From: Pavan Nikhilesh <pbhagavat...@marvell.com>
Use rte_mempool for DMA chunk pool to allow using mempool cache.
Signed-off-by: Pavan Nikhilesh <pbhagavat...@marvell.com>
---
Depends-on: 29324
drivers/common/cnxk/roc_dpi.c | 95 +++++--------------------
drivers/common/cnxk/roc_dpi.h | 28 +-------
drivers/common/cnxk/roc_dpi_priv.h | 3 -
drivers/common/cnxk/roc_platform.c | 1 +
drivers/common/cnxk/roc_platform.h | 2 +
drivers/common/cnxk/version.map | 1 +
drivers/dma/cnxk/cnxk_dmadev.c | 108 +++++++++++++++++++++--------
drivers/dma/cnxk/cnxk_dmadev.h | 10 ++-
8 files changed, 110 insertions(+), 138 deletions(-)
diff --git a/drivers/common/cnxk/roc_dpi.c b/drivers/common/cnxk/roc_dpi.c
index 0e2f803077..9cb479371a 100644
--- a/drivers/common/cnxk/roc_dpi.c
+++ b/drivers/common/cnxk/roc_dpi.c
@@ -1,14 +1,14 @@
/* SPDX-License-Identifier: BSD-3-Clause
* Copyright(C) 2021 Marvell.
*/
+
+#include "roc_api.h"
+#include "roc_priv.h"
#include <fcntl.h>
#include <sys/stat.h>
#include <sys/types.h>
#include <unistd.h>
-#include "roc_api.h"
-#include "roc_priv.h"
-
#define DPI_PF_MBOX_SYSFS_ENTRY "dpi_device_config"
static inline int
@@ -52,17 +52,12 @@ roc_dpi_disable(struct roc_dpi *dpi)
}
int
-roc_dpi_configure(struct roc_dpi *roc_dpi)
+roc_dpi_configure(struct roc_dpi *roc_dpi, uint32_t chunk_sz, uint64_t aura,
uint64_t chunk_base)
{
struct plt_pci_device *pci_dev;
- const struct plt_memzone *dpi_mz;
dpi_mbox_msg_t mbox_msg;
- struct npa_pool_s pool;
- struct npa_aura_s aura;
- int rc, count, buflen;
- uint64_t aura_handle;
- plt_iova_t iova;
- char name[32];
+ uint64_t reg;
+ int rc;
if (!roc_dpi) {
plt_err("roc_dpi is NULL");
@@ -70,80 +65,31 @@ roc_dpi_configure(struct roc_dpi *roc_dpi)
}
pci_dev = roc_dpi->pci_dev;
- memset(&pool, 0, sizeof(struct npa_pool_s));
- pool.nat_align = 1;
-
- memset(&aura, 0, sizeof(aura));
- rc = roc_npa_pool_create(&aura_handle, DPI_CMD_QUEUE_SIZE,
- DPI_CMD_QUEUE_BUFS, &aura, &pool, 0);
- if (rc) {
- plt_err("Failed to create NPA pool, err %d\n", rc);
- return rc;
- }
-
- snprintf(name, sizeof(name), "dpimem%d:%d:%d:%d", pci_dev->addr.domain,
pci_dev->addr.bus,
- pci_dev->addr.devid, pci_dev->addr.function);
- buflen = DPI_CMD_QUEUE_SIZE * DPI_CMD_QUEUE_BUFS;
- dpi_mz = plt_memzone_reserve_aligned(name, buflen, 0,
DPI_CMD_QUEUE_SIZE);
- if (dpi_mz == NULL) {
- plt_err("dpi memzone reserve failed");
- rc = -ENOMEM;
- goto err1;
- }
-
- roc_dpi->mz = dpi_mz;
- iova = dpi_mz->iova;
- for (count = 0; count < DPI_CMD_QUEUE_BUFS; count++) {
- roc_npa_aura_op_free(aura_handle, 0, iova);
- iova += DPI_CMD_QUEUE_SIZE;
- }
-
- roc_dpi->chunk_base = (void *)roc_npa_aura_op_alloc(aura_handle, 0);
- if (!roc_dpi->chunk_base) {
- plt_err("Failed to alloc buffer from NPA aura");
- rc = -ENOMEM;
- goto err2;
- }
- roc_dpi->chunk_next = (void *)roc_npa_aura_op_alloc(aura_handle, 0);
- if (!roc_dpi->chunk_next) {
- plt_err("Failed to alloc buffer from NPA aura");
- rc = -ENOMEM;
- goto err2;
- }
-
- roc_dpi->aura_handle = aura_handle;
- /* subtract 2 as they have already been alloc'ed above */
