<qu1j0t3> i've never been able to transfer a domain out of Quicknames. afaik the facility doesn't even exist
<lain> qu1j0t3: that's against ICANN terms, and they can be shut down if that is actually the case
<lain> though they are allowed to make you /call them/ I think
scrts has quit [Ping timeout: 255 seconds]
scrts has joined ##openfpga
promach__ has joined ##openfpga
promach__ has quit [Remote host closed the connection]
digshadow has quit [Ping timeout: 258 seconds]
scrts has quit [Ping timeout: 255 seconds]
scrts has joined ##openfpga
_whitelogger has joined ##openfpga
amclain has quit [Quit: Leaving]
scrts has quit [Ping timeout: 240 seconds]
scrts has joined ##openfpga
scrts has quit [Ping timeout: 268 seconds]
scrts has joined ##openfpga
digshadow has joined ##openfpga
lain has quit [Quit: WeeChat 1.7.1]
<rqou> azonenberg: did you know the kicad pcb calculator can save? :P
<azonenberg> Nope
<azonenberg> lol
<rqou> yeah
<rqou> but (at least on my machine) the part of the UI you need to activate saving is hidden
<rqou> because one of the dialog boxes is too small such that the OK button isn't visible
<rqou> azonenberg: it saves the list of regulators on the "Regulators" tab
<awygle> but is there a speedrunning exploit related to saving...?
lain has joined ##openfpga
<rqou> awygle: no, but saving blocks the UI thread
eduardo__ has quit [Ping timeout: 248 seconds]
eduardo__ has joined ##openfpga
m_w has quit [Quit: leaving]
<azonenberg> So after much pain and suffering debugging unrelated stuff
<azonenberg> apparently Tj of my greenpak is 23.58C
<azonenberg> +/- 1C
<azonenberg> now i just have to add the code for converting the ADC code to temps rather than doing it by hand in a calculator
pie_ has joined ##openfpga
<rqou> oh huh i found a kicad bug/feature
<rqou> i'm pretty sure i didn't introduce this
Hootch has joined ##openfpga
<rqou> azonenberg: do you happen to have a kicad "legacy" board file or know what that even is?
<azonenberg> the .pcb file format
<azonenberg> vs .kicad_pcb
<azonenberg> from version 3 and before iirc?
<rqou> got one i can test stuff on?
<azonenberg> um...
<azonenberg> let me see
* azonenberg digs around into old svn checkout from google code
digshadow has quit [Quit: Leaving.]
<azonenberg> actually no it's .brd
<azonenberg> not .pcb
<azonenberg> This one is from april 2014
<azonenberg> it's actually the coolrunner devboard i gave you
Hootch has quit [Quit: Leaving]
<rqou> azonenberg: what's a legacy schematic?
<azonenberg> That's the current format i think
<azonenberg> The new kicad_sch format is planned for an upcoming refactoring
<azonenberg> but i dont think the .sch format has changed much if at all
<rqou> ah but it doesn't exist yet?
<azonenberg> Yeah
<azonenberg> basically they completely revamped the pcb editor over the last ~2 yrs with the push-and-shove, opengl renderer, etc
<azonenberg> the sch is getting a similar facelift at a time TBD
<azonenberg> But pcb needed it more
scrts has quit [Ping timeout: 248 seconds]
scrts has joined ##openfpga
<rqou> er
<rqou> azonenberg: all the pads are missing
<rqou> on your ancient kicad board
<rqou> unless i broke it?
<azonenberg> idk if they included pcb footprints in the .brd file back then
<azonenberg> the modern kicad_pcb file includes copies of the library footprints
<azonenberg> the schematic just has links to the library which means a sch file is useless w/o the library
<azonenberg> The refamped kicad_sch will probably include copies of the library symbols for this reason
<rqou> can you send me your library for that legacy board?
<azonenberg> it's a lot of stuff
<azonenberg> would take me a while to track them all down
<azonenberg> maybe tomorrow?
<rqou> wait what
<rqou> it's not one file?
<rqou> also, has the ability to save legacy files been deleted?
