svcrdma: harden parsed chunk list against malformed wire values

The RPC/RDMA transport stores wire-supplied chunk positions, lengths, and segment counts verbatim in the parsed chunk list. Consumer functions compute gap lengths and sub-range offsets from these values using unsigned subtraction. A malicious or buggy peer can supply values that cause these subtractions to underflow, exposing slab memory to the Reply channel or driving oversized allocations.

The fix proceeds in two layers. Consumer functions in svc_rdma_rw.c gain bounds checks against the saved inline body length before each unsigned subtraction, closing the immediate underflow paths. The Read list decoder in svc_rdma_recvfrom.c gains a segment-length cap against the receive context's page budget, and the existing page- overrun guard in svc_rdma_build_read_segment() is corrected to release the rw context it has already acquired. These consumer-side fixes are backportable independently.

pcl_for_each_segment() uses an inclusive upper bound that underflows when ch_segcount is zero, turning a zero-segment Write or Reply chunk into an unbounded memory walk. The macro is changed to a half-open bound that naturally produces an empty iteration for ch_segcount == 0. The decoder then also rejects zero-segment chunks at the wire boundary, and reorders pcl_alloc_write() so that only fully-populated chunks appear on the list. The macro fix remains as defense in depth and is safe to backport to trees without the decoder change.

A consolidation pass validates Read chunk positions and overlap invariants once, immediately after decoding, so that future PCL consumers inherit the guarantee without replicating per-site checks.

---
Chris Mason (3):
svcrdma: Fix offset arithmetic in read_chunk_range
svcrdma: fix pcl_for_each_segment for empty chunks
svcrdma: reject Write/Reply chunks with segcount 0

Chuck Lever (3):
svcrdma: validate Read chunk positions before reconstruction
svcrdma: reject oversized Read segments at decode time
svcrdma: Validate Read chunk positions at decode time

 include/linux/sunrpc/svc_rdma_pcl.h     |  4 ++-
 net/sunrpc/xprtrdma/svc_rdma_pcl.c      | 63 ++++++++++++++++++++++++++++++---
 net/sunrpc/xprtrdma/svc_rdma_recvfrom.c | 14 ++++++--
 net/sunrpc/xprtrdma/svc_rdma_rw.c       | 52 ++++++++++++++++++++-------
 4 files changed, 113 insertions(+), 20 deletions(-)

---
base-commit: 887d478bb2115cec0be8caae58bad4d4b3109b1a
change-id: 20260524-rpc-kernel-bugs-fb537a0615ec

Best regards,

Chuck Lever <[email protected]>

From: Chuck Lever <[email protected]>

The RPC/RDMA Read chunk position field is supplied by the remote client and stored verbatim in the parsed chunk list. xdr_count_read_segments() checks only 4-byte alignment; it never compares the position against the received inline body length.

In the single-chunk path, svc_rdma_read_complete_one() splits the head and tail kvecs at ch_position. A position past the inline body underflows the tail length, exposing adjacent slab memory to the upper XDR decoder.

In the multi-chunk path, svc_rdma_read_multiple_chunks() computes gap lengths between chunks as unsigned subtractions from ch_position. Overlapping Read chunks cause these subtractions to underflow. A final position past the inline body likewise underflows the trailing gap length. svc_rdma_copy_inline_range() then copies past the receive buffer into request pages that are returned to the client through the Reply channel.

Bound inline-range copies in svc_rdma_copy_inline_range() against the decoded inline RPC body saved in rc_saved_arg. Reject a single Read chunk positioned beyond that body, and reject multi-chunk lists where accumulated read bytes exceed the next chunk's position. Apply the same position and overlap checks in the call-chunk interleaving path.

