Skip to contentSkip to conversation

RDMA: detect and handle CoCo DMA bounce buffering

PATCH Applied Jiri Pirko 3 participants 6 messages May 17 – May 29 InfiniBand

lore
Share & tools
lore

Jiri Pirko

author
#0 lore raw

v3 cover RDMA: detect and handle CoCo DMA bounce buffering

From: Jiri Pirko <[email protected]>

In Confidential Computing (CoCo) guests, the DMA mapping layer redirects all device DMA through swiotlb bounce buffers to keep guest memory encrypted. This is transparent for regular devices because the CPU copies data between the bounce buffer and the real buffer on every DMA map/unmap cycle.

RDMA breaks this model. Once a memory region is registered, the device accesses the underlying pages directly for an extended period without CPU involvement. The swiotlb layer never gets a chance to synchronize, so the device operates on bounce buffer memory while the application works with its own pages - the two never see each other's updates.

This series adds detection and handling of this condition. A new IB_UVERBS_DEVICE_CC_DMA_BOUNCE flag is exposed in device_cap_flags_ex so userspace libraries can detect the situation and switch to dmabuf-based memory registration using "system_cc_shared" heap where available. Plain ib_umem_get() is made to fail early with -EOPNOTSUPP to prevent silent misfunction.

---
See individual patches for changelog.

v2: https://lore.kernel.org/all/[email protected]/
v1: https://lore.kernel.org/all/[email protected]/

based on top of:
https://lore.kernel.org/all/[email protected]/

Jiri Pirko (2):
RDMA/uverbs: expose CoCo DMA bounce requirement to userspace RDMA/umem: block plain userspace memory registration under CoCo bounce

 drivers/infiniband/core/device.c     | 9 +++++++++
 drivers/infiniband/core/umem.c       | 3 +++
 drivers/infiniband/core/uverbs_cmd.c | 2 ++
 include/rdma/ib_verbs.h              | 3 +++
 include/uapi/rdma/ib_user_verbs.h    | 2 ++
 5 files changed, 19 insertions(+)
Signature · 1 line
2.54.0
author
#1 lore raw

PATCH v3 1/2 RDMA/uverbs: expose CoCo DMA bounce requirement to userspace

From: Jiri Pirko <[email protected]>

In CoCo guests, guest memory is encrypted and untrusted (T=0) devices cannot DMA to it directly; such transfers must go through unencrypted bounce buffers. RDMA registers user pages for direct device access, bypassing the DMA layer and thus any bouncing, so registered memory does not work in this configuration.

Until trusted (T=1) device detection is available, conservatively flag every device attached to a CoCo guest. Expose the condition to userspace as IB_UVERBS_DEVICE_CC_DMA_BOUNCE in device_cap_flags_ex so applications can avoid memory registration and fall back to copying buffers through send/recv.

---
v2->v3:
- Dropped is_swiotlb_force_bounce()/swiotlb.h; keyed off CC_ATTR_GUEST_MEM_ENCRYPT alone.
- Added comment noting T=1 detection should narrow the check.
- Rewrote log: dropped SWIOTLB rationale, explained T=0 assumption; fixed device_cap_flags_ex typo.

---
 drivers/infiniband/core/device.c     | 9 +++++++++
 drivers/infiniband/core/uverbs_cmd.c | 2 ++
 include/rdma/ib_verbs.h              | 3 +++
 include/uapi/rdma/ib_user_verbs.h    | 2 ++
 4 files changed, 16 insertions(+)
drivers/infiniband/core/device.c +9 -0
index b89efaaa81ec..21ada0fe9059 100644
@@ -42,6 +42,7 @@
 #include <linux/security.h>
 #include <linux/notifier.h>
 #include <linux/hashtable.h>
+#include <linux/cc_platform.h>
 #include <rdma/rdma_netlink.h>
 #include <rdma/ib_addr.h>
 #include <rdma/ib_cache.h>
@@ -1419,6 +1420,14 @@ int ib_register_device(struct ib_device *device, const char *name,
 	 */
 	WARN_ON(dma_device && !dma_device->dma_parms);
 	device->dma_device = dma_device;
+	/*
+	 * In a CoCo guest every device is currently assumed to be untrusted
+	 * (T=0) and therefore subject to DMA bouncing. Once trusted (T=1)
+	 * device detection is wired up, narrow this check to exclude such
+	 * devices.
+	 */
+	if (dma_device && cc_platform_has(CC_ATTR_GUEST_MEM_ENCRYPT))
+		device->cc_dma_bounce = 1;
 
