Re: Xorg startup with serial console

On Thu, Aug 13, 2020 at 8:31 AM Michael <[email protected]> wrote:
>
> Hello,
>
> On Wed, 12 Aug 2020 08:35:24 +0200
> Martin Husemann <[email protected]> wrote:
>
> > On Wed, Aug 12, 2020 at 02:23:31AM -0400, Michael wrote:
> > > Depends on the driver. And I doubt this claim holds up on non-x86.
> >
> > I'd read that claim as: anything with a drikms driver.
> >
> > BTW: after Connors' recent hints I now have in my blade 2500:
> >
> > ppb0 at pci1 dev 2 function 0: Pericom Semiconductors PI7C9X111SL PCIe to PCI Reverse Bridge (rev. 0x02)
> > ppb0: PCI Express capability version 1 <PCI/PCI-X to PCI-E Bridge> x1 @ 2.5GT/s
> > pci2 at ppb0 bus 1
> > pci2: i/o space, memory space enabled
> > genfb0 at pci2 dev 0 function 0: NVIDIA product 128b (rev. 0xa1)
> > NVIDIA product 0e0f (mixed mode multimedia, revision 0xa1) at pci2 dev 0 function 1 not configured
> > [..]
> > radeonfb0 at pci4 dev 2 function 0: ATI Technologies Radeon 7000/VE QY (rev. 0x00)
> > [..]
> > genfb1 at pci3 dev 4 function 0: NVIDIA product 0326 (rev. 0xa1)
> >
> > .. so I might be looking into nouveau and the pmap side of drm. The second
> > nvidia was just for fun - always nice to have a spare real PCI graphics card.
>
> Nice!
> And good luck with drm vs. iommu...
> While you're at it - could you try an xhci with that?
>
> I guess I'll stick a graphics card into the v210 and see what happens.
>
> have fun
> Michael


Hello Martin,

nice that it works for your machine. As i wrote, i have only used the
PCIe to PCI bridge on x86.

Since you reported success to get a non F-Code card booted up and
recognized, i wanted to try a similar thing by plugging a geforce
750ti PCI-E without bridge into my Sun Ultra 25 to see how it would
behave. The system did not pass the power on check and went into a
fault state followed by a reset loop.

Searching again the internet for clues, people are reporting that each
card in a Sun system needs to have Sun F-Code on it or the system will
not boot.

So i wonder, why your blade 2500 is coming up at all?
Might it be because of the bridge that somehow gets accepted by
openboot and the non F-Code card is ignored during this stage (then
later initialized by NetBSD), allowing the system to boot up?



Regards,
Connor