Merge branch 'xfs-6.15-zoned_devices' into XFS-for-linus-6.15-merge

Merge Zoned allocator for XFS.

Signed-off-by: Carlos Maiolino <[email protected]>

Author: cmaiolino

Documentation/filesystems/iomap/design.rst

@@ -246,6 +246,10 @@ The fields are as follows:
    * **IOMAP_F_PRIVATE**: Starting with this value, the upper bits can
      be set by the filesystem for its own purposes.
 
+   * **IOMAP_F_ANON_WRITE**: Indicates that (write) I/O does not have a target
+     block assigned to it yet and the file system will do that in the bio
+     submission handler, splitting the I/O as needed.
+
    These flags can be set by iomap itself during file operations.
    The filesystem should supply an ``->iomap_end`` function if it needs
    to observe these flags:
@@ -352,6 +356,11 @@ operations:
    ``IOMAP_NOWAIT`` is often set on behalf of ``IOCB_NOWAIT`` or
    ``RWF_NOWAIT``.
 
+ * ``IOMAP_DONTCACHE`` is set when the caller wishes to perform a
+   buffered file I/O and would like the kernel to drop the pagecache
+   after the I/O completes, if it isn't already being used by another
+   thread.
+
 If it is necessary to read existing file contents from a `different
 <https://lore.kernel.org/all/[email protected]/>`_
 device or address range on a device, the filesystem should return that

Documentation/filesystems/iomap/operations.rst

@@ -131,6 +131,8 @@ These ``struct kiocb`` flags are significant for buffered I/O with iomap:
 
  * ``IOCB_NOWAIT``: Turns on ``IOMAP_NOWAIT``.
 
+ * ``IOCB_DONTCACHE``: Turns on ``IOMAP_DONTCACHE``.
+
 Internal per-Folio State
 ------------------------
 
@@ -283,7 +285,7 @@ The ``ops`` structure must be specified and is as follows:
  struct iomap_writeback_ops {
      int (*map_blocks)(struct iomap_writepage_ctx *wpc, struct inode *inode,
                        loff_t offset, unsigned len);
-     int (*prepare_ioend)(struct iomap_ioend *ioend, int status);
+     int (*submit_ioend)(struct iomap_writepage_ctx *wpc, int status);
      void (*discard_folio)(struct folio *folio, loff_t pos);
  };
 
@@ -306,13 +308,12 @@ The fields are as follows:
     purpose.
     This function must be supplied by the filesystem.
 
-  - ``prepare_ioend``: Enables filesystems to transform the writeback
-    ioend or perform any other preparatory work before the writeback I/O
-    is submitted.
+  - ``submit_ioend``: Allows the file systems to hook into writeback bio
+    submission.
     This might include pre-write space accounting updates, or installing
     a custom ``->bi_end_io`` function for internal purposes, such as
     deferring the ioend completion to a workqueue to run metadata update
-    transactions from process context.
+    transactions from process context before submitting the bio.
     This function is optional.
 
   - ``discard_folio``: iomap calls this function after ``->map_blocks``
@@ -341,7 +342,7 @@ This can happen in interrupt or process context, depending on the
 storage device.
 
 Filesystems that need to update internal bookkeeping (e.g. unwritten
-extent conversions) should provide a ``->prepare_ioend`` function to
+extent conversions) should provide a ``->submit_ioend`` function to
 set ``struct iomap_end::bio::bi_end_io`` to its own function.
 This function should call ``iomap_finish_ioends`` after finishing its
 own work (e.g. unwritten extent conversion).

fs/dax.c

@@ -1258,19 +1258,19 @@ static vm_fault_t dax_pmd_load_hole(struct xa_state *xas, struct vm_fault *vmf,
 }
 #endif /* CONFIG_FS_DAX_PMD */
 
-static s64 dax_unshare_iter(struct iomap_iter *iter)
+static int dax_unshare_iter(struct iomap_iter *iter)
 {
 	struct iomap *iomap = &iter->iomap;
 	const struct iomap *srcmap = iomap_iter_srcmap(iter);
 	loff_t copy_pos = iter->pos;
 	u64 copy_len = iomap_length(iter);
 	u32 mod;
 	int id = 0;
-	s64 ret = 0;
+	s64 ret;
 	void *daddr = NULL, *saddr = NULL;
 
 	if (!iomap_want_unshare_iter(iter))
-		return iomap_length(iter);
+		return iomap_iter_advance_full(iter);
 
