6 comments:
File src/device/oprom/realmode/x86.c:
if (!vbe_mode_info_valid())
return;
This is redundant now. We know it is valid, because we just set it. In
fill_lb_framebuffer(), it was only necessary because we didn't know if
this code ran.
Patch Set #15, Line 372: mode_info.vesa.blue_mask_size);
Passing a struct would really make this less error-prone. I guess we
could even write it like this:
fb_add_framebuffer_info_ex((struct lb_framebuffer){
.bits_per_pixel = modeinfo.vesa.bits_per_pixel,
...
});
File src/drivers/intel/fsp1_1/fsp_gop.c:
Patch Set #15, Line 32: BOOT_STATE_INIT_ENTRY(BS_DEV_INIT, BS_ON_EXIT, fill_framebuffer_info, NULL);
How about calling it from fsp_run_silicon_init() along with the
gfx_set_init_done() call there?
File src/drivers/intel/fsp2_0/graphics.c:
Patch Set #15, Line 119: BOOT_STATE_INIT_ENTRY(BS_DEV_INIT, BS_ON_EXIT, fill_framebuffer_info, NULL);
At end of fsp_silicon_init()?
File src/drivers/intel/gma/hires_fb/gma-gfx_init.adb:
Patch Set #15, Line 35: end fb_add_framebuffer_info;
What's the purpose of this function? Can't we just call the C function
directly, below?
Patch Set #15, Line 106: end if;
Would be nicer to have only one `if` and call the C function directly, e.g.
if (linear_fb_addr /= 0) then
c_fb_add_framebuffer_info
(fb_addr => Interfaces.C.size_t (linear_fb_addr),
x_resolution => word32 (fb.Width),
y_resolution => word32 (fb.Height),
bytes_per_line => word32 (fb.Stride) * 4,
bits_per_pixel => 32);
lightup_ok := 1;
else
lightup_ok := 0;
end if;
To view, visit change 39003. To unsubscribe, or for help writing mail filters, visit settings.