham writes, but I think there are some good ideas in this one http://www.paulgraham.com/hundred.html 2012-06-17 14:34 it's a PDP-11 assembler which thinks it's a compiler 2012-06-17 14:34 Using function typedefs, instead of pointers-to-function, allows using the typedef for the prototypes. 2012-06-17 14:35 Therefore throwing a bad declaration. 2012-06-17 14:35 in case of bad types. 2012-06-17 14:36 as here: http://sprunge.us/LNaW 2012-06-17 14:37 the line 3 can't be typed if the typedef were of function pointer. 2012-06-17 14:38 maybe there is a good reason why most people use typedef of function pointers, but I don't know it still. 2012-06-17 14:42 wolfspraul: for a language you have grammar, syntax, etc and a parser which will take care of generating a AST from code written in that language. In migen you don't have that, but rather construct the AST by hand 2012-06-17 14:45 well, you'll have it later for special cases. but keeping the low-level stuff accessible is good - we need it for many things... 2012-06-17 14:51 but migen intentions is not been a language is it? i understand as the result of frustating of generating SoC by hand, now implmented in a "friendly" scripting language no? 2012-06-17 14:52 s/frustating/frustation 2012-06-17 14:52 kristianpaul meant: "but migen intentions is not been a language is it? i understand as the result of frustation of generating SoC by hand, now implmented in a "friendly" scripting language no?" 2012-06-17 14:52 migen is a "toolbox" for generating large synchronous systems. anything that makes SoC design great can go into it. 2012-06-17 14:53 oh, so is not milkymist centric at all 2012-06-17 14:54 no. the milkymist specific stuff is in the milkymist-ng repository. 2012-06-17 14:54 hmm 2012-06-17 14:56 here's another project using migen: https://github.com/brandonhamilton/rhino-tools 2012-06-17 14:59 oh, migen include dsp ASTs as well? 2012-06-17 15:00 lekernel: well, I agree partially on http://www.paulgraham.com/hundred.html POV 2012-06-17 15:00 if your question is whether migen flow can be used for DSP, then yes 2012-06-17 15:00 that's why I dislike C 2012-06-17 15:00 the "AST" (FHDL) is just like verilog... and can do pretty much everything verilog does, as long as you have a single clock 2012-06-17 15:01 C is all about premature optimization and concept tainted by hw-related axioms/paradigms 2012-06-17 15:01 DocScrutinizer05: and therefore you do assembler instead? ;) 2012-06-17 15:02 actually assembler is more honest in that regard 2012-06-17 15:03 DocScrutinizer05: you code in forth as well? 2012-06-17 15:03 I used to 2012-06-17 15:03 in the early 80s 2012-06-17 15:04 pretty nice lang, after a week to get accustomed to it 2012-06-17 15:06 'rhino'... I had some bad time with java rhino. 2012-06-17 15:06 we used to program a Z80 based realime video manipulation hardware in a lang 99% overlapping with forth 2012-06-17 15:07 the whole forth code had to get interpreted 25 times per second, on a fixed pace (today you call that realtime) 2012-06-17 15:09 basically the "mainloop" had the central wait on VSYNC 2012-06-17 15:13 lekernel: regarding this proposition in Graham about wasting cycles for simplicity - I had a similar discussion just some 12h ago: http://mg.pov.lt/maemo-irclog/%23maemo.2012-06-17.log.html#t2012-06-17T03:03:22 http://mg.pov.lt/maemo-irclog/%23maemo.2012-06-17.log.html#t2012-06-17T03:16:31 http://mg.pov.lt/maemo-irclog/%23maemo.2012-06-17.log.html#t2012-06-17T03:24:02 2012-06-17 15:38 a fellow hacker, who reverse engineers SGS2 RIL (modem driver), just found a backdoor 2012-06-17 15:38 if a specially-crafted incoming CSD call arrives, it passes a root shell to it 2012-06-17 15:38 why am I not surprised 2012-06-17 15:50 whitequark: that's inside the BP? 2012-06-17 15:51 SGS = samsung galaxy something? 2012-06-17 15:52 kyak has quit [] 2012-06-17 15:52 viric: yes 2012-06-17 15:52 lindi-: no 2012-06-17 15:53 and what is CSD? 2012-06-17 15:53 circuit switched data? 2012-06-17 15:53 whitequark: is that part of some android phones? 