[PATCH v8 0/2] RDMA/rxe: copy send WQE to kernel buffer before processing

[PATCH v8 0/2] RDMA/rxe: copy send WQE to kernel buffer before processing

[Tristan Madani]

The SoftRoCE send queue is mmapped into userspace. Both the requester and
completer read WQE fields directly from shared memory, so userspace can
modify them between kernel reads.

This series copies each send WQE to a kernel-private buffer before
processing in both the requester and completer paths, with field-level
writeback to the shared queue and cache invalidation on completion,
retry, and QP reset.

Patch 1 covers the requester path (req_next_wqe).
Patch 2 covers the completer path (get_wqe).

Changes since v7:
  - Patch 1/2: reject atomic write WQEs with oversized resid to
    prevent OOB read in finish_packet() (sashiko finding).
  - Patch 2/2: write back DMA progress before invalidating completer
    cache on error retry, so req_retry() resumes multi-packet reads
    from the correct offset (sashiko finding).
  - Patch 2/2: invalidate completer cache in COMPST_RNR_RETRY to
    prevent stale state after RNR timer fires (sashiko finding).

Changes since v6:
  - Fixed all sashiko review findings (2 Critical + 9 High).
  - FORTIFY_SOURCE: memcpy into enclosing struct, not inner member.
  - Copy size: max(max_sge * sizeof(ib_sge), max_inline).
  - Opcode bound: IB_WR_REG_MR (was IB_WR_ATOMIC_WRITE).
  - Requester: removed silent num_sge clamp, use mask=0 for invalid opcodes.
  - Completer: validate num_sge + cur_sge, preserve smp_load_acquire state.
  - Both paths: DMA progress writeback (resid, cur_sge, sge_offset).
  - Completer: *wqe_p = NULL for posted state, invalidate on retry.
  - Addressed Zhu Yanjun review feedback.

Changes since v5:
  - Added opcode bounds check and num_sge clamp after copy.

Changes since v4:
  - Resent from the correct sender address.

Changes since v3:
  - Copy full queue element (max_sge SGEs) instead of per-WQE size.
  - Invalidate on QP reset, ERR flush, RESET state check.

Changes since v2:
  - WRITE_ONCE/smp_store_release for field writeback.

Changes since v1:
  - Split into per-path patches. Cache across multi-packet ops.

Tristan Madani (2):
  RDMA/rxe: copy send WQE to kernel buffer before processing
  RDMA/rxe: copy send WQE to kernel buffer in completer path

 drivers/infiniband/sw/rxe/rxe_comp.c  | 75 +++++++++++++++++++++++++-
 drivers/infiniband/sw/rxe/rxe_qp.c    |  2 +
 drivers/infiniband/sw/rxe/rxe_req.c   | 71 ++++++++++++++++++++++---
 drivers/infiniband/sw/rxe/rxe_verbs.h | 12 +++++
 4 files changed, 151 insertions(+), 9 deletions(-)

-- 
2.47.3

[PATCH v8 1/2] RDMA/rxe: copy send WQE to kernel buffer before processing

[Tristan Madani]

The rxe send queue is mapped into userspace via mmap. The requester
processes Work Queue Entries (WQEs) directly from this shared buffer
without first copying them to kernel memory. Userspace can modify WQE
fields (num_sge, sge_offset, SGE entries) between kernel reads,
leading to inconsistent state in copy_data().

This is the send-path counterpart to the receive-path fixes:
  - commit 22b8fbded65b8 ("RDMA/rxe: Fix TOCTOU heap overflow in
    get_srq_wqe")
  - commit d6ab440240a04 ("RDMA/rxe: Copy WQE to local buffer in
    non-SRQ receive path")

Fix by copying the send WQE to a kernel-private buffer in req_next_wqe()
before processing. The copy uses the full queue element data size
(max of max_sge SGEs and max_inline bytes) so that both SGE entries
and inline data are always captured.

After copying, reject out-of-range opcodes and oversized atomic write
residuals by setting mask to 0 so that next_opcode() transitions the
QP to error state through the normal error path. This covers all valid
opcodes up to IB_WR_REG_MR while preventing OOB access in both the
wr_opcode_mask() lookup table and finish_packet() for atomic writes.

The local copy is reused across multi-packet sends to preserve DMA
progress state (cur_sge, sge_offset, resid). Field updates including
DMA progress are written back to the shared queue using WRITE_ONCE()
for individual fields and smp_store_release() for state transitions,
so the completer and userspace observe consistent values without the
tearing risk of bulk memcpy(). DMA progress writeback ensures retries
can resume from the correct offset.

The copy is invalidated when the WQE index advances (last packet sent,
error, local ops, UD oversized), when a retry resets WQE state, on QP
reset, or when the QP enters error state for flush.