- roc_dpi->pool_size_m1 = (DPI_CMD_QUEUE_SIZE >> 3) - 2;
+ roc_dpi_disable(roc_dpi);
+ reg = plt_read64(roc_dpi->rbase + DPI_VDMA_SADDR);
+ while (!(reg & BIT_ULL(63)))
+ reg = plt_read64(roc_dpi->rbase + DPI_VDMA_SADDR);
plt_write64(0x0, roc_dpi->rbase + DPI_VDMA_REQQ_CTL);
- plt_write64(((uint64_t)(roc_dpi->chunk_base) >> 7) << 7,
- roc_dpi->rbase + DPI_VDMA_SADDR);
+ plt_write64(chunk_base, roc_dpi->rbase + DPI_VDMA_SADDR);
mbox_msg.u[0] = 0;
mbox_msg.u[1] = 0;
/* DPI PF driver expects vfid starts from index 0 */
mbox_msg.s.vfid = roc_dpi->vfid;
mbox_msg.s.cmd = DPI_QUEUE_OPEN;
- mbox_msg.s.csize = DPI_CMD_QUEUE_SIZE;
- mbox_msg.s.aura = roc_npa_aura_handle_to_aura(aura_handle);
+ mbox_msg.s.csize = chunk_sz;
+ mbox_msg.s.aura = aura;
mbox_msg.s.sso_pf_func = idev_sso_pffunc_get();
mbox_msg.s.npa_pf_func = idev_npa_pffunc_get();
rc = send_msg_to_pf(&pci_dev->addr, (const char *)&mbox_msg,
sizeof(dpi_mbox_msg_t));
- if (rc < 0) {
+ if (rc < 0)
plt_err("Failed to send mbox message %d to DPI PF, err %d",
mbox_msg.s.cmd, rc);
- goto err2;
- }
return rc;
-
-err2:
- plt_memzone_free(dpi_mz);
-err1:
- roc_npa_pool_destroy(aura_handle);
- return rc;
}
int
@@ -153,11 +99,9 @@ roc_dpi_dev_init(struct roc_dpi *roc_dpi)
uint16_t vfid;
roc_dpi->rbase = pci_dev->mem_resource[0].addr;
- vfid = ((pci_dev->addr.devid & 0x1F) << 3) |
- (pci_dev->addr.function & 0x7);
+ vfid = ((pci_dev->addr.devid & 0x1F) << 3) | (pci_dev->addr.function &
0x7);
vfid -= 1;
roc_dpi->vfid = vfid;
- plt_spinlock_init(&roc_dpi->chunk_lock);
return 0;
}
@@ -180,14 +124,9 @@ roc_dpi_dev_fini(struct roc_dpi *roc_dpi)
mbox_msg.s.vfid = roc_dpi->vfid;
mbox_msg.s.cmd = DPI_QUEUE_CLOSE;
- rc = send_msg_to_pf(&pci_dev->addr, (const char *)&mbox_msg,
- sizeof(dpi_mbox_msg_t));
+ rc = send_msg_to_pf(&pci_dev->addr, (const char *)&mbox_msg,
sizeof(dpi_mbox_msg_t));
if (rc < 0)
- plt_err("Failed to send mbox message %d to DPI PF, err %d",
- mbox_msg.s.cmd, rc);
-
- roc_npa_pool_destroy(roc_dpi->aura_handle);
- plt_memzone_free(roc_dpi->mz);
+ plt_err("Failed to send mbox message %d to DPI PF, err %d",
mbox_msg.s.cmd, rc);
return rc;
}
diff --git a/drivers/common/cnxk/roc_dpi.h b/drivers/common/cnxk/roc_dpi.h
index 2f061b07c5..4ebde5b8a6 100644
--- a/drivers/common/cnxk/roc_dpi.h
+++ b/drivers/common/cnxk/roc_dpi.h
@@ -5,41 +5,17 @@
#ifndef _ROC_DPI_H_
#define _ROC_DPI_H_
-struct roc_dpi_args {
- uint8_t num_ssegs;
- uint8_t num_dsegs;
- uint8_t comp_type;
- uint8_t direction;
- uint8_t sdevice;
- uint8_t ddevice;
- uint8_t swap;
- uint8_t use_lock : 1;
- uint8_t tt : 7;
- uint16_t func;
- uint16_t grp;
- uint32_t tag;
- uint64_t comp_ptr;
-};
-
struct roc_dpi {
- /* Input parameters */
struct plt_pci_device *pci_dev;
- /* End of Input parameters */
- const struct plt_memzone *mz;
uint8_t *rbase;
uint16_t vfid;
- uint16_t pool_size_m1;
- uint16_t chunk_head;
- uint64_t *chunk_base;
- uint64_t *chunk_next;
- uint64_t aura_handle;
- plt_spinlock_t chunk_lock;
} __plt_cache_aligned;
int __roc_api roc_dpi_dev_init(struct roc_dpi *roc_dpi);
int __roc_api roc_dpi_dev_fini(struct roc_dpi *roc_dpi);
-int __roