Skip to contentSkip to conversation

s390/vmem: minor cleanup

PATCH Applied Alexander Gordeev 3 participants 10 messages Nov 10, 2020 – Nov 17, 2020

Subsystems (2)
lore
Share & tools
lore

Alexander Gordeev

author maintainer
#0 lore raw

cover s390/vmem: minor cleanup

Hi David,

Nothing really fancy here, just couple of cleanups

Alexander Gordeev (3):
s390/vmem: remove redundant check
s390/vmem: fix possible memory overwrite s390/vmem: make variable and function names consistent

 arch/s390/mm/vmem.c | 42 ++++++++++++++++++++++--------------------
 1 file changed, 22 insertions(+), 20 deletions(-)
Signature · 1 line
2.26.0
author maintainer
#1 lore raw

PATCH 1/3 s390/vmem: remove redundant check

---
 arch/s390/mm/vmem.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/s390/mm/vmem.c +1 -1
index b239f2ba93b09..56ab9bb770f3a 100644
@@ -223,7 +223,7 @@ static int __ref modify_pmd_table(pud_t *pud, unsigned long addr,
 		if (!add) {
 			if (pmd_none(*pmd))
 				continue;
-			if (pmd_large(*pmd) && !add) {
+			if (pmd_large(*pmd)) {
 				if (IS_ALIGNED(addr, PMD_SIZE) &&
 				    IS_ALIGNED(next, PMD_SIZE)) {
 					if (!direct)
Signature · 1 line
2.26.0
Reviewed-by on v1 1/3
#2 lore raw

Re: [PATCH 1/3] s390/vmem: remove redundant check

Quoted 11 lines · Alexander Gordeev
---
  arch/s390/mm/vmem.c | 2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)
arch/s390/mm/vmem.c +0 -0
index b239f2ba93b09..56ab9bb770f3a 100644

arch/s390/mm/vmem.c static int __ref modify_pmd_table(pud_t *pud, unsigned long addr, · line 229

Quoted 8 lines
arch/s390/mm/vmem.c +1 -1
@@ -223,7 +223,7 @@ static int __ref modify_pmd_table(pud_t *pud, unsigned long addr,
  		if (!add) {
  			if (pmd_none(*pmd))
  				continue;
-			if (pmd_large(*pmd) && !add) {
+			if (pmd_large(*pmd)) {
  				if (IS_ALIGNED(addr, PMD_SIZE) &&
  				    IS_ALIGNED(next, PMD_SIZE)) {
  					if (!direct)
Signature · 3 lines
Thanks,

David / dhildenb
maintainer Applied
#3 lore raw

Re: [PATCH 1/3] s390/vmem: remove redundant check

In reply to Alexander Gordeev
Quoted 10 lines · Alexander Gordeev
---
 arch/s390/mm/vmem.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/s390/mm/vmem.c +0 -0
index b239f2ba93b09..56ab9bb770f3a 100644

arch/s390/mm/vmem.c static int __ref modify_pmd_table(pud_t *pud, unsigned long addr, · line 226

Quoted 5 lines
arch/s390/mm/vmem.c +1 -1
@@ -223,7 +223,7 @@ static int __ref modify_pmd_table(pud_t *pud, unsigned long addr,
 		if (!add) {
 			if (pmd_none(*pmd))
 				continue;
-			if (pmd_large(*pmd) && !add) {
+			if (pmd_large(*pmd)) {

Applied, thanks!

author maintainer
#4 lore raw

PATCH 2/3 s390/vmem: fix possible memory overwrite

In reply to Alexander Gordeev

If for whatever reason the sub-PMD region to be used is less than struct page size (e.g in the future), then it is possible to overwrite beyond the region size.

---
 arch/s390/mm/vmem.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
arch/s390/mm/vmem.c +3 -1
index 56ab9bb770f3a..d7f25884061f4 100644
@@ -91,13 +91,15 @@ static void vmemmap_flush_unused_pmd(void)
 
 static void __vmemmap_use_sub_pmd(unsigned long start, unsigned long end)
 {
+	unsigned long size = min(end - start, sizeof(struct page));
+
 	/*
 	 * As we expect to add in the same granularity as we remove, it's
 	 * sufficient to mark only some piece used to block the memmap page from
 	 * getting removed (just in case the memmap never gets initialized,
 	 * e.g., because the memory block never gets onlined).
 	 */
-	memset(__va(start), 0, sizeof(struct page));
+	memset(__va(start), 0, size);
 }
 
 static void vmemmap_use_sub_pmd(unsigned long start, unsigned long end)
Signature · 1 line
2.26.0
#5 lore raw

Re: [PATCH 2/3] s390/vmem: fix possible memory overwrite

Quoted 21 lines · Alexander Gordeev

If for whatever reason the sub-PMD region to be used is less than struct page size (e.g in the future), then it is possible to overwrite beyond the region size.