 	/*
 	 * Extend the file range to be aligned to fsblock/pagesize, because
@@ -1300,14 +1300,14 @@ static s64 dax_unshare_iter(struct iomap_iter *iter)
 	if (ret < 0)
 		goto out_unlock;
 
-	if (copy_mc_to_kernel(daddr, saddr, copy_len) == 0)
-		ret = iomap_length(iter);
-	else
+	if (copy_mc_to_kernel(daddr, saddr, copy_len) != 0)
 		ret = -EIO;
 
 out_unlock:
 	dax_read_unlock(id);
-	return dax_mem2blk_err(ret);
+	if (ret < 0)
+		return dax_mem2blk_err(ret);
+	return iomap_iter_advance_full(iter);
 }
 
 int dax_file_unshare(struct inode *inode, loff_t pos, loff_t len,
@@ -1326,7 +1326,7 @@ int dax_file_unshare(struct inode *inode, loff_t pos, loff_t len,
 
 	iter.len = min(len, size - pos);
 	while ((ret = iomap_iter(&iter, ops)) > 0)
-		iter.processed = dax_unshare_iter(&iter);
+		iter.status = dax_unshare_iter(&iter);
 	return ret;
 }
 EXPORT_SYMBOL_GPL(dax_file_unshare);
@@ -1354,51 +1354,52 @@ static int dax_memzero(struct iomap_iter *iter, loff_t pos, size_t size)
 	return ret;
 }
 
-static s64 dax_zero_iter(struct iomap_iter *iter, bool *did_zero)
+static int dax_zero_iter(struct iomap_iter *iter, bool *did_zero)
 {
 	const struct iomap *iomap = &iter->iomap;
 	const struct iomap *srcmap = iomap_iter_srcmap(iter);
-	loff_t pos = iter->pos;
 	u64 length = iomap_length(iter);
-	s64 written = 0;
+	int ret;
 
 	/* already zeroed?  we're done. */
 	if (srcmap->type == IOMAP_HOLE || srcmap->type == IOMAP_UNWRITTEN)
-		return length;
+		return iomap_iter_advance(iter, &length);
 
 	/*
 	 * invalidate the pages whose sharing state is to be changed
 	 * because of CoW.
 	 */
 	if (iomap->flags & IOMAP_F_SHARED)
 		invalidate_inode_pages2_range(iter->inode->i_mapping,
-					      pos >> PAGE_SHIFT,
-					      (pos + length - 1) >> PAGE_SHIFT);
+				iter->pos >> PAGE_SHIFT,
+				(iter->pos + length - 1) >> PAGE_SHIFT);
 
 	do {
+		loff_t pos = iter->pos;
 		unsigned offset = offset_in_page(pos);
-		unsigned size = min_t(u64, PAGE_SIZE - offset, length);
 		pgoff_t pgoff = dax_iomap_pgoff(iomap, pos);
-		long rc;
 		int id;
 
+		length = min_t(u64, PAGE_SIZE - offset, length);
+
 		id = dax_read_lock();
-		if (IS_ALIGNED(pos, PAGE_SIZE) && size == PAGE_SIZE)
-			rc = dax_zero_page_range(iomap->dax_dev, pgoff, 1);
+		if (IS_ALIGNED(pos, PAGE_SIZE) && length == PAGE_SIZE)
+			ret = dax_zero_page_range(iomap->dax_dev, pgoff, 1);
 		else
-			rc = dax_memzero(iter, pos, size);
+			ret = dax_memzero(iter, pos, length);
 		dax_read_unlock(id);
 