---
 net/sunrpc/xprtrdma/svc_rdma_rw.c | 38 ++++++++++++++++++++++++++++++--------
 1 file changed, 30 insertions(+), 8 deletions(-)

index c535e6de9654..eb4bc56ed387 100644
@@ -1065,7 +1065,7 @@ static int svc_rdma_build_read_chunk(struct svc_rqst *rqstp,
  * svc_rdma_copy_inline_range - Copy part of the inline content into pages
  * @rqstp: RPC transaction context
  * @head: context for ongoing I/O
- * @offset: offset into the Receive buffer of region to copy
+ * @offset: offset into the inline content of region to copy
  * @remaining: length of region to copy
  *
  * Take a page at a time from rqstp->rq_pages and copy the inline

@@ -1082,9 +1082,13 @@ static int svc_rdma_copy_inline_range(struct svc_rqst *rqstp,
 				      unsigned int offset,
 				      unsigned int remaining)
 {
-	unsigned char *dst, *src = head->rc_recv_buf;
+	unsigned char *dst, *src = head->rc_saved_arg.head[0].iov_base;
+	unsigned int inline_len = head->rc_saved_arg.head[0].iov_len;
 	unsigned int page_no, numpages;
 
+	if (offset > inline_len || remaining > inline_len - offset)
+		return -EINVAL;
+
 	numpages = PAGE_ALIGN(head->rc_pageoff + remaining) >> PAGE_SHIFT;
 	for (page_no = 0; page_no < numpages; page_no++) {
 		unsigned int page_len;

@@ -1135,9 +1139,10 @@ svc_rdma_read_multiple_chunks(struct svc_rqst *rqstp,
 {
 	const struct svc_rdma_pcl *pcl = &head->rc_read_pcl;
 	struct svc_rdma_chunk *chunk, *next;
-	unsigned int start, length;
+	unsigned int inline_len, start, length;
 	int ret;
 
+	inline_len = head->rc_saved_arg.head[0].iov_len;
 	start = 0;
 	chunk = pcl_first_chunk(pcl);
 	length = chunk->ch_position;

@@ -1155,6 +1160,8 @@ svc_rdma_read_multiple_chunks(struct svc_rqst *rqstp,
 			break;
 
 		start += length;
+		if (head->rc_readbytes > next->ch_position)
+			return -EINVAL;
 		length = next->ch_position - head->rc_readbytes;
 		ret = svc_rdma_copy_inline_range(rqstp, head, start, length);
 		if (ret < 0)

@@ -1162,7 +1169,9 @@ svc_rdma_read_multiple_chunks(struct svc_rqst *rqstp,
 	}
 
 	start += length;
-	length = head->rc_byte_len - start;
+	if (start > inline_len)
+		return -EINVAL;
+	length = inline_len - start;
 	return svc_rdma_copy_inline_range(rqstp, head, start, length);
 }
 

@@ -1187,8 +1196,12 @@ svc_rdma_read_multiple_chunks(struct svc_rqst *rqstp,
 static int svc_rdma_read_data_item(struct svc_rqst *rqstp,
 				   struct svc_rdma_recv_ctxt *head)
 {
-	return svc_rdma_build_read_chunk(rqstp, head,
-					 pcl_first_chunk(&head->rc_read_pcl));
+	struct svc_rdma_chunk *chunk = pcl_first_chunk(&head->rc_read_pcl);
+
+	if (chunk->ch_position > head->rc_saved_arg.head[0].iov_len)
+		return -EINVAL;
+
+	return svc_rdma_build_read_chunk(rqstp, head, chunk);
 }
 
 /**

@@ -1257,14 +1270,17 @@ static int svc_rdma_read_call_chunk(struct svc_rqst *rqstp,
 			pcl_first_chunk(&head->rc_call_pcl);
 	const struct svc_rdma_pcl *pcl = &head->rc_read_pcl;
 	struct svc_rdma_chunk *chunk, *next;
-	unsigned int start, length;
+	unsigned int call_len, start, length;
 	int ret;
 
 	if (pcl_is_empty(pcl))
 		return svc_rdma_build_read_chunk(rqstp, head, call_chunk);
 
+	call_len = call_chunk->ch_length;
 	start = 0;
 	chunk = pcl_first_chunk(pcl);
+	if (chunk->ch_position > call_len)
+		return -EINVAL;
 	length = chunk->ch_position;
 	ret = svc_rdma_read_chunk_range(rqstp, head, call_chunk,
 					start, length);

