Skip to contentSkip to conversation

net/mlx5e: xsk: Discard unaligned XSK frames on striding RQ

PATCH Applied Maxim Mikityanskiy 4 participants 7 messages Jul 29, 2022 – Aug 2, 2022

Subsystems (4)
lore
Share & tools
lore
#0 lore raw

PATCH 1/1 net/mlx5e: xsk: Discard unaligned XSK frames on striding RQ

Striding RQ uses MTT page mapping, where each page corresponds to an XSK frame. MTT pages have alignment requirements, and XSK frames don't have any alignment guarantees in the unaligned mode. Frames with improper alignment must be discarded, otherwise the packet data will be written at a wrong address.

Fixes 282c0c798f8e ("net/mlx5e: Allow XSK frames smaller than a page")
---
 .../net/ethernet/mellanox/mlx5/core/en/xsk/rx.h    | 14 ++++++++++++++
 include/net/xdp_sock_drv.h                         | 11 +++++++++++
 2 files changed, 25 insertions(+)
drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.h +14 -0
index a8cfab4a393c..cc18d97d8ee0 100644
@@ -7,6 +7,8 @@
 #include "en.h"
 #include <net/xdp_sock_drv.h>
 
+#define MLX5E_MTT_PTAG_MASK 0xfffffffffffffff8ULL
+
 /* RX data path */
 
 struct sk_buff *mlx5e_xsk_skb_from_cqe_mpwrq_linear(struct mlx5e_rq *rq,
@@ -21,6 +23,7 @@ struct sk_buff *mlx5e_xsk_skb_from_cqe_linear(struct mlx5e_rq *rq,
 static inline int mlx5e_xsk_page_alloc_pool(struct mlx5e_rq *rq,
 					    struct mlx5e_dma_info *dma_info)
 {
+retry:
 	dma_info->xsk = xsk_buff_alloc(rq->xsk_pool);
 	if (!dma_info->xsk)
 		return -ENOMEM;
@@ -32,6 +35,17 @@ static inline int mlx5e_xsk_page_alloc_pool(struct mlx5e_rq *rq,
 	 */
 	dma_info->addr = xsk_buff_xdp_get_frame_dma(dma_info->xsk);
 
+	/* MTT page mapping has alignment requirements. If they are not
+	 * satisfied, leak the descriptor so that it won't come again, and try
+	 * to allocate a new one.
+	 */
+	if (rq->wq_type == MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ) {
+		if (unlikely(dma_info->addr & ~MLX5E_MTT_PTAG_MASK)) {
+			xsk_buff_discard(dma_info->xsk);
+			goto retry;
+		}
+	}
+
 	return 0;
 }
 
include/net/xdp_sock_drv.h +11 -0
index 4aa031849668..0774ce97c2f1 100644
@@ -95,6 +95,13 @@ static inline void xsk_buff_free(struct xdp_buff *xdp)
 	xp_free(xskb);
 }
 
+static inline void xsk_buff_discard(struct xdp_buff *xdp)
+{
+	struct xdp_buff_xsk *xskb = container_of(xdp, struct xdp_buff_xsk, xdp);
+
+	xp_release(xskb);
+}
+
 static inline void xsk_buff_set_size(struct xdp_buff *xdp, u32 size)
 {
 	xdp->data = xdp->data_hard_start + XDP_PACKET_HEADROOM;
@@ -238,6 +245,10 @@ static inline void xsk_buff_free(struct xdp_buff *xdp)
 {
 }
 
+static inline void xsk_buff_discard(struct xdp_buff *xdp)
+{
+}
+
 static inline void xsk_buff_set_size(struct xdp_buff *xdp, u32 size)
 {
 }
Signature · 1 line
2.25.1
#1 lore raw

Any comments on this patch, or can it be merged?

Saeed reviewed the mlx5 part.

Björn, Magnus, Maciej, anything to say about the XSK drv part?

