diff mbox series

[v9,33/43] x86/compressed: Add SEV-SNP feature detection/setup

Message ID 20220128171804.569796-34-brijesh.singh@amd.com
State New
Headers show
Series Add AMD Secure Nested Paging (SEV-SNP) Guest Support | expand

Commit Message

Brijesh Singh Jan. 28, 2022, 5:17 p.m. UTC
From: Michael Roth <michael.roth@amd.com>

Initial/preliminary detection of SEV-SNP is done via the Confidential
Computing blob. Check for it prior to the normal SEV/SME feature
initialization, and add some sanity checks to confirm it agrees with
SEV-SNP CPUID/MSR bits.

Signed-off-by: Michael Roth <michael.roth@amd.com>
Signed-off-by: Brijesh Singh <brijesh.singh@amd.com>
---
 arch/x86/boot/compressed/sev.c | 118 ++++++++++++++++++++++++++++++++-
 arch/x86/include/asm/sev.h     |   3 +
 2 files changed, 120 insertions(+), 1 deletion(-)

Comments

Borislav Petkov Feb. 6, 2022, 4:41 p.m. UTC | #1
On Fri, Jan 28, 2022 at 11:17:54AM -0600, Brijesh Singh wrote:
> +static struct cc_setup_data *get_cc_setup_data(struct boot_params *bp)
> +{
> +	struct setup_data *hdr = (struct setup_data *)bp->hdr.setup_data;
> +
> +	while (hdr) {
> +		if (hdr->type == SETUP_CC_BLOB)
> +			return (struct cc_setup_data *)hdr;
> +		hdr = (struct setup_data *)hdr->next;
> +	}
> +
> +	return NULL;
> +}

Merge that function into its only caller.

...

> +static struct cc_blob_sev_info *snp_find_cc_blob(struct boot_params *bp)

static function, no need for the "snp_" prefix. Please audit all your
patches for that and remove that prefix from all static functions.

> +{
> +	struct cc_blob_sev_info *cc_info;
> +
> +	cc_info = snp_find_cc_blob_efi(bp);
> +	if (cc_info)
> +		goto found_cc_info;
> +
> +	cc_info = snp_find_cc_blob_setup_data(bp);
> +	if (!cc_info)
> +		return NULL;
> +
> +found_cc_info:
> +	if (cc_info->magic != CC_BLOB_SEV_HDR_MAGIC)
> +		sev_es_terminate(SEV_TERM_SET_GEN, GHCB_SNP_UNSUPPORTED);
> +
> +	return cc_info;
> +}
> +
> +bool snp_init(struct boot_params *bp)
> +{
> +	struct cc_blob_sev_info *cc_info;
> +
> +	if (!bp)
> +		return false;
> +
> +	cc_info = snp_find_cc_blob(bp);
> +	if (!cc_info)
> +		return false;
> +
> +	/*
> +	 * Pass run-time kernel a pointer to CC info via boot_params so EFI
> +	 * config table doesn't need to be searched again during early startup
> +	 * phase.
> +	 */
> +	bp->cc_blob_address = (u32)(unsigned long)cc_info;
> +
> +	/*
> +	 * Indicate SEV-SNP based on presence of SEV-SNP-specific CC blob.
> +	 * Subsequent checks will verify SEV-SNP CPUID/MSR bits.
> +	 */

Put that comment over the function name.

Thx.
Michael Roth Feb. 8, 2022, 1:50 p.m. UTC | #2
On Sun, Feb 06, 2022 at 05:41:26PM +0100, Borislav Petkov wrote:
> On Fri, Jan 28, 2022 at 11:17:54AM -0600, Brijesh Singh wrote:
> > +static struct cc_setup_data *get_cc_setup_data(struct boot_params *bp)
> > +{
> > +	struct setup_data *hdr = (struct setup_data *)bp->hdr.setup_data;
> > +
> > +	while (hdr) {
> > +		if (hdr->type == SETUP_CC_BLOB)
> > +			return (struct cc_setup_data *)hdr;
> > +		hdr = (struct setup_data *)hdr->next;
> > +	}
> > +
> > +	return NULL;
> > +}
> 
> Merge that function into its only caller.
> 
> ...
> 
> > +static struct cc_blob_sev_info *snp_find_cc_blob(struct boot_params *bp)
> 
> static function, no need for the "snp_" prefix. Please audit all your
> patches for that and remove that prefix from all static functions.
> 

I'm a little unsure how to handle some of these others cases:

We have this which is defined in sev-shared.c and used "externally"
by kernel/sev.c or boot/compressed/sev.c:

  snp_setup_cpuid_table()

I'm assuming that would be considered 'non-static' since it would need
to be exported if sev-shared.c was compiled separately instead of
directly #include'd.

