From: Yan Zhao <yan.y.zhao@intel.com>
To: Sagi Shahar <sagis@google.com>
Cc: <pbonzini@redhat.com>, <seanjc@google.com>,
<linux-kernel@vger.kernel.org>, <kvm@vger.kernel.org>,
<x86@kernel.org>, <rick.p.edgecombe@intel.com>,
<dave.hansen@intel.com>, <kas@kernel.org>, <tabba@google.com>,
<ackerleytng@google.com>, <quic_eberman@quicinc.com>,
<michael.roth@amd.com>, <david@redhat.com>,
<vannapurve@google.com>, <vbabka@suse.cz>,
<thomas.lendacky@amd.com>, <pgonda@google.com>,
<zhiquan1.li@intel.com>, <fan.du@intel.com>, <jun.miao@intel.com>,
<ira.weiny@intel.com>, <isaku.yamahata@intel.com>,
<xiaoyao.li@intel.com>, <binbin.wu@linux.intel.com>,
<chao.p.peng@intel.com>
Subject: Re: [RFC PATCH v2 18/23] x86/virt/tdx: Do not perform cache flushes unless CLFLUSH_BEFORE_ALLOC is set
Date: Tue, 12 Aug 2025 14:37:40 +0800 [thread overview]
Message-ID: <aJrhNFmLFBOP2TVK@yzhao56-desk.sh.intel.com> (raw)
In-Reply-To: <CAAhR5DEZZfX0=9QwBrXhC+1fp1Z0w4Xbb3mXcn0OuW+45tsLwA@mail.gmail.com>
On Mon, Aug 11, 2025 at 04:10:41PM -0500, Sagi Shahar wrote:
> On Thu, Aug 7, 2025 at 4:47 AM Yan Zhao <yan.y.zhao@intel.com> wrote:
> >
> > From: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
> >
> > The TDX module enumerates with a TDX_FEATURES0 bit if an explicit cache
> > flush is necessary when switching KeyID for a page, like before
> > handing the page over to a TD.
> >
> > Currently, none of the TDX-capable platforms have this bit enabled.
> >
> > Moreover, cache flushing with TDH.PHYMEM.PAGE.WBINVD fails if
> > Dynamic PAMT is active and the target page is not 4k. The SEAMCALL only
> > supports 4k pages and will fail if there is no PAMT_4K for the HPA.
I actually couldn't observe this failure in my side with DPAMT + hugepage
(without shutdown optimization).
> > Avoid performing these cache flushes unless the CLFLUSH_BEFORE_ALLOC bit
> > of TDX_FEATURES0 is set.
> >
> > Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
> > Signed-off-by: Yan Zhao <yan.y.zhao@intel.com>
> > ---
> > RFC v2:
> > - Pulled from
> > git://git.kernel.org/pub/scm/linux/kernel/git/kas/linux.git tdx/dpamt-huge.
> > - Rebased on top of TDX huge page RFC v2 (Yan)
> > ---
> > arch/x86/include/asm/tdx.h | 1 +
> > arch/x86/virt/vmx/tdx/tdx.c | 19 +++++++++++++------
> > 2 files changed, 14 insertions(+), 6 deletions(-)
> >
> > diff --git a/arch/x86/include/asm/tdx.h b/arch/x86/include/asm/tdx.h
> > index f1bd74348b34..c058a82d4a97 100644
> > --- a/arch/x86/include/asm/tdx.h
> > +++ b/arch/x86/include/asm/tdx.h
> > @@ -15,6 +15,7 @@
> >
> > /* Bit definitions of TDX_FEATURES0 metadata field */
> > #define TDX_FEATURES0_NO_RBP_MOD BIT_ULL(18)
> > +#define TDX_FEATURES0_CLFLUSH_BEFORE_ALLOC BIT_ULL(23)
> > #define TDX_FEATURES0_DYNAMIC_PAMT BIT_ULL(36)
> >
> > #ifndef __ASSEMBLER__
> > diff --git a/arch/x86/virt/vmx/tdx/tdx.c b/arch/x86/virt/vmx/tdx/tdx.c
> > index 9ed585bde062..b7a0ee0f4a50 100644
> > --- a/arch/x86/virt/vmx/tdx/tdx.c
> > +++ b/arch/x86/virt/vmx/tdx/tdx.c
> > @@ -1648,14 +1648,13 @@ static inline u64 tdx_tdvpr_pa(struct tdx_vp *td)
> > return page_to_phys(td->tdvpr_page);
> > }
> >
> > -/*
> > - * The TDX module exposes a CLFLUSH_BEFORE_ALLOC bit to specify whether
> > - * a CLFLUSH of pages is required before handing them to the TDX module.
