Skip to contentSkip to conversation

efi/x86: Some mixed mode fixes

PATCH In review Ard Biesheuvel 2 participants 6 messages Feb 13, 2020 – Feb 14, 2020

Subsystems (2)
lore
Share & tools
lore

Ard Biesheuvel

author maintainer
#0 lore raw

cover efi/x86: Some mixed mode fixes

Hans reports that v5.6-rc1 triggers a WARN() when booting in mixed mode, which appears to be due to the fact that a GUID argument allocated on the vmap'ed stack and passed by reference to the EFI runtime services is no longer aligned to its size, which is what the mixed mode runtime wrappers expect.

Let's fix this in a way that doesn't burden the caller, and copy the GUID into a suitably aligned buffer in the runtime wrappers. (#1)

Then, patches #2 and #3 are additional cleanup and correctness fixes for the mixed mode runtime wrappers.

Cc Hans de Goede <[email protected]>
Cc Arvind Sankar <[email protected]>
Cc Andy Lutomirski <[email protected]>

Ard Biesheuvel (3):
efi/x86: align GUIDs to their size in the mixed mode runtime wrapper efi/x86: remove support for EFI time and counter services in mixed
mode
efi/x86: Handle by-ref arguments covering multiple pages in mixed mode

 arch/x86/platform/efi/efi_64.c | 151 +++++++-------------
 1 file changed, 52 insertions(+), 99 deletions(-)
Signature · 1 line
2.17.1
author maintainer
#1 lore raw

PATCH 1/3 efi/x86: align GUIDs to their size in the mixed mode runtime wrapper

The mixed mode runtime wrappers are fragile when it comes to how the memory referred to by its pointer arguments are laid out in memory, due to the fact that it translates these addresses to physical addresses that the runtime services can dereference when running in 1:1 mode.

As a quick (i.e., backportable) fix, copy GUID pointer arguments to the local stack into a buffer that is naturally aligned to its size, so that is guaranteed to cover only one physical page.

Note that on x86, we cannot rely on the stack pointer being aligned the way the compiler expects, so we need to allocate an 8-byte aligned buffer of sufficient size, and copy the GUID into that buffer at an offset that is aligned to 16 bytes.

---
 arch/x86/platform/efi/efi_64.c | 25 ++++++++++++++++----
 1 file changed, 21 insertions(+), 4 deletions(-)
arch/x86/platform/efi/efi_64.c +21 -4
index fa8506e76bbe..543edfdcd1b9 100644
@@ -658,6 +658,8 @@ static efi_status_t
 efi_thunk_get_variable(efi_char16_t *name, efi_guid_t *vendor,
 		       u32 *attr, unsigned long *data_size, void *data)
 {
+	u8 buf[24] __aligned(8);
+	efi_guid_t *vnd = PTR_ALIGN((efi_guid_t *)buf, sizeof(*vnd));
 	efi_status_t status;
 	u32 phys_name, phys_vendor, phys_attr;
 	u32 phys_data_size, phys_data;
@@ -665,8 +667,10 @@ efi_thunk_get_variable(efi_char16_t *name, efi_guid_t *vendor,
 
 	spin_lock_irqsave(&efi_runtime_lock, flags);
 
+	*vnd = *vendor;
+
 	phys_data_size = virt_to_phys_or_null(data_size);
-	phys_vendor = virt_to_phys_or_null(vendor);
+	phys_vendor = virt_to_phys_or_null(vnd);
 	phys_name = virt_to_phys_or_null_size(name, efi_name_size(name));
 	phys_attr = virt_to_phys_or_null(attr);
 	phys_data = virt_to_phys_or_null_size(data, *data_size);
@@ -683,14 +687,18 @@ static efi_status_t
 efi_thunk_set_variable(efi_char16_t *name, efi_guid_t *vendor,
 		       u32 attr, unsigned long data_size, void *data)
 {
+	u8 buf[24] __aligned(8);
+	efi_guid_t *vnd = PTR_ALIGN((efi_guid_t *)buf, sizeof(*vnd));
 	u32 phys_name, phys_vendor, phys_data;
 	efi_status_t status;
 	unsigned long flags;
 
 	spin_lock_irqsave(&efi_runtime_lock, flags);
 