2012-06-17 15:53 RIL is a service which translates AT commands from dialer and stuff to the modem IPC (thankfully, it runs in its own small compartment and cannot e.g. influence RAM of the AP) 2012-06-17 15:53 yeah, CSD is circuit switched data 2012-06-17 15:53 RIL is generally a part of any Android phone, but different vendors provide different RILs 2012-06-17 15:53 there is a reference, FOSS one 2012-06-17 15:53 and Samsung ships this one with a backdoor in their phones 2012-06-17 15:54 quite an improvement over the foss. 2012-06-17 15:54 well, that's why we are writing a FOSS RIL 2012-06-17 15:54 whitequark: interesting, do you need operator help to initiate such a CSD call? 2012-06-17 15:54 lindi-: in Russia, I need to sign a contract (_very_ expensive) to be able to receive CSD calls at all 2012-06-17 15:55 obviously the operator can still initiate it at their will 2012-06-17 15:55 what's a csd call about? 2012-06-17 15:55 but no, other users can't do that 2012-06-17 15:55 whitequark: yeah but could I initiate such a call? 2012-06-17 15:55 viric: dialup through gsm 2012-06-17 15:55 lindi-: no, not while I'm in Russia with my current operator 2012-06-17 15:55 but I'm pretty sure it is perfectly possible in other countries/operators 2012-06-17 15:55 whitequark: yeah but I'm not in Russia 2012-06-17 15:56 lindi-: doesn't matter, incoming CSD calls are barred by my operator 2012-06-17 15:56 what's different between dialup through gsm, and usual voice calls? 2012-06-17 15:56 viric: just like the difference between voice and data over plain old telephony 2012-06-17 15:56 voice call is voice, and CSD is an analog modem 2012-06-17 15:57 well, it's not exactly this way, but pretty close 2012-06-17 15:57 an analog modem over gsm? 2012-06-17 15:57 kinda 2012-06-17 15:57 it's about making something analog work over something digital? 2012-06-17 15:57 who may want to use that? 2012-06-17 15:57 I'm not very familiar with this technology, it was already dead when I got my first cellphone 2012-06-17 15:57 well 2012-06-17 15:57 you didn't have EDGE and GPRS back then 2012-06-17 15:57 and you got to use CSD 2012-06-17 15:57 whitequark: (backdoor) DUH! 2012-06-17 15:58 ah is that about the 9600 bps internet connection? 2012-06-17 15:58 viric: 9600 is fast. 2012-06-17 15:58 DocScrutinizer05: it's not counting all other ways you could get control 2012-06-17 15:58 before gprs, I had 9600baud in my mobile phone, with a WAP browser 2012-06-17 15:58 DocScrutinizer05: e.g. it sprintf()s a filename into a string and then system()s 2012-06-17 15:58 viric: thats csd. 2012-06-17 15:58 Ah ok 2012-06-17 15:59 I remember its cost was calculated per minutes, not per amount of data transferred :) 2012-06-17 15:59 csd is the '9k6 data' service in 2g (gsm) 2012-06-17 15:59 viric: yeah, bloody expensive 2012-06-17 15:59 everything else came later. 2012-06-17 15:59 ah, perfect. 2012-06-17 16:00 hm that's why gprs mobile phones say allow to choose: "a) Only use GPRS b) fallback to GSM in case of lack of GPRS" 2012-06-17 16:00 everything else has much higher latency also. csd is/was much better than mobile ip now. 2012-06-17 16:00 I used it very rarely. 2012-06-17 16:01 DocScrutinizer05: (the system() is from Android side and not modem), but it's as solid as swiss cheese 2012-06-17 16:01 SGS2 is a funny machine, it does almost nothing to prevent your tinkering with it 2012-06-17 16:01 blobs are not obfuscated nor even optimized 2012-06-17 16:01 i.e. a simple IDA run gives away all the details 2012-06-17 16:01 they didn't even strip them. 2012-06-17 16:02 viric: it's a service tag on the data connection via GSM. There are tags for voice, data, fax 2012-06-17 16:02 just like on ISDN 2012-06-17 16:02 i.e. you have all symbols AND DEBUG INFO. 2012-06-17 16:02 ok 2012-06-17 16:02 as I already said, modem bootloader allows you to read/write RAM and execute arbitrary code on the BP 2012-06-17 16:02 whitequark: isn't it tricky about the linking, addresses, ...? 2012-06-17 16:02 whitequark: why do you wonder? 2012-06-17 16:03 roh: about what? 2012-06-17 16:03 about debug symbols etc. 2012-06-17 16:03 viric: nope, RIL is just a linux .so 2012-06-17 16:03