On 09/07/2018 18:54, Dong Wei wrote:

Grant,

In section 4.2, the /Firmware directory is mentioned. What are the reasons why this new hierarchy of directories are created? Can we not use the existing /EFI hierarchy as shown at http://www.uefi.org/registry? I may have missed the rational when discussed.

Because it doesn't contain EFI executables, or have anything to do with the UEFI spec at all. A different hierarchy was selected because it insulates firmware implementation details from the hierarchy used by an OS to install EFI binaries.

If we do need to create a new /Firmware hierarchy, we need to register that with the UEFI Forum as the UEFI Forum owns the ESP name space to avoid conflicts. EBBR cannot single-handedly create new hierarchy.

I'm okay with that. Care to propose suitable language for the EFI spec?

Also, each SoC vendor would need to make request to the UEFI Forum to get their vendor subdirectories under /Firmware as well.

Is that true? I lifted the description of the /FIRMWARE directory straight from UEFI section 13.3.1.3:

"Applications that are loaded by other applications or drivers are not required to be stored in any specific location in the EFI system partition. The choice of the subdirectory name is up to the vendor, but all vendors must pick names that do not collide with any other vendor’s subdirectory name. This applies to system manufacturers, operating system vendors, BIOS vendors, and third party tool vendors, or any other vendor that wishes to install files on an EFI system partition."

Seems to me that the vendors have freedom to chose a name.

Cheers, g.

• DW

-----Original Message----- From: arm.ebbr-discuss-bounces@arm.com arm.ebbr-discuss-bounces@arm.com On Behalf Of Grant Likely Sent: Friday, July 6, 2018 12:13 PM To: boot-architecture@lists.linaro.org; arm.ebbr-discuss arm.ebbr-discuss@arm.com Subject: [Arm.ebbr-discuss] EBBR v0.6-pre1 Released

I've tagged the prerelease in preparation for a wider v0.6 RFC release next week. Please review and comment:

https://github.com/glikely/ebbr/releases/tag/v0.6-pre1

(I've linked to the copy on my personal ebbr fork because I'm having trouble getting Travis-ci to deploy to the official repo. It will take a bit of effort to work out what is wrong)

g. _______________________________________________ Arm.ebbr-discuss mailing list Arm.ebbr-discuss@arm.com

IMPORTANT NOTICE: The contents of this email and any attachments are confidential and may also be privileged. If you are not the intended recipient, please notify the sender immediately and do not disclose the contents to any other person, use it for any purpose, or store or copy the information in any medium. Thank you.

On 09/07/2018 19:47, Dong Wei wrote:

Grant,

We don't need to change anything in the UEFI Spec. All we need to do is to request the addition of /Firmware directory in the registry. You can send the request to the USWG chair. There is a link on that page to send the request.

Okay, I'll do that.

Created a github issue to track the topic:

https://github.com/ARM-software/ebbr/issues/25

g.

Vendors do have the freedom to choose a name, but the name needs to be requested to the USWG chair so that USWG can check whether the name collides with any existing names and whether the request is legit. Sometimes vendors may need to change the proposed name if it collides with the existing one.

• DW

-----Original Message----- From: Grant Likely Sent: Monday, July 9, 2018 11:17 AM To: Dong Wei Dong.Wei@arm.com; boot-architecture@lists.linaro.org; arm.ebbr-discuss arm.ebbr-discuss@arm.com Subject: Re: [Arm.ebbr-discuss] EBBR v0.6-pre1 Released

On 09/07/2018 18:54, Dong Wei wrote:

...

Grant,

In section 4.2, the /Firmware directory is mentioned. What are the reasons why this new hierarchy of directories are created? Can we not use the existing /EFI hierarchy as shown at http://www.uefi.org/registry? I may have missed the rational when discussed.

Because it doesn't contain EFI executables, or have anything to do with the UEFI spec at all. A different hierarchy was selected because it insulates firmware implementation details from the hierarchy used by an OS to install EFI binaries.

...

If we do need to create a new /Firmware hierarchy, we need to register that with the UEFI Forum as the UEFI Forum owns the ESP name space to avoid conflicts. EBBR cannot single-handedly create new hierarchy.

I'm okay with that. Care to propose suitable language for the EFI spec?

...

Also, each SoC vendor would need to make request to the UEFI Forum to get their vendor subdirectories under /Firmware as well.

Is that true? I lifted the description of the /FIRMWARE directory straight from UEFI section 13.3.1.3:

  "Applications that are loaded by other applications or drivers are
  not required to be stored in any specific location in the EFI system
  partition. The choice of the subdirectory name is up to the vendor,
  but all vendors must pick names that do not collide with any other
  vendor’s subdirectory name. This applies to system manufacturers,
  operating system vendors, BIOS vendors, and third party tool
  vendors, or any other vendor that wishes to install files on an EFI
  system partition."

Seems to me that the vendors have freedom to chose a name.

Cheers, g.

...

• DW

-----Original Message----- From: arm.ebbr-discuss-bounces@arm.com arm.ebbr-discuss-bounces@arm.com On Behalf Of Grant Likely Sent: Friday, July 6, 2018 12:13 PM To: boot-architecture@lists.linaro.org; arm.ebbr-discuss arm.ebbr-discuss@arm.com Subject: [Arm.ebbr-discuss] EBBR v0.6-pre1 Released

I've tagged the prerelease in preparation for a wider v0.6 RFC release next week. Please review and comment:

https://github.com/glikely/ebbr/releases/tag/v0.6-pre1

(I've linked to the copy on my personal ebbr fork because I'm having trouble getting Travis-ci to deploy to the official repo. It will take a bit of effort to work out what is wrong)

g. _______________________________________________ Arm.ebbr-discuss mailing list Arm.ebbr-discuss@arm.com

IMPORTANT NOTICE: The contents of this email and any attachments are confidential and may also be privileged. If you are not the intended recipient, please notify the sender immediately and do not disclose the contents to any other person, use it for any purpose, or store or copy the information in any medium. Thank you.