<azonenberg> It's not one file, the library is a bunch of files each with symbols in them
pie_ has quit [Remote host closed the connection]
<azonenberg> i generally only have one for footprints, but many for schematic
pie_ has joined ##openfpga
<azonenberg> But i also import third party symbols in their own libraries (mostly like graphic logos for things like RoHS or ESD-sensitive)
<rqou> ah ok
<rqou> i'll just say "i smoke-tested it" for now
<azonenberg> And the new PCB library format is a directory per lib and one file per symbol
<azonenberg> Which means less parsing in the app and more in the OS
<azonenberg> which makes sense
<rqou> also, something is wrong somewhere in my setup and i seem to be getting repeated gpu crashes
<azonenberg> But it means they're not directly equivalent
<rqou> i probably need to downgrade my nvidia drivers
<azonenberg> Also, i imagine support for saving legacy stuff died a long time back
<azonenberg> kicad has never been backward compatible, this wasn't a design goal at all
<azonenberg> The rule is, if you create a file in the current version it should work on future versions
<azonenberg> But there's no expectation that a future file will be sane in the current version
<azonenberg> aaanyway bedtime
scrts has quit [Ping timeout: 248 seconds]
scrts has joined ##openfpga
Hootch has joined ##openfpga
scrts has quit [Ping timeout: 260 seconds]
Jarth has joined ##openfpga
scrts has joined ##openfpga
m_t has joined ##openfpga
scrts has quit [Ping timeout: 260 seconds]
scrts has joined ##openfpga
<mithro> azonenberg: KiCad creates a -cache.lib where it keeps copies of schematic stuff in
m_t_ has joined ##openfpga
m_t has quit [Ping timeout: 240 seconds]
eightdot has quit [Ping timeout: 240 seconds]
Jarth has quit [Quit: Leaving]
Marex has quit [Ping timeout: 248 seconds]
Marex has joined ##openfpga
eightdot has joined ##openfpga
m_t_ has quit [Quit: Leaving]
scrts has quit [Ping timeout: 260 seconds]
scrts has joined ##openfpga
<qu1j0t3> not totally random pie_ seems very much on topic
<qu1j0t3> pie_: Nice
* pie_ wants a setup like that :x
<qu1j0t3> nice no-gloves action
<pie_> heh
zino has quit [Remote host closed the connection]
zino has joined ##openfpga
scrts has quit [Ping timeout: 240 seconds]
scrts has joined ##openfpga
digshadow has joined ##openfpga
amclain has joined ##openfpga
scrts has quit [Ping timeout: 255 seconds]
scrts has joined ##openfpga
digshadow has quit [Ping timeout: 240 seconds]
mifune has joined ##openfpga
mifune has joined ##openfpga
scrts has quit [Ping timeout: 240 seconds]
scrts has joined ##openfpga
<balrog> do you need a xilinx jtag adapter to read an xc9000-series CPLD or can you use pretty much any?
<azonenberg> balrog: anything should work, although you obviously need software support
<azonenberg> i'm not familiar with the dumping algorithm for them
<azonenberg> mithro: yes but its only a cache
<azonenberg> if the library changes it will update the cache
digshadow has joined ##openfpga
<balrog> looks like an XC9536
digshadow has quit [Quit: Leaving.]
scrts has quit [Ping timeout: 255 seconds]
scrts has joined ##openfpga
digshadow has joined ##openfpga
scrts has quit [Ping timeout: 240 seconds]
scrts has joined ##openfpga
pie_ has quit [Ping timeout: 240 seconds]
<openfpga-github> [openfpga] azonenberg pushed 1 new commit to master: https://git.io/v7TVm
<openfpga-github> openfpga/master c79bb87 Andrew Zonenberg: Now loading LF/ring oscillator bitstreams
<openfpga-github> [openfpga] azonenberg pushed 1 new commit to master: https://git.io/v7TVX
<openfpga-github> openfpga/master b884453 Andrew Zonenberg: Fixed typo
<azonenberg> Also did the buildbot break again?
