Closed shaos closed 12 months ago
Some speeds are worse, but still passing (before and after):
< Info: Max frequency for clock '$glbnet$clk48m': 48.62 MHz (PASS at 48.00 MHz)
< Info: Max frequency for clock '$glbnet$vid_pixelclk': 70.83 MHz (PASS at 30.00 MHz)
< Info: Max frequency for clock '$glbnet$clkint': 451.06 MHz (PASS at 25.00 MHz)
< Info: Max frequency for clock '$glbnet$hdmi_encoder.clk_125MHz': 346.86 MHz (PASS at 140.02 MHz)
< Info: Max frequency for clock '$glbnet$clk96m': 686.81 MHz (PASS at 48.00 MHz)
< Info: Max frequency for clock '$glbnet$clk24m': 136.22 MHz (PASS at 48.00 MHz)
> Info: Max frequency for clock '$glbnet$clk48m': 48.72 MHz (PASS at 48.00 MHz)
> Info: Max frequency for clock '$glbnet$vid_pixelclk': 68.95 MHz (PASS at 30.00 MHz)
> Info: Max frequency for clock '$glbnet$clkint': 279.02 MHz (PASS at 25.00 MHz)
> Info: Max frequency for clock '$glbnet$hdmi_encoder.clk_125MHz': 336.25 MHz (PASS at 140.02 MHz)
> Info: Max frequency for clock '$glbnet$clk96m': 626.57 MHz (PASS at 48.00 MHz)
> Info: Max frequency for clock '$glbnet$clk24m': 341.88 MHz (PASS at 48.00 MHz)
Added example app-draw720 to show the issue - it should look like this on the TV if SOC is fixed (all characters are readable):
Another screen with fixed SOC: And this is "before" picture:
Sorry for letting this linger for a while; I don't really have time to check if this works or if it has side effects. I'll take your word for it and simply merge; if it breaks other things we'll hear it soon enough :P Thanks for contributing!
Thanks! :)
See attached image to see a difference in framebuffer image: Basically it looked like every ODD line displayed twice and EVEN lines are completely ignored by DMA So the fix is making actual resolution for framebuffer 480x320 (as expected) instead of effective 480x160 (see above).
Note: This fix also mostly fixing this issue: https://github.com/Spritetm/hadbadge2019_fpgasoc/issues/143