On Tue, Jul 25, 2023 at 07:34:39PM +0100, Abdellatif El Khlifi wrote:
Hi Tom,
On Tue, Jul 25, 2023 at 09:47:24AM -0400, Tom Rini wrote:
On Tue, Jul 25, 2023 at 10:26:16AM +0100, Abdellatif El Khlifi wrote:
Hi Tom,
On Mon, Jul 24, 2023 at 07:50:07PM -0400, Tom Rini wrote:
On Thu, Jul 13, 2023 at 02:28:37PM +0100, Abdellatif El Khlifi wrote:
Adding support for Arm FF-A v1.0 (Arm Firmware Framework for Armv8-A) [A].
FF-A specifies interfaces that enable a pair of software execution environments aka partitions to communicate with each other. A partition could be a VM in the Normal or Secure world, an application in S-EL0, or a Trusted OS in S-EL1.
FF-A is a discoverable bus and similar to architecture features. FF-A bus is discovered using ARM_SMCCC_FEATURES mechanism performed by the PSCI driver.
...
Changelog of changes:
v15:
So, this does not pass CI: https://source.denx.de/u-boot/u-boot/-/jobs/662303 https://source.denx.de/u-boot/u-boot/-/jobs/662304#L2234
And it's on me for merging this before letting the CI run I kicked complete, sorry. But I'm reverting this now.
The v15 patchset has been built successfully on next using this SHA: [1].
The tested platforms are the following: [2].
Is there any action from my side please ?
[2]: tested platforms
sandbox64 sandbox corstone1000-fvp corstone1000-mps3 qemu_arm64
Yes, you need to go and fix all of the problems that CI is showing:
- On 64bit, espresso7420 is building cmd/armffa.c and pointing out an unused variable.
- Maybe related to why that platform is building the code, a ton of 32bit platforms are building the code and showing warnings. And also that your Kconfig logic is wrong.
- Finally, the MAINTAINERS file needs to be alphabetical.
Thanks for the details. I fixed all the issues and will send them shortly in v16 (a minor change).
espresso7420 builds fine and Arm 32-bit boards no longer automatically pick up FF-A.
Thanks.
For the alphabetical order in MAINTAINERS, please help me undertand where the issue is. The FF-A part is already in alphabetical order:
ARM CORTINA ACCESS CAxxxx
The UUID test portion was not, however.