* [PATCH 01/21] pNFS: Fix CB_NOTIFY_DEVICEID CHANGE to consume ndc_immediate
2026-08-13 20:42 [PATCH 00/21] NFS: flexfiles device notifications and caching for wide striped layouts Benjamin Coddington
@ 2026-08-13 20:42 ` Benjamin Coddington
2026-08-13 20:42 ` [PATCH 02/21] NFSv4/flexfiles: Use the full 64-bit offset for read DS selection Benjamin Coddington
` (19 subsequent siblings)
20 siblings, 0 replies; 22+ messages in thread
From: Benjamin Coddington @ 2026-08-13 20:42 UTC (permalink / raw)
To: Trond Myklebust, Anna Schumaker
Cc: linux-nfs, Jonathan Curley, Mike Snitzer, Jeff Layton
decode_devicenotify_args() gated consuming the trailing ndc_immediate
boolean on cbd_layout_type instead of cbd_notify_type. cbd_layout_type
holds the pNFS layout type (e.g. 4 for flexfiles) and is never equal to
NOTIFY_DEVICEID4_CHANGE (2), so ndc_immediate was never consumed: it was
silently lost for a single CHANGE, and a multi-item cnda_changes<> array
misaligned after the first item and failed with NFS4ERR_INVAL/BADXDR.
Key the decode off the notification type.
Fixes: 1be5683b03a7 ("pnfs: CB_NOTIFY_DEVICEID")
Cc: stable@vger.kernel.org
Assisted-by: Claude:claude-opus-4-8
Signed-off-by: Benjamin Coddington <bcodding@hammerspace.com>
---
fs/nfs/callback_xdr.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/fs/nfs/callback_xdr.c b/fs/nfs/callback_xdr.c
index 4382baddc9ee..04af8b9078a0 100644
--- a/fs/nfs/callback_xdr.c
+++ b/fs/nfs/callback_xdr.c
@@ -312,7 +312,7 @@ __be32 decode_devicenotify_args(struct svc_rqst *rqstp,
memcpy(dev->cbd_dev_id.data, p, NFS4_DEVICEID4_SIZE);
p += XDR_QUADLEN(NFS4_DEVICEID4_SIZE);
- if (dev->cbd_layout_type == NOTIFY_DEVICEID4_CHANGE) {
+ if (dev->cbd_notify_type == NOTIFY_DEVICEID4_CHANGE) {
p = xdr_inline_decode(xdr, sizeof(uint32_t));
if (unlikely(p == NULL)) {
status = htonl(NFS4ERR_BADXDR);
--
2.53.0
^ permalink raw reply related [flat|nested] 22+ messages in thread* [PATCH 02/21] NFSv4/flexfiles: Use the full 64-bit offset for read DS selection
2026-08-13 20:42 [PATCH 00/21] NFS: flexfiles device notifications and caching for wide striped layouts Benjamin Coddington
2026-08-13 20:42 ` [PATCH 01/21] pNFS: Fix CB_NOTIFY_DEVICEID CHANGE to consume ndc_immediate Benjamin Coddington
@ 2026-08-13 20:42 ` Benjamin Coddington
2026-08-13 20:42 ` [PATCH 03/21] NFSv4/flexfiles: Bound page coalescing on the absolute stripe offset Benjamin Coddington
` (18 subsequent siblings)
20 siblings, 0 replies; 22+ messages in thread
From: Benjamin Coddington @ 2026-08-13 20:42 UTC (permalink / raw)
To: Trond Myklebust, Anna Schumaker
Cc: linux-nfs, Jonathan Curley, Mike Snitzer, Jeff Layton
The read data-server selection helpers declared their offset parameter
as u32, truncating the file offset before nfs4_ff_layout_calc_dss_id()
computes the stripe index. For striped reads at offsets >= 4 GiB this
probes DS availability and sizes coalescing against the wrong stripe's
data server; the read itself recomputes dss_id from the full offset, so
data is not corrupted. The write path already uses the full offset.
Widen the offset parameter to u64 in the five read-selection helpers.
Fixes: 4934ccbeaed3 ("NFSv4/flexfiles: Read path updates for striped layouts")
Cc: stable@vger.kernel.org
Assisted-by: Claude:claude-opus-4-8
Signed-off-by: Benjamin Coddington <bcodding@hammerspace.com>
---
fs/nfs/flexfilelayout/flexfilelayout.c | 10 +++++-----
1 file changed, 5 insertions(+), 5 deletions(-)
diff --git a/fs/nfs/flexfilelayout/flexfilelayout.c b/fs/nfs/flexfilelayout/flexfilelayout.c
index c4aa995026f6..cce56faa8b70 100644
--- a/fs/nfs/flexfilelayout/flexfilelayout.c
+++ b/fs/nfs/flexfilelayout/flexfilelayout.c
@@ -878,7 +878,7 @@ ff_layout_mark_ds_reachable(struct pnfs_layout_segment *lseg, u32 idx, u32 dss_i
static struct nfs4_pnfs_ds *
ff_layout_choose_ds_for_read(struct pnfs_layout_segment *lseg,
u32 start_idx, u32 *best_idx,
- u32 offset, u32 *dss_id,
+ u64 offset, u32 *dss_id,
bool check_device)
{
struct nfs4_ff_layout_segment *fls = FF_LAYOUT_LSEG(lseg);
@@ -914,7 +914,7 @@ ff_layout_choose_ds_for_read(struct pnfs_layout_segment *lseg,
static struct nfs4_pnfs_ds *
ff_layout_choose_any_ds_for_read(struct pnfs_layout_segment *lseg,
u32 start_idx, u32 *best_idx,
- u32 offset, u32 *dss_id)
+ u64 offset, u32 *dss_id)
{
return ff_layout_choose_ds_for_read(lseg, start_idx, best_idx,
offset, dss_id, false);
@@ -923,7 +923,7 @@ ff_layout_choose_any_ds_for_read(struct pnfs_layout_segment *lseg,
static struct nfs4_pnfs_ds *
ff_layout_choose_valid_ds_for_read(struct pnfs_layout_segment *lseg,
u32 start_idx, u32 *best_idx,
- u32 offset, u32 *dss_id)
+ u64 offset, u32 *dss_id)
{
return ff_layout_choose_ds_for_read(lseg, start_idx, best_idx,
offset, dss_id, true);
@@ -932,7 +932,7 @@ ff_layout_choose_valid_ds_for_read(struct pnfs_layout_segment *lseg,
static struct nfs4_pnfs_ds *
ff_layout_choose_best_ds_for_read(struct pnfs_layout_segment *lseg,
u32 start_idx, u32 *best_idx,
- u32 offset, u32 *dss_id)
+ u64 offset, u32 *dss_id)
{
struct nfs4_pnfs_ds *ds;
@@ -947,7 +947,7 @@ ff_layout_choose_best_ds_for_read(struct pnfs_layout_segment *lseg,
static struct nfs4_pnfs_ds *
ff_layout_get_ds_for_read(struct nfs_pageio_descriptor *pgio,
u32 *best_idx,
- u32 offset,
+ u64 offset,
u32 *dss_id)
{
struct pnfs_layout_segment *lseg = pgio->pg_lseg;
--
2.53.0
^ permalink raw reply related [flat|nested] 22+ messages in thread* [PATCH 03/21] NFSv4/flexfiles: Bound page coalescing on the absolute stripe offset
2026-08-13 20:42 [PATCH 00/21] NFS: flexfiles device notifications and caching for wide striped layouts Benjamin Coddington
2026-08-13 20:42 ` [PATCH 01/21] pNFS: Fix CB_NOTIFY_DEVICEID CHANGE to consume ndc_immediate Benjamin Coddington
2026-08-13 20:42 ` [PATCH 02/21] NFSv4/flexfiles: Use the full 64-bit offset for read DS selection Benjamin Coddington
@ 2026-08-13 20:42 ` Benjamin Coddington
2026-08-13 20:43 ` [PATCH 04/21] NFSv4/filelayout: Anchor page coalescing on pattern_offset Benjamin Coddington
` (17 subsequent siblings)
20 siblings, 0 replies; 22+ messages in thread
From: Benjamin Coddington @ 2026-08-13 20:42 UTC (permalink / raw)
To: Trond Myklebust, Anna Schumaker
Cc: linux-nfs, Jonathan Curley, Mike Snitzer, Jeff Layton
ff_layout_pg_test() was copied from the files layout, which partitions
data servers relative to pattern_offset and bounds coalescing with a
segment-relative offset. Flexfiles has no pattern_offset:
nfs4_ff_layout_calc_dss_id() partitions on the absolute file offset
(RFC 8435 Section 6). So when a layout segment's offset is not
stripe-unit-aligned, the segment-relative coalescing window is shifted
off the absolute stripe grid and a coalesced I/O can straddle a stripe
boundary, sending the bytes past it to the wrong data server.
Bound coalescing on the absolute offset to match calc_dss_id(). Aligned
segments, including all whole-file layouts, are unchanged.
Fixes: 4934ccbeaed3 ("NFSv4/flexfiles: Read path updates for striped layouts")
Cc: stable@vger.kernel.org
Assisted-by: Claude:claude-opus-4-8
Signed-off-by: Benjamin Coddington <bcodding@hammerspace.com>
---
fs/nfs/flexfilelayout/flexfilelayout.c | 7 +++----
1 file changed, 3 insertions(+), 4 deletions(-)
diff --git a/fs/nfs/flexfilelayout/flexfilelayout.c b/fs/nfs/flexfilelayout/flexfilelayout.c
index cce56faa8b70..692ad84047d4 100644
--- a/fs/nfs/flexfilelayout/flexfilelayout.c
+++ b/fs/nfs/flexfilelayout/flexfilelayout.c
@@ -996,7 +996,6 @@ ff_layout_pg_test(struct nfs_pageio_descriptor *pgio, struct nfs_page *prev,
unsigned int size;
u64 p_stripe, r_stripe;
u32 stripe_offset;
- u64 segment_offset = pgio->pg_lseg->pls_range.offset;
u32 stripe_unit = FF_LAYOUT_LSEG(pgio->pg_lseg)->stripe_unit;
/* calls nfs_generic_pg_test */
@@ -1008,8 +1007,8 @@ ff_layout_pg_test(struct nfs_pageio_descriptor *pgio, struct nfs_page *prev,
/* see if req and prev are in the same stripe */
if (prev) {
- p_stripe = (u64)req_offset(prev) - segment_offset;
- r_stripe = (u64)req_offset(req) - segment_offset;
+ p_stripe = (u64)req_offset(prev);
+ r_stripe = (u64)req_offset(req);
do_div(p_stripe, stripe_unit);
do_div(r_stripe, stripe_unit);
@@ -1018,7 +1017,7 @@ ff_layout_pg_test(struct nfs_pageio_descriptor *pgio, struct nfs_page *prev,
}
/* calculate remaining bytes in the current stripe */
- div_u64_rem((u64)req_offset(req) - segment_offset,
+ div_u64_rem((u64)req_offset(req),
stripe_unit,
&stripe_offset);
WARN_ON_ONCE(stripe_offset > stripe_unit);
--
2.53.0
^ permalink raw reply related [flat|nested] 22+ messages in thread* [PATCH 04/21] NFSv4/filelayout: Anchor page coalescing on pattern_offset
2026-08-13 20:42 [PATCH 00/21] NFS: flexfiles device notifications and caching for wide striped layouts Benjamin Coddington
` (2 preceding siblings ...)
2026-08-13 20:42 ` [PATCH 03/21] NFSv4/flexfiles: Bound page coalescing on the absolute stripe offset Benjamin Coddington
@ 2026-08-13 20:43 ` Benjamin Coddington
2026-08-13 20:43 ` [PATCH 05/21] NFSv4/flexfiles: Reference the device node across DS setup Benjamin Coddington
` (16 subsequent siblings)
20 siblings, 0 replies; 22+ messages in thread
From: Benjamin Coddington @ 2026-08-13 20:43 UTC (permalink / raw)
To: Trond Myklebust, Anna Schumaker
Cc: linux-nfs, Jonathan Curley, Mike Snitzer, Jeff Layton
filelayout_pg_test() bounds page coalescing to a single stripe unit
using an offset relative to pls_range.offset, but the data server is
selected by nfs4_fl_calc_j_index() using an offset relative to
pattern_offset. When a segment's pattern_offset and range offset are
not congruent modulo the stripe unit, the coalescing window is shifted
off the DS-selection grid, so a coalesced I/O can straddle a stripe
boundary and send the bytes past it to the wrong data server.
Anchor coalescing on pattern_offset to match nfs4_fl_calc_j_index().
Stripe-congruent segments, including the common whole-file pattern_offset
0 case, are unchanged.
Fixes: c6194271f94b ("pnfs: filelayout: support non page aligned layouts")
Cc: stable@vger.kernel.org
Assisted-by: Claude:claude-opus-4-8
Signed-off-by: Benjamin Coddington <bcodding@hammerspace.com>
---
fs/nfs/filelayout/filelayout.c | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/fs/nfs/filelayout/filelayout.c b/fs/nfs/filelayout/filelayout.c
index 72e20b56fbc7..0d53277f972e 100644
--- a/fs/nfs/filelayout/filelayout.c
+++ b/fs/nfs/filelayout/filelayout.c
@@ -796,7 +796,7 @@ filelayout_pg_test(struct nfs_pageio_descriptor *pgio, struct nfs_page *prev,
unsigned int size;
u64 p_stripe, r_stripe;
u32 stripe_offset;
- u64 segment_offset = pgio->pg_lseg->pls_range.offset;
+ u64 pattern_offset = FILELAYOUT_LSEG(pgio->pg_lseg)->pattern_offset;
u32 stripe_unit = FILELAYOUT_LSEG(pgio->pg_lseg)->stripe_unit;
/* calls nfs_generic_pg_test */
@@ -808,8 +808,8 @@ filelayout_pg_test(struct nfs_pageio_descriptor *pgio, struct nfs_page *prev,
/* see if req and prev are in the same stripe */
if (prev) {
- p_stripe = (u64)req_offset(prev) - segment_offset;
- r_stripe = (u64)req_offset(req) - segment_offset;
+ p_stripe = (u64)req_offset(prev) - pattern_offset;
+ r_stripe = (u64)req_offset(req) - pattern_offset;
do_div(p_stripe, stripe_unit);
do_div(r_stripe, stripe_unit);
@@ -818,7 +818,7 @@ filelayout_pg_test(struct nfs_pageio_descriptor *pgio, struct nfs_page *prev,
}
/* calculate remaining bytes in the current stripe */
- div_u64_rem((u64)req_offset(req) - segment_offset,
+ div_u64_rem((u64)req_offset(req) - pattern_offset,
stripe_unit,
&stripe_offset);
WARN_ON_ONCE(stripe_offset > stripe_unit);
--
2.53.0
^ permalink raw reply related [flat|nested] 22+ messages in thread* [PATCH 05/21] NFSv4/flexfiles: Reference the device node across DS setup
2026-08-13 20:42 [PATCH 00/21] NFS: flexfiles device notifications and caching for wide striped layouts Benjamin Coddington
` (3 preceding siblings ...)
2026-08-13 20:43 ` [PATCH 04/21] NFSv4/filelayout: Anchor page coalescing on pattern_offset Benjamin Coddington
@ 2026-08-13 20:43 ` Benjamin Coddington
2026-08-13 20:43 ` [PATCH 06/21] NFSv4/flexfiles: Carry the device node reference across each I/O Benjamin Coddington
` (15 subsequent siblings)
20 siblings, 0 replies; 22+ messages in thread
From: Benjamin Coddington @ 2026-08-13 20:43 UTC (permalink / raw)
To: Trond Myklebust, Anna Schumaker
Cc: linux-nfs, Jonathan Curley, Mike Snitzer, Jeff Layton
The flexfiles I/O setup paths read the mirror's pinned per-stripe device
node (mirror->dss[dss_id].mirror_ds) repeatedly and locklessly, relying
on the mirror's lifetime pin. To prepare for re-resolving that pointer
in place on CB_NOTIFY_DEVICEID CHANGE, readers must hold their own
reference on the node they are using rather than trusting the pin.
Replace ff_layout_init_mirror_ds() with ff_layout_get_mirror_ds(), which
resolves on first use as before but returns the node with its own
reference held. Thread the referenced node through
nfs4_ff_layout_prepare_ds() and the DS selection helpers so each setup
path snapshots the node once, and drop the reference when setup is done.
No functional change: the pointer is still resolved once and pinned for
the life of the mirror.
