Skip to contentSkip to conversation

Support TDX guests on Hyper-V (the x86/tdx part)

PATCH In review Dexuan Cui 5 participants 12 messages Aug 11, 2023 – Sep 7, 2023

Subsystems (4)
lore
Share & tools
lore

Dexuan Cui

author maintainer
#0 lore raw

v10 cover Support TDX guests on Hyper-V (the x86/tdx part)

The two patches can apply cleanly to today's tip.git's master branch.

Dave kindly re-wrote the changelog for the first patch, and suggested a better version of the second patch. See [1].

I integrated Dave's comments into this v10. Dave, thank you!

Please review.

[1] v9: https://lwn.net/ml/linux-kernel/[email protected]/

Dexuan Cui (2):
x86/tdx: Retry partially-completed page conversion hypercalls x86/tdx: Support vmalloc() for tdx_enc_status_changed()

 arch/x86/coco/tdx/tdx.c           | 88 ++++++++++++++++++++++++++-----
 arch/x86/include/asm/shared/tdx.h |  2 +
 2 files changed, 78 insertions(+), 12 deletions(-)
Signature · 1 line
2.25.1
author maintainer
#1 lore raw

PATCH v10 1/2 x86/tdx: Retry partially-completed page conversion hypercalls

TDX guest memory is private by default and the VMM may not access it. However, in cases where the guest needs to share data with the VMM, the guest and the VMM can coordinate to make memory shared between them.

The guest side of this protocol includes the "MapGPA" hypercall. This call takes a guest physical address range. The hypercall spec (aka. the GHCI) says that the MapGPA call is allowed to return partial progress in mapping this range and indicate that fact with a special error code. A guest that sees such partial progress is expected to retry the operation for the portion of the address range that was not completed.

Hyper-V does this partial completion dance when set_memory_decrypted() is called to "decrypt" swiotlb bounce buffers that can be up to 1GB in size. It is evidently the only VMM that does this, which is why nobody noticed this until now.

---
 arch/x86/coco/tdx/tdx.c           | 64 +++++++++++++++++++++++++------
 arch/x86/include/asm/shared/tdx.h |  2 +
 2 files changed, 54 insertions(+), 12 deletions(-)

Changes in v10:
Dave kindly re-wrote the changelog. No other changes.

arch/x86/coco/tdx/tdx.c +52 -12
index 1d6b863c42b00..746075d20cd2d 100644
@@ -703,14 +703,15 @@ static bool tdx_cache_flush_required(void)
 }
 
 /*
- * Inform the VMM of the guest's intent for this physical page: shared with
- * the VMM or private to the guest.  The VMM is expected to change its mapping
- * of the page in response.
+ * Notify the VMM about page mapping conversion. More info about ABI
+ * can be found in TDX Guest-Host-Communication Interface (GHCI),
+ * section "TDG.VP.VMCALL<MapGPA>".
  */
-static bool tdx_enc_status_changed(unsigned long vaddr, int numpages, bool enc)
+static bool tdx_map_gpa(phys_addr_t start, phys_addr_t end, bool enc)
 {
-	phys_addr_t start = __pa(vaddr);
-	phys_addr_t end   = __pa(vaddr + numpages * PAGE_SIZE);
+	/* Retrying the hypercall a second time should succeed; use 3 just in case */
+	const int max_retries_per_page = 3;
+	int retry_count = 0;
 
 	if (!enc) {
 		/* Set the shared (decrypted) bits: */
@@ -718,12 +719,51 @@ static bool tdx_enc_status_changed(unsigned long vaddr, int numpages, bool enc)
 		end   |= cc_mkdec(0);
 	}
 