<azonenberg> i don't remember seeing it logging stuff
<azonenberg> openfpga-bb: status
<openfpga-bb> openfpga: idle, last build 762h25m38s ago: failed cppcheck_openfpga test_openfpga_normal make_openfpga_ubsan
<azonenberg> openfpga-bb: status
<openfpga-bb> openfpga: idle, last build 762h25m55s ago: failed cppcheck_openfpga test_openfpga_normal make_openfpga_ubsan
<azonenberg> openfpga-bb: force build openfpga
<openfpga-bb> build #106 forced
<openfpga-bb> I'll give a shout when the build finishes
pie_ has joined ##openfpga
<azonenberg> openfpga-bb: status
<openfpga-bb> openfpga: building(building yosys)
<azonenberg> whitequark: so it seems like the commit trigger to do a rebuild isn't working
<whitequark> i think so yeah
lexano_ has joined ##openfpga
<azonenberg> any idea what's up with that?
lexano has quit [Ping timeout: 268 seconds]
<openfpga-bb> build #106 of openfpga is complete: Failure [failed cppcheck_openfpga test_openfpga_normal make_openfpga_ubsan] Build details are at https://openfpga-dashboard.antikernel.net/builders/openfpga/builds/106
<openfpga-github> [openfpga] azonenberg pushed 1 new commit to master: https://git.io/v7ToD
<openfpga-github> openfpga/master b504a02 Andrew Zonenberg: Fixed memory leak found by cppcheck
pie_ has quit [Read error: Connection reset by peer]
<azonenberg> rqou: hmm
<azonenberg> any idea whats up with this? as far as it not correctly finding the slg46140 to test with
pie_ has joined ##openfpga
<rqou> er, what's the problem exactly?
<rqou> are there multiple boards involved?
<azonenberg> yes
<azonenberg> i have a 46537, a 46620, and a 46140
<openfpga-github> [openfpga] azonenberg pushed 1 new commit to master: https://git.io/v7Toj
<openfpga-github> openfpga/master d2096bf Andrew Zonenberg: Removed possible uninitilaized speed if process is typo'd
<azonenberg> and it's not finding the 46140 for some reason
<azonenberg> let me re-build and verify the other stuff is fixed
<azonenberg> openfpga-bb: force build openfpga
<openfpga-bb> build #107 forced
<openfpga-bb> I'll give a shout when the build finishes
<azonenberg> And then i have to track down the auto-builder failing
<azonenberg> CI testing is no use if it doesn't actually continually build
<openfpga-bb> build #107 of openfpga is complete: Failure [failed cppcheck_openfpga test_openfpga_normal make_openfpga_ubsan] Build details are at https://openfpga-dashboard.antikernel.net/builders/openfpga/builds/107
<azonenberg> can you fix that hidapi warning?
mifune has quit [Ping timeout: 260 seconds]
<openfpga-github> [openfpga] azonenberg pushed 1 new commit to master: https://git.io/v7TK3
<openfpga-github> openfpga/master 9303e0c Andrew Zonenberg: Fixed questionable cast in legacy code to shut up cppcheck. Shouldn't matter as this code isn't actually being built or used right now.
<rqou> ugh
<rqou> that part of the code is totally different in my no-libudev PR
<azonenberg> ubsan is complaining about my random number generator overflowing integers
<azonenberg> how are you supposed to fix this?
<rqou> wait
<rqou> i thought unsigned overflow is defined?