> > - * Be conservative and make the code simpler by doing the CLFLUSH
> > - * unconditionally.
> > - */
> > static void tdx_clflush_page(struct page *page)
> > {
> > + u64 tdx_features0 = tdx_sysinfo.features.tdx_features0;
> > +
> > + if (tdx_features0 & TDX_FEATURES0_CLFLUSH_BEFORE_ALLOC)
> > + return;
>
> Isn't the logic here and below reversed? If
> TDX_FEATURES0_CLFLUSH_BEFORE_ALLOC bit is set, we want to perform the
> clflush()
Yes, I think so.
As my test machine has boot_cpu_has_bug(X86_BUG_TDX_PW_MCE) returning true, I
thought it was right to perform clflush() and overlooked this logical error.
> > clflush_cache_range(page_to_virt(page), PAGE_SIZE);
> > }
> >
> > @@ -2030,8 +2029,12 @@ EXPORT_SYMBOL_GPL(tdh_phymem_cache_wb);
> >
> > u64 tdh_phymem_page_wbinvd_tdr(struct tdx_td *td)
> > {
> > + u64 tdx_features0 = tdx_sysinfo.features.tdx_features0;
> > struct tdx_module_args args = {};
> >
> > + if (tdx_features0 & TDX_FEATURES0_CLFLUSH_BEFORE_ALLOC)
> > + return 0;
> > +
> > args.rcx = mk_keyed_paddr(tdx_global_keyid, td->tdr_page);
> >
> > return seamcall(TDH_PHYMEM_PAGE_WBINVD, &args);
> > @@ -2041,10 +2044,14 @@ EXPORT_SYMBOL_GPL(tdh_phymem_page_wbinvd_tdr);
> > u64 tdh_phymem_page_wbinvd_hkid(u64 hkid, struct folio *folio,
> > unsigned long start_idx, unsigned long npages)
> > {
> > + u64 tdx_features0 = tdx_sysinfo.features.tdx_features0;
> > struct page *start = folio_page(folio, start_idx);
> > struct tdx_module_args args = {};
> > u64 err;
> >
> > + if (tdx_features0 & TDX_FEATURES0_CLFLUSH_BEFORE_ALLOC)
> > + return 0;
> > +
> > if (start_idx + npages > folio_nr_pages(folio))
> > return TDX_OPERAND_INVALID;
> >
> > --
> > 2.43.2
> >
> >
>
next prev parent reply other threads:[~2025-08-12 6:38 UTC|newest]
Thread overview: 129+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-08-07 9:39 [RFC PATCH v2 00/23] KVM: TDX huge page support for private memory Yan Zhao
2025-08-07 9:41 ` [RFC PATCH v2 01/23] x86/tdx: Enhance tdh_mem_page_aug() to support huge pages Yan Zhao
2025-08-07 9:41 ` [RFC PATCH v2 02/23] x86/virt/tdx: Add SEAMCALL wrapper tdh_mem_page_demote() Yan Zhao
2025-09-01 8:55 ` Binbin Wu
2025-09-01 9:08 ` Yan Zhao
2025-09-02 16:56 ` Edgecombe, Rick P
2025-09-02 17:37 ` Sean Christopherson
2025-09-02 17:45 ` Edgecombe, Rick P
2025-09-04 9:31 ` Yan Zhao
2025-11-11 9:15 ` Huang, Kai
2025-11-12 8:06 ` Yan Zhao
2025-11-14 9:14 ` Binbin Wu
2025-11-14 9:21 ` Yan Zhao
2025-08-07 9:42 ` [RFC PATCH v2 03/23] x86/tdx: Enhance tdh_phymem_page_wbinvd_hkid() to invalidate huge pages Yan Zhao
2025-11-11 9:23 ` Huang, Kai
2025-11-12 8:43 ` Yan Zhao
2025-11-12 10:29 ` Huang, Kai
2025-11-13 2:35 ` Yan Zhao
2025-11-13 7:37 ` Huang, Kai
2025-11-13 9:03 ` Yan Zhao
2025-11-13 15:26 ` Dave Hansen
2025-11-14 1:21 ` Yan Zhao
2025-12-10 1:14 ` Vishal Annapurve
