ARM64: kernel: implement ACPI parking protocol
A message that is not in the archive (sent off-list or not yet received): CAPw-ZTmsLqgh6W6R0uLgQjueV7XjaGOD49CYCRYufJ43zT8Kyg@mail.gmail.com
PATCH v2 1/1 ARM64: kernel: implement ACPI parking protocol
Hi Loc,
[CC'ed Jeremy - I know he is testing this on m400]
On Fri, Dec 18, 2015 at 06:12:20PM -0800, Loc Ho wrote:
[...]
Quoted 16 lines
Quoted 6 lines
Quoted 5 lines
I put it together to make sure that it is there if any platform compliant with it relies on it and more importantly to prevent non-compliant implentations from trickling into the kernel, basically to prevent broken firmware implementations from being considered compliant and "working".
Quoted 9 lines
That's a good point. However, AFAIK, AMD are using PSCI already in their firmware implementation and that's what we recommend anyway if EL3 is available.
So this patch is mainly aimed at X-Gene since other implementations I'm aware of have EL3 already. What are the chances of Applied Micro fixing their firmware and actually using this patch? (not a question for you; I'm hoping some Applied engineers would comment).
Quoted 6 lines
Sorry that I missed this. This FW is now fixed and tested with Lorenzo patch. Though some minor manual patch will be required with recent kernel. I was going to get some one to ack the patch but got tied up with other matter. As for availability of the FW, it is currently internally only and will be available sometime next year for the next FW release. If any one interested in the FW, you can contact me.
I guess that by "some minor manual patch" you meant you had to rebase the patch on top of mainline, I can post an updated version for you to test.
Mind giving a tested-by if FW has been updated to comply with the specs and tested with my patch applied ?
Thanks,
Lorenzo
PATCH v2 1/1 ARM64: kernel: implement ACPI parking protocol
Hi,
Quoted 30 lines
[CC'ed Jeremy - I know he is testing this on m400]
On Fri, Dec 18, 2015 at 06:12:20PM -0800, Loc Ho wrote:
[...]
Quoted 16 lines
Quoted 6 lines
Quoted 5 lines
I put it together to make sure that it is there if any platform compliant with it relies on it and more importantly to prevent non-compliant implentations from trickling into the kernel, basically to prevent broken firmware implementations from being considered compliant and "working".
Quoted 9 lines
That's a good point. However, AFAIK, AMD are using PSCI already in their firmware implementation and that's what we recommend anyway if EL3 is available.
So this patch is mainly aimed at X-Gene since other implementations I'm aware of have EL3 already. What are the chances of Applied Micro fixing their firmware and actually using this patch? (not a question for you; I'm hoping some Applied engineers would comment).
Quoted 6 lines
Sorry that I missed this. This FW is now fixed and tested with Lorenzo patch. Though some minor manual patch will be required with recent kernel. I was going to get some one to ack the patch but got tied up with other matter. As for availability of the FW, it is currently internally only and will be available sometime next year for the next FW release. If any one interested in the FW, you can contact me.
Quoted 3 lines
I guess that by "some minor manual patch" you meant you had to rebase the patch on top of mainline, I can post an updated version for you to test.
Yes... It don't apply cleanly with newer kernel.
Quoted 3 lines
Mind giving a tested-by if FW has been updated to comply with the specs and tested with my patch applied ?
Yes... Let me know when an newer version is available.
-Loc
PATCH v2 1/1 ARM64: kernel: implement ACPI parking protocol
Hi Lorenzo,
Quoted 36 lines · Loc Ho
Hi,
Quoted 30 lines
[CC'ed Jeremy - I know he is testing this on m400]
On Fri, Dec 18, 2015 at 06:12:20PM -0800, Loc Ho wrote:
[...]
Quoted 16 lines
Quoted 6 lines
Quoted 5 lines
I put it together to make sure that it is there if any platform compliant with it relies on it and more importantly to prevent non-compliant implentations from trickling into the kernel, basically to prevent broken firmware implementations from being considered compliant and "working".
Quoted 9 lines
That's a good point. However, AFAIK, AMD are using PSCI already in their firmware implementation and that's what we recommend anyway if EL3 is available.
So this patch is mainly aimed at X-Gene since other implementations I'm aware of have EL3 already. What are the chances of Applied Micro fixing their firmware and actually using this patch? (not a question for you; I'm hoping some Applied engineers would comment).
Quoted 6 lines
Sorry that I missed this. This FW is now fixed and tested with Lorenzo patch. Though some minor manual patch will be required with recent kernel. I was going to get some one to ack the patch but got tied up with other matter. As for availability of the FW, it is currently internally only and will be available sometime next year for the next FW release. If any one interested in the FW, you can contact me.
Quoted 3 lines
I guess that by "some minor manual patch" you meant you had to rebase the patch on top of mainline, I can post an updated version for you to test.
Quoted 7 lines
Yes... It don't apply cleanly with newer kernel.
Quoted 3 lines
Mind giving a tested-by if FW has been updated to comply with the specs and tested with my patch applied ?
Yes... Let me know when an newer version is available.
I am following up here. Do you have an ETA on another version against latest 4.5.rc1?
-Loc