@@ -1281,6 +1297,10 @@ static int svc_rdma_read_call_chunk(struct svc_rqst *rqstp,
 			break;
 
 		start += length;
+		if (next->ch_position > call_len)
+			return -EINVAL;
+		if (head->rc_readbytes > next->ch_position)
+			return -EINVAL;
 		length = next->ch_position - head->rc_readbytes;
 		ret = svc_rdma_read_chunk_range(rqstp, head, call_chunk,
 						start, length);

@@ -1289,7 +1309,9 @@ static int svc_rdma_read_call_chunk(struct svc_rqst *rqstp,
 	}
 
 	start += length;
-	length = call_chunk->ch_length - start;
+	if (start > call_len)
+		return -EINVAL;
+	length = call_len - start;
 	return svc_rdma_read_chunk_range(rqstp, head, call_chunk,
 					 start, length);
 }

2.54.0

From: Chris Mason <[email protected]>

svc_rdma_read_chunk_range() walks a Read chunk's segment list to build a sub-range starting at byte offset and spanning length bytes for a Position-Zero or Call chunk. Two arithmetic defects in the per-segment loop produce wrong DMA lengths and a u32 underflow:

    pcl_for_each_segment(segment, chunk) {
            if (offset > segment->rs_length) {
                    offset -= segment->rs_length;
                    continue;
            }

            dummy.rs_handle = segment->rs_handle;
            dummy.rs_length = min_t(u32, length,
                                    segment->rs_length) - offset;
            dummy.rs_offset = segment->rs_offset + offset;

First, the skip predicate uses '>' instead of '>='. When offset equals the segment's full rs_length, the segment is fully consumed and should be skipped, but the loop falls through into the body. The resulting dummy.rs_length is min_t(u32, length, rs_length) - rs_length, which underflows to a near-UINT_MAX u32 when length is smaller than rs_length, or is zero otherwise.

Second, the length formula subtracts offset from the min_t() result rather than from segment->rs_length before the cap. For offset > 0 the segment's residual is rs_length - offset, not rs_length, so the cap must be applied to the residual. With the current bracketing, whenever length is smaller than rs_length - offset the per-segment length becomes length - offset instead of length, silently dropping offset bytes from the rebuilt chunk. Combined with the boundary case above it also enables the u32 underflow path, which propagates a huge nr_bvec into svc_rdma_build_read_segment() and a multi-MiB kmalloc_array_node() in svc_rdma_get_rw_ctxt().

Additionally, svc_rdma_read_call_chunk() can invoke this function with length == 0 when the last Read chunk ends exactly at the end of the Call chunk. With the corrected >= predicate, every segment is skipped and the function returns the initial -EINVAL, rejecting a valid request. Return success immediately when length is zero. Also break out of the loop once length is fully consumed to avoid passing zero-length segments to svc_rdma_build_read_segment().

Fix by using '>=' so a fully-consumed segment is skipped, by moving '- offset' inside min_t() so the cap is applied to the segment's residual length, by returning success for zero-length requests, and by stopping iteration when the requested range has been consumed.

---
 net/sunrpc/xprtrdma/svc_rdma_rw.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

index eb4bc56ed387..182bd577e0b7 100644
@@ -1227,17 +1227,20 @@ static int svc_rdma_read_chunk_range(struct svc_rqst *rqstp,
 	const struct svc_rdma_segment *segment;
 	int ret;
 
+	if (!length)
+		return 0;
+
 	ret = -EINVAL;
 	pcl_for_each_segment(segment, chunk) {
 		struct svc_rdma_segment dummy;
 
-		if (offset > segment->rs_length) {
+		if (offset >= segment->rs_length) {
 			offset -= segment->rs_length;
 			continue;
 		}
 
 		dummy.rs_handle = segment->rs_handle;
-		dummy.rs_length = min_t(u32, length, segment->rs_length) - offset;
+		dummy.rs_length = min_t(u32, length, segment->rs_length - offset);
 		dummy.rs_offset = segment->rs_offset + offset;
 
 		ret = svc_rdma_build_read_segment(rqstp, head, &dummy);

