BUG: Bad page state in process swapper on new imx8qm board

[Oliver Graute]

On 17/10/19, Peng Fan wrote:
> 
> 
> > -----Original Message-----
> > From: Oliver Graute <oliver.graute at gmail.com>
> > Sent: 2019年10月17日 15:34
> > To: kernelnewbies at kernelnewbies.org; Aisheng Dong
> > <aisheng.dong at nxp.com>; Peng Fan <peng.fan at nxp.com>
> > Subject: Re: BUG: Bad page state in process swapper on new imx8qm board
> > 
> > On 16/10/19, Oliver Graute wrote:
> > > Hello list,
> > >
> > > I try to bootup up a new imx8qm congatec board and I have written a
> > > dts file for it and applied some imx8qm related patches which are not
> > > mainline yet but working fine on another imx8qm board (same cpu,
> > > different board vendor).
> > >
> > > The Kernel starts to boot. But unfortunately its stucked after few
> > > seconds with a lot memory/swapper issues.
> > >
> > > BUG: Bad page state in process swapper
> > >
> > > Some clue what`s going on here?
> > 
> > So after some more digging I assume that this error is related to a missing
> > "reserved-memory" node in my devicetree. Now I need to find out how to
> > split up my memory the right way for this imx8qm congatec board.
> 
> Which uboot release are you using? Did you include M4 image in flash.bin?

I'am using the u-boot master branch with my board patches. Together with
the M4 Image. This particular problem is solved now by adding the right
reserved-memory statement and linux,cma {} statement.

No more kernel crashes yet. Now the kernel starts systemd and is stucked
there. I get no login prompt. It looks for me like getty can't find any
serial to attach. But the kernel is still running and writing to serial
lpuart32.

Best regards,

Oliver