Show Idle (>14 d.) Chans
← 2020-12-03 | 2020-12-05 → ↓
verisimilitude: Say, care to see something interesting?
verisimilitude: I noticed this domain in mine HTTP referrers earlier, and wanted to see who was mentioning me.
verisimilitude: It was surprising, to learn someone, for some reason, decided to cover my website with Cloudflare filth so.
verisimilitude: People who don't use TLS are constantly told that people could intercept the pages and inject JavaScript. Well, this is the first time I've ever seen JavaScript injected on my website, and it was done by Cloudflare. This is ironic, but plenty of people would be too stupid to see that, and think it's helpful instead, I suppose.
verisimilitude: So, someone's making more work for me, since I need to decide whom to contact to have this removed.
adlai: /away/ empty
adlai: verisimilitude: it is impossible to prevent that kind of rehosting; a different approach is to include a brief text at a different place within each article that alert people to the possibility that they are reading a repackaging
adlai: if it is in an identical place each time, it is trivial to remove; and the ultimate 'preventative cure', in this case, is to publish privately only to your trusted friends, which defeats the entire purpose of publication.
snsabot: Logged on 2020-11-26 14:12:01 verisimilitude: I try to avoid caring about what I can't control, trinque.
qonfluent: Hello all, I've just gotten back in to x86 BIOS development from a long hiatus and would like some assistance with the Sage Smartprobe if anyone has a moment to humor me.
verisimilitude: I've no issue, adlai, when some Chinese takes mine articles to host in some fake newspaper, because I can't stop that. This is different, however, if only because it's with a US business, and as I get the impression this is some fool trying to be ``helpful''.
verisimilitude: Hello, qonfluent. I enjoy low-level programming, but that's not my particular niche, so I can't help.
qonfluent: Yeah, I figure it'll be a while before someone comes around who has one. I think this is probably the best place to find help with it, probably the highest concentration of smartprobe owners on the planet. My particular problem is that I can't write to memory. Register reads/writes and code stepping work fine, but writing to memory to test my code
qonfluent: without using up flash write cycles isn't working. The function "monitor CarOnReset,c" is supposed to allow that if I'm not mistaken, but it's been far too long.
qonfluent: What type of low-level programming do you enjoy, if you don't mind?
gregorynyssa: http://logs.nosuchlabs.com/log/asciilifeform/2020-11-02#1024128 << the computing industry in China is dominated by people who attended university in America and who harbor pro-American sentiments.
snsabot: Logged on 2020-11-02 13:32:13 asciilifeform: mats: it aint clear to me why 'chinese, national' cpu has to be made at enemy's tmsc.
gregorynyssa: the "Chinese with American degrees" form their own social stratum and subculture within PRC society.
asciilifeform: qonfluent: i'm the author of the sage guide you appear to have read. could help, if you were to stay put for a bit !
asciilifeform: qonfluent: i personally have not used the cache-as-ram mode in sage, incidentally.
asciilifeform: qonfluent: come back & plz post, in particular, what exactly the irons were that you were using. i may be able to replicate your setup.
asciilifeform: http://logs.nosuchlabs.com/log/asciilifeform/2020-12-04#1025826 << if the thing snarfs from your www blindly, rather than simply storing plagiarized text, and has obvious refer header -- you can always config your box to serve up goatse (or whatever else) to the viewers
snsabot: Logged on 2020-12-04 01:54:55 verisimilitude: I noticed this domain in mine HTTP referrers earlier, and wanted to see who was mentioning me.
thimbronion: For the record, I too have a smartprobe - but don't have anything to use it on.
asciilifeform: thimbronion: get e.g. apu1
asciilifeform: thimbronion: and hm, how the hell didja end up w/ a probe, if had no reason..
thimbronion: asciilifeform: I may try to find one - I unfortunately originally purchaced an apu2
asciilifeform: thimbronion: orig. vendor still sells'em
gregorynyssa: I was reading some of the old logs. this conversation was remarkable: http://logs.nosuchlabs.com/log/trilema/2015-06-24
gregorynyssa: maybe we can return
gregorynyssa: to Gutenberg's era in the future, by discovering the next metal after Silicon.
gregorynyssa: sorry, "silicon" should not be capitalized.
thimbronion: asciilifeform: I wanted to do some bios hacking, but it turned out I didn't have enough shiva hands.