Assisted-by: Claude:claude-fable-5
Signed-off-by: Benjamin Coddington <bcodding@hammerspace.com>
---
fs/nfs/flexfilelayout/flexfilelayout.c | 140 ++++++++++++++--------
fs/nfs/flexfilelayout/flexfilelayout.h | 16 ++-
fs/nfs/flexfilelayout/flexfilelayoutdev.c | 81 ++++++++-----
3 files changed, 149 insertions(+), 88 deletions(-)
diff --git a/fs/nfs/flexfilelayout/flexfilelayout.c b/fs/nfs/flexfilelayout/flexfilelayout.c
index 692ad84047d4..2bdc74a9a3eb 100644
--- a/fs/nfs/flexfilelayout/flexfilelayout.c
+++ b/fs/nfs/flexfilelayout/flexfilelayout.c
@@ -875,7 +875,7 @@ ff_layout_mark_ds_reachable(struct pnfs_layout_segment *lseg, u32 idx, u32 dss_i
nfs4_mark_deviceid_available(devid);
}
-static struct nfs4_pnfs_ds *
+static struct nfs4_ff_layout_ds *
ff_layout_choose_ds_for_read(struct pnfs_layout_segment *lseg,
u32 start_idx, u32 *best_idx,
u64 offset, u32 *dss_id,
@@ -883,7 +883,9 @@ ff_layout_choose_ds_for_read(struct pnfs_layout_segment *lseg,
{
struct nfs4_ff_layout_segment *fls = FF_LAYOUT_LSEG(lseg);
struct nfs4_ff_layout_mirror *mirror;
- struct nfs4_pnfs_ds *ds = ERR_PTR(-EAGAIN);
+ struct nfs4_ff_layout_ds *mirror_ds;
+ struct nfs4_ff_layout_ds *ret = ERR_PTR(-EAGAIN);
+ struct nfs4_pnfs_ds *ds;
u32 idx;
/* mirrors are initially sorted by efficiency */
@@ -893,25 +895,32 @@ ff_layout_choose_ds_for_read(struct pnfs_layout_segment *lseg,
fls->stripe_unit,
fls->mirror_array[idx]->dss_count,
offset);
- ds = nfs4_ff_layout_prepare_ds(lseg, mirror, *dss_id, false);
- if (IS_ERR(ds))
+ mirror_ds = ff_layout_get_mirror_ds(lseg->pls_layout, mirror,
+ *dss_id);
+ ds = nfs4_ff_layout_prepare_ds(lseg, mirror, mirror_ds,
+ *dss_id, false);
+ if (IS_ERR(ds)) {
+ nfs4_ff_layout_put_deviceid(mirror_ds);
+ ret = ERR_CAST(ds);
continue;
+ }
if (check_device &&
- nfs4_test_deviceid_unavailable(&mirror->dss[*dss_id].mirror_ds->id_node)) {
+ nfs4_test_deviceid_unavailable(&mirror_ds->id_node)) {
+ nfs4_ff_layout_put_deviceid(mirror_ds);
// reinitialize the error state in case if this is the last iteration
- ds = ERR_PTR(-EINVAL);
+ ret = ERR_PTR(-EINVAL);
continue;
}
*best_idx = idx;
- break;
+ return mirror_ds;
}
- return ds;
+ return ret;
}
-static struct nfs4_pnfs_ds *
+static struct nfs4_ff_layout_ds *
ff_layout_choose_any_ds_for_read(struct pnfs_layout_segment *lseg,
u32 start_idx, u32 *best_idx,
u64 offset, u32 *dss_id)
@@ -920,7 +929,7 @@ ff_layout_choose_any_ds_for_read(struct pnfs_layout_segment *lseg,
offset, dss_id, false);
}
-static struct nfs4_pnfs_ds *
+static struct nfs4_ff_layout_ds *
ff_layout_choose_valid_ds_for_read(struct pnfs_layout_segment *lseg,
u32 start_idx, u32 *best_idx,
u64 offset, u32 *dss_id)
@@ -929,34 +938,36 @@ ff_layout_choose_valid_ds_for_read(struct pnfs_layout_segment *lseg,
offset, dss_id, true);
}
-static struct nfs4_pnfs_ds *
+static struct nfs4_ff_layout_ds *
ff_layout_choose_best_ds_for_read(struct pnfs_layout_segment *lseg,
u32 start_idx, u32 *best_idx,
u64 offset, u32 *dss_id)
{
- struct nfs4_pnfs_ds *ds;
+ struct nfs4_ff_layout_ds *mirror_ds;
- ds = ff_layout_choose_valid_ds_for_read(lseg, start_idx, best_idx,
- offset, dss_id);
- if (!IS_ERR(ds))
- return ds;
+ mirror_ds = ff_layout_choose_valid_ds_for_read(lseg, start_idx,
+ best_idx, offset,
+ dss_id);
+ if (!IS_ERR(mirror_ds))
+ return mirror_ds;
return ff_layout_choose_any_ds_for_read(lseg, start_idx, best_idx,
offset, dss_id);
}
-static struct nfs4_pnfs_ds *
+static struct nfs4_ff_layout_ds *
ff_layout_get_ds_for_read(struct nfs_pageio_descriptor *pgio,
u32 *best_idx,
u64 offset,
u32 *dss_id)
{
struct pnfs_layout_segment *lseg = pgio->pg_lseg;
- struct nfs4_pnfs_ds *ds;
+ struct nfs4_ff_layout_ds *mirror_ds;
- ds = ff_layout_choose_best_ds_for_read(lseg, pgio->pg_mirror_idx,
- best_idx, offset, dss_id);
- if (!IS_ERR(ds) || !pgio->pg_mirror_idx)
- return ds;
+ mirror_ds = ff_layout_choose_best_ds_for_read(lseg,
+ pgio->pg_mirror_idx,
+ best_idx, offset, dss_id);
+ if (!IS_ERR(mirror_ds) || !pgio->pg_mirror_idx)
+ return mirror_ds;
return ff_layout_choose_best_ds_for_read(lseg, 0, best_idx,
offset, dss_id);
}
@@ -1031,8 +1042,7 @@ ff_layout_pg_init_read(struct nfs_pageio_descriptor *pgio,
struct nfs_page *req)
{
struct nfs_pgio_mirror *pgm;
- struct nfs4_ff_layout_mirror *mirror;
- struct nfs4_pnfs_ds *ds;
+ struct nfs4_ff_layout_ds *mirror_ds;
u32 ds_idx, dss_id;
if (NFS_SERVER(pgio->pg_inode)->flags &
@@ -1054,9 +1064,9 @@ ff_layout_pg_init_read(struct nfs_pageio_descriptor *pgio,
/* Reset wb_nio, since getting layout segment was successful */
req->wb_nio = 0;
- ds = ff_layout_get_ds_for_read(pgio, &ds_idx,
- req_offset(req), &dss_id);
- if (IS_ERR(ds)) {
+ mirror_ds = ff_layout_get_ds_for_read(pgio, &ds_idx,
+ req_offset(req), &dss_id);
+ if (IS_ERR(mirror_ds)) {
if (!ff_layout_no_fallback_to_mds(pgio->pg_lseg))
goto out_mds;
pnfs_generic_pg_cleanup(pgio);
@@ -1065,9 +1075,9 @@ ff_layout_pg_init_read(struct nfs_pageio_descriptor *pgio,
goto retry;
}
- mirror = FF_LAYOUT_COMP(pgio->pg_lseg, ds_idx);
pgm = &pgio->pg_mirrors[0];
- pgm->pg_bsize = mirror->dss[dss_id].mirror_ds->ds_versions[0].rsize;
+ pgm->pg_bsize = mirror_ds->ds_versions[0].rsize;
+ nfs4_ff_layout_put_deviceid(mirror_ds);
pgio->pg_mirror_idx = ds_idx;
return;
@@ -1102,6 +1112,7 @@ ff_layout_pg_init_write(struct nfs_pageio_descriptor *pgio,
struct nfs_page *req)
{
struct nfs4_ff_layout_mirror *mirror;
+ struct nfs4_ff_layout_ds *mirror_ds;
struct nfs_pgio_mirror *pgm;
struct nfs4_pnfs_ds *ds;
u32 i, dss_id;
@@ -1133,9 +1144,12 @@ ff_layout_pg_init_write(struct nfs_pageio_descriptor *pgio,
FF_LAYOUT_LSEG(pgio->pg_lseg)->stripe_unit,
mirror->dss_count,
req_offset(req));
+ mirror_ds = ff_layout_get_mirror_ds(pgio->pg_lseg->pls_layout,
+ mirror, dss_id);
ds = nfs4_ff_layout_prepare_ds(pgio->pg_lseg, mirror,
- dss_id, true);
+ mirror_ds, dss_id, true);
if (IS_ERR(ds)) {
+ nfs4_ff_layout_put_deviceid(mirror_ds);
if (!ff_layout_no_fallback_to_mds(pgio->pg_lseg))
goto out_mds;
pnfs_generic_pg_cleanup(pgio);
@@ -1144,7 +1158,8 @@ ff_layout_pg_init_write(struct nfs_pageio_descriptor *pgio,
goto retry;
}
pgm = &pgio->pg_mirrors[i];
- pgm->pg_bsize = mirror->dss[dss_id].mirror_ds->ds_versions[0].wsize;
+ pgm->pg_bsize = mirror_ds->ds_versions[0].wsize;
+ nfs4_ff_layout_put_deviceid(mirror_ds);
}
if (NFS_SERVER(pgio->pg_inode)->flags &
@@ -1276,14 +1291,16 @@ static void ff_layout_resend_pnfs_read(struct nfs_pgio_header *hdr)
u32 idx = hdr->pgio_mirror_idx + 1;
u32 new_idx = 0;
u32 dss_id = 0;
- struct nfs4_pnfs_ds *ds;
+ struct nfs4_ff_layout_ds *mirror_ds;
- ds = ff_layout_choose_any_ds_for_read(hdr->lseg, idx, &new_idx,
- hdr->args.offset, &dss_id);
- if (IS_ERR(ds))
+ mirror_ds = ff_layout_choose_any_ds_for_read(hdr->lseg, idx, &new_idx,
+ hdr->args.offset, &dss_id);
+ if (IS_ERR(mirror_ds)) {
pnfs_error_mark_layout_for_return(hdr->inode, hdr->lseg);
- else
+ } else {
+ nfs4_ff_layout_put_deviceid(mirror_ds);
ff_layout_send_layouterror(hdr->lseg);
+ }
pnfs_read_resend_pnfs(hdr, new_idx);
}
@@ -2144,6 +2161,7 @@ ff_layout_read_pagelist(struct nfs_pgio_header *hdr)
struct rpc_clnt *ds_clnt;
struct nfsd_file *localio;
struct nfs4_ff_layout_mirror *mirror;
+ struct nfs4_ff_layout_ds *mirror_ds;
const struct cred *ds_cred;
loff_t offset = hdr->args.offset;
u32 idx = hdr->pgio_mirror_idx;
@@ -2161,22 +2179,24 @@ ff_layout_read_pagelist(struct nfs_pgio_header *hdr)
FF_LAYOUT_LSEG(lseg)->stripe_unit,
mirror->dss_count,
offset);
- ds = nfs4_ff_layout_prepare_ds(lseg, mirror, dss_id, false);
+ mirror_ds = ff_layout_get_mirror_ds(lseg->pls_layout, mirror, dss_id);
+ ds = nfs4_ff_layout_prepare_ds(lseg, mirror, mirror_ds, dss_id, false);
if (IS_ERR(ds)) {
ds_fatal_error = nfs_error_is_fatal(PTR_ERR(ds));
goto out_failed;
}
- ds_clnt = nfs4_ff_find_or_create_ds_client(mirror, ds->ds_clp,
- hdr->inode, dss_id);
+ ds_clnt = nfs4_ff_find_or_create_ds_client(mirror_ds, ds->ds_clp,
+ hdr->inode);
if (IS_ERR(ds_clnt))
goto out_failed;
- ds_cred = ff_layout_get_ds_cred(mirror, &lseg->pls_range, hdr->cred, dss_id);
+ ds_cred = ff_layout_get_ds_cred(mirror, &lseg->pls_range, hdr->cred,
+ mirror_ds, dss_id);
if (!ds_cred)
goto out_failed;
- vers = nfs4_ff_layout_ds_version(mirror, dss_id);
+ vers = nfs4_ff_layout_ds_version(mirror_ds);
dprintk("%s USE DS: %s cl_count %d vers %d\n", __func__,
ds->ds_remotestr, refcount_read(&ds->ds_clp->cl_count), vers);
@@ -2188,7 +2208,8 @@ ff_layout_read_pagelist(struct nfs_pgio_header *hdr)
if (fh)
hdr->args.fh = fh;
- nfs4_ff_layout_select_ds_stateid(mirror, dss_id, &hdr->args.stateid);
+ nfs4_ff_layout_select_ds_stateid(mirror, mirror_ds, dss_id,
+ &hdr->args.stateid);
/*
* Note that if we ever decide to split across DSes,
@@ -2211,9 +2232,11 @@ ff_layout_read_pagelist(struct nfs_pgio_header *hdr)
&ff_layout_read_call_ops_v4,
0, RPC_TASK_SOFTCONN, localio);
put_cred(ds_cred);
+ nfs4_ff_layout_put_deviceid(mirror_ds);
return PNFS_ATTEMPTED;
out_failed:
+ nfs4_ff_layout_put_deviceid(mirror_ds);
if (ff_layout_avoid_mds_available_ds(lseg) && !ds_fatal_error)
return PNFS_TRY_AGAIN;
if (ff_layout_no_fallback_to_mds(lseg)) {
@@ -2239,6 +2262,7 @@ ff_layout_write_pagelist(struct nfs_pgio_header *hdr, int sync)
struct rpc_clnt *ds_clnt;
struct nfsd_file *localio;
struct nfs4_ff_layout_mirror *mirror;
+ struct nfs4_ff_layout_ds *mirror_ds;
const struct cred *ds_cred;
loff_t offset = hdr->args.offset;
int vers;
@@ -2252,22 +2276,24 @@ ff_layout_write_pagelist(struct nfs_pgio_header *hdr, int sync)
FF_LAYOUT_LSEG(lseg)->stripe_unit,
mirror->dss_count,
offset);
- ds = nfs4_ff_layout_prepare_ds(lseg, mirror, dss_id, true);
+ mirror_ds = ff_layout_get_mirror_ds(lseg->pls_layout, mirror, dss_id);
+ ds = nfs4_ff_layout_prepare_ds(lseg, mirror, mirror_ds, dss_id, true);
if (IS_ERR(ds)) {
ds_fatal_error = nfs_error_is_fatal(PTR_ERR(ds));
goto out_failed;
}
- ds_clnt = nfs4_ff_find_or_create_ds_client(mirror, ds->ds_clp,
- hdr->inode, dss_id);
+ ds_clnt = nfs4_ff_find_or_create_ds_client(mirror_ds, ds->ds_clp,
+ hdr->inode);
if (IS_ERR(ds_clnt))
goto out_failed;
- ds_cred = ff_layout_get_ds_cred(mirror, &lseg->pls_range, hdr->cred, dss_id);
+ ds_cred = ff_layout_get_ds_cred(mirror, &lseg->pls_range, hdr->cred,
+ mirror_ds, dss_id);
if (!ds_cred)
goto out_failed;
- vers = nfs4_ff_layout_ds_version(mirror, dss_id);
+ vers = nfs4_ff_layout_ds_version(mirror_ds);
dprintk("%s ino %llu sync %d req %zu@%llu DS: %s cl_count %d vers %d\n",
__func__, hdr->inode->i_ino, sync, (size_t) hdr->args.count,
@@ -2282,7 +2308,8 @@ ff_layout_write_pagelist(struct nfs_pgio_header *hdr, int sync)
if (fh)
hdr->args.fh = fh;
- nfs4_ff_layout_select_ds_stateid(mirror, dss_id, &hdr->args.stateid);
+ nfs4_ff_layout_select_ds_stateid(mirror, mirror_ds, dss_id,
+ &hdr->args.stateid);
/*
* Note that if we ever decide to split across DSes,
@@ -2304,9 +2331,11 @@ ff_layout_write_pagelist(struct nfs_pgio_header *hdr, int sync)
&ff_layout_write_call_ops_v4,
sync, RPC_TASK_SOFTCONN, localio);
put_cred(ds_cred);
+ nfs4_ff_layout_put_deviceid(mirror_ds);
return PNFS_ATTEMPTED;
out_failed:
+ nfs4_ff_layout_put_deviceid(mirror_ds);
if (ff_layout_avoid_mds_available_ds(lseg) && !ds_fatal_error)
return PNFS_TRY_AGAIN;
if (ff_layout_no_fallback_to_mds(lseg)) {
@@ -2341,6 +2370,7 @@ static int ff_layout_initiate_commit(struct nfs_commit_data *data, int how)
struct rpc_clnt *ds_clnt;
struct nfsd_file *localio;
struct nfs4_ff_layout_mirror *mirror;
+ struct nfs4_ff_layout_ds *mirror_ds = NULL;
const struct cred *ds_cred;
u32 idx, dss_id;
int vers, ret;
@@ -2353,20 +2383,22 @@ static int ff_layout_initiate_commit(struct nfs_commit_data *data, int how)
idx = calc_mirror_idx_from_commit(lseg, data->ds_commit_index);
mirror = FF_LAYOUT_COMP(lseg, idx);
dss_id = calc_dss_id_from_commit(lseg, data->ds_commit_index);
- ds = nfs4_ff_layout_prepare_ds(lseg, mirror, dss_id, true);
+ mirror_ds = ff_layout_get_mirror_ds(lseg->pls_layout, mirror, dss_id);
+ ds = nfs4_ff_layout_prepare_ds(lseg, mirror, mirror_ds, dss_id, true);
if (IS_ERR(ds))
goto out_err;
- ds_clnt = nfs4_ff_find_or_create_ds_client(mirror, ds->ds_clp,
- data->inode, dss_id);
+ ds_clnt = nfs4_ff_find_or_create_ds_client(mirror_ds, ds->ds_clp,
+ data->inode);
if (IS_ERR(ds_clnt))
goto out_err;
- ds_cred = ff_layout_get_ds_cred(mirror, &lseg->pls_range, data->cred, dss_id);
+ ds_cred = ff_layout_get_ds_cred(mirror, &lseg->pls_range, data->cred,
+ mirror_ds, dss_id);
if (!ds_cred)
goto out_err;
- vers = nfs4_ff_layout_ds_version(mirror, dss_id);
+ vers = nfs4_ff_layout_ds_version(mirror_ds);
dprintk("%s ino %llu, how %d cl_count %d vers %d\n", __func__,
data->inode->i_ino, how, refcount_read(&ds->ds_clp->cl_count),
@@ -2392,8 +2424,10 @@ static int ff_layout_initiate_commit(struct nfs_commit_data *data, int how)
&ff_layout_commit_call_ops_v4,
how, RPC_TASK_SOFTCONN, localio);
put_cred(ds_cred);
+ nfs4_ff_layout_put_deviceid(mirror_ds);
return ret;
out_err:
+ nfs4_ff_layout_put_deviceid(mirror_ds);
pnfs_generic_prepare_to_resend_writes(data);
pnfs_generic_commit_release(data);
return -EAGAIN;
diff --git a/fs/nfs/flexfilelayout/flexfilelayout.h b/fs/nfs/flexfilelayout/flexfilelayout.h
index a5bd00f69e82..a65cdffe325b 100644
--- a/fs/nfs/flexfilelayout/flexfilelayout.h
+++ b/fs/nfs/flexfilelayout/flexfilelayout.h
@@ -207,9 +207,9 @@ ff_layout_no_read_on_rw(struct pnfs_layout_segment *lseg)
}
static inline int
-nfs4_ff_layout_ds_version(const struct nfs4_ff_layout_mirror *mirror, u32 dss_id)
+nfs4_ff_layout_ds_version(const struct nfs4_ff_layout_ds *mirror_ds)
{
- return mirror->dss[dss_id].mirror_ds->ds_versions[0].version;
+ return mirror_ds->ds_versions[0].version;
}
static inline u32
@@ -245,23 +245,29 @@ struct nfs_fh *
nfs4_ff_layout_select_ds_fh(struct nfs4_ff_layout_mirror *mirror, u32 dss_id);
void
nfs4_ff_layout_select_ds_stateid(const struct nfs4_ff_layout_mirror *mirror,
+ const struct nfs4_ff_layout_ds *mirror_ds,
u32 dss_id,
nfs4_stateid *stateid);
+struct nfs4_ff_layout_ds *
+ff_layout_get_mirror_ds(struct pnfs_layout_hdr *lo,
+ struct nfs4_ff_layout_mirror *mirror,
+ u32 dss_id);
struct nfs4_pnfs_ds *
nfs4_ff_layout_prepare_ds(struct pnfs_layout_segment *lseg,
struct nfs4_ff_layout_mirror *mirror,
+ struct nfs4_ff_layout_ds *mirror_ds,
u32 dss_id,
bool fail_return);
struct rpc_clnt *
-nfs4_ff_find_or_create_ds_client(struct nfs4_ff_layout_mirror *mirror,
+nfs4_ff_find_or_create_ds_client(const struct nfs4_ff_layout_ds *mirror_ds,
struct nfs_client *ds_clp,
- struct inode *inode,
- u32 dss_id);
+ struct inode *inode);
const struct cred *ff_layout_get_ds_cred(struct nfs4_ff_layout_mirror *mirror,
const struct pnfs_layout_range *range,
const struct cred *mdscred,
+ const struct nfs4_ff_layout_ds *mirror_ds,
u32 dss_id);
bool ff_layout_avoid_mds_available_ds(struct pnfs_layout_segment *lseg);
bool ff_layout_avoid_read_on_rw(struct pnfs_layout_segment *lseg);
diff --git a/fs/nfs/flexfilelayout/flexfilelayoutdev.c b/fs/nfs/flexfilelayout/flexfilelayoutdev.c
index 1109462a9699..b44281406011 100644
--- a/fs/nfs/flexfilelayout/flexfilelayoutdev.c
+++ b/fs/nfs/flexfilelayout/flexfilelayoutdev.c
@@ -316,24 +316,42 @@ nfs4_ff_layout_select_ds_fh(struct nfs4_ff_layout_mirror *mirror, u32 dss_id)
void
nfs4_ff_layout_select_ds_stateid(const struct nfs4_ff_layout_mirror *mirror,
+ const struct nfs4_ff_layout_ds *mirror_ds,
u32 dss_id,
nfs4_stateid *stateid)
{
- if (nfs4_ff_layout_ds_version(mirror, dss_id) == 4)
+ if (nfs4_ff_layout_ds_version(mirror_ds) == 4)
nfs4_stateid_copy(stateid, &mirror->dss[dss_id].stateid);
}
-static bool
-ff_layout_init_mirror_ds(struct pnfs_layout_hdr *lo,
- struct nfs4_ff_layout_mirror *mirror,
- u32 dss_id)
+/**
+ * ff_layout_get_mirror_ds - get a reference to a stripe's device node
+ * @lo: the layout header
+ * @mirror: layout mirror
+ * @dss_id: DS stripe id within the mirror
+ *
+ * Resolve the stripe's deviceid on first use and pin the resulting node
+ * on the mirror; every caller additionally gets its own reference so the
+ * node stays valid for as long as the caller uses it, independent of the
+ * mirror's pin.
+ *
+ * Returns a referenced device node on success, an ERR_PTR otherwise.
+ */
+struct nfs4_ff_layout_ds *
+ff_layout_get_mirror_ds(struct pnfs_layout_hdr *lo,
+ struct nfs4_ff_layout_mirror *mirror,
+ u32 dss_id)
{
+ struct nfs4_ff_layout_ds *mirror_ds;
+
if (mirror == NULL)
- goto outerr;
- if (mirror->dss[dss_id].mirror_ds == NULL) {
+ return ERR_PTR(-ENODEV);
+
+ mirror_ds = mirror->dss[dss_id].mirror_ds;
+ if (mirror_ds == NULL) {
struct nfs4_deviceid_node *node;
- struct nfs4_ff_layout_ds *mirror_ds = ERR_PTR(-ENODEV);
+ mirror_ds = ERR_PTR(-ENODEV);
node = nfs4_find_get_deviceid(NFS_SERVER(lo->plh_inode),
&mirror->dss[dss_id].devid, lo->plh_lc_cred,
GFP_KERNEL);
@@ -344,20 +362,23 @@ ff_layout_init_mirror_ds(struct pnfs_layout_hdr *lo,
if (cmpxchg(&mirror->dss[dss_id].mirror_ds, NULL, mirror_ds) &&
mirror_ds != ERR_PTR(-ENODEV))
nfs4_put_deviceid_node(node);
- }
- if (IS_ERR(mirror->dss[dss_id].mirror_ds))
- goto outerr;
+ mirror_ds = mirror->dss[dss_id].mirror_ds;
+ }
- return true;
-outerr:
- return false;
+ if (IS_ERR(mirror_ds))
+ return mirror_ds;
+ if (!atomic_inc_not_zero(&mirror_ds->id_node.ref))
+ return ERR_PTR(-ENODEV);
+ return mirror_ds;
}
/**
* nfs4_ff_layout_prepare_ds - prepare a DS connection for an RPC call
* @lseg: the layout segment we're operating on
* @mirror: layout mirror describing the DS to use
+ * @mirror_ds: referenced device node for the stripe, from
+ * ff_layout_get_mirror_ds() (may be an ERR_PTR)
* @dss_id: DS stripe id to select stripe to use
* @fail_return: return layout on connect failure?
*
@@ -375,6 +396,7 @@ ff_layout_init_mirror_ds(struct pnfs_layout_hdr *lo,
struct nfs4_pnfs_ds *
nfs4_ff_layout_prepare_ds(struct pnfs_layout_segment *lseg,
struct nfs4_ff_layout_mirror *mirror,
+ struct nfs4_ff_layout_ds *mirror_ds,
u32 dss_id,
bool fail_return)
{
@@ -384,10 +406,10 @@ nfs4_ff_layout_prepare_ds(struct pnfs_layout_segment *lseg,
unsigned int max_payload;
int status = -EAGAIN;
- if (!ff_layout_init_mirror_ds(lseg->pls_layout, mirror, dss_id))
+ if (IS_ERR_OR_NULL(mirror_ds))
goto noconnect;
- ds = mirror->dss[dss_id].mirror_ds->ds;
+ ds = mirror_ds->ds;
if (READ_ONCE(ds->ds_clp))
goto out;
/* matching smp_wmb() in _nfs4_pnfs_v3/4_ds_connect */
@@ -396,10 +418,10 @@ nfs4_ff_layout_prepare_ds(struct pnfs_layout_segment *lseg,
/* FIXME: For now we assume the server sent only one version of NFS
* to use for the DS.
*/
- status = nfs4_pnfs_ds_connect(s, ds, &mirror->dss[dss_id].mirror_ds->id_node,
+ status = nfs4_pnfs_ds_connect(s, ds, &mirror_ds->id_node,
dataserver_timeo, dataserver_retrans,
- mirror->dss[dss_id].mirror_ds->ds_versions[0].version,
- mirror->dss[dss_id].mirror_ds->ds_versions[0].minor_version);
+ mirror_ds->ds_versions[0].version,
+ mirror_ds->ds_versions[0].minor_version);
/* connect success, check rsize/wsize limit */
if (!status) {
@@ -412,10 +434,10 @@ nfs4_ff_layout_prepare_ds(struct pnfs_layout_segment *lseg,
max_payload =
nfs_block_size(rpc_max_payload(ds->ds_clp->cl_rpcclient),
NULL);
- if (mirror->dss[dss_id].mirror_ds->ds_versions[0].rsize > max_payload)
- mirror->dss[dss_id].mirror_ds->ds_versions[0].rsize = max_payload;
- if (mirror->dss[dss_id].mirror_ds->ds_versions[0].wsize > max_payload)
- mirror->dss[dss_id].mirror_ds->ds_versions[0].wsize = max_payload;
+ if (mirror_ds->ds_versions[0].rsize > max_payload)
+ mirror_ds->ds_versions[0].rsize = max_payload;
+ if (mirror_ds->ds_versions[0].wsize > max_payload)
+ mirror_ds->ds_versions[0].wsize = max_payload;
goto out;
}
noconnect:
@@ -435,11 +457,12 @@ const struct cred *
ff_layout_get_ds_cred(struct nfs4_ff_layout_mirror *mirror,
const struct pnfs_layout_range *range,
const struct cred *mdscred,
+ const struct nfs4_ff_layout_ds *mirror_ds,
u32 dss_id)
{
const struct cred *cred;
- if (mirror && !mirror->dss[dss_id].mirror_ds->ds_versions[0].tightly_coupled) {
+ if (mirror && !mirror_ds->ds_versions[0].tightly_coupled) {
cred = ff_layout_get_mirror_cred(mirror, range->iomode, dss_id);
if (!cred)
cred = get_cred(mdscred);
@@ -451,20 +474,18 @@ ff_layout_get_ds_cred(struct nfs4_ff_layout_mirror *mirror,
/**
* nfs4_ff_find_or_create_ds_client - Find or create a DS rpc client
- * @mirror: pointer to the mirror
+ * @mirror_ds: device node for the stripe
* @ds_clp: nfs_client for the DS
* @inode: pointer to inode
- * @dss_id: DS stripe id
*
* Find or create a DS rpc client with th MDS server rpc client auth flavor
* in the nfs_client cl_ds_clients list.
