<openfpga-github>
[openfpga] azonenberg pushed 1 new commit to master: https://git.io/vXeqe
<openfpga-github>
openfpga/master fddece6 Andrew Zonenberg: gpcosim: added initial version of verilog PLI co-simulation wrapper. Doesn't do anything useful yet.
<cr1901_modern>
azonenberg: Well, I've only used iverilog. Just heard that verilator is also good.
<azonenberg>
well in theory the PLI module is portable across sims
<azonenberg>
So, i guess we'll find out if i move :p
<openfpga-github>
[openfpga] azonenberg force-pushed gh-pages from 44d70c3 to 835517b: https://git.io/v6vmV
<openfpga-github>
openfpga/gh-pages 835517b Travis CI User: Update documentation
<travis-ci>
azonenberg/openfpga#157 (master - fddece6 : Andrew Zonenberg): The build passed.
<rqou>
and then there are super gross hacks like VPI for GHDL
<rqou>
which ironically works better than the alternative
<rqou>
GHDL has a PLI interface that they invented called AVHPI
<rqou>
it's VHPI but with an Ada interface rather than a C interface
<rqou>
aka nobody's going to deal with that
<rqou>
somebody really needs to design a "actually works cosimulation interface"
<rqou>
maybe ##openfpga can? :P
<azonenberg>
lol
<rqou>
btw GHDL has this amazing sentence in its documentation: "This function must be called once, and returns 0 at the end of the simulation. In case of failure, this function does not return. This has to be fixed."
<azonenberg>
rqou: fwiw my last project along these lines used ISim
<azonenberg>
which does not even support PLI
<azonenberg>
so i had to use $fread/$fwrite with pipes
<azonenberg>
:p
<rqou>
i hate EDA tools
<rqou>
I didn't even realize those functions existed
<rqou>
aaand I just realized GHDL's documentation doesn't actually mention it supports VPI
<rqou>
you have to Use The Source(TM) to find out
wareya has left ##openfpga [##openfpga]
massi has joined ##openfpga
<azonenberg>
lol eda is a nightmare
<azonenberg>
why do you think we're all here? :p
Bike has quit [Quit: leaving]
rqou has quit [Quit: ZNC 1.6.3+deb1 - http://znc.in]
rqou has joined ##openfpga
* nats`
is here because he's a spy for the account of the Grand EDA Master From Hell
cosmobird has quit [Ping timeout: 244 seconds]
<rqou>
holy **** let's encrypt now supports IDNs
<rqou>
need to update my certs now
cosmobird has joined ##openfpga
cosmobird has quit [Ping timeout: 256 seconds]
<rqou>
aaand it doesn't work
<rqou>
good job let's encrypt
<plaes>
:)
<rqou>
'Name does not end in a public suffix'
<rqou>
where the subject in question is "xn--j6w193g" aka "香港"
<rqou>
maybe whitequark will be sad now? :P
<rqou>
anyways, good thing that firefox refuses to propagate HKPK from private browsing into normal
<rqou>
*HPKP
clifford has quit [Ping timeout: 244 seconds]
<cr1901_modern>
let's encrypt is not user friendly. And I KNOW they're really trying.
<cr1901_modern>
Can't tell if this is b/c security is legit hard, or this is a side effect of professionals willingly making it inaccessible for years.
<rqou>
this has nothing to do with user friendliness
<cr1901_modern>
Oh, that's good. 10/10 would break again
<rqou>
hey, the .hk registrar bundles a free .香港 with all .hk domains
<rqou>
so I can test all your IDN code
<rqou>
(although I actually bought this domain specifically for the .香港 and not the .hk)
<rqou>
I specifically got this domain so I could get the 龘 character
<rqou>
it's one of the BMP characters with the most number of strokes
<rqou>
there's some more chinese characters with more strokes in the astral planes, but last time I tried HKIRC doesn't allow astral plane characters
clifford has joined ##openfpga
doomlord has joined ##openfpga
cosmobird has joined ##openfpga
<rqou>
offtopic: I have brought back to life the old registry of tuples of 16-bit integers at f055.cc (now redirects to f055.io)
<rqou>
hooray for archive.org
doomlord has quit [Ping timeout: 265 seconds]
eightdot has quit [Ping timeout: 260 seconds]
eightdot has joined ##openfpga
azonenberg_work has quit [Ping timeout: 250 seconds]
tecepe has quit [Ping timeout: 260 seconds]
tecepe has joined ##openfpga
tecepe has quit [Ping timeout: 260 seconds]
tecepe has joined ##openfpga
amclain has joined ##openfpga
m_w has joined ##openfpga
digshadow has quit [Ping timeout: 250 seconds]
Bike has joined ##openfpga
mifune has joined ##openfpga
mifune has quit [Client Quit]
azonenberg_work has joined ##openfpga
tecepe has quit [Remote host closed the connection]
tecepe has joined ##openfpga
<DocScrutinizer05>
>>you would need at least 800 millizonenbergs to crack it<< [#kicad] :-D
<DocScrutinizer05>
<c4757p> Physical Unclonable Function. basically a way to make a secure signature of a piece of hardware that can't be copied onto another piece of similar hardware without serious effort <xzcvczx> #define serious effort? like decapping sort of effort or? <cyborgar_> you would need at least 800 millizonenbergs to crack it
<azonenberg_work>
lol
<azonenberg_work>
some PUFs are pretty easy to clone gievn the right gear
<azonenberg_work>
for example, SRAM that powers up to a certain value
<azonenberg_work>
you can use backside photon emission etc to read it out
<azonenberg_work>
then given a new chip you can FIB the ram to weaken one half or the other of the sram inverter loop
<azonenberg_work>
so it still works but is biased in one direction or the other
<azonenberg_work>
somebody did a paper on that a while ago
forrestv has quit [Ping timeout: 245 seconds]
forrestv has joined ##openfpga
massi has quit [Remote host closed the connection]
cosmobird has quit [Remote host closed the connection]
<rqou>
how many millizonenbergs do we need to mount more backside attacks?