thimbronion: asciilifeform: If you know a good home for it, I will happily donate it.
shinohai: apu2 is good for trb node, neh?
thimbronion: shinohai: good enough I imagine. Can't remember much about it at this point.
shinohai: thimbronion: if you desire to sell it outright, will mention on my telegram channel, as a couple of folks might be interested to buy (if priced in BTC).
thimbronion: shinohai: lemme dig it out of storage and we can proceed from there.
shinohai: kk thanx.
snsabot: Logged on 2020-12-04 03:22:02 qonfluent: What type of low-level programming do you enjoy, if you don't mind?
snsabot: Logged on 2020-12-04 09:00:40 asciilifeform: http://logs.nosuchlabs.com/log/asciilifeform/2020-12-04#1025826 << if the thing snarfs from your www blindly, rather than simply storing plagiarized text, and has obvious refer header -- you can always config your box to serve up goatse (or whatever else) to the viewers
snsabot: Logged on 2020-12-04 10:26:35 thimbronion: asciilifeform: I wanted to do some bios hacking, but it turned out I didn't have enough shiva hands.
asciilifeform: ( keep in mind that the boxen where probe worked, afaik all had published working builds of 'coreboot' )
asciilifeform: http://logs.nosuchlabs.com/log/asciilifeform/2020-12-04#1025861 << usable as general-purpose small x64 box. ( asciilifeform even has a node on one ) but dunwork w/ probe (see faq re why)
snsabot: Logged on 2020-12-04 10:28:03 shinohai: apu2 is good for trb node, neh?
asciilifeform: http://logs.nosuchlabs.com/log/asciilifeform/2020-12-04#1025855 << ~which~ conversation ? you linked to a date w/ almost MB of log...
snsabot: Logged on 2020-12-04 09:29:42 gregorynyssa: I was reading some of the old logs. this conversation was remarkable: http://logs.nosuchlabs.com/log/trilema/2015-06-24
asciilifeform: gregorynyssa: observe that most log links are to specific line.
asciilifeform: $ticker btc usd
btcinfobot: Current BTC price in USD: $18978.88
asciilifeform: !w poll
watchglass: Polling 16 nodes...
watchglass: 205.134.172.6:8333 : (172-6.core.ai.net) Alive: (0.022s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=659941
watchglass: 54.39.156.171:8333 : (ns562940.ip-54-39-156.net) Alive: (0.053s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=659941
watchglass: 205.134.172.4:8333 : (172-4.core.ai.net) Alive: (0.093s) V=70001 (/therealbitcoin.org:0.7.0.1/) Jumpers=0x1 (TRB-Compat.) Blocks=659941
watchglass: 205.134.172.27:8333 : Alive: (0.084s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=659941 (Operator: asciilifeform)
watchglass: 205.134.172.26:8333 : Alive: (0.082s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=659522
watchglass: 205.134.172.28:8333 : Alive: (0.084s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Return Addr=0.0.0.0:8333 Blocks=659606 (Operator: whaack)
watchglass: 208.94.240.42:8333 : Alive: (0.211s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=659941
watchglass: 143.202.160.10:8333 : Alive: (0.246s) V=70001 (/therealbitcoin.org:0.7.0.1/) Jumpers=0x1 (TRB-Compat.) Blocks=659941
watchglass: 176.9.59.199:8333 : (static.199.59.9.176.clients.your-server.de) Alive: (0.344s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=391683 (Operator: jurov)
watchglass: 192.151.158.26:8333 : Alive: (0.410s) V=70001 (/therealbitcoin.org:0.7.0.1/) Jumpers=0x1 (TRB-Compat.) Blocks=659941
watchglass: 213.109.238.156:8333 : Alive: (0.390s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=659940
watchglass: 185.85.38.54:8333 : (tlapnet-38-54.cust.tlapnet.cz) Alive: (0.631s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=659941
watchglass: 84.16.46.130:8333 : Could not connect!
watchglass: 185.163.46.29:8333 : Violated BTC Protocol: Bad header length!
watchglass: 71.114.46.209:8333 : Violated BTC Protocol: Bad header length! (Operator: asciilifeform)
watchglass: 103.36.92.112:8333 : Violated BTC Protocol: Bad header length!