2025-12-10 1:18 ` Yan Zhao
2025-12-10 1:30 ` Vishal Annapurve
2025-12-10 1:55 ` Yan Zhao
2025-12-31 19:37 ` Vishal Annapurve
2026-01-06 10:37 ` Yan Zhao
2025-08-07 9:42 ` [RFC PATCH v2 04/23] KVM: TDX: Introduce tdx_clear_folio() to clear " Yan Zhao
2025-09-02 2:56 ` Binbin Wu
2025-09-03 9:51 ` Yan Zhao
2025-09-03 11:19 ` Binbin Wu
2025-09-04 2:53 ` Yan Zhao
2025-08-07 9:42 ` [RFC PATCH v2 05/23] x86/tdx: Enhance tdh_phymem_page_reclaim() to support " Yan Zhao
2025-11-17 2:09 ` Binbin Wu
2025-11-17 4:05 ` Yan Zhao
2025-08-07 9:42 ` [RFC PATCH v2 06/23] KVM: TDX: Do not hold page refcount on private guest pages Yan Zhao
2025-08-07 9:42 ` [RFC PATCH v2 07/23] KVM: x86/mmu: Disallow page merging (huge page adjustment) for mirror root Yan Zhao
2025-08-07 9:43 ` [RFC PATCH v2 08/23] KVM: x86/tdp_mmu: Alloc external_spt page for mirror page table splitting Yan Zhao
2025-11-11 9:52 ` Huang, Kai
2025-11-12 9:29 ` Yan Zhao
2025-08-07 9:43 ` [RFC PATCH v2 09/23] KVM: x86/tdp_mmu: Add split_external_spt hook called during write mmu_lock Yan Zhao
2025-11-11 10:06 ` Huang, Kai
2025-11-13 3:16 ` Yan Zhao
2025-11-17 8:53 ` Binbin Wu
2025-11-17 9:09 ` Yan Zhao
2025-08-07 9:43 ` [RFC PATCH v2 10/23] KVM: TDX: Enable huge page splitting under write kvm->mmu_lock Yan Zhao
2025-11-11 10:20 ` Huang, Kai
2025-11-13 5:53 ` Yan Zhao
2025-11-17 9:17 ` Binbin Wu
2025-11-17 9:26 ` Yan Zhao
2025-12-09 23:49 ` Sagi Shahar
2025-12-09 23:54 ` Edgecombe, Rick P
2025-12-10 0:28 ` Sagi Shahar
2025-12-10 0:50 ` Yan Zhao
2025-12-10 17:16 ` Sagi Shahar
2025-12-10 19:49 ` Edgecombe, Rick P
2025-12-11 2:10 ` Yan Zhao
2025-08-07 9:43 ` [RFC PATCH v2 11/23] KVM: x86: Reject splitting huge pages under shared mmu_lock for mirror root Yan Zhao
2025-09-03 3:30 ` Binbin Wu
2025-08-07 9:43 ` [RFC PATCH v2 12/23] KVM: x86/mmu: Introduce kvm_split_cross_boundary_leafs() Yan Zhao
2025-09-03 6:57 ` Binbin Wu
2025-09-03 9:44 ` Yan Zhao
2025-11-11 10:42 ` Huang, Kai
2025-11-13 8:54 ` Yan Zhao
2025-11-13 11:02 ` Huang, Kai
2025-11-13 11:40 ` Huang, Kai
2025-11-14 6:09 ` Yan Zhao
2025-11-18 0:14 ` Huang, Kai
2025-11-18 6:30 ` Yan Zhao
2025-11-18 8:59 ` Binbin Wu
2025-11-18 10:49 ` Huang, Kai
2025-11-19 3:41 ` Yan Zhao
2026-01-06 10:35 ` Yan Zhao
2025-11-19 6:23 ` Yan Zhao
2025-11-19 6:31 ` Yan Zhao
2025-08-07 9:44 ` [RFC PATCH v2 13/23] KVM: x86: Introduce hugepage_set_guest_inhibit() Yan Zhao
2025-08-07 9:44 ` [RFC PATCH v2 14/23] KVM: TDX: Split and inhibit huge mappings if a VMExit carries level info Yan Zhao
2025-09-03 7:36 ` Binbin Wu
2025-09-03 9:37 ` Yan Zhao
2025-11-11 10:55 ` Huang, Kai
2025-11-14 1:42 ` Yan Zhao
2025-11-18 0:26 ` Huang, Kai
2025-11-18 2:44 ` Yan Zhao
2025-11-11 11:05 ` Huang, Kai
2025-11-14 7:22 ` Yan Zhao
2025-11-18 1:04 ` Huang, Kai
2025-11-18 2:20 ` Yan Zhao
