[v3] SUPPORT: downgrade Physical CPU Hotplug to Experimental

[PATCH v3] SUPPORT: downgrade Physical CPU Hotplug to Experimental

Posted by Stefano Stabellini 7 months ago

From: Stefano Stabellini <stefano.stabellini@amd.com>

The feature is not commonly used, and we don't have hardware to test it,
not in OSSTest, not in Gitlab, and not even ad-hoc manually by community
members. We could use QEMU to test it, but even that it is known not to
work.

Also take the opportunity to rename the feature to "ACPI CPU Hotplug"
for clarity.

Signed-off-by: Stefano Stabellini <stefano.stabellini@amd.com>
---
Changes in v3:
- add note to CHANGELOG.md
---
 CHANGELOG.md | 2 ++
 SUPPORT.md   | 4 ++--
 2 files changed, 4 insertions(+), 2 deletions(-)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 24636b8eaf..e33cf4e1b1 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -13,6 +13,8 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/)
    known user doesn't use it properly, leading to in-guest breakage.
  - The "dom0" option is now supported on Arm and "sve=" sub-option can be used
    to enable dom0 guest to use SVE/SVE2 instructions.
+ - Physical CPU Hotplug downgraded to Experimental and renamed "ACPI CPU
+   Hotplug" for clarity
 
 ### Added
  - On x86, support for features new in Intel Sapphire Rapids CPUs:
diff --git a/SUPPORT.md b/SUPPORT.md
index 3461f5cf2f..3472b6edfa 100644
--- a/SUPPORT.md
+++ b/SUPPORT.md
@@ -46,9 +46,9 @@ For the Cortex A77 r0p0 - r1p0, see Errata 1508412.
 
 ## Host hardware support
 
-### Physical CPU Hotplug
+### ACPI CPU Hotplug
 
-    Status, x86: Supported
+    Status, x86: Experimental
 
 ### Physical Memory
 
-- 
2.25.1

Re: [PATCH v3] SUPPORT: downgrade Physical CPU Hotplug to Experimental

Posted by Roger Pau Monné 7 months ago

On Wed, Sep 27, 2023 at 04:20:04PM -0700, Stefano Stabellini wrote:
> From: Stefano Stabellini <stefano.stabellini@amd.com>
> 
> The feature is not commonly used, and we don't have hardware to test it,
> not in OSSTest, not in Gitlab, and not even ad-hoc manually by community
> members. We could use QEMU to test it, but even that it is known not to
> work.

I think this last sentence is ambiguous.  QEMU ACPI CPU hotplug
implementation does work AFAIK, it's Xen implementation of ACPI
hotplug that explodes (or so I've been told).  I would drop or reword
the "but even that it is known not to work." part of the sentence.

> 
> Also take the opportunity to rename the feature to "ACPI CPU Hotplug"
> for clarity.
> 
> Signed-off-by: Stefano Stabellini <stefano.stabellini@amd.com>

Acked-by: Roger Pau Monné <roger.pau@citrix.com>

Thanks, Roger.

Re: [PATCH v3] SUPPORT: downgrade Physical CPU Hotplug to Experimental

Posted by Jan Beulich 7 months ago

On 28.09.2023 10:32, Roger Pau Monné wrote:
> On Wed, Sep 27, 2023 at 04:20:04PM -0700, Stefano Stabellini wrote:
>> From: Stefano Stabellini <stefano.stabellini@amd.com>
>>
>> The feature is not commonly used, and we don't have hardware to test it,
>> not in OSSTest, not in Gitlab, and not even ad-hoc manually by community
>> members. We could use QEMU to test it, but even that it is known not to
>> work.
> 
> I think this last sentence is ambiguous.  QEMU ACPI CPU hotplug
> implementation does work AFAIK, it's Xen implementation of ACPI
> hotplug that explodes (or so I've been told).  I would drop or reword
> the "but even that it is known not to work." part of the sentence.

Maybe simply add "... on our end"?

Jan

Re: [PATCH v3] SUPPORT: downgrade Physical CPU Hotplug to Experimental

Posted by Roger Pau Monné 7 months ago

On Thu, Sep 28, 2023 at 11:32:22AM +0200, Jan Beulich wrote:
> On 28.09.2023 10:32, Roger Pau Monné wrote:
> > On Wed, Sep 27, 2023 at 04:20:04PM -0700, Stefano Stabellini wrote:
> >> From: Stefano Stabellini <stefano.stabellini@amd.com>
> >>
> >> The feature is not commonly used, and we don't have hardware to test it,
> >> not in OSSTest, not in Gitlab, and not even ad-hoc manually by community
> >> members. We could use QEMU to test it, but even that it is known not to
> >> work.
> > 
> > I think this last sentence is ambiguous.  QEMU ACPI CPU hotplug
> > implementation does work AFAIK, it's Xen implementation of ACPI
> > hotplug that explodes (or so I've been told).  I would drop or reword
> > the "but even that it is known not to work." part of the sentence.
> 
> Maybe simply add "... on our end"?