*/
struct rpc_clnt *
-nfs4_ff_find_or_create_ds_client(struct nfs4_ff_layout_mirror *mirror,
- struct nfs_client *ds_clp, struct inode *inode,
- u32 dss_id)
+nfs4_ff_find_or_create_ds_client(const struct nfs4_ff_layout_ds *mirror_ds,
+ struct nfs_client *ds_clp, struct inode *inode)
{
- switch (mirror->dss[dss_id].mirror_ds->ds_versions[0].version) {
+ switch (mirror_ds->ds_versions[0].version) {
case 3:
/* For NFSv3 DS, flavor is set when creating DS connections */
return ds_clp->cl_rpcclient;
--
2.53.0
^ permalink raw reply related [flat|nested] 22+ messages in thread* [PATCH 06/21] NFSv4/flexfiles: Carry the device node reference across each I/O
2026-08-13 20:42 [PATCH 00/21] NFS: flexfiles device notifications and caching for wide striped layouts Benjamin Coddington
` (4 preceding siblings ...)
2026-08-13 20:43 ` [PATCH 05/21] NFSv4/flexfiles: Reference the device node across DS setup Benjamin Coddington
@ 2026-08-13 20:43 ` Benjamin Coddington
2026-08-13 20:43 ` [PATCH 07/21] NFSv4/flexfiles: Hold a device node reference for layoutstats encoding Benjamin Coddington
` (14 subsequent siblings)
20 siblings, 0 replies; 22+ messages in thread
From: Benjamin Coddington @ 2026-08-13 20:43 UTC (permalink / raw)
To: Trond Myklebust, Anna Schumaker
Cc: linux-nfs, Jonathan Curley, Mike Snitzer, Jeff Layton
Rather than dropping the device node reference when DS setup completes,
transfer it to the in-flight I/O: carry it on nfs_pgio_header and
nfs_commit_data as ds_dev, and release it when the header or commit
data is released, alongside the lseg reference.
Convert the completion paths to use the carried node instead of
re-reading the mirror's pinned pointer: DS error tracking, marking the
deviceid available/unavailable, and deleting the deviceid on connection
errors now act on the node the I/O was actually sent to. Once a CHANGE
notification can re-point the pinned pointer mid-flight, this keeps
error attribution on the old device rather than its replacement (and
removes a NULL dereference had the pointer been reset to NULL).
FF_LAYOUT_DEVID_NODE() is now unused; remove it.
Assisted-by: Claude:claude-fable-5
Signed-off-by: Benjamin Coddington <bcodding@hammerspace.com>
---
fs/nfs/flexfilelayout/flexfilelayout.c | 82 ++++++++++-------------
fs/nfs/flexfilelayout/flexfilelayout.h | 15 +----
fs/nfs/flexfilelayout/flexfilelayoutdev.c | 11 +--
fs/nfs/pnfs.c | 2 +
fs/nfs/pnfs.h | 8 +++
fs/nfs/pnfs_nfs.c | 1 +
include/linux/nfs_xdr.h | 2 +
7 files changed, 58 insertions(+), 63 deletions(-)
diff --git a/fs/nfs/flexfilelayout/flexfilelayout.c b/fs/nfs/flexfilelayout/flexfilelayout.c
index 2bdc74a9a3eb..b52323e639a2 100644
--- a/fs/nfs/flexfilelayout/flexfilelayout.c
+++ b/fs/nfs/flexfilelayout/flexfilelayout.c
@@ -857,24 +857,6 @@ nfs4_ff_layout_stat_io_end_write(struct rpc_task *task,
spin_unlock(&mirror->lock);
}
-static void
-ff_layout_mark_ds_unreachable(struct pnfs_layout_segment *lseg, u32 idx, u32 dss_id)
-{
- struct nfs4_deviceid_node *devid = FF_LAYOUT_DEVID_NODE(lseg, idx, dss_id);
-
- if (devid)
- nfs4_mark_deviceid_unavailable(devid);
-}
-
-static void
-ff_layout_mark_ds_reachable(struct pnfs_layout_segment *lseg, u32 idx, u32 dss_id)
-{
- struct nfs4_deviceid_node *devid = FF_LAYOUT_DEVID_NODE(lseg, idx, dss_id);
-
- if (devid)
- nfs4_mark_deviceid_available(devid);
-}
-
static struct nfs4_ff_layout_ds *
ff_layout_choose_ds_for_read(struct pnfs_layout_segment *lseg,
u32 start_idx, u32 *best_idx,
@@ -1333,11 +1315,10 @@ static int ff_layout_async_handle_error_v4(struct rpc_task *task,
struct nfs4_state *state,
struct nfs_client *clp,
struct pnfs_layout_segment *lseg,
- u32 idx, u32 dss_id)
+ struct nfs4_deviceid_node *devid)
{
struct pnfs_layout_hdr *lo = lseg->pls_layout;
struct inode *inode = lo->plh_inode;
- struct nfs4_deviceid_node *devid = FF_LAYOUT_DEVID_NODE(lseg, idx, dss_id);
struct nfs4_slot_table *tbl = &clp->cl_session->fc_slot_table;
switch (op_status) {
@@ -1409,8 +1390,9 @@ static int ff_layout_async_handle_error_v4(struct rpc_task *task,
case -ENODEV:
dprintk("%s DS connection error %d\n", __func__,
task->tk_status);
- nfs4_delete_deviceid(devid->ld, devid->nfs_client,
- &devid->deviceid);
+ if (devid)
+ nfs4_delete_deviceid(devid->ld, devid->nfs_client,
+ &devid->deviceid);
rpc_wake_up(&tbl->slot_tbl_waitq);
break;
default:
@@ -1434,9 +1416,8 @@ static int ff_layout_async_handle_error_v3(struct rpc_task *task,
u32 op_status,
struct nfs_client *clp,
struct pnfs_layout_segment *lseg,
- u32 idx, u32 dss_id)
+ struct nfs4_deviceid_node *devid)
{
- struct nfs4_deviceid_node *devid = FF_LAYOUT_DEVID_NODE(lseg, idx, dss_id);
switch (op_status) {
case NFS_OK:
@@ -1482,8 +1463,9 @@ static int ff_layout_async_handle_error_v3(struct rpc_task *task,
default:
dprintk("%s DS connection error %d\n", __func__,
task->tk_status);
- nfs4_delete_deviceid(devid->ld, devid->nfs_client,
- &devid->deviceid);
+ if (devid)
+ nfs4_delete_deviceid(devid->ld, devid->nfs_client,
+ &devid->deviceid);
}
out_reset_to_pnfs:
/* FIXME: Need to prevent infinite looping here. */
@@ -1500,12 +1482,13 @@ static int ff_layout_async_handle_error(struct rpc_task *task,
struct nfs4_state *state,
struct nfs_client *clp,
struct pnfs_layout_segment *lseg,
- u32 idx, u32 dss_id)
+ struct nfs4_deviceid_node *devid)
{
int vers = clp->cl_nfs_mod->rpc_vers->number;
if (task->tk_status >= 0) {
- ff_layout_mark_ds_reachable(lseg, idx, dss_id);
+ if (devid)
+ nfs4_mark_deviceid_available(devid);
return 0;
}
@@ -1516,10 +1499,10 @@ static int ff_layout_async_handle_error(struct rpc_task *task,
switch (vers) {
case 3:
return ff_layout_async_handle_error_v3(task, op_status, clp,
- lseg, idx, dss_id);
+ lseg, devid);
case 4:
return ff_layout_async_handle_error_v4(task, op_status, state,
- clp, lseg, idx, dss_id);
+ clp, lseg, devid);
default:
/* should never happen */
WARN_ON_ONCE(1);
@@ -1528,6 +1511,7 @@ static int ff_layout_async_handle_error(struct rpc_task *task,
}
static void ff_layout_io_track_ds_error(struct pnfs_layout_segment *lseg,
+ struct nfs4_deviceid_node *devid,
u32 idx, u32 dss_id, u64 offset, u64 length,
u32 *op_status, int opnum, int error)
{
@@ -1566,8 +1550,8 @@ static void ff_layout_io_track_ds_error(struct pnfs_layout_segment *lseg,
mirror = FF_LAYOUT_COMP(lseg, idx);
err = ff_layout_track_ds_error(FF_LAYOUT_FROM_HDR(lseg->pls_layout),
- mirror, dss_id, offset, length, status, opnum,
- nfs_io_gfp_mask());
+ mirror, devid, dss_id, offset, length,
+ status, opnum, nfs_io_gfp_mask());
switch (status) {
case NFS4ERR_DELAY:
@@ -1575,7 +1559,8 @@ static void ff_layout_io_track_ds_error(struct pnfs_layout_segment *lseg,
case NFS4ERR_PERM:
break;
case NFS4ERR_NXIO:
- ff_layout_mark_ds_unreachable(lseg, idx, dss_id);
+ if (devid)
+ nfs4_mark_deviceid_unavailable(devid);
/*
* Don't return the layout if this is a read and we still
* have layouts to try
@@ -1603,7 +1588,7 @@ static int ff_layout_read_done_cb(struct rpc_task *task,
int err;
if (task->tk_status < 0) {
- ff_layout_io_track_ds_error(hdr->lseg,
+ ff_layout_io_track_ds_error(hdr->lseg, hdr->ds_dev,
hdr->pgio_mirror_idx, dss_id,
hdr->args.offset, hdr->args.count,
&hdr->res.op_status, OP_READ,
@@ -1614,8 +1599,7 @@ static int ff_layout_read_done_cb(struct rpc_task *task,
err = ff_layout_async_handle_error(task, hdr->res.op_status,
hdr->args.context->state,
hdr->ds_clp, hdr->lseg,
- hdr->pgio_mirror_idx,
- dss_id);
+ hdr->ds_dev);
trace_nfs4_pnfs_read(hdr, err);
clear_bit(NFS_IOHDR_RESEND_PNFS, &hdr->flags);
@@ -1808,7 +1792,7 @@ static int ff_layout_write_done_cb(struct rpc_task *task,
int err;
if (task->tk_status < 0) {
- ff_layout_io_track_ds_error(hdr->lseg,
+ ff_layout_io_track_ds_error(hdr->lseg, hdr->ds_dev,
hdr->pgio_mirror_idx, dss_id,
hdr->args.offset, hdr->args.count,
&hdr->res.op_status, OP_WRITE,
@@ -1819,8 +1803,7 @@ static int ff_layout_write_done_cb(struct rpc_task *task,
err = ff_layout_async_handle_error(task, hdr->res.op_status,
hdr->args.context->state,
hdr->ds_clp, hdr->lseg,
- hdr->pgio_mirror_idx,
- dss_id);
+ hdr->ds_dev);
trace_nfs4_pnfs_write(hdr, err);
clear_bit(NFS_IOHDR_RESEND_PNFS, &hdr->flags);
@@ -1862,7 +1845,7 @@ static int ff_layout_commit_done_cb(struct rpc_task *task,
u32 dss_id = calc_dss_id_from_commit(data->lseg, data->ds_commit_index);
if (task->tk_status < 0) {
- ff_layout_io_track_ds_error(data->lseg, idx, dss_id,
+ ff_layout_io_track_ds_error(data->lseg, data->ds_dev, idx, dss_id,
data->args.offset, data->args.count,
&data->res.op_status, OP_COMMIT,
task->tk_status);
@@ -1870,8 +1853,8 @@ static int ff_layout_commit_done_cb(struct rpc_task *task,
}
err = ff_layout_async_handle_error(task, data->res.op_status,
- NULL, data->ds_clp, data->lseg, idx,
- dss_id);
+ NULL, data->ds_clp, data->lseg,
+ data->ds_dev);
trace_nfs4_pnfs_commit_ds(data, err);
switch (err) {
@@ -2226,13 +2209,16 @@ ff_layout_read_pagelist(struct nfs_pgio_header *hdr)
ff_layout_read_record_layoutstats_start(&hdr->task, hdr);
}
+ /* Transfer the device node reference to the I/O; put on release */
+ pnfs_put_ds_dev(hdr->ds_dev);
+ hdr->ds_dev = &mirror_ds->id_node;
+
/* Perform an asynchronous read to ds */
nfs_initiate_pgio(ds_clnt, hdr, ds_cred, ds->ds_clp->rpc_ops,
vers == 3 ? &ff_layout_read_call_ops_v3 :
&ff_layout_read_call_ops_v4,
0, RPC_TASK_SOFTCONN, localio);
put_cred(ds_cred);
- nfs4_ff_layout_put_deviceid(mirror_ds);
return PNFS_ATTEMPTED;
out_failed:
@@ -2325,13 +2311,16 @@ ff_layout_write_pagelist(struct nfs_pgio_header *hdr, int sync)
ff_layout_write_record_layoutstats_start(&hdr->task, hdr);
}
+ /* Transfer the device node reference to the I/O; put on release */
+ pnfs_put_ds_dev(hdr->ds_dev);
+ hdr->ds_dev = &mirror_ds->id_node;
+
/* Perform an asynchronous write */
nfs_initiate_pgio(ds_clnt, hdr, ds_cred, ds->ds_clp->rpc_ops,
vers == 3 ? &ff_layout_write_call_ops_v3 :
&ff_layout_write_call_ops_v4,
sync, RPC_TASK_SOFTCONN, localio);
put_cred(ds_cred);
- nfs4_ff_layout_put_deviceid(mirror_ds);
return PNFS_ATTEMPTED;
out_failed:
@@ -2419,12 +2408,15 @@ static int ff_layout_initiate_commit(struct nfs_commit_data *data, int how)
ff_layout_commit_record_layoutstats_start(&data->task, data);
}
+ /* Transfer the device node reference to the commit; put on release */
+ pnfs_put_ds_dev(data->ds_dev);
+ data->ds_dev = &mirror_ds->id_node;
+
ret = nfs_initiate_commit(ds_clnt, data, ds->ds_clp->rpc_ops,
vers == 3 ? &ff_layout_commit_call_ops_v3 :
&ff_layout_commit_call_ops_v4,
how, RPC_TASK_SOFTCONN, localio);
put_cred(ds_cred);
- nfs4_ff_layout_put_deviceid(mirror_ds);
return ret;
out_err:
nfs4_ff_layout_put_deviceid(mirror_ds);
diff --git a/fs/nfs/flexfilelayout/flexfilelayout.h b/fs/nfs/flexfilelayout/flexfilelayout.h
index a65cdffe325b..ceed1a9dd697 100644
--- a/fs/nfs/flexfilelayout/flexfilelayout.h
+++ b/fs/nfs/flexfilelayout/flexfilelayout.h
@@ -162,20 +162,6 @@ FF_LAYOUT_COMP(struct pnfs_layout_segment *lseg, u32 idx)
return NULL;
}
-static inline struct nfs4_deviceid_node *
-FF_LAYOUT_DEVID_NODE(struct pnfs_layout_segment *lseg, u32 idx, u32 dss_id)
-{
- struct nfs4_ff_layout_mirror *mirror = FF_LAYOUT_COMP(lseg, idx);
-
- if (mirror != NULL) {
- struct nfs4_ff_layout_ds *mirror_ds = mirror->dss[dss_id].mirror_ds;
-
- if (!IS_ERR_OR_NULL(mirror_ds))
- return &mirror_ds->id_node;
- }
- return NULL;
-}
-
static inline u32
FF_LAYOUT_MIRROR_COUNT(struct pnfs_layout_segment *lseg)
{
@@ -232,6 +218,7 @@ void nfs4_ff_layout_put_deviceid(struct nfs4_ff_layout_ds *mirror_ds);
void nfs4_ff_layout_free_deviceid(struct nfs4_ff_layout_ds *mirror_ds);
int ff_layout_track_ds_error(struct nfs4_flexfile_layout *flo,
struct nfs4_ff_layout_mirror *mirror,
+ const struct nfs4_deviceid_node *devid,
u32 dss_id, u64 offset, u64 length, int status,
enum nfs_opnum4 opnum, gfp_t gfp_flags);
void ff_layout_send_layouterror(struct pnfs_layout_segment *lseg);
diff --git a/fs/nfs/flexfilelayout/flexfilelayoutdev.c b/fs/nfs/flexfilelayout/flexfilelayoutdev.c
index b44281406011..d140e46581e7 100644
--- a/fs/nfs/flexfilelayout/flexfilelayoutdev.c
+++ b/fs/nfs/flexfilelayout/flexfilelayoutdev.c
@@ -255,6 +255,7 @@ ff_layout_add_ds_error_locked(struct nfs4_flexfile_layout *flo,
int ff_layout_track_ds_error(struct nfs4_flexfile_layout *flo,
struct nfs4_ff_layout_mirror *mirror,
+ const struct nfs4_deviceid_node *devid,
u32 dss_id, u64 offset, u64 length, int status,
enum nfs_opnum4 opnum, gfp_t gfp_flags)
{
@@ -263,7 +264,7 @@ int ff_layout_track_ds_error(struct nfs4_flexfile_layout *flo,
if (status == 0)
return 0;
- if (IS_ERR_OR_NULL(mirror->dss[dss_id].mirror_ds))
+ if (devid == NULL)
return -EINVAL;
dserr = kmalloc_obj(*dserr, gfp_flags);
@@ -276,8 +277,7 @@ int ff_layout_track_ds_error(struct nfs4_flexfile_layout *flo,
dserr->status = status;
dserr->opnum = opnum;
nfs4_stateid_copy(&dserr->stateid, &mirror->dss[dss_id].stateid);
- memcpy(&dserr->deviceid, &mirror->dss[dss_id].mirror_ds->id_node.deviceid,
- NFS4_DEVICEID4_SIZE);
+ memcpy(&dserr->deviceid, &devid->deviceid, NFS4_DEVICEID4_SIZE);
spin_lock(&flo->generic_hdr.plh_inode->i_lock);
ff_layout_add_ds_error_locked(flo, dserr);
@@ -442,7 +442,10 @@ nfs4_ff_layout_prepare_ds(struct pnfs_layout_segment *lseg,
}
noconnect:
ff_layout_track_ds_error(FF_LAYOUT_FROM_HDR(lseg->pls_layout),
- mirror, dss_id, lseg->pls_range.offset,
+ mirror,
+ IS_ERR_OR_NULL(mirror_ds) ?
+ NULL : &mirror_ds->id_node,
+ dss_id, lseg->pls_range.offset,
lseg->pls_range.length, NFS4ERR_NXIO,
OP_ILLEGAL, GFP_NOIO);
ff_layout_send_layouterror(lseg);
diff --git a/fs/nfs/pnfs.c b/fs/nfs/pnfs.c
index 7715e2bd5871..a21128321c0a 100644
--- a/fs/nfs/pnfs.c
+++ b/fs/nfs/pnfs.c
@@ -3104,6 +3104,7 @@ pnfs_do_write(struct nfs_pageio_descriptor *desc,
static void pnfs_writehdr_free(struct nfs_pgio_header *hdr)
{
+ pnfs_put_ds_dev(hdr->ds_dev);
pnfs_put_lseg(hdr->lseg);
nfs_pgio_header_free(hdr);
}
@@ -3249,6 +3250,7 @@ pnfs_do_read(struct nfs_pageio_descriptor *desc, struct nfs_pgio_header *hdr)
static void pnfs_readhdr_free(struct nfs_pgio_header *hdr)
{
+ pnfs_put_ds_dev(hdr->ds_dev);
pnfs_put_lseg(hdr->lseg);
nfs_pgio_header_free(hdr);
}
diff --git a/fs/nfs/pnfs.h b/fs/nfs/pnfs.h
index eb39859c216c..bdce7f930c6a 100644
--- a/fs/nfs/pnfs.h
+++ b/fs/nfs/pnfs.h
@@ -384,6 +384,14 @@ void nfs4_init_deviceid_node(struct nfs4_deviceid_node *, struct nfs_server *,
const struct nfs4_deviceid *);
bool nfs4_put_deviceid_node(struct nfs4_deviceid_node *);
void nfs4_mark_deviceid_available(struct nfs4_deviceid_node *node);
+
+/* Put the device node reference carried by an in-flight I/O, if any */
+static inline void pnfs_put_ds_dev(struct nfs4_deviceid_node *dev)
+{
+ if (dev)
+ nfs4_put_deviceid_node(dev);
+}
+
void nfs4_mark_deviceid_unavailable(struct nfs4_deviceid_node *node);
bool nfs4_test_deviceid_unavailable(struct nfs4_deviceid_node *node);
void nfs4_deviceid_purge_client(const struct nfs_client *);
diff --git a/fs/nfs/pnfs_nfs.c b/fs/nfs/pnfs_nfs.c
index 648c95b78eea..f40368f839d0 100644
--- a/fs/nfs/pnfs_nfs.c
+++ b/fs/nfs/pnfs_nfs.c
@@ -55,6 +55,7 @@ void pnfs_generic_commit_release(void *calldata)
struct nfs_commit_data *data = calldata;
data->completion_ops->completion(data);
+ pnfs_put_ds_dev(data->ds_dev);
pnfs_put_lseg(data->lseg);
nfs_put_client(data->ds_clp);
nfs_commitdata_release(data);
diff --git a/include/linux/nfs_xdr.h b/include/linux/nfs_xdr.h
index 11c5b31cfc7d..aefb4715c809 100644
--- a/include/linux/nfs_xdr.h
+++ b/include/linux/nfs_xdr.h
@@ -1687,6 +1687,7 @@ struct nfs_pgio_header {
struct nfs_client *ds_clp; /* pNFS data server */
u32 ds_commit_idx; /* ds index if ds_clp is set */
u32 pgio_mirror_idx;/* mirror index in pgio layer */
+ struct nfs4_deviceid_node *ds_dev; /* device node ref held across the I/O */
};
struct nfs_mds_commit_info {
@@ -1725,6 +1726,7 @@ struct nfs_commit_data {
struct nfs_open_context *context;
struct pnfs_layout_segment *lseg;
struct nfs_client *ds_clp; /* pNFS data server */
+ struct nfs4_deviceid_node *ds_dev; /* device node ref held across the commit */
int ds_commit_index;
loff_t lwb;
const struct rpc_call_ops *mds_ops;
--
2.53.0
^ permalink raw reply related [flat|nested] 22+ messages in thread* [PATCH 07/21] NFSv4/flexfiles: Hold a device node reference for layoutstats encoding
2026-08-13 20:42 [PATCH 00/21] NFS: flexfiles device notifications and caching for wide striped layouts Benjamin Coddington
` (5 preceding siblings ...)
