[dpdk-dev] [PATCH 1/2] net/mlx4: support externally allocated static memory

[Yongseok Koh]

When MLX PMD registers memory for DMA, it accesses the global memseg list
of DPDK to maximize the range of registration so that LKey search can be
more efficient. Granularity of MR registration is per page.

Externally allocated memory shouldn't be used for DMA because it can't be
searched in the memseg list and free event can't be tracked by DPDK. If it
is used, the following error will occur:

	net_mlx5: port 0 unable to find virtually contiguous chunk for
	address (0x5600017587c0). rte_memseg_contig_walk() failed.

There's a pending patchset [1] which enables externally allocated memory.
Once it is merged, users can register their own memory out of EAL then that
will resolve this issue.

Meanwhile, if the external memory is static (allocated on startup and never
freed), such memory can also be registered by little tweak in the code.

[1] http://patches.dpdk.org/project/dpdk/list/?series=1415

This patch is not a bug fix but needs to be included in stable versions.

Fixes: 9797bfcce1c9 ("net/mlx4: add new memory region support")
Cc: stable at dpdk.org
Cc: "Damjan Marion (damarion)" <damarion at cisco.com>
Cc: Ed Warnicke <hagbard at gmail.com>

Signed-off-by: Yongseok Koh <yskoh at mellanox.com>
---
 drivers/net/mlx4/mlx4_mr.c   | 149 +++++++++++++++++++++++++++++++++++++++++++
 drivers/net/mlx4/mlx4_rxtx.h |  35 +++++++++-
 2 files changed, 183 insertions(+), 1 deletion(-)

diff --git a/drivers/net/mlx4/mlx4_mr.c b/drivers/net/mlx4/mlx4_mr.c
index d23d3c613..bee858643 100644
--- a/drivers/net/mlx4/mlx4_mr.c
+++ b/drivers/net/mlx4/mlx4_mr.c
@@ -289,6 +289,23 @@ mr_find_next_chunk(struct mlx4_mr *mr, struct mlx4_mr_cache *entry,
 	uintptr_t end = 0;
 	uint32_t idx = 0;
 
+	/* MR for external memory doesn't have memseg list. */
+	if (mr->msl == NULL) {
+		struct ibv_mr *ibv_mr = mr->ibv_mr;
+
+		assert(mr->ms_bmp_n == 1);
+		assert(mr->ms_n == 1);
+		assert(base_idx == 0);
+		/*
+		 * Can't search it from memseg list but get it directly from
+		 * verbs MR as there's only one chunk.
+		 */
+		entry->start = (uintptr_t)ibv_mr->addr;
+		entry->end = (uintptr_t)ibv_mr->addr + mr->ibv_mr->length;
+		entry->lkey = rte_cpu_to_be_32(mr->ibv_mr->lkey);
+		/* Returning 1 ends iteration. */
+		return 1;
+	}
 	for (idx = base_idx; idx < mr->ms_bmp_n; ++idx) {
 		if (rte_bitmap_get(mr->ms_bmp, idx)) {
 			const struct rte_memseg_list *msl;
@@ -809,6 +826,7 @@ mlx4_mr_mem_event_free_cb(struct rte_eth_dev *dev, const void *addr, size_t len)
 		mr = mr_lookup_dev_list(dev, &entry, start);
 		if (mr == NULL)
 			continue;
+		assert(mr->msl); /* Can't be external memory. */
 		ms = rte_mem_virt2memseg((void *)start, msl);
 		assert(ms != NULL);
 		assert(msl->page_sz == ms->hugepage_sz);
@@ -1055,6 +1073,133 @@ mlx4_mr_flush_local_cache(struct mlx4_mr_ctrl *mr_ctrl)
 	      (void *)mr_ctrl, mr_ctrl->cur_gen);
 }
 