Fixes: 8700e3e7c485 ("Soft RoCE driver")
Cc: stable@vger.kernel.org
Signed-off-by: Tristan Madani <tristan@talencesecurity.com>
---

Changes since v7:
  - Reject atomic write WQEs with oversized dma.resid to prevent
    OOB read in finish_packet() (sashiko finding).

Changes since v6:
  - Fixed all sashiko review findings (2 Critical + 9 High).
  - FORTIFY_SOURCE: memcpy into enclosing struct, not inner member.
  - Copy size: max(max_sge * sizeof(ib_sge), max_inline).
  - Opcode bound: IB_WR_REG_MR (was IB_WR_ATOMIC_WRITE).
  - Removed silent num_sge clamp, use mask=0 for invalid opcodes.
  - DMA progress writeback (resid, cur_sge, sge_offset).
  - Addressed Zhu Yanjun review feedback.

Changes since v5:
  - Added opcode bounds check and num_sge clamp after copy.

Changes since v4:
  - Resent from the correct sender address.

Changes since v3:
  - Copy full queue element (max_sge SGEs) instead of per-WQE size.
  - Invalidate on QP reset, ERR flush, RESET state check.

Changes since v2:
  - WRITE_ONCE/smp_store_release for field writeback.

Changes since v1:
  - Split into per-path patches. Cache across multi-packet ops.
---
 drivers/infiniband/sw/rxe/rxe_qp.c    |  1 +
 drivers/infiniband/sw/rxe/rxe_req.c   | 71 +++++++++++++++++++++++++--
 drivers/infiniband/sw/rxe/rxe_verbs.h |  6 +++
 3 files changed, 73 insertions(+), 5 deletions(-)

diff --git a/drivers/infiniband/sw/rxe/rxe_qp.c b/drivers/infiniband/sw/rxe/rxe_qp.c
index 311f285d78a6b..77606c4a039b1 100644
--- a/drivers/infiniband/sw/rxe/rxe_qp.c
+++ b/drivers/infiniband/sw/rxe/rxe_qp.c
@@ -581,6 +581,7 @@ static void rxe_qp_reset(struct rxe_qp *qp)
 	qp->req.need_retry = 0;
 	qp->req.wait_for_rnr_timer = 0;
 	qp->req.noack_pkts = 0;
+	qp->req.send_wqe_valid = false;
 	qp->resp.msn = 0;
 	qp->resp.opcode = -1;
 	qp->resp.drop_msg = 0;
diff --git a/drivers/infiniband/sw/rxe/rxe_req.c b/drivers/infiniband/sw/rxe/rxe_req.c
index 24f5c044363f7..d5a0c46eb8a25 100644
--- a/drivers/infiniband/sw/rxe/rxe_req.c
+++ b/drivers/infiniband/sw/rxe/rxe_req.c
@@ -161,6 +161,26 @@ static void req_check_sq_drain_done(struct rxe_qp *qp)
 	spin_unlock_irqrestore(&qp->state_lock, flags);
 }
 
+static void rxe_req_writeback_wqe(struct rxe_qp *qp)
+{
+	struct rxe_send_wqe *shared = qp->req.shared_wqe;
+	struct rxe_send_wqe *local = &qp->req.send_wqe.wqe;
+
+	if (!qp->req.send_wqe_valid || !shared)
+		return;
+
+	WRITE_ONCE(shared->status, local->status);
+	WRITE_ONCE(shared->first_psn, local->first_psn);
+	WRITE_ONCE(shared->last_psn, local->last_psn);
+	WRITE_ONCE(shared->mask, local->mask);
+	WRITE_ONCE(shared->has_rd_atomic, local->has_rd_atomic);
+	WRITE_ONCE(shared->dma.resid, local->dma.resid);
+	WRITE_ONCE(shared->dma.cur_sge, local->dma.cur_sge);
+	WRITE_ONCE(shared->dma.sge_offset, local->dma.sge_offset);
+	/* State must be last so the completer sees prior updates */
+	smp_store_release(&shared->state, local->state);
+}
+
 static struct rxe_send_wqe *__req_next_wqe(struct rxe_qp *qp)
 {
 	struct rxe_queue *q = qp->sq.queue;
@@ -193,7 +213,35 @@ static struct rxe_send_wqe *req_next_wqe(struct rxe_qp *qp)
 	}
 	spin_unlock_irqrestore(&qp->state_lock, flags);
 
