Re: [PATCH v2 2/2] drivers: firmware: efi: install new fdt in configuration table

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Tue, 11 Dec 2018 at 13:29, Pankaj Bansal <pankaj.bansal@xxxxxxx> wrote:
>
>
>
> > -----Original Message-----
> > From: Ard Biesheuvel [mailto:ard.biesheuvel@xxxxxxxxxx]
> > Sent: Tuesday, December 11, 2018 5:55 PM
> > To: Pankaj Bansal <pankaj.bansal@xxxxxxx>
> > Cc: Bhupesh Sharma <bhsharma@xxxxxxxxxx>; Mark Rutland
> > <mark.rutland@xxxxxxx>; Leif Lindholm <leif.lindholm@xxxxxxxxxx>; Grant
> > Likely <grant.likely@xxxxxxx>; Varun Sethi <V.Sethi@xxxxxxx>; Udit Kumar
> > <udit.kumar@xxxxxxx>; linux-efi <linux-efi@xxxxxxxxxxxxxxx>
> > Subject: Re: [PATCH v2 2/2] drivers: firmware: efi: install new fdt in
> > configuration table
> >
> > On Tue, 11 Dec 2018 at 13:22, Pankaj Bansal <pankaj.bansal@xxxxxxx> wrote:
> > > > -----Original Message-----
> > > > From: Bhupesh Sharma [mailto:bhsharma@xxxxxxxxxx]
> > > > Sent: Tuesday, December 11, 2018 4:25 PM
> > > > To: Pankaj Bansal <pankaj.bansal@xxxxxxx>
> > > > Cc: Ard Biesheuvel <ard.biesheuvel@xxxxxxxxxx>; Mark Rutland
> > > > <mark.rutland@xxxxxxx>; Leif Lindholm <leif.lindholm@xxxxxxxxxx>;
> > > > Grant Likely <grant.likely@xxxxxxx>; Varun Sethi <V.Sethi@xxxxxxx>;
> > > > Udit Kumar <udit.kumar@xxxxxxx>; linux-efi@xxxxxxxxxxxxxxx
> > > > Subject: Re: [PATCH v2 2/2] drivers: firmware: efi: install new fdt
> > > > in configuration table
> > > >
> > > > Hi Pankaj,
> > > >
> > ..
> > > > Can you please share an example, as the above description is not
> > > > very clear to me. May be you can include a dt property that you are
> > > > trying to fix via kernel and what happens in the kernel driver when it is not in
> > a expected state.
> > >
> > > We already do. The "status = "okay";" or "status = "disabled";" is added to the
> > device node in dts file.
> > > Based on this the device structure is created or not created in kernel when
> > booting.
> > >
> > > >
> > > > Also may be you can share why the boot firmware is not able to set a
> > > > correct state of the same.
> > >
> > > The correct state of device would depend on the user supplied parameters and
> > boot time configuration.
> > > Boot firmware is able to set the "status" in fdt file in exit boot services.
> >
> > But why not before? Why does it have to wait until ExitBootServices() to do this?
>
> We attempt to apply the user supplied parameters in ExitBootServices.

What does 'user supplied' mean? And why can't you apply them earlier?

> If it fails, then the device state is un deterministic. If it passed, then device can be used in kernel.
> Once there parameters are applied, regardless of they failed or passed, the boot firmware cannot use the device.
> So we have no choice but to apply these parameters when we no longer wish to use the device in boot firmware.
>

This is incorrect. Setting the DT status property does absolutely
nothing until long after ExitBootServices() completes. So if you want
to set the device status, you need to do it before invoking the
kernel.



[Index of Archives]     [Linux ARM Kernel]     [Linux ARM]     [Linux Omap]     [Fedora ARM]     [IETF Annouce]     [Security]     [Bugtraq]     [Linux OMAP]     [Linux MIPS]     [ECOS]     [Asterisk Internet PBX]     [Linux API]

  Powered by Linux