Quoted 83 lines · Maxim Mikityanskiy

Striding RQ uses MTT page mapping, where each page corresponds to an XSK frame. MTT pages have alignment requirements, and XSK frames don't have any alignment guarantees in the unaligned mode. Frames with improper alignment must be discarded, otherwise the packet data will be written at a wrong address.

Fixes 282c0c798f8e ("net/mlx5e: Allow XSK frames smaller than a page")
---
 .../net/ethernet/mellanox/mlx5/core/en/xsk/rx.h    | 14 ++++++++++++++
 include/net/xdp_sock_drv.h                         | 11 +++++++++++
 2 files changed, 25 insertions(+)
drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.h +14 -0
index a8cfab4a393c..cc18d97d8ee0 100644
@@ -7,6 +7,8 @@
 #include "en.h"
 #include <net/xdp_sock_drv.h>
 
+#define MLX5E_MTT_PTAG_MASK 0xfffffffffffffff8ULL
+
 /* RX data path */
 
 struct sk_buff *mlx5e_xsk_skb_from_cqe_mpwrq_linear(struct mlx5e_rq *rq,
@@ -21,6 +23,7 @@ struct sk_buff *mlx5e_xsk_skb_from_cqe_linear(struct mlx5e_rq *rq,
 static inline int mlx5e_xsk_page_alloc_pool(struct mlx5e_rq *rq,
 					    struct mlx5e_dma_info *dma_info)
 {
+retry:
 	dma_info->xsk = xsk_buff_alloc(rq->xsk_pool);
 	if (!dma_info->xsk)
 		return -ENOMEM;
@@ -32,6 +35,17 @@ static inline int mlx5e_xsk_page_alloc_pool(struct mlx5e_rq *rq,
 	 */
 	dma_info->addr = xsk_buff_xdp_get_frame_dma(dma_info->xsk);
 
+	/* MTT page mapping has alignment requirements. If they are not
+	 * satisfied, leak the descriptor so that it won't come again, and try
+	 * to allocate a new one.
+	 */
+	if (rq->wq_type == MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ) {
+		if (unlikely(dma_info->addr & ~MLX5E_MTT_PTAG_MASK)) {
+			xsk_buff_discard(dma_info->xsk);
+			goto retry;
+		}
+	}
+
 	return 0;
 }
 
include/net/xdp_sock_drv.h +11 -0
index 4aa031849668..0774ce97c2f1 100644
@@ -95,6 +95,13 @@ static inline void xsk_buff_free(struct xdp_buff *xdp)
 	xp_free(xskb);
 }
 
+static inline void xsk_buff_discard(struct xdp_buff *xdp)
+{
+	struct xdp_buff_xsk *xskb = container_of(xdp, struct xdp_buff_xsk, xdp);
+
+	xp_release(xskb);
+}
+
 static inline void xsk_buff_set_size(struct xdp_buff *xdp, u32 size)
 {
 	xdp->data = xdp->data_hard_start + XDP_PACKET_HEADROOM;
@@ -238,6 +245,10 @@ static inline void xsk_buff_free(struct xdp_buff *xdp)
 {
 }
 
+static inline void xsk_buff_discard(struct xdp_buff *xdp)
+{
+}
+
 static inline void xsk_buff_set_size(struct xdp_buff *xdp, u32 size)
 {
 }
#2 lore raw
In reply to Maxim Mikityanskiy
Quoted 5 lines · Maxim Mikityanskiy

Striding RQ uses MTT page mapping, where each page corresponds to an XSK frame. MTT pages have alignment requirements, and XSK frames don't have any alignment guarantees in the unaligned mode. Frames with improper alignment must be discarded, otherwise the packet data will be written at a wrong address.

Hey Maxim,
can you explain what MTT stands for?

