[v2] x86/svm: Add support for Bus Lock Threshold

[PATCH v2 3/3] CHANGELOG: Note the new SVM bus-lock intercept

Posted by Alejandro Vallejo 4 days, 17 hours ago

Signed-off-by: Alejandro Vallejo <alejandro.garciavallejo@amd.com>
---
 CHANGELOG.md | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 53d92a2597..07c1745f22 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -9,6 +9,9 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/)
 ### Changed
 
 ### Added
+ - On x86:
+     - AMD bus-lock detect, used by Xen to mitigate (by rate-limiting) the
+       system wide impact of an HVM guest misusing atomic instructions.
 
 ### Removed
  - On x86:
-- 
2.43.0

Re: [PATCH v2 3/3] CHANGELOG: Note the new SVM bus-lock intercept

Posted by Andrew Cooper 4 days, 13 hours ago

On 21/01/2026 2:28 pm, Alejandro Vallejo wrote:
> Signed-off-by: Alejandro Vallejo <alejandro.garciavallejo@amd.com>
> ---
>  CHANGELOG.md | 3 +++
>  1 file changed, 3 insertions(+)
>
> diff --git a/CHANGELOG.md b/CHANGELOG.md
> index 53d92a2597..07c1745f22 100644
> --- a/CHANGELOG.md
> +++ b/CHANGELOG.md
> @@ -9,6 +9,9 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/)
>  ### Changed
>  
>  ### Added
> + - On x86:
> +     - AMD bus-lock detect, used by Xen to mitigate (by rate-limiting) the
> +       system wide impact of an HVM guest misusing atomic instructions.
>  

A little too much copy&paste from the SPR section.  This wants
unindenting by one level.

Also, this text most importantly needs to identify which hardware the
feature exists in.  I've reworked it to:

diff --git a/CHANGELOG.md b/CHANGELOG.md
index 53d92a259731..9c3bf0411ccc 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -9,6 +9,10 @@ The format is based on [Keep a
Changelog](https://keepachangelog.com/en/1.0.0/)
 ### Changed
 
 ### Added
+ - On x86:
+   - Support for Bus Lock Threshold on AMD Zen5 and later CPUs, used by
Xen to
+     mitigate (by rate-limiting) the system wide impact of an HVM guest
+     misusing atomic instructions.
 
 ### Removed
  - On x86:


The internet suggests that it's Zen5 rather than Zen4.

Also I personally put the CHANGELOG update in the commit which finally
implements the feature, rather than doing it after the fact.  This is
more helpful when git blame-ing CHANGELOG to find things.

~Andrew

Re: [PATCH v2 3/3] CHANGELOG: Note the new SVM bus-lock intercept

Posted by Alejandro Vallejo 3 days, 22 hours ago

On Wed Jan 21, 2026 at 7:36 PM CET, Andrew Cooper wrote:
> On 21/01/2026 2:28 pm, Alejandro Vallejo wrote:
>> Signed-off-by: Alejandro Vallejo <alejandro.garciavallejo@amd.com>
>> ---
>>  CHANGELOG.md | 3 +++
>>  1 file changed, 3 insertions(+)
>>
>> diff --git a/CHANGELOG.md b/CHANGELOG.md
>> index 53d92a2597..07c1745f22 100644
>> --- a/CHANGELOG.md
>> +++ b/CHANGELOG.md
>> @@ -9,6 +9,9 @@ The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/)
>>  ### Changed
>>  
>>  ### Added
>> + - On x86:
>> +     - AMD bus-lock detect, used by Xen to mitigate (by rate-limiting) the
>> +       system wide impact of an HVM guest misusing atomic instructions.
>>  
>
> A little too much copy&paste from the SPR section.  This wants
> unindenting by one level.
>
> Also, this text most importantly needs to identify which hardware the
> feature exists in.  I've reworked it to:
>
> diff --git a/CHANGELOG.md b/CHANGELOG.md
> index 53d92a259731..9c3bf0411ccc 100644
> --- a/CHANGELOG.md
> +++ b/CHANGELOG.md
> @@ -9,6 +9,10 @@ The format is based on [Keep a
> Changelog](https://keepachangelog.com/en/1.0.0/)
>  ### Changed
>  
>  ### Added
> + - On x86:
> +   - Support for Bus Lock Threshold on AMD Zen5 and later CPUs, used by
> Xen to
> +     mitigate (by rate-limiting) the system wide impact of an HVM guest
> +     misusing atomic instructions.
>  
>  ### Removed
>   - On x86:
>
>
> The internet suggests that it's Zen5 rather than Zen4.
>
> Also I personally put the CHANGELOG update in the commit which finally
> implements the feature, rather than doing it after the fact.  This is
> more helpful when git blame-ing CHANGELOG to find things.
>
> ~Andrew

Sure, fixing up the prior patch sounds good to me.

Cheers,
Alejandro