-	/*
-	 * Notify the VMM about page mapping conversion. More info about ABI
-	 * can be found in TDX Guest-Host-Communication Interface (GHCI),
-	 * section "TDG.VP.VMCALL<MapGPA>"
-	 */
-	if (_tdx_hypercall(TDVMCALL_MAP_GPA, start, end - start, 0, 0))
+	while (retry_count < max_retries_per_page) {
+		struct tdx_hypercall_args args = {
+			.r10 = TDX_HYPERCALL_STANDARD,
+			.r11 = TDVMCALL_MAP_GPA,
+			.r12 = start,
+			.r13 = end - start };
+
+		u64 map_fail_paddr;
+		u64 ret = __tdx_hypercall_ret(&args);
+
+		if (ret != TDVMCALL_STATUS_RETRY)
+			return !ret;
+		/*
+		 * The guest must retry the operation for the pages in the
+		 * region starting at the GPA specified in R11. R11 comes
+		 * from the untrusted VMM. Sanity check it.
+		 */
+		map_fail_paddr = args.r11;
+		if (map_fail_paddr < start || map_fail_paddr >= end)
+			return false;
+
+		/* "Consume" a retry without forward progress */
+		if (map_fail_paddr == start) {
+			retry_count++;
+			continue;
+		}
+
+		start = map_fail_paddr;
+		retry_count = 0;
+	}
+
+	return false;
+}
+
+/*
+ * Inform the VMM of the guest's intent for this physical page: shared with
+ * the VMM or private to the guest.  The VMM is expected to change its mapping
+ * of the page in response.
+ */
+static bool tdx_enc_status_changed(unsigned long vaddr, int numpages, bool enc)
+{
+	phys_addr_t start = __pa(vaddr);
+	phys_addr_t end   = __pa(vaddr + numpages * PAGE_SIZE);
+
+	if (!tdx_map_gpa(start, end, enc))
 		return false;
 
 	/* shared->private conversion requires memory to be accepted before use */
arch/x86/include/asm/shared/tdx.h +2 -0
index 7513b3bb69b7e..22ee23a3f24a6 100644
@@ -24,6 +24,8 @@
 #define TDVMCALL_MAP_GPA		0x10001
 #define TDVMCALL_REPORT_FATAL_ERROR	0x10003
 
+#define TDVMCALL_STATUS_RETRY		1
+
 #ifndef __ASSEMBLY__
 
 /*
Signature · 1 line
2.25.1
Reviewed-by Tested-by on v10 1/2
#2 lore raw

Re: [PATCH v10 1/2] x86/tdx: Retry partially-completed page conversion hypercalls

Quoted 13 lines · Dexuan Cui

TDX guest memory is private by default and the VMM may not access it. However, in cases where the guest needs to share data with the VMM, the guest and the VMM can coordinate to make memory shared between them.

The guest side of this protocol includes the "MapGPA" hypercall. This call takes a guest physical address range. The hypercall spec (aka. the GHCI) says that the MapGPA call is allowed to return partial progress in mapping this range and indicate that fact with a special error code. A guest that sees such partial progress is expected to retry the operation for the portion of the address range that was not completed.

Quoted 4 lines

Hyper-V does this partial completion dance when set_memory_decrypted() is called to "decrypt" swiotlb bounce buffers that can be up to 1GB in size. It is evidently the only VMM that does this, which is why nobody noticed this until now.

Now TDX KVM + TDX qemu supports partial completion because TD guest can pass very large range. e.g. 1GB order. I tested this patch with (patched) TDX KVM/qemu.

Signature · 1 line
Isaku Yamahata <[email protected]>
author maintainer
#3 lore raw

RE: [PATCH v10 1/2] x86/tdx: Retry partially-completed page conversion hypercalls

Quoted 26 lines

From: Isaku Yamahata <[email protected]>
Sent: Monday, August 14, 2023 12:04 PM
To: Dexuan Cui <[email protected]>
[...]

Quoted 13 lines · Dexuan Cui

TDX guest memory is private by default and the VMM may not access it. However, in cases where the guest needs to share data with the VMM, the guest and the VMM can coordinate to make memory shared between them.

