rohitksingh has quit [Remote host closed the connection]
rohitksingh has joined #yosys
<shapr>
awygle: 45mph is the speed limit
* shapr
is tired and silly
<awygle>
lol hey shapr
<shapr>
howdy awygle, how you doin?
<awygle>
i've been better
<shapr>
oh, what's up?
<awygle>
but hanging in there
<awygle>
i live in the Seattle area, so things are a bit crazy here atm
<shapr>
you dropped off irc for a year or so, life happenings?
<shapr>
ohh
<awygle>
i did do that yeah. mental health stuff, which i'm happy to report is _much_ better
<shapr>
glad to hear that
<awygle>
how have you been?
<shapr>
similar happenings, on call made me crazy, I quit my job in April, spent six months at recurse center, and just now have a Haskell job in Boston
<shapr>
I've been here a month, it's a blast
<awygle>
nice!
<shapr>
But I'd also like to get yosys doing xilinx 7 stuff on this NeTV2 I got from bunnie's campaign
<shapr>
though I hear from daveshah et al that xray won't be done for months
<shapr>
and especially not for the 100T chip in this NeTV2
rohitksingh has quit [Ping timeout: 265 seconds]
<shapr>
awygle: did you change jobs in the past year or so?
<awygle>
it's been about 3 years since i did that i think
rohitksingh has joined #yosys
<awygle>
although i may have updated my linkedin in the last year or so :p
* shapr
looks
<awygle>
purely out of laziness
<shapr>
ack, I haven't looked at linkedin in some time
* shapr
digs out from under a pile of notifications
az0re has joined #yosys
ZipCPU has quit [Ping timeout: 255 seconds]
<awygle>
where does ecppack come from? trellis, presumably?
<ZirconiumX>
Yeah
<awygle>
it doesn't seem to handle 12Fs correctly
<awygle>
in that... it doesn't handle them :p
<ZirconiumX>
...Because 12Fs are 25Fs
<awygle>
this is because they're 25Fs, yes
<awygle>
but they have different IDCODEs
<ZirconiumX>
Which you handle at the programming stage
<awygle>
ecppack outputs an SVF
<awygle>
which expects the correct IDCODE
<awygle>
afaict
<awygle>
at any rate programming with openocd is failing with "WANT: <25F IDCODE> READ: <12F IDCODE>"
<awygle>
updating trellis to see if that fixes it
<awygle>
it does not
<awygle>
overriding the IDCODE in the ecppack command line does work correctly, so that's good
emeb has quit [Quit: Leaving.]
<awygle>
actually i think the problem is in nextpnr so i'll update that too...
* shapr
asks awygle to connect on the embedded network
<shapr>
yay
<shapr>
I'm gonna bet I have the largest number of linked in endorsements, for cartwheels
<awygle>
updating nextpnr does not fix it, filed a bug report
<awygle>
adding the override to my local copy of nmigen and moving on with my life
proteusguy has quit [Ping timeout: 265 seconds]
<whitequark>
it's weird that the FOSS toolchain approach to market segmentation is "expose the lie to everyone who needs to use the chips"
<whitequark>
there's no "use 12f" switch. every single person who wants to use 12f's has to go through discovering this, rather painfully too
<whitequark>
nmigen has a very nice table that makes it *especially* obvious in the platform code
<awygle>
yep
<awygle>
do you want me to upstream this into nmigen? think i can make it fit into the template so it's non-disruptive
<whitequark>
if it's a nextpnr bug it should be fixed in nextpnr
<whitequark>
it could be also fixed in nmigen temporarily to reduce the disruptiveness
<whitequark>
but with emphasis on former
<awygle>
definitely
<awygle>
guess i'll wait and see whether anybody has a comment on the nextpnr issue. i'm utterly unfamiliar with that codebase and don't really want to try to push this uphill there right this second.
<whitequark>
ime nextpnr bugs are fixed exceptionally quickly
<whitequark>
so i don't usually bother with workarounds
<awygle>
mk, noted. i'll hope to have the same experience
<awygle>
hm now how do i program attached flash with openocd instead of the internal sram cells...
proteusguy has joined #yosys
rohitksingh has quit [Ping timeout: 240 seconds]
<whitequark>
i think you might need a proxy bitstream
citypw has joined #yosys
<awygle>
looking like it yeah
Degi has quit [Ping timeout: 256 seconds]
Degi has joined #yosys
cr1901_modern has quit [Ping timeout: 255 seconds]
cr1901_modern has joined #yosys
thardin has quit [Ping timeout: 268 seconds]
ZipCPU has joined #yosys
dh73 has joined #yosys
rohitksingh has joined #yosys
_whitelogger has joined #yosys
strongsaxophone has joined #yosys
dh73 has quit [Quit: Leaving.]