And then there's also these which are static helpers that are only used
within sev-shared.c:

  snp_cpuid_info_get_ptr()
  snp_cpuid_calc_xsave_size()
  snp_cpuid_get_validated_func()
  snp_cpuid_check_range()
  snp_cpuid_hv()
  snp_cpuid_postprocess()
  snp_cpuid()

but in those cases it seems useful to keep them grouped under the
snp_cpuid_* prefix since they become ambiguous otherwise, and
just using cpuid_* as a prefix (or suffix/etc) makes it unclear
that they are only used for SNP and not for general CPUID handling.
Should we leave those as-is?
Borislav Petkov Feb. 8, 2022, 3:02 p.m. UTC | #3
On Tue, Feb 08, 2022 at 07:50:09AM -0600, Michael Roth wrote:
> I'm assuming that would be considered 'non-static' since it would need
> to be exported if sev-shared.c was compiled separately instead of
> directly #include'd.

No, that one can lose the prefix too. We'll cross that bridge when we
get to it.

> And then there's also these which are static helpers that are only used
> within sev-shared.c:
> 
>   snp_cpuid_info_get_ptr()

So looking at that one - and it felt weird reading that code because it
said "cpuid_info" but that's not an "info" - that should be:

struct snp_cpuid_table {
        u32 count;
        u32 __reserved1;
        u64 __reserved2;
        struct snp_cpuid_fn fn[SNP_CPUID_COUNT_MAX];
} __packed;

just call it what it is - a SNP CPUID *table*.

And then you can have

	const struct snp_cpuid_table *cpuid_tbl = snp_cpuid_get_table();

and that makes it crystal clear what this does.

>   snp_cpuid_calc_xsave_size()
>   snp_cpuid_get_validated_func()
>
>   snp_cpuid_check_range()

You can merge that small function into its only call site and put a
comment above the code:

	/* Check function is within the supported CPUID leafs ranges */

>   snp_cpuid_hv()
>   snp_cpuid_postprocess()
>   snp_cpuid()
> 
> but in those cases it seems useful to keep them grouped under the
> snp_cpuid_* prefix since they become ambiguous otherwise, and
> just using cpuid_* as a prefix (or suffix/etc) makes it unclear
> that they are only used for SNP and not for general CPUID handling.
> Should we leave those as-is?

Yap, the rest make sense to denote to what functionality they belong to.

Thx.
diff mbox series

Patch

diff --git a/arch/x86/boot/compressed/sev.c b/arch/x86/boot/compressed/sev.c
index 1b80c1d0ea1f..04cabff015ba 100644
--- a/arch/x86/boot/compressed/sev.c
+++ b/arch/x86/boot/compressed/sev.c
@@ -286,6 +286,13 @@  static void enforce_vmpl0(void)
 void sev_enable(struct boot_params *bp)
 {
 	unsigned int eax, ebx, ecx, edx;
+	bool snp;
+
+	/*
+	 * Setup/preliminary detection of SEV-SNP. This will be sanity-checked
+	 * against CPUID/MSR values later.
+	 */
+	snp = snp_init(bp);
 
 	/* Check for the SME/SEV support leaf */
 	eax = 0x80000000;
@@ -306,8 +313,11 @@  void sev_enable(struct boot_params *bp)
 	ecx = 0;
 	native_cpuid(&eax, &ebx, &ecx, &edx);
 	/* Check whether SEV is supported */
-	if (!(eax & BIT(1)))
+	if (!(eax & BIT(1))) {
+		if (snp)
+			error("SEV-SNP support indicated by CC blob, but not CPUID.");
 		return;
+	}
 
 	/* Set the SME mask if this is an SEV guest. */
 	sev_status = rd_sev_status_msr();
@@ -332,5 +342,111 @@  void sev_enable(struct boot_params *bp)
 		enforce_vmpl0();
 	}
 
