Skip to contentSkip to conversation

rds: zero per-item info buffer before handing it to visitors

PATCH Superseded Michael Bommarito 4 participants 5 messages Apr 17 – Apr 23

Subsystems (5)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 rds: zero per-item info buffer before handing it to visitors

Yet another from my "clanker." This only applies to people who don't use CONFIG_INIT_STACK_ALL_ZERO, but I presume that's still enough people that it's worth backporting since it can be chained through leaked addresses to defeat KASLR.

rds_for_each_conn_info() and rds_walk_conn_path_info() both hand a caller-allocated on-stack u64 buffer to a per-connection visitor and then copy the full item_len bytes back to user space via rds_info_copy() regardless of how much of the buffer the visitor actually wrote.

rds_ib_conn_info_visitor() and rds6_ib_conn_info_visitor() only write a subset of their output struct when the underlying rds_connection is not in state RDS_CONN_UP (src/dst addr, tos, sl and the two GIDs via explicit memsets). Several u32 fields (max_send_wr, max_recv_wr, max_send_sge, rdma_mr_max, rdma_mr_size, cache_allocs) and the 2-byte alignment hole between sl and cache_allocs remain as whatever stack contents preceded the visitor call and are then memcpy_to_user()'d out to user space.

struct rds_info_rdma_connection and struct rds6_info_rdma_connection are the only rds_info_* structs in include/uapi/linux/rds.h that are not marked __attribute__((packed)), so they have a real alignment hole. The other info visitors (rds_conn_info_visitor, rds6_conn_info_visitor, rds_tcp_tc_info, ...) write all fields of their packed output struct today and are not known to be vulnerable, but a future visitor that adds a conditional write-path would have the same bug.

Reproduction on a kernel built without CONFIG_INIT_STACK_ALL_ZERO=y:
a local unprivileged user opens AF_RDS, sets SO_RDS_TRANSPORT=IB, binds to a local address on an RDMA-capable netdev (rxe soft-RoCE on any netdev is sufficient), sendto()'s any peer on the same subnet (fails cleanly but installs an rds_connection in the global hash in RDS_CONN_CONNECTING), then calls getsockopt(SOL_RDS, RDS_INFO_IB_CONNECTIONS). The returned 68-byte item contains 26 bytes of stack garbage including kernel text/data pointers:

    0..7   0a 63 00 01 0a 63 00 02     src=10.99.0.1 dst=10.99.0.2
    8..39  00 ...                      gids (memset-zeroed)
    40..47 e0 92 a3 81 ff ff ff ff     kernel pointer (max_send_wr)
    48..55 7f 37 b5 81 ff ff ff ff     kernel pointer (rdma_mr_max)
    56..59 01 00 08 00                 rdma_mr_size (garbage)
    60..61 00 00                       tos, sl
    62..63 00 00                       alignment padding
    64..67 18 00 00 00                 cache_allocs (garbage)

Fix by zeroing the per-item buffer in both rds_for_each_conn_info() and rds_walk_conn_path_info() before invoking the visitor. This covers the IPv4/IPv6 IB visitors and hardens all current and future visitors against the same class of bug.

No functional change for visitors that fully populate their output.

Fixes ec16227e1414 ("RDS/IB: Infiniband transport")
Assisted-by Claude:claude-opus-4-7
---
 net/rds/connection.c | 14 ++++++++++++++
 1 file changed, 14 insertions(+)
net/rds/connection.c +14 -0
index 412441aaa298..c10b7ed06c49 100644
@@ -701,6 +701,13 @@ void rds_for_each_conn_info(struct socket *sock, unsigned int len,
 	     i++, head++) {
 		hlist_for_each_entry_rcu(conn, head, c_hash_node) {
 
+			/* Zero the per-item buffer before handing it to the
+			 * visitor so any field the visitor does not write -
+			 * including implicit alignment padding - cannot leak
+			 * stack contents to user space via rds_info_copy().
+			 */
+			memset(buffer, 0, item_len);
+
 			/* XXX no c_lock usage.. */
 			if (!visitor(conn, buffer))
 				continue;
@@ -750,6 +757,13 @@ static void rds_walk_conn_path_info(struct socket *sock, unsigned int len,
 			 */
 			cp = conn->c_path;
 