Quoted 15 lines
Fixes 282c0c798f8e ("net/mlx5e: Allow XSK frames smaller than a page")
---
 .../net/ethernet/mellanox/mlx5/core/en/xsk/rx.h    | 14 ++++++++++++++
 include/net/xdp_sock_drv.h                         | 11 +++++++++++
 2 files changed, 25 insertions(+)
drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.h +0 -0
index a8cfab4a393c..cc18d97d8ee0 100644

drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.h line 10

Quoted 4 lines
drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.h +1 -0
@@ -7,6 +7,8 @@
 #include "en.h"
 #include <net/xdp_sock_drv.h>
 
+#define MLX5E_MTT_PTAG_MASK 0xfffffffffffffff8ULL

What if PAGE_SIZE != 4096 ? Is aligned mode with 2k frame fine for MTT case?

Quoted 16 lines
drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.h +2 -0
+
 /* RX data path */
 
 struct sk_buff *mlx5e_xsk_skb_from_cqe_mpwrq_linear(struct mlx5e_rq *rq,
@@ -21,6 +23,7 @@ struct sk_buff *mlx5e_xsk_skb_from_cqe_linear(struct mlx5e_rq *rq,
 static inline int mlx5e_xsk_page_alloc_pool(struct mlx5e_rq *rq,
 					    struct mlx5e_dma_info *dma_info)
 {
+retry:
 	dma_info->xsk = xsk_buff_alloc(rq->xsk_pool);
 	if (!dma_info->xsk)
 		return -ENOMEM;
@@ -32,6 +35,17 @@ static inline int mlx5e_xsk_page_alloc_pool(struct mlx5e_rq *rq,
 	 */
 	dma_info->addr = xsk_buff_xdp_get_frame_dma(dma_info->xsk);
 

drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.h static inline int mlx5e_xsk_page_alloc_pool(struct mlx5e_rq *rq, · line 47

Quoted 10 lines
drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.h +10 -0
+	/* MTT page mapping has alignment requirements. If they are not
+	 * satisfied, leak the descriptor so that it won't come again, and try
+	 * to allocate a new one.
+	 */
+	if (rq->wq_type == MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ) {
+		if (unlikely(dma_info->addr & ~MLX5E_MTT_PTAG_MASK)) {
+			xsk_buff_discard(dma_info->xsk);
+			goto retry;
+		}
+	}

I don't know your hardware much, but how would this work out performance wise? Are there any config combos (page size vs chunk size in unaligned mode) that you would forbid during pool attach to queue or would you better allow anything?

Also would be helpful if you would describe the use case you're fixing.

Thanks!

Quoted 36 lines
drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.h +1 -0
+
 	return 0;
 }
 
include/net/xdp_sock_drv.h +11 -0
index 4aa031849668..0774ce97c2f1 100644
@@ -95,6 +95,13 @@ static inline void xsk_buff_free(struct xdp_buff *xdp)
 	xp_free(xskb);
 }
 
+static inline void xsk_buff_discard(struct xdp_buff *xdp)
+{
+	struct xdp_buff_xsk *xskb = container_of(xdp, struct xdp_buff_xsk, xdp);
+
+	xp_release(xskb);
+}
+
 static inline void xsk_buff_set_size(struct xdp_buff *xdp, u32 size)
 {
 	xdp->data = xdp->data_hard_start + XDP_PACKET_HEADROOM;
@@ -238,6 +245,10 @@ static inline void xsk_buff_free(struct xdp_buff *xdp)
 {
 }
 
+static inline void xsk_buff_discard(struct xdp_buff *xdp)
+{
+}
+
 static inline void xsk_buff_set_size(struct xdp_buff *xdp, u32 size)
 {
 }
Signature · 1 line
2.25.1
#3 lore raw

First of all, this patch is a temporary kludge. I found a bug in the current implementation of the unaligned mode: frames not aligned at least to 8 are misplaced. There is a proper fix in the driver, but it will be pushed to net-next, because it's huge. In the meanwhile, this workaround that drops packets not aligned to 8 will go to stable kernels.

