[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <64eaba98-e1c9-4667-9825-a64e157d4dfb@intel.com>
Date: Tue, 29 Oct 2024 12:22:31 +1300
From: "Huang, Kai" <kai.huang@...el.com>
To: Dan Williams <dan.j.williams@...el.com>, <dave.hansen@...el.com>,
<kirill.shutemov@...ux.intel.com>, <tglx@...utronix.de>, <bp@...en8.de>,
<peterz@...radead.org>, <mingo@...hat.com>, <hpa@...or.com>,
<seanjc@...gle.com>, <pbonzini@...hat.com>
CC: <x86@...nel.org>, <linux-kernel@...r.kernel.org>, <kvm@...r.kernel.org>,
<rick.p.edgecombe@...el.com>, <isaku.yamahata@...el.com>,
<adrian.hunter@...el.com>, <nik.borisov@...e.com>
Subject: Re: [PATCH v6 08/10] x86/virt/tdx: Reduce TDMR's reserved areas by
using CMRs to find memory holes
>> /*
>> * Start looking for reserved blocks at the
>> * beginning of the TDMR.
>> */
>> prev_end = tdmr->base;
>> - list_for_each_entry(tmb, tmb_list, list) {
>> + for (i = 0; i < sysinfo_cmr->num_cmrs; i++) {
>> u64 start, end;
>>
>> - start = PFN_PHYS(tmb->start_pfn);
>> - end = PFN_PHYS(tmb->end_pfn);
>> + start = sysinfo_cmr->cmr_base[i];
>> + end = start + sysinfo_cmr->cmr_size[i];
>
> This had me go check the inclusive vs exclusive range comparisons. Even
> though it is not in this patch I think tdmr_populate_rsvd_holes() needs
> this fixup:
>
> diff --git a/arch/x86/virt/vmx/tdx/tdx.c b/arch/x86/virt/vmx/tdx/tdx.c
> index 4e2b2e2ac9f9..b5026edf1eeb 100644
> --- a/arch/x86/virt/vmx/tdx/tdx.c
> +++ b/arch/x86/virt/vmx/tdx/tdx.c
> @@ -776,7 +776,7 @@ static int tdmr_populate_rsvd_holes(struct list_head *tmb_list,
> break;
>
> /* Exclude regions before this TDMR */
> - if (end < tdmr->base)
> + if (end <= tdmr->base)
> continue;
>
> /*
>
> ...because a CMR that ends at tdmr->base is still "before" the TDMR.
I think you are right. Thanks for catching this.
But in practice this will not cause any problem because the check right
after it:
/*
* Skip over memory areas that
* have already been dealt with.
*/
if (start <= prev_end) {
prev_end = end;
continue;
}
.. will always be true and effectively skip this region.
So it is just a matter of 'skipping the region one step earlier or later'.
>
> As that's a separate fixup you can add for this patch.
Yeah I agree logically this fixup is needed. I'll send out as a
separate patch and see.
>
> Reviewed-by: Dan Williams <dan.j.williams@...el.com>
Thanks!
Powered by blists - more mailing lists