-	wqe->mask = wr_opcode_mask(wqe->wr.opcode, qp);
+	/* Reuse the existing kernel-private copy if still valid */
+	if (qp->req.send_wqe_valid && qp->req.shared_wqe == wqe)
+		return &qp->req.send_wqe.wqe;
+
+	/* Copy WQE from userspace-mapped shared queue to kernel-private
+	 * buffer. Copy size matches the queue element data area.
+	 */
+	memcpy(&qp->req.send_wqe, wqe, sizeof(*wqe) +
+	       max_t(int, qp->sq.max_sge * sizeof(struct ib_sge),
+		    qp->sq.max_inline));
+	qp->req.shared_wqe = wqe;
+	qp->req.send_wqe_valid = true;
+
+	wqe = &qp->req.send_wqe.wqe;
+
+	/* Prevent OOB in wr_opcode_mask() for corrupted opcodes.
+	 * Mask 0 causes next_opcode() to fail, moving QP to error state.
+	 */
+	if (unlikely(wqe->wr.opcode > IB_WR_REG_MR))
+		wqe->mask = 0;
+	else
+		wqe->mask = wr_opcode_mask(wqe->wr.opcode, qp);
+
+	/* Atomic writes carry exactly sizeof(u64) payload.
+	 * Reject oversized resid to prevent OOB read in finish_packet().
+	 */
+	if (unlikely((wqe->mask & RXE_ATOMIC_WRITE_MASK) &&
+		     wqe->dma.resid > sizeof(u64)))
+		wqe->mask = 0;
 	return wqe;
 }
 
@@ -582,9 +630,13 @@ static void update_state(struct rxe_qp *qp, struct rxe_pkt_info *pkt)
 {
 	qp->req.opcode = pkt->opcode;
 
-	if (pkt->mask & RXE_END_MASK)
+	rxe_req_writeback_wqe(qp);
+
+	if (pkt->mask & RXE_END_MASK) {
 		qp->req.wqe_index = queue_next_index(qp->sq.queue,
 						     qp->req.wqe_index);
+		qp->req.send_wqe_valid = false;
+	}
 
 	qp->need_req_skb = 0;
 
@@ -634,7 +686,9 @@ static int rxe_do_local_ops(struct rxe_qp *qp, struct rxe_send_wqe *wqe)
 
 	wqe->state = wqe_state_done;
 	wqe->status = IB_WC_SUCCESS;
+	rxe_req_writeback_wqe(qp);
 	qp->req.wqe_index = queue_next_index(qp->sq.queue, qp->req.wqe_index);
+	qp->req.send_wqe_valid = false;
 
 	return 0;
 }
@@ -666,6 +720,7 @@ int rxe_requester(struct rxe_qp *qp)
 		wqe = __req_next_wqe(qp);
 		spin_unlock_irqrestore(&qp->state_lock, flags);
 		if (wqe) {
+			qp->req.send_wqe_valid = false;
 			wqe->status = IB_WC_WR_FLUSH_ERR;
 			goto err;
 		} else {
@@ -681,6 +736,7 @@ int rxe_requester(struct rxe_qp *qp)
 		qp->req.wait_psn = 0;
 		qp->req.need_retry = 0;
 		qp->req.wait_for_rnr_timer = 0;
+		qp->req.send_wqe_valid = false;
 		spin_unlock_irqrestore(&qp->state_lock, flags);
 		goto exit;
 	}
@@ -695,6 +751,7 @@ int rxe_requester(struct rxe_qp *qp)
 	if (unlikely(qp->req.need_retry && !qp->req.wait_for_rnr_timer)) {
 		req_retry(qp);
 		qp->req.need_retry = 0;
+		qp->req.send_wqe_valid = false;
 	}
 
 	wqe = req_next_wqe(qp);
@@ -772,10 +829,12 @@ int rxe_requester(struct rxe_qp *qp)
 			wqe->last_psn = qp->req.psn;
 			qp->req.psn = (qp->req.psn + 1) & BTH_PSN_MASK;
 			qp->req.opcode = IB_OPCODE_UD_SEND_ONLY;
-			qp->req.wqe_index = queue_next_index(qp->sq.queue,
-						       qp->req.wqe_index);
 			wqe->state = wqe_state_done;
 			wqe->status = IB_WC_SUCCESS;
+			rxe_req_writeback_wqe(qp);
+			qp->req.wqe_index = queue_next_index(qp->sq.queue,
+						       qp->req.wqe_index);
+			qp->req.send_wqe_valid = false;
 			goto done;
 		}
 		payload = mtu;
@@ -839,8 +898,10 @@ int rxe_requester(struct rxe_qp *qp)
 	goto out;
 err:
 	/* update wqe_index for each wqe completion */
-	qp->req.wqe_index = queue_next_index(qp->sq.queue, qp->req.wqe_index);
 	wqe->state = wqe_state_error;
+	rxe_req_writeback_wqe(qp);
+	qp->req.wqe_index = queue_next_index(qp->sq.queue, qp->req.wqe_index);
+	qp->req.send_wqe_valid = false;
 	rxe_qp_error(qp);
 exit:
 	ret = -EAGAIN;
diff --git a/drivers/infiniband/sw/rxe/rxe_verbs.h b/drivers/infiniband/sw/rxe/rxe_verbs.h
index 0f5ffd94643f9..a22dfc6e5ae3c 100644
--- a/drivers/infiniband/sw/rxe/rxe_verbs.h
+++ b/drivers/infiniband/sw/rxe/rxe_verbs.h
@@ -114,6 +114,12 @@ struct rxe_req_info {
 	int			wait_for_rnr_timer;
 	int			noack_pkts;
 	int			again;
+	struct rxe_send_wqe	*shared_wqe;
+	bool			send_wqe_valid;
+	struct {
+		struct rxe_send_wqe	wqe;
+		struct ib_sge		sge[RXE_MAX_SGE];
+	} send_wqe;
 };
 
 struct rxe_comp_info {
-- 
2.47.3

Re: [PATCH v8 1/2] RDMA/rxe: copy send WQE to kernel buffer before processing

[sashiko-bot]

