git: 3630506b9dae - main - loader: implement framebuffer console

[Dan Mack]

On Wed, 6 Jan 2021, Cy Schubert wrote:

> In message <alpine.GSO.2.20.2101061138230.66723 at localhost.local>, Dan Mack
> writ
> es:
>> On Wed, 6 Jan 2021, Dan Mack wrote:
>>
>>> On Wed, 6 Jan 2021, Cy Schubert wrote:
>>>
>>>> In message <51035B54-FF30-41EE-BC4A-3D4E0ABD1B13 at me.com>, Toomas Soome
>>>> writes:
>>>>>
>>>>>
>>>>>> On 6. Jan 2021, at 12:33, Peter Jeremy <peter at rulingia.com> wrote:
>>>>>>
>>>>>> On 2021-Jan-02 20:09:25 +0000, Toomas Soome <tsoome at FreeBSD.org> wrote:
>>>>>>> The branch main has been updated by tsoome:
>>>>>>>
>>>>>>> URL:
>>>>>>> https://cgit.FreeBSD.org/src/commit/?id=3630506b9daec9167a89bc4525638
>>>>> ea45a00769e
>>>>>>
>>>>>> Note that this commit adds a hard build dependency on vtfontcvt(8)
>>>>>> but that tool is depends on the WITH_VT build option.  I suspect
>>>>>> that stand/fonts should be likewise dependent on WITH_VT
>>>>>>
>>>>>> --
>>>>>> Peter Jeremy
>>>>>
>>>>>
>>>>> Thanks for pointing this out. We actually need to build vtfontcvt
>>>>> uncondition
>>>>> ally, in addition to font files, we also need it to generate C code for
>>>>> built
>>>>>  in font (used when there are no /boot/fonts).
>>>>
>>>> Is this why my consoles are blank during boot? My laptop is blank during
>>>> boot until X starts. My three machines downstairs don't run X and remain
>>>> blank. There is no way to enable the display.
>>>>
>>
>>
>> Sorry my reply got garbled ...  My console is an hdmi connected monitor.
>> The display works until FreeBSD boots and then it's just blank.  Something
>> is broke; I am doing full buildworld / buildkernel (GENERIC builds) from
>> the tip of main.
>
> All four machines here exhibit the same problem. Some are VGA connected, others HDMI. Each has a different video card. The three downstairs are servers, no X, using the video cards in text mode. The laptop has an HD3000, uses X, X works but boot messages, like the three downstairs are black holes.
>
> I suspect it's some commit from Jan 5 as messages did display then. This morning, nothing. The puzzling thing is that the commits between Jan 5 and now look innocuous.

I noticed that commit babda0952f8355a89b3241d5e8943c7da0fa4f6b has some 
more framebuffer related changes so I'm rebuilding world with the latest 
to see if anything changes.

Dan