00:00 <agg> But yes I too really want a cortex-m and some FPGA fabric. It exists but not quite right yet I guess.

00:06 <TD-Linux> I didn't see any way to use it for capture

00:06 <TD-Linux> the compare lines for it are output only

00:07 <TD-Linux> the one obnoxious thing is that the hrtim isn't actually emulatatble with a fpga

00:08 <TD-Linux> (except, maybe, with some very clever manual placement)

00:14 <agg> it can't capture the CHA1 etc outputs but can capture the sub-timer counter values based on any of the external events or output set/reset events, so e.g. can capture the comparator triggering

00:14 <agg> i use it to read back how long a pulse was that's terminated by a per-pulse current limit comparator

00:15 <TD-Linux> oh huh, I'll have to reread the datasheet again then

00:15 <agg> this is on the f334, i think it changes a bit on some other chips

00:15 <TD-Linux> I was looking at the g7 version

00:15 <agg> not used it on the h7

00:15 <TD-Linux> h7 doesn't have it :(

00:15 <agg> there's a g7 now??

00:16 <TD-Linux> er sorry I meant g4

00:16 <agg> i think the g474 and also some h7s have hrtim

00:16 <agg> yea, h7x3 for example does

00:17 <TD-Linux> oh cool

00:17 <agg> how did you find the g4? i've not used one before but it seems like one day i'll probably have to :p

00:18 <TD-Linux> I have a project that currently uses the the f7 (without hrtimer) and I wanted hrtimer. the closest I could pick is g4, but that sacrifices clock speed and gives me some useless stuff like cordic accelerator

00:18 <agg> it's wild that it has a CORDIC peripheral

00:18 <agg> i don't think any f7 has an hrtim

00:19 <TD-Linux> yeah. I didn't realize the h7 was an option for some reason

00:19 <agg> the FMAC looks kinda cool. the g4 in general seems a pretty wicked lowish speed dsp/motor/power control chip

00:21 <TD-Linux> assuming the packaging is ok I'll probably swap next version of this board to h7 then

00:23 <agg> they are pretty wild chips

00:23 <TD-Linux> the FMAC and CORDIC are kind of cool but it's awkward how they are separate memory mapped controllers. the FMAC in particular is pretty redundant with the DSP instructions

00:23 <agg> clock speed is up at like 400MHz, some are dual core too

00:23 <TD-Linux> in fact I think m4f dsp instructions would probably outperform it

02:53 <cr1901_modern1> mmicko: Found this by chance... this is nuts and I love it :'D https://github.com/mmicko/micropython/tree/mc68000

23:22 <emeb_mac> Having some fun with this UP5k hooked to an RPi - loading the bitstream with a shell script that uses sysfs to drive the GPIO results in proper load & execute of of code from SPRAM, but the HFOSC is not properly calibrated.

23:24 <emeb_mac> But, loading the bitstream via a C prog that uses the gpio character device results in properly calibrated HFOSC but the CPU isn't running properly (sorry - not SPRAM, BRAM)

23:24 <emeb_mac> So, down to comparing the differences in timing of the two approaches.

23:26 <emeb_mac> Really proud of myself for inadvertently finding two unique ways to fail.

23:36 <pie_> I dont know where I could possibly ask this; so I looted an big ol CRT monitor that turns on and shows its "no signal" stuff, but I cant seem to be able to enable my vga output on my laptop

23:36 <pie_> edp1 is the built in monitor, and dp1 and dp1 and hdmi1 and hdmi2 are grayed out in arandr and thats all i have

23:46 <pie_> protip, the lenovo thinkpad dock has a vga port. it seems to take precedence. doh.

23:58 <gruetzkopf> yeah, there's a video switch in the laptop