The guest side of this protocol includes the "MapGPA" hypercall. This call takes a guest physical address range. The hypercall spec (aka. the GHCI) says that the MapGPA call is allowed to return partial progress in mapping this range and indicate that fact with a special error code. A guest that sees such partial progress is expected to retry the operation for the portion of the address range that was not completed.

Quoted 4 lines

Hyper-V does this partial completion dance when set_memory_decrypted() is called to "decrypt" swiotlb bounce buffers that can be up to 1GB in size. It is evidently the only VMM that does this, which is why nobody noticed this until now.

Quoted 7 lines

Now TDX KVM + TDX qemu supports partial completion because TD guest can pass
very large range. e.g. 1GB order. I tested this patch with (patched) TDX KVM/qemu.

Thanks Isaku for reviewing and testing the patch!

@Dave, may I know if the 2 updated patches look good to you?

#4 lore raw

Re: [PATCH v10 1/2] x86/tdx: Retry partially-completed page conversion hypercalls

In reply to Dexuan Cui
Quoted 13 lines · Dexuan Cui

TDX guest memory is private by default and the VMM may not access it. However, in cases where the guest needs to share data with the VMM, the guest and the VMM can coordinate to make memory shared between them.

The guest side of this protocol includes the "MapGPA" hypercall. This call takes a guest physical address range. The hypercall spec (aka. the GHCI) says that the MapGPA call is allowed to return partial progress in mapping this range and indicate that fact with a special error code. A guest that sees such partial progress is expected to retry the operation for the portion of the address range that was not completed.

Quoted 4 lines

Hyper-V does this partial completion dance when set_memory_decrypted() is called to "decrypt" swiotlb bounce buffers that can be up to 1GB in size. It is evidently the only VMM that does this, which is why nobody noticed this until now.

Sorry for late commenting.

Nit:

IMHO this patch is doing two separate things together:

1) convert tdx_enc_status_changed() to tdx_map_gpa() to take physical address.
2) Handle MapGPA() retry

The reason of doing 1), IIUC, is hidden in the second patch, that hyperv guest code is using vzalloc(). I.e., handle MapGPA() retry doesn't strictly require to change API to take PA rather than VA.

So to me it's better to split this into two patches and give properly justification to each of them.

Also, see below for the retry ...

Quoted 26 lines
---
 arch/x86/coco/tdx/tdx.c           | 64 +++++++++++++++++++++++++------
 arch/x86/include/asm/shared/tdx.h |  2 +
 2 files changed, 54 insertions(+), 12 deletions(-)

Changes in v10:
Dave kindly re-wrote the changelog. No other changes.

arch/x86/coco/tdx/tdx.c +2 -3
index 1d6b863c42b00..746075d20cd2d 100644
@@ -703,14 +703,15 @@ static bool tdx_cache_flush_required(void)
 }
 
 /*
- * Inform the VMM of the guest's intent for this physical page: shared with
- * the VMM or private to the guest.  The VMM is expected to change its mapping
- * of the page in response.
+ * Notify the VMM about page mapping conversion. More info about ABI
+ * can be found in TDX Guest-Host-Communication Interface (GHCI),

arch/x86/coco/tdx/tdx.c static bool tdx_cache_flush_required(void) · line 714

Quoted 10 lines
arch/x86/coco/tdx/tdx.c +5 -3
+ * section "TDG.VP.VMCALL<MapGPA>".
  */
-static bool tdx_enc_status_changed(unsigned long vaddr, int numpages, bool enc)
+static bool tdx_map_gpa(phys_addr_t start, phys_addr_t end, bool enc)
 {
-	phys_addr_t start = __pa(vaddr);
-	phys_addr_t end   = __pa(vaddr + numpages * PAGE_SIZE);
+	/* Retrying the hypercall a second time should succeed; use 3 just in case */
+	const int max_retries_per_page = 3;
+	int retry_count = 0;

... I tried to dig the full history, but sorry if I am still missing something.