+			/* Zero the per-item buffer for the same reason as
+			 * rds_for_each_conn_info(): any byte the visitor
+			 * does not write (including alignment padding) must
+			 * not leak stack contents via rds_info_copy().
+			 */
+			memset(buffer, 0, item_len);
+
 			/* XXX no cp_lock usage.. */
 			if (!visitor(cp, buffer))
 				continue;
Signature · 1 line
2.53.0
#1 lore raw
Quoted 47 lines · Michael Bommarito

Yet another from my "clanker." This only applies to people who don't use CONFIG_INIT_STACK_ALL_ZERO, but I presume that's still enough people that it's worth backporting since it can be chained through leaked addresses to defeat KASLR.

rds_for_each_conn_info() and rds_walk_conn_path_info() both hand a caller-allocated on-stack u64 buffer to a per-connection visitor and then copy the full item_len bytes back to user space via rds_info_copy() regardless of how much of the buffer the visitor actually wrote.

rds_ib_conn_info_visitor() and rds6_ib_conn_info_visitor() only write a subset of their output struct when the underlying rds_connection is not in state RDS_CONN_UP (src/dst addr, tos, sl and the two GIDs via explicit memsets). Several u32 fields (max_send_wr, max_recv_wr, max_send_sge, rdma_mr_max, rdma_mr_size, cache_allocs) and the 2-byte alignment hole between sl and cache_allocs remain as whatever stack contents preceded the visitor call and are then memcpy_to_user()'d out to user space.

struct rds_info_rdma_connection and struct rds6_info_rdma_connection are the only rds_info_* structs in include/uapi/linux/rds.h that are not marked __attribute__((packed)), so they have a real alignment hole. The other info visitors (rds_conn_info_visitor, rds6_conn_info_visitor, rds_tcp_tc_info, ...) write all fields of their packed output struct today and are not known to be vulnerable, but a future visitor that adds a conditional write-path would have the same bug.

Reproduction on a kernel built without CONFIG_INIT_STACK_ALL_ZERO=y:
a local unprivileged user opens AF_RDS, sets SO_RDS_TRANSPORT=IB, binds to a local address on an RDMA-capable netdev (rxe soft-RoCE on any netdev is sufficient), sendto()'s any peer on the same subnet (fails cleanly but installs an rds_connection in the global hash in RDS_CONN_CONNECTING), then calls getsockopt(SOL_RDS, RDS_INFO_IB_CONNECTIONS). The returned 68-byte item contains 26 bytes of stack garbage including kernel text/data pointers:

    0..7   0a 63 00 01 0a 63 00 02     src=10.99.0.1 dst=10.99.0.2
    8..39  00 ...                      gids (memset-zeroed)
    40..47 e0 92 a3 81 ff ff ff ff     kernel pointer (max_send_wr)
    48..55 7f 37 b5 81 ff ff ff ff     kernel pointer (rdma_mr_max)
    56..59 01 00 08 00                 rdma_mr_size (garbage)
    60..61 00 00                       tos, sl
    62..63 00 00                       alignment padding
    64..67 18 00 00 00                 cache_allocs (garbage)
Quoted 8 lines

Fix by zeroing the per-item buffer in both rds_for_each_conn_info() and rds_walk_conn_path_info() before invoking the visitor. This covers the IPv4/IPv6 IB visitors and hardens all current and future visitors against the same class of bug.

No functional change for visitors that fully populate their output.