<azonenberg> rqou: i'll check that PR shortly
<azonenberg> rqou: i thought it was too
<azonenberg> but ubsan doesnt like it
<azonenberg> I'm just gonna __attribute__((no_sanitize("integer")))
<azonenberg> :p
<azonenberg> on that function
<openfpga-github> [openfpga] azonenberg pushed 1 new commit to master: https://git.io/v7TKV
<openfpga-github> openfpga/master 8eff19d Andrew Zonenberg: Added attribute to tell ubsan that the RNG's integer overflow is intentional
<azonenberg> openfpga-bb: force build openfpga
<openfpga-bb> build #108 forced
<openfpga-bb> I'll give a shout when the build finishes
<azonenberg> Sooo at this point the FIXME list short term to get a clean build is
<azonenberg> me: test no-udev PR
scrts has quit [Ping timeout: 240 seconds]
<azonenberg> whitequark: figure out why the auto-builds are not working (or point me in the right direction to do that myself)
<azonenberg> aaand the no-sanitize attribute gives warnings on gcc since it's not understood
<azonenberg> great :p
scrts has joined ##openfpga
<whitequark> azonenberg: re attribute: let me look up the right thing to do about the attribute
<whitequark> the reason i made it error on unsigned overflow is uh
<azonenberg> I just added a #ifdef __clang__
<openfpga-github> [openfpga] azonenberg pushed 1 new commit to master: https://git.io/v7T6M
<openfpga-github> openfpga/master fde41d2 Andrew Zonenberg: Only apply no_sanitize attribute in Clang, not G++
<azonenberg> that should fix it for now
<azonenberg> And i'm OK with erroring on unsigned overflow
<whitequark> ah ok
<azonenberg> as long as i can turn it off in explicitly requested cases
<whitequark> btw gcc also has ubsan
<azonenberg> Which i just did
<azonenberg> it's possible my gcc version doesn't have it or something
<azonenberg> becuse it warned about not knowing what the attribute was?
<whitequark> azonenberg: oh
<whitequark> you can use bitwise not
<azonenberg> ?
<azonenberg> I'm still on debian 8 on my desktop
<azonenberg> Have to upgrade to 9 but putting it off since i have other stuff going on :p
<openfpga-bb> build #108 of openfpga is complete: Failure [failed cppcheck_openfpga test_openfpga_normal test_openfpga_ubsan test_openfpga_asan make_openfpga_analyzer] Build details are at https://openfpga-dashboard.antikernel.net/builders/openfpga/builds/108
<azonenberg> whitequark: fele free to fix it
<rqou> wait what
<rqou> that's not the right file at all
<azonenberg> Going to test your no-udev PR and if that works then try to debug the multi-devkit issue
<azonenberg> on it
<whitequark> azonenberg: ok can you remind me tomorrow about the hook issue
<whitequark> I need to depart soon and then sleep
<azonenberg> Lol ok
<azonenberg> this isnt my priority, i'm supposed to be doing research for work today
<azonenberg> But my research involves (among other things) adding libgreenpak4 bitstream import support
<rqou> ugh, so i'm trying to detect more locale problem in my patched kicad
<rqou> so i'm trying to build an EBCDIC locale :P
<rqou> but apparently that basically no longer works
<rqou> i wonder why? :P
Hootch has quit [Quit: Leaving]
<azonenberg> lol
<azonenberg> wait, ebcdic still exists?
<azonenberg> like, outside ibm mainframes?
<rqou> sure, $ zless /usr/share/i18n/charmaps/EBCDIC-US.gz
<rqou> but afaik it doesn't really work
<azonenberg> lol
<azonenberg> honestly i really do not think anyone will care if your patched kicad does not support ebcdic
<rqou> no that's not what i'm doing this for
<rqou> i'm trying to make sure i haven't forgotten any places where a locale-sensitive function was accidentally used
<rqou> a good way to try and test that is a totally f*cked up locale like EBCDIC
<azonenberg> Lol
<rqou> except i just looked and apparently glibc's int->string conversion is hardcoded to the character set glibc was built for
<azonenberg> so how about like utf16 arabic
<rqou> afaict locales only affect isdigit
<azonenberg> that should be odd enough to get noticed
<azonenberg> if anything is totally busticated
<rqou> ok, so what i'm fixing right now is mostly decimal separators
<rqou> because strtod/%f are locale-sensitive
<rqou> and these are used in file parsing
<rqou> but also used by wx
<rqou> and locales are process-global
<rqou> cue mass hilarity and nasal demons all around
<rqou> but various other bits of parsing use functions like strcasecmp
<rqou> which is also affected by locales
<rqou> ok, i just smoke-tested with a locale i manually constructed that's really f*cked up
<rqou> smoke-testing seems to work; going to test a bit more extensively
<rqou> azonenberg: so you know how turkish has i->İ and ı->I?