<rqou>
XNU kernel is a much bigger pile of duct tape than I thought
<azonenberg_work>
lol
<rqou>
i'm no longer sure which of Linux/XNU/NT have more duct tape in the kernel
<azonenberg_work>
they all suck
<azonenberg_work>
only question is how hard in which ways
<azonenberg_work>
any idea what the free/open BSD kernel looks like these days?
<rqou>
XNU seemed very secure until people started actually poking it :P
<azonenberg_work>
lool
<rqou>
i've never looked at the BSD kernels
<cr1901_modern>
azonenberg_work: NetBSD is the only kernel I've really looked at. MUCH easier to understand than Linux once you get used to the directory structure. Free is similar
<cr1901_modern>
they both use bus_space_map/read/write fcns to abstract away device driver I/O, so in theory the same driver can be used on different platforms w/ different bus architectures w/o changes
<rqou>
doesn't linux also have that?
<cr1901_modern>
Not that I'm aware
<rqou>
maybe only the PCI bus is abstracted
<rqou>
no wait
<rqou>
at least some of the driver IO has to be abstracted
<rqou>
because the USB HCI drivers are shared across bus architectures
<cr1901_modern>
Well, even if that's true, my original point still stands :P. I've had a much easier time looking at BSD kernel code than Linux
<rqou>
quite possible
<rqou>
you hang out in #j-core, have you seen the giant mess that is arch/sh in linux? :P
<cr1901_modern>
No. And don't really want to, tbh. I'm waiting for the J4 patent to expire, then I'll attempt a Net port
<cr1901_modern>
depending on how I feel and how confident I am
<rqou>
no nommu support on netbsd?
<cr1901_modern>
NetBSD doesn't support any system without an MMU
<cr1901_modern>
it's a "open project", but nommu skeeves me in the first place so I'm not the one to add it
<rqou>
hmm i wonder how possible it is to build a linux 4.8 for sh4/dreamcast
<rqou>
i actually obtained a dreamcast and booted a ~2003 era linux build on it, but I have no idea how to put together a new one
<cr1901_modern>
rqou: Yea I think you're right. Linux's device abstraction facilities are all over the place, like DRM and drivers/tty/serial/serial_core.c
<rqou>
ugh tty
<rqou>
iirc the tty subsystem is awful
<rqou>
it was also one of the last components that was holding up deleting the Big Kernel Lock
carl0s has joined ##openfpga
<azonenberg_work>
rqou: ?
<azonenberg_work>
Big Kernel Lock?
<cr1901_modern>
Some old legacy thing that prevented the kernel from being multithreaded IIRC
<rqou>
during the last push to get rid of it the tty subsystem was migrated to a Big TTY Mutex that was global only within the tty subsystem
<azonenberg_work>
lol
<azonenberg_work>
So the entire tty subsystem still has a global lock"
<carl0s>
Hi all :), cyrozap, just downloaded all the TRM and Registers TRM from Cypress to try to understand all the memorymaps, i just want to be sure to not work on something that you are already working, what family of PSoC are you working on (libopencm3 support)? i would like to start on 4200 fam.
<rqou>
i have no idea
<cr1901_modern>
I like atomics for lock-free data structures, but sadly I'm not exactly wonderful at using them. Or implementing them.
<cr1901_modern>
Or even understanding them.
<rqou>
last time I looked at the tty/pty/termios/job control stuff, I was amazed that it worked
<qu1j0t3>
:)
* qu1j0t3
should get a T-shirt, HOW DID THIS EVER WORK AT ALL
<cr1901_modern>
It is a perfect metaphor for how errors always cascade in ways that waste more of your time and energy, as opposed to letting you just get work done.
<qu1j0t3>
I remember that episode. It's awesome.
<cr1901_modern>
It's one of my favorites for that scene alone
<qu1j0t3>
yep
digshadow has joined ##openfpga
<balrog>
ohai qu1j0t3
<qu1j0t3>
balrog: o/
<whitequark>
anyone pinged me here?
digshadow has quit [Ping timeout: 245 seconds]
<cr1901_modern>
rqou was asking about your domain
<rqou>
i was?
<rqou>
no, I was joking about my domain
<cr1901_modern>
Oh. Well if fucking search worked on my IRC client, I could check
<cr1901_modern>
lmao
<rqou>
just use grep :P
<azonenberg_work>
whitequark: so what's the status on the buildbot? What has to be done on my end to set things up? What if anything do you need to do to set it up? When do you expect to have time to do that?
<whitequark>
azonenberg_work: all should be set on your end. I think I'll be able to set it maybe later today
<whitequark>
ah, can you make a repo for the buildbot config?
<azonenberg_work>
i dont have my github password saved on the work box
<azonenberg_work>
Will do it when i get home
m_w has quit [Quit: leaving]
digshadow has joined ##openfpga
digshadow has quit [Client Quit]
carl0s has quit [Quit: Leaving]
digshadow has joined ##openfpga
digshadow has quit [Ping timeout: 260 seconds]
digshadow has joined ##openfpga
digshadow has quit [Client Quit]
digshadow has joined ##openfpga
digshadow has quit [Quit: Leaving.]
digshadow has joined ##openfpga
digshadow has quit [Client Quit]
maaku has quit [Ping timeout: 260 seconds]
maaku has joined ##openfpga
azonenberg_work has quit [Ping timeout: 244 seconds]