2025-11-18 9:44 ` Huang, Kai
2025-11-19 2:58 ` Yan Zhao
2025-11-19 5:51 ` Binbin Wu
2025-11-19 6:29 ` Yan Zhao
2025-11-19 6:39 ` Binbin Wu
2025-08-07 9:44 ` [RFC PATCH v2 15/23] KVM: Change the return type of gfn_handler_t() from bool to int Yan Zhao
2025-08-07 9:44 ` [RFC PATCH v2 16/23] KVM: x86: Split cross-boundary mirror leafs for KVM_SET_MEMORY_ATTRIBUTES Yan Zhao
2025-08-07 9:45 ` [RFC PATCH v2 17/23] KVM: guest_memfd: Split for punch hole and private-to-shared conversion Yan Zhao
2025-09-04 7:58 ` Binbin Wu
2025-09-04 9:48 ` Yan Zhao
2025-09-04 11:07 ` Yan Zhao
2025-10-01 6:21 ` Ackerley Tng
2025-10-13 0:18 ` Yan Zhao
2025-10-01 8:00 ` Ackerley Tng
2025-10-13 0:45 ` Yan Zhao
2025-08-07 9:45 ` [RFC PATCH v2 18/23] x86/virt/tdx: Do not perform cache flushes unless CLFLUSH_BEFORE_ALLOC is set Yan Zhao
2025-08-11 21:10 ` Sagi Shahar
2025-08-12 6:37 ` Yan Zhao [this message]
2025-09-04 8:16 ` Binbin Wu
2025-09-04 9:50 ` Yan Zhao
2025-09-05 9:05 ` Binbin Wu
2025-09-05 15:41 ` Edgecombe, Rick P
2025-09-15 6:05 ` Yan Zhao
2025-08-07 9:45 ` [RFC PATCH v2 19/23] KVM: TDX: Pass down pfn to split_external_spt() Yan Zhao
2025-09-04 8:30 ` Binbin Wu
2025-08-07 9:45 ` [RFC PATCH v2 20/23] KVM: TDX: Handle Dynamic PAMT in tdh_mem_page_demote() Yan Zhao
2025-08-07 9:46 ` [RFC PATCH v2 21/23] KVM: TDX: Preallocate PAMT pages to be used in split path Yan Zhao
2025-09-04 9:17 ` Binbin Wu
2025-09-04 9:58 ` Yan Zhao
2025-12-05 6:14 ` Sagi Shahar
2025-12-08 5:49 ` Yan Zhao
2025-12-11 1:42 ` Vishal Annapurve
2025-12-11 2:36 ` Yan Zhao
2025-08-07 9:46 ` [RFC PATCH v2 22/23] KVM: TDX: Handle Dynamic PAMT on page split Yan Zhao
2025-08-14 5:31 ` Vishal Annapurve
2025-08-14 18:29 ` Vishal Annapurve
2025-08-18 4:19 ` Yan Zhao
2025-08-07 9:46 ` [RFC PATCH v2 23/23] KVM: TDX: Turn on PG_LEVEL_2M after TD is RUNNABLE Yan Zhao
2025-11-11 11:25 ` Huang, Kai
2025-11-14 8:34 ` Yan Zhao
2025-11-18 0:56 ` Huang, Kai
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=aJrhNFmLFBOP2TVK@yzhao56-desk.sh.intel.com \
--to=yan.y.zhao@intel.com \
--cc=ackerleytng@google.com \
--cc=binbin.wu@linux.intel.com \
--cc=chao.p.peng@intel.com \
--cc=dave.hansen@intel.com \
--cc=david@redhat.com \
--cc=fan.du@intel.com \
--cc=ira.weiny@intel.com \
--cc=isaku.yamahata@intel.com \
--cc=jun.miao@intel.com \
--cc=kas@kernel.org \
--cc=kvm@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=michael.roth@amd.com \
--cc=pbonzini@redhat.com \
--cc=pgonda@google.com \
--cc=quic_eberman@quicinc.com \
--cc=rick.p.edgecombe@intel.com \
--cc=sagis@google.com \
--cc=seanjc@google.com \
--cc=tabba@google.com \
--cc=thomas.lendacky@amd.com \
--cc=vannapurve@google.com \
--cc=vbabka@suse.cz \
--cc=x86@kernel.org \
--cc=xiaoyao.li@intel.com \
--cc=zhiquan1.li@intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox