Incorrect color rendering on S8+ (dream2lte) [xfce4]

Describe your issue

S8+ (probably S8 as well, cannot test) is rendering colors incorrectly.

What's the expected behavior?

Colors should be rendered correctly as seen here, and here (Screenshot from VNC)

What's the current behavior?

Colors are being rendered incorrectly (Reds and Blues are getting swapped), as can be seen here.

How to reproduce your issue?

Simply install pmos with any (supported) Desktop Env using pmbootstrap.

What device are you using?

samsung-dream2lte, Samsung Galaxy S8+ SM-G955F (Exynos)

On what postmarketOS version did you encounter the issue?

• edge (master branch)
• stable (v20.05 branch)

On what environment did you encounter the issue?

Phone environments

• Phosh
• Plasma Mobile (Using software renderer)
• Sxmo

Other

• No environment
• GNOME
• KDE Plasma
• Kodi
• MATE
• Shelli
• Sway
• Weston
• Xfce4
• fbkeyboard
• i3wm

How did you get postmarketOS image?

• from https://images.postmarketos.org
• I built it using pmbootstrap
• It was preinstalled on my device

mentioned in merge request !1725 (closed)

By Raghav Vashisht on 2020-12-27T19:21:51

added category::device type::bug labels

I'll see if I can reproduce it on my S8+ tomorrow. I did not have this problem last time I flashed pmos (around 2 weeks ago).

There are some patches in the repo that solves very similar issues, for example https://gitlab.postmarketos.org/postmarketos/pmaports/-/blob/master/device/testing/linux-motorola-nash/99_framebuffer.patch, so that's worth investigating!

By Henrik Grimler on 2020-12-27T23:25:51

I have the same problem. I have probably always had it but just not noticed it.

By Henrik Grimler on 2020-12-28T10:10:07

I hadn't noticed it either until I played a youtube video lol.

By Raghav Vashisht on 2020-12-28T10:10:07

CC: @drebrez (you made the device ports)

There are some patches in the repo that solves very similar issues, for example https://gitlab.postmarketos.org/postmarketos/pmaports/-/blob/master/device/testing/linux-motorola-nash/99_framebuffer.patch, so that's worth investigating!

Yes, this can be fixed in the kernel source. Related wiki page (feel free to extend and update, some of this might be outdated.)

By Oliver Smith on 2020-12-28T16:50:48

Alright, here's the patch that fixes it

diff --git a/drivers/video/fbdev/exynos/dpu/decon_dsi.c b/drivers/video/fbdev/exynos/dpu/decon_dsi.c
index eeaa2ba19778..3e597a145432 100644
--- a/drivers/video/fbdev/exynos/dpu/decon_dsi.c
+++ b/drivers/video/fbdev/exynos/dpu/decon_dsi.c
@@ -862,11 +862,11 @@ int decon_check_var(struct fb_var_screeninfo *var, struct fb_info *info)
 	case 24:
 		/* our 24bpp is unpacked, so 32bpp */
 		var->bits_per_pixel	= 32;
-		var->red.offset		= 16;
+		var->red.offset		= 0;
 		var->red.length		= 8;
 		var->green.offset	= 8;
 		var->green.length	= 8;
-		var->blue.offset	= 0;
+		var->blue.offset	= 16;
 		var->blue.length	= 8;
 		break;
 

There are some other places where RGB/BGR mode is mentioned/set, but none of them seem to affect the colors. Changing the offsets here work though, I'll tidy up my branch and open a PR tomorrow.

By Henrik Grimler on 2020-12-30T23:34:03

closed