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Critical] RDMA/rxe: broken atomic write resid check in req_next_wqe() allows OOB read
--

commit c2bf5201e4676c1f9f9acad4282470fc916075a3
Author: Tristan Madani <tristmd@gmail.com>

RDMA/rxe: copy send WQE to kernel buffer before processing

This commit copies send WQEs from the userspace-mapped send queue into a
kernel-private buffer in req_next_wqe() before processing, writing updated
fields back to the shared queue via rxe_req_writeback_wqe().

> diff --git a/drivers/infiniband/sw/rxe/rxe_req.c b/drivers/infiniband/sw/rxe/rxe_req.c
> index 4f353abd76f55..45e0e97d17926 100644
> --- a/drivers/infiniband/sw/rxe/rxe_req.c
> +++ b/drivers/infiniband/sw/rxe/rxe_req.c
[ ... ]
> @@ -193,7 +213,35 @@ static struct rxe_send_wqe *req_next_wqe(struct rxe_qp *qp)
>  	}
>  	spin_unlock_irqrestore(&qp->state_lock, flags);
>  
> -	wqe->mask = wr_opcode_mask(wqe->wr.opcode, qp);
> +	/* Reuse the existing kernel-private copy if still valid */
> +	if (qp->req.send_wqe_valid && qp->req.shared_wqe == wqe)
> +		return &qp->req.send_wqe.wqe;
> +
> +	/* Copy WQE from userspace-mapped shared queue to kernel-private
> +	 * buffer. Copy size matches the queue element data area.
> +	 */
> +	memcpy(&qp->req.send_wqe, wqe, sizeof(*wqe) +
> +	       max_t(int, qp->sq.max_sge * sizeof(struct ib_sge),
> +		    qp->sq.max_inline));
> +	qp->req.shared_wqe = wqe;
> +	qp->req.send_wqe_valid = true;
> +
> +	wqe = &qp->req.send_wqe.wqe;
> +
> +	/* Prevent OOB in wr_opcode_mask() for corrupted opcodes.
> +	 * Mask 0 causes next_opcode() to fail, moving QP to error state.
> +	 */
> +	if (unlikely(wqe->wr.opcode > IB_WR_REG_MR))
> +		wqe->mask = 0;
> +	else
> +		wqe->mask = wr_opcode_mask(wqe->wr.opcode, qp);
> +
> +	/* Atomic writes carry exactly sizeof(u64) payload.
> +	 * Reject oversized resid to prevent OOB read in finish_packet().
> +	 */
> +	if (unlikely((wqe->mask & RXE_ATOMIC_WRITE_MASK) &&
> +		     wqe->dma.resid > sizeof(u64)))
> +		wqe->mask = 0;
>  	return wqe;
>  }

[Severity: Critical]
Does this atomic write resid check in req_next_wqe() in
drivers/infiniband/sw/rxe/rxe_req.c actually reject an IB_WR_ATOMIC_WRITE WQE
posted with wqe->dma.resid > sizeof(u64), or can finish_packet() still read
out of bounds past qp->req.send_wqe?

First, wqe->mask is populated by wr_opcode_mask(), which returns flags from
enum rxe_wr_mask (bits 0..7, where WR_ATOMIC_WRITE_MASK is BIT(7)). Testing
wqe->mask against RXE_ATOMIC_WRITE_MASK from enum rxe_hdr_mask (BIT(27))
always evaluates to false.

Second, even if wqe->mask is set to 0, rxe_requester() and next_opcode() do
not reject a WQE based on wqe->mask == 0 when wqe->wr.opcode is
IB_WR_ATOMIC_WRITE. In rxe_requester():

	if (wqe->mask & WR_LOCAL_OP_MASK) {
		...
	}
	...
	opcode = next_opcode(qp, wqe, wqe->wr.opcode);
	if (unlikely(opcode < 0)) {
		wqe->status = IB_WC_LOC_QP_OP_ERR;
		goto err;
	}

	mask = rxe_opcode[opcode].mask;

next_opcode() -> next_opcode_rc() switches directly on wqe->wr.opcode without
checking wqe->mask:

	case IB_WR_ATOMIC_WRITE:
		return IB_OPCODE_RC_ATOMIC_WRITE;

rxe_requester() then sets mask and pkt.mask from rxe_opcode[opcode].mask
(which has RXE_ATOMIC_WRITE_MASK set), clamps payload to mtu (up to 4096
bytes), and calls finish_packet():

	if (pkt->mask & RXE_ATOMIC_WRITE_MASK) {
		memcpy(payload_addr(pkt), wqe->dma.atomic_wr, payload);
		wqe->dma.resid -= payload;
	}

Would this read up to 4096 bytes starting at
qp->req.send_wqe.wqe.dma.atomic_wr, reading past the 512-byte
qp->req.send_wqe.sge array into adjacent memory?

