On 09/06/2013 08:31 AM, Tomasz Nowicki wrote:
Values set within all APEI tables have no hardware meaning and are valid combining with kernel hacks. These changes allow only to confirm that APEI tables (HEST, ERST, BERT) are supported by kernel and no issues were observed on arm/arm64 architectures.
Tomasz Nowicki (6): acpi, apei, ghes: Enable APEI firmware first mode by APEI bit. acpi, apei, hed: Add Hardware Error Device (HED) to SB bus. acpi, apei, einj: Fill ACPI tables to enable EINJ driver. acpi, apei: Fix EINJ and HEST tables. acpi, apei, bert: Point to the same error block status memory region as HEST points. acpi, apei, erst: Point to region where we can pretend registers, persistent sotrage.
platforms/exynos5250-arndale.acpi/bert.asl | 4 +- platforms/exynos5250-arndale.acpi/dsdt.asl | 27 +++++ platforms/exynos5250-arndale.acpi/einj.asl | 2 +- platforms/exynos5250-arndale.acpi/erst.asl | 34 +++---- platforms/exynos5250-arndale.acpi/hest.asl | 148 +--------------------------- platforms/foundation-v8.acpi/bert.asl | 4 +- platforms/foundation-v8.acpi/dsdt.asl | 27 +++++ platforms/foundation-v8.acpi/einj.asl | 2 +- platforms/foundation-v8.acpi/erst.asl | 34 +++---- platforms/foundation-v8.acpi/hest.asl | 148 +--------------------------- 10 files changed, 98 insertions(+), 332 deletions(-)
One comment only: in the future, would you mind checking each of these tables and incrementing the OEM Revision number as you make changes? It is not really essential, but it may help in hunting things down in the future. To be fair, I haven't been consistent with this either but I am trying to be.
Otherwise, nicely done.
Acked-by: Al Stone al.stone@linaro.org