Using 3 is fine if "Retrying the hypercall a second time should succeed" is always true. I assume this is because hyperV is able to handle large amount of pages in one call?

That being said, this is purely hypervisor implementation specific. Here IIUC Linux is trying to define a non-spec-based value to retry, which can happen to work for hyperv *current* implementation. I am not sure whether it's a good idea? For instance, what happens if hyperv is changed in the future to reduce the number of pages it can handle in one call?

Is there any hyperv specification to define how many pages it can handle in one call?

What's more, given this function only takes a random range of pages, it makes even more strange to use hard-coded retry here. Looks a more reasonable way is to let the caller who knows how many pages are going to be converted, and *ideally*, also knows the which hypervisor is running underneath, to determine how many pages to be converted in one call.

For instance, any hyperv specific guest code can safely assume hyperv is able to handle how many pages thus can determine how many pages to try in one call.

Just my 2cents. And feel free to ignore if all others are fine with the current solution in this patch.

Quoted 9 lines
arch/x86/coco/tdx/tdx.c +0 -2
 	if (!enc) {
 		/* Set the shared (decrypted) bits: */
@@ -718,12 +719,51 @@ static bool tdx_enc_status_changed(unsigned long vaddr, int numpages, bool enc)
 		end   |= cc_mkdec(0);
 	}
 
-	/*
-	 * Notify the VMM about page mapping conversion. More info about ABI

arch/x86/coco/tdx/tdx.c static bool tdx_enc_status_changed(unsigned long vaddr, int numpages, bool enc) · line 727

Quoted 10 lines
arch/x86/coco/tdx/tdx.c +6 -4
-	 * can be found in TDX Guest-Host-Communication Interface (GHCI),
-	 * section "TDG.VP.VMCALL<MapGPA>"
-	 */
-	if (_tdx_hypercall(TDVMCALL_MAP_GPA, start, end - start, 0, 0))
+	while (retry_count < max_retries_per_page) {
+		struct tdx_hypercall_args args = {
+			.r10 = TDX_HYPERCALL_STANDARD,
+			.r11 = TDVMCALL_MAP_GPA,
+			.r12 = start,
+			.r13 = end - start };

Nit:

Is it a better style to move '}' to a new line?

arch/x86/coco/tdx/tdx.c static bool tdx_enc_status_changed(unsigned long vaddr, int numpages, bool enc) · line 730

Quoted 3 lines
arch/x86/coco/tdx/tdx.c +3 -0
+
+		u64 map_fail_paddr;
+		u64 ret = __tdx_hypercall_ret(&args);

Nit:

Looks people prefers reverse-christmas-tree style. Perhaps separate the 'ret' declaration out.

Quoted 52 lines
arch/x86/coco/tdx/tdx.c +36 -0
+
+		if (ret != TDVMCALL_STATUS_RETRY)
+			return !ret;
+		/*
+		 * The guest must retry the operation for the pages in the
+		 * region starting at the GPA specified in R11. R11 comes
+		 * from the untrusted VMM. Sanity check it.
+		 */
+		map_fail_paddr = args.r11;
+		if (map_fail_paddr < start || map_fail_paddr >= end)
+			return false;
+
+		/* "Consume" a retry without forward progress */
+		if (map_fail_paddr == start) {
+			retry_count++;
+			continue;
+		}
+
+		start = map_fail_paddr;
+		retry_count = 0;
+	}
+
+	return false;
+}
+
+/*
+ * Inform the VMM of the guest's intent for this physical page: shared with
+ * the VMM or private to the guest.  The VMM is expected to change its mapping
+ * of the page in response.
+ */
+static bool tdx_enc_status_changed(unsigned long vaddr, int numpages, bool enc)
+{
+	phys_addr_t start = __pa(vaddr);
+	phys_addr_t end   = __pa(vaddr + numpages * PAGE_SIZE);
+
+	if (!tdx_map_gpa(start, end, enc))
 		return false;
 