+/**
+ * Called during rte_mempool_mem_iter() by mlx4_mr_update_ext_mp().
+ *
+ * Externally allocated chunk is registered and a MR is created for the chunk.
+ * The MR object is added to the global list. If memseg list of a MR object
+ * (mr->msl) is null, the MR object can be regarded as externally allocated
+ * memory.
+ *
+ * Once external memory is registered, it should be static. If the memory is
+ * freed and the virtual address range has different physical memory mapped
+ * again, it may cause crash on device due to the wrong translation entry. PMD
+ * can't track the free event of the external memory for now.
+ */
+static void
+mlx4_mr_update_ext_mp_cb(struct rte_mempool *mp, void *opaque,
+			 struct rte_mempool_memhdr *memhdr,
+			 unsigned mem_idx __rte_unused)
+{
+	struct mr_update_mp_data *data = opaque;
+	struct rte_eth_dev *dev = data->dev;
+	struct priv *priv = dev->data->dev_private;
+	struct mlx4_mr_ctrl *mr_ctrl = data->mr_ctrl;
+	struct mlx4_mr *mr = NULL;
+	uintptr_t addr = (uintptr_t)memhdr->addr;
+	size_t len = memhdr->len;
+	struct mlx4_mr_cache entry;
+	uint32_t lkey;
+
+	/* If already registered, it should return. */
+	rte_rwlock_read_lock(&priv->mr.rwlock);
+	lkey = mr_lookup_dev(dev, &entry, addr);
+	rte_rwlock_read_unlock(&priv->mr.rwlock);
+	if (lkey != UINT32_MAX)
+		return;
+	mr = rte_zmalloc_socket(NULL,
+				RTE_ALIGN_CEIL(sizeof(*mr),
+					       RTE_CACHE_LINE_SIZE),
+				RTE_CACHE_LINE_SIZE, mp->socket_id);
+	if (mr == NULL) {
+		WARN("port %u unable to allocate memory for a new MR of"
+		     " mempool (%s).",
+		     dev->data->port_id, mp->name);
+		data->ret = -1;
+		return;
+	}
+	DEBUG("port %u register MR for chunk #%d of mempool (%s)",
+	      dev->data->port_id, mem_idx, mp->name);
+	mr->ibv_mr = mlx4_glue->reg_mr(priv->pd, (void *)addr, len,
+				       IBV_ACCESS_LOCAL_WRITE);
+	if (mr->ibv_mr == NULL) {
+		WARN("port %u fail to create a verbs MR for address (%p)",
+		     dev->data->port_id, (void *)addr);
+		rte_free(mr);
+		data->ret = -1;
+		return;
+	}
+	mr->msl = NULL; /* Mark it is external memory. */
+	mr->ms_bmp = NULL;
+	mr->ms_n = 1;
+	mr->ms_bmp_n = 1;
+	rte_rwlock_write_lock(&priv->mr.rwlock);
+	LIST_INSERT_HEAD(&priv->mr.mr_list, mr, mr);
+	DEBUG("port %u MR CREATED (%p) for external memory %p:\n"
+	      "  [0x%" PRIxPTR ", 0x%" PRIxPTR "),"
+	      " lkey=0x%x base_idx=%u ms_n=%u, ms_bmp_n=%u",
+	      dev->data->port_id, (void *)mr, (void *)addr,
+	      addr, addr + len, rte_cpu_to_be_32(mr->ibv_mr->lkey),
+	      mr->ms_base_idx, mr->ms_n, mr->ms_bmp_n);
+	/* Insert to the global cache table. */
+	mr_insert_dev_cache(dev, mr);
+	rte_rwlock_write_unlock(&priv->mr.rwlock);
+	/* Insert to the local cache table */
+	mlx4_mr_addr2mr_bh(dev, mr_ctrl, addr);
+}
+
+/**
+ * Register MR for entire memory chunks in a Mempool having externally allocated
+ * memory and fill in local cache.
+ *
+ * @param dev
+ *   Pointer to Ethernet device.
+ * @param mr_ctrl
+ *   Pointer to per-queue MR control structure.
+ * @param mp
+ *   Pointer to registering Mempool.
+ *
+ * @return
+ *   0 on success, -1 on failure.
+ */
+static uint32_t
+mlx4_mr_update_ext_mp(struct rte_eth_dev *dev, struct mlx4_mr_ctrl *mr_ctrl,
+		      struct rte_mempool *mp)
+{
+	struct mr_update_mp_data data = {
+		.dev = dev,
+		.mr_ctrl = mr_ctrl,
+		.ret = 0,
+	};
+
+	rte_mempool_mem_iter(mp, mlx4_mr_update_ext_mp_cb, &data);
+	return data.ret;
+}
+
+/**
+ * Register MR entire memory chunks in a Mempool having externally allocated
+ * memory and search LKey of the address to return.
+ *
+ * @param dev
+ *   Pointer to Ethernet device.
+ * @param addr
+ *   Search key.
+ * @param mp
+ *   Pointer to registering Mempool where addr belongs.
+ *
+ * @return
+ *   LKey for address on success, UINT32_MAX on failure.
+ */
+uint32_t
+mlx4_tx_update_ext_mp(struct txq *txq, uintptr_t addr, struct rte_mempool *mp)
+{
+	struct mlx4_mr_ctrl *mr_ctrl = &txq->mr_ctrl;
+	struct priv *priv = txq->priv;
+
+	mlx4_mr_update_ext_mp(priv->dev, mr_ctrl, mp);
+	return mlx4_tx_addr2mr_bh(txq, addr);
+}
+
 /* Called during rte_mempool_mem_iter() by mlx4_mr_update_mp(). */
 static void
 mlx4_mr_update_mp_cb(struct rte_mempool *mp __rte_unused, void *opaque,
@@ -1098,6 +1243,10 @@ mlx4_mr_update_mp(struct rte_eth_dev *dev, struct mlx4_mr_ctrl *mr_ctrl,
 	};
 