Fixes ec16227e1414 ("RDS/IB: Infiniband transport")

LGTM. Reviewed-by: Sharath Srinivasan <[email protected]>

Thanks,
Sharath

Quoted 38 lines
Assisted-by Claude:claude-opus-4-7
---
 net/rds/connection.c | 14 ++++++++++++++
 1 file changed, 14 insertions(+)
net/rds/connection.c +14 -0
index 412441aaa298..c10b7ed06c49 100644
@@ -701,6 +701,13 @@ void rds_for_each_conn_info(struct socket *sock, unsigned int len,
 	     i++, head++) {
 		hlist_for_each_entry_rcu(conn, head, c_hash_node) {
 
+			/* Zero the per-item buffer before handing it to the
+			 * visitor so any field the visitor does not write -
+			 * including implicit alignment padding - cannot leak
+			 * stack contents to user space via rds_info_copy().
+			 */
+			memset(buffer, 0, item_len);
+
 			/* XXX no c_lock usage.. */
 			if (!visitor(conn, buffer))
 				continue;
@@ -750,6 +757,13 @@ static void rds_walk_conn_path_info(struct socket *sock, unsigned int len,
 			 */
 			cp = conn->c_path;
 
+			/* Zero the per-item buffer for the same reason as
+			 * rds_for_each_conn_info(): any byte the visitor
+			 * does not write (including alignment padding) must
+			 * not leak stack contents via rds_info_copy().
+			 */
+			memset(buffer, 0, item_len);
+
 			/* XXX no cp_lock usage.. */
 			if (!visitor(cp, buffer))
 				continue;
maintainer Reviewed-by on v1 1/1
#2 lore raw
In reply to Michael Bommarito
Quoted 52 lines · Michael Bommarito

Yet another from my "clanker." This only applies to people who don't use CONFIG_INIT_STACK_ALL_ZERO, but I presume that's still enough people that it's worth backporting since it can be chained through leaked addresses to defeat KASLR.

rds_for_each_conn_info() and rds_walk_conn_path_info() both hand a caller-allocated on-stack u64 buffer to a per-connection visitor and then copy the full item_len bytes back to user space via rds_info_copy() regardless of how much of the buffer the visitor actually wrote.

rds_ib_conn_info_visitor() and rds6_ib_conn_info_visitor() only write a subset of their output struct when the underlying rds_connection is not in state RDS_CONN_UP (src/dst addr, tos, sl and the two GIDs via explicit memsets). Several u32 fields (max_send_wr, max_recv_wr, max_send_sge, rdma_mr_max, rdma_mr_size, cache_allocs) and the 2-byte alignment hole between sl and cache_allocs remain as whatever stack contents preceded the visitor call and are then memcpy_to_user()'d out to user space.

struct rds_info_rdma_connection and struct rds6_info_rdma_connection are the only rds_info_* structs in include/uapi/linux/rds.h that are not marked __attribute__((packed)), so they have a real alignment hole. The other info visitors (rds_conn_info_visitor, rds6_conn_info_visitor, rds_tcp_tc_info, ...) write all fields of their packed output struct today and are not known to be vulnerable, but a future visitor that adds a conditional write-path would have the same bug.

Reproduction on a kernel built without CONFIG_INIT_STACK_ALL_ZERO=y:
a local unprivileged user opens AF_RDS, sets SO_RDS_TRANSPORT=IB, binds to a local address on an RDMA-capable netdev (rxe soft-RoCE on any netdev is sufficient), sendto()'s any peer on the same subnet (fails cleanly but installs an rds_connection in the global hash in RDS_CONN_CONNECTING), then calls getsockopt(SOL_RDS, RDS_INFO_IB_CONNECTIONS). The returned 68-byte item contains 26 bytes of stack garbage including kernel text/data pointers:

    0..7   0a 63 00 01 0a 63 00 02     src=10.99.0.1 dst=10.99.0.2
    8..39  00 ...                      gids (memset-zeroed)
    40..47 e0 92 a3 81 ff ff ff ff     kernel pointer (max_send_wr)
    48..55 7f 37 b5 81 ff ff ff ff     kernel pointer (rdma_mr_max)
    56..59 01 00 08 00                 rdma_mr_size (garbage)
    60..61 00 00                       tos, sl
    62..63 00 00                       alignment padding
    64..67 18 00 00 00                 cache_allocs (garbage)

Fix by zeroing the per-item buffer in both rds_for_each_conn_info() and rds_walk_conn_path_info() before invoking the visitor. This covers the IPv4/IPv6 IB visitors and hardens all current and future visitors against the same class of bug.