 	/* shared->private conversion requires memory to be accepted before use */
arch/x86/include/asm/shared/tdx.h +2 -0
index 7513b3bb69b7e..22ee23a3f24a6 100644
@@ -24,6 +24,8 @@
 #define TDVMCALL_MAP_GPA		0x10001
 #define TDVMCALL_REPORT_FATAL_ERROR	0x10003
 
+#define TDVMCALL_STATUS_RETRY		1
+
 #ifndef __ASSEMBLY__
 
 /*
#5 lore raw

Re: [PATCH v10 1/2] x86/tdx: Retry partially-completed page conversion hypercalls

Quoted 28 lines · Kai
Quoted 13 lines · Dexuan Cui

TDX guest memory is private by default and the VMM may not access it. However, in cases where the guest needs to share data with the VMM, the guest and the VMM can coordinate to make memory shared between them.

The guest side of this protocol includes the "MapGPA" hypercall. This call takes a guest physical address range. The hypercall spec (aka. the GHCI) says that the MapGPA call is allowed to return partial progress in mapping this range and indicate that fact with a special error code. A guest that sees such partial progress is expected to retry the operation for the portion of the address range that was not completed.

Quoted 4 lines

Hyper-V does this partial completion dance when set_memory_decrypted() is called to "decrypt" swiotlb bounce buffers that can be up to 1GB in size. It is evidently the only VMM that does this, which is why nobody noticed this until now.

Sorry for late commenting.

Nit:

IMHO this patch is doing two separate things together:

1) convert tdx_enc_status_changed() to tdx_map_gpa() to take physical address.
2) Handle MapGPA() retry

Quoted 6 lines

The reason of doing 1), IIUC, is hidden in the second patch, that hyperv guest code is using vzalloc(). I.e., handle MapGPA() retry doesn't strictly require to change API to take PA rather than VA.

So to me it's better to split this into two patches and give properly justification to each of them.

Sorry I realized I missed one point in the retry logic below, so feel free to ignore this comment.

Quoted 61 lines

Also, see below for the retry ...

Quoted 26 lines
---
 arch/x86/coco/tdx/tdx.c           | 64 +++++++++++++++++++++++++------
 arch/x86/include/asm/shared/tdx.h |  2 +
 2 files changed, 54 insertions(+), 12 deletions(-)

Changes in v10:
Dave kindly re-wrote the changelog. No other changes.

arch/x86/coco/tdx/tdx.c +2 -3
index 1d6b863c42b00..746075d20cd2d 100644
@@ -703,14 +703,15 @@ static bool tdx_cache_flush_required(void)
 }
 
 /*
- * Inform the VMM of the guest's intent for this physical page: shared with
- * the VMM or private to the guest.  The VMM is expected to change its mapping
- * of the page in response.
+ * Notify the VMM about page mapping conversion. More info about ABI
+ * can be found in TDX Guest-Host-Communication Interface (GHCI),

arch/x86/coco/tdx/tdx.c static bool tdx_cache_flush_required(void) · line 714

Quoted 10 lines
arch/x86/coco/tdx/tdx.c +5 -3
+ * section "TDG.VP.VMCALL<MapGPA>".
  */
-static bool tdx_enc_status_changed(unsigned long vaddr, int numpages, bool enc)
+static bool tdx_map_gpa(phys_addr_t start, phys_addr_t end, bool enc)
 {
-	phys_addr_t start = __pa(vaddr);
-	phys_addr_t end   = __pa(vaddr + numpages * PAGE_SIZE);
+	/* Retrying the hypercall a second time should succeed; use 3 just in case */
+	const int max_retries_per_page = 3;
+	int retry_count = 0;

... I tried to dig the full history, but sorry if I am still missing something.

Using 3 is fine if "Retrying the hypercall a second time should succeed" is always true. I assume this is because hyperV is able to handle large amount of pages in one call?