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261030165044.2799619-1-tristan@talencesecurity.com?part=1

[PATCH v8 2/2] RDMA/rxe: copy send WQE to kernel buffer in completer path

[Tristan Madani]

The completer reads send WQEs from the same userspace-mapped shared
queue as the requester. Even after the requester copies the WQE to a
kernel-private buffer (previous patch), the completer still reads
directly from shared memory, leaving it exposed to the same TOCTOU
races.

Fix by copying the WQE in get_wqe() to a completer-private buffer
using the full queue element data size. The local copy is reused
when the same WQE is being processed with unchanged state, which
preserves DMA progress across multi-packet RDMA READ responses.
A fresh copy is taken when the state changes or a new WQE appears.

After copying, reject invalid opcodes (above IB_WR_REG_MR), out-of-
bounds num_sge, and invalid cur_sge by transitioning the QP to error
state through COMPST_ERROR. This prevents OOB SGE walks in copy_data()
from corrupted DMA indices.

The state read from the shared queue is preserved via smp_load_acquire()
and restored after the memcpy to prevent a concurrent userspace write
from overwriting it with a stale value.

DMA progress (resid, cur_sge, sge_offset) is written back alongside
completion status and rd_atomic state on completion and before retry
invalidation, ensuring retries can resume from the correct offset
rather than restarting from zero.

The cache is invalidated on QP reset, when a completion advances to
the next WQE, on error retry, and on RNR retry to ensure fresh state
after the requester retransmits.

Fixes: 8700e3e7c485 ("Soft RoCE driver")
Cc: stable@vger.kernel.org
Signed-off-by: Tristan Madani <tristan@talencesecurity.com>
---

Changes since v7:
  - Write back DMA progress before invalidating completer cache on
    error retry, so req_retry() resumes multi-packet reads from the
    correct offset (sashiko finding).
  - Invalidate completer cache in COMPST_RNR_RETRY to prevent stale
    PSN comparison after RNR timer fires (sashiko finding).

Changes since v6:
  - Fixed all sashiko review findings (2 Critical + 9 High).
  - FORTIFY_SOURCE: memcpy into enclosing struct, not inner member.
  - Copy size: max(max_sge * sizeof(ib_sge), max_inline).
  - Opcode bound: IB_WR_REG_MR (was IB_WR_ATOMIC_WRITE).
  - Validate num_sge + cur_sge, preserve smp_load_acquire state.
  - DMA progress writeback (resid, cur_sge, sge_offset).
  - *wqe_p = NULL for posted state, invalidate on retry.
  - Addressed Zhu Yanjun review feedback.

Changes since v5:
  - Added opcode bounds check and num_sge clamp after copy.

Changes since v4:
  - Resent from the correct sender address.

Changes since v3:
  - Copy full queue element (max_sge SGEs) instead of per-WQE size.
  - Invalidate on QP reset, ERR flush, RESET state check.

Changes since v2:
  - WRITE_ONCE/smp_store_release for field writeback.

Changes since v1:
  - Split into per-path patches. Cache across multi-packet ops.
 drivers/infiniband/sw/rxe/rxe_comp.c  | 75 ++++++++++++++++++++++++++-
 drivers/infiniband/sw/rxe/rxe_qp.c    |  1 +
 drivers/infiniband/sw/rxe/rxe_verbs.h |  6 +++
 3 files changed, 80 insertions(+), 2 deletions(-)

diff --git a/drivers/infiniband/sw/rxe/rxe_comp.c b/drivers/infiniband/sw/rxe/rxe_comp.c
index 1390e861bd1d7..b8bc710dc75e9 100644
--- a/drivers/infiniband/sw/rxe/rxe_comp.c
+++ b/drivers/infiniband/sw/rxe/rxe_comp.c
@@ -137,22 +137,87 @@ void rxe_comp_queue_pkt(struct rxe_qp *qp, struct sk_buff *skb)
 	rxe_sched_task(&qp->send_task);
 }
 
