virtiofs: keep a list of free dax memory ranges

Divide the dax memory range into fixed size ranges (2MB for now) and put
them in a list. This will track free ranges. Once an inode requires a
free range, we will take one from here and put it in interval-tree
of ranges assigned to inode.

Signed-off-by: Vivek Goyal <vgoyal@redhat.com>
Signed-off-by: Peng Tao <tao.peng@linux.alibaba.com>
Signed-off-by: Miklos Szeredi <mszeredi@redhat.com>

1 files changed, 90 insertions, 1 deletions

diff --git a/fs/fuse/dax.c b/fs/fuse/dax.c
index 9660d01f49a5..031106020f75 100644
--- a/fs/fuse/dax.c
+++ b/fs/fuse/dax.c
@@ -7,20 +7,104 @@
 #include "fuse_i.h"
 
 #include <linux/dax.h>
+#include <linux/pfn_t.h>
+
+/* Default memory range size, 2MB */
+#define FUSE_DAX_SHIFT	21
+#define FUSE_DAX_SZ	(1 << FUSE_DAX_SHIFT)
+#define FUSE_DAX_PAGES	(FUSE_DAX_SZ / PAGE_SIZE)
+
+/** Translation information for file offsets to DAX window offsets */
+struct fuse_dax_mapping {
+	/* Will connect in fcd->free_ranges to keep track of free memory */
+	struct list_head list;
+
+	/** Position in DAX window */
+	u64 window_offset;
+
+	/** Length of mapping, in bytes */
+	loff_t length;
+};
 
 struct fuse_conn_dax {
 	/* DAX device */
 	struct dax_device *dev;
+
+	/* DAX Window Free Ranges */
+	long nr_free_ranges;
+	struct list_head free_ranges;
 };
 
+static void fuse_free_dax_mem_ranges(struct list_head *mem_list)
+{
+	struct fuse_dax_mapping *range, *temp;
+
+	/* Free All allocated elements */
+	list_for_each_entry_safe(range, temp, mem_list, list) {
+		list_del(&range->list);
+		kfree(range);
+	}
+}
+
 void fuse_dax_conn_free(struct fuse_conn *fc)
 {
-	kfree(fc->dax);
+	if (fc->dax) {
+		fuse_free_dax_mem_ranges(&fc->dax->free_ranges);
+		kfree(fc->dax);
+	}
+}
+
+static int fuse_dax_mem_range_init(struct fuse_conn_dax *fcd)
+{
+	long nr_pages, nr_ranges;
+	void *kaddr;
+	pfn_t pfn;
+	struct fuse_dax_mapping *range;
+	int ret, id;
+	size_t dax_size = -1;
+	unsigned long i;
+
+	INIT_LIST_HEAD(&fcd->free_ranges);
+	id = dax_read_lock();
+	nr_pages = dax_direct_access(fcd->dev, 0, PHYS_PFN(dax_size), &kaddr,
+				     &pfn);
+	dax_read_unlock(id);
+	if (nr_pages < 0) {
+		pr_debug("dax_direct_access() returned %ld\n", nr_pages);
+		return nr_pages;
+	}
+
+	nr_ranges = nr_pages/FUSE_DAX_PAGES;
+	pr_debug("%s: dax mapped %ld pages. nr_ranges=%ld\n",
+		__func__, nr_pages, nr_ranges);
+
+	for (i = 0; i < nr_ranges; i++) {
+		range = kzalloc(sizeof(struct fuse_dax_mapping), GFP_KERNEL);
+		ret = -ENOMEM;
+		if (!range)
+			goto out_err;
+
+		/* TODO: This offset only works if virtio-fs driver is not
+		 * having some memory hidden at the beginning. This needs
+		 * better handling
+		 */
+		range->window_offset = i * FUSE_DAX_SZ;
+		range->length = FUSE_DAX_SZ;
+		list_add_tail(&range->list, &fcd->free_ranges);
+	}
+
+	fcd->nr_free_ranges = nr_ranges;
+	return 0;
+out_err:
+	/* Free All allocated elements */
+	fuse_free_dax_mem_ranges(&fcd->free_ranges);
+	return ret;
 }
 
 int fuse_dax_conn_alloc(struct fuse_conn *fc, struct dax_device *dax_dev)
 {
 	struct fuse_conn_dax *fcd;
+	int err;
 
 	if (!dax_dev)
 		return 0;
@@ -30,6 +114,11 @@ int fuse_dax_conn_alloc(struct fuse_conn *fc, struct dax_device *dax_dev)
 		return -ENOMEM;
 
 	fcd->dev = dax_dev;
+	err = fuse_dax_mem_range_init(fcd);
+	if (err) {
+		kfree(fcd);
+		return err;
+	}
 
 	fc->dax = fcd;
 	return 0;