Quoted 9 lines · Maciej Fijalkowski
Quoted 5 lines · Maxim Mikityanskiy

Striding RQ uses MTT page mapping, where each page corresponds to an XSK frame. MTT pages have alignment requirements, and XSK frames don't have any alignment guarantees in the unaligned mode. Frames with improper alignment must be discarded, otherwise the packet data will be written at a wrong address.

Hey Maxim,
can you explain what MTT stands for?

MTT is Memory Translation Table, it's a mechanism for virtual mapping in the NIC. It's essentially a table of pages, where each virtual page maps to a physical page.

Quoted 19 lines
Quoted 19 lines
Fixes 282c0c798f8e ("net/mlx5e: Allow XSK frames smaller than a page")
---
 .../net/ethernet/mellanox/mlx5/core/en/xsk/rx.h    | 14 ++++++++++++++
 include/net/xdp_sock_drv.h                         | 11 +++++++++++
 2 files changed, 25 insertions(+)
drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.h +0 -0
index a8cfab4a393c..cc18d97d8ee0 100644
@@ -7,6 +7,8 @@
 #include "en.h"
 #include <net/xdp_sock_drv.h>
 
Quoted 4 lines

drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.h line 10

Quoted 1 line
drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.h +1 -0
+#define MLX5E_MTT_PTAG_MASK 0xfffffffffffffff8ULL

What if PAGE_SIZE != 4096 ? Is aligned mode with 2k frame fine for MTT case?

PAGE_SIZE doesn't affect this value. Aligned mode doesn't suffer from this bug, because 2k or bigger frames are all aligned to 8.

Quoted 17 lines
Quoted 17 lines
drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.h +2 -0
+
 /* RX data path */
 
 struct sk_buff *mlx5e_xsk_skb_from_cqe_mpwrq_linear(struct mlx5e_rq *rq,
@@ -21,6 +23,7 @@ struct sk_buff *mlx5e_xsk_skb_from_cqe_linear(struct mlx5e_rq *rq,
 static inline int mlx5e_xsk_page_alloc_pool(struct mlx5e_rq *rq,
 					    struct mlx5e_dma_info *dma_info)
 {
+retry:
 	dma_info->xsk = xsk_buff_alloc(rq->xsk_pool);
 	if (!dma_info->xsk)
 		return -ENOMEM;
@@ -32,6 +35,17 @@ static inline int mlx5e_xsk_page_alloc_pool(struct mlx5e_rq *rq,
 	 */
 	dma_info->addr = xsk_buff_xdp_get_frame_dma(dma_info->xsk);
 
Quoted 15 lines

drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.h static inline int mlx5e_xsk_page_alloc_pool(struct mlx5e_rq *rq, · line 47

Quoted 10 lines
drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.h +10 -0
+	/* MTT page mapping has alignment requirements. If they are not
+	 * satisfied, leak the descriptor so that it won't come again, and try
+	 * to allocate a new one.
+	 */
+	if (rq->wq_type == MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ) {
+		if (unlikely(dma_info->addr & ~MLX5E_MTT_PTAG_MASK)) {
+			xsk_buff_discard(dma_info->xsk);
+			goto retry;
+		}
+	}

I don't know your hardware much, but how would this work out performance wise? Are there any config combos (page size vs chunk size in unaligned mode) that you would forbid during pool attach to queue or would you better allow anything?

This issue isn't related to page or frame sizes, but rather to frame locations. As far as I understand, frames can be located at any places in the unaligned mode (even at odd addresses), regardless of their size. Frames whose addr % 8 != 0 don't really work with MTT, but it's not something that can be enforced on attach. Enforcing it in xp_alloc won't be any faster either (well, only a tiny bit, because of one fewer function call).

In any case, next kernels will get another page mapping mechanism, which supports arbitrary addresses, and it's almost as fast as MTT, as the preliminary testing shows. It will be used for the unaligned XSK, this kludge will be removed altogether, and I also plan to remove xsk_buff_discard.

Quoted 1 line

Also would be helpful if you would describe the use case you're fixing.

Sure - described in the beginning of the email.