That being said, this is purely hypervisor implementation specific. Here IIUC Linux is trying to define a non-spec-based value to retry, which can happen to work for hyperv *current* implementation. I am not sure whether it's a good idea? For instance, what happens if hyperv is changed in the future to reduce the number of pages it can handle in one call?

Is there any hyperv specification to define how many pages it can handle in one call?

What's more, given this function only takes a random range of pages, it makes even more strange to use hard-coded retry here. Looks a more reasonable way is to let the caller who knows how many pages are going to be converted, and *ideally*, also knows the which hypervisor is running underneath, to determine how many pages to be converted in one call.

Quoted 7 lines

For instance, any hyperv specific guest code can safely assume hyperv is able to handle how many pages thus can determine how many pages to try in one call.

Just my 2cents. And feel free to ignore if all others are fine with the current solution in this patch.

Ah, reading patch again I missed the fact that retry is only consumed when no
forward progress is made. If there's any page has been converted by the
hypervisor, retry_count is reset to 0, and this function will loop until all pages are converted. So feel free to ignore my above comments.

But is it better to explicitly call this out in the comment?

/*
* When the hypercall made no forward progress, retrying the hypercall a second
* time should succeed to make some progress. Use 3 just in case. */

Also, this "retry w/o forward progress" seems a little bit odd, i.e., it is a special case when hypervisor cannot convert any. Have you seen any in practice? How true can we say "retrying a second time should succeed"?

Quoted 53 lines
Quoted 52 lines
arch/x86/coco/tdx/tdx.c +36 -0
+
+		if (ret != TDVMCALL_STATUS_RETRY)
+			return !ret;
+		/*
+		 * The guest must retry the operation for the pages in the
+		 * region starting at the GPA specified in R11. R11 comes
+		 * from the untrusted VMM. Sanity check it.
+		 */
+		map_fail_paddr = args.r11;
+		if (map_fail_paddr < start || map_fail_paddr >= end)
+			return false;
+
+		/* "Consume" a retry without forward progress */
+		if (map_fail_paddr == start) {
+			retry_count++;
+			continue;
+		}
+
+		start = map_fail_paddr;
+		retry_count = 0;
+	}
+
+	return false;
+}
+
+/*
+ * Inform the VMM of the guest's intent for this physical page: shared with
+ * the VMM or private to the guest.  The VMM is expected to change its mapping
+ * of the page in response.
+ */
+static bool tdx_enc_status_changed(unsigned long vaddr, int numpages, bool enc)
+{
+	phys_addr_t start = __pa(vaddr);
+	phys_addr_t end   = __pa(vaddr + numpages * PAGE_SIZE);
+
+	if (!tdx_map_gpa(start, end, enc))
 		return false;
 
 	/* shared->private conversion requires memory to be accepted before use */
arch/x86/include/asm/shared/tdx.h +2 -0
index 7513b3bb69b7e..22ee23a3f24a6 100644
@@ -24,6 +24,8 @@
 #define TDVMCALL_MAP_GPA		0x10001
 #define TDVMCALL_REPORT_FATAL_ERROR	0x10003
 
+#define TDVMCALL_STATUS_RETRY		1
+
 #ifndef __ASSEMBLY__
 
 /*
Reviewed-by on v10 1/2
#6 lore raw

Re: [PATCH v10 1/2] x86/tdx: Retry partially-completed page conversion hypercalls

In reply to Dexuan Cui
Quoted 18 lines · Dexuan Cui

TDX guest memory is private by default and the VMM may not access it. However, in cases where the guest needs to share data with the VMM, the guest and the VMM can coordinate to make memory shared between them.

The guest side of this protocol includes the "MapGPA" hypercall. This call takes a guest physical address range. The hypercall spec (aka. the GHCI) says that the MapGPA call is allowed to return partial progress in mapping this range and indicate that fact with a special error code. A guest that sees such partial progress is expected to retry the operation for the portion of the address range that was not completed.