_whitelogger has joined #yosys
rohitksingh has quit [Ping timeout: 255 seconds]
rohitksingh has joined #yosys
rohitksingh has quit [Ping timeout: 256 seconds]
rohitksingh has joined #yosys
_whitelogger has joined #yosys
rohitksingh has quit [Ping timeout: 240 seconds]
rohitksingh has joined #yosys
_whitelogger has joined #yosys
pie_[bnc] has joined #yosys
<tnt>
Huh yosys sees a path (when doing ltp) from a SB_RAM40_4K write enable to the read data port ...
<tnt>
(which means in a full design where the read data influences what you do next, it finds a loop ...)
dys has quit [Ping timeout: 258 seconds]
<corecode>
is the write enable a write/nread?
<tnt>
Shouldn't matter, they're registered. write enable is not going to influence the read data until the next cycle.
<sensille>
could also be that if read and write address are the same, the write data shows up on the read port in the same cycle
<tnt>
Also read / write ports don't even use the same clock ..
<tnt>
All write are registered.
<corecode>
oh i didn't read right
<corecode>
the bram, not the spram
<corecode>
i'll have to watch your video to find out more :)
rohitksingh has quit [Ping timeout: 255 seconds]
<tnt>
Actually, I'm not even sure if ltp knows about clock / registers etc at all ... I might be expecting too much.
<daveshah>
No it doesn't
<daveshah>
You have to use the Yosys selection syntax to exclude these cell types
rohitksingh has joined #yosys
<tnt>
daveshah: you don't happend to have an example ? syntax is rather obscure. I tried "ltp */t:SB_LUT4" but that doenst' seem to work.
rohitksingh_ has joined #yosys
rohitksingh has quit [Read error: No route to host]
marex-cloud has joined #yosys
<daveshah>
I think if it is a flattened design the */ isn't needed
<daveshah>
I don't have an example to hand though
rohitksingh_ has quit [Remote host closed the connection]
<tnt>
Ok, I think I got it: ltp t:SB_DFF* t:SB_SPRAM256KA t:SB_RAM40_4KNW %% %n
rohitksingh has joined #yosys
<tnt>
tx for the tip !
rohitksingh has quit [Remote host closed the connection]
rohitksingh has joined #yosys
rohitksingh has quit [Remote host closed the connection]
<awygle>
daveshah: any chance you can confirm whether https://github.com/YosysHQ/nextpnr/issues/409 is a nextpnr issue or a trellis issue? just wanna make sure i filed it in the right place
<daveshah>
awygle: it's both, now that I'm free from political control from Edmund I'll deal with it
<awygle>
okay, cool :) thanks
<awygle>
(no rush or anything obv)
<daveshah>
It certainly wasn't an oversight previously, it just got forgotten about
<daveshah>
subsequent to my departure from seda
<awygle>
sure
X-Scale` has joined #yosys
X-Scale has quit [Ping timeout: 256 seconds]
X-Scale` is now known as X-Scale
cr1901_modern has quit [Disconnected by services]
cr1901_modern has joined #yosys
cr1901_modern has quit [Read error: Connection reset by peer]
cr1901_modern has joined #yosys
<Degi>
Hm it still gives the error, more specifically "fata: empty string is not a valid pathspec. please use . instead if you meant to match all paths" while executing git checkout "$(make --silent -C "${srcdir}/yosys" echo-abc-rev)"
cr1901_modern has quit [Ping timeout: 256 seconds]
<ZirconiumX>
Try a `make clean` before rerunning `make`
<Degi>
Well I deleted the whole directory and redownloaded it, I guess the pkgbuild is broken
<Degi>
Hm I'll just delete the echo-abc-rev from the pkgbuild and see what happens
<Degi>
Well now it works
<Degi>
Yeah somebody added that in an update 5 hours ago which apparently broke the AUR package.
dormito has quit [Ping timeout: 265 seconds]
dormito has joined #yosys
m4ssi has quit [Remote host closed the connection]
<lambda>
Degi: my PR adding echo-abc-rev was accepted, then thasti updated the PKGBUILD to use it, then it was reverted.
<Degi>
Hm oh. Yeah I just deleted that part from the PKGBUILD and now it works just fine
dh73 has quit [Read error: Connection reset by peer]