Quoted 5 lines

No functional change for visitors that fully populate their output.

Fixes ec16227e1414 ("RDS/IB: Infiniband transport")
Assisted-by Claude:claude-opus-4-7

Hi Micheal,

The change looks fine to me. Since this is a bug fix, you'll want to cc stable tree [email protected], and note the target tree and component in the subject line like this:

[PATCH net v2] net/rds: zero per-item info buffer before handing it to visitors

Other than that, the patch looks good to me. Thanks Micheal.

Allison

Quoted 36 lines
---
 net/rds/connection.c | 14 ++++++++++++++
 1 file changed, 14 insertions(+)
net/rds/connection.c +14 -0
index 412441aaa298..c10b7ed06c49 100644
@@ -701,6 +701,13 @@ void rds_for_each_conn_info(struct socket *sock, unsigned int len,
 	     i++, head++) {
 		hlist_for_each_entry_rcu(conn, head, c_hash_node) {
 
+			/* Zero the per-item buffer before handing it to the
+			 * visitor so any field the visitor does not write -
+			 * including implicit alignment padding - cannot leak
+			 * stack contents to user space via rds_info_copy().
+			 */
+			memset(buffer, 0, item_len);
+
 			/* XXX no c_lock usage.. */
 			if (!visitor(conn, buffer))
 				continue;
@@ -750,6 +757,13 @@ static void rds_walk_conn_path_info(struct socket *sock, unsigned int len,
 			 */
 			cp = conn->c_path;
 
+			/* Zero the per-item buffer for the same reason as
+			 * rds_for_each_conn_info(): any byte the visitor
+			 * does not write (including alignment padding) must
+			 * not leak stack contents via rds_info_copy().
+			 */
+			memset(buffer, 0, item_len);
+
 			/* XXX no cp_lock usage.. */
 			if (!visitor(cp, buffer))
 				continue;
#3 lore raw

PATCH v2 1/1 net/rds: zero per-item info buffer before handing it to visitors

In reply to Michael Bommarito

rds_for_each_conn_info() and rds_walk_conn_path_info() both hand a caller-allocated on-stack u64 buffer to a per-connection visitor and then copy the full item_len bytes back to user space via rds_info_copy() regardless of how much of the buffer the visitor actually wrote.

rds_ib_conn_info_visitor() and rds6_ib_conn_info_visitor() only write a subset of their output struct when the underlying rds_connection is not in state RDS_CONN_UP (src/dst addr, tos, sl and the two GIDs via explicit memsets). Several u32 fields (max_send_wr, max_recv_wr, max_send_sge, rdma_mr_max, rdma_mr_size, cache_allocs) and the 2-byte alignment hole between sl and cache_allocs remain as whatever stack contents preceded the visitor call and are then memcpy_to_user()'d out to user space.

struct rds_info_rdma_connection and struct rds6_info_rdma_connection are the only rds_info_* structs in include/uapi/linux/rds.h that are not marked __attribute__((packed)), so they have a real alignment hole. The other info visitors (rds_conn_info_visitor, rds6_conn_info_visitor, rds_tcp_tc_info, ...) write all fields of their packed output struct today and are not known to be vulnerable, but a future visitor that adds a conditional write-path would have the same bug.

Reproduction on a kernel built without CONFIG_INIT_STACK_ALL_ZERO=y:
a local unprivileged user opens AF_RDS, sets SO_RDS_TRANSPORT=IB, binds to a local address on an RDMA-capable netdev (rxe soft-RoCE on any netdev is sufficient), sendto()'s any peer on the same subnet (fails cleanly but installs an rds_connection in the global hash in RDS_CONN_CONNECTING), then calls getsockopt(SOL_RDS, RDS_INFO_IB_CONNECTIONS). The returned 68-byte item contains 26 bytes of stack garbage including kernel text/data pointers:

