Maturity Of Linux Frame Buffer Backend

I have been developing some basic programs that use the Linux Frame Buffer as the backend, and found the results to fall short of expectations. Material theme isn’t being automatically applied, all text isn’t legible, text isn’t aligned properly, and all graphics have an EGA resolution (very pixilated) look to them which is horrible.

  1. What testing has been done with the Linux Frame Buffer backend?
  2. Does the Linux Frame Buffer backend have feature parity with other backends?
  3. Where is the documentation covering the Linux Frame Buffer backend?
  4. What code samples are available that use the Linux Frame Buffer backend?

Very disappointing to see that the PC Simulator is being recommended for Embedded Linux testing, which isn’t suitable for the following reasons:

  • All results in the PC Simulator aren’t similar to results on an Embedded Linux device (eg Raspberry Pi)
  • Some Embedded Linux devices are inexpensive, and can be easily used for testing (eg Raspberry Pi)
  • PC Simulator doesn’t accurately reflect a real Embedded Linux environment
  • Cannot do remote testing using a real Embedded Linux device (eg Raspberry Pi)
  • The environment that the PC simulator uses heavily relies on XOrg which is a different backend from the Linux Frame Buffer backend

What comes into question is the maturity of the Linux Frame Buffer backend. Some urgent work is required to sort out some significant issues with the Linux Frame Buffer backend. If this cannot be done in a reasonable period of time then it is best to put the Linux Frame Buffer backend on hold, and remove some text from the website on supporting the Linux Frame Buffer as a backend.

@kisvegabor - I have submitted an issue in the LVGL Issue Tracker on the pixelated graphics displayed with the Linux Frame Buffer.

Hi,

I commented in the GitHub issue. It should answer most of the questions here too.

Is there a sample project that targets the Raspberry Pi where a Computer Monitor is used as the display? What configuration settings need to be used when the Raspberry Pi video driver is used? Is there any special flushing logic that needs to be used to deal with the quirks of the Raspberry Pi video driver (eg to avoid graphics corruption similar to what is shown here)?

I’ve tried it a few years ago, with the SDL driver and it worked well.
I can’t comment on the fbdev on RP :frowning: