<davidc__>
I'm actually surprised you saw anything on your packet counters with 0x5D as SFD
<adamgreig>
yea, god knows what the switch thought was going on
<adamgreig>
well - actually no
<adamgreig>
the SFD being wrong like that just meant it saw the sfd a bit early
<adamgreig>
since it still got the (01) to (11) transition
<adamgreig>
just the destination mac address would have been 7FFFFFFF or something, instead of broadcast
<adamgreig>
(and there'd be an extra nibble at the end?)
<adamgreig>
(and the preamble would have been a nibble short but I guess it doesn't care)
<adamgreig>
wikipedia describes the sfd as 0xd5 "in ethernet LSB first bit ordering" :p
<adamgreig>
mac now happily transmitting and receiving packets to/from brams, yay
<adamgreig>
800 LUs for the mac plus a little wrapper to transmit dummy packets and flash leds on valid ones etc
<adamgreig>
but the nextpnr gui makes it look like a lot more than 10% of the chip is in use
<adamgreig>
guessing I wasn't seeing invalid crc counter going up because it was dropping the runt packets instead
Bike has joined ##openfpga
azonenberg_work has quit [Ping timeout: 246 seconds]
<davidc__>
adamgreig: nice! Just making one for fun, or is this one smaller than some of the other open source cores?
<adamgreig>
This is just for fun and learning, I haven't put any effort into optimising it and deliberately avoided looking at open source ones until mine is working
<adamgreig>
Though definitely I need to do that soon before whatever terrible bad habits I'm developing get too firmly cemented :p
<davidc__>
gotcha! I was just curious; I'd done some streaming ethernet processing stuff in the past that I should really open source one of these days
pie__ has joined ##openfpga
pie___ has quit [Remote host closed the connection]
azonenberg_work has joined ##openfpga
azonenberg_work has quit [Ping timeout: 272 seconds]
genii has quit [Remote host closed the connection]
azonenberg_work has joined ##openfpga
unixb0y has quit [Ping timeout: 268 seconds]
unixb0y_ has joined ##openfpga
azonenberg_work has quit [Ping timeout: 240 seconds]
rohitksingh has quit [Read error: Connection reset by peer]
rohitksingh has joined ##openfpga
_whitelogger has joined ##openfpga
jhol has joined ##openfpga
rohitksingh has quit [Ping timeout: 268 seconds]
futarisIRCcloud has joined ##openfpga
zkms has quit [Quit: WeeChat 2.2]
zkms has joined ##openfpga
GuzTech has joined ##openfpga
rohitksingh_work has quit [Read error: Connection reset by peer]
rohitksingh_work has joined ##openfpga
futarisIRCcloud has quit [Quit: Connection closed for inactivity]
Bike has joined ##openfpga
rohitksingh_work has quit [Read error: Connection reset by peer]
<cpresser>
does anyone know a reasonable cheap standalone jtag progammer? like upload binary and config (tap id), and then just a button to programm a chip
<cpresser>
i could do this with a raspberry or similar, but I rather want to buy a ready made device with a manual. its supposed to be used by unskilled workers in a production line
Miyu has joined ##openfpga
Flea86 has joined ##openfpga
<plaes>
cpresser: I have used ft2232 with simple openocd script
<plaes>
but it was SWD, not JTAG
Bike is now known as Bicyclidine
<ZipCPU>
cpresser: Have you looked at libxsvg?
<ZipCPU>
It might need a little bit of TLC to get it running on a Pi, but it does include instructions for doing so
<ZipCPU>
I've also used it to load FT2232 devices as well
genii has joined ##openfpga
<cr1901_modern>
libxsvg is pretty cool. It is unfortunately extremely slow on Windoze and before I was able to diagnose why I managed to accidentally destroy the only CPLD I had to test on it :(
<cr1901_modern>
well actually I have a good idea nowadays why it was so slow (libusb overhead is extremely bad for small xfers on Windoze).
<plaes>
you mean libxsvf?
<cr1901_modern>
plaes: Yes, oops :)
<cr1901_modern>
Although I am all for lib Extremely-Smol Scalable Vector Graphics
<cr1901_modern>
eXtremely-smol*
Flea86 has quit [Quit: Goodbye and thanks for all the dirty sand ;-)]
pie___ has quit [Remote host closed the connection]
pie__ has joined ##openfpga
<cpresser>
plaes: yep, thats what i am currently using. raspberry-pi and openocd. but: people are stupid and break that all the time. like remove the sd-card and ask why its not working anymore... stuff like that
<ZipCPU>
foolproof? ??? Something that will keep folks from removing SD-cards and power mid-operation? Somehow I don't think additional hardware will keep someone from being that foolish
GuzTech has quit [Quit: Leaving]
emeb has joined ##openfpga
pie___ has joined ##openfpga
pie__ has quit [Remote host closed the connection]
<cpresser>
according to douglas adams, there is no such thing as foolproof :) fools are really resourcefull.
<plaes>
yeah, you can make a foolproof thing, but then a smart guy comes...
<tnt>
cpresser: if it's for production programming, you could easily make an embedded SVF player reading from SPI flash using any microcontroller or small fpga.
<tnt>
Is there a way to define symbol from the command line in yosys ?
<tnt>
Like in gcc you can -DXXXX-YYY ?
m_w has joined ##openfpga
<daveshah>
You can use -D, but you have to use it on read_verilog
<tnt>
Ooooohhhh, right, I didn't think of that.
pie__ has joined ##openfpga
Hamilton has joined ##openfpga
pie___ has quit [Ping timeout: 260 seconds]
<tnt>
wtf ... hyperram has a 10 ns clock period, it's using ddr so you get new data every 5 ns. And the clock-to-out is "somwhere" between 1 and 8 ns ... so ... you have no idea where you data is vs the clock edge basically.
<daveshah>
tnt: Isn't that what DQS is for?
<daveshah>
*RWDS for Hyperram
<tnt>
Sort of I guess ... but then they also have variants with PSC,PSC# for "DDR Center Aligned Read Strobe (DCARS)" ...
<sorear>
wtf luke still hasn't been banned from the riscv list
<TD-Linux>
glad I'm not on that list tbh
<sorear>
he is singlehandedly making riscv a dramatically worse project to be involved with
<q3k>
who's luke
<daveshah>
q3k: lkcl
<daveshah>
I'm guessing anyway
<whitequark>
sorear: have you escalated that with riscv list admins?
<q3k>
what's the tl;dr about them making risc-v a worse project to be involved with?
<sorear>
idk who's actually a list admin, I brought it up with some foundation people last year
<sorear>
I'm not an eloquent person and I can't describe it much better than "very obnoxious"
<TD-Linux>
tbh it's not that bad compared to some of my mailing lists
<TD-Linux>
I would describe it as "impedance mismatch between authoritative tone and degree of clue"
<sorear>
that's a good take
* qu1j0t3
feels described
<sorear>
the *easiest* thing to object to is that he's publically objected to codes of conduct
<sorear>
but that needs to be understood as a symptom