+	*vnd = *vendor;
+
 	phys_name = virt_to_phys_or_null_size(name, efi_name_size(name));
-	phys_vendor = virt_to_phys_or_null(vendor);
+	phys_vendor = virt_to_phys_or_null(vnd);
 	phys_data = virt_to_phys_or_null_size(data, data_size);
 
 	/* If data_size is > sizeof(u32) we've got problems */
@@ -707,6 +715,8 @@ efi_thunk_set_variable_nonblocking(efi_char16_t *name, efi_guid_t *vendor,
 				   u32 attr, unsigned long data_size,
 				   void *data)
 {
+	u8 buf[24] __aligned(8);
+	efi_guid_t *vnd = PTR_ALIGN((efi_guid_t *)buf, sizeof(*vnd));
 	u32 phys_name, phys_vendor, phys_data;
 	efi_status_t status;
 	unsigned long flags;
@@ -714,8 +724,10 @@ efi_thunk_set_variable_nonblocking(efi_char16_t *name, efi_guid_t *vendor,
 	if (!spin_trylock_irqsave(&efi_runtime_lock, flags))
 		return EFI_NOT_READY;
 
+	*vnd = *vendor;
+
 	phys_name = virt_to_phys_or_null_size(name, efi_name_size(name));
-	phys_vendor = virt_to_phys_or_null(vendor);
+	phys_vendor = virt_to_phys_or_null(vnd);
 	phys_data = virt_to_phys_or_null_size(data, data_size);
 
 	/* If data_size is > sizeof(u32) we've got problems */
@@ -732,14 +744,18 @@ efi_thunk_get_next_variable(unsigned long *name_size,
 			    efi_char16_t *name,
 			    efi_guid_t *vendor)
 {
+	u8 buf[24] __aligned(8);
+	efi_guid_t *vnd = PTR_ALIGN((efi_guid_t *)buf, sizeof(*vnd));
 	efi_status_t status;
 	u32 phys_name_size, phys_name, phys_vendor;
 	unsigned long flags;
 
 	spin_lock_irqsave(&efi_runtime_lock, flags);
 
+	*vnd = *vendor;
+
 	phys_name_size = virt_to_phys_or_null(name_size);
-	phys_vendor = virt_to_phys_or_null(vendor);
+	phys_vendor = virt_to_phys_or_null(vnd);
 	phys_name = virt_to_phys_or_null_size(name, *name_size);
 
 	status = efi_thunk(get_next_variable, phys_name_size,
@@ -747,6 +763,7 @@ efi_thunk_get_next_variable(unsigned long *name_size,
 
 	spin_unlock_irqrestore(&efi_runtime_lock, flags);
 
+	*vendor = *vnd;
 	return status;
 }
 
Signature · 1 line
2.17.1
author maintainer
#4 lore raw

PATCH 2/3 efi/x86: remove support for EFI time and counter services in mixed mode

In reply to Ard Biesheuvel

Mixed mode calls at runtime are rather tricky with vmap'ed stack, as we can no longer assume that data passed in by the callers of the EFI runtime wrapper routines is contiguous in physical memory.

We need to fix this, but before we do, let's drop the implementations of routines that we know are never used on x86, i.e., the RTC related ones. Given that UEFI rev2.8 permits any runtime service to return EFI_UNSUPPORTED at runtime, let's return that instead.

As get_next_high_mono_count() is never used at all, even on other architectures, let's make that return EFI_UNSUPPORTED as well.

