Hi Charles
Thanks for the latest merge.
I would like to address this elder post and update it with the current 
state of my explorations, and encourage a test of axis,
with the latest rev 2.1 soc image.

On Friday, 1 September 2017 17:58:59 UTC+2, Charles Steinkuehler wrote:
>
> On 9/1/2017 10:43 AM, [email protected] <javascript:> wrote: 
> > 
> > On 01/09/17 15:45, Charles Steinkuehler wrote: 
> >> ...but since it's already working, I think programming the FPGA via 
> >> overlays should remain supported for folks who aren't trying to use 
> >> things like HDMI. 
> > 
> > My thoughts for what they are worth. 
> > 
> > I would have no intention of ever using HDMI from a board with 1GB 
> > SDRAM and no GPU. 
> > 
> http://www.terasic.com.tw/cgi-bin/page/archive.pl?Language=English&CategoryNo=205&No=1046&PartNo=2
>  
>

Yes no gpu however the older vip framereader core is able to act as bus 
master and utilize the fpga to hps sdram bus,
which is a great advancement over the old style bit-banged framebuffers you 
are refering to.
even in the rev 2.0 image without this enhancement axis worked adequately. 
 

> > 
> > I am not sure we should even encourage it by making it available! 
> > Just asks for loads of "Axis locks up when I load my half billion line 
> > printer file" threads. 
>

This needs to be tested (again) on he new image.
 

>
> I would never recommend using the HDMI out for Axis, that's just 
> asking for problems.  Some of the other light-weight UI's might be OK, 
> but I doubt it.  People forget how slow CPU bit-banged displays can be 
> (and most of the young folk have never even used one). 
>
>
There has been some advancement with sddm - kwin and lxqt

 

> > It certainly needs to be able to be disabled or not used, with 
> > preferably the inconvenience of reboot etc. being for those seeking to 
> > use HDMI :-P 
>
> I think HDMI should be disabled by default, or possibly in a text-only 
> console mode (I could actually see that being useful).  I agree 
> enabling any sort of graphics output ought to require at least a 
> reboot, and possibly a compile (or installation) of a different FPGA 
> image. 
>
>
This has been solved via different device trees for the de10 nano:
only those containing _fb_ activate the framebuffer
_fb_hd is for 192x1080 resolution, _fb_ has 1024x760 resolution matching 
the bitfiles in the #95 pr commit.

If you prefer also not to have the detktop gui related files and run 
console only, you should be able to use the deo_nano sd image with a non 
_hd_ devicetree.
(look at the devicetree utilized upon first-boot in the de10_hdmi 
instructions, works fine without loading the fpga upon boot).
((I have not tested a machinekit fpga load run in this setup, it is 
Supposed to work  ))

I think there are some open-source text-mode displays that could be 
> compiled into the FPGA that have Linux drivers and could be tied to 
> the HDMI output... 
>
> ...but again, I'm not particularly worried about getting HDMI output 
> working at all.  I think effort is currently better spent on things 
> like fixing up the hm2_soc_ol driver and automating the uSD image 
> creation. 
>
> Yes agree it seem like the deb package building is failing ... atm.
:-)

 

> -- 
> Charles Steinkuehler 
> [email protected] <javascript:> 
>

-- 
website: http://www.machinekit.io blog: http://blog.machinekit.io github: 
https://github.com/machinekit
--- 
You received this message because you are subscribed to the Google Groups 
"Machinekit" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
Visit this group at https://groups.google.com/group/machinekit.
For more options, visit https://groups.google.com/d/optout.

Reply via email to