+/* Write back completer WQE fields to shared memory */
+static void rxe_comp_writeback_wqe(struct rxe_qp *qp)
+{
+	struct rxe_send_wqe *shared = qp->comp.shared_wqe;
+	struct rxe_send_wqe *local = &qp->comp.comp_wqe.wqe;
+
+	if (!qp->comp.comp_wqe_valid || !shared)
+		return;
+
+	WRITE_ONCE(shared->status, local->status);
+	WRITE_ONCE(shared->has_rd_atomic, local->has_rd_atomic);
+	WRITE_ONCE(shared->dma.resid, local->dma.resid);
+	WRITE_ONCE(shared->dma.cur_sge, local->dma.cur_sge);
+	WRITE_ONCE(shared->dma.sge_offset, local->dma.sge_offset);
+}
+
 static inline enum comp_state get_wqe(struct rxe_qp *qp,
 				      struct rxe_pkt_info *pkt,
 				      struct rxe_send_wqe **wqe_p)
 {
 	struct rxe_send_wqe *wqe;
+	u32 state;
 
 	/* we come here whether or not we found a response packet to see if
 	 * there are any posted WQEs
 	 */
 	wqe = queue_head(qp->sq.queue, QUEUE_TYPE_FROM_CLIENT);
-	*wqe_p = wqe;
 
 	/* no WQE or requester has not started it yet */
-	if (!wqe || wqe->state == wqe_state_posted)
+	if (!wqe) {
+		*wqe_p = NULL;
 		return pkt ? COMPST_DONE : COMPST_EXIT;
+	}
+
+	/* Pairs with smp_store_release() in rxe_req_writeback_wqe() */
+	state = smp_load_acquire(&wqe->state);
+	if (state == wqe_state_posted) {
+		*wqe_p = NULL;
+		return pkt ? COMPST_DONE : COMPST_EXIT;
+	}
+
+	/* Reuse existing local copy if still processing same WQE
+	 * with unchanged state (preserves DMA progress for multi-packet ops)
+	 */
+	if (qp->comp.comp_wqe_valid && qp->comp.shared_wqe == wqe &&
+	    qp->comp.comp_wqe.wqe.state == state) {
+		wqe = &qp->comp.comp_wqe.wqe;
+		*wqe_p = wqe;
+		goto check_state;
+	}
+
+	/* Copy shared WQE to kernel-private buffer.
+	 * Copy size matches the queue element data area.
+	 */
+	qp->comp.shared_wqe = wqe;
+	memcpy(&qp->comp.comp_wqe, wqe,
+	       sizeof(*wqe) + max_t(int,
+				qp->sq.max_sge * sizeof(struct ib_sge),
+				qp->sq.max_inline));
+	qp->comp.comp_wqe_valid = true;
+
+	/* Preserve the state we observed via smp_load_acquire()
+	 * since memcpy may have overwritten it with a stale read.
+	 */
+	qp->comp.comp_wqe.wqe.state = state;
+
+	/* Reject corrupted opcode or SGE indices, error the QP */
+	if (unlikely(qp->comp.comp_wqe.wqe.wr.opcode > IB_WR_REG_MR) ||
+	    qp->comp.comp_wqe.wqe.dma.num_sge > qp->sq.max_sge ||
+	    (qp->comp.comp_wqe.wqe.dma.resid &&
+	     qp->comp.comp_wqe.wqe.dma.cur_sge >= qp->sq.max_sge)) {
+		qp->comp.comp_wqe.wqe.status = IB_WC_LOC_QP_OP_ERR;
+		wqe = &qp->comp.comp_wqe.wqe;
+		*wqe_p = wqe;
+		return COMPST_ERROR;
+	}
+
+	wqe = &qp->comp.comp_wqe.wqe;
+	*wqe_p = wqe;
 
+check_state:
 	/* WQE does not require an ack */
 	if (wqe->state == wqe_state_done)
 		return COMPST_COMP_WQE;
@@ -454,6 +519,8 @@ static void do_complete(struct rxe_qp *qp, struct rxe_send_wqe *wqe)
 	if (post)
 		make_send_cqe(qp, wqe, &cqe);
 
+	rxe_comp_writeback_wqe(qp);
+	qp->comp.comp_wqe_valid = false;
 	queue_advance_consumer(qp->sq.queue, QUEUE_TYPE_FROM_CLIENT);
 