On 11/07/2018 06:33, Udit Kumar wrote:

Hi Grant

Few comments

1/ 1.1 Introduction

...

For example, an Arm A-class embedded networking platform

to

...

For example, an Arm A-class embedded platform

done

2/ 1.2 Guiding Principals

...

EBBR as a specification defines requirements

to

...

EBBR as a specification define requirements

The original is correct grammer. The alternative would be "EBBR as a specification will define requirements...", but that changes the tense of the entire paragraph.

3.1/ 1.3 Scope Please see, if we can remove reference to SBBR here. In case, we wish to put this reference then SBBR does not may about hardware requirement

...

with SBBR having stricter requirements on hardware

to

...

with SBSA having stricter requirements on hardware

I do want to position EBBR as compared to SBBR, particularly because SBBR compliance should always mean also EBBR compliant. I don't ever want a scenario where an SBBR system cannot be EBBR compliant. That would be an unnecessary fork of the ecosystem.

I've changed the whole paragraph to this:

""" This specification is similar to the Arm Server Base Boot Requirements specification [SBBR]_ in that it defines the firmware interface presented to an operating system. SBBR is targeted at the server ecosystem and places strict requirements on the platform to ensure cross vendor interoperability. EBBR on the other hand allows more flexibility to support embedded designs which do not fit within the SBBR model. """

3.2/ 1.3 Scope Below may not be a valid example, SBBR recommends to use sperate storage but does not mandate it. In other discussion Dong Wei said this is left on implementation

...

For example, an embedded system may use a single eMMC storage device to hold both firmware and operating system images

How about this then?

""" For example, a platform that isn't SBBR compliant because the SoC is only supported using Devicetree could be EBBR compliant, but not SBBR compliant. """

4/ 2.4.3 Configuration Tables

...

A UEFI system that complies with this specification may provide the additional tables via the EFI Configuration Table.

Please help with some example tables here.

I don't have any examples to give. The only point of this sentence is that EBBR does not preclude having additional tables populated in the UEFI configuration table, which could be anything.

This language was originally in SBBR if I remember correctly.

...

As stated above, EBBR systems must not provide both ACPI and Devicetree tables at the same time. Systems that support both interfaces must provide a configuration mechanism to select either ACPI or Devicetree

Could we reword this please, first we said, either ACPI or devicetree. Later we are saying selection to be done.

Both are true. A platform can include both ACPI & DT, but only one can be presented to software at a time.

Also at what time, ACPI or devicetree to be selected build time or runtime

Doesn't matter. Implementer can choose. All that matters is that by the time we're running executables only one or the other is presented. The language as written seems clear to me. Can you propose an alternate wording?

5/ FIRMWARE STORAGE Please elaborate ESP at first place, this is used instead of doing this later

...

of the storage used for OS partitions and the ESP

To

...

of the storage used for OS partitions and the ESP (EFI System Partition)

Fixed

6/ 4.1 Partitioning of Shared Storage

...

Warning: A future issue of this specification will remove the MBR

We are putting a requirement on hardware here. Do we want to update boot ROM to understand GPT in future hardware ?

Yes. We want to push future platforms to understand GPT if firmware is being loaded from the same block device or LU as the OS.

7/ 5.2 Required Runtime Services Please add a note for Update capsule too, Similar to Get/Set Time, Update capsule can return EFI_UNSUPPORTED

All of appendix A is probably going to be removed in the next release because it duplicates things that are already in UEFI. Instead, the next release of the spec will probably need to have exceptions to the spec due to U-Boot not being able to implement everything that is required yet.

Regards Udit

...

-----Original Message----- From: arm.ebbr-discuss-bounces@arm.com [mailto:arm.ebbr-discuss- bounces@arm.com] On Behalf Of Grant Likely Sent: Saturday, July 7, 2018 12:43 AM To: boot-architecture@lists.linaro.org; arm.ebbr-discuss <arm.ebbr- discuss@arm.com> Subject: [Arm.ebbr-discuss] EBBR v0.6-pre1 Released

I've tagged the prerelease in preparation for a wider v0.6 RFC release next week. Please review and comment:

https://emea01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub. com%2Fglikely%2Febbr%2Freleases%2Ftag%2Fv0.6- pre1&amp;data=02%7C01%7Cudit.kumar%40nxp.com%7Ccc2fd094bcb9462f79 8908d5e3747e4a%7C686ea1d3bc2b4c6fa92cd99c5c301635%7C0%7C0%7C6366 65011834763051&amp;sdata=or0ngRAQ937f37XgY8vUPFEaX86YigH2a5J122v98 z0%3D&amp;reserved=0

(I've linked to the copy on my personal ebbr fork because I'm having trouble getting Travis-ci to deploy to the official repo. It will take a bit of effort to work out what is wrong)

g. _______________________________________________ Arm.ebbr-discuss mailing list Arm.ebbr-discuss@arm.com

IMPORTANT NOTICE: The contents of this email and any attachments are confidential and may also be privileged. If you are not the intended recipient, please notify the sender immediately and do not disclose the contents to any other person, use it for any purpose, or store or copy the information in any medium. Thank you.