2026-08-13 20:43 ` [PATCH 06/21] NFSv4/flexfiles: Carry the device node reference across each I/O Benjamin Coddington
@ 2026-08-13 20:43 ` Benjamin Coddington
2026-08-13 20:43 ` [PATCH 08/21] NFSv4/flexfiles: Make the pinned device node pointer RCU-managed Benjamin Coddington
` (13 subsequent siblings)
20 siblings, 0 replies; 22+ messages in thread
From: Benjamin Coddington @ 2026-08-13 20:43 UTC (permalink / raw)
To: Trond Myklebust, Anna Schumaker
Cc: linux-nfs, Jonathan Curley, Mike Snitzer, Jeff Layton
ff_layout_mirror_prepare_stats() records a stripe's device under i_lock,
but the layoutstats/layoutreturn XDR encode runs later and dereferences
the mirror's pinned device node (for the DS netaddr) with no reference
of its own. Once a CHANGE notification can re-point the pinned node,
that read becomes use-after-free.
Take a node reference at prepare time and carry it with the stripe in a
per-devinfo nfs4_ff_layoutstat_priv, released in
ff_layout_free_layoutstats(). The priv entries live alongside the
devinfo array: co-allocated for LAYOUTSTATS, an additional member of
nfs4_flexfile_layoutreturn_args for LAYOUTRETURN.
Assisted-by: Claude:claude-fable-5
Signed-off-by: Benjamin Coddington <bcodding@hammerspace.com>
---
fs/nfs/flexfilelayout/flexfilelayout.c | 53 +++++++++++++++++---------
fs/nfs/flexfilelayout/flexfilelayout.h | 11 ++++++
2 files changed, 47 insertions(+), 17 deletions(-)
diff --git a/fs/nfs/flexfilelayout/flexfilelayout.c b/fs/nfs/flexfilelayout/flexfilelayout.c
index b52323e639a2..c07409056292 100644
--- a/fs/nfs/flexfilelayout/flexfilelayout.c
+++ b/fs/nfs/flexfilelayout/flexfilelayout.c
@@ -44,10 +44,11 @@ static void ff_layout_read_record_layoutstats_done(struct rpc_task *task,
static int
ff_layout_mirror_prepare_stats(struct pnfs_layout_hdr *lo,
struct nfs42_layoutstat_devinfo *devinfo,
+ struct nfs4_ff_layoutstat_priv *priv,
int dev_limit, enum nfs4_ff_op_type type);
static void ff_layout_encode_ff_layoutupdate(struct xdr_stream *xdr,
const struct nfs42_layoutstat_devinfo *devinfo,
- struct nfs4_ff_layout_ds_stripe *dss_info);
+ struct nfs4_ff_layoutstat_priv *priv);
static struct pnfs_layout_hdr *
ff_layout_alloc_layout_hdr(struct inode *inode, gfp_t gfp_flags)
@@ -2703,7 +2704,7 @@ ff_layout_prepare_layoutreturn(struct nfs4_layoutreturn_args *args)
spin_lock(&args->inode->i_lock);
ff_args->num_dev = ff_layout_mirror_prepare_stats(
- &ff_layout->generic_hdr, &ff_args->devinfo[0],
+ &ff_layout->generic_hdr, &ff_args->devinfo[0], &ff_args->priv[0],
ARRAY_SIZE(ff_args->devinfo), NFS4_FF_OP_LAYOUTRETURN);
spin_unlock(&args->inode->i_lock);
@@ -2878,10 +2879,11 @@ ff_layout_encode_io_latency(struct xdr_stream *xdr,
static void
ff_layout_encode_ff_layoutupdate(struct xdr_stream *xdr,
const struct nfs42_layoutstat_devinfo *devinfo,
- struct nfs4_ff_layout_ds_stripe *dss_info)
+ struct nfs4_ff_layoutstat_priv *priv)
{
+ struct nfs4_ff_layout_ds_stripe *dss_info = priv->dss_info;
struct nfs4_pnfs_ds_addr *da;
- struct nfs4_pnfs_ds *ds = dss_info->mirror_ds->ds;
+ struct nfs4_pnfs_ds *ds = priv->mirror_ds->ds;
struct nfs_fh *fh = &dss_info->fh_versions[0];
__be32 *p;
@@ -2928,10 +2930,10 @@ ff_layout_encode_layoutstats(struct xdr_stream *xdr, const void *args,
static void
ff_layout_free_layoutstats(struct nfs4_xdr_opaque_data *opaque)
{
- struct nfs4_ff_layout_ds_stripe *dss_info = opaque->data;
- struct nfs4_ff_layout_mirror *mirror = dss_info->mirror;
+ struct nfs4_ff_layoutstat_priv *priv = opaque->data;
- ff_layout_put_mirror(mirror);
+ nfs4_ff_layout_put_deviceid(priv->mirror_ds);
+ ff_layout_put_mirror(priv->dss_info->mirror);
}
static const struct nfs4_xdr_opaque_ops layoutstat_ops = {
@@ -2942,12 +2944,13 @@ static const struct nfs4_xdr_opaque_ops layoutstat_ops = {
static int
ff_layout_mirror_prepare_stats(struct pnfs_layout_hdr *lo,
struct nfs42_layoutstat_devinfo *devinfo,
+ struct nfs4_ff_layoutstat_priv *priv,
int dev_limit, enum nfs4_ff_op_type type)
{
struct nfs4_flexfile_layout *ff_layout = FF_LAYOUT_FROM_HDR(lo);
struct nfs4_ff_layout_mirror *mirror;
struct nfs4_ff_layout_ds_stripe *dss_info;
- struct nfs4_deviceid_node *dev;
+ struct nfs4_ff_layout_ds *mirror_ds;
int i = 0, dss_id;
list_for_each_entry(mirror, &ff_layout->mirrors, mirrors) {
@@ -2955,7 +2958,8 @@ ff_layout_mirror_prepare_stats(struct pnfs_layout_hdr *lo,
dss_info = &mirror->dss[dss_id];
if (i >= dev_limit)
break;
- if (IS_ERR_OR_NULL(dss_info->mirror_ds))
+ mirror_ds = dss_info->mirror_ds;
+ if (IS_ERR_OR_NULL(mirror_ds))
continue;
if (!test_and_clear_bit(NFS4_FF_MIRROR_STAT_AVAIL,
&mirror->flags) &&
@@ -2964,9 +2968,14 @@ ff_layout_mirror_prepare_stats(struct pnfs_layout_hdr *lo,
/* mirror refcount put in cleanup_layoutstats */
if (!refcount_inc_not_zero(&mirror->ref))
continue;
- dev = &dss_info->mirror_ds->id_node;
+ /*
+ * The mirror's pin holds the node while we're under
+ * i_lock; take a reference for the encode, put in
+ * ff_layout_free_layoutstats().
+ */
+ atomic_inc(&mirror_ds->id_node.ref);
memcpy(&devinfo->dev_id,
- &dev->deviceid,
+ &mirror_ds->id_node.deviceid,
NFS4_DEVICEID4_SIZE);
devinfo->offset = 0;
devinfo->length = NFS4_MAX_UINT64;
@@ -2982,9 +2991,12 @@ ff_layout_mirror_prepare_stats(struct pnfs_layout_hdr *lo,
spin_unlock(&mirror->lock);
devinfo->layout_type = LAYOUT_FLEX_FILES;
devinfo->ld_private.ops = &layoutstat_ops;
- devinfo->ld_private.data = &mirror->dss[dss_id];
+ priv->dss_info = dss_info;
+ priv->mirror_ds = mirror_ds;
+ devinfo->ld_private.data = priv;
devinfo++;
+ priv++;
i++;
}
}
@@ -2995,21 +3007,28 @@ static int ff_layout_prepare_layoutstats(struct nfs42_layoutstat_args *args)
{
struct pnfs_layout_hdr *lo;
struct nfs4_flexfile_layout *ff_layout;
+ struct nfs4_ff_layoutstat_priv *priv;
const int dev_count = PNFS_LAYOUTSTATS_MAXDEV;
- /* For now, send at most PNFS_LAYOUTSTATS_MAXDEV statistics */
- args->devinfo = kmalloc_objs(*args->devinfo, dev_count,
- nfs_io_gfp_mask());
+ /*
+ * For now, send at most PNFS_LAYOUTSTATS_MAXDEV statistics.
+ * The per-devinfo private entries are co-allocated after the
+ * devinfo array and freed along with it.
+ */
+ args->devinfo = kmalloc(dev_count * (sizeof(*args->devinfo) +
+ sizeof(*priv)),
+ nfs_io_gfp_mask());
if (!args->devinfo)
return -ENOMEM;
+ priv = (struct nfs4_ff_layoutstat_priv *)&args->devinfo[dev_count];
spin_lock(&args->inode->i_lock);
lo = NFS_I(args->inode)->layout;
if (lo && pnfs_layout_is_valid(lo)) {
ff_layout = FF_LAYOUT_FROM_HDR(lo);
args->num_dev = ff_layout_mirror_prepare_stats(
- &ff_layout->generic_hdr, &args->devinfo[0], dev_count,
- NFS4_FF_OP_LAYOUTSTATS);
+ &ff_layout->generic_hdr, &args->devinfo[0], priv,
+ dev_count, NFS4_FF_OP_LAYOUTSTATS);
} else
args->num_dev = 0;
spin_unlock(&args->inode->i_lock);
diff --git a/fs/nfs/flexfilelayout/flexfilelayout.h b/fs/nfs/flexfilelayout/flexfilelayout.h
index ceed1a9dd697..d6ec80cf8a6e 100644
--- a/fs/nfs/flexfilelayout/flexfilelayout.h
+++ b/fs/nfs/flexfilelayout/flexfilelayout.h
@@ -124,9 +124,20 @@ struct nfs4_flexfile_layout {
unsigned long flags;
};
+/*
+ * Per-devinfo private data for a layoutstats/layoutreturn encode: the
+ * stripe the stats describe plus a reference on its device node so the
+ * node (and its DS addresses) stay valid until the XDR encode runs.
+ */
+struct nfs4_ff_layoutstat_priv {
+ struct nfs4_ff_layout_ds_stripe *dss_info;
+ struct nfs4_ff_layout_ds *mirror_ds;
+};
+
struct nfs4_flexfile_layoutreturn_args {
struct list_head errors;
struct nfs42_layoutstat_devinfo devinfo[FF_LAYOUTSTATS_MAXDEV];
+ struct nfs4_ff_layoutstat_priv priv[FF_LAYOUTSTATS_MAXDEV];
unsigned int num_errors;
unsigned int num_dev;
struct page *pages[1];
--
2.53.0
^ permalink raw reply related [flat|nested] 22+ messages in thread* [PATCH 08/21] NFSv4/flexfiles: Make the pinned device node pointer RCU-managed
2026-08-13 20:42 [PATCH 00/21] NFS: flexfiles device notifications and caching for wide striped layouts Benjamin Coddington
` (6 preceding siblings ...)
2026-08-13 20:43 ` [PATCH 07/21] NFSv4/flexfiles: Hold a device node reference for layoutstats encoding Benjamin Coddington
@ 2026-08-13 20:43 ` Benjamin Coddington
2026-08-13 20:43 ` [PATCH 09/21] pNFS: Add a reresolve_deviceid layout driver hook Benjamin Coddington
` (12 subsequent siblings)
20 siblings, 0 replies; 22+ messages in thread
From: Benjamin Coddington @ 2026-08-13 20:43 UTC (permalink / raw)
To: Trond Myklebust, Anna Schumaker
Cc: linux-nfs, Jonathan Curley, Mike Snitzer, Jeff Layton
Annotate mirror->dss[dss_id].mirror_ds as __rcu and convert the
remaining readers, completing the preparation for re-pointing the pinned
node while I/O is in flight:
- ff_layout_get_mirror_ds() takes its reference under rcu_read_lock()
with atomic_inc_not_zero(), retrying if it races a reset; the resolve
path takes the caller's reference before publishing the node so a
concurrent reset cannot free it under the caller. This is the shape
of get_huge_zero_folio(): try-ref, else create and install with
cmpxchg (retrying on a lost race) holding one reference for the
pointer and one for the caller, torn down elsewhere by xchg + put.
- The availability scans hold rcu_read_lock() across the walk; they
only test flags on the RCU-freed node.
- ff_layout_cancel_io() takes a reference around the cancel/disconnect
calls, which may block.
- ff_layout_mirror_prepare_stats() runs under i_lock, which a future
re-pointing walk must also hold.
- ff_layout_free_mirror() tears down the last reference; no concurrency.
The pointer is still only ever set once per mirror lifetime, so there is
no behavior change; this commit makes the subsequent in-place re-resolve
on CB_NOTIFY_DEVICEID CHANGE safe to introduce.
Assisted-by: Claude:claude-fable-5
Signed-off-by: Benjamin Coddington <bcodding@hammerspace.com>
---
fs/nfs/flexfilelayout/flexfilelayout.c | 29 ++++--
fs/nfs/flexfilelayout/flexfilelayout.h | 2 +-
fs/nfs/flexfilelayout/flexfilelayoutdev.c | 112 ++++++++++++++--------
3 files changed, 92 insertions(+), 51 deletions(-)
diff --git a/fs/nfs/flexfilelayout/flexfilelayout.c b/fs/nfs/flexfilelayout/flexfilelayout.c
index c07409056292..dfc0c298ab4b 100644
--- a/fs/nfs/flexfilelayout/flexfilelayout.c
+++ b/fs/nfs/flexfilelayout/flexfilelayout.c
@@ -313,7 +313,9 @@ static void ff_layout_free_mirror(struct nfs4_ff_layout_mirror *mirror)
cred = rcu_access_pointer(mirror->dss[dss_id].rw_cred);
put_cred(cred);
nfs_close_local_fh(&mirror->dss[dss_id].nfl);
- nfs4_ff_layout_put_deviceid(mirror->dss[dss_id].mirror_ds);
+ /* the last reference to the mirror is gone; no concurrency */
+ nfs4_ff_layout_put_deviceid(rcu_dereference_protected(
+ mirror->dss[dss_id].mirror_ds, 1));
}
kfree(mirror->dss);
@@ -2476,22 +2478,29 @@ static void ff_layout_cancel_io(struct pnfs_layout_segment *lseg)
for (idx = 0; idx < flseg->mirror_array_cnt; idx++) {
mirror = flseg->mirror_array[idx];
for (dss_id = 0; dss_id < mirror->dss_count; dss_id++) {
- mirror_ds = mirror->dss[dss_id].mirror_ds;
- if (IS_ERR_OR_NULL(mirror_ds))
+ rcu_read_lock();
+ mirror_ds = rcu_dereference(mirror->dss[dss_id].mirror_ds);
+ if (IS_ERR_OR_NULL(mirror_ds) ||
+ !atomic_inc_not_zero(&mirror_ds->id_node.ref)) {
+ rcu_read_unlock();
continue;
- ds = mirror->dss[dss_id].mirror_ds->ds;
+ }
+ rcu_read_unlock();
+ ds = mirror_ds->ds;
if (!ds)
- continue;
+ goto next;
ds_clp = ds->ds_clp;
if (!ds_clp)
- continue;
+ goto next;
clnt = ds_clp->cl_rpcclient;
if (!clnt)
- continue;
+ goto next;
if (!rpc_cancel_tasks(clnt, -EAGAIN,
ff_layout_match_io, lseg))
- continue;
+ goto next;
rpc_clnt_disconnect(clnt);
+next:
+ nfs4_ff_layout_put_deviceid(mirror_ds);
}
}
}
@@ -2958,7 +2967,9 @@ ff_layout_mirror_prepare_stats(struct pnfs_layout_hdr *lo,
dss_info = &mirror->dss[dss_id];
if (i >= dev_limit)
break;
- mirror_ds = dss_info->mirror_ds;
+ mirror_ds = rcu_dereference_protected(
+ dss_info->mirror_ds,
+ lockdep_is_held(&lo->plh_inode->i_lock));
if (IS_ERR_OR_NULL(mirror_ds))
continue;
if (!test_and_clear_bit(NFS4_FF_MIRROR_STAT_AVAIL,
diff --git a/fs/nfs/flexfilelayout/flexfilelayout.h b/fs/nfs/flexfilelayout/flexfilelayout.h
index d6ec80cf8a6e..54e87847ee45 100644
--- a/fs/nfs/flexfilelayout/flexfilelayout.h
+++ b/fs/nfs/flexfilelayout/flexfilelayout.h
@@ -79,7 +79,7 @@ struct nfs4_ff_layout_ds_stripe {
struct nfs4_ff_layout_mirror *mirror;
struct nfs4_deviceid devid;
u32 efficiency;
- struct nfs4_ff_layout_ds *mirror_ds;
+ struct nfs4_ff_layout_ds __rcu *mirror_ds;
u32 fh_versions_cnt;
struct nfs_fh *fh_versions;
nfs4_stateid stateid;
diff --git a/fs/nfs/flexfilelayout/flexfilelayoutdev.c b/fs/nfs/flexfilelayout/flexfilelayoutdev.c
index d140e46581e7..4ab55f0874d5 100644
--- a/fs/nfs/flexfilelayout/flexfilelayoutdev.c
+++ b/fs/nfs/flexfilelayout/flexfilelayoutdev.c
@@ -342,35 +342,54 @@ ff_layout_get_mirror_ds(struct pnfs_layout_hdr *lo,
struct nfs4_ff_layout_mirror *mirror,
u32 dss_id)
{
- struct nfs4_ff_layout_ds *mirror_ds;
+ struct nfs4_ff_layout_ds *mirror_ds, *old;
+ struct nfs4_deviceid_node *node;
if (mirror == NULL)
return ERR_PTR(-ENODEV);
- mirror_ds = mirror->dss[dss_id].mirror_ds;
- if (mirror_ds == NULL) {
- struct nfs4_deviceid_node *node;
-
- mirror_ds = ERR_PTR(-ENODEV);
- node = nfs4_find_get_deviceid(NFS_SERVER(lo->plh_inode),
- &mirror->dss[dss_id].devid, lo->plh_lc_cred,
- GFP_KERNEL);
- if (node)
- mirror_ds = FF_LAYOUT_MIRROR_DS(node);
-
- /* check for race with another call to this function */
- if (cmpxchg(&mirror->dss[dss_id].mirror_ds, NULL, mirror_ds) &&
- mirror_ds != ERR_PTR(-ENODEV))
- nfs4_put_deviceid_node(node);
-
- mirror_ds = mirror->dss[dss_id].mirror_ds;
+retry:
+ rcu_read_lock();
+ mirror_ds = rcu_dereference(mirror->dss[dss_id].mirror_ds);
+ if (mirror_ds && !IS_ERR(mirror_ds) &&
+ atomic_inc_not_zero(&mirror_ds->id_node.ref)) {
+ rcu_read_unlock();
+ return mirror_ds;
}
-
+ rcu_read_unlock();
if (IS_ERR(mirror_ds))
return mirror_ds;
- if (!atomic_inc_not_zero(&mirror_ds->id_node.ref))
- return ERR_PTR(-ENODEV);
- return mirror_ds;
+ if (mirror_ds != NULL)
+ /* raced with a reset; the field is being re-pointed */
+ goto retry;
+
+ node = nfs4_find_get_deviceid(NFS_SERVER(lo->plh_inode),
+ &mirror->dss[dss_id].devid, lo->plh_lc_cred,
+ GFP_KERNEL);
+ if (node) {
+ mirror_ds = FF_LAYOUT_MIRROR_DS(node);
+ /*
+ * The caller's reference, taken while we hold the only
+ * reference so a concurrent reset of the installed pointer
+ * cannot free the node under us.
+ */
+ atomic_inc(&node->ref);
+ } else
+ mirror_ds = ERR_PTR(-ENODEV);
+
+ /* check for race with another call to this function */
+ old = unrcu_pointer(cmpxchg(&mirror->dss[dss_id].mirror_ds,
+ NULL, RCU_INITIALIZER(mirror_ds)));
+ if (old == NULL)
+ /* installed: one reference on the mirror, one for the caller */
+ return mirror_ds;
+
+ /* lost the race; use the winner's node instead */
+ if (node) {
+ nfs4_put_deviceid_node(node);
+ nfs4_put_deviceid_node(node);
+ }
+ goto retry;
}
/**
@@ -593,49 +612,60 @@ unsigned int ff_layout_fetch_ds_ioerr(struct pnfs_layout_hdr *lo,
static bool ff_read_layout_has_available_ds(struct pnfs_layout_segment *lseg)
{
struct nfs4_ff_layout_mirror *mirror;
- struct nfs4_deviceid_node *devid;
+ struct nfs4_ff_layout_ds *mirror_ds;
+ bool ret = false;
u32 idx, dss_id;
+ rcu_read_lock();
for (idx = 0; idx < FF_LAYOUT_MIRROR_COUNT(lseg); idx++) {
mirror = FF_LAYOUT_COMP(lseg, idx);
if (!mirror)
continue;
for (dss_id = 0; dss_id < mirror->dss_count; dss_id++) {
- if (!mirror->dss[dss_id].mirror_ds)
- return true;
- if (IS_ERR(mirror->dss[dss_id].mirror_ds))
+ mirror_ds = rcu_dereference(mirror->dss[dss_id].mirror_ds);
+ if (!mirror_ds) {
+ ret = true;
+ goto out;
+ }
+ if (IS_ERR(mirror_ds))
continue;
- devid = &mirror->dss[dss_id].mirror_ds->id_node;
- if (!nfs4_test_deviceid_unavailable(devid))
- return true;
+ if (!nfs4_test_deviceid_unavailable(&mirror_ds->id_node)) {
+ ret = true;
+ goto out;
+ }
}
}
-
- return false;
+out:
+ rcu_read_unlock();
+ return ret;
}
static bool ff_rw_layout_has_available_ds(struct pnfs_layout_segment *lseg)
{
struct nfs4_ff_layout_mirror *mirror;
- struct nfs4_deviceid_node *devid;
+ struct nfs4_ff_layout_ds *mirror_ds;
+ bool ret = false;
u32 idx, dss_id;
+ rcu_read_lock();
for (idx = 0; idx < FF_LAYOUT_MIRROR_COUNT(lseg); idx++) {
mirror = FF_LAYOUT_COMP(lseg, idx);
if (!mirror)
- return false;
+ goto out;
for (dss_id = 0; dss_id < mirror->dss_count; dss_id++) {
- if (IS_ERR(mirror->dss[dss_id].mirror_ds))
- return false;
- if (!mirror->dss[dss_id].mirror_ds)
+ mirror_ds = rcu_dereference(mirror->dss[dss_id].mirror_ds);
+ if (IS_ERR(mirror_ds))
+ goto out;
+ if (!mirror_ds)
continue;
- devid = &mirror->dss[dss_id].mirror_ds->id_node;
- if (nfs4_test_deviceid_unavailable(devid))
- return false;
+ if (nfs4_test_deviceid_unavailable(&mirror_ds->id_node))
+ goto out;
}
}
-
- return FF_LAYOUT_MIRROR_COUNT(lseg) != 0;
+ ret = FF_LAYOUT_MIRROR_COUNT(lseg) != 0;
+out:
+ rcu_read_unlock();
+ return ret;
}
static bool ff_layout_has_available_ds(struct pnfs_layout_segment *lseg)
--
2.53.0
^ permalink raw reply related [flat|nested] 22+ messages in thread* [PATCH 09/21] pNFS: Add a reresolve_deviceid layout driver hook
2026-08-13 20:42 [PATCH 00/21] NFS: flexfiles device notifications and caching for wide striped layouts Benjamin Coddington
` (7 preceding siblings ...)
