[edk2-devel] [PATCH v2 0/2] Add support for scanning Option ROMs

Tue Nov 10 15:01:51 UTC 2020

Hi,

> If the assumption in this env is that PCI enumeration has already been
done, would it make

more sense for the component that performed PCI enumeration to collect the
PCI Option ROM

contents and pass them into the UefiPayLoadPkg in memory buffers and have
the PCI Platform Protocol provide the PCI Option ROM images from these
memory buffers?

We did not plan to add extra memory buffers to pass PCI Option ROMs. Is
there any benefit to creating it?

> In your UefiPayloadPkg change, you have an assumption that the option ROM BAR
still contain the correct base address. Is it always true in real world? Do
you think maybe the boot loader should supply a MMIO space range so
UeifPayloadPkg can use for option rom decoding?

We could update UefiPayloadPkg/BuildAndIntegrationInstructions.txt and make
this a requirement for the boot loader to make all option ROM BARs contain
a valid base address.

Thanks,
Marcello

On Wed, Sep 23, 2020 at 1:58 AM Kinney, Michael D <
michael.d.kinney at intel.com> wrote:

> Marcello,
>
>
>
> If the assumption in this env is that PCI enumeration has already been
> done, would it make
>
> more sense for the component that performed PCI enumeration to collect the
> PCI Option ROM
>
> contents and pass them into the UefiPayLoadPkg in memory buffers and have
> the PCI Platform Protocol
>
> provide the PCI Option ROM images from these memory buffers?
>
>
>
> Thanks,
>
>
>
> Mike
>
>
>
> *From:* devel at edk2.groups.io <devel at edk2.groups.io> *On Behalf Of *Marcello
> Sylvester Bauer
> *Sent:* Monday, September 21, 2020 6:25 AM
> *To:* Ni, Ray <ray.ni at intel.com>
> *Cc:* devel at edk2.groups.io
> *Subject:* Re: [edk2-devel] [PATCH v2 0/2] Add support for scanning
> Option ROMs
>
>
>
> hi Ray,
>
> > I guess it is because the PciIoDevice->AllOpRomProcessed is set which
> causes GetPciRom() is skipped.
>
> exactly. Therefore we need this for the gPciPlatformProtocol
> implementation to work.
> I will add this information as a code comment.
>
> > In your UefiPayloadPkg change, you have an assumption that the option
> ROM BAR still contain the correct base address.
> > Is it always true in real world? Do you think maybe the boot loader
> should supply a MMIO space range so UeifPayloadPkg
> > can use for option rom decoding?
>
> Every option ROM in coreboot does have it's own memory space and can not
> be provided as MMIO space range. Probably we could
> add some kind of validation to verify the validity of the base address.
>
> thanks,
> Marcello
>
>
>
> On Wed, Sep 16, 2020 at 10:28 AM Ni, Ray <ray.ni at intel.com> wrote:
>
> Why running it will disable the ability of PciPlatform code to scan for
> ROMs?
>
> I guess it is because the PciIoDevice->AllOpRomProcessed is set which
> causes GetPciRom() is skipped.
>
> Can you explain more in the code comment?
>
> > -----Original Message-----
> > From: devel at edk2.groups.io <devel at edk2.groups.io> On Behalf Of Marcello
> > Sylvester Bauer
> > Sent: Tuesday, September 15, 2020 8:26 PM
> > To: devel at edk2.groups.io
> > Subject: [edk2-devel] [PATCH v2 0/2] Add support for scanning Option ROMs
> >
> > Fix Option ROM enumeration and support scanning.
> >
> > v2:
> > * add correct Maintainer and Reviewer to Cc
> > * PciPlatformDxe:
> >   - Update description
> >   - add function description
> >
> > Branch: https://github.com/9elements/edk2-1/tree/UefiPayloadPkg-
> > Option_ROMs
> > PR: https://github.com/tianocore/edk2/pull/926
> >
> > Patrick Rudolph (2):
> >   MdeModulePkg: Fix OptionROM scanning
> >   UefiPayloadPkg: Scan for Option ROMs
> >
> >  UefiPayloadPkg/UefiPayloadPkgIa32.dsc                 |   1 +
> >  UefiPayloadPkg/UefiPayloadPkgIa32X64.dsc              |   1 +
> >  UefiPayloadPkg/UefiPayloadPkg.fdf                     |   1 +
> >  UefiPayloadPkg/PciPlatformDxe/PciPlatformDxe.inf      |  46 +++
> >  UefiPayloadPkg/PciPlatformDxe/PciPlatformDxe.h        |  19 +
> >  MdeModulePkg/Bus/Pci/PciBusDxe/PciEnumeratorSupport.c |  10 +-
> >  UefiPayloadPkg/PciPlatformDxe/PciPlatformDxe.c        | 426
> > ++++++++++++++++++++
> >  7 files changed, 500 insertions(+), 4 deletions(-)
> >  create mode 100644 UefiPayloadPkg/PciPlatformDxe/PciPlatformDxe.inf
> >  create mode 100644 UefiPayloadPkg/PciPlatformDxe/PciPlatformDxe.h
> >  create mode 100644 UefiPayloadPkg/PciPlatformDxe/PciPlatformDxe.c
> >
> > --
> > 2.28.0
> >
> >
> >
>
>
>
>
> --
>
> *[Marcello Sylvester Bauer]*
>
>
>
> 9elements Agency GmbH, Kortumstraße 19-21, 44787 Bochum, Germany
>
> Email:  [DEINE EMAIL ADDRESSE]
> <https://static.9elements.com/email_signatur.html>
>
> Phone:  *+49 234 68 94 188 <+492346894188>*
>
> Mobile:  *+49 1722847618 <+491722847618>*
>
>
>
> Sitz der Gesellschaft: Bochum
>
> Handelsregister: Amtsgericht Bochum, HRB 17519
>
> Geschäftsführung: Sebastian Deutsch, Eray Basar
>
>
> Datenschutzhinweise nach Art. 13 DSGVO <https://9elements.com/privacy>
>
> 
>

-- 
*[Marcello Sylvester Bauer]*

9elements Agency GmbH, Kortumstraße 19-21, 44787 Bochum, Germany
Email:  [DEINE EMAIL ADDRESSE]
<https://static.9elements.com/email_signatur.html>
Phone:  *+49 234 68 94 188 <+492346894188>*
Mobile:  *+49 1722847618 <+491722847618>*

Sitz der Gesellschaft: Bochum
Handelsregister: Amtsgericht Bochum, HRB 17519
Geschäftsführung: Sebastian Deutsch, Eray Basar

Datenschutzhinweise nach Art. 13 DSGVO <https://9elements.com/privacy>

-=-=-=-=-=-=-=-=-=-=-=-
Groups.io Links: You receive all messages sent to this group.
View/Reply Online (#67230): https://edk2.groups.io/g/devel/message/67230
Mute This Topic: https://groups.io/mt/76863515/1813853
Group Owner: devel+owner at edk2.groups.io
Unsubscribe: https://edk2.groups.io/g/devel/unsub [edk2-devel-archive at redhat.com]
-=-=-=-=-=-=-=-=-=-=-=-

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listman.redhat.com/archives/edk2-devel-archive/attachments/20201110/2e229aea/attachment.htm>