---
 arch/x86/platform/efi/efi_64.c | 81 ++------------------
 1 file changed, 5 insertions(+), 76 deletions(-)
arch/x86/platform/efi/efi_64.c +5 -76
index 543edfdcd1b9..ae398587f264 100644
@@ -568,85 +568,25 @@ efi_thunk_set_virtual_address_map(unsigned long memory_map_size,
 
 static efi_status_t efi_thunk_get_time(efi_time_t *tm, efi_time_cap_t *tc)
 {
-	efi_status_t status;
-	u32 phys_tm, phys_tc;
-	unsigned long flags;
-
-	spin_lock(&rtc_lock);
-	spin_lock_irqsave(&efi_runtime_lock, flags);
-
-	phys_tm = virt_to_phys_or_null(tm);
-	phys_tc = virt_to_phys_or_null(tc);
-
-	status = efi_thunk(get_time, phys_tm, phys_tc);
-
-	spin_unlock_irqrestore(&efi_runtime_lock, flags);
-	spin_unlock(&rtc_lock);
-
-	return status;
+	return EFI_UNSUPPORTED;
 }
 
 static efi_status_t efi_thunk_set_time(efi_time_t *tm)
 {
-	efi_status_t status;
-	u32 phys_tm;
-	unsigned long flags;
-
-	spin_lock(&rtc_lock);
-	spin_lock_irqsave(&efi_runtime_lock, flags);
-
-	phys_tm = virt_to_phys_or_null(tm);
-
-	status = efi_thunk(set_time, phys_tm);
-
-	spin_unlock_irqrestore(&efi_runtime_lock, flags);
-	spin_unlock(&rtc_lock);
-
-	return status;
+	return EFI_UNSUPPORTED;
 }
 
 static efi_status_t
 efi_thunk_get_wakeup_time(efi_bool_t *enabled, efi_bool_t *pending,
 			  efi_time_t *tm)
 {
-	efi_status_t status;
-	u32 phys_enabled, phys_pending, phys_tm;
-	unsigned long flags;
-
-	spin_lock(&rtc_lock);
-	spin_lock_irqsave(&efi_runtime_lock, flags);
-
-	phys_enabled = virt_to_phys_or_null(enabled);
-	phys_pending = virt_to_phys_or_null(pending);
-	phys_tm = virt_to_phys_or_null(tm);
-
-	status = efi_thunk(get_wakeup_time, phys_enabled,
-			     phys_pending, phys_tm);
-
-	spin_unlock_irqrestore(&efi_runtime_lock, flags);
-	spin_unlock(&rtc_lock);
-
-	return status;
+	return EFI_UNSUPPORTED;
 }
 
 static efi_status_t
 efi_thunk_set_wakeup_time(efi_bool_t enabled, efi_time_t *tm)
 {
-	efi_status_t status;
-	u32 phys_tm;
-	unsigned long flags;
-
-	spin_lock(&rtc_lock);
-	spin_lock_irqsave(&efi_runtime_lock, flags);
-
-	phys_tm = virt_to_phys_or_null(tm);
-
-	status = efi_thunk(set_wakeup_time, enabled, phys_tm);
-
-	spin_unlock_irqrestore(&efi_runtime_lock, flags);
-	spin_unlock(&rtc_lock);
-
-	return status;
+	return EFI_UNSUPPORTED;
 }
 
 static unsigned long efi_name_size(efi_char16_t *name)
@@ -770,18 +710,7 @@ efi_thunk_get_next_variable(unsigned long *name_size,
 static efi_status_t
 efi_thunk_get_next_high_mono_count(u32 *count)
 {
-	efi_status_t status;
-	u32 phys_count;
-	unsigned long flags;
-
-	spin_lock_irqsave(&efi_runtime_lock, flags);
-
-	phys_count = virt_to_phys_or_null(count);
-	status = efi_thunk(get_next_high_mono_count, phys_count);
-
-	spin_unlock_irqrestore(&efi_runtime_lock, flags);
-
-	return status;
+	return EFI_UNSUPPORTED;
 }
 
 static void
Signature · 1 line
2.17.1
author maintainer
#5 lore raw

PATCH 3/3 efi/x86: Handle by-ref arguments covering multiple pages in mixed mode

In reply to Ard Biesheuvel

We currently apply some stringent checks on the variable name and buffer arguments passed to the EFI runtime services, and WARN() if they are allocated in the vmalloc space and are not aligned to their size - which must be a power of two - since in that case, it is not guaranteed that they will not cross a page boundary.