2026-08-13 20:43 ` [PATCH 08/21] NFSv4/flexfiles: Make the pinned device node pointer RCU-managed Benjamin Coddington
@ 2026-08-13 20:43 ` Benjamin Coddington
2026-08-13 20:43 ` [PATCH 10/21] NFSv4/flexfiles: Implement in-place device re-resolve on CHANGE Benjamin Coddington
` (11 subsequent siblings)
20 siblings, 0 replies; 22+ messages in thread
From: Benjamin Coddington @ 2026-08-13 20:43 UTC (permalink / raw)
To: Trond Myklebust, Anna Schumaker
Cc: linux-nfs, Jonathan Curley, Mike Snitzer, Jeff Layton
RFC 8881 Section 12.2.10 lets a server change a deviceid's mapping under
live layouts by sending CB_NOTIFY_DEVICEID CHANGE instead of recalling
the layouts, but the client's only response today is to unhash the
cached device, which never reaches references pinned inside a layout
driver's segments.
Add a reresolve_deviceid hook to pnfs_layoutdriver_type and a generic
driver, pnfs_layout_reresolve_deviceid_byclid(), that walks every layout
on every server of the client and invokes the hook under the layout
inode's i_lock. Because the final put of a device node can sleep (it
may tear down the DS nfs_client), the hook must not drop references
itself: it collects the nodes it un-pins on a list through the new
nfs4_deviceid_node.put_list member, and the generic driver puts them
once all locks are dropped.
No driver implements the hook yet, and nothing calls the walker: no
behavior change.
Assisted-by: Claude:claude-fable-5
Signed-off-by: Benjamin Coddington <bcodding@hammerspace.com>
---
fs/nfs/pnfs.c | 68 +++++++++++++++++++++++++++++++++++++++++++++++++++
fs/nfs/pnfs.h | 19 ++++++++++++++
2 files changed, 87 insertions(+)
diff --git a/fs/nfs/pnfs.c b/fs/nfs/pnfs.c
index a21128321c0a..3020eee50918 100644
--- a/fs/nfs/pnfs.c
+++ b/fs/nfs/pnfs.c
@@ -2876,6 +2876,74 @@ pnfs_layout_return_unused_byclid(struct nfs_client *clp,
&range);
}
+struct pnfs_reresolve_deviceid_args {
+ const struct pnfs_layoutdriver_type *ld;
+ const struct nfs4_deviceid *id;
+ bool immediate;
+ struct list_head put_list;
+};
+
+static int pnfs_layout_reresolve_deviceid_byserver(struct nfs_server *server,
+ void *data)
+{
+ struct pnfs_reresolve_deviceid_args *args = data;
+ struct pnfs_layout_hdr *lo;
+ struct inode *inode;
+
+ if (server->pnfs_curr_ld != args->ld)
+ return 0;
+
+ rcu_read_lock();
+ list_for_each_entry_rcu(lo, &server->layouts, plh_layouts) {
+ inode = lo->plh_inode;
+ if (!inode)
+ continue;
+ spin_lock(&inode->i_lock);
+ if (lo->plh_inode == inode && pnfs_layout_is_valid(lo))
+ args->ld->reresolve_deviceid(lo, args->id,
+ args->immediate,
+ &args->put_list);
+ spin_unlock(&inode->i_lock);
+ }
+ rcu_read_unlock();
+ return 0;
+}
+
+/*
+ * pnfs_layout_reresolve_deviceid_byclid - re-point live layouts at a
+ * changed deviceid mapping (CB_NOTIFY_DEVICEID CHANGE)
+ *
+ * Walk every layout of @clp's servers using @ld and invoke the driver's
+ * reresolve_deviceid hook for @id. The hook runs under each layout
+ * inode's i_lock and defers its device node puts to the put_list, which
+ * is drained here once all locks are dropped (the final put can sleep).
+ */
+void
+pnfs_layout_reresolve_deviceid_byclid(struct nfs_client *clp,
+ const struct pnfs_layoutdriver_type *ld,
+ const struct nfs4_deviceid *id,
+ bool immediate)
+{
+ struct pnfs_reresolve_deviceid_args args = {
+ .ld = ld,
+ .id = id,
+ .immediate = immediate,
+ .put_list = LIST_HEAD_INIT(args.put_list),
+ };
+ struct nfs4_deviceid_node *node, *tmp;
+
+ if (!ld->reresolve_deviceid)
+ return;
+
+ nfs_client_for_each_server(clp,
+ pnfs_layout_reresolve_deviceid_byserver, &args);
+
+ list_for_each_entry_safe(node, tmp, &args.put_list, put_list) {
+ list_del(&node->put_list);
+ nfs4_put_deviceid_node(node);
+ }
+}
+
/* Check if we have we have a valid layout but if there isn't an intersection
* between the request and the pgio->pg_lseg, put this pgio->pg_lseg away.
*/
diff --git a/fs/nfs/pnfs.h b/fs/nfs/pnfs.h
index bdce7f930c6a..9ec0ffebe3c4 100644
--- a/fs/nfs/pnfs.h
+++ b/fs/nfs/pnfs.h
@@ -170,6 +170,17 @@ struct pnfs_layoutdriver_type {
struct nfs4_deviceid_node * (*alloc_deviceid_node)
(struct nfs_server *server, struct pnfs_device *pdev,
gfp_t gfp_flags);
+ /*
+ * Re-resolve any of @lo's references to the changed deviceid @id
+ * (CB_NOTIFY_DEVICEID CHANGE). Called under @lo's inode i_lock;
+ * must not sleep. Device node references given up here must be
+ * collected on @put_list (via node->put_list) for the caller to
+ * put once all locks are dropped.
+ */
+ void (*reresolve_deviceid)(struct pnfs_layout_hdr *lo,
+ const struct nfs4_deviceid *id,
+ bool immediate,
+ struct list_head *put_list);
int (*prepare_layoutreturn) (struct nfs4_layoutreturn_args *);
@@ -353,6 +364,10 @@ void pnfs_error_mark_layout_for_return(struct inode *inode,
struct pnfs_layout_segment *lseg);
void pnfs_layout_return_unused_byclid(struct nfs_client *clp,
enum pnfs_iomode iomode);
+void pnfs_layout_reresolve_deviceid_byclid(struct nfs_client *clp,
+ const struct pnfs_layoutdriver_type *ld,
+ const struct nfs4_deviceid *id,
+ bool immediate);
int pnfs_layout_handle_reboot(struct nfs_client *clp);
/* nfs4_deviceid_flags */
@@ -373,6 +388,10 @@ struct nfs4_deviceid_node {
struct nfs4_deviceid deviceid;
struct rcu_head rcu;
atomic_t ref;
+ /* deferred-put linkage for in-place re-resolve; owned by whoever
+ * exchanged this node out of its pinning pointer
+ */
+ struct list_head put_list;
};
struct nfs4_deviceid_node *
--
2.53.0
^ permalink raw reply related [flat|nested] 22+ messages in thread* [PATCH 10/21] NFSv4/flexfiles: Implement in-place device re-resolve on CHANGE
2026-08-13 20:42 [PATCH 00/21] NFS: flexfiles device notifications and caching for wide striped layouts Benjamin Coddington
` (8 preceding siblings ...)
2026-08-13 20:43 ` [PATCH 09/21] pNFS: Add a reresolve_deviceid layout driver hook Benjamin Coddington
@ 2026-08-13 20:43 ` Benjamin Coddington
2026-08-13 20:43 ` [PATCH 11/21] NFSv4: Dispatch CB_NOTIFY_DEVICEID CHANGE to an in-place refresh Benjamin Coddington
` (10 subsequent siblings)
20 siblings, 0 replies; 22+ messages in thread
From: Benjamin Coddington @ 2026-08-13 20:43 UTC (permalink / raw)
To: Trond Myklebust, Anna Schumaker
Cc: linux-nfs, Jonathan Curley, Mike Snitzer, Jeff Layton
Implement the reresolve_deviceid hook: walk the layout's mirrors and,
for every stripe whose raw deviceid matches the changed one, exchange
the pinned device node out for NULL. In-flight I/O completes on the old
node through its own reference; the next I/O to the stripe re-resolves
via ff_layout_get_mirror_ds() and picks up the server's new mapping
with a fresh GETDEVICEINFO. The un-pinned references are handed back
on the walker's list to be put outside the locks.
Matching uses the raw deviceid decoded from the layout (dss[].devid), so
stripes whose node was never resolved, or already exchanged, are
untouched. Nothing dispatches CHANGE to the walker yet: no behavior
change.
Assisted-by: Claude:claude-fable-5
Signed-off-by: Benjamin Coddington <bcodding@hammerspace.com>
---
fs/nfs/flexfilelayout/flexfilelayout.c | 36 ++++++++++++++++++++++++++
1 file changed, 36 insertions(+)
diff --git a/fs/nfs/flexfilelayout/flexfilelayout.c b/fs/nfs/flexfilelayout/flexfilelayout.c
index dfc0c298ab4b..bb27bbea6c95 100644
--- a/fs/nfs/flexfilelayout/flexfilelayout.c
+++ b/fs/nfs/flexfilelayout/flexfilelayout.c
@@ -2505,6 +2505,41 @@ static void ff_layout_cancel_io(struct pnfs_layout_segment *lseg)
}
}
+/*
+ * The server changed the mapping for deviceid @id (CB_NOTIFY_DEVICEID
+ * CHANGE). Un-pin every stripe device node resolved from @id in @lo's
+ * mirrors: in-flight I/O completes on the old node via its own
+ * reference, and the next I/O to the stripe re-resolves, picking up the
+ * new mapping with a fresh GETDEVICEINFO (the stale cache entry has
+ * already been unhashed).
+ *
+ * Called under @lo's inode i_lock; the un-pinned references are handed
+ * to the caller on @head to put outside the lock.
+ */
+static void ff_layout_reresolve_deviceid(struct pnfs_layout_hdr *lo,
+ const struct nfs4_deviceid *id,
+ bool immediate,
+ struct list_head *head)
+{
+ struct nfs4_flexfile_layout *flo = FF_LAYOUT_FROM_HDR(lo);
+ struct nfs4_ff_layout_mirror *mirror;
+ struct nfs4_ff_layout_ds *old;
+ u32 dss_id;
+
+ list_for_each_entry(mirror, &flo->mirrors, mirrors) {
+ for (dss_id = 0; dss_id < mirror->dss_count; dss_id++) {
+ if (memcmp(&mirror->dss[dss_id].devid, id,
+ sizeof(*id)) != 0)
+ continue;
+ old = unrcu_pointer(
+ xchg(&mirror->dss[dss_id].mirror_ds, NULL));
+ if (IS_ERR_OR_NULL(old))
+ continue;
+ list_add(&old->id_node.put_list, head);
+ }
+ }
+}
+
static struct pnfs_ds_commit_info *
ff_layout_get_ds_info(struct inode *inode)
{
@@ -3088,6 +3123,7 @@ static struct pnfs_layoutdriver_type flexfilelayout_type = {
.pg_write_ops = &ff_layout_pg_write_ops,
.get_ds_info = ff_layout_get_ds_info,
.free_deviceid_node = ff_layout_free_deviceid_node,
+ .reresolve_deviceid = ff_layout_reresolve_deviceid,
.read_pagelist = ff_layout_read_pagelist,
.write_pagelist = ff_layout_write_pagelist,
.alloc_deviceid_node = ff_layout_alloc_deviceid_node,
--
2.53.0
^ permalink raw reply related [flat|nested] 22+ messages in thread* [PATCH 11/21] NFSv4: Dispatch CB_NOTIFY_DEVICEID CHANGE to an in-place refresh
2026-08-13 20:42 [PATCH 00/21] NFS: flexfiles device notifications and caching for wide striped layouts Benjamin Coddington
` (9 preceding siblings ...)
2026-08-13 20:43 ` [PATCH 10/21] NFSv4/flexfiles: Implement in-place device re-resolve on CHANGE Benjamin Coddington
@ 2026-08-13 20:43 ` Benjamin Coddington
2026-08-13 20:43 ` [PATCH 12/21] NFSv4/flexfiles: Honor ndc_immediate on CB_NOTIFY_DEVICEID CHANGE Benjamin Coddington
` (9 subsequent siblings)
20 siblings, 0 replies; 22+ messages in thread
From: Benjamin Coddington @ 2026-08-13 20:43 UTC (permalink / raw)
To: Trond Myklebust, Anna Schumaker
Cc: linux-nfs, Jonathan Curley, Mike Snitzer, Jeff Layton
nfs4_callback_devicenotify() treated CHANGE identically to DELETE: both
only unhashed the cached device, so references pinned under live
layouts kept sending I/O to the old mapping until the layouts were
freed. Per RFC 8881 Section 12.2.10, CHANGE exists precisely so a
server can change a mapping without recalling the layouts.
For CHANGE, after unhashing the stale cache entry (so re-resolution
cannot re-pin it), invoke the layout driver's re-resolve walker to
re-point pinned references under live layouts. DELETE is unchanged,
and drivers without a reresolve_deviceid hook see no change.
Assisted-by: Claude:claude-fable-5
Signed-off-by: Benjamin Coddington <bcodding@hammerspace.com>
---
fs/nfs/callback_proc.c | 9 +++++++++
1 file changed, 9 insertions(+)
diff --git a/fs/nfs/callback_proc.c b/fs/nfs/callback_proc.c
index 10f2354ba304..64c994790d3f 100644
--- a/fs/nfs/callback_proc.c
+++ b/fs/nfs/callback_proc.c
@@ -390,7 +390,16 @@ __be32 nfs4_callback_devicenotify(void *argp, void *resp,
if (!ld)
continue;
}
+ /*
+ * Unhash the cached device first so re-resolution cannot
+ * re-pin the stale node, then re-point any references
+ * pinned under live layouts (RFC 8881 Section 12.2.10).
+ */
nfs4_delete_deviceid(ld, cps->clp, &dev->cbd_dev_id);
+ if (dev->cbd_notify_type == NOTIFY_DEVICEID4_CHANGE)
+ pnfs_layout_reresolve_deviceid_byclid(cps->clp, ld,
+ &dev->cbd_dev_id,
+ dev->cbd_immediate);
}
pnfs_put_layoutdriver(ld);
out:
--
2.53.0
^ permalink raw reply related [flat|nested] 22+ messages in thread* [PATCH 12/21] NFSv4/flexfiles: Honor ndc_immediate on CB_NOTIFY_DEVICEID CHANGE
2026-08-13 20:42 [PATCH 00/21] NFS: flexfiles device notifications and caching for wide striped layouts Benjamin Coddington
` (10 preceding siblings ...)
2026-08-13 20:43 ` [PATCH 11/21] NFSv4: Dispatch CB_NOTIFY_DEVICEID CHANGE to an in-place refresh Benjamin Coddington
@ 2026-08-13 20:43 ` Benjamin Coddington
2026-08-13 20:43 ` [PATCH 13/21] pNFS: Discard a GETDEVICEINFO reply that raced a CHANGE notification Benjamin Coddington
` (8 subsequent siblings)
20 siblings, 0 replies; 22+ messages in thread
From: Benjamin Coddington @ 2026-08-13 20:43 UTC (permalink / raw)
To: Trond Myklebust, Anna Schumaker
Cc: linux-nfs, Jonathan Curley, Mike Snitzer, Jeff Layton
When a CHANGE notification carries ndc_immediate, RFC 8881 Section
20.12 says the change is enforced immediately and the client might not
be able to complete pending I/O. In addition to un-pinning the stripe's
device node, mark the old node unavailable so paths still holding a
reference stop selecting it; I/O the server rejects is re-driven onto
the freshly resolved mapping by the existing DS error handling.
The walk can also exchange out a node that already carries the new
mapping: a re-resolve that completed between the unhash and the walk
reaching the stripe, raced by this walk or by the walk of a later
CHANGE for the same deviceid. Ripping such a node out is harmless (it
is still hashed, so the next I/O re-pins it from the cache), but it
must not be marked unavailable. Stale vintages are distinguishable by
hashed-ness: every superseded node was unhashed before its walk began
and is never re-inserted, while a fresh node is inserted before it is
installed. Only mark nodes that are no longer hashed.
Without ndc_immediate, pending I/O drains on the old mapping and only
new I/O re-resolves, as before.
Assisted-by: Claude:claude-fable-5
Signed-off-by: Benjamin Coddington <bcodding@hammerspace.com>
---
fs/nfs/flexfilelayout/flexfilelayout.c | 17 +++++++++++++++++
1 file changed, 17 insertions(+)
diff --git a/fs/nfs/flexfilelayout/flexfilelayout.c b/fs/nfs/flexfilelayout/flexfilelayout.c
index bb27bbea6c95..f042551dd23a 100644
--- a/fs/nfs/flexfilelayout/flexfilelayout.c
+++ b/fs/nfs/flexfilelayout/flexfilelayout.c
@@ -2535,6 +2535,23 @@ static void ff_layout_reresolve_deviceid(struct pnfs_layout_hdr *lo,
xchg(&mirror->dss[dss_id].mirror_ds, NULL));
if (IS_ERR_OR_NULL(old))
continue;
+ /*
+ * ndc_immediate: the server is enforcing the change
+ * now and pending I/O may not complete. Mark the
+ * old node unavailable so anyone still holding a
+ * reference stops selecting it; I/O the server
+ * rejects is re-driven onto the new mapping by the
+ * existing DS error handling.
+ *
+ * Only stale vintages are unhashed: a node still
+ * hashed was fetched after the stale cache entry
+ * was removed and already carries the new mapping
+ * (it can be exchanged out here by a walk that
+ * lost a race with the re-resolve, or with the
+ * walk of a later CHANGE); don't mark it.
+ */
+ if (immediate && hlist_unhashed(&old->id_node.node))
+ nfs4_mark_deviceid_unavailable(&old->id_node);
list_add(&old->id_node.put_list, head);
}
}
--
2.53.0
^ permalink raw reply related [flat|nested] 22+ messages in thread* [PATCH 13/21] pNFS: Discard a GETDEVICEINFO reply that raced a CHANGE notification
2026-08-13 20:42 [PATCH 00/21] NFS: flexfiles device notifications and caching for wide striped layouts Benjamin Coddington
` (11 preceding siblings ...)
2026-08-13 20:43 ` [PATCH 12/21] NFSv4/flexfiles: Honor ndc_immediate on CB_NOTIFY_DEVICEID CHANGE Benjamin Coddington
@ 2026-08-13 20:43 ` Benjamin Coddington
2026-08-13 20:43 ` [PATCH 14/21] pNFS: Add deviceid reference query and collection walkers Benjamin Coddington
` (7 subsequent siblings)
20 siblings, 0 replies; 22+ messages in thread
From: Benjamin Coddington @ 2026-08-13 20:43 UTC (permalink / raw)
To: Trond Myklebust, Anna Schumaker
Cc: linux-nfs, Jonathan Curley, Mike Snitzer, Jeff Layton
RFC 8881 Section 18.40.4: a GETDEVICEINFO reply in flight while the
server changes the device mapping may carry the pre-change mapping; if
it is inserted into the cache after the CHANGE notification unhashed
the stale entry, the client re-caches stale data.
Track a change epoch, bumped when a CHANGE notification is processed
before the stale entry is unhashed. nfs4_find_get_deviceid() snapshots
the epoch before issuing GETDEVICEINFO and, serialized against the
unhash by nfs4_deviceid_lock at insert time, discards the reply and
refetches if the epoch moved. A stale insert that instead precedes the
unhash is removed by the unhash itself, so both orderings end with the
post-change mapping.
Assisted-by: Claude:claude-fable-5
Signed-off-by: Benjamin Coddington <bcodding@hammerspace.com>
---
fs/nfs/callback_proc.c | 4 ++++
fs/nfs/pnfs.h | 1 +
fs/nfs/pnfs_dev.c | 25 +++++++++++++++++++++++++
3 files changed, 30 insertions(+)
diff --git a/fs/nfs/callback_proc.c b/fs/nfs/callback_proc.c
index 64c994790d3f..1721c6dd2250 100644
--- a/fs/nfs/callback_proc.c
+++ b/fs/nfs/callback_proc.c
@@ -394,7 +394,11 @@ __be32 nfs4_callback_devicenotify(void *argp, void *resp,
* Unhash the cached device first so re-resolution cannot
* re-pin the stale node, then re-point any references
* pinned under live layouts (RFC 8881 Section 12.2.10).
+ * The epoch bump lets an in-flight GETDEVICEINFO detect
+ * that its reply may predate the change.
*/
+ if (dev->cbd_notify_type == NOTIFY_DEVICEID4_CHANGE)
+ nfs4_deviceid_bump_change_epoch();
nfs4_delete_deviceid(ld, cps->clp, &dev->cbd_dev_id);
if (dev->cbd_notify_type == NOTIFY_DEVICEID4_CHANGE)
pnfs_layout_reresolve_deviceid_byclid(cps->clp, ld,
diff --git a/fs/nfs/pnfs.h b/fs/nfs/pnfs.h
index 9ec0ffebe3c4..f3c55c86c256 100644
--- a/fs/nfs/pnfs.h
+++ b/fs/nfs/pnfs.h
@@ -399,6 +399,7 @@ nfs4_find_get_deviceid(struct nfs_server *server,
const struct nfs4_deviceid *id, const struct cred *cred,
gfp_t gfp_mask);
void nfs4_delete_deviceid(const struct pnfs_layoutdriver_type *, const struct nfs_client *, const struct nfs4_deviceid *);
+void nfs4_deviceid_bump_change_epoch(void);
void nfs4_init_deviceid_node(struct nfs4_deviceid_node *, struct nfs_server *,
const struct nfs4_deviceid *);
bool nfs4_put_deviceid_node(struct nfs4_deviceid_node *);
diff --git a/fs/nfs/pnfs_dev.c b/fs/nfs/pnfs_dev.c
index 274abdd6d5f3..0c39ae8325b3 100644
--- a/fs/nfs/pnfs_dev.c
+++ b/fs/nfs/pnfs_dev.c
@@ -181,6 +181,22 @@ __nfs4_find_get_deviceid(struct nfs_server *server,
return d;
}
+/*
+ * Bumped when a CB_NOTIFY_DEVICEID CHANGE is processed. A GETDEVICEINFO
+ * reply fetched across a change may carry the pre-change mapping; the
+ * epoch lets nfs4_find_get_deviceid() detect that and refetch. The
+ * counter is ordered by nfs4_deviceid_lock: the notification handler
+ * bumps it before unhashing the stale entry under the lock, so an
+ * insert serialized after the unhash observes the new epoch.