 	rte_mempool_mem_iter(mp, mlx4_mr_update_mp_cb, &data);
+	if (data.ret < 0 && rte_errno == ENXIO) {
+		/* Mempool may have externally allocated memory. */
+		return mlx4_mr_update_ext_mp(dev, mr_ctrl, mp);
+	}
 	return data.ret;
 }
 
diff --git a/drivers/net/mlx4/mlx4_rxtx.h b/drivers/net/mlx4/mlx4_rxtx.h
index ffa8abfca..1be060cda 100644
--- a/drivers/net/mlx4/mlx4_rxtx.h
+++ b/drivers/net/mlx4/mlx4_rxtx.h
@@ -163,6 +163,26 @@ void mlx4_tx_queue_release(void *dpdk_txq);
 void mlx4_mr_flush_local_cache(struct mlx4_mr_ctrl *mr_ctrl);
 uint32_t mlx4_rx_addr2mr_bh(struct rxq *rxq, uintptr_t addr);
 uint32_t mlx4_tx_addr2mr_bh(struct txq *txq, uintptr_t addr);
+uint32_t mlx4_tx_update_ext_mp(struct txq *txq, uintptr_t addr,
+			       struct rte_mempool *mp);
+
+/**
+ * Get Memory Pool (MP) from mbuf. If mbuf is indirect, the pool from which the
+ * cloned mbuf is allocated is returned instead.
+ *
+ * @param buf
+ *   Pointer to mbuf.
+ *
+ * @return
+ *   Memory pool where data is located for given mbuf.
+ */
+static struct rte_mempool *
+mlx4_mb2mp(struct rte_mbuf *buf)
+{
+	if (unlikely(RTE_MBUF_INDIRECT(buf)))
+		return rte_mbuf_from_indirect(buf)->pool;
+	return buf->pool;
+}
 
 /**
  * Query LKey from a packet buffer for Rx. No need to flush local caches for Rx
@@ -222,6 +242,19 @@ mlx4_tx_addr2mr(struct txq *txq, uintptr_t addr)
 	return mlx4_tx_addr2mr_bh(txq, addr);
 }
 
-#define mlx4_tx_mb2mr(rxq, mb) mlx4_tx_addr2mr(rxq, (uintptr_t)((mb)->buf_addr))
+static __rte_always_inline uint32_t
+mlx4_tx_mb2mr(struct txq *txq, struct rte_mbuf *mb)
+{
+	uintptr_t addr = (uintptr_t)mb->buf_addr;
+	uint32_t lkey = mlx4_tx_addr2mr(txq, addr);
+
+	if (likely(lkey != UINT32_MAX))
+		return lkey;
+	if (rte_errno == ENXIO) {
+		/* Mempool may have externally allocated memory. */
+		lkey = mlx4_tx_update_ext_mp(txq, addr, mlx4_mb2mp(mb));
+	}
+	return lkey;
+}
 
 #endif /* MLX4_RXTX_H_ */
-- 
2.11.0