---
  arch/s390/mm/vmem.c | 4 +++-
  1 file changed, 3 insertions(+), 1 deletion(-)
arch/s390/mm/vmem.c +2 -0
index 56ab9bb770f3a..d7f25884061f4 100644
@@ -91,13 +91,15 @@ static void vmemmap_flush_unused_pmd(void)
 
  static void __vmemmap_use_sub_pmd(unsigned long start, unsigned long end)
  {
+	unsigned long size = min(end - start, sizeof(struct page));
+
  	/*

arch/s390/mm/vmem.c static void vmemmap_flush_unused_pmd(void) · line 105

Quoted 10 lines
arch/s390/mm/vmem.c +1 -1
  	 * As we expect to add in the same granularity as we remove, it's
  	 * sufficient to mark only some piece used to block the memmap page from
  	 * getting removed (just in case the memmap never gets initialized,
  	 * e.g., because the memory block never gets onlined).
  	 */
-	memset(__va(start), 0, sizeof(struct page));
+	memset(__va(start), 0, size);
  }
 
  static void vmemmap_use_sub_pmd(unsigned long start, unsigned long end)

I don't really see a need for that. Can you spell out one possible configuration that would trigger that in the future? It's sounds very unlikely and I have the feeling there might be more to change at other points.

Signature · 3 lines
Thanks,

David / dhildenb
author maintainer
#6 lore raw

Re: [PATCH 2/3] s390/vmem: fix possible memory overwrite

Quoted 27 lines · David Hildenbrand
Quoted 26 lines · Alexander Gordeev

If for whatever reason the sub-PMD region to be used is less than struct page size (e.g in the future), then it is possible to overwrite beyond the region size.

---
 arch/s390/mm/vmem.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
arch/s390/mm/vmem.c +3 -1
index 56ab9bb770f3a..d7f25884061f4 100644
@@ -91,13 +91,15 @@ static void vmemmap_flush_unused_pmd(void)
 static void __vmemmap_use_sub_pmd(unsigned long start, unsigned long end)
 {
+	unsigned long size = min(end - start, sizeof(struct page));
+
 	/*
 	 * As we expect to add in the same granularity as we remove, it's
 	 * sufficient to mark only some piece used to block the memmap page from
 	 * getting removed (just in case the memmap never gets initialized,
 	 * e.g., because the memory block never gets onlined).
 	 */
-	memset(__va(start), 0, sizeof(struct page));
+	memset(__va(start), 0, size);
Quoted 8 lines

arch/s390/mm/vmem.c static void vmemmap_flush_unused_pmd(void)

Quoted 3 lines
arch/s390/mm/vmem.c +0 -0
 }
 static void vmemmap_use_sub_pmd(unsigned long start, unsigned long end)
 

I don't really see a need for that. Can you spell out one possible configuration that would trigger that in the future? It's sounds very unlikely and I have the feeling there might be more to change at other points.

No configuration in mind. But dependency on struct page is the only obstacle that prevents the whole thing to become generic (unless I am missing something). Moreover, the memset() would not be needed also - just a single non-PAGE_UNUSED word within [start..end) should be enough.

Quoted 5 lines
Signature · 3 lines
Thanks,

David / dhildenb
maintainer
#7 lore raw

Re: [PATCH 2/3] s390/vmem: fix possible memory overwrite

Quoted 21 lines · Alexander Gordeev
Quoted 12 lines
Quoted 12 lines

static void __vmemmap_use_sub_pmd(unsigned long start, unsigned long end) {

arch/s390/mm/vmem.c +3 -1
+	unsigned long size = min(end - start, sizeof(struct page));
+
 	/*
 	 * As we expect to add in the same granularity as we remove, it's
 	 * sufficient to mark only some piece used to block the memmap page from
 	 * getting removed (just in case the memmap never gets initialized,
 	 * e.g., because the memory block never gets onlined).
 	 */
-	memset(__va(start), 0, sizeof(struct page));
+	memset(__va(start), 0, size);
Quoted 8 lines
Quoted 3 lines

}
static void vmemmap_use_sub_pmd(unsigned long start, unsigned long end)

I don't really see a need for that. Can you spell out one possible configuration that would trigger that in the future? It's sounds very unlikely and I have the feeling there might be more to change at other points.

Quoted 5 lines

No configuration in mind. But dependency on struct page is the only obstacle that prevents the whole thing to become generic (unless I am missing something). Moreover, the memset() would not be needed also - just a single non-PAGE_UNUSED word within [start..end) should be enough.