 	if (post)
@@ -788,6 +855,8 @@ int rxe_completer(struct rxe_qp *qp)
 							RXE_CNT_COMP_RETRY);
 					qp->req.need_retry = 1;
 					qp->comp.started_retry = 1;
+					rxe_comp_writeback_wqe(qp);
+					qp->comp.comp_wqe_valid = false;
 					qp->req.again = 1;
 				}
 				goto done;
@@ -809,6 +878,8 @@ int rxe_completer(struct rxe_qp *qp)
 				 * rnr timer has fired
 				 */
 				qp->req.wait_for_rnr_timer = 1;
+				rxe_comp_writeback_wqe(qp);
+				qp->comp.comp_wqe_valid = false;
 				rxe_dbg_qp(qp, "set rnr nak timer\n");
 				// TODO who protects from destroy_qp??
 				mod_timer(&qp->rnr_nak_timer,
diff --git a/drivers/infiniband/sw/rxe/rxe_qp.c b/drivers/infiniband/sw/rxe/rxe_qp.c
index 77606c4a039b1..450254a131017 100644
--- a/drivers/infiniband/sw/rxe/rxe_qp.c
+++ b/drivers/infiniband/sw/rxe/rxe_qp.c
@@ -582,6 +582,7 @@ static void rxe_qp_reset(struct rxe_qp *qp)
 	qp->req.wait_for_rnr_timer = 0;
 	qp->req.noack_pkts = 0;
 	qp->req.send_wqe_valid = false;
+	qp->comp.comp_wqe_valid = false;
 	qp->resp.msn = 0;
 	qp->resp.opcode = -1;
 	qp->resp.drop_msg = 0;
diff --git a/drivers/infiniband/sw/rxe/rxe_verbs.h b/drivers/infiniband/sw/rxe/rxe_verbs.h
index a22dfc6e5ae3c..6205dbc29dff6 100644
--- a/drivers/infiniband/sw/rxe/rxe_verbs.h
+++ b/drivers/infiniband/sw/rxe/rxe_verbs.h
@@ -130,6 +130,12 @@ struct rxe_comp_info {
 	int			started_retry;
 	u32			retry_cnt;
 	u32			rnr_retry;
+	struct rxe_send_wqe	*shared_wqe;
+	bool			comp_wqe_valid;
+	struct {
+		struct rxe_send_wqe	wqe;
+		struct ib_sge		sge[RXE_MAX_SGE];
+	} comp_wqe;
 };
 
 /* responder states */
-- 
2.47.3

Re: [PATCH v8 2/2] RDMA/rxe: copy send WQE to kernel buffer in completer path

[sashiko-bot]

Thank you for your contribution! Sashiko AI review found 2 potential issue(s) to consider:
- [High] RDMA/rxe: stale comp_wqe cache across RNR retry in rxe_completer()
- [High] RDMA/rxe: unvalidated shared state and status in get_wqe() hit WARN_ON_ONCE
--

commit 249927ff74c8311cd6d2280c79ed9338f558aefd
Author: Tristan Madani <tristmd@gmail.com>

RDMA/rxe: copy send WQE to kernel buffer in completer path

This commit copies send WQEs from the userspace-mapped shared queue into a
completer-private buffer in get_wqe() to prevent TOCTOU races, reuses the
local copy while processing multi-packet responses, validates the opcode and
SGE fields, and writes back DMA progress on completion and retry.

> diff --git a/drivers/infiniband/sw/rxe/rxe_comp.c b/drivers/infiniband/sw/rxe/rxe_comp.c
> index 1390e861bd1d7..b8bc710dc75e9 100644
> --- a/drivers/infiniband/sw/rxe/rxe_comp.c
> +++ b/drivers/infiniband/sw/rxe/rxe_comp.c
> @@ -137,22 +137,87 @@ void rxe_comp_queue_pkt(struct rxe_qp *qp, struct sk_buff *skb)

[ ... ]

> +	/* Pairs with smp_store_release() in rxe_req_writeback_wqe() */
> +	state = smp_load_acquire(&wqe->state);
> +	if (state == wqe_state_posted) {
> +		*wqe_p = NULL;
> +		return pkt ? COMPST_DONE : COMPST_EXIT;
> +	}
> +
> +	/* Reuse existing local copy if still processing same WQE
> +	 * with unchanged state (preserves DMA progress for multi-packet ops)
> +	 */
> +	if (qp->comp.comp_wqe_valid && qp->comp.shared_wqe == wqe &&
> +	    qp->comp.comp_wqe.wqe.state == state) {
> +		wqe = &qp->comp.comp_wqe.wqe;
> +		*wqe_p = wqe;
> +		goto check_state;
> +	}
> +
> +	/* Copy shared WQE to kernel-private buffer.
> +	 * Copy size matches the queue element data area.
> +	 */
> +	qp->comp.shared_wqe = wqe;
> +	memcpy(&qp->comp.comp_wqe, wqe,
> +	       sizeof(*wqe) + max_t(int,
> +				qp->sq.max_sge * sizeof(struct ib_sge),
> +				qp->sq.max_inline));
> +	qp->comp.comp_wqe_valid = true;
> +
> +	/* Preserve the state we observed via smp_load_acquire()
> +	 * since memcpy may have overwritten it with a stale read.
> +	 */
> +	qp->comp.comp_wqe.wqe.state = state;
> +
> +	/* Reject corrupted opcode or SGE indices, error the QP */
> +	if (unlikely(qp->comp.comp_wqe.wqe.wr.opcode > IB_WR_REG_MR) ||
> +	    qp->comp.comp_wqe.wqe.dma.num_sge > qp->sq.max_sge ||
> +	    (qp->comp.comp_wqe.wqe.dma.resid &&
> +	     qp->comp.comp_wqe.wqe.dma.cur_sge >= qp->sq.max_sge)) {
> +		qp->comp.comp_wqe.wqe.status = IB_WC_LOC_QP_OP_ERR;
> +		wqe = &qp->comp.comp_wqe.wqe;
> +		*wqe_p = wqe;
> +		return COMPST_ERROR;
> +	}
> +
> +	wqe = &qp->comp.comp_wqe.wqe;
> +	*wqe_p = wqe;
>  
> +check_state:
>  	/* WQE does not require an ack */
>  	if (wqe->state == wqe_state_done)
>  		return COMPST_COMP_WQE;

[Severity: High]
Can userspace trigger the WARN_ON_ONCE(wqe->status == IB_WC_SUCCESS) in
rxe_completer() or force get_wqe() to overwrite qp->comp.comp_wqe mid-read?

