Skip to contentSkip to conversation

x86/boot/e820: Make __e820__mapped_all() external

RFC PATCH In review Aaron Tomlin 3 participants 8 messages Jun 11, 2022 – Sep 29, 2022 x86

lore
Share & tools
lore
#0 lore raw

RFC PATCH 1/3 x86/boot/e820: Make __e820__mapped_all() external

No functional change.

This patch allows __e820__mapped_all() to be available for external use, in preparation to enhance the error message generated by arch_rmrr_sanity_check().

---
 arch/x86/include/asm/e820/api.h | 1 +
 arch/x86/kernel/e820.c          | 2 +-
 2 files changed, 2 insertions(+), 1 deletion(-)
arch/x86/include/asm/e820/api.h +1 -0
index e8f58ddd06d9..bf78daa08897 100644
@@ -13,6 +13,7 @@ extern unsigned long pci_mem_start;
 extern bool e820__mapped_raw_any(u64 start, u64 end, enum e820_type type);
 extern bool e820__mapped_any(u64 start, u64 end, enum e820_type type);
 extern bool e820__mapped_all(u64 start, u64 end, enum e820_type type);
+struct e820_entry *__e820__mapped_all(u64 start, u64 end, enum e820_type type);
 
 extern void e820__range_add   (u64 start, u64 size, enum e820_type type);
 extern u64  e820__range_update(u64 start, u64 size, enum e820_type old_type, enum e820_type new_type);
arch/x86/kernel/e820.c +1 -1
index f267205f2d5a..09b1c863a766 100644
@@ -109,7 +109,7 @@ EXPORT_SYMBOL_GPL(e820__mapped_any);
  * Note: this function only works correctly once the E820 table is sorted and
  * not-overlapping (at least for the range specified), which is the case normally.
  */
-static struct e820_entry *__e820__mapped_all(u64 start, u64 end,
+struct e820_entry *__e820__mapped_all(u64 start, u64 end,
 					     enum e820_type type)
 {
 	int i;
Signature · 1 line
2.34.1
#1 lore raw

RFC PATCH 2/3 x86/boot/e820: Make e820_type_to_string() external

No functional change.

This patch allows e820_type_to_string() to be available for external use, in preparation to enhance the error message generated by arch_rmrr_sanity_check().

---
 arch/x86/include/asm/e820/api.h | 1 +
 arch/x86/kernel/e820.c          | 2 +-
 2 files changed, 2 insertions(+), 1 deletion(-)
arch/x86/include/asm/e820/api.h +1 -0
index bf78daa08897..ceb301e591de 100644
@@ -43,6 +43,7 @@ extern void e820__reallocate_tables(void);
 extern void e820__register_nosave_regions(unsigned long limit_pfn);
 
 extern int  e820__get_entry_type(u64 start, u64 end);
+const char *e820_type_to_string(struct e820_entry *entry);
 
 /*
  * Returns true iff the specified range [start,end) is completely contained inside
arch/x86/kernel/e820.c +1 -1
index 09b1c863a766..95b994cf80cd 100644
@@ -1071,7 +1071,7 @@ void __init e820__finish_early_params(void)
 	}
 }
 
-static const char *__init e820_type_to_string(struct e820_entry *entry)
+const char *__init e820_type_to_string(struct e820_entry *entry)
 {
 	switch (entry->type) {
 	case E820_TYPE_RESERVED_KERN:	/* Fall-through: */
Signature · 1 line
2.34.1
#2 lore raw

RFC PATCH 3/3 iommu/vt-d: Show region type in arch_rmrr_sanity_check()

In reply to Aaron Tomlin

This patch will attempt to describe the region type in the event that a given RMRR entry is not within a reserved region.