<rqou> i constructed a locale where a->B
<azonenberg> loool
<rqou> and b->C
<rqou> and so on
<azonenberg> rot13-US
<azonenberg> gogogoo
<rqou> but this only affects toupper/tolower
<rqou> which apparently isn't actually called that often
<rqou> you know, f*ck it i'll just set breakpoints on strto[x] and strcasecmp to see if they get called
<azonenberg> lol
<rqou> wait i can just look at the symbols to see what gets imported
<azonenberg> lol
<rqou> hmm these are all pretty overcomplicated
<rqou> i guess i'll try the traditional g
<rqou> grep
<rqou> ugh this is impossible to test
<rqou> glibc's locale support doesn't work super well for f*cked up locales
<rqou> and triggering the right codepaths in Kicad is "fun"
<rqou> ugh and there's a bajillion different reimplementations of isdigit in kicad
<rqou> also aargh i just discovered something else
<rqou> kicad needs a locale-independent strftime
<rqou> fucking kicad
<openfpga-github> [openfpga] azonenberg pushed 1 new commit to master: https://git.io/v7TDF
<openfpga-github> openfpga/master 96420f3 Andrew Zonenberg: Greenpak4BitstreamEntity / Greenpak4Flipflop: correct loading of netlists that use FF Q/nQ output
pie_ has quit [Read error: Connection reset by peer]
pie_ has joined ##openfpga
scrts has quit [Ping timeout: 268 seconds]
scrts has joined ##openfpga
<pie_> north korea, russia, trump, god help us
<pie_> alternatively you may replace trump with usa
* pie_ goes back to watching some dude play horizon zero dawn
<rqou> wait what?
* rqou goes to check whatthefuckjusthappenedtoday.com
<azonenberg> lol
<azonenberg> i think everyone is watching that now
<rqou> oh, sean spicer is gone
<azonenberg> Among other things yes
<rqou> in other geopolitical news, anybody notice that china and india are arguing about their border again?
<rqou> apparently it's been five weeks of a standoff already
<rqou> but nobody noticed because of russia and trump
<azonenberg> lol surprise surprise
scrts has quit [Ping timeout: 260 seconds]
scrts has joined ##openfpga
<pie_> rqou, nothing really i just heard some news today ive been avoiding catching up on
<pie_> what is this trump jr business?
<pie_> poor kid. or something.
* pie_ doesnt even know how old he is
<cr1901> Trump Jr. ain't a kid. I assume you mean Barron?
<pie_> no idea.ok so the clusterf. seems to be about dude meeting with a russian lawyer?
<cr1901> That would be _indeed_ be Trump Jr, and he's near 40 now
<awygle> yeah. met the lawyer to get dirt on hillary for the campaign. actively illegal. lots of reporters were like "Did this happen??" and Trump Jr was like "yeah totally, here's the email transcript on twitter"
<cr1901> But Repubs are too busy killing poor people and exempting themselves from said policies that would kill poor ppl to care
<rqou> solution: just don't be poor :P
<cr1901> :/
<pie_> might be illegal but it doesnt seem like something to make a big deal out of...
<cr1901> pie_: Well, once upon a time, doing shit like that got a president to resign. It's more about the hypocrisy
<pie_> oh.
<cr1901> that the actual action
<cr1901> than*
<pie_> didnt know thats resignation material
<awygle> the lawyer works for the russian government. it was essentially as though Putin was like "hey Donny I can install you as president, that cool?"
<pie_> so i guess its a question of open hypocrist
<pie_> *crisy
<pie_> awygle, hm. putting it that way.
<pie_> wait i didnt realize this was before the election. of course that would make more sense.
<awygle> lots of muddy waters of course. and arguably not super relevant to open FPGA toolchains :P
<pie_> yeah sorry for the offtopic :P
<cr1901> Pretty sure we stopped enforcing on-topic about a week after this room opened
<pie_> lol
<rqou> pie_: how's politics in hungary going? :P
* cr1901 knows, because he was there XD
<pie_> rqou, i havent a fucking clue. still refugees probably.
<cr1901> pie_: didnt know thats resignation material <-- depends on who you ask. Look up Watergate if intersted
<pie_> cr1901, i still dont get what thats about
<pie_> well ill just look it up again. i probably just forgot.