 	ret = setup_device(device);
 	if (ret)
drivers/infiniband/core/uverbs_cmd.c +2 -0
index 8eed017091b0..2269f636bf58 100644
@@ -3579,6 +3579,8 @@ static int ib_uverbs_ex_query_device(struct uverbs_attr_bundle *attrs)
 	resp.timestamp_mask = attr.timestamp_mask;
 	resp.hca_core_clock = attr.hca_core_clock;
 	resp.device_cap_flags_ex = attr.device_cap_flags;
+	if (ib_dev->cc_dma_bounce)
+		resp.device_cap_flags_ex |= IB_UVERBS_DEVICE_CC_DMA_BOUNCE;
 	resp.rss_caps.supported_qpts = attr.rss_caps.supported_qpts;
 	resp.rss_caps.max_rwq_indirection_tables =
 		attr.rss_caps.max_rwq_indirection_tables;
include/rdma/ib_verbs.h +3 -0
index 167fb924f0cf..d06071b87d96 100644
@@ -275,6 +275,7 @@ enum ib_device_cap_flags {
 	IB_DEVICE_FLUSH_GLOBAL = IB_UVERBS_DEVICE_FLUSH_GLOBAL,
 	IB_DEVICE_FLUSH_PERSISTENT = IB_UVERBS_DEVICE_FLUSH_PERSISTENT,
 	IB_DEVICE_ATOMIC_WRITE = IB_UVERBS_DEVICE_ATOMIC_WRITE,
+	IB_DEVICE_CC_DMA_BOUNCE = IB_UVERBS_DEVICE_CC_DMA_BOUNCE,
 };
 
 enum ib_kernel_cap_flags {
@@ -2950,6 +2951,8 @@ struct ib_device {
 	u16                          kverbs_provider:1;
 	/* CQ adaptive moderation (RDMA DIM) */
 	u16                          use_cq_dim:1;
+	/* CoCo guest with DMA bounce buffering required */
+	u16                          cc_dma_bounce:1;
 	u8                           node_type;
 	u32			     phys_port_cnt;
 	struct ib_device_attr        attrs;
include/uapi/rdma/ib_user_verbs.h +2 -0
index 3b7bd99813e9..d2aeadb6d2f9 100644
@@ -1368,6 +1368,8 @@ enum ib_uverbs_device_cap_flags {
 	IB_UVERBS_DEVICE_FLUSH_PERSISTENT = 1ULL << 39,
 	/* Atomic write attributes */
 	IB_UVERBS_DEVICE_ATOMIC_WRITE = 1ULL << 40,
+	/* CoCo guest with DMA bounce buffering required */
+	IB_UVERBS_DEVICE_CC_DMA_BOUNCE = 1ULL << 41,
 };
 
 enum ib_uverbs_raw_packet_caps {
Signature · 1 line
2.54.0
author
#2 lore raw

PATCH v3 2/2 RDMA/umem: block plain userspace memory registration under CoCo bounce

In reply to Jiri Pirko

From: Jiri Pirko <[email protected]>

When a device requires DMA bounce buffering inside a Confidential Computing guest, __ib_umem_get_va() cannot work. The DMA mapping layer redirects all mappings through swiotlb bounce buffers, so the device receives DMA addresses pointing to bounce buffer memory rather than the user's pages. Since RDMA devices access registered memory directly without CPU involvement, there is no opportunity for swiotlb to synchronize between the bounce buffer and the original pages.

The registration would already fail later on, since the umem mapping is requested with DMA_ATTR_REQUIRE_COHERENT and gets rejected under is_swiotlb_force_bounce() with -EIO. Fail early with -EOPNOTSUPP instead, so the user gets a specific error code to react to.

---
v1->v2:
- updated patch description with mention of DMA_ATTR_REQUIRE_COHERENT

---
 drivers/infiniband/core/umem.c | 3 +++
 1 file changed, 3 insertions(+)
drivers/infiniband/core/umem.c +3 -0
index eb1de32bab9d..b32bc2a5d7d0 100644
@@ -167,6 +167,9 @@ static struct ib_umem *__ib_umem_get_va(struct ib_device *device,
 	int pinned, ret;
 	unsigned int gup_flags = FOLL_LONGTERM;
 
+	if (device->cc_dma_bounce)
+		return ERR_PTR(-EOPNOTSUPP);
+
 	/*
 	 * If the combination of the addr and size requested for this memory
 	 * region causes an integer overflow, return error.
Signature · 1 line
2.54.0
#3 lore raw

Re: [PATCH rdma-next v3 2/2] RDMA/umem: block plain userspace memory registration under CoCo bounce

Quoted 28 lines · Jiri Pirko

From: Jiri Pirko <[email protected]>

When a device requires DMA bounce buffering inside a Confidential Computing guest, __ib_umem_get_va() cannot work. The DMA mapping layer redirects all mappings through swiotlb bounce buffers, so the device receives DMA addresses pointing to bounce buffer memory rather than the user's pages. Since RDMA devices access registered memory directly without CPU involvement, there is no opportunity for swiotlb to synchronize between the bounce buffer and the original pages.