-		if (rc < 0)
-			return rc;
-		pos += size;
-		length -= size;
-		written += size;
+		if (ret < 0)
+			return ret;
+
+		ret = iomap_iter_advance(iter, &length);
+		if (ret)
+			return ret;
 	} while (length > 0);
 
 	if (did_zero)
 		*did_zero = true;
-	return written;
+	return ret;
 }
 
 int dax_zero_range(struct inode *inode, loff_t pos, loff_t len, bool *did_zero,
@@ -1413,7 +1414,7 @@ int dax_zero_range(struct inode *inode, loff_t pos, loff_t len, bool *did_zero,
 	int ret;
 
 	while ((ret = iomap_iter(&iter, ops)) > 0)
-		iter.processed = dax_zero_iter(&iter, did_zero);
+		iter.status = dax_zero_iter(&iter, did_zero);
 	return ret;
 }
 EXPORT_SYMBOL_GPL(dax_zero_range);
@@ -1431,8 +1432,7 @@ int dax_truncate_page(struct inode *inode, loff_t pos, bool *did_zero,
 }
 EXPORT_SYMBOL_GPL(dax_truncate_page);
 
-static loff_t dax_iomap_iter(const struct iomap_iter *iomi,
-		struct iov_iter *iter)
+static int dax_iomap_iter(struct iomap_iter *iomi, struct iov_iter *iter)
 {
 	const struct iomap *iomap = &iomi->iomap;
 	const struct iomap *srcmap = iomap_iter_srcmap(iomi);
@@ -1451,8 +1451,10 @@ static loff_t dax_iomap_iter(const struct iomap_iter *iomi,
 		if (pos >= end)
 			return 0;
 
-		if (iomap->type == IOMAP_HOLE || iomap->type == IOMAP_UNWRITTEN)
-			return iov_iter_zero(min(length, end - pos), iter);
+		if (iomap->type == IOMAP_HOLE || iomap->type == IOMAP_UNWRITTEN) {
+			done = iov_iter_zero(min(length, end - pos), iter);
+			return iomap_iter_advance(iomi, &done);
+		}
 	}
 
 	/*
@@ -1485,7 +1487,7 @@ static loff_t dax_iomap_iter(const struct iomap_iter *iomi,
 	}
 
 	id = dax_read_lock();
-	while (pos < end) {
+	while ((pos = iomi->pos) < end) {
 		unsigned offset = pos & (PAGE_SIZE - 1);
 		const size_t size = ALIGN(length + offset, PAGE_SIZE);
 		pgoff_t pgoff = dax_iomap_pgoff(iomap, pos);
@@ -1535,18 +1537,16 @@ static loff_t dax_iomap_iter(const struct iomap_iter *iomi,
 			xfer = dax_copy_to_iter(dax_dev, pgoff, kaddr,
 					map_len, iter);
 
-		pos += xfer;
-		length -= xfer;
-		done += xfer;
-
-		if (xfer == 0)
+		length = xfer;
+		ret = iomap_iter_advance(iomi, &length);
+		if (!ret && xfer == 0)
 			ret = -EFAULT;
 		if (xfer < map_len)
 			break;
 	}
 	dax_read_unlock(id);
 
-	return done ? done : ret;
+	return ret;
 }
 
 /**
@@ -1586,7 +1586,7 @@ dax_iomap_rw(struct kiocb *iocb, struct iov_iter *iter,
 		iomi.flags |= IOMAP_NOWAIT;
 
 	while ((ret = iomap_iter(&iomi, ops)) > 0)
-		iomi.processed = dax_iomap_iter(&iomi, iter);
+		iomi.status = dax_iomap_iter(&iomi, iter);
 
 	done = iomi.pos - iocb->ki_pos;
 	iocb->ki_pos = iomi.pos;
@@ -1757,7 +1757,7 @@ static vm_fault_t dax_iomap_pte_fault(struct vm_fault *vmf, pfn_t *pfnp,
 
 	while ((error = iomap_iter(&iter, ops)) > 0) {
 		if (WARN_ON_ONCE(iomap_length(&iter) < PAGE_SIZE)) {
-			iter.processed = -EIO;	/* fs corruption? */
+			iter.status = -EIO;	/* fs corruption? */
 			continue;
 		}
 
@@ -1769,8 +1769,10 @@ static vm_fault_t dax_iomap_pte_fault(struct vm_fault *vmf, pfn_t *pfnp,
 			ret |= VM_FAULT_MAJOR;
 		}
 
