On 6 July 2016 at 15:04, Ard Biesheuvel ard.biesheuvel@linaro.org wrote:
(+ Ryan)
On 6 July 2016 at 15:29, Leif Lindholm leif.lindholm@linaro.org wrote:
On Wed, Jul 06, 2016 at 01:13:15PM +0200, Ard Biesheuvel wrote:
On 6 July 2016 at 12:39, Graeme Gregory graeme.gregory@linaro.org wrote:
On Wed, Jul 06, 2016 at 11:55:22AM +0200, Ard Biesheuvel wrote:
On 5 July 2016 at 22:18, Graeme Gregory graeme.gregory@linaro.org wrote:
.asl files can end up being modified to later versions of spec when FADT is tied to an earlier version. This can cause builds to break unexpectedly on tools upgrades. Convert to aslc to fix the version of ACPI in use for this platform to 6.1
Contributed-under: TianoCore Contribution Agreement 1.0 Signed-off-by: Graeme Gregory graeme.gregory@linaro.org
Platforms/ARM/VExpress/AcpiTables/AcpiTables.inf | 43 +++++++ Platforms/ARM/VExpress/AcpiTables/Dbg2.asl | 72 ++++++++++++
As discussed off line, this file is not referenced by the .inf so it can be removed.
Platforms/ARM/VExpress/AcpiTables/Dsdt.asl | 144 +++++++++++++++++++++++ Platforms/ARM/VExpress/AcpiTables/Fadt.aslc | 86 ++++++++++++++ Platforms/ARM/VExpress/AcpiTables/FvpPlatform.h | 46 ++++++++ Platforms/ARM/VExpress/AcpiTables/Gtdt.aslc | 92 +++++++++++++++ Platforms/ARM/VExpress/AcpiTables/Madt.aslc | 91 ++++++++++++++
It looks like the MADT describes a GICv3 unconditionally? This is fine, and my intention was to build runtime detection on top of these patches. But we should probably mention the v2/v3 change in the commit log, since the code it replaces describes GICv2 only.
I can make both changes when committing, if necessary.
That is fine with me if there are no other changes. I based the files off your patch to go to GICv3 and just forgot to mention it.
No, other than that the patches look fine to me, and they are working as expected afaict So unless Leif has any comments, I'll proceed and commit them (with the above changes applied)
If you're happy to do the changes, I'm happy for it to go in.
OK, I will push them later today or tomorrow.
@Ryan: as a head's up, this series will require Graeme's patch
6ff71a134f5b EmbeddedPkg/AcpiLib: add GICC table init macro for ACPI 6.0
on the EDK2 side, which I committed this morning.
Thanks for the heads up, knowing that will save me some trouble.
Thanks, Ard.