[PATCH 0/6] Some trivial cleanups for shmem

[PATCH 0/6] Some trivial cleanups for shmem

[Baolin Wang]

Hi,

Patch 1 - Patch 5 do some trivial cleanups and refactoring for shmem.
Patch 6 adds myself as shmem reviewer.

Baolin Wang (6):
  mm: shmem: drop the unused macro
  mm: shmem: remove 'fadvise()' comments
  mm: shmem: remove duplicate error validation
  mm: shmem: change the return value of shmem_find_swap_entries()
  mm: shmem: factor out the within_size logic into a new helper
  MAINTAINERS: add myself as shmem reviewer

 MAINTAINERS |  1 +
 mm/shmem.c  | 76 ++++++++++++++++++++++++++---------------------------
 2 files changed, 39 insertions(+), 38 deletions(-)

-- 
2.39.3

[PATCH 1/6] mm: shmem: drop the unused macro

[Baolin Wang]

Drop the unused 'BLOCKS_PER_PAGE' macro.

Signed-off-by: Baolin Wang <baolin.wang@linux.alibaba.com>
---
 mm/shmem.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/mm/shmem.c b/mm/shmem.c
index 745f130bfb4c..ddf800357e7a 100644
--- a/mm/shmem.c
+++ b/mm/shmem.c
@@ -86,7 +86,6 @@ static struct vfsmount *shm_mnt __ro_after_init;
 
 #include "internal.h"
 
-#define BLOCKS_PER_PAGE  (PAGE_SIZE/512)
 #define VM_ACCT(size)    (PAGE_ALIGN(size) >> PAGE_SHIFT)
 
 /* Pretend that each entry is of this size in directory's i_size */
-- 
2.39.3

[PATCH 2/6] mm: shmem: remove 'fadvise()' comments

[Baolin Wang]

Similar to commit 255ff62d1586 ("docs: tmpfs: drop 'fadvise()' from the
documentation"), fadvise() has no HUGEPAGE advise currently. Remove
the confusing fadvise() comments.

Signed-off-by: Baolin Wang <baolin.wang@linux.alibaba.com>
---
 mm/shmem.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/mm/shmem.c b/mm/shmem.c
index ddf800357e7a..b7aef4f0a427 100644
--- a/mm/shmem.c
+++ b/mm/shmem.c
@@ -525,9 +525,9 @@ static bool shmem_confirm_swap(struct address_space *mapping,
  *	enables huge pages for the mount;
  * SHMEM_HUGE_WITHIN_SIZE:
  *	only allocate huge pages if the page will be fully within i_size,
- *	also respect fadvise()/madvise() hints;
+ *	also respect madvise() hints;
  * SHMEM_HUGE_ADVISE:
- *	only allocate huge pages if requested with fadvise()/madvise();
+ *	only allocate huge pages if requested with madvise();
  */
 
 #define SHMEM_HUGE_NEVER	0
-- 
2.39.3

[PATCH 3/6] mm: shmem: remove duplicate error validation

[Baolin Wang]

Remove duplicate error code checks for 'start' and 'end', as the
get_order_from_str() will only return -EINVAL if the cmdline string
is configured incorrectly.

Signed-off-by: Baolin Wang <baolin.wang@linux.alibaba.com>
---
 mm/shmem.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/mm/shmem.c b/mm/shmem.c
index b7aef4f0a427..b764ad336598 100644
--- a/mm/shmem.c
+++ b/mm/shmem.c
@@ -5650,19 +5650,19 @@ static int __init setup_thp_shmem(char *str)
 								 THP_ORDERS_ALL_FILE_DEFAULT);
 			}
 
-			if (start == -EINVAL) {
+			if (start < 0) {
 				pr_err("invalid size %s in thp_shmem boot parameter\n",
 				       start_size);
 				goto err;
 			}
 
-			if (end == -EINVAL) {
+			if (end < 0) {
 				pr_err("invalid size %s in thp_shmem boot parameter\n",
 				       end_size);
 				goto err;
 			}
 
-			if (start < 0 || end < 0 || start > end)
+			if (start > end)
 				goto err;
 
 			nr = end - start + 1;
-- 
2.39.3

[PATCH 4/6] mm: shmem: change the return value of shmem_find_swap_entries()

[Baolin Wang]

The shmem_find_swap_entries() originally returned the index corresponding to
the swap entry, but no callers used this return value. It should return the
number of entries that were found like other functions, which can be used by
the callers.