+ */
+static atomic_t nfs4_deviceid_change_epoch = ATOMIC_INIT(0);
+
+void
+nfs4_deviceid_bump_change_epoch(void)
+{
+ atomic_inc(&nfs4_deviceid_change_epoch);
+}
+
struct nfs4_deviceid_node *
nfs4_find_get_deviceid(struct nfs_server *server,
const struct nfs4_deviceid *id, const struct cred *cred,
@@ -188,11 +204,14 @@ nfs4_find_get_deviceid(struct nfs_server *server,
{
long hash = nfs4_deviceid_hash(id);
struct nfs4_deviceid_node *d, *new;
+ int epoch;
+retry:
d = __nfs4_find_get_deviceid(server, id, hash);
if (d)
goto found;
+ epoch = atomic_read(&nfs4_deviceid_change_epoch);
new = nfs4_get_device_info(server, id, cred, gfp_mask);
if (!new) {
trace_nfs4_find_deviceid(server, id, -ENOENT);
@@ -200,6 +219,12 @@ nfs4_find_get_deviceid(struct nfs_server *server,
}
spin_lock(&nfs4_deviceid_lock);
+ if (atomic_read(&nfs4_deviceid_change_epoch) != epoch) {
+ /* a mapping changed while we fetched; ours may be stale */
+ spin_unlock(&nfs4_deviceid_lock);
+ server->pnfs_curr_ld->free_deviceid_node(new);
+ goto retry;
+ }
d = __nfs4_find_get_deviceid(server, id, hash);
if (d) {
spin_unlock(&nfs4_deviceid_lock);
--
2.53.0
^ permalink raw reply related [flat|nested] 22+ messages in thread* [PATCH 14/21] pNFS: Add deviceid reference query and collection walkers
2026-08-13 20:42 [PATCH 00/21] NFS: flexfiles device notifications and caching for wide striped layouts Benjamin Coddington
` (12 preceding siblings ...)
2026-08-13 20:43 ` [PATCH 13/21] pNFS: Discard a GETDEVICEINFO reply that raced a CHANGE notification Benjamin Coddington
@ 2026-08-13 20:43 ` Benjamin Coddington
2026-08-13 20:43 ` [PATCH 15/21] NFSv4/pnfs: Recover revoked layouts on a deleted deviceID Benjamin Coddington
` (6 subsequent siblings)
20 siblings, 0 replies; 22+ messages in thread
From: Benjamin Coddington @ 2026-08-13 20:43 UTC (permalink / raw)
To: Trond Myklebust, Anna Schumaker
Cc: linux-nfs, Jonathan Curley, Mike Snitzer, Jeff Layton
The CB_NOTIFY_DEVICEID DELETE race recovery (RFC 8881 Section
18.40.4) needs to ask whether any live layout still references a
deviceID, and to enumerate those layouts for TEST_STATEID. Add a
layout_references_deviceid hook (sibling of reresolve_deviceid; the
flexfiles implementation memcmps each mirror stripe's decoded devid,
valid independent of the pinned device node) and two walkers over
the byserver pattern:
- pnfs_layout_deviceid_referenced_byclid(): boolean existence query,
early-stopping, entirely under i_lock.
- pnfs_layout_collect_deviceid_refs(): collects each matching layout
with the hdr pinned, the inode grabbed with its superblock active
(a pinned hdr does not hold its inode -- same discipline as the
bulk-destroy walker), and the layout stateid and cred snapshotted
under i_lock, so the caller can issue sleeping RPCs against the
collection.
No callers yet; no behavior change.
Assisted-by: Claude:claude-fable-5
Signed-off-by: Benjamin Coddington <bcodding@hammerspace.com>
---
fs/nfs/flexfilelayout/flexfilelayout.c | 17 +++
fs/nfs/pnfs.c | 155 +++++++++++++++++++++++++
fs/nfs/pnfs.h | 29 +++++
3 files changed, 201 insertions(+)
diff --git a/fs/nfs/flexfilelayout/flexfilelayout.c b/fs/nfs/flexfilelayout/flexfilelayout.c
index f042551dd23a..5b88584f4163 100644
--- a/fs/nfs/flexfilelayout/flexfilelayout.c
+++ b/fs/nfs/flexfilelayout/flexfilelayout.c
@@ -2505,6 +2505,22 @@ static void ff_layout_cancel_io(struct pnfs_layout_segment *lseg)
}
}
+/* Called under @lo's inode i_lock. */
+static bool ff_layout_references_deviceid(struct pnfs_layout_hdr *lo,
+ const struct nfs4_deviceid *id)
+{
+ struct nfs4_flexfile_layout *flo = FF_LAYOUT_FROM_HDR(lo);
+ struct nfs4_ff_layout_mirror *mirror;
+ u32 dss_id;
+
+ list_for_each_entry(mirror, &flo->mirrors, mirrors)
+ for (dss_id = 0; dss_id < mirror->dss_count; dss_id++)
+ if (memcmp(&mirror->dss[dss_id].devid, id,
+ sizeof(*id)) == 0)
+ return true;
+ return false;
+}
+
/*
* The server changed the mapping for deviceid @id (CB_NOTIFY_DEVICEID
* CHANGE). Un-pin every stripe device node resolved from @id in @lo's
@@ -3141,6 +3157,7 @@ static struct pnfs_layoutdriver_type flexfilelayout_type = {
.get_ds_info = ff_layout_get_ds_info,
.free_deviceid_node = ff_layout_free_deviceid_node,
.reresolve_deviceid = ff_layout_reresolve_deviceid,
+ .layout_references_deviceid = ff_layout_references_deviceid,
.read_pagelist = ff_layout_read_pagelist,
.write_pagelist = ff_layout_write_pagelist,
.alloc_deviceid_node = ff_layout_alloc_deviceid_node,
diff --git a/fs/nfs/pnfs.c b/fs/nfs/pnfs.c
index 3020eee50918..101fd23f2547 100644
--- a/fs/nfs/pnfs.c
+++ b/fs/nfs/pnfs.c
@@ -2944,6 +2944,161 @@ pnfs_layout_reresolve_deviceid_byclid(struct nfs_client *clp,
}
}
+struct pnfs_deviceid_ref_args {
+ const struct pnfs_layoutdriver_type *ld;
+ const struct nfs4_deviceid *id;
+ struct list_head *result;
+ bool found;
+};
+
+static int pnfs_layout_deviceid_referenced_byserver(
+ struct nfs_server *server, void *data)
+{
+ struct pnfs_deviceid_ref_args *args = data;
+ struct pnfs_layout_hdr *lo;
+ struct inode *inode;
+
+ if (server->pnfs_curr_ld != args->ld)
+ return 0;
+
+ rcu_read_lock();
+ list_for_each_entry_rcu(lo, &server->layouts, plh_layouts) {
+ inode = lo->plh_inode;
+ if (!inode)
+ continue;
+ spin_lock(&inode->i_lock);
+ if (lo->plh_inode == inode && pnfs_layout_is_valid(lo) &&
+ args->ld->layout_references_deviceid(lo, args->id))
+ args->found = true;
+ spin_unlock(&inode->i_lock);
+ if (args->found)
+ break;
+ }
+ rcu_read_unlock();
+ return args->found;
+}
+
+/*
+ * pnfs_layout_deviceid_referenced_byclid - does any live layout of
+ * @clp's servers using @ld still reference deviceid @id?
+ */
+bool
+pnfs_layout_deviceid_referenced_byclid(struct nfs_client *clp,
+ const struct pnfs_layoutdriver_type *ld,
+ const struct nfs4_deviceid *id)
+{
+ struct pnfs_deviceid_ref_args args = {
+ .ld = ld,
+ .id = id,
+ };
+
+ if (!ld->layout_references_deviceid)
+ return false;
+
+ nfs_client_for_each_server(clp,
+ pnfs_layout_deviceid_referenced_byserver, &args);
+ return args.found;
+}
+
+static int pnfs_layout_collect_deviceid_refs_byserver(
+ struct nfs_server *server, void *data)
+{
+ struct pnfs_deviceid_ref_args *args = data;
+ struct nfs4_deviceid_ref *ref;
+ struct pnfs_layout_hdr *lo;
+ struct inode *inode;
+ bool matched;
+
+ if (server->pnfs_curr_ld != args->ld)
+ return 0;
+
+ rcu_read_lock();
+ list_for_each_entry_rcu(lo, &server->layouts, plh_layouts) {
+ inode = lo->plh_inode;
+ if (!inode ||
+ test_bit(NFS_LAYOUT_INODE_FREEING, &lo->plh_flags))
+ continue;
+
+ ref = kzalloc_obj(*ref, GFP_ATOMIC);
+ if (!ref)
+ break; /* act on what was collected */
+
+ spin_lock(&inode->i_lock);
+ matched = lo->plh_inode == inode && pnfs_layout_is_valid(lo) &&
+ args->ld->layout_references_deviceid(lo, args->id);
+ if (matched) {
+ /* a valid layout's lsegs hold hdr references, so
+ * this cannot become the last reference
+ */
+ pnfs_get_layout_hdr(lo);
+ ref->lo = lo;
+ nfs4_stateid_copy(&ref->stateid, &lo->plh_stateid);
+ ref->cred = get_cred(lo->plh_lc_cred);
+ }
+ spin_unlock(&inode->i_lock);
+
+ if (!matched) {
+ kfree(ref);
+ continue;
+ }
+ /* the pinned hdr does not hold the inode: grab it (and
+ * keep the superblock active) for use across RPCs
+ */
+ ref->inode = nfs_igrab_and_active(inode);
+ if (!ref->inode) {
+ pnfs_put_layout_hdr(lo);
+ put_cred(ref->cred);
+ kfree(ref);
+ continue;
+ }
+ list_add_tail(&ref->node, args->result);
+ }
+ rcu_read_unlock();
+ return 0;
+}
+
+/*
+ * pnfs_layout_collect_deviceid_refs - collect live layouts
+ * referencing a deviceid
+ *
+ * Collect every valid layout of @clp's servers using @ld whose
+ * layout_references_deviceid hook matches @id onto @result as
+ * nfs4_deviceid_ref entries safe to use across sleeping RPCs.
+ * Release with pnfs_layout_put_deviceid_refs().
+ */
+void
+pnfs_layout_collect_deviceid_refs(struct nfs_client *clp,
+ const struct pnfs_layoutdriver_type *ld,
+ const struct nfs4_deviceid *id,
+ struct list_head *result)
+{
+ struct pnfs_deviceid_ref_args args = {
+ .ld = ld,
+ .id = id,
+ .result = result,
+ };
+
+ if (!ld->layout_references_deviceid)
+ return;
+
+ nfs_client_for_each_server(clp,
+ pnfs_layout_collect_deviceid_refs_byserver, &args);
+}
+
+void
+pnfs_layout_put_deviceid_refs(struct list_head *result)
+{
+ struct nfs4_deviceid_ref *ref, *tmp;
+
+ list_for_each_entry_safe(ref, tmp, result, node) {
+ list_del(&ref->node);
+ put_cred(ref->cred);
+ pnfs_put_layout_hdr(ref->lo);
+ nfs_iput_and_deactive(ref->inode);
+ kfree(ref);
+ }
+}
+
/* Check if we have we have a valid layout but if there isn't an intersection
* between the request and the pgio->pg_lseg, put this pgio->pg_lseg away.
*/
diff --git a/fs/nfs/pnfs.h b/fs/nfs/pnfs.h
index f3c55c86c256..96ab15c5c6ad 100644
--- a/fs/nfs/pnfs.h
+++ b/fs/nfs/pnfs.h
@@ -181,6 +181,12 @@ struct pnfs_layoutdriver_type {
const struct nfs4_deviceid *id,
bool immediate,
struct list_head *put_list);
+ /*
+ * Does @lo hold any reference to deviceid @id? Called under
+ * @lo's inode i_lock; must not sleep.
+ */
+ bool (*layout_references_deviceid)(struct pnfs_layout_hdr *lo,
+ const struct nfs4_deviceid *id);
int (*prepare_layoutreturn) (struct nfs4_layoutreturn_args *);
@@ -368,6 +374,29 @@ void pnfs_layout_reresolve_deviceid_byclid(struct nfs_client *clp,
const struct pnfs_layoutdriver_type *ld,
const struct nfs4_deviceid *id,
bool immediate);
+bool pnfs_layout_deviceid_referenced_byclid(struct nfs_client *clp,
+ const struct pnfs_layoutdriver_type *ld,
+ const struct nfs4_deviceid *id);
+
+/*
+ * One live layout referencing a deviceID, collected for the
+ * CB_NOTIFY_DEVICEID DELETE recovery: the hdr is pinned, the inode
+ * igrab'd with its superblock active, and the layout stateid and
+ * cred snapshotted for TEST_STATEID.
+ */
+struct nfs4_deviceid_ref {
+ struct list_head node;
+ struct pnfs_layout_hdr *lo;
+ struct inode *inode;
+ nfs4_stateid stateid;
+ const struct cred *cred;
+};
+
+void pnfs_layout_collect_deviceid_refs(struct nfs_client *clp,
+ const struct pnfs_layoutdriver_type *ld,
+ const struct nfs4_deviceid *id,
+ struct list_head *result);
+void pnfs_layout_put_deviceid_refs(struct list_head *result);
int pnfs_layout_handle_reboot(struct nfs_client *clp);
/* nfs4_deviceid_flags */
--
2.53.0
^ permalink raw reply related [flat|nested] 22+ messages in thread* [PATCH 15/21] NFSv4/pnfs: Recover revoked layouts on a deleted deviceID
2026-08-13 20:42 [PATCH 00/21] NFS: flexfiles device notifications and caching for wide striped layouts Benjamin Coddington
` (13 preceding siblings ...)
2026-08-13 20:43 ` [PATCH 14/21] pNFS: Add deviceid reference query and collection walkers Benjamin Coddington
@ 2026-08-13 20:43 ` Benjamin Coddington
2026-08-13 20:43 ` [PATCH 16/21] NFSv4/pnfs: Confirm a deviceID delete via GETDEVICEINFO Benjamin Coddington
` (5 subsequent siblings)
20 siblings, 0 replies; 22+ messages in thread
From: Benjamin Coddington @ 2026-08-13 20:43 UTC (permalink / raw)
To: Trond Myklebust, Anna Schumaker
Cc: linux-nfs, Jonathan Curley, Mike Snitzer, Jeff Layton
RFC 8881 Section 18.40.4: when CB_NOTIFY_DEVICEID deletes a deviceID
the client believes live layouts still reference, layouts referring
to it may have been revoked -- TEST_STATEID each referring layout's
stateid and recover the revoked ones (mark the layout stateid
invalid, free the lsegs, FREE_STATEID to acknowledge).
The callback thread cannot issue fore-channel RPCs, so suspects are
queued on the nfs_client (dedup'd, holding a layoutdriver reference)
and resolved by a new state-manager step keyed on
NFS4CLNT_DEVICEID_DELETE. The worker re-collects the referring
layouts via the deviceid-ref walker rather than trusting a carried
snapshot, so layouts legitimately returned or recalled in between
are skipped. If every referring layout turned out revoked, the
delete is confirmed by the revocations themselves and the cached
device is dropped; a layout the server still considers valid leaves
the device cached (verifying the delete with GETDEVICEINFO comes
next).
Nothing enqueues suspects yet, so no behavior change.
Assisted-by: Claude:claude-fable-5
Signed-off-by: Benjamin Coddington <bcodding@hammerspace.com>
---
fs/nfs/nfs4_fs.h | 2 ++
fs/nfs/nfs4client.c | 2 ++
fs/nfs/nfs4proc.c | 70 +++++++++++++++++++++++++++++++++++++++
fs/nfs/nfs4state.c | 3 ++
fs/nfs/pnfs.c | 67 +++++++++++++++++++++++++++++++++++++
fs/nfs/pnfs.h | 19 +++++++++++
include/linux/nfs_fs_sb.h | 2 ++
7 files changed, 165 insertions(+)
diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h
index b48e5b87cb2a..d642aca0adc3 100644
--- a/fs/nfs/nfs4_fs.h
+++ b/fs/nfs/nfs4_fs.h
@@ -52,6 +52,7 @@ enum nfs4_client_state {
NFS4CLNT_RECALL_ANY_LAYOUT_READ,
NFS4CLNT_RECALL_ANY_LAYOUT_RW,
NFS4CLNT_DELEGRETURN_DELAYED,
+ NFS4CLNT_DEVICEID_DELETE,
};
#define NFS4_RENEW_TIMEOUT 0x01
@@ -493,6 +494,7 @@ int nfs41_discover_server_trunking(struct nfs_client *clp,
struct nfs_client **, const struct cred *);
extern void nfs4_schedule_session_recovery(struct nfs4_session *, int);
extern void nfs41_notify_server(struct nfs_client *);
+extern void nfs4_deviceid_delete_recover_run(struct nfs_client *clp);
bool nfs4_check_serverowner_major_id(struct nfs41_server_owner *o1,
struct nfs41_server_owner *o2);
diff --git a/fs/nfs/nfs4client.c b/fs/nfs/nfs4client.c
index 71c271a1700a..6a2f7522179c 100644
--- a/fs/nfs/nfs4client.c
+++ b/fs/nfs/nfs4client.c
@@ -217,6 +217,7 @@ struct nfs_client *nfs4_alloc_client(const struct nfs_client_initdata *cl_init)
clp->cl_last_renewal = jiffies;
init_waitqueue_head(&clp->cl_lock_waitq);
INIT_LIST_HEAD(&clp->pending_cb_stateids);
+ INIT_LIST_HEAD(&clp->cl_deviceid_deletes);
if (cl_init->minorversion != 0)
__set_bit(NFS_CS_INFINITE_SLOTS, &clp->cl_flags);
@@ -285,6 +286,7 @@ static void nfs4_shutdown_client(struct nfs_client *clp)
nfs4_kill_renewd(clp);
clp->cl_mvops->shutdown_client(clp);
nfs4_destroy_callback(clp);
+ pnfs_deviceid_delete_queue_free(clp);
if (__test_and_clear_bit(NFS_CS_IDMAP, &clp->cl_res_state))
nfs_idmap_delete(clp);
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index 5709c6fea85b..aa1414d2d891 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -10441,6 +10441,76 @@ static int nfs41_free_stateid(struct nfs_server *server,
return ret;
}
+/*
+ * A CB_NOTIFY_DEVICEID DELETE named a deviceID that live layouts still
+ * referenced -- a conformant server never does this (RFC 8881 Section
+ * 20.12), so run the Section 18.40.4 recovery: TEST_STATEID each
+ * referring layout; recover revoked layouts (mark the layout stateid
+ * invalid, free the lsegs, FREE_STATEID). If every referring layout
+ * turned out revoked, the delete is confirmed by the revocations and
+ * the cached device is dropped. A layout the server still considers
+ * valid means the delete cannot be trusted; leave the device cached.
+ */
+static void nfs4_deviceid_delete_recover(struct nfs_client *clp,
+ const struct pnfs_layoutdriver_type *ld,
+ const struct nfs4_deviceid *id)
+{
+ LIST_HEAD(layouts);
+ struct nfs4_deviceid_ref *ref;
+ bool revoked = false;
+ bool referenced = false;
+
+ pnfs_layout_collect_deviceid_refs(clp, ld, id, &layouts);
+
+ list_for_each_entry(ref, &layouts, node) {
+ struct pnfs_layout_hdr *lo = ref->lo;
+ struct inode *inode = ref->inode;
+ LIST_HEAD(head);
+ int status;
+
+ status = nfs41_test_stateid(NFS_SERVER(inode), &ref->stateid,
+ ref->cred);
+ switch (status) {
+ case NFS_OK:
+ referenced = true;
+ break;
+ case -NFS4ERR_ADMIN_REVOKED:
+ case -NFS4ERR_DELEG_REVOKED:
+ case -NFS4ERR_EXPIRED:
+ case -NFS4ERR_BAD_STATEID:
+ spin_lock(&inode->i_lock);
+ if (pnfs_layout_is_valid(lo) &&
+ nfs4_stateid_match_other(&ref->stateid,
+ &lo->plh_stateid))
+ pnfs_mark_layout_stateid_invalid(lo, &head);
+ spin_unlock(&inode->i_lock);
+ pnfs_free_lseg_list(&head);
+ nfs41_free_stateid(NFS_SERVER(inode), &ref->stateid,
+ ref->cred, true);
+ revoked = true;
+ break;
+ default:
+ /* inconclusive; leave the device alone */
+ break;
+ }
+ }
+ pnfs_layout_put_deviceid_refs(&layouts);
+
+ if (revoked && !referenced)
+ nfs4_delete_deviceid(ld, clp, id);
+}
+
+void nfs4_deviceid_delete_recover_run(struct nfs_client *clp)
+{
+ struct nfs4_deviceid_delete *dd;
+
+ while ((dd = pnfs_deviceid_delete_dequeue(clp)) != NULL) {
+ nfs4_deviceid_delete_recover(clp, dd->ld, &dd->id);
+ pnfs_put_layoutdriver(dd->ld);
+ kfree(dd);
+ }
+}
+
static void
nfs41_free_lock_state(struct nfs_server *server, struct nfs4_lock_state *lsp)
{
diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c
index 305a772e5497..fcdb4b55c98a 100644
--- a/fs/nfs/nfs4state.c
+++ b/fs/nfs/nfs4state.c
@@ -2643,6 +2643,9 @@ static void nfs4_state_manager(struct nfs_client *clp)
set_bit(NFS4CLNT_RUN_MANAGER, &clp->cl_state);
}
nfs4_layoutreturn_any_run(clp);
+ if (test_and_clear_bit(NFS4CLNT_DEVICEID_DELETE,
+ &clp->cl_state))
+ nfs4_deviceid_delete_recover_run(clp);
clear_bit(NFS4CLNT_RECALL_RUNNING, &clp->cl_state);
}
diff --git a/fs/nfs/pnfs.c b/fs/nfs/pnfs.c
index 101fd23f2547..368d065cdb20 100644
--- a/fs/nfs/pnfs.c
+++ b/fs/nfs/pnfs.c
@@ -3099,6 +3099,73 @@ pnfs_layout_put_deviceid_refs(struct list_head *result)
}
}
+/*
+ * pnfs_deviceid_delete_mark - queue a suspect deviceID delete
+ *
+ * A CB_NOTIFY_DEVICEID DELETE named @id while a live layout still
+ * references it. Queue it for the state manager, which runs the
+ * RFC 8881 Section 18.40.4 recovery (TEST_STATEID the referring
+ * layouts, then confirm or refute the delete). Duplicates of an
+ * already-queued suspect are dropped.