Hyper-V does this partial completion dance when set_memory_decrypted() is called to "decrypt" swiotlb bounce buffers that can be up to 1GB in size. It is evidently the only VMM that does this, which is why nobody noticed this until now.

Quoted 4 lines

Is there any reason that this needs to go into the stable trees? If so,
Fixes: and Cc:stable@ tags would be nice.

author maintainer
#7 lore raw

PATCH v10 2/2 x86/tdx: Support vmalloc() for tdx_enc_status_changed()

In reply to Dexuan Cui

When a TDX guest runs on Hyper-V, the hv_netvsc driver's netvsc_init_buf() allocates buffers using vzalloc(), and needs to share the buffers with the host OS by calling set_memory_decrypted(), which is not working for vmalloc() yet. Add the support by handling the pages one by one.

---
 arch/x86/coco/tdx/tdx.c | 36 ++++++++++++++++++++++++++++++------
 1 file changed, 30 insertions(+), 6 deletions(-)

Changes in v10:
Dave kindly improved tdx_enc_status_changed():
Call tdx_enc_status_changed_phys() only once.
Make the change concise and more readable
See https://lwn.net/ml/linux-kernel/[email protected]/

arch/x86/coco/tdx/tdx.c +30 -6
index 746075d20cd2d..38044bb32c498 100644
@@ -7,6 +7,7 @@
 #include <linux/cpufeature.h>
 #include <linux/export.h>
 #include <linux/io.h>
+#include <linux/mm.h>
 #include <asm/coco.h>
 #include <asm/tdx.h>
 #include <asm/vmx.h>
@@ -753,6 +754,19 @@ static bool tdx_map_gpa(phys_addr_t start, phys_addr_t end, bool enc)
 	return false;
 }
 
+static bool tdx_enc_status_changed_phys(phys_addr_t start, phys_addr_t end,
+					bool enc)
+{
+	if (!tdx_map_gpa(start, end, enc))
+		return false;
+
+	/* shared->private conversion requires memory to be accepted before use */
+	if (enc)
+		return tdx_accept_memory(start, end);
+
+	return true;
+}
+
 /*
  * Inform the VMM of the guest's intent for this physical page: shared with
  * the VMM or private to the guest.  The VMM is expected to change its mapping
@@ -760,15 +774,25 @@ static bool tdx_map_gpa(phys_addr_t start, phys_addr_t end, bool enc)
  */
 static bool tdx_enc_status_changed(unsigned long vaddr, int numpages, bool enc)
 {
-	phys_addr_t start = __pa(vaddr);
-	phys_addr_t end   = __pa(vaddr + numpages * PAGE_SIZE);
+	unsigned long start = vaddr;
+	unsigned long end = start + numpages * PAGE_SIZE;
+	unsigned long step = end - start;
+	unsigned long addr;
 
-	if (!tdx_map_gpa(start, end, enc))
+	if (offset_in_page(start) != 0)
 		return false;
 
-	/* shared->private conversion requires memory to be accepted before use */
-	if (enc)
-		return tdx_accept_memory(start, end);
+	/* Step through page-by-page for vmalloc() mappings: */
+	if (is_vmalloc_addr((void *)vaddr))
+		step = PAGE_SIZE;
+
+	for (addr = start; addr < end; addr += step) {
+		phys_addr_t start_pa = slow_virt_to_phys((void *)addr);
+		phys_addr_t end_pa   = start_pa + step;
+
+		if (!tdx_enc_status_changed_phys(start_pa, end_pa, enc))
+			return false;
+	}
 
 	return true;
 }
Signature · 1 line
2.25.1
maintainer Reviewed-by on v10 2/2
#8 lore raw

Re: [PATCH v10 2/2] x86/tdx: Support vmalloc() for tdx_enc_status_changed()

Quoted 11 lines · Dexuan Cui

When a TDX guest runs on Hyper-V, the hv_netvsc driver's netvsc_init_buf()
allocates buffers using vzalloc(), and needs to share the buffers with the
host OS by calling set_memory_decrypted(), which is not working for vmalloc() yet. Add the support by handling the pages one by one.