Quoted 39 lines

Thanks!

Quoted 36 lines
drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.h +1 -0
+
 	return 0;
 }
 
include/net/xdp_sock_drv.h +11 -0
index 4aa031849668..0774ce97c2f1 100644
@@ -95,6 +95,13 @@ static inline void xsk_buff_free(struct xdp_buff *xdp)
 	xp_free(xskb);
 }
 
+static inline void xsk_buff_discard(struct xdp_buff *xdp)
+{
+	struct xdp_buff_xsk *xskb = container_of(xdp, struct xdp_buff_xsk, xdp);
+
+	xp_release(xskb);
+}
+
 static inline void xsk_buff_set_size(struct xdp_buff *xdp, u32 size)
 {
 	xdp->data = xdp->data_hard_start + XDP_PACKET_HEADROOM;
@@ -238,6 +245,10 @@ static inline void xsk_buff_free(struct xdp_buff *xdp)
 {
 }
 
+static inline void xsk_buff_discard(struct xdp_buff *xdp)
+{
+}
+
 static inline void xsk_buff_set_size(struct xdp_buff *xdp, u32 size)
 {
 }
Signature · 1 line
2.25.1
maintainer
#4 lore raw
Quoted 96 lines · Maxim Mikityanskiy

First of all, this patch is a temporary kludge. I found a bug in the current implementation of the unaligned mode: frames not aligned at least to 8 are misplaced. There is a proper fix in the driver, but it will be pushed to net-next, because it's huge. In the meanwhile, this workaround that drops packets not aligned to 8 will go to stable kernels.

Quoted 9 lines · Maciej Fijalkowski
Quoted 5 lines · Maxim Mikityanskiy

Striding RQ uses MTT page mapping, where each page corresponds to an XSK frame. MTT pages have alignment requirements, and XSK frames don't have any alignment guarantees in the unaligned mode. Frames with improper alignment must be discarded, otherwise the packet data will be written at a wrong address.

Hey Maxim,
can you explain what MTT stands for?

MTT is Memory Translation Table, it's a mechanism for virtual mapping in the NIC. It's essentially a table of pages, where each virtual page maps to a physical page.

Quoted 19 lines
Quoted 19 lines
Fixes 282c0c798f8e ("net/mlx5e: Allow XSK frames smaller than a page")
---
 .../net/ethernet/mellanox/mlx5/core/en/xsk/rx.h    | 14 ++++++++++++++
 include/net/xdp_sock_drv.h                         | 11 +++++++++++
 2 files changed, 25 insertions(+)
drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.h +0 -0
index a8cfab4a393c..cc18d97d8ee0 100644
@@ -7,6 +7,8 @@
 #include "en.h"
 #include <net/xdp_sock_drv.h>
 
Quoted 4 lines

drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.h line 10

Quoted 1 line
drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.h +1 -0
+#define MLX5E_MTT_PTAG_MASK 0xfffffffffffffff8ULL

What if PAGE_SIZE != 4096 ? Is aligned mode with 2k frame fine for MTT case?

PAGE_SIZE doesn't affect this value. Aligned mode doesn't suffer from this bug, because 2k or bigger frames are all aligned to 8.

Quoted 17 lines
Quoted 17 lines
drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.h +2 -0
+
 /* RX data path */
 
 struct sk_buff *mlx5e_xsk_skb_from_cqe_mpwrq_linear(struct mlx5e_rq *rq,
@@ -21,6 +23,7 @@ struct sk_buff *mlx5e_xsk_skb_from_cqe_linear(struct mlx5e_rq *rq,
 static inline int mlx5e_xsk_page_alloc_pool(struct mlx5e_rq *rq,
 					    struct mlx5e_dma_info *dma_info)
 {
+retry:
 	dma_info->xsk = xsk_buff_alloc(rq->xsk_pool);
 	if (!dma_info->xsk)
 		return -ENOMEM;
@@ -32,6 +35,17 @@ static inline int mlx5e_xsk_page_alloc_pool(struct mlx5e_rq *rq,
 	 */
 	dma_info->addr = xsk_buff_xdp_get_frame_dma(dma_info->xsk);
 