<cr1901> Very TL;DR: President hired ppl to break into a secure location containing confidential info on a political opponent. This isn't new, but he got caught.
<pie_> hm watergate still seems a category above. though ties to kremlin is a prety good point.
rvense has quit [Remote host closed the connection]
* cr1901 puts on his "America Was Never Great" hat :)
<pie_> cr1901, ;)
<pie_> just gotta have good pr eh
<rqou> are there any countries that are great?
<pie_> "major global players were never great" hat :C
* pie_ brofists rqou melancholically
<rqou> pie_: countries that have a bent crown are great :P :P
<pie_> :'D
<cr1901> rqou: Hmmm... maybe Switzerland? They manage to mind their own business ;)
digshadow has quit [Quit: Leaving.]
<pie_> hm.
<pie_> cr1901, re resignation material. i actually find it fucky how its the getting caught thats the issue
<pie_> --> do bad shit but dont get caught
<pie_> its worse than a pipe dream to expect otherwise but it seems almost more honest and less hypocritical to admit to it half the time
<pie_> or idk....
<cr1901> pie_: Well, to put it glibly, that's America for you
<pie_> thats everybody though, no?
<pie_> isnt that how this works everywhere?
<awygle> whelp i have a 9-5 job again. that'll be fun.
<azonenberg> awygle: Were you funemployed? Consulting? or what
<awygle> azonenberg: weakly-failing self-employment
<awygle> which i've been wrapping up for the last 2-3 months
<azonenberg> Lol i know how that goes
<azonenberg> Sooo let's see, right now libgreenpak4 bitstream loading is coming together nicely
<rqou> alright, i've once again not gotten anything done today
<azonenberg> Still have to add support for DAC, DELAY, DCMP, ACMP, VREF, SHREG, COUNT, and one type of IOB
<rqou> welcome to funemployed :P
<awygle> on the plus side i'll have more time to write GPU-accelerated FPGA routing algorithms :P
<azonenberg> Lol
<rqou> awygle: plz 2 fix arachne-pnr kthx :P :P
<awygle> rqou: working on it
<awygle> actually azonenberg i wanted to ask you about your vision for non-crossbar FPGA routing. not so much algorithmically, but organizationally
<rqou> we need some (preferably not XML :P ) way of describing fpga fabrics, yes
<azonenberg> i would not be opposed to making tables of some sort for greenpak devices vs hard coding ctor calls
<azonenberg> maybe use json since we already have a json parser in there
<azonenberg> we'll still need manually written code for the individual tiles
<azonenberg> but we should be able to add a new device to the family without much custom coding
<rqou> meanwhile xc2bit's knowledge of what is connected to what is just "latent" throughout the code :P
<rqou> needs a balance between that and the xilinx "everything is data-driven all the time"
<azonenberg> Yeah
<rqou> alright, i've wasted enough time reading wikipedia today
<azonenberg> gp4par needs some rewriting as well
<azonenberg> The knowledge of what goes where is entirely in the graph
* rqou goes and grabs a really really late lunch
<azonenberg> but this is currently in a giant pair of functions in gp4par
<azonenberg> this should get moved to libgreenpak4 in the greenpak4device class or something
<azonenberg> because i don't want the gp4par executable to do anything but call into a library
<azonenberg> to ease integration of gp4par with other stuff in the future
pie__ has joined ##openfpga
pie_ has quit [Remote host closed the connection]
<rqou> except your libraries have a c++ abi
<openfpga-github> [openfpga] azonenberg pushed 2 new commits to master: https://git.io/v7kvX
<openfpga-github> openfpga/master 59c095e Andrew Zonenberg: Fixed missing \n at end of log message
<openfpga-github> openfpga/master b160414 Andrew Zonenberg: Added bitstream loading support for DELAY and SHREG
<azonenberg> rqou: yes, and? nothing wrong with C++ libraries as long as your app uses the same ABI
<azonenberg> Which is easy to ensure if you compile the lib on the same compiler
<rqou> which is fine except on a Certain Platform
<rqou> that likes to have many many copies of the CRT and the C++ runtime