Reviewed-by Kuppuswamy Sathyanarayanan
Quoted 5 lines

<[email protected]>

---
 arch/x86/coco/tdx/tdx.c | 36 ++++++++++++++++++++++++++++++------

 1 file changed, 30 insertions(+), 6 deletions(-)

Only small comment, it is possible to have huge vmalloc's now, which would mean this would do 512 TDVMCALL_MAP_GPA calls instead of 1 when encountering a huge vmalloc mapping. If this used lookup_address() directly instead of slow_virt_to_phys(), it could catch this case. I don't think there are any cases of huge vmallocs today that would get passed into set_memory_en/decrypted(), so would only be future proofing.

author maintainer
#9 lore raw

RE: [PATCH v10 2/2] x86/tdx: Support vmalloc() for tdx_enc_status_changed()

Quoted 21 lines

From: Edgecombe, Rick P <[email protected]>
Sent: Tuesday, September 5, 2023 9:25 AM [...]

Quoted 11 lines · Dexuan Cui

When a TDX guest runs on Hyper-V, the hv_netvsc driver's netvsc_init_buf()
allocates buffers using vzalloc(), and needs to share the buffers with the
host OS by calling set_memory_decrypted(), which is not working for vmalloc() yet. Add the support by handling the pages one by one.

Reviewed-by Kuppuswamy Sathyanarayanan
Quoted 5 lines

<[email protected]>

---
 arch/x86/coco/tdx/tdx.c | 36 ++++++++++++++++++++++++++++++------

 1 file changed, 30 insertions(+), 6 deletions(-)
Quoted 1 line

Thanks!

Quoted 7 lines

Only small comment, it is possible to have huge vmalloc's now, which would mean this would do 512 TDVMCALL_MAP_GPA calls instead of 1 when encountering a huge vmalloc mapping. If this used lookup_address() directly instead of slow_virt_to_phys(), it could catch this case. I don't think there are any cases of huge vmallocs today that would get passed into set_memory_en/decrypted(), so would only be future proofing.

Thanks for the suggestion! So I think let's keep the code as-is for simplicity. We can enhance the code in future when it's necessary.

Acked-by on v10 2/2
#10 lore raw

Re: [PATCH v10 2/2] x86/tdx: Support vmalloc() for tdx_enc_status_changed()

In reply to Dexuan Cui
Quoted 10 lines · Dexuan Cui

When a TDX guest runs on Hyper-V, the hv_netvsc driver's netvsc_init_buf() allocates buffers using vzalloc(), and needs to share the buffers with the host OS by calling set_memory_decrypted(), which is not working for vmalloc() yet. Add the support by handling the pages one by one.

One nit below ...

[...]

arch/x86/coco/tdx/tdx.c static bool tdx_map_gpa(phys_addr_t start, phys_addr_t end, bool enc) · line 783

Quoted 4 lines
arch/x86/coco/tdx/tdx.c +1 -1
-	if (!tdx_map_gpa(start, end, enc))
+	if (offset_in_page(start) != 0)
 		return false;

... "!= 0" isn't needed.

Or should we even WARN()? IIUC by reaching here the caller should already verified both address and size are page aligned, but I didn't do full check.

Reviewed-by on v10 2/2
#11 lore raw

Re: [PATCH v10 2/2] x86/tdx: Support vmalloc() for tdx_enc_status_changed()

In reply to Dexuan Cui
Quoted 10 lines · Dexuan Cui

When a TDX guest runs on Hyper-V, the hv_netvsc driver's netvsc_init_buf() allocates buffers using vzalloc(), and needs to share the buffers with the host OS by calling set_memory_decrypted(), which is not working for vmalloc() yet. Add the support by handling the pages one by one.