In get_wqe() in drivers/infiniband/sw/rxe/rxe_comp.c, state is loaded from the
userspace-mapped shared queue and wqe (including wqe->status) is copied into
qp->comp.comp_wqe. At check_state, get_wqe() then checks:

	/* WQE caused an error */
	if (wqe->state == wqe_state_error)
		return COMPST_ERROR;

When COMPST_ERROR runs in rxe_completer():

		case COMPST_ERROR:
			WARN_ON_ONCE(wqe->status == IB_WC_SUCCESS);
			do_complete(qp, wqe);
			rxe_qp_error(qp);
			goto exit;

If userspace writes wqe->state = wqe_state_error with wqe->status =
IB_WC_SUCCESS in the mmap'd send queue ring (or overwrites shared->status = 0
after rxe_req_writeback_wqe() sets wqe_state_error), get_wqe() returns
COMPST_ERROR without ensuring wqe->status != IB_WC_SUCCESS. This hits
WARN_ON_ONCE(wqe->status == IB_WC_SUCCESS) in rxe_completer() and causes
do_complete() to see IB_WC_SUCCESS, skipping mandatory error CQE generation
for unsignaled WRs.

In addition, get_wqe() only reuses qp->comp.comp_wqe when
qp->comp.comp_wqe.wqe.state == state. If userspace toggles wqe->state in the
shared ring between response packets of a multi-packet RDMA READ (for example
between wqe_state_pending and wqe_state_processing), won't get_wqe() discard
the cached local copy and overwrite qp->comp.comp_wqe from untrusted shared
memory on each packet?

[ ... ]

> @@ -809,6 +878,8 @@ int rxe_completer(struct rxe_qp *qp)
>  				 * rnr timer has fired
>  				 */
>  				qp->req.wait_for_rnr_timer = 1;
> +				rxe_comp_writeback_wqe(qp);
> +				qp->comp.comp_wqe_valid = false;
>  				rxe_dbg_qp(qp, "set rnr nak timer\n");

[Severity: High]
Can qp->comp.comp_wqe become stale across an RNR retry if rxe_completer()
runs again before rnr_nak_timer fires?

In rxe_completer() in drivers/infiniband/sw/rxe/rxe_comp.c, COMPST_RNR_RETRY
sets qp->req.wait_for_rnr_timer = 1, writes back and invalidates
qp->comp.comp_wqe_valid, and arms qp->rnr_nak_timer, while qp->req.need_retry
remains 0 until rnr_nak_timer() fires.

However, the early exit check at the top of rxe_completer() only checks
qp->req.need_retry:

	if (qp->req.need_retry)
		goto exit;

If rxe_completer() runs again before rnr_nak_timer expires (for example on
the next iteration of rxe_sender(), or when send_task is scheduled by
rxe_post_send(), rxe_comp_queue_pkt(), or rxe_skb_tx_dtor()), it enters
COMPST_GET_WQE -> get_wqe(). Because req_retry() has not run yet, the shared
WQE still has state == wqe_state_pending (or wqe_state_processing), so
get_wqe() copies the pre-retry WQE (with its old first_psn and last_psn) back
into qp->comp.comp_wqe and sets qp->comp.comp_wqe_valid = true.

When rnr_nak_timer() later fires, rxe_requester() in
drivers/infiniband/sw/rxe/rxe_req.c runs:

	if (unlikely(qp->req.need_retry && !qp->req.wait_for_rnr_timer)) {
		req_retry(qp);
		qp->req.need_retry = 0;
		qp->req.send_wqe_valid = false;
	}

This does not clear qp->comp.comp_wqe_valid, and rxe_requester() immediately
retransmits the WQE with updated PSN bounds and sets shared->state back to
wqe_state_pending. When the retried response arrives, get_wqe() sees
qp->comp.comp_wqe_valid == true, qp->comp.shared_wqe == wqe, and
qp->comp.comp_wqe.wqe.state == state (both wqe_state_pending), so it reuses
the stale pre-retry comp_wqe. For a partially completed multi-packet RDMA
READ where first_psn advanced on retry, won't check_ack() compare pkt->psn
against the stale wqe->first_psn and transition the QP to COMPST_ERROR?

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20261030165044.2799619-1-tristan@talencesecurity.com?part=2