+ */
+void pnfs_deviceid_delete_mark(struct nfs_client *clp,
+ const struct pnfs_layoutdriver_type *ld,
+ const struct nfs4_deviceid *id)
+{
+ struct nfs4_deviceid_delete *dd, *new;
+
+ new = kzalloc_obj(*new, GFP_KERNEL);
+ if (!new)
+ return; /* lost notification; recovery waits for the next */
+ new->ld = pnfs_find_layoutdriver(ld->id);
+ if (!new->ld) {
+ kfree(new);
+ return;
+ }
+ memcpy(&new->id, id, sizeof(new->id));
+
+ spin_lock(&clp->cl_lock);
+ list_for_each_entry(dd, &clp->cl_deviceid_deletes, list) {
+ if (dd->ld == new->ld &&
+ !memcmp(&dd->id, &new->id, sizeof(dd->id))) {
+ spin_unlock(&clp->cl_lock);
+ pnfs_put_layoutdriver(new->ld);
+ kfree(new);
+ return;
+ }
+ }
+ list_add_tail(&new->list, &clp->cl_deviceid_deletes);
+ spin_unlock(&clp->cl_lock);
+
+ set_bit(NFS4CLNT_DEVICEID_DELETE, &clp->cl_state);
+ nfs4_schedule_state_manager(clp);
+}
+
+struct nfs4_deviceid_delete *pnfs_deviceid_delete_dequeue(
+ struct nfs_client *clp)
+{
+ struct nfs4_deviceid_delete *dd = NULL;
+
+ spin_lock(&clp->cl_lock);
+ if (!list_empty(&clp->cl_deviceid_deletes)) {
+ dd = list_first_entry(&clp->cl_deviceid_deletes,
+ struct nfs4_deviceid_delete, list);
+ list_del(&dd->list);
+ }
+ spin_unlock(&clp->cl_lock);
+ return dd;
+}
+
+void pnfs_deviceid_delete_queue_free(struct nfs_client *clp)
+{
+ struct nfs4_deviceid_delete *dd;
+
+ while ((dd = pnfs_deviceid_delete_dequeue(clp)) != NULL) {
+ pnfs_put_layoutdriver(dd->ld);
+ kfree(dd);
+ }
+}
+
/* Check if we have we have a valid layout but if there isn't an intersection
* between the request and the pgio->pg_lseg, put this pgio->pg_lseg away.
*/
diff --git a/fs/nfs/pnfs.h b/fs/nfs/pnfs.h
index 96ab15c5c6ad..1ce1cbc32f77 100644
--- a/fs/nfs/pnfs.h
+++ b/fs/nfs/pnfs.h
@@ -397,6 +397,25 @@ void pnfs_layout_collect_deviceid_refs(struct nfs_client *clp,
const struct nfs4_deviceid *id,
struct list_head *result);
void pnfs_layout_put_deviceid_refs(struct list_head *result);
+
+/*
+ * A CB_NOTIFY_DEVICEID DELETE naming a deviceID that live layouts
+ * still reference (RFC 8881 Section 18.40.4). Queued on
+ * nfs_client.cl_deviceid_deletes under cl_lock for the state manager
+ * to resolve; holds a layoutdriver reference.
+ */
+struct nfs4_deviceid_delete {
+ struct list_head list;
+ const struct pnfs_layoutdriver_type *ld;
+ struct nfs4_deviceid id;
+};
+
+void pnfs_deviceid_delete_mark(struct nfs_client *clp,
+ const struct pnfs_layoutdriver_type *ld,
+ const struct nfs4_deviceid *id);
+struct nfs4_deviceid_delete *pnfs_deviceid_delete_dequeue(
+ struct nfs_client *clp);
+void pnfs_deviceid_delete_queue_free(struct nfs_client *clp);
int pnfs_layout_handle_reboot(struct nfs_client *clp);
/* nfs4_deviceid_flags */
diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h
index 34d294774f8c..c21220f26675 100644
--- a/include/linux/nfs_fs_sb.h
+++ b/include/linux/nfs_fs_sb.h
@@ -101,6 +101,8 @@ struct nfs_client {
/* The flags used for obtaining the clientid during EXCHANGE_ID */
u32 cl_exchange_flags;
struct nfs4_session *cl_session; /* shared session */
+ /* CB_NOTIFY_DEVICEID DELETE suspects, protected by cl_lock */
+ struct list_head cl_deviceid_deletes;
bool cl_preserve_clid;
struct nfs41_server_owner *cl_serverowner;
struct nfs41_server_scope *cl_serverscope;
--
2.53.0
^ permalink raw reply related [flat|nested] 22+ messages in thread* [PATCH 16/21] NFSv4/pnfs: Confirm a deviceID delete via GETDEVICEINFO
2026-08-13 20:42 [PATCH 00/21] NFS: flexfiles device notifications and caching for wide striped layouts Benjamin Coddington
` (14 preceding siblings ...)
2026-08-13 20:43 ` [PATCH 15/21] NFSv4/pnfs: Recover revoked layouts on a deleted deviceID Benjamin Coddington
@ 2026-08-13 20:43 ` Benjamin Coddington
2026-08-13 20:43 ` [PATCH 17/21] NFSv4/pnfs: Dispatch CB_NOTIFY_DEVICEID DELETE to race recovery Benjamin Coddington
` (4 subsequent siblings)
20 siblings, 0 replies; 22+ messages in thread
From: Benjamin Coddington @ 2026-08-13 20:43 UTC (permalink / raw)
To: Trond Myklebust, Anna Schumaker
Cc: linux-nfs, Jonathan Curley, Mike Snitzer, Jeff Layton
RFC 8881 Section 18.40.4: if TEST_STATEID says at least one layout
referring to the deleted deviceID is still valid, the delete cannot
be trusted -- verify it with GETDEVICEINFO. The device really being
gone while the server also considers a referring layout valid means
the server is faulty; recover by re-establishing the client ID
(nfs4_schedule_lease_recovery drives the prescribed EXCHANGE_ID) and
drop the cached device. Any other answer -- including the device
existing, i.e. an erroneous DELETE -- keeps the cached device and
the layout intact.
The raw-status probe calls nfs4_proc_getdeviceinfo() directly
because nfs4_get_device_info() swallows the RPC status and cannot
distinguish NFS4ERR_NOENT from a transient failure. A one-page
reply buffer is enough: a device too large for it fails with
something other than -ENOENT, which still proves existence.
Still nothing enqueues suspects; no behavior change.
Assisted-by: Claude:claude-fable-5
Signed-off-by: Benjamin Coddington <bcodding@hammerspace.com>
---
fs/nfs/nfs4proc.c | 57 +++++++++++++++++++++++++++++++++++++++++------
1 file changed, 50 insertions(+), 7 deletions(-)
diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c
index aa1414d2d891..13dc6d5ee904 100644
--- a/fs/nfs/nfs4proc.c
+++ b/fs/nfs/nfs4proc.c
@@ -10441,6 +10441,36 @@ static int nfs41_free_stateid(struct nfs_server *server,
return ret;
}
+/*
+ * Ask the server whether a deviceID exists, surfacing the raw status
+ * (nfs4_get_device_info() swallows it). A one-page reply buffer is
+ * enough for the answer: a device too large for it fails with
+ * something other than -ENOENT, which still proves existence.
+ */
+static int nfs4_deviceid_validate(struct nfs_server *server,
+ const struct pnfs_layoutdriver_type *ld,
+ const struct nfs4_deviceid *id, const struct cred *cred)
+{
+ struct pnfs_device pdev;
+ struct page *page;
+ int status;
+
+ page = alloc_page(GFP_KERNEL);
+ if (!page)
+ return -ENOMEM;
+
+ memset(&pdev, 0, sizeof(pdev));
+ memcpy(&pdev.dev_id, id, sizeof(pdev.dev_id));
+ pdev.layout_type = ld->id;
+ pdev.pages = &page;
+ pdev.pglen = PAGE_SIZE;
+ pdev.maxcount = PAGE_SIZE - nfs41_maxgetdevinfo_overhead;
+
+ status = nfs4_proc_getdeviceinfo(server, &pdev, cred);
+ __free_page(page);
+ return status;
+}
+
/*
* A CB_NOTIFY_DEVICEID DELETE named a deviceID that live layouts still
* referenced -- a conformant server never does this (RFC 8881 Section
@@ -10449,16 +10479,21 @@ static int nfs41_free_stateid(struct nfs_server *server,
* invalid, free the lsegs, FREE_STATEID). If every referring layout
* turned out revoked, the delete is confirmed by the revocations and
* the cached device is dropped. A layout the server still considers
- * valid means the delete cannot be trusted; leave the device cached.
+ * valid means the delete cannot be trusted: verify it with
+ * GETDEVICEINFO. The device really being gone under a valid layout
+ * means the server is faulty -- recover by re-establishing the client
+ * ID (Section 18.40.4 prescribes EXCHANGE_ID). Any other answer
+ * (including the device existing: an erroneous DELETE) keeps the
+ * cached device.
*/
static void nfs4_deviceid_delete_recover(struct nfs_client *clp,
const struct pnfs_layoutdriver_type *ld,
const struct nfs4_deviceid *id)
{
LIST_HEAD(layouts);
- struct nfs4_deviceid_ref *ref;
+ struct nfs4_deviceid_ref *ref, *confirm = NULL;
bool revoked = false;
- bool referenced = false;
+ int status;
pnfs_layout_collect_deviceid_refs(clp, ld, id, &layouts);
@@ -10466,13 +10501,13 @@ static void nfs4_deviceid_delete_recover(struct nfs_client *clp,
struct pnfs_layout_hdr *lo = ref->lo;
struct inode *inode = ref->inode;
LIST_HEAD(head);
- int status;
status = nfs41_test_stateid(NFS_SERVER(inode), &ref->stateid,
ref->cred);
switch (status) {
case NFS_OK:
- referenced = true;
+ if (!confirm)
+ confirm = ref;
break;
case -NFS4ERR_ADMIN_REVOKED:
case -NFS4ERR_DELEG_REVOKED:
@@ -10494,10 +10529,18 @@ static void nfs4_deviceid_delete_recover(struct nfs_client *clp,
break;
}
}
- pnfs_layout_put_deviceid_refs(&layouts);
- if (revoked && !referenced)
+ if (confirm) {
+ status = nfs4_deviceid_validate(NFS_SERVER(confirm->inode),
+ ld, id, confirm->cred);
+ if (status == -ENOENT) {
+ nfs4_schedule_lease_recovery(clp);
+ nfs4_delete_deviceid(ld, clp, id);
+ }
+ } else if (revoked) {
nfs4_delete_deviceid(ld, clp, id);
+ }
+ pnfs_layout_put_deviceid_refs(&layouts);
}
void nfs4_deviceid_delete_recover_run(struct nfs_client *clp)
--
2.53.0
^ permalink raw reply related [flat|nested] 22+ messages in thread* [PATCH 17/21] NFSv4/pnfs: Dispatch CB_NOTIFY_DEVICEID DELETE to race recovery
2026-08-13 20:42 [PATCH 00/21] NFS: flexfiles device notifications and caching for wide striped layouts Benjamin Coddington
` (15 preceding siblings ...)
2026-08-13 20:43 ` [PATCH 16/21] NFSv4/pnfs: Confirm a deviceID delete via GETDEVICEINFO Benjamin Coddington
@ 2026-08-13 20:43 ` Benjamin Coddington
2026-08-13 20:43 ` [PATCH 18/21] NFSv4/pnfs: Grow the deviceid cache hash table Benjamin Coddington
` (3 subsequent siblings)
20 siblings, 0 replies; 22+ messages in thread
From: Benjamin Coddington @ 2026-08-13 20:43 UTC (permalink / raw)
To: Trond Myklebust, Anna Schumaker
Cc: linux-nfs, Jonathan Curley, Mike Snitzer, Jeff Layton
Turn on the RFC 8881 Section 18.40.4 DELETE handling: a DELETE for a
deviceID that no live layout references keeps today's cheap behavior
(drop the cached device, no state-manager wake). A DELETE for a
deviceID that live layouts still reference -- which a conformant
server never sends (Section 20.12) -- is deferred to the state
manager, which TEST_STATEIDs the referring layouts, recovers revoked
ones, and confirms or refutes the delete with GETDEVICEINFO.
The deferred case no longer unhashes the device immediately: if the
recovery concludes the DELETE was erroneous (the deviceID still
exists and a referring layout is still valid), the client keeps
using the cached device.
Assisted-by: Claude:claude-fable-5
Signed-off-by: Benjamin Coddington <bcodding@hammerspace.com>
---
fs/nfs/callback_proc.c | 31 +++++++++++++++++++++++--------
1 file changed, 23 insertions(+), 8 deletions(-)
diff --git a/fs/nfs/callback_proc.c b/fs/nfs/callback_proc.c
index 1721c6dd2250..df4cbe7670e7 100644
--- a/fs/nfs/callback_proc.c
+++ b/fs/nfs/callback_proc.c
@@ -391,19 +391,34 @@ __be32 nfs4_callback_devicenotify(void *argp, void *resp,
continue;
}
/*
- * Unhash the cached device first so re-resolution cannot
- * re-pin the stale node, then re-point any references
- * pinned under live layouts (RFC 8881 Section 12.2.10).
- * The epoch bump lets an in-flight GETDEVICEINFO detect
- * that its reply may predate the change.
+ * CHANGE: unhash the cached device first so re-resolution
+ * cannot re-pin the stale node, then re-point any
+ * references pinned under live layouts (RFC 8881 Section
+ * 12.2.10). The epoch bump lets an in-flight
+ * GETDEVICEINFO detect that its reply may predate the
+ * change.
+ *
+ * DELETE: a deviceID still referenced by a live layout
+ * must never be deleted (Section 20.12), so a referenced
+ * one may be racing layout revocation -- defer to the
+ * state manager for the Section 18.40.4 recovery (this
+ * thread cannot issue the fore-channel RPCs it needs).
+ * Unreferenced: "no harm is done" -- drop the cached
+ * device as before.
*/
- if (dev->cbd_notify_type == NOTIFY_DEVICEID4_CHANGE)
+ if (dev->cbd_notify_type == NOTIFY_DEVICEID4_CHANGE) {
nfs4_deviceid_bump_change_epoch();
- nfs4_delete_deviceid(ld, cps->clp, &dev->cbd_dev_id);
- if (dev->cbd_notify_type == NOTIFY_DEVICEID4_CHANGE)
+ nfs4_delete_deviceid(ld, cps->clp, &dev->cbd_dev_id);
pnfs_layout_reresolve_deviceid_byclid(cps->clp, ld,
&dev->cbd_dev_id,
dev->cbd_immediate);
+ } else if (pnfs_layout_deviceid_referenced_byclid(cps->clp,
+ ld, &dev->cbd_dev_id)) {
+ pnfs_deviceid_delete_mark(cps->clp, ld,
+ &dev->cbd_dev_id);
+ } else {
+ nfs4_delete_deviceid(ld, cps->clp, &dev->cbd_dev_id);
+ }
}
pnfs_put_layoutdriver(ld);
out:
--
2.53.0
^ permalink raw reply related [flat|nested] 22+ messages in thread* [PATCH 18/21] NFSv4/pnfs: Grow the deviceid cache hash table
2026-08-13 20:42 [PATCH 00/21] NFS: flexfiles device notifications and caching for wide striped layouts Benjamin Coddington
` (16 preceding siblings ...)
2026-08-13 20:43 ` [PATCH 17/21] NFSv4/pnfs: Dispatch CB_NOTIFY_DEVICEID DELETE to race recovery Benjamin Coddington
@ 2026-08-13 20:43 ` Benjamin Coddington
2026-08-13 20:43 ` [PATCH 19/21] NFSv4/pnfs: Re-home the data-server cache onto hash buckets Benjamin Coddington
` (2 subsequent siblings)
20 siblings, 0 replies; 22+ messages in thread
From: Benjamin Coddington @ 2026-08-13 20:43 UTC (permalink / raw)
To: Trond Myklebust, Anna Schumaker
Cc: linux-nfs, Jonathan Curley, Mike Snitzer, Jeff Layton
The global deviceid cache has 32 buckets shared by every server and
layout type. pNFS striping deployments put hundreds to ~1000
distinct deviceIDs in play per mount (mirrors x stripes), a load
factor of ~31: resolution and re-resolution bursts walk long chains.
Grow to 256 buckets (2KB fixed): load factor ~4 at 1000 devices.
The existing byte-folding hash distributes fine at any power-of-two
size (the multiplier is odd), so bucket count is the only lever.
Alternatives considered and set aside: per-client tables do not help
the headline single-mount striping case, and a resizing rhashtable
is unwarranted for a burst-only (never per-I/O) lookup path.
Assisted-by: Claude:claude-fable-5
Signed-off-by: Benjamin Coddington <bcodding@hammerspace.com>
---
fs/nfs/pnfs_dev.c | 10 +++++++++-
1 file changed, 9 insertions(+), 1 deletion(-)
diff --git a/fs/nfs/pnfs_dev.c b/fs/nfs/pnfs_dev.c
index 0c39ae8325b3..e09f25a26004 100644
--- a/fs/nfs/pnfs_dev.c
+++ b/fs/nfs/pnfs_dev.c
@@ -40,8 +40,16 @@
/*
* Device ID RCU cache. A device ID is unique per server and layout type.
+ *
+ * 256 buckets keep chains short for pNFS striping deployments with
+ * hundreds to ~1000 distinct deviceIDs (load factor ~4 at 1000) at a
+ * fixed cost of 2KB. The hash folds every deviceID byte through
+ * x = x * 37 + byte; 37 is odd, so the low-bit mask distributes
+ * deviceIDs differing in any single byte perfectly at any
+ * power-of-two table size -- bucket count, not the hash, is the
+ * scaling lever.
*/
-#define NFS4_DEVICE_ID_HASH_BITS 5
+#define NFS4_DEVICE_ID_HASH_BITS 8
#define NFS4_DEVICE_ID_HASH_SIZE (1 << NFS4_DEVICE_ID_HASH_BITS)
#define NFS4_DEVICE_ID_HASH_MASK (NFS4_DEVICE_ID_HASH_SIZE - 1)
--
2.53.0
^ permalink raw reply related [flat|nested] 22+ messages in thread* [PATCH 19/21] NFSv4/pnfs: Re-home the data-server cache onto hash buckets
2026-08-13 20:42 [PATCH 00/21] NFS: flexfiles device notifications and caching for wide striped layouts Benjamin Coddington
` (17 preceding siblings ...)
2026-08-13 20:43 ` [PATCH 18/21] NFSv4/pnfs: Grow the deviceid cache hash table Benjamin Coddington
@ 2026-08-13 20:43 ` Benjamin Coddington
2026-08-13 20:43 ` [PATCH 20/21] NFSv4/pnfs: Key the data-server cache by its address set Benjamin Coddington
2026-08-13 20:43 ` [PATCH 21/21] NFSv4/flexfiles: Add a dataserver_nconnect cap Benjamin Coddington
20 siblings, 0 replies; 22+ messages in thread
From: Benjamin Coddington @ 2026-08-13 20:43 UTC (permalink / raw)
To: Trond Myklebust, Anna Schumaker
Cc: linux-nfs, Jonathan Curley, Mike Snitzer, Jeff Layton
Mechanical conversion of the per-net data-server cache from a single
list_head to an array of hlist buckets: nfs4_pnfs_ds.ds_node becomes
an hlist_node, netns init/teardown cover every bucket, and removal
uses hlist_del_init (which needs no bucket reference).
Insertion still targets bucket 0 and lookup still scans every entry,
so behavior and cost are unchanged -- keying the buckets by the DS
address set comes next. Splitting the structural churn from the
keying keeps a bisect able to tell a list-conversion bug from a
hash-key bug.