asciilifeform: !w probe 71.114.46.209 8333
watchglass: 71.114.46.209:8333 : (pool-71-114-46-209.washdc.fios.verizon.net) Alive: (0.097s) V=99999 (/therealbitcoin.org:0.9.99.99/) Jumpers=0x1 (TRB-Compat.) Blocks=659940
thimbronion: http://logs.nosuchlabs.com/log/asciilifeform/2020-12-04#1025870 << I don't recall my thinking at the time (this was back in 2016), but I had put some effort into building an amd/tyan box with ecc memory, and wanted to take things a step further and install an opensource bios. Maybe I just thought it would be cool to debug the boot process...
snsabot: Logged on 2020-12-04 12:56:11 asciilifeform: http://logs.nosuchlabs.com/log/asciilifeform/2020-12-04#1025859 << curious, what didja have in mind?
asciilifeform: thimbronion: you don't need a sageprobe simply to install coreboot etc. it is strictly for gnarly dev work w/ same, i.e. when you MUST have breakpoints, single-step, etc, in bios.
thimbronion: asciilifeform: I am aware.
asciilifeform: feedbot is drinking on the job, apparently ?
snsabot: Logged on 2020-12-04 08:55:46 asciilifeform: qonfluent: i'm the author of the sage guide you appear to have read. could help, if you were to stay put for a bit !
qonfluent: I'm using the Gizmo 1 board with the probe, connected via USB to my host system
qonfluent: http://logs.nosuchlabs.com/log/asciilifeform/2020-12-04#1025866 Oh interesting, I was reading through this yesterday, you've got some really interesting ideas on here. The bit-level memory array based computer seems like a particularly interesting research topic.
snsabot: Logged on 2020-12-04 11:46:06 verisimilitude: http://logs.nosuchlabs.com/log/asciilifeform/2020-12-04#1025840 I enjoy using my machine code development tool of mine own design to write and document machine code, qonfluent.
asciilifeform: qonfluent: i assume you have already found the manual ?
asciilifeform: covers 'car mode'. again will note, i have not used this mode personally.
qonfluent: Yeah, that's what I've been using for reference for the various commands. I've also looked at the raw packets from the device, and when I do the write it seems to generate the correct responses.
asciilifeform: afaik it will work on the gizmo-1 ('g-series') chipset.
asciilifeform: qonfluent: make sure yer using gdb 7 or above if you have 'bitness' mode transitions in your process (and if you're going from 'car' to 16bit to 32 to 64... you are)
asciilifeform: this is the #1 cause of barfola when using 'sage', historically
qonfluent: I'm using version 9.2 of GDB, but all the code I'm currently running(just a cli followed by a jmp) is in real mode. I can't change the code at the moment since I don't want to change the actual firmware(I don't have a flasher, so if I break the actual BIOS I would have no way of restoring it)
asciilifeform: qonfluent: iirc you want 'CarOnReset' ~only~ if you intend to fill up the cache and then run. otherwise (if merely stepping) you want instead 'HaltOnReset'.
asciilifeform: qonfluent: buy a chinese spi eeprom flasher, the kind w/ crocodile clip. then can let go of the phobia of 'bricking'
asciilifeform: ( if you have the spare change, i rec this box : it can not only reflash, but comes w/ a hose that has a mini pcb which can solder ~instead~ of the rom, or stuff into sockets, it ~emulates~ spi eeproms and 'reflashes' via usb in <1s
asciilifeform: ( under no circumstances buy the 'em100' 'non-pro' -- there is no published driver proggy for that one afaik )
asciilifeform: theoretically this is unnecessary on amd-g boards, can use 'sage' + 'car' -- but as qonfluent has found, it is finicky
qonfluent: Yeah, I might end up doing that. I'd prefer for time reasons to be able to write the BIOS to the cache instead though, to get a quicker turn around time. The last time I started on this project, I recall being able to just write bytes to the reset vector area and the space below it to test new ideas and code
asciilifeform: qonfluent: if aint a seekrit, what are you baking that needs to run before dram init ?
asciilifeform: ( or for that matter, before exit from bios )
asciilifeform: cuz asciilifeform , when fiddling w/ x64 os , simply hbreak *0x7c00 .
asciilifeform: well, and :
asciilifeform: restore rom.bin binary 0x7c00
asciilifeform: set $rip=0x7c00
asciilifeform: ^ in a gdb script. and threw in makefile.
qonfluent: My goal is to merge the BIOS/boot loader/OS into a single object. Direct from boot, it'll scan the various devices, set up drivers, and then drop in to an interpreter/REPL
asciilifeform: and this was enuff to osdev.