However, we then simply proceed with the call, which could cause data corruption if the next physical page belongs to a mapping that is entirely unrelated.

Given that with vmap'ed stacks, this condition is much more likely to trigger, let's relax the condition a bit, but fail the runtime service call if it triggers.

---
 arch/x86/platform/efi/efi_64.c | 45 +++++++++++---------
 1 file changed, 26 insertions(+), 19 deletions(-)
arch/x86/platform/efi/efi_64.c +26 -19
index ae398587f264..d19a2edd63cb 100644
@@ -180,7 +180,7 @@ void efi_sync_low_kernel_mappings(void)
 static inline phys_addr_t
 virt_to_phys_or_null_size(void *va, unsigned long size)
 {
-	bool bad_size;
+	phys_addr_t pa;
 
 	if (!va)
 		return 0;
@@ -188,16 +188,13 @@ virt_to_phys_or_null_size(void *va, unsigned long size)
 	if (virt_addr_valid(va))
 		return virt_to_phys(va);
 
-	/*
-	 * A fully aligned variable on the stack is guaranteed not to
-	 * cross a page bounary. Try to catch strings on the stack by
-	 * checking that 'size' is a power of two.
-	 */
-	bad_size = size > PAGE_SIZE || !is_power_of_2(size);
+	pa = slow_virt_to_phys(va);
 
-	WARN_ON(!IS_ALIGNED((unsigned long)va, size) || bad_size);
+	/* check if the object crosses a page boundary */
+	if (WARN_ON((pa ^ (pa + size - 1)) & PAGE_MASK))
+		return 0;
 
-	return slow_virt_to_phys(va);
+	return pa;
 }
 
 #define virt_to_phys_or_null(addr)				\
@@ -615,8 +612,11 @@ efi_thunk_get_variable(efi_char16_t *name, efi_guid_t *vendor,
 	phys_attr = virt_to_phys_or_null(attr);
 	phys_data = virt_to_phys_or_null_size(data, *data_size);
 
-	status = efi_thunk(get_variable, phys_name, phys_vendor,
-			   phys_attr, phys_data_size, phys_data);
+	if (!phys_name || (data && !phys_data))
+		status = EFI_INVALID_PARAMETER;
+	else
+		status = efi_thunk(get_variable, phys_name, phys_vendor,
+				   phys_attr, phys_data_size, phys_data);
 
 	spin_unlock_irqrestore(&efi_runtime_lock, flags);
 
@@ -641,9 +641,11 @@ efi_thunk_set_variable(efi_char16_t *name, efi_guid_t *vendor,
 	phys_vendor = virt_to_phys_or_null(vnd);
 	phys_data = virt_to_phys_or_null_size(data, data_size);
 
-	/* If data_size is > sizeof(u32) we've got problems */
-	status = efi_thunk(set_variable, phys_name, phys_vendor,
-			   attr, data_size, phys_data);
+	if (!phys_name || !phys_data)
+		status = EFI_INVALID_PARAMETER;
+	else
+		status = efi_thunk(set_variable, phys_name, phys_vendor,
+				   attr, data_size, phys_data);
 
 	spin_unlock_irqrestore(&efi_runtime_lock, flags);
 
@@ -670,9 +672,11 @@ efi_thunk_set_variable_nonblocking(efi_char16_t *name, efi_guid_t *vendor,
 	phys_vendor = virt_to_phys_or_null(vnd);
 	phys_data = virt_to_phys_or_null_size(data, data_size);
 
