diff mbox series

[RFC,4/4] nfs/blocklayout: Use bulk page allocation APIs

Message ID 20240619173929.177818-10-cel@kernel.org (mailing list archive)
State New
Headers show
Series Snapshot of fixes for SCSI PR key registration | expand

Commit Message

Chuck Lever June 19, 2024, 5:39 p.m. UTC 
From: Chuck Lever <chuck.lever@oracle.com>

nfs4_get_device_info() frequently requests more than a few pages
when provisioning a nfs4_deviceid_node object. Make this more
efficient by using alloc_pages_bulk_array(). This API is known to be
several times faster than an open-coded loop around alloc_page().

release_pages() is folio-enabled so it is also more efficient than
repeatedly invoking __free_pages().

Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
---
 fs/nfs/pnfs_dev.c | 15 ++++++---------
 1 file changed, 6 insertions(+), 9 deletions(-)

Comments

Christoph Hellwig June 20, 2024, 4:44 a.m. UTC | #1 
On Wed, Jun 19, 2024 at 01:39:34PM -0400, cel@kernel.org wrote:
> From: Chuck Lever <chuck.lever@oracle.com>
> 
> nfs4_get_device_info() frequently requests more than a few pages
> when provisioning a nfs4_deviceid_node object.

Hmm.  Looks like nfs4_get_device_info uses max_resp_sz to size
the buffer.  In theory the max_deviceinfo_size field in
struct pnfs_layoutdriver_type caps it, but that isn't actually set
anywhere at all.

I guess we can't really do anything to size this better, but at least
for blocklayout where we usually get a single device or a handful
of stripes this is quite an overallocation.

And all that is just to pass it to xdr_init_decode_pages, which
is using it basically as a scratchpad.
diff mbox series

Patch

diff --git a/fs/nfs/pnfs_dev.c b/fs/nfs/pnfs_dev.c
index 178001c90156..26a78d69acab 100644
--- a/fs/nfs/pnfs_dev.c
+++ b/fs/nfs/pnfs_dev.c
@@ -101,9 +101,8 @@  nfs4_get_device_info(struct nfs_server *server,
 	struct nfs4_deviceid_node *d = NULL;
 	struct pnfs_device *pdev = NULL;
 	struct page **pages = NULL;
+	int rc, i, max_pages;
 	u32 max_resp_sz;
-	int max_pages;
-	int rc, i;
 
 	/*
 	 * Use the session max response size as the basis for setting
@@ -125,11 +124,9 @@  nfs4_get_device_info(struct nfs_server *server,
 	if (!pages)
 		goto out_free_pdev;
 
-	for (i = 0; i < max_pages; i++) {
-		pages[i] = alloc_page(gfp_flags);
-		if (!pages[i])
-			goto out_free_pages;
-	}
+	i = alloc_pages_bulk_array(GFP_KERNEL, max_pages, pages);
+	if (i != max_pages)
+		goto out_free_pages;
 
 	memcpy(&pdev->dev_id, dev_id, sizeof(*dev_id));
 	pdev->layout_type = server->pnfs_curr_ld->id;
@@ -154,8 +151,8 @@  nfs4_get_device_info(struct nfs_server *server,
 		set_bit(NFS_DEVICEID_NOCACHE, &d->flags);
 
 out_free_pages:
-	while (--i >= 0)
-		__free_page(pages[i]);
+	if (i)
+		release_pages(pages, i);
 	kfree(pages);
 out_free_pdev:
 	kfree(pdev);