No functional changes.

Signed-off-by: Baolin Wang <baolin.wang@linux.alibaba.com>
---
 mm/shmem.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/mm/shmem.c b/mm/shmem.c
index b764ad336598..c243d814f2b0 100644
--- a/mm/shmem.c
+++ b/mm/shmem.c
@@ -1379,9 +1379,9 @@ static void shmem_evict_inode(struct inode *inode)
 #endif
 }
 
-static int shmem_find_swap_entries(struct address_space *mapping,
-				   pgoff_t start, struct folio_batch *fbatch,
-				   pgoff_t *indices, unsigned int type)
+static unsigned int shmem_find_swap_entries(struct address_space *mapping,
+				pgoff_t start, struct folio_batch *fbatch,
+				pgoff_t *indices, unsigned int type)
 {
 	XA_STATE(xas, &mapping->i_pages, start);
 	struct folio *folio;
@@ -1414,7 +1414,7 @@ static int shmem_find_swap_entries(struct address_space *mapping,
 	}
 	rcu_read_unlock();
 
-	return xas.xa_index;
+	return folio_batch_count(fbatch);
 }
 
 /*
@@ -1461,8 +1461,8 @@ static int shmem_unuse_inode(struct inode *inode, unsigned int type)
 
 	do {
 		folio_batch_init(&fbatch);
-		shmem_find_swap_entries(mapping, start, &fbatch, indices, type);
-		if (folio_batch_count(&fbatch) == 0) {
+		if (!shmem_find_swap_entries(mapping, start, &fbatch,
+					     indices, type)) {
 			ret = 0;
 			break;
 		}
-- 
2.39.3

[PATCH 5/6] mm: shmem: factor out the within_size logic into a new helper

[Baolin Wang]

Factor out the within_size logic into a new helper to remove duplicate
code.

Suggested-by: David Hildenbrand <david@redhat.com>
Signed-off-by: Baolin Wang <baolin.wang@linux.alibaba.com>
---
 mm/shmem.c | 53 +++++++++++++++++++++++++++--------------------------
 1 file changed, 27 insertions(+), 26 deletions(-)

diff --git a/mm/shmem.c b/mm/shmem.c
index c243d814f2b0..671f63063fd4 100644
--- a/mm/shmem.c
+++ b/mm/shmem.c
@@ -590,6 +590,28 @@ shmem_mapping_size_orders(struct address_space *mapping, pgoff_t index, loff_t w
 	return order > 0 ? BIT(order + 1) - 1 : 0;
 }
 
+static unsigned int shmem_get_orders_within_size(struct inode *inode,
+		unsigned long within_size_orders, pgoff_t index,
+		loff_t write_end)
+{
+	pgoff_t aligned_index;
+	unsigned long order;
+	loff_t i_size;
+
+	order = highest_order(within_size_orders);
+	while (within_size_orders) {
+		aligned_index = round_up(index + 1, 1 << order);
+		i_size = max(write_end, i_size_read(inode));
+		i_size = round_up(i_size, PAGE_SIZE);
+		if (i_size >> PAGE_SHIFT >= aligned_index)
+			return within_size_orders;
+
+		order = next_order(&within_size_orders, order);
+	}
+
+	return 0;
+}
+
 static unsigned int shmem_huge_global_enabled(struct inode *inode, pgoff_t index,
 					      loff_t write_end, bool shmem_huge_force,
 					      struct vm_area_struct *vma,
@@ -598,9 +620,6 @@ static unsigned int shmem_huge_global_enabled(struct inode *inode, pgoff_t index
 	unsigned int maybe_pmd_order = HPAGE_PMD_ORDER > MAX_PAGECACHE_ORDER ?
 		0 : BIT(HPAGE_PMD_ORDER);
 	unsigned long within_size_orders;
-	unsigned int order;
-	pgoff_t aligned_index;
-	loff_t i_size;
 
 	if (!S_ISREG(inode->i_mode))
 		return 0;
@@ -634,16 +653,11 @@ static unsigned int shmem_huge_global_enabled(struct inode *inode, pgoff_t index
 			within_size_orders = shmem_mapping_size_orders(inode->i_mapping,
 								       index, write_end);
 