@@ -1246,6 +1249,8 @@ static int svc_rdma_read_chunk_range(struct svc_rqst *rqstp,
 
 		head->rc_readbytes += dummy.rs_length;
 		length -= dummy.rs_length;
+		if (!length)
+			break;
 		offset = 0;
 	}
 	return ret;

2.54.0

From: Chuck Lever <[email protected]>

The RPC/RDMA Read list decoder stores wire-supplied segment lengths without validation. xdr_count_read_segments() checks 4-byte alignment for non-zero position values but does not cap the segment length.

An oversized rs_length reaches svc_rdma_build_read_segment(), which derives nr_bvec from it and can drive a large dynamic bvec allocation before verifying that enough rq_pages remain. If the post-allocation page-overrun guard fires, the freshly acquired rw context is not returned, leaking the resource.

Reject any segment whose length exceeds the receive context's page budget during Read list decoding, consistent with how xdr_check_write_chunk() bounds Write segment counts against rc_maxpages. Also return the rw context on the existing post-allocation overrun path in svc_rdma_build_read_segment(), keeping that defensive guard balanced.

---
 net/sunrpc/xprtrdma/svc_rdma_recvfrom.c | 2 ++
 net/sunrpc/xprtrdma/svc_rdma_rw.c       | 5 +++--
 2 files changed, 5 insertions(+), 2 deletions(-)

index fe9bf0371b6e..15c1d8ae5301 100644
@@ -458,6 +458,8 @@ static bool xdr_count_read_segments(struct svc_rdma_recv_ctxt *rctxt, __be32 *p)
 
 		xdr_decode_read_segment(p, &position, &handle,
 					    &length, &offset);
+		if (length > rctxt->rc_maxpages << PAGE_SHIFT)
+			return false;
 		if (position) {
 			if (position & 3)
 				return false;

index 182bd577e0b7..587e4cd29303 100644
@@ -1013,7 +1013,7 @@ static int svc_rdma_build_read_segment(struct svc_rqst *rqstp,
 		len -= seg_len;
 
 		if (len && ((head->rc_curpage + 1) > rqstp->rq_maxpages))
-			goto out_overrun;
+			goto out_put;
 	}
 
 	ret = svc_rdma_rw_ctx_init(rdma, ctxt, segment->rs_offset,

@@ -1027,7 +1027,8 @@ static int svc_rdma_build_read_segment(struct svc_rqst *rqstp,
 	cc->cc_sqecount += ret;
 	return 0;
 
-out_overrun:
+out_put:
+	svc_rdma_put_rw_ctxt(rdma, ctxt);
 	trace_svcrdma_page_overrun_err(&cc->cc_cid, head->rc_curpage);
 	return -EINVAL;
 }

2.54.0

From: Chris Mason <[email protected]>

When a parsed chunk list contains a chunk whose ch_segcount is zero, pcl_for_each_segment computes its inclusive upper bound as &chunk->ch_segments[ch_segcount - 1]. ch_segcount is u32, so the subtraction wraps to 0xFFFFFFFF and the bound lands far past the ch_segments flex array. The loop body then walks unrelated memory at sizeof(struct svc_rdma_segment) stride until it faults.

A zero-segcount chunk is reachable from the wire:
xdr_check_write_chunk() only rejects segcount values greater than rc_maxpages, and pcl_alloc_write() links a freshly allocated chunk onto rc_write_pcl/rc_reply_pcl before its segment-fill loop runs, so a Write or Reply chunk advertising zero segments leaves ch_segcount == 0 on the list. When the transport has negotiated Send-With-Invalidate, svc_rdma_get_inv_rkey() iterates all four PCLs with pcl_for_each_segment and dereferences segment->rs_handle on each iteration, turning the underflow into an out-of-bounds read and a general protection fault.

