Hi Robin, Thank you for the review. On 8/7/2017 7:17 AM, Robin Murphy wrote: > Hi Neil, > > On 04/08/17 20:59, Neil Leeder wrote: [...] >> + res[1].start = pmcg->base_address + SZ_64K; > > Ugh, I see there's a nasty spec hole here - IORT only defines one base > address, but SMMUv3 says *both* pages have imp-def base addresses. I > guess this assumption of them being in consecutive 64K regions holds for > your platform, but as things stand it doesn't appear possible to rely on > it generally :( > > I'll follow up internally to see if we need to get IORT and/or SBSA > updated or clarified. > Thanks. I'll wait for the outcome before submitting a new patch.
[...] >> diff --git a/include/acpi/actbl2.h b/include/acpi/actbl2.h >> index 707dda74..2169b6f 100644 >> --- a/include/acpi/actbl2.h >> +++ b/include/acpi/actbl2.h > > Hopefully these changes are merely here for reference, but just in case > it needs to be said, they should go as a separate patch via ACPICA > (presumably there's also some corresponding iASL stuff to compile PMCG > nodes in the first place). > Yes, I'll submit this to ACPICA once the IORT addresses get figured out. Neil -- Qualcomm Datacenter Technologies, Inc. as an affiliate of Qualcomm Technologies Inc. Qualcomm Technologies, Inc. is a member of the Code Aurora Forum, a Linux Foundation Collaborative Project.