    0..7   0a 63 00 01 0a 63 00 02     src=10.99.0.1 dst=10.99.0.2
    8..39  00 ...                      gids (memset-zeroed)
    40..47 e0 92 a3 81 ff ff ff ff     kernel pointer (max_send_wr)
    48..55 7f 37 b5 81 ff ff ff ff     kernel pointer (rdma_mr_max)
    56..59 01 00 08 00                 rdma_mr_size (garbage)
    60..61 00 00                       tos, sl
    62..63 00 00                       alignment padding
    64..67 18 00 00 00                 cache_allocs (garbage)

Fix by zeroing the per-item buffer in both rds_for_each_conn_info() and rds_walk_conn_path_info() before invoking the visitor. This covers the IPv4/IPv6 IB visitors and hardens all current and future visitors against the same class of bug.

No functional change for visitors that fully populate their output.

Changes in v2:
- retarget at the net tree (subject prefix "[PATCH net v2]", net/rds: prefix in the title)
- add Cc: [email protected]
- pick up Reviewed-by tags from Sharath Srinivasan and Allison Henderson

Fixes ec16227e1414 ("RDS/IB: Infiniband transport")
Assisted-by Claude:claude-opus-4-7
---
 net/rds/connection.c | 14 ++++++++++++++
 1 file changed, 14 insertions(+)
net/rds/connection.c +14 -0
index 412441aaa298..c10b7ed06c49 100644
@@ -701,6 +701,13 @@ void rds_for_each_conn_info(struct socket *sock, unsigned int len,
 	     i++, head++) {
 		hlist_for_each_entry_rcu(conn, head, c_hash_node) {
 
+			/* Zero the per-item buffer before handing it to the
+			 * visitor so any field the visitor does not write -
+			 * including implicit alignment padding - cannot leak
+			 * stack contents to user space via rds_info_copy().
+			 */
+			memset(buffer, 0, item_len);
+
 			/* XXX no c_lock usage.. */
 			if (!visitor(conn, buffer))
 				continue;
@@ -750,6 +757,13 @@ static void rds_walk_conn_path_info(struct socket *sock, unsigned int len,
 			 */
 			cp = conn->c_path;
 
+			/* Zero the per-item buffer for the same reason as
+			 * rds_for_each_conn_info(): any byte the visitor
+			 * does not write (including alignment padding) must
+			 * not leak stack contents via rds_info_copy().
+			 */
+			memset(buffer, 0, item_len);
+
 			/* XXX no cp_lock usage.. */
 			if (!visitor(cp, buffer))
 				continue;
Signature · 1 line
2.53.0
#4 lore raw

Re: [PATCH net v2] net/rds: zero per-item info buffer before handing it to visitors

This patch was applied to netdev/net.git (main) by Jakub Kicinski <[email protected]>:

Hello:

This patch was applied to netdev/net.git (main)
by Jakub Kicinski <[email protected]>:

Quoted 15 lines · On Sat, 18 Apr 2026 10:10:47 -0400 you wrote:

rds_for_each_conn_info() and rds_walk_conn_path_info() both hand a caller-allocated on-stack u64 buffer to a per-connection visitor and then copy the full item_len bytes back to user space via rds_info_copy() regardless of how much of the buffer the visitor actually wrote.

rds_ib_conn_info_visitor() and rds6_ib_conn_info_visitor() only write a subset of their output struct when the underlying rds_connection is not in state RDS_CONN_UP (src/dst addr, tos, sl and the two GIDs via explicit memsets). Several u32 fields (max_send_wr, max_recv_wr, max_send_sge, rdma_mr_max, rdma_mr_size, cache_allocs) and the 2-byte alignment hole between sl and cache_allocs remain as whatever stack contents preceded the visitor call and are then memcpy_to_user()'d out to user space.

Quoted 1 line

[...]

Here is the summary with links:
- [net,v2] net/rds: zero per-item info buffer before handing it to visitors

    https://git.kernel.org/netdev/net/c/c88eb7e8d839

You are awesome, thank you!

Signature · 2 lines
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html