    xdr_check_write_list / xdr_check_reply_chunk
      pcl_alloc_write()
        chunk = pcl_alloc_chunk(...)  /* ch_segcount = 0 */
        list_add_tail(&chunk->ch_list, &pcl->cl_chunks)
        /* fill loop iterates zero times for wire segcount 0 */

    svc_rdma_get_inv_rkey()
      pcl_for_each_chunk(rc_write_pcl)
        pcl_for_each_segment(segment, chunk)
          pos <= &ch_segments[0u - 1u]  /* 0xFFFFFFFF */
          segment->rs_handle            /* OOB read -> GPF */

Fix by switching the macro to a half-open upper bound that uses ch_segcount directly. For ch_segcount == 0 the loop start equals the loop end and the body is skipped; for ch_segcount > 0 the iteration range is unchanged. All six existing call sites in net/sunrpc/xprtrdma/svc_rdma_recvfrom.c and net/sunrpc/xprtrdma/svc_rdma_rw.c remain correct under the new bound, so no caller changes are needed.

---
 include/linux/sunrpc/svc_rdma_pcl.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

index 7516ad0fae80..655681cf8fed 100644
@@ -97,7 +97,7 @@ pcl_next_chunk(const struct svc_rdma_pcl *pcl, struct svc_rdma_chunk *chunk)
  */
 #define pcl_for_each_segment(pos, chunk) \
 	for (pos = &(chunk)->ch_segments[0]; \
-	     pos <= &(chunk)->ch_segments[(chunk)->ch_segcount - 1]; \
+	     pos < &(chunk)->ch_segments[(chunk)->ch_segcount]; \
 	     pos++)
 
 /**

2.54.0

From: Chris Mason <[email protected]>

A peer can send a Write or Reply chunk whose segcount field is zero. xdr_check_write_chunk() only rejects segcount > rc_maxpages, so zero passes the range check, and xdr_inline_decode(stream, 0) returns the current (non-NULL) cursor without advancing. The function returns true and pcl_alloc_write() then links a struct svc_rdma_chunk with ch_segcount == 0 onto rc_write_pcl or rc_reply_pcl.

An earlier patch in this series made pcl_for_each_segment() safe for ch_segcount == 0, so this no longer drives the memory walk it used to. Rejecting the malformed frame at the decode boundary is still worthwhile as defense in depth: it keeps degenerate zero-segment chunks off the parsed chunk lists entirely, so any future consumer that walks ch_segments directly cannot observe one, and it makes the zero-floor easy to backport to trees where the macro change is more intrusive. RFC 8166 has no meaning for a Write/Reply chunk that describes no remote buffer, so no legitimate client is affected.

xdr_check_reply_chunk() funnels Reply chunks through xdr_check_write_chunk() and inherits the same rejection.

pcl_alloc_write() also links each chunk onto the parsed chunk list before filling its segment array. If a future change weakens the segcount-0 rejection, an incomplete chunk is visible to consumers during the fill loop. Reorder so that list_add_tail() follows the segment fill loop, ensuring only fully-populated chunks appear on the list.

---
 net/sunrpc/xprtrdma/svc_rdma_pcl.c      | 2 +-
 net/sunrpc/xprtrdma/svc_rdma_recvfrom.c | 9 ++++++---
 2 files changed, 7 insertions(+), 4 deletions(-)

index 1f8f7dad8b6f..18d1045799ce 100644
@@ -213,7 +213,6 @@ bool pcl_alloc_write(struct svc_rdma_recv_ctxt *rctxt,
 		chunk = pcl_alloc_chunk(segcount, 0);
 		if (!chunk)
 			return false;
-		list_add_tail(&chunk->ch_list, &pcl->cl_chunks);
 
 		for (j = 0; j < segcount; j++) {
 			segment = &chunk->ch_segments[j];