The registration would already fail later on, since the umem mapping is requested with DMA_ATTR_REQUIRE_COHERENT and gets rejected under is_swiotlb_force_bounce() with -EIO. Fail early with -EOPNOTSUPP instead, so the user gets a specific error code to react to.

---
v1->v2:
- updated patch description with mention of DMA_ATTR_REQUIRE_COHERENT

---
 drivers/infiniband/core/umem.c | 3 +++
 1 file changed, 3 insertions(+)
drivers/infiniband/core/umem.c +0 -0
index eb1de32bab9d..b32bc2a5d7d0 100644

drivers/infiniband/core/umem.c static struct ib_umem *__ib_umem_get_va(struct ib_device *device, · line 172

Quoted 6 lines
drivers/infiniband/core/umem.c +3 -0
@@ -167,6 +167,9 @@ static struct ib_umem *__ib_umem_get_va(struct ib_device *device,
 	int pinned, ret;
 	unsigned int gup_flags = FOLL_LONGTERM;
 
+	if (device->cc_dma_bounce)
+		return ERR_PTR(-EOPNOTSUPP);
+

The series looks reasonable, but I cannot apply it yet because __ib_umem_get_va() has not been merged.

Thanks

Quoted 6 lines
drivers/infiniband/core/umem.c +0 -0
 	/*
 	 * If the combination of the addr and size requested for this memory
 	 * region causes an integer overflow, return error.
Signature · 1 line
2.54.0
author
#4 lore raw

Re: [PATCH rdma-next v3 2/2] RDMA/umem: block plain userspace memory registration under CoCo bounce

Quoted 32 lines · Sun, May 17, 2026 at 06:17:12PM +0200, [email protected] wrote:
Quoted 31 lines · Jiri Pirko

From: Jiri Pirko <[email protected]>

When a device requires DMA bounce buffering inside a Confidential Computing guest, __ib_umem_get_va() cannot work. The DMA mapping layer redirects all mappings through swiotlb bounce buffers, so the device receives DMA addresses pointing to bounce buffer memory rather than the user's pages. Since RDMA devices access registered memory directly without CPU involvement, there is no opportunity for swiotlb to synchronize between the bounce buffer and the original pages.

The registration would already fail later on, since the umem mapping is requested with DMA_ATTR_REQUIRE_COHERENT and gets rejected under is_swiotlb_force_bounce() with -EIO. Fail early with -EOPNOTSUPP instead, so the user gets a specific error code to react to.

---
v1->v2:
- updated patch description with mention of DMA_ATTR_REQUIRE_COHERENT

---
 drivers/infiniband/core/umem.c | 3 +++
 1 file changed, 3 insertions(+)
drivers/infiniband/core/umem.c +0 -0
index eb1de32bab9d..b32bc2a5d7d0 100644
@@ -167,6 +167,9 @@ static struct ib_umem *__ib_umem_get_va(struct ib_device *device,
 	int pinned, ret;
 	unsigned int gup_flags = FOLL_LONGTERM;
 
Quoted 6 lines

drivers/infiniband/core/umem.c static struct ib_umem *__ib_umem_get_va(struct ib_device *device, · line 172

Quoted 3 lines
drivers/infiniband/core/umem.c +3 -0
+	if (device->cc_dma_bounce)
+		return ERR_PTR(-EOPNOTSUPP);
+

The series looks reasonable, but I cannot apply it yet because __ib_umem_get_va() has not been merged.

Correct. Cover letter says:
based on top of:
https://lore.kernel.org/all/[email protected]/

maintainer Applied
#5 lore raw
In reply to Jiri Pirko
Quoted 3 lines · Jiri Pirko

Jiri Pirko (2):
RDMA/uverbs: expose CoCo DMA bounce requirement to userspace RDMA/umem: block plain userspace memory registration under CoCo bounce

Applied to for-next

Thanks,
Jason