WFM.

Thanks, Roger.

Re: [PATCH v3] SUPPORT: downgrade Physical CPU Hotplug to Experimental

Posted by Stefano Stabellini 7 months ago

On Thu, 28 Sep 2023, Roger Pau Monné wrote:
> On Thu, Sep 28, 2023 at 11:32:22AM +0200, Jan Beulich wrote:
> > On 28.09.2023 10:32, Roger Pau Monné wrote:
> > > On Wed, Sep 27, 2023 at 04:20:04PM -0700, Stefano Stabellini wrote:
> > >> From: Stefano Stabellini <stefano.stabellini@amd.com>
> > >>
> > >> The feature is not commonly used, and we don't have hardware to test it,
> > >> not in OSSTest, not in Gitlab, and not even ad-hoc manually by community
> > >> members. We could use QEMU to test it, but even that it is known not to
> > >> work.
> > > 
> > > I think this last sentence is ambiguous.  QEMU ACPI CPU hotplug
> > > implementation does work AFAIK, it's Xen implementation of ACPI
> > > hotplug that explodes (or so I've been told).  I would drop or reword
> > > the "but even that it is known not to work." part of the sentence.
> > 
> > Maybe simply add "... on our end"?
> 
> WFM.

Just to be clear we are going for:

"The feature is not commonly used, and we don't have hardware to test it,
not in OSSTest, not in Gitlab, and not even ad-hoc manually by community
members. We could use QEMU to test it, but even that it is known not to
work on our end."

Can the change be done on commit?

Re: [PATCH v3] SUPPORT: downgrade Physical CPU Hotplug to Experimental

Posted by Roger Pau Monné 6 months, 4 weeks ago

On Thu, Sep 28, 2023 at 01:48:59PM -0700, Stefano Stabellini wrote:
> On Thu, 28 Sep 2023, Roger Pau Monné wrote:
> > On Thu, Sep 28, 2023 at 11:32:22AM +0200, Jan Beulich wrote:
> > > On 28.09.2023 10:32, Roger Pau Monné wrote:
> > > > On Wed, Sep 27, 2023 at 04:20:04PM -0700, Stefano Stabellini wrote:
> > > >> From: Stefano Stabellini <stefano.stabellini@amd.com>
> > > >>
> > > >> The feature is not commonly used, and we don't have hardware to test it,
> > > >> not in OSSTest, not in Gitlab, and not even ad-hoc manually by community
> > > >> members. We could use QEMU to test it, but even that it is known not to
> > > >> work.
> > > > 
> > > > I think this last sentence is ambiguous.  QEMU ACPI CPU hotplug
> > > > implementation does work AFAIK, it's Xen implementation of ACPI
> > > > hotplug that explodes (or so I've been told).  I would drop or reword
> > > > the "but even that it is known not to work." part of the sentence.
> > > 
> > > Maybe simply add "... on our end"?
> > 
> > WFM.
> 
> Just to be clear we are going for:
> 
> "The feature is not commonly used, and we don't have hardware to test it,
> not in OSSTest, not in Gitlab, and not even ad-hoc manually by community
> members. We could use QEMU to test it, but even that it is known not to
> work on our end."
> 
> Can the change be done on commit?

LGTM.

Feel free to add my Reviewed-by: Roger Pau Monné
<roger.pau@citrix.com>.

Thanks.

Re: [PATCH v3] SUPPORT: downgrade Physical CPU Hotplug to Experimental

Posted by Henry Wang 7 months ago

Hi Stefano,

> On Sep 28, 2023, at 07:20, Stefano Stabellini <sstabellini@kernel.org> wrote:
> 
> From: Stefano Stabellini <stefano.stabellini@amd.com>
> 
> The feature is not commonly used, and we don't have hardware to test it,
> not in OSSTest, not in Gitlab, and not even ad-hoc manually by community
> members. We could use QEMU to test it, but even that it is known not to
> work.
> 
> Also take the opportunity to rename the feature to "ACPI CPU Hotplug"
> for clarity.
> 
> Signed-off-by: Stefano Stabellini <stefano.stabellini@amd.com>

Release-acked-by: Henry Wang <Henry.Wang@arm.com>

Kind regards,
Henry