-		order = highest_order(within_size_orders);
-		while (within_size_orders) {
-			aligned_index = round_up(index + 1, 1 << order);
-			i_size = max(write_end, i_size_read(inode));
-			i_size = round_up(i_size, PAGE_SIZE);
-			if (i_size >> PAGE_SHIFT >= aligned_index)
-				return within_size_orders;
+		within_size_orders = shmem_get_orders_within_size(inode, within_size_orders,
+								  index, write_end);
+		if (within_size_orders > 0)
+			return within_size_orders;
 
-			order = next_order(&within_size_orders, order);
-		}
 		fallthrough;
 	case SHMEM_HUGE_ADVISE:
 		if (vm_flags & VM_HUGEPAGE)
@@ -1756,10 +1770,7 @@ unsigned long shmem_allowable_huge_orders(struct inode *inode,
 	unsigned long mask = READ_ONCE(huge_shmem_orders_always);
 	unsigned long within_size_orders = READ_ONCE(huge_shmem_orders_within_size);
 	unsigned long vm_flags = vma ? vma->vm_flags : 0;
-	pgoff_t aligned_index;
 	unsigned int global_orders;
-	loff_t i_size;
-	int order;
 
 	if (thp_disabled_by_hw() || (vma && vma_thp_disabled(vma, vm_flags)))
 		return 0;
@@ -1785,17 +1796,7 @@ unsigned long shmem_allowable_huge_orders(struct inode *inode,
 		return READ_ONCE(huge_shmem_orders_inherit);
 
 	/* Allow mTHP that will be fully within i_size. */
-	order = highest_order(within_size_orders);
-	while (within_size_orders) {
-		aligned_index = round_up(index + 1, 1 << order);
-		i_size = round_up(i_size_read(inode), PAGE_SIZE);
-		if (i_size >> PAGE_SHIFT >= aligned_index) {
-			mask |= within_size_orders;
-			break;
-		}
-
-		order = next_order(&within_size_orders, order);
-	}
+	mask |= shmem_get_orders_within_size(inode, within_size_orders, index, 0);
 
 	if (vm_flags & VM_HUGEPAGE)
 		mask |= READ_ONCE(huge_shmem_orders_madvise);
-- 
2.39.3

[PATCH 6/6] MAINTAINERS: add myself as shmem reviewer

[Baolin Wang]

In the past year, I've primarily focused on shmem and added several features
to it, such as support for mTHP, large folio swap-out and swap-in support,
mTHP collapse support, skipping swapcache, and tmpfs support for large folios,
and so on. Meanwhile I've also been helping with testing and reviewing shmem
related patches.

So I am willing to continue assisting with testing and reviewing shmem
related patches. Let me be Cc'd on patches related to shmem.

Signed-off-by: Baolin Wang <baolin.wang@linux.alibaba.com>
---
 MAINTAINERS | 1 +
 1 file changed, 1 insertion(+)

diff --git a/MAINTAINERS b/MAINTAINERS
index 577592e3af82..7fe4ea237afe 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -23883,6 +23883,7 @@ F:	drivers/hwmon/tmp513.c
 
 TMPFS (SHMEM FILESYSTEM)
 M:	Hugh Dickins <hughd@google.com>
+R:	Baolin Wang <baolin.wang@linux.alibaba.com>
 L:	linux-mm@kvack.org
 S:	Maintained
 F:	include/linux/shmem_fs.h
-- 
2.39.3