[PATCH v3 01/14] x86/virt/tdx: Add SEAMCALL wrappers for TDX KeyID management

From: Paolo Bonzini
Date: Wed Jan 15 2025 - 11:09:50 EST


From: Rick Edgecombe <rick.p.edgecombe@xxxxxxxxx>

Intel TDX protects guest VMs from malicious host and certain physical
attacks. Pre-TDX Intel hardware has support for a memory encryption
architecture called MK-TME, which repurposes several high bits of
physical address as "KeyID". TDX ends up with reserving a sub-range of
MK-TME KeyIDs as "TDX private KeyIDs".

Like MK-TME, these KeyIDs can be associated with an ephemeral key. For TDX
this association is done by the TDX module. It also has its own tracking
for which KeyIDs are in use. To do this ephemeral key setup and manipulate
the TDX module's internal tracking, KVM will use the following SEAMCALLs:
TDH.MNG.KEY.CONFIG: Mark the KeyID as in use, and initialize its
ephemeral key.
TDH.MNG.KEY.FREEID: Mark the KeyID as not in use.

These SEAMCALLs both operate on TDR structures, which are setup using the
previously added TDH.MNG.CREATE SEAMCALL. KVM's use of these operations
will go like:
- tdx_guest_keyid_alloc()
- Initialize TD and TDR page with TDH.MNG.CREATE (not yet-added), passing
KeyID
- TDH.MNG.KEY.CONFIG to initialize the key
- TD runs, teardown is started
- TDH.MNG.KEY.FREEID
- tdx_guest_keyid_free()

Don't try to combine the tdx_guest_keyid_alloc() and TDH.MNG.KEY.CONFIG
operations because TDH.MNG.CREATE and some locking need to be done in the
middle. Don't combine TDH.MNG.KEY.FREEID and tdx_guest_keyid_free() so they
are symmetrical with the creation path.

So implement tdh_mng_key_config() and tdh_mng_key_freeid() as separate
functions than tdx_guest_keyid_alloc() and tdx_guest_keyid_free().

The TDX module provides SEAMCALLs to hand pages to the TDX module for
storing TDX controlled state. SEAMCALLs that operate on this state are
directed to the appropriate TD VM using references to the pages originally
provided for managing the TD's state. So the host kernel needs to track
these pages, both as an ID for specifying which TD to operate on, and to
allow them to be eventually reclaimed. The TD VM associated pages are
called TDR (Trust Domain Root) and TDCS (Trust Domain Control Structure).

Introduce "struct tdx_td" for holding references to pages provided to the
TDX module for this TD VM associated state. Don't plan for any TD
associated state that is controlled by KVM to live in this struct. Only
expect it to hold data for concepts specific to the TDX architecture, for
which there can't already be preexisting storage for in KVM.

Add both the TDR page and an array of TDCS pages, even though the SEAMCALL
wrappers will only need to know about the TDR pages for directing the
SEAMCALLs to the right TD. Adding the TDCS pages to this struct will let
all of the TD VM associated pages handed to the TDX module be tracked in
one location. For a type to specify physical pages, use KVM's hpa_t type.
Do this for KVM's benefit This is the common type used to hold physical
addresses in KVM, so will make interoperability easier.

Co-developed-by: Sean Christopherson <sean.j.christopherson@xxxxxxxxx>
Signed-off-by: Sean Christopherson <sean.j.christopherson@xxxxxxxxx>
Signed-off-by: Isaku Yamahata <isaku.yamahata@xxxxxxxxx>
Signed-off-by: Kai Huang <kai.huang@xxxxxxxxx>
Signed-off-by: Rick Edgecombe <rick.p.edgecombe@xxxxxxxxx>
Reviewed-by: Binbin Wu <binbin.wu@xxxxxxxxxxxxxxx>
Reviewed-by: Yuan Yao <yuan.yao@xxxxxxxxx>
Message-ID: <20241203010317.827803-2-rick.p.edgecombe@xxxxxxxxx>
Acked-by: Dave Hansen <dave.hansen@xxxxxxxxxxxxxxx>
Signed-off-by: Paolo Bonzini <pbonzini@xxxxxxxxxx>
---
arch/x86/include/asm/tdx.h | 12 ++++++++++++
arch/x86/virt/vmx/tdx/tdx.c | 25 +++++++++++++++++++++++++
arch/x86/virt/vmx/tdx/tdx.h | 16 +++++++++-------
3 files changed, 46 insertions(+), 7 deletions(-)