Assisted-by: Claude:claude-fable-5
Signed-off-by: Benjamin Coddington <bcodding@hammerspace.com>
---
fs/nfs/client.c | 6 ++++--
fs/nfs/netns.h | 5 ++++-
fs/nfs/pnfs.h | 2 +-
fs/nfs/pnfs_nfs.c | 17 ++++++++++-------
4 files changed, 19 insertions(+), 11 deletions(-)
diff --git a/fs/nfs/client.c b/fs/nfs/client.c
index 4dcb91ab3039..10eeb14d251b 100644
--- a/fs/nfs/client.c
+++ b/fs/nfs/client.c
@@ -1283,7 +1283,8 @@ void nfs_clients_init(struct net *net)
INIT_LIST_HEAD(&nn->nfs_volume_list);
#if IS_ENABLED(CONFIG_NFS_V4)
idr_init(&nn->cb_ident_idr);
- INIT_LIST_HEAD(&nn->nfs4_data_server_cache);
+ for (int i = 0; i < NFS4_DS_CACHE_HASH_SIZE; i++)
+ INIT_HLIST_HEAD(&nn->nfs4_data_server_cache[i]);
spin_lock_init(&nn->nfs4_data_server_lock);
#endif /* CONFIG_NFS_V4 */
spin_lock_init(&nn->nfs_client_lock);
@@ -1303,7 +1304,8 @@ void nfs_clients_exit(struct net *net)
WARN_ON_ONCE(!list_empty(&nn->nfs_client_list));
WARN_ON_ONCE(!list_empty(&nn->nfs_volume_list));
#if IS_ENABLED(CONFIG_NFS_V4)
- WARN_ON_ONCE(!list_empty(&nn->nfs4_data_server_cache));
+ for (int i = 0; i < NFS4_DS_CACHE_HASH_SIZE; i++)
+ WARN_ON_ONCE(!hlist_empty(&nn->nfs4_data_server_cache[i]));
#endif /* CONFIG_NFS_V4 */
}
diff --git a/fs/nfs/netns.h b/fs/nfs/netns.h
index 36658579100d..f21118554272 100644
--- a/fs/nfs/netns.h
+++ b/fs/nfs/netns.h
@@ -31,7 +31,10 @@ struct nfs_net {
unsigned short nfs_callback_tcpport;
unsigned short nfs_callback_tcpport6;
int cb_users[NFS4_MAX_MINOR_VERSION + 1];
- struct list_head nfs4_data_server_cache;
+#define NFS4_DS_CACHE_HASH_BITS 8
+#define NFS4_DS_CACHE_HASH_SIZE (1 << NFS4_DS_CACHE_HASH_BITS)
+ /* hashed by nfs4_ds_addrs_hash() over the DS's address set */
+ struct hlist_head nfs4_data_server_cache[NFS4_DS_CACHE_HASH_SIZE];
spinlock_t nfs4_data_server_lock;
#endif /* CONFIG_NFS_V4 */
struct nfs_netns_client *nfs_client;
diff --git a/fs/nfs/pnfs.h b/fs/nfs/pnfs.h
index 1ce1cbc32f77..e0a7ae5fe8e9 100644
--- a/fs/nfs/pnfs.h
+++ b/fs/nfs/pnfs.h
@@ -57,7 +57,7 @@ struct nfs4_pnfs_ds_addr {
};
struct nfs4_pnfs_ds {
- struct list_head ds_node; /* nfs4_pnfs_dev_hlist dev_dslist */
+ struct hlist_node ds_node; /* nfs_net nfs4_data_server_cache */
char *ds_remotestr; /* comma sep list of addrs */
struct list_head ds_addrs;
const struct net *ds_net;
diff --git a/fs/nfs/pnfs_nfs.c b/fs/nfs/pnfs_nfs.c
index f40368f839d0..e15e0059c56b 100644
--- a/fs/nfs/pnfs_nfs.c
+++ b/fs/nfs/pnfs_nfs.c
@@ -604,16 +604,19 @@ _same_data_server_addrs_locked(const struct list_head *dsaddrs1,
}
/*
- * Lookup DS by addresses. nfs4_ds_cache_lock is held
+ * Lookup DS by addresses. nfs4_data_server_lock is held
*/
static struct nfs4_pnfs_ds *
_data_server_lookup_locked(const struct nfs_net *nn, const struct list_head *dsaddrs)
{
struct nfs4_pnfs_ds *ds;
- list_for_each_entry(ds, &nn->nfs4_data_server_cache, ds_node)
- if (_same_data_server_addrs_locked(&ds->ds_addrs, dsaddrs))
- return ds;
+ for (int i = 0; i < NFS4_DS_CACHE_HASH_SIZE; i++)
+ hlist_for_each_entry(ds, &nn->nfs4_data_server_cache[i],
+ ds_node)
+ if (_same_data_server_addrs_locked(&ds->ds_addrs,
+ dsaddrs))
+ return ds;
return NULL;
}
@@ -659,7 +662,7 @@ void nfs4_pnfs_ds_put(struct nfs4_pnfs_ds *ds)
struct nfs_net *nn = net_generic(ds->ds_net, nfs_net_id);
if (refcount_dec_and_lock(&ds->ds_count, &nn->nfs4_data_server_lock)) {
- list_del_init(&ds->ds_node);
+ hlist_del_init(&ds->ds_node);
spin_unlock(&nn->nfs4_data_server_lock);
destroy_ds(ds);
}
@@ -745,10 +748,10 @@ nfs4_pnfs_ds_add(const struct net *net, struct list_head *dsaddrs, gfp_t gfp_fla
list_splice_init(dsaddrs, &ds->ds_addrs);
ds->ds_remotestr = remotestr;
refcount_set(&ds->ds_count, 1);
- INIT_LIST_HEAD(&ds->ds_node);
+ INIT_HLIST_NODE(&ds->ds_node);
ds->ds_net = net;
ds->ds_clp = NULL;
- list_add(&ds->ds_node, &nn->nfs4_data_server_cache);
+ hlist_add_head(&ds->ds_node, &nn->nfs4_data_server_cache[0]);
dprintk("%s add new data server %s\n", __func__,
ds->ds_remotestr);
} else {
--
2.53.0
^ permalink raw reply related [flat|nested] 22+ messages in thread* [PATCH 20/21] NFSv4/pnfs: Key the data-server cache by its address set
2026-08-13 20:42 [PATCH 00/21] NFS: flexfiles device notifications and caching for wide striped layouts Benjamin Coddington
` (18 preceding siblings ...)
2026-08-13 20:43 ` [PATCH 19/21] NFSv4/pnfs: Re-home the data-server cache onto hash buckets Benjamin Coddington
@ 2026-08-13 20:43 ` Benjamin Coddington
2026-08-13 20:43 ` [PATCH 21/21] NFSv4/flexfiles: Add a dataserver_nconnect cap Benjamin Coddington
20 siblings, 0 replies; 22+ messages in thread
From: Benjamin Coddington @ 2026-08-13 20:43 UTC (permalink / raw)
To: Trond Myklebust, Anna Schumaker
Cc: linux-nfs, Jonathan Curley, Mike Snitzer, Jeff Layton
Populating the data-server cache was quadratic: every GETDEVICEINFO
decode scanned the whole per-net cache under one spinlock, with an
O(n*m) address-set comparison per entry. A striping mount with
hundreds to ~1000 distinct data servers pays that in a burst at
first access and again on notification-driven re-resolution.
Hash each DS to a bucket keyed on its address set: per-address
jhash over exactly the fields same_sockaddr() compares, combined by
addition so multipath ordering cannot change the bucket, and the
existing comparator as the in-bucket tiebreaker. Lookup and insert
now touch one bucket; teardown is unchanged (hlist_del_init needs no
bucket).
The comparator also tightens from subset to set-equality to agree
with the order-independent key. The subset test was itself a latent
bug: two data servers where one's address set contained the other's
were wrongly merged into whichever was cached first. A flexfiles or
files-layout device decodes a deterministic multipath list per
GETDEVICEINFO, so equal sets are what servers actually hand out.
Assisted-by: Claude:claude-fable-5
Signed-off-by: Benjamin Coddington <bcodding@hammerspace.com>
---
fs/nfs/pnfs_nfs.c | 73 +++++++++++++++++++++++++++++++++++++++++------
1 file changed, 64 insertions(+), 9 deletions(-)
diff --git a/fs/nfs/pnfs_nfs.c b/fs/nfs/pnfs_nfs.c
index e15e0059c56b..a94f4a2933cd 100644
--- a/fs/nfs/pnfs_nfs.c
+++ b/fs/nfs/pnfs_nfs.c
@@ -15,6 +15,8 @@
#include "nfs4session.h"
#include "internal.h"
+#include <linux/hash.h>
+#include <linux/jhash.h>
#include "pnfs.h"
#include "netns.h"
#include "nfs4trace.h"
@@ -577,8 +579,10 @@ same_sockaddr(struct sockaddr *addr1, struct sockaddr *addr2)
}
/*
- * Checks if 'dsaddrs1' contains a subset of 'dsaddrs2'. If it does,
- * declare a match.
+ * Checks if 'dsaddrs1' and 'dsaddrs2' contain the same set of
+ * addresses. If they do, declare a match. Equal element counts
+ * plus a subset test give set equality (neither side can hold
+ * duplicates: each list decodes one device's multipath addresses).
*/
static bool
_same_data_server_addrs_locked(const struct list_head *dsaddrs1,
@@ -588,6 +592,10 @@ _same_data_server_addrs_locked(const struct list_head *dsaddrs1,
struct sockaddr *sa1, *sa2;
bool match = false;
+ if (list_count_nodes((struct list_head *)dsaddrs1) !=
+ list_count_nodes((struct list_head *)dsaddrs2))
+ return false;
+
list_for_each_entry(da1, dsaddrs1, da_node) {
sa1 = (struct sockaddr *)&da1->da_addr;
match = false;
@@ -603,6 +611,53 @@ _same_data_server_addrs_locked(const struct list_head *dsaddrs1,
return match;
}
+/*
+ * Hash exactly the fields same_sockaddr() compares: family, address
+ * bytes and port. The v6 scope id is deliberately left out -- it
+ * only discriminates link-local addresses, and hashing it would be
+ * harmless but pointless: unequal link-local addresses that share a
+ * bucket are still separated by the comparator.
+ */
+static u32
+nfs4_ds_addr_hash(const struct sockaddr *sa)
+{
+ u32 h = sa->sa_family;
+
+ switch (sa->sa_family) {
+ case AF_INET: {
+ const struct sockaddr_in *a = (const struct sockaddr_in *)sa;
+
+ h = jhash(&a->sin_addr.s_addr, sizeof(a->sin_addr.s_addr), h);
+ h = jhash(&a->sin_port, sizeof(a->sin_port), h);
+ break;
+ }
+ case AF_INET6: {
+ const struct sockaddr_in6 *a = (const struct sockaddr_in6 *)sa;
+
+ h = jhash(&a->sin6_addr, sizeof(a->sin6_addr), h);
+ h = jhash(&a->sin6_port, sizeof(a->sin6_port), h);
+ break;
+ }
+ }
+ return h;
+}
+
+/*
+ * Bucket index for a DS's address set. Per-address hashes combine
+ * by addition so the multipath list order cannot change the bucket,
+ * matching the order-independent set comparison above.
+ */
+static u32
+nfs4_ds_addrs_hash(const struct list_head *dsaddrs)
+{
+ const struct nfs4_pnfs_ds_addr *da;
+ u32 h = 0;
+
+ list_for_each_entry(da, dsaddrs, da_node)
+ h += nfs4_ds_addr_hash((const struct sockaddr *)&da->da_addr);
+ return hash_32(h, NFS4_DS_CACHE_HASH_BITS);
+}
+
/*
* Lookup DS by addresses. nfs4_data_server_lock is held
*/
@@ -610,13 +665,11 @@ static struct nfs4_pnfs_ds *
_data_server_lookup_locked(const struct nfs_net *nn, const struct list_head *dsaddrs)
{
struct nfs4_pnfs_ds *ds;
+ u32 bucket = nfs4_ds_addrs_hash(dsaddrs);
- for (int i = 0; i < NFS4_DS_CACHE_HASH_SIZE; i++)
- hlist_for_each_entry(ds, &nn->nfs4_data_server_cache[i],
- ds_node)
- if (_same_data_server_addrs_locked(&ds->ds_addrs,
- dsaddrs))
- return ds;
+ hlist_for_each_entry(ds, &nn->nfs4_data_server_cache[bucket], ds_node)
+ if (_same_data_server_addrs_locked(&ds->ds_addrs, dsaddrs))
+ return ds;
return NULL;
}
@@ -751,7 +804,9 @@ nfs4_pnfs_ds_add(const struct net *net, struct list_head *dsaddrs, gfp_t gfp_fla
INIT_HLIST_NODE(&ds->ds_node);
ds->ds_net = net;
ds->ds_clp = NULL;
- hlist_add_head(&ds->ds_node, &nn->nfs4_data_server_cache[0]);
+ hlist_add_head(&ds->ds_node,
+ &nn->nfs4_data_server_cache[
+ nfs4_ds_addrs_hash(&ds->ds_addrs)]);
dprintk("%s add new data server %s\n", __func__,
ds->ds_remotestr);
} else {
--
2.53.0
^ permalink raw reply related [flat|nested] 22+ messages in thread* [PATCH 21/21] NFSv4/flexfiles: Add a dataserver_nconnect cap
2026-08-13 20:42 [PATCH 00/21] NFS: flexfiles device notifications and caching for wide striped layouts Benjamin Coddington
` (19 preceding siblings ...)
2026-08-13 20:43 ` [PATCH 20/21] NFSv4/pnfs: Key the data-server cache by its address set Benjamin Coddington
@ 2026-08-13 20:43 ` Benjamin Coddington
20 siblings, 0 replies; 22+ messages in thread
From: Benjamin Coddington @ 2026-08-13 20:43 UTC (permalink / raw)
To: Trond Myklebust, Anna Schumaker
Cc: linux-nfs, Jonathan Curley, Mike Snitzer, Jeff Layton
Data-server clients inherit the MDS nconnect setting. A striping
mount multiplies that by every distinct data server: at ~1000 DSes
and nconnect=16 the client opens ~16k sockets plus their sunrpc slot
tables, and striping already spreads I/O across the data servers, so
a high per-DS transport count buys little for that workload.
Add a dataserver_nconnect module parameter to the flexfiles layout
driver alongside its existing dataserver_timeo and dataserver_retrans
knobs, and thread the value through nfs4_pnfs_ds_connect() to both
the v3 and v4 data-server client setup paths. The default of 0
preserves today's inherit-from-MDS behavior, so the cap is opt-in
and non-regressing. The files layout passes 0, unchanged.
Assisted-by: Claude:claude-fable-5
Signed-off-by: Benjamin Coddington <bcodding@hammerspace.com>
---
fs/nfs/filelayout/filelayoutdev.c | 2 +-
fs/nfs/flexfilelayout/flexfilelayoutdev.c | 7 +++++++
fs/nfs/internal.h | 3 ++-
fs/nfs/nfs3client.c | 9 +++++++--
fs/nfs/nfs4client.c | 5 ++++-
fs/nfs/pnfs.h | 3 ++-
fs/nfs/pnfs_nfs.c | 20 +++++++++++++-------
7 files changed, 36 insertions(+), 13 deletions(-)
diff --git a/fs/nfs/filelayout/filelayoutdev.c b/fs/nfs/filelayout/filelayoutdev.c
index 7226989ee4d5..35c7bcfaa773 100644
--- a/fs/nfs/filelayout/filelayoutdev.c
+++ b/fs/nfs/filelayout/filelayoutdev.c
@@ -279,7 +279,7 @@ nfs4_fl_prepare_ds(struct pnfs_layout_segment *lseg, u32 ds_idx)
goto out_test_devid;
status = nfs4_pnfs_ds_connect(s, ds, devid, dataserver_timeo,
- dataserver_retrans, 4,
+ dataserver_retrans, 0, 4,
s->nfs_client->cl_minorversion);
if (status) {
nfs4_mark_deviceid_unavailable(devid);
diff --git a/fs/nfs/flexfilelayout/flexfilelayoutdev.c b/fs/nfs/flexfilelayout/flexfilelayoutdev.c
index 4ab55f0874d5..a870c9303bf8 100644
--- a/fs/nfs/flexfilelayout/flexfilelayoutdev.c
+++ b/fs/nfs/flexfilelayout/flexfilelayoutdev.c
@@ -20,6 +20,7 @@
static unsigned int dataserver_timeo = NFS_DEF_TCP_TIMEO;
static unsigned int dataserver_retrans;
+static unsigned int dataserver_nconnect;
static bool ff_layout_has_available_ds(struct pnfs_layout_segment *lseg);
@@ -439,6 +440,7 @@ nfs4_ff_layout_prepare_ds(struct pnfs_layout_segment *lseg,
*/
status = nfs4_pnfs_ds_connect(s, ds, &mirror_ds->id_node,
dataserver_timeo, dataserver_retrans,
+ dataserver_nconnect,
mirror_ds->ds_versions[0].version,
mirror_ds->ds_versions[0].minor_version);
@@ -696,3 +698,8 @@ module_param(dataserver_timeo, uint, 0644);
MODULE_PARM_DESC(dataserver_timeo, "The time (in tenths of a second) the "
"NFSv4.1 client waits for a response from a "
" data server before it retries an NFS request.");
+module_param(dataserver_nconnect, uint, 0644);
+MODULE_PARM_DESC(dataserver_nconnect, "The maximum number of connections "
+ "the NFSv4.1 client opens to each data server, "
+ "capping the value inherited from the MDS nconnect "
+ "mount option. 0 (default) applies no cap.");
diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h
index e4533f583632..21007ac2eb0a 100644
--- a/fs/nfs/internal.h
+++ b/fs/nfs/internal.h
@@ -250,6 +250,7 @@ extern struct nfs_client *nfs4_set_ds_client(struct nfs_server *mds_srv,
int ds_addrlen, int ds_proto,
unsigned int ds_timeo,
unsigned int ds_retrans,
+ unsigned int ds_nconnect,
u32 minor_version);
extern struct rpc_clnt *nfs4_find_or_create_ds_client(struct nfs_client *,
struct inode *);
@@ -258,7 +259,7 @@ extern void nfs4_session_limit_xasize(struct nfs_server *server);
extern struct nfs_client *nfs3_set_ds_client(struct nfs_server *mds_srv,
const struct sockaddr_storage *ds_addr, int ds_addrlen,
int ds_proto, unsigned int ds_timeo,
- unsigned int ds_retrans);
+ unsigned int ds_retrans, unsigned int ds_nconnect);
#ifdef CONFIG_PROC_FS
extern int __init nfs_fs_proc_init(void);
extern void nfs_fs_proc_exit(void);
diff --git a/fs/nfs/nfs3client.c b/fs/nfs/nfs3client.c
index 5d97c1d38bb6..cf2f7be4b435 100644
--- a/fs/nfs/nfs3client.c
+++ b/fs/nfs/nfs3client.c
@@ -84,7 +84,8 @@ struct nfs_server *nfs3_clone_server(struct nfs_server *source,
*/
struct nfs_client *nfs3_set_ds_client(struct nfs_server *mds_srv,
const struct sockaddr_storage *ds_addr, int ds_addrlen,
- int ds_proto, unsigned int ds_timeo, unsigned int ds_retrans)
+ int ds_proto, unsigned int ds_timeo, unsigned int ds_retrans,
+ unsigned int ds_nconnect)
{
struct rpc_timeout ds_timeout;
unsigned long connect_timeout = ds_timeo * (ds_retrans + 1) * HZ / 10;
@@ -124,8 +125,12 @@ struct nfs_client *nfs3_set_ds_client(struct nfs_server *mds_srv,
fallthrough;
case XPRT_TRANSPORT_RDMA:
case XPRT_TRANSPORT_TCP:
- if (mds_clp->cl_nconnect > 1)
+ if (mds_clp->cl_nconnect > 1) {
cl_init.nconnect = mds_clp->cl_nconnect;
+ if (ds_nconnect)
+ cl_init.nconnect = min(cl_init.nconnect,
+ ds_nconnect);
+ }
}
if (mds_srv->flags & NFS_MOUNT_NORESVPORT)
diff --git a/fs/nfs/nfs4client.c b/fs/nfs/nfs4client.c
index 6a2f7522179c..56a2a9c57cb2 100644
--- a/fs/nfs/nfs4client.c
+++ b/fs/nfs/nfs4client.c
@@ -793,7 +793,7 @@ static int nfs4_set_client(struct nfs_server *server,
struct nfs_client *nfs4_set_ds_client(struct nfs_server *mds_srv,
const struct sockaddr_storage *ds_addr, int ds_addrlen,
int ds_proto, unsigned int ds_timeo, unsigned int ds_retrans,
- u32 minor_version)
+ unsigned int ds_nconnect, u32 minor_version)
{
struct rpc_timeout ds_timeout;
struct nfs_client *mds_clp = mds_srv->nfs_client;
@@ -831,6 +831,9 @@ struct nfs_client *nfs4_set_ds_client(struct nfs_server *mds_srv,
case XPRT_TRANSPORT_TCP:
if (mds_clp->cl_nconnect > 1) {
cl_init.nconnect = mds_clp->cl_nconnect;
+ if (ds_nconnect)
+ cl_init.nconnect = min(cl_init.nconnect,
+ ds_nconnect);
cl_init.max_connect = NFS_MAX_TRANSPORTS;
}
}
diff --git a/fs/nfs/pnfs.h b/fs/nfs/pnfs.h
index e0a7ae5fe8e9..0ae5329afdab 100644
--- a/fs/nfs/pnfs.h
+++ b/fs/nfs/pnfs.h
@@ -497,7 +497,8 @@ struct nfs4_pnfs_ds *nfs4_pnfs_ds_add(const struct net *net,
void nfs4_pnfs_v3_ds_connect_unload(void);
int nfs4_pnfs_ds_connect(struct nfs_server *mds_srv, struct nfs4_pnfs_ds *ds,
struct nfs4_deviceid_node *devid, unsigned int timeo,
- unsigned int retrans, u32 version, u32 minor_version);
+ unsigned int retrans, unsigned int nconnect,
+ u32 version, u32 minor_version);
struct nfs4_pnfs_ds_addr *nfs4_decode_mp_ds_addr(struct net *net,
struct xdr_stream *xdr,
gfp_t gfp_flags);
diff --git a/fs/nfs/pnfs_nfs.c b/fs/nfs/pnfs_nfs.c
index a94f4a2933cd..4740d41ef1c5 100644
--- a/fs/nfs/pnfs_nfs.c
+++ b/fs/nfs/pnfs_nfs.c
@@ -842,7 +842,8 @@ static struct nfs_client *(*get_v3_ds_connect)(
int ds_addrlen,
int ds_proto,
unsigned int ds_timeo,
- unsigned int ds_retrans);
+ unsigned int ds_retrans,
+ unsigned int ds_nconnect);
static bool load_v3_ds_connect(void)
{
@@ -865,7 +866,8 @@ void nfs4_pnfs_v3_ds_connect_unload(void)
static int _nfs4_pnfs_v3_ds_connect(struct nfs_server *mds_srv,
struct nfs4_pnfs_ds *ds,
unsigned int timeo,
- unsigned int retrans)
+ unsigned int retrans,
+ unsigned int nconnect)
{
struct nfs_client *clp = ERR_PTR(-EIO);
struct nfs_client *mds_clp = mds_srv->nfs_client;
@@ -917,7 +919,7 @@ static int _nfs4_pnfs_v3_ds_connect(struct nfs_server *mds_srv,
ds_proto = XPRT_TRANSPORT_TCP_TLS;
clp = get_v3_ds_connect(mds_srv, &da->da_addr, da->da_addrlen,
- ds_proto, timeo, retrans);
+ ds_proto, timeo, retrans, nconnect);
if (IS_ERR(clp))
continue;
clp->cl_rpcclient->cl_softerr = 0;
@@ -940,6 +942,7 @@ static int _nfs4_pnfs_v4_ds_connect(struct nfs_server *mds_srv,
struct nfs4_pnfs_ds *ds,
unsigned int timeo,
unsigned int retrans,
+ unsigned int nconnect,
u32 minor_version)
{
struct nfs_client *clp = ERR_PTR(-EIO);
@@ -1030,7 +1033,8 @@ static int _nfs4_pnfs_v4_ds_connect(struct nfs_server *mds_srv,
clp = nfs4_set_ds_client(mds_srv, &da->da_addr,
da->da_addrlen, ds_proto,
- timeo, retrans, minor_version);
+ timeo, retrans, nconnect,
+ minor_version);
if (IS_ERR(clp))
continue;
@@ -1063,7 +1067,8 @@ static int _nfs4_pnfs_v4_ds_connect(struct nfs_server *mds_srv,
*/
int nfs4_pnfs_ds_connect(struct nfs_server *mds_srv, struct nfs4_pnfs_ds *ds,
struct nfs4_deviceid_node *devid, unsigned int timeo,
- unsigned int retrans, u32 version, u32 minor_version)
+ unsigned int retrans, unsigned int nconnect,
+ u32 version, u32 minor_version)
{
int err;
@@ -1082,11 +1087,12 @@ int nfs4_pnfs_ds_connect(struct nfs_server *mds_srv, struct nfs4_pnfs_ds *ds,
switch (version) {
case 3:
- err = _nfs4_pnfs_v3_ds_connect(mds_srv, ds, timeo, retrans);
+ err = _nfs4_pnfs_v3_ds_connect(mds_srv, ds, timeo, retrans,
+ nconnect);
break;
case 4:
err = _nfs4_pnfs_v4_ds_connect(mds_srv, ds, timeo, retrans,
- minor_version);
+ nconnect, minor_version);
break;
default:
dprintk("%s: unsupported DS version %d\n", __func__, version);
--
2.53.0
^ permalink raw reply related [flat|nested] 22+ messages in thread