-		if (!(ret & VM_FAULT_ERROR))
-			iter.processed = PAGE_SIZE;
+		if (!(ret & VM_FAULT_ERROR)) {
+			u64 length = PAGE_SIZE;
+			iter.status = iomap_iter_advance(&iter, &length);
+		}
 	}
 
 	if (iomap_errp)
@@ -1883,8 +1885,10 @@ static vm_fault_t dax_iomap_pmd_fault(struct vm_fault *vmf, pfn_t *pfnp,
 			continue; /* actually breaks out of the loop */
 
 		ret = dax_fault_iter(vmf, &iter, pfnp, &xas, &entry, true);
-		if (ret != VM_FAULT_FALLBACK)
-			iter.processed = PMD_SIZE;
+		if (ret != VM_FAULT_FALLBACK) {
+			u64 length = PMD_SIZE;
+			iter.status = iomap_iter_advance(&iter, &length);
+		}
 	}
 
 unlock_entry:
@@ -1999,20 +2003,21 @@ vm_fault_t dax_finish_sync_fault(struct vm_fault *vmf, unsigned int order,
 }
 EXPORT_SYMBOL_GPL(dax_finish_sync_fault);
 
-static loff_t dax_range_compare_iter(struct iomap_iter *it_src,
+static int dax_range_compare_iter(struct iomap_iter *it_src,
 		struct iomap_iter *it_dest, u64 len, bool *same)
 {
 	const struct iomap *smap = &it_src->iomap;
 	const struct iomap *dmap = &it_dest->iomap;
 	loff_t pos1 = it_src->pos, pos2 = it_dest->pos;
+	u64 dest_len;
 	void *saddr, *daddr;
 	int id, ret;
 
 	len = min(len, min(smap->length, dmap->length));
 
 	if (smap->type == IOMAP_HOLE && dmap->type == IOMAP_HOLE) {
 		*same = true;
-		return len;
+		goto advance;
 	}
 
 	if (smap->type == IOMAP_HOLE || dmap->type == IOMAP_HOLE) {
@@ -2035,7 +2040,13 @@ static loff_t dax_range_compare_iter(struct iomap_iter *it_src,
 	if (!*same)
 		len = 0;
 	dax_read_unlock(id);
-	return len;
+
+advance:
+	dest_len = len;
+	ret = iomap_iter_advance(it_src, &len);
+	if (!ret)
+		ret = iomap_iter_advance(it_dest, &dest_len);
+	return ret;
 
 out_unlock:
 	dax_read_unlock(id);
@@ -2058,15 +2069,15 @@ int dax_dedupe_file_range_compare(struct inode *src, loff_t srcoff,
 		.len		= len,
 		.flags		= IOMAP_DAX,
 	};
-	int ret, compared = 0;
+	int ret, status;
 
 	while ((ret = iomap_iter(&src_iter, ops)) > 0 &&
 	       (ret = iomap_iter(&dst_iter, ops)) > 0) {
-		compared = dax_range_compare_iter(&src_iter, &dst_iter,
+		status = dax_range_compare_iter(&src_iter, &dst_iter,
 				min(src_iter.len, dst_iter.len), same);
-		if (compared < 0)
+		if (status < 0)
 			return ret;
-		src_iter.processed = dst_iter.processed = compared;
+		src_iter.status = dst_iter.status = status;
 	}
 	return ret;
 }

fs/gfs2/bmap.c

@@ -1300,7 +1300,8 @@ static int gfs2_block_zero_range(struct inode *inode, loff_t from,
 				 unsigned int length)
 {
 	BUG_ON(current->journal_info);
-	return iomap_zero_range(inode, from, length, NULL, &gfs2_iomap_ops);
+	return iomap_zero_range(inode, from, length, NULL, &gfs2_iomap_ops,
+			NULL);
 }
 
 #define GFS2_JTRUNC_REVOKES 8192

fs/iomap/Makefile

@@ -12,6 +12,7 @@ iomap-y				+= trace.o \
 				   iter.o
 iomap-$(CONFIG_BLOCK)		+= buffered-io.o \
 				   direct-io.o \
+				   ioend.o \
 				   fiemap.o \
 				   seek.o
 iomap-$(CONFIG_SWAP)		+= swapfile.o