Similar to the sysfs files for mds and tsx_async_abort, that for
mmio_stale_data also provides the best effort mitigation mode which
invokes the mitigation instructions without a guarantee they clear the
CPU buffers to address virtualized scenarios. Adds description for the
mode in the mmio_stale_data's page to make it self-contained.

Signed-off-by: Takahiro Itazuri <itazur@amazon.com>
---
 .../hw-vuln/processor_mmio_stale_data.rst     | 19 ++++++++++++++++++-
 1 file changed, 18 insertions(+), 1 deletion(-)

diff --git a/Documentation/admin-guide/hw-vuln/processor_mmio_stale_data.rst b/Documentation/admin-guide/hw-vuln/processor_mmio_stale_data.rst
index c98fd1190..c1c96eeb8 100644
--- a/Documentation/admin-guide/hw-vuln/processor_mmio_stale_data.rst
+++ b/Documentation/admin-guide/hw-vuln/processor_mmio_stale_data.rst
@@ -226,7 +226,7 @@ The possible values in this file are:
        - The processor is vulnerable, but no mitigation enabled
      * - 'Vulnerable: Clear CPU buffers attempted, no microcode'
        - The processor is vulnerable, but microcode is not updated. The
-         mitigation is enabled on a best effort basis.
+         mitigation is enabled on a best effort basis. See :ref:`best_effort`.
      * - 'Mitigation: Clear CPU buffers'
        - The processor is vulnerable and the CPU buffer clearing mitigation is
          enabled.
@@ -254,6 +254,23 @@ the above information:
   'SMT Host state unknown'  Kernel runs in a VM, Host SMT state unknown
   ========================  ===========================================
 
+.. _best_effort:
+
+Best effort mitigation mode
+^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+If the processor is vulnerable, but the availability of the microcode-based
+mitigation mechanism is not advertised via CPUID the kernel selects a best
+effort mitigation mode.  This mode invokes the mitigation instructions
+without a guarantee that they clear the CPU buffers.
+
+This is done to address virtualization scenarios where the host has the
+microcode update applied, but the hypervisor is not yet updated to expose the
+CPUID to the guest. If the host has updated microcode the protection takes
+effect; otherwise a few CPU cycles are wasted pointlessly.
+
+The state in the mmio_stale_data sysfs file reflects this situation accordingly.
+
 References
 ----------
 .. [#f1] Affected Processors
-- 
2.40.1

Takahiro Itazuri <itazur@amazon.com> writes:

> Similar to the sysfs files for mds and tsx_async_abort, that for
> mmio_stale_data also provides the best effort mitigation mode which
> invokes the mitigation instructions without a guarantee they clear the
> CPU buffers to address virtualized scenarios. Adds description for the
> mode in the mmio_stale_data's page to make it self-contained.
>
> Signed-off-by: Takahiro Itazuri <itazur@amazon.com>
> ---
>  .../hw-vuln/processor_mmio_stale_data.rst     | 19 ++++++++++++++++++-
>  1 file changed, 18 insertions(+), 1 deletion(-)
>
> diff --git a/Documentation/admin-guide/hw-vuln/processor_mmio_stale_data.rst b/Documentation/admin-guide/hw-vuln/processor_mmio_stale_data.rst
> index c98fd1190..c1c96eeb8 100644
> --- a/Documentation/admin-guide/hw-vuln/processor_mmio_stale_data.rst
> +++ b/Documentation/admin-guide/hw-vuln/processor_mmio_stale_data.rst
> @@ -226,7 +226,7 @@ The possible values in this file are:
>         - The processor is vulnerable, but no mitigation enabled
>       * - 'Vulnerable: Clear CPU buffers attempted, no microcode'
>         - The processor is vulnerable, but microcode is not updated. The
> -         mitigation is enabled on a best effort basis.
> +         mitigation is enabled on a best effort basis. See :ref:`best_effort`.
>       * - 'Mitigation: Clear CPU buffers'
>         - The processor is vulnerable and the CPU buffer clearing mitigation is
>           enabled.
> @@ -254,6 +254,23 @@ the above information:
>    'SMT Host state unknown'  Kernel runs in a VM, Host SMT state unknown
>    ========================  ===========================================
>  
> +.. _best_effort:
> +
> +Best effort mitigation mode
> +^^^^^^^^^^^^^^^^^^^^^^^^^^^
> +
> +If the processor is vulnerable, but the availability of the microcode-based
> +mitigation mechanism is not advertised via CPUID the kernel selects a best
> +effort mitigation mode.  This mode invokes the mitigation instructions
> +without a guarantee that they clear the CPU buffers.
> +
> +This is done to address virtualization scenarios where the host has the
> +microcode update applied, but the hypervisor is not yet updated to expose the
> +CPUID to the guest. If the host has updated microcode the protection takes
> +effect; otherwise a few CPU cycles are wasted pointlessly.
> +
> +The state in the mmio_stale_data sysfs file reflects this situation accordingly.
> +

So the change seems OK, but why not just describe the "best effort" mode
in the place where you refer to it, rather than creating a forward
reference?

Thanks,

jon

Jonathan Corbet <corbet@lwn.net> wrote:
> So the change seems OK, but why not just describe the "best effort" mode
> in the place where you refer to it, rather than creating a forward
> reference?

tbh, I just followed the docs for mds and tsx_async_abort. Of course, I
can do it! Would it better to do the same thing for mds and
tsx_async_abort docs as well?

Best regards,
Takahiro Itazuri