asciilifeform: qonfluent: how much does this need custom replacement for ordinary 'coreboot' at the early stages ?
asciilifeform: cuz afaik it doesn't (can work on abolishing the traditional msdos-compat. part of 'coreboot' later, it doesn't much change the os init coad )
qonfluent: It probably doesn't strictly need to not be coreboot, I'd just prefer to write the initialization as well. Ideally I want to have every byte accounted for and have nothing I don't understand running on the system if that makes sense
asciilifeform: qonfluent: good luck w/ that on any known x64 chipset. most of the heft of 'coreboot' in my builds is amd's boilerplate, but w/out which machine does not quite run
asciilifeform: it may be possible to set up working irq's w/out the acpi garbage, however. but afaik 0 info publicly avail. re how.
asciilifeform: bringing up amd cpu -- or even the realtek nic -- will certainly feature a good many kB of 'magic numbers no one knows meaning of' , sadly.
asciilifeform: this is in fact largely why asciilifeform lost interest in x86 osdev.
asciilifeform: whatever you come up with, in fact will not be meaningfully opensource. (and the chance that it won't boot on recent irons from shop, is ~100% )
asciilifeform: i do not like to discourage folx, but this summarizes asciilifeform's experience with subj.
qonfluent: Yeah, I'm not interested in this as much beyond a personal hobby. I'll likely put the source up somewhere, but unless someone else has the Gizmo specifically, it won't be useful.
asciilifeform: right, but even as hobby, can end up demoralizing. qonfluent are you aware of 'movitz' ?
asciilifeform: it was just about exactly this. but boots on 0 currently-avail. irons. for the reason i described.
snsabot: Logged on 2020-12-04 17:50:49 qonfluent: My goal is to merge the BIOS/boot loader/OS into a single object. Direct from boot, it'll scan the various devices, set up drivers, and then drop in to an interpreter/REPL
qonfluent: I feel as though I've heard of it before, but it has been a while
asciilifeform: it is rather like writing in beach sand. the lack of anything resembling meaningful standardization in x64 irons, turns erry attempt at os into either a) bloat monstrosity a la linux kernel b) a curio, that readers in general cannot even boot, even with effort
asciilifeform: take 'gizmo'. afaik neither v1 nor v2 is any longer available commercially.
asciilifeform: i have both, but should they release their magic smoke, that's it
asciilifeform had been hoarding 'apu1' boxes, but still only has enuff for ~self~ 'lifetime supply', but sadly not to hand out..
qonfluent: Yeah, that's what I've run in to as well. I have an FPGA as well, but it's been even longer since I last worked in an HDL.
asciilifeform: qonfluent: sadly, very similar biz model in fpga world.
snsabot: Logged on 2020-08-20 18:09:04 asciilifeform: Aerthean: xilinx's biz model requires the periodic breaking of compat.
qonfluent: Of course, probably to an even greater extent. It's sad, but unless I can get together the cash to spin out custom silicon it's about as close as I could get. Sucks that everything has to run through shitware top to bottom. While I might not be able to remove the stink, at least I can try to make a raft haha
asciilifeform trying to make rafts since '07...
qonfluent: Yeah, reading your site has been a big inspiration for that same feeling I had, the desire to escape the layers of garbage and get something sane to work from. And of course it's where I heard about the Gizmo. I'll have to dig through some old hard drives, I'm at least half sure I have the answer I'm looking for burried on one of them from my last
qonfluent: round of this
asciilifeform: qonfluent: see also this thread.
snsabot: Logged on 2020-08-22 14:14:33 gregorynyssa: asciilifeform: suppose we don't use any FPGA. suppose you directly owned a fabrication-facility and could perform your own runs on whim. how low does the process-width have to be, for it to be worth your time?
asciilifeform: ( recurrent subj. in fact. )
qonfluent: Exactly, it's a non-starter. Even the cheapest FPGA-to-silicon route is barely in reach for most companies, and comes with its own set of limitations.
asciilifeform: qonfluent: actually, the ~only~ si i'd consider making, is a large homogeneous fpga . but afaik it is ~impossible to actually finance such a project , unless yer already rockefeller.
← 2020-12-03 | 2020-12-05 → ↑
Random(asciilifeform) | Download hourly DB snapshot | Get Source Code