@@ -225,6 +224,7 @@ bool pcl_alloc_write(struct svc_rdma_recv_ctxt *rctxt,
 			chunk->ch_length += segment->rs_length;
 			chunk->ch_segcount++;
 		}
+		list_add_tail(&chunk->ch_list, &pcl->cl_chunks);
 	}
 	return true;
 }

index 15c1d8ae5301..f6a7533a7555 100644
@@ -510,10 +510,13 @@ static bool xdr_check_write_chunk(struct svc_rdma_recv_ctxt *rctxt)
 		return false;
 
 	/* Before trusting the segcount value enough to use it in
-	 * a computation, perform a simple range check. This is an
-	 * arbitrary but sensible limit (ie, not architectural).
+	 * a computation, perform a simple range check. A zero
+	 * segcount describes no remote buffer and is rejected so
+	 * downstream consumers never see a degenerate ch_segcount==0
+	 * chunk. The upper bound is an arbitrary but sensible limit
+	 * (ie, not architectural).
 	 */
-	if (unlikely(segcount > rctxt->rc_maxpages))
+	if (segcount == 0 || unlikely(segcount > rctxt->rc_maxpages))
 		return false;
 
 	p = xdr_inline_decode(&rctxt->rc_stream,

2.54.0

From: Chuck Lever <[email protected]>

Read chunk position and length validation is currently scattered across three consumer functions: svc_rdma_read_data_item(), svc_rdma_read_multiple_chunks(), and svc_rdma_read_call_chunk(). Each independently guards against the same class of unsigned arithmetic underflow from untrusted wire values. Any new consumer of the parsed Read chunk list must replicate these checks or risk re-introducing the defects fixed by earlier patches in this series.

Add pcl_check_read_chunk_positions() to consolidate position and length validation into a single post-decode pass, called from svc_rdma_xdr_decode_req() after all three chunk lists have been parsed and the inline body length is known. The pass verifies three properties:

- Each Read chunk's inline-body offset (its unreduced-stream position minus the cumulative length of preceding Read chunks) falls within the inline body length, or within the Call chunk length for interleaved reads.

- Adjacent Read chunk positions do not overlap: cumulative read bytes at each transition do not exceed the next position.

- Each chunk length does not exceed the receive context's page budget.

Malformed frames are rejected before reaching any consumer. The existing consumer-side guards remain as defense in depth.

---
 include/linux/sunrpc/svc_rdma_pcl.h     |  2 ++
 net/sunrpc/xprtrdma/svc_rdma_pcl.c      | 61 +++++++++++++++++++++++++++++++--
 net/sunrpc/xprtrdma/svc_rdma_recvfrom.c |  3 ++
 3 files changed, 63 insertions(+), 3 deletions(-)

index 655681cf8fed..6346d8cf2587 100644
@@ -119,6 +119,8 @@ extern bool pcl_alloc_call(struct svc_rdma_recv_ctxt *rctxt, __be32 *p);
 extern bool pcl_alloc_read(struct svc_rdma_recv_ctxt *rctxt, __be32 *p);
 extern bool pcl_alloc_write(struct svc_rdma_recv_ctxt *rctxt,
 			    struct svc_rdma_pcl *pcl, __be32 *p);
+extern bool pcl_check_read_chunk_positions(struct svc_rdma_recv_ctxt *rctxt,
+					   unsigned int inline_len);
 extern int pcl_process_nonpayloads(const struct svc_rdma_pcl *pcl,
 				   const struct xdr_buf *xdr,
 				   int (*actor)(const struct xdr_buf *,

index f6a7533a7555..d64b5f78ce8a 100644
@@ -724,6 +724,9 @@ static int svc_rdma_xdr_decode_req(struct xdr_buf *rq_arg,
 
 	rq_arg->head[0].iov_base = rctxt->rc_stream.p;
 	hdr_len = xdr_stream_pos(&rctxt->rc_stream);
+	if (!pcl_check_read_chunk_positions(rctxt,
+					    rq_arg->head[0].iov_len - hdr_len))
+		goto out_inval;
 	rq_arg->head[0].iov_len -= hdr_len;
 	rq_arg->len -= hdr_len;
 	trace_svcrdma_decode_rqst(rctxt, rdma_argp, hdr_len);

2.54.0