Re: [PATCH v6 12/21] x86/virt/tdx: Add placeholder to construct TDMRs to cover all TDX memory regions

[Huang, Kai]

On Fri, 2022-10-28 at 02:21 +0000, Huang, Kai wrote:
> On Thu, 2022-10-27 at 08:31 -0700, Andi Kleen wrote:
> > > +/* Calculate the actual TDMR_INFO size */
> > > +static inline int cal_tdmr_size(void)
> > > +{
> > > +	int tdmr_sz;
> > > +
> > > +	/*
> > > +	 * The actual size of TDMR_INFO depends on the maximum number
> > > +	 * of reserved areas.
> > > +	 */
> > > +	tdmr_sz = sizeof(struct tdmr_info);
> > > +	tdmr_sz += sizeof(struct tdmr_reserved_area) *
> > > +		   tdx_sysinfo.max_reserved_per_tdmr;
> > 
> > 
> > would seem safer to have a overflow check here.
> > 
> > 
> 
> How about below?
> 
> --- a/arch/x86/virt/vmx/tdx/tdx.c
> +++ b/arch/x86/virt/vmx/tdx/tdx.c
> @@ -614,6 +614,14 @@ static inline int cal_tdmr_size(void)
>         tdmr_sz += sizeof(struct tdmr_reserved_area) *
>                    tdx_sysinfo.max_reserved_per_tdmr;
>  
> +       /*
> +        * Do simple check against overflow, and return 0 (invalid)
> +        * TDMR_INFO size if it happened.  Also WARN() as it should
> +        * should never happen in reality.
> +        */
> +       if (WARN_ON_ONCE(tdmr_sz < 0))
> +               return 0;
> +
>         /*
>          * TDX requires each TDMR_INFO to be 512-byte aligned.  Always
>          * round up TDMR_INFO size to the 512-byte boundary.
> @@ -623,19 +631,27 @@ static inline int cal_tdmr_size(void)
>  
>  static struct tdmr_info *alloc_tdmr_array(int *array_sz)
>  {
> +       int sz;
> +
>         /*
>          * TDX requires each TDMR_INFO to be 512-byte aligned.
>          * Use alloc_pages_exact() to allocate all TDMRs at once.
>          * Each TDMR_INFO will still be 512-byte aligned since
>          * cal_tdmr_size() always return 512-byte aligned size.
>          */
> -       *array_sz = cal_tdmr_size() * tdx_sysinfo.max_tdmrs;
> +       sz = cal_tdmr_size() * tdx_sysinfo.max_tdmrs;
> +
> +       /* Overflow */
> +       if (!sz || WARN_ON_ONCE(sz < 0))
> +               return NULL;
> +
> +       *array_sz = sz;
>  
>         /*
>          * Zero the buffer so 'struct tdmr_info::size' can be
>          * used to determine whether a TDMR is valid.
>          */
> -       return alloc_pages_exact(*array_sz, GFP_KERNEL | __GFP_ZERO);
> +       return alloc_pages_exact(sz, GFP_KERNEL | __GFP_ZERO);
>  }
> 
> 
> Btw, should I use alloc_contig_pages() instead of alloc_pages_exact() as IIUC
> the latter should fail if the size is larger than 4MB?  In reality, the entire
> array only takes dozens of KBs, though.

Hi Andi,

Could you take a look whether this is OK?

Also could you take a look my replies to your other comments?

Thanks!