Quoted 15 lines

drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.h static inline int mlx5e_xsk_page_alloc_pool(struct mlx5e_rq *rq, · line 47

Quoted 10 lines
drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.h +10 -0
+	/* MTT page mapping has alignment requirements. If they are not
+	 * satisfied, leak the descriptor so that it won't come again, and try
+	 * to allocate a new one.
+	 */
+	if (rq->wq_type == MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ) {
+		if (unlikely(dma_info->addr & ~MLX5E_MTT_PTAG_MASK)) {
+			xsk_buff_discard(dma_info->xsk);
+			goto retry;
+		}
+	}

I don't know your hardware much, but how would this work out performance wise? Are there any config combos (page size vs chunk size in unaligned mode) that you would forbid during pool attach to queue or would you better allow anything?

This issue isn't related to page or frame sizes, but rather to frame locations. As far as I understand, frames can be located at any places in the unaligned mode (even at odd addresses), regardless of their size. Frames whose addr % 8 != 0 don't really work with MTT, but it's not something that can be enforced on attach. Enforcing it in xp_alloc won't be any faster either (well, only a tiny bit, because of one fewer function call).

In any case, next kernels will get another page mapping mechanism, which supports arbitrary addresses, and it's almost as fast as MTT, as the preliminary testing shows. It will be used for the unaligned XSK, this kludge will be removed altogether, and I also plan to remove xsk_buff_discard.

Quoted 3 lines
Quoted 1 line

Also would be helpful if you would describe the use case you're fixing.

Sure - described in the beginning of the email.

@Maciej: are you satisfied by Maxim's answers?

/P

maintainer Reviewed-by
#5 lore raw
Quoted 91 lines · Paolo Abeni
Quoted 90 lines · Maxim Mikityanskiy

First of all, this patch is a temporary kludge. I found a bug in the current implementation of the unaligned mode: frames not aligned at least to 8 are misplaced. There is a proper fix in the driver, but it will be pushed to net-next, because it's huge. In the meanwhile, this workaround that drops packets not aligned to 8 will go to stable kernels.

Quoted 9 lines · Maciej Fijalkowski
Quoted 5 lines · Maxim Mikityanskiy

Striding RQ uses MTT page mapping, where each page corresponds to an XSK frame. MTT pages have alignment requirements, and XSK frames don't have any alignment guarantees in the unaligned mode. Frames with improper alignment must be discarded, otherwise the packet data will be written at a wrong address.

Hey Maxim,
can you explain what MTT stands for?

MTT is Memory Translation Table, it's a mechanism for virtual mapping in the NIC. It's essentially a table of pages, where each virtual page maps to a physical page.

Quoted 19 lines
Quoted 19 lines
Fixes 282c0c798f8e ("net/mlx5e: Allow XSK frames smaller than a page")
---
 .../net/ethernet/mellanox/mlx5/core/en/xsk/rx.h    | 14 ++++++++++++++
 include/net/xdp_sock_drv.h                         | 11 +++++++++++
 2 files changed, 25 insertions(+)
drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.h +0 -0
index a8cfab4a393c..cc18d97d8ee0 100644
@@ -7,6 +7,8 @@
 #include "en.h"
 #include <net/xdp_sock_drv.h>
 
Quoted 4 lines

drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.h line 10

Quoted 1 line
drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.h +1 -0
+#define MLX5E_MTT_PTAG_MASK 0xfffffffffffffff8ULL

What if PAGE_SIZE != 4096 ? Is aligned mode with 2k frame fine for MTT case?

PAGE_SIZE doesn't affect this value. Aligned mode doesn't suffer from this bug, because 2k or bigger frames are all aligned to 8.