---
 arch/x86/include/asm/iommu.h | 9 ++++++---
 arch/x86/kernel/e820.c       | 5 +++--
 2 files changed, 9 insertions(+), 5 deletions(-)
arch/x86/include/asm/iommu.h +6 -3
index bf1ed2ddc74b..d21366644520 100644
@@ -17,12 +17,15 @@ arch_rmrr_sanity_check(struct acpi_dmar_reserved_memory *rmrr)
 {
 	u64 start = rmrr->base_address;
 	u64 end = rmrr->end_address + 1;
+	struct e820_entry *entry;
 
-	if (e820__mapped_all(start, end, E820_TYPE_RESERVED))
+	entry = __e820__mapped_all(start, end, 0);
+
+	if (entry && entry->type == E820_TYPE_RESERVED)
 		return 0;
 
-	pr_err(FW_BUG "No firmware reserved region can cover this RMRR [%#018Lx-%#018Lx], contact BIOS vendor for fixes\n",
-	       start, end - 1);
+	pr_err(FW_BUG "No firmware reserved region can cover this RMRR [%s: %#018Lx-%#018Lx], contact BIOS vendor for fixes\n",
+	       e820_type_to_string(entry), start, end - 1);
 	return -EINVAL;
 }
 
arch/x86/kernel/e820.c +3 -2
index 95b994cf80cd..165e9a444bb9 100644
@@ -1073,7 +1073,7 @@ void __init e820__finish_early_params(void)
 
 const char *__init e820_type_to_string(struct e820_entry *entry)
 {
-	switch (entry->type) {
+	switch (entry && entry->type) {
 	case E820_TYPE_RESERVED_KERN:	/* Fall-through: */
 	case E820_TYPE_RAM:		return "System RAM";
 	case E820_TYPE_ACPI:		return "ACPI Tables";
@@ -1083,8 +1083,9 @@ const char *__init e820_type_to_string(struct e820_entry *entry)
 	case E820_TYPE_PMEM:		return "Persistent Memory";
 	case E820_TYPE_RESERVED:	return "Reserved";
 	case E820_TYPE_SOFT_RESERVED:	return "Soft Reserved";
-	default:			return "Unknown E820 type";
+	default:			break;
 	}
+	return "Unknown E820 type";
 }
 
 static unsigned long __init e820_type_to_iomem_type(struct e820_entry *entry)
Signature · 1 line
2.34.1
#3 lore raw

Re: [RFC PATCH 3/3] iommu/vt-d: Show region type in arch_rmrr_sanity_check()

Quoted 2 lines · Aaron Tomlin

This patch will attempt to describe the region type in the event that a given RMRR entry is not within a reserved region.

Any thoughts?

Kind regards,

Signature · 1 line
Aaron Tomlin
#4 lore raw

Re: [RFC PATCH 3/3] iommu/vt-d: Show region type in arch_rmrr_sanity_check()

In reply to Aaron Tomlin
Quoted 2 lines · Aaron Tomlin

This patch will attempt to describe the region type in the event that a given RMRR entry is not within a reserved region.

Hmm, is this useful information for the user? You'd hope the firmware vendor knows the memory map already, but either way, is it particularly likely that anyone would be noticing and caring about this warning in a context where they couldn't just scroll further up the log and cross-reference the full memory map listing? If so, it might be worth clarifying what that use-case is, since as it stands there doesn't seem to be much justification for the "why" here.

Quoted 34 lines
---
  arch/x86/include/asm/iommu.h | 9 ++++++---
  arch/x86/kernel/e820.c       | 5 +++--
  2 files changed, 9 insertions(+), 5 deletions(-)
arch/x86/include/asm/iommu.h +6 -3
index bf1ed2ddc74b..d21366644520 100644
@@ -17,12 +17,15 @@ arch_rmrr_sanity_check(struct acpi_dmar_reserved_memory *rmrr)
  {
  	u64 start = rmrr->base_address;
  	u64 end = rmrr->end_address + 1;
+	struct e820_entry *entry;
 
-	if (e820__mapped_all(start, end, E820_TYPE_RESERVED))
+	entry = __e820__mapped_all(start, end, 0);
+
+	if (entry && entry->type == E820_TYPE_RESERVED)
  		return 0;
 