+	if (snp && !(sev_status & MSR_AMD64_SEV_SNP_ENABLED))
+		error("SEV-SNP supported indicated by CC blob, but not SEV status MSR.");
+
 	sme_me_mask = BIT_ULL(ebx & 0x3f);
 }
+
+/* Search for Confidential Computing blob in the EFI config table. */
+static struct cc_blob_sev_info *snp_find_cc_blob_efi(struct boot_params *bp)
+{
+	unsigned long cfg_table_pa;
+	unsigned int cfg_table_len;
+	int ret;
+
+	ret = efi_get_conf_table(bp, &cfg_table_pa, &cfg_table_len);
+	if (ret)
+		return NULL;
+
+	return (struct cc_blob_sev_info *)efi_find_vendor_table(bp, cfg_table_pa,
+								cfg_table_len,
+								EFI_CC_BLOB_GUID);
+}
+
+struct cc_setup_data {
+	struct setup_data header;
+	u32 cc_blob_address;
+};
+
+static struct cc_setup_data *get_cc_setup_data(struct boot_params *bp)
+{
+	struct setup_data *hdr = (struct setup_data *)bp->hdr.setup_data;
+
+	while (hdr) {
+		if (hdr->type == SETUP_CC_BLOB)
+			return (struct cc_setup_data *)hdr;
+		hdr = (struct setup_data *)hdr->next;
+	}
+
+	return NULL;
+}
+
+/*
+ * Search for a Confidential Computing blob passed in as a setup_data entry
+ * via the Linux Boot Protocol.
+ */
+static struct cc_blob_sev_info *snp_find_cc_blob_setup_data(struct boot_params *bp)
+{
+	struct cc_setup_data *sd;
+
+	sd = get_cc_setup_data(bp);
+	if (!sd)
+		return NULL;
+
+	return (struct cc_blob_sev_info *)(unsigned long)sd->cc_blob_address;
+}
+
+/*
+ * Initial set up of SEV-SNP relies on information provided by the
+ * Confidential Computing blob, which can be passed to the boot kernel
+ * by firmware/bootloader in the following ways:
+ *
+ * - via an entry in the EFI config table
+ * - via a setup_data structure, as defined by the Linux Boot Protocol
+ *
+ * Scan for the blob in that order.
+ */
+static struct cc_blob_sev_info *snp_find_cc_blob(struct boot_params *bp)
+{
+	struct cc_blob_sev_info *cc_info;
+
+	cc_info = snp_find_cc_blob_efi(bp);
+	if (cc_info)
+		goto found_cc_info;
+
+	cc_info = snp_find_cc_blob_setup_data(bp);
+	if (!cc_info)
+		return NULL;
+
+found_cc_info:
+	if (cc_info->magic != CC_BLOB_SEV_HDR_MAGIC)
+		sev_es_terminate(SEV_TERM_SET_GEN, GHCB_SNP_UNSUPPORTED);
+
+	return cc_info;
+}
+
+bool snp_init(struct boot_params *bp)
+{
+	struct cc_blob_sev_info *cc_info;
+
+	if (!bp)
+		return false;
+
+	cc_info = snp_find_cc_blob(bp);
+	if (!cc_info)
+		return false;
+
+	/*
+	 * Pass run-time kernel a pointer to CC info via boot_params so EFI
+	 * config table doesn't need to be searched again during early startup
+	 * phase.
+	 */
+	bp->cc_blob_address = (u32)(unsigned long)cc_info;
+
+	/*
+	 * Indicate SEV-SNP based on presence of SEV-SNP-specific CC blob.
+	 * Subsequent checks will verify SEV-SNP CPUID/MSR bits.
+	 */
+	return true;
+}
diff --git a/arch/x86/include/asm/sev.h b/arch/x86/include/asm/sev.h
index 1a7e21bb6eea..4e3909042001 100644
--- a/arch/x86/include/asm/sev.h
+++ b/arch/x86/include/asm/sev.h
@@ -11,6 +11,7 @@ 
 #include <linux/types.h>
 #include <asm/insn.h>
 #include <asm/sev-common.h>
+#include <asm/bootparam.h>
 
 #define GHCB_PROTOCOL_MIN	1ULL
 #define GHCB_PROTOCOL_MAX	2ULL
@@ -151,6 +152,7 @@  void __init snp_prep_memory(unsigned long paddr, unsigned int sz, enum psc_op op
 void snp_set_memory_shared(unsigned long vaddr, unsigned int npages);
 void snp_set_memory_private(unsigned long vaddr, unsigned int npages);
 void snp_set_wakeup_secondary_cpu(void);
+bool snp_init(struct boot_params *bp);
 #else
 static inline void sev_es_ist_enter(struct pt_regs *regs) { }
 static inline void sev_es_ist_exit(void) { }
@@ -168,6 +170,7 @@  static inline void __init snp_prep_memory(unsigned long paddr, unsigned int sz,
 static inline void snp_set_memory_shared(unsigned long vaddr, unsigned int npages) { }
 static inline void snp_set_memory_private(unsigned long vaddr, unsigned int npages) { }
 static inline void snp_set_wakeup_secondary_cpu(void) { }
+static inline bool snp_init(struct boot_params *bp) { return false; }
 #endif
 
 #endif