-	/* If data_size is > sizeof(u32) we've got problems */
-	status = efi_thunk(set_variable, phys_name, phys_vendor,
-			   attr, data_size, phys_data);
+	if (!phys_name || !phys_data)
+		status = EFI_INVALID_PARAMETER;
+	else
+		status = efi_thunk(set_variable, phys_name, phys_vendor,
+				   attr, data_size, phys_data);
 
 	spin_unlock_irqrestore(&efi_runtime_lock, flags);
 
@@ -698,8 +702,11 @@ efi_thunk_get_next_variable(unsigned long *name_size,
 	phys_vendor = virt_to_phys_or_null(vnd);
 	phys_name = virt_to_phys_or_null_size(name, *name_size);
 
-	status = efi_thunk(get_next_variable, phys_name_size,
-			   phys_name, phys_vendor);
+	if (!phys_name)
+		status = EFI_INVALID_PARAMETER;
+	else
+		status = efi_thunk(get_next_variable, phys_name_size,
+				   phys_name, phys_vendor);
 
 	spin_unlock_irqrestore(&efi_runtime_lock, flags);
 
Signature · 1 line
2.17.1
Tested-by on v1 1/3
#2 lore raw

Re: [PATCH 1/3] efi/x86: align GUIDs to their size in the mixed mode runtime wrapper

In reply to Ard Biesheuvel

Hi,

Quoted 9 lines · Ard Biesheuvel

The mixed mode runtime wrappers are fragile when it comes to how the memory referred to by its pointer arguments are laid out in memory, due to the fact that it translates these addresses to physical addresses that the runtime services can dereference when running in 1:1 mode.

As a quick (i.e., backportable) fix, copy GUID pointer arguments to the local stack into a buffer that is naturally aligned to its size, so that is guaranteed to cover only one physical page.

Quoted 7 lines

Note that on x86, we cannot rely on the stack pointer being aligned the way the compiler expects, so we need to allocate an 8-byte aligned buffer of sufficient size, and copy the GUID into that buffer at an offset that is aligned to 16 bytes.

I can confirm that this fixes the WARN_ON triggering I was seeing, thanks:

Regards,

Hans

Quoted 99 lines
---
  arch/x86/platform/efi/efi_64.c | 25 ++++++++++++++++----
  1 file changed, 21 insertions(+), 4 deletions(-)
arch/x86/platform/efi/efi_64.c +21 -4
index fa8506e76bbe..543edfdcd1b9 100644
@@ -658,6 +658,8 @@ static efi_status_t
  efi_thunk_get_variable(efi_char16_t *name, efi_guid_t *vendor,
  		       u32 *attr, unsigned long *data_size, void *data)
  {
+	u8 buf[24] __aligned(8);
+	efi_guid_t *vnd = PTR_ALIGN((efi_guid_t *)buf, sizeof(*vnd));
  	efi_status_t status;
  	u32 phys_name, phys_vendor, phys_attr;
  	u32 phys_data_size, phys_data;
@@ -665,8 +667,10 @@ efi_thunk_get_variable(efi_char16_t *name, efi_guid_t *vendor,
 
  	spin_lock_irqsave(&efi_runtime_lock, flags);
 
+	*vnd = *vendor;
+
  	phys_data_size = virt_to_phys_or_null(data_size);
-	phys_vendor = virt_to_phys_or_null(vendor);
+	phys_vendor = virt_to_phys_or_null(vnd);
  	phys_name = virt_to_phys_or_null_size(name, efi_name_size(name));
  	phys_attr = virt_to_phys_or_null(attr);
  	phys_data = virt_to_phys_or_null_size(data, *data_size);
@@ -683,14 +687,18 @@ static efi_status_t
  efi_thunk_set_variable(efi_char16_t *name, efi_guid_t *vendor,
  		       u32 attr, unsigned long data_size, void *data)
  {
+	u8 buf[24] __aligned(8);
+	efi_guid_t *vnd = PTR_ALIGN((efi_guid_t *)buf, sizeof(*vnd));
  	u32 phys_name, phys_vendor, phys_data;
  	efi_status_t status;
  	unsigned long flags;
 
  	spin_lock_irqsave(&efi_runtime_lock, flags);
 