Well, I agree with David here - so not applying.

author maintainer
#8 lore raw

PATCH 3/3 s390/vmem: make variable and function names consistent

In reply to Alexander Gordeev

Rename some variable and functions to better clarify what they are and what they do.

---
 arch/s390/mm/vmem.c | 36 ++++++++++++++++++------------------
 1 file changed, 18 insertions(+), 18 deletions(-)
arch/s390/mm/vmem.c +18 -18
index d7f25884061f4..4bb198db6aa01 100644
@@ -76,20 +76,20 @@ static void vmem_pte_free(unsigned long *table)
 
 /*
  * The unused vmemmap range, which was not yet memset(PAGE_UNUSED) ranges
- * from unused_pmd_start to next PMD_SIZE boundary.
+ * from unused_sub_pmd_start to next PMD_SIZE boundary.
  */
-static unsigned long unused_pmd_start;
+static unsigned long unused_sub_pmd_start;
 
-static void vmemmap_flush_unused_pmd(void)
+static void vmemmap_flush_unused_sub_pmd(void)
 {
-	if (!unused_pmd_start)
+	if (!unused_sub_pmd_start)
 		return;
-	memset(__va(unused_pmd_start), PAGE_UNUSED,
-	       ALIGN(unused_pmd_start, PMD_SIZE) - unused_pmd_start);
-	unused_pmd_start = 0;
+	memset(__va(unused_sub_pmd_start), PAGE_UNUSED,
+	       ALIGN(unused_sub_pmd_start, PMD_SIZE) - unused_sub_pmd_start);
+	unused_sub_pmd_start = 0;
 }
 
-static void __vmemmap_use_sub_pmd(unsigned long start, unsigned long end)
+static void vmemmap_mark_sub_pmd_used(unsigned long start, unsigned long end)
 {
 	unsigned long size = min(end - start, sizeof(struct page));
 
@@ -108,24 +108,24 @@ static void vmemmap_use_sub_pmd(unsigned long start, unsigned long end)
 	 * We only optimize if the new used range directly follows the
 	 * previously unused range (esp., when populating consecutive sections).
 	 */
-	if (unused_pmd_start == start) {
-		unused_pmd_start = end;
-		if (likely(IS_ALIGNED(unused_pmd_start, PMD_SIZE)))
-			unused_pmd_start = 0;
+	if (unused_sub_pmd_start == start) {
+		unused_sub_pmd_start = end;
+		if (likely(IS_ALIGNED(unused_sub_pmd_start, PMD_SIZE)))
+			unused_sub_pmd_start = 0;
 		return;
 	}
-	vmemmap_flush_unused_pmd();
-	__vmemmap_use_sub_pmd(start, end);
+	vmemmap_flush_unused_sub_pmd();
+	vmemmap_mark_sub_pmd_used(start, end);
 }
 
 static void vmemmap_use_new_sub_pmd(unsigned long start, unsigned long end)
 {
 	void *page = __va(ALIGN_DOWN(start, PMD_SIZE));
 
-	vmemmap_flush_unused_pmd();
+	vmemmap_flush_unused_sub_pmd();
 
 	/* Could be our memmap page is filled with PAGE_UNUSED already ... */
-	__vmemmap_use_sub_pmd(start, end);
+	vmemmap_mark_sub_pmd_used(start, end);
 
 	/* Mark the unused parts of the new memmap page PAGE_UNUSED. */
 	if (!IS_ALIGNED(start, PMD_SIZE))
@@ -136,7 +136,7 @@ static void vmemmap_use_new_sub_pmd(unsigned long start, unsigned long end)
 	 * unused range in the populated PMD.
 	 */
 	if (!IS_ALIGNED(end, PMD_SIZE))
-		unused_pmd_start = end;
+		unused_sub_pmd_start = end;
 }
 
 /* Returns true if the PMD is completely unused and can be freed. */
@@ -144,7 +144,7 @@ static bool vmemmap_unuse_sub_pmd(unsigned long start, unsigned long end)
 {
 	void *page = __va(ALIGN_DOWN(start, PMD_SIZE));
 
-	vmemmap_flush_unused_pmd();
+	vmemmap_flush_unused_sub_pmd();
 	memset(__va(start), PAGE_UNUSED, end - start);
 	return !memchr_inv(page, PAGE_UNUSED, PMD_SIZE);
 }
Signature · 1 line
2.26.0
maintainer Applied
#9 lore raw

Re: [PATCH 3/3] s390/vmem: make variable and function names consistent

Quoted 7 lines · Alexander Gordeev

Rename some variable and functions to better clarify what they are and what they do.

---
 arch/s390/mm/vmem.c | 36 ++++++++++++++++++------------------
 1 file changed, 18 insertions(+), 18 deletions(-)

Applied, thanks.