Re: [PATCH v4 0/7] Rework x86 page table walks

[Don Porter]

On 7/23/24 23:39, Richard Henderson wrote:
> Hi Don.
>
> In addition to the other issues, this really needs to be broken up 
> into many more patches.
>
> Every patch should do *one* thing:
>
>   - Code motion
>   - Introduce an API
>   - Introduce target-specific support for an API
>   - Use an API to implement a monitor command
>   - etc
>
> Patch 3, 'Add an "info pg" command ...' attempts to do all of these at 
> once, and so is at least 2000 lines too long, which makes the whole 
> thing extraordinarily hard to review.
>
>
> r~

Hi Richard (and others),

I just wanted to acknowledge that this and the other feedback makes 
sense and I can apply them and send a new version, with the exception of 
what to do about KVM guest register state for Intel VT.

Thanks,

Don