-	pr_err(FW_BUG "No firmware reserved region can cover this RMRR [%#018Lx-%#018Lx], contact BIOS vendor for fixes\n",
-	       start, end - 1);
+	pr_err(FW_BUG "No firmware reserved region can cover this RMRR [%s: %#018Lx-%#018Lx], contact BIOS vendor for fixes\n",
+	       e820_type_to_string(entry), start, end - 1);
  	return -EINVAL;
  }
 
arch/x86/kernel/e820.c +0 -0
index 95b994cf80cd..165e9a444bb9 100644

arch/x86/kernel/e820.c void __init e820__finish_early_params(void) · line 1076

Quoted 5 lines
arch/x86/kernel/e820.c +1 -1
@@ -1073,7 +1073,7 @@ void __init e820__finish_early_params(void)
 
  const char *__init e820_type_to_string(struct e820_entry *entry)
  {
-	switch (entry->type) {
+	switch (entry && entry->type) {

Have you tested this for anything other than E820_TYPE_RAM? I think sufficiently up-to-date compilers should warn you here anyway.

Robin.

Quoted 4 lines
arch/x86/kernel/e820.c +0 -0
  	case E820_TYPE_RESERVED_KERN:	/* Fall-through: */
  	case E820_TYPE_RAM:		return "System RAM";
  	case E820_TYPE_ACPI:		return "ACPI Tables";

arch/x86/kernel/e820.c const char *__init e820_type_to_string(struct e820_entry *entry) · line 1091

Quoted 10 lines
arch/x86/kernel/e820.c +2 -1
@@ -1083,8 +1083,9 @@ const char *__init e820_type_to_string(struct e820_entry *entry)
  	case E820_TYPE_PMEM:		return "Persistent Memory";
  	case E820_TYPE_RESERVED:	return "Reserved";
  	case E820_TYPE_SOFT_RESERVED:	return "Soft Reserved";
-	default:			return "Unknown E820 type";
+	default:			break;
  	}
+	return "Unknown E820 type";
  }
 
  static unsigned long __init e820_type_to_iomem_type(struct e820_entry *entry)

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu

#5 lore raw

Re: [RFC PATCH 3/3] iommu/vt-d: Show region type in arch_rmrr_sanity_check()

Quoted 11 lines · Robin Murphy
Quoted 2 lines · Aaron Tomlin

This patch will attempt to describe the region type in the event that a given RMRR entry is not within a reserved region.

Hmm, is this useful information for the user? You'd hope the firmware vendor knows the memory map already, but either way, is it particularly likely that anyone would be noticing and caring about this warning in a context where they couldn't just scroll further up the log and cross-reference the full memory map listing? If so, it might be worth clarifying what that use-case is, since as it stands there doesn't seem to be much justification for the "why" here.

Hi Robin,

Thanks for looking at this.

Honestly, the only justification for the modification/or proposed changes is to have more insight when this statement is provided in total isolation and the RAM map listing (as per e820__print_table()) is no longer available to reference.

Quoted 4 lines
Quoted 4 lines
arch/x86/kernel/e820.c +0 -0
index 95b994cf80cd..165e9a444bb9 100644
Quoted 8 lines

arch/x86/kernel/e820.c void __init e820__finish_early_params(void) · line 1075

Quoted 5 lines
arch/x86/kernel/e820.c +1 -1
@@ -1073,7 +1073,7 @@ void __init e820__finish_early_params(void)
  const char *__init e820_type_to_string(struct e820_entry *entry)
  {
-	switch (entry->type) {
+	switch (entry && entry->type) {

Have you tested this for anything other than E820_TYPE_RAM? I think sufficiently up-to-date compilers should warn you here anyway.

I have not.

Kind regards,

Signature · 1 line
Aaron Tomlin
#6 lore raw
In reply to Aaron Tomlin
Quoted 6 lines · Aaron Tomlin

No functional change.

This patch allows __e820__mapped_all() to be available for external use, in preparation to enhance the error message generated by arch_rmrr_sanity_check().

Any more feedback?

Kind regards,

Signature · 1 line
Aaron Tomlin
#7 lore raw
Quoted 9 lines · Aaron Tomlin
Quoted 6 lines · Aaron Tomlin

No functional change.

This patch allows __e820__mapped_all() to be available for external use, in preparation to enhance the error message generated by arch_rmrr_sanity_check().

Any more feedback?

[Adding [email protected] and [email protected] on Cc]

Signature · 1 line
Aaron Tomlin