+	*vnd = *vendor;
+
  	phys_name = virt_to_phys_or_null_size(name, efi_name_size(name));
-	phys_vendor = virt_to_phys_or_null(vendor);
+	phys_vendor = virt_to_phys_or_null(vnd);
  	phys_data = virt_to_phys_or_null_size(data, data_size);
 
  	/* If data_size is > sizeof(u32) we've got problems */
@@ -707,6 +715,8 @@ efi_thunk_set_variable_nonblocking(efi_char16_t *name, efi_guid_t *vendor,
  				   u32 attr, unsigned long data_size,
  				   void *data)
  {
+	u8 buf[24] __aligned(8);
+	efi_guid_t *vnd = PTR_ALIGN((efi_guid_t *)buf, sizeof(*vnd));
  	u32 phys_name, phys_vendor, phys_data;
  	efi_status_t status;
  	unsigned long flags;
@@ -714,8 +724,10 @@ efi_thunk_set_variable_nonblocking(efi_char16_t *name, efi_guid_t *vendor,
  	if (!spin_trylock_irqsave(&efi_runtime_lock, flags))
  		return EFI_NOT_READY;
 
+	*vnd = *vendor;
+
  	phys_name = virt_to_phys_or_null_size(name, efi_name_size(name));
-	phys_vendor = virt_to_phys_or_null(vendor);
+	phys_vendor = virt_to_phys_or_null(vnd);
  	phys_data = virt_to_phys_or_null_size(data, data_size);
 
  	/* If data_size is > sizeof(u32) we've got problems */
@@ -732,14 +744,18 @@ efi_thunk_get_next_variable(unsigned long *name_size,
  			    efi_char16_t *name,
  			    efi_guid_t *vendor)
  {
+	u8 buf[24] __aligned(8);
+	efi_guid_t *vnd = PTR_ALIGN((efi_guid_t *)buf, sizeof(*vnd));
  	efi_status_t status;
  	u32 phys_name_size, phys_name, phys_vendor;
  	unsigned long flags;
 
  	spin_lock_irqsave(&efi_runtime_lock, flags);
 
+	*vnd = *vendor;
+
  	phys_name_size = virt_to_phys_or_null(name_size);
-	phys_vendor = virt_to_phys_or_null(vendor);
+	phys_vendor = virt_to_phys_or_null(vnd);
  	phys_name = virt_to_phys_or_null_size(name, *name_size);
 
  	status = efi_thunk(get_next_variable, phys_name_size,
@@ -747,6 +763,7 @@ efi_thunk_get_next_variable(unsigned long *name_size,
 
  	spin_unlock_irqrestore(&efi_runtime_lock, flags);
 
+	*vendor = *vnd;
  	return status;
  }
 
author maintainer
#3 lore raw

Re: [PATCH 1/3] efi/x86: align GUIDs to their size in the mixed mode runtime wrapper

Quoted 18 lines · Hans de Goede

Hi,

Quoted 14 lines · Ard Biesheuvel

The mixed mode runtime wrappers are fragile when it comes to how the memory referred to by its pointer arguments are laid out in memory, due to the fact that it translates these addresses to physical addresses that the runtime services can dereference when running in 1:1 mode.

As a quick (i.e., backportable) fix, copy GUID pointer arguments to the local stack into a buffer that is naturally aligned to its size, so that is guaranteed to cover only one physical page.

Note that on x86, we cannot rely on the stack pointer being aligned the way the compiler expects, so we need to allocate an 8-byte aligned buffer of sufficient size, and copy the GUID into that buffer at an offset that is aligned to 16 bytes.

Quoted 8 lines
Quoted 2 lines

I can confirm that this fixes the WARN_ON triggering I was seeing, thanks:

Thanks Hans

I'll let these soak a bit in efi/urgent before I send them to tip.