Quoted 17 lines
Quoted 17 lines
drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.h +2 -0
+
 /* RX data path */
 
 struct sk_buff *mlx5e_xsk_skb_from_cqe_mpwrq_linear(struct mlx5e_rq *rq,
@@ -21,6 +23,7 @@ struct sk_buff *mlx5e_xsk_skb_from_cqe_linear(struct mlx5e_rq *rq,
 static inline int mlx5e_xsk_page_alloc_pool(struct mlx5e_rq *rq,
 					    struct mlx5e_dma_info *dma_info)
 {
+retry:
 	dma_info->xsk = xsk_buff_alloc(rq->xsk_pool);
 	if (!dma_info->xsk)
 		return -ENOMEM;
@@ -32,6 +35,17 @@ static inline int mlx5e_xsk_page_alloc_pool(struct mlx5e_rq *rq,
 	 */
 	dma_info->addr = xsk_buff_xdp_get_frame_dma(dma_info->xsk);
 
Quoted 15 lines

drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.h static inline int mlx5e_xsk_page_alloc_pool(struct mlx5e_rq *rq, · line 47

Quoted 10 lines
drivers/net/ethernet/mellanox/mlx5/core/en/xsk/rx.h +10 -0
+	/* MTT page mapping has alignment requirements. If they are not
+	 * satisfied, leak the descriptor so that it won't come again, and try
+	 * to allocate a new one.
+	 */
+	if (rq->wq_type == MLX5_WQ_TYPE_LINKED_LIST_STRIDING_RQ) {
+		if (unlikely(dma_info->addr & ~MLX5E_MTT_PTAG_MASK)) {
+			xsk_buff_discard(dma_info->xsk);
+			goto retry;
+		}
+	}

I don't know your hardware much, but how would this work out performance wise? Are there any config combos (page size vs chunk size in unaligned mode) that you would forbid during pool attach to queue or would you better allow anything?

This issue isn't related to page or frame sizes, but rather to frame locations. As far as I understand, frames can be located at any places in the unaligned mode (even at odd addresses), regardless of their size. Frames whose addr % 8 != 0 don't really work with MTT, but it's not something that can be enforced on attach. Enforcing it in xp_alloc won't be any faster either (well, only a tiny bit, because of one fewer function call).

Quoted 5 lines
Quoted 5 lines

In any case, next kernels will get another page mapping mechanism, which supports arbitrary addresses, and it's almost as fast as MTT, as the preliminary testing shows. It will be used for the unaligned XSK, this kludge will be removed altogether, and I also plan to remove xsk_buff_discard.

Ok makes sense. I only misread the mask though, so maybe use GENMASK_ULL(63, 3) ? Also, saying explicitly about MTT's requirement issues (8) in the commit message probably wouldn't make me to misread it :p

Besides:

Quoted 6 lines
Quoted 4 lines
Quoted 2 lines

Also would be helpful if you would describe the use case you're fixing.

Sure - described in the beginning of the email.

@Maciej: are you satisfied by Maxim's answers?

Yep!

Quoted 3 lines

/P

#6 lore raw
In reply to Maxim Mikityanskiy
This patch was applied to netdev/net.git (master) by Paolo Abeni <[email protected]>:

Hello:

This patch was applied to netdev/net.git (master)
by Paolo Abeni <[email protected]>:

Quoted 6 lines · On Fri, 29 Jul 2022 15:13:56 +0300 you wrote:

Striding RQ uses MTT page mapping, where each page corresponds to an XSK frame. MTT pages have alignment requirements, and XSK frames don't have any alignment guarantees in the unaligned mode. Frames with improper alignment must be discarded, otherwise the packet data will be written at a wrong address.

Quoted 6 lines
Fixes 282c0c798f8e ("net/mlx5e: Allow XSK frames smaller than a page")

[...]

Here is the summary with links:
- [net] net/mlx5e: xsk: Discard unaligned XSK frames on striding RQ

    https://git.kernel.org/netdev/net/c/8eaa1d110800

You are awesome, thank you!

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