diff --git a/arch/x86/include/asm/tdx.h b/arch/x86/include/asm/tdx.h
index eba178996d84..5045ab1c3d5b 100644
--- a/arch/x86/include/asm/tdx.h
+++ b/arch/x86/include/asm/tdx.h
@@ -116,6 +116,18 @@ static inline u64 sc_retry(sc_func_t func, u64 fn,
int tdx_cpu_enable(void);
int tdx_enable(void);
const char *tdx_dump_mce_info(struct mce *m);
+
+struct tdx_td {
+ /* TD root structure: */
+ struct page *tdr_page;
+
+ int tdcs_nr_pages;
+ /* TD control structure: */
+ struct page **tdcs_pages;
+};
+
+u64 tdh_mng_key_config(struct tdx_td *td);
+u64 tdh_mng_key_freeid(struct tdx_td *td);
#else
static inline void tdx_init(void) { }
static inline int tdx_cpu_enable(void) { return -ENODEV; }
diff --git a/arch/x86/virt/vmx/tdx/tdx.c b/arch/x86/virt/vmx/tdx/tdx.c
index 7fdb37387886..1ffbdb840004 100644
--- a/arch/x86/virt/vmx/tdx/tdx.c
+++ b/arch/x86/virt/vmx/tdx/tdx.c
@@ -1456,3 +1456,28 @@ void __init tdx_init(void)

check_tdx_erratum();
}
+
+static inline u64 tdx_tdr_pa(struct tdx_td *td)
+{
+ return page_to_phys(td->tdr_page);
+}
+
+u64 tdh_mng_key_config(struct tdx_td *td)
+{
+ struct tdx_module_args args = {
+ .rcx = tdx_tdr_pa(td),
+ };
+
+ return seamcall(TDH_MNG_KEY_CONFIG, &args);
+}
+EXPORT_SYMBOL_GPL(tdh_mng_key_config);
+
+u64 tdh_mng_key_freeid(struct tdx_td *td)
+{
+ struct tdx_module_args args = {
+ .rcx = tdx_tdr_pa(td),
+ };
+
+ return seamcall(TDH_MNG_KEY_FREEID, &args);
+}
+EXPORT_SYMBOL_GPL(tdh_mng_key_freeid);
diff --git a/arch/x86/virt/vmx/tdx/tdx.h b/arch/x86/virt/vmx/tdx/tdx.h
index 4e3d533cdd61..5579317f67ab 100644
--- a/arch/x86/virt/vmx/tdx/tdx.h
+++ b/arch/x86/virt/vmx/tdx/tdx.h
@@ -15,13 +15,15 @@
/*
* TDX module SEAMCALL leaf functions
*/
-#define TDH_PHYMEM_PAGE_RDMD 24
-#define TDH_SYS_KEY_CONFIG 31
-#define TDH_SYS_INIT 33
-#define TDH_SYS_RD 34
-#define TDH_SYS_LP_INIT 35
-#define TDH_SYS_TDMR_INIT 36
-#define TDH_SYS_CONFIG 45
+#define TDH_MNG_KEY_CONFIG 8
+#define TDH_MNG_KEY_FREEID 20
+#define TDH_PHYMEM_PAGE_RDMD 24
+#define TDH_SYS_KEY_CONFIG 31
+#define TDH_SYS_INIT 33
+#define TDH_SYS_RD 34
+#define TDH_SYS_LP_INIT 35
+#define TDH_SYS_TDMR_INIT 36
+#define TDH_SYS_CONFIG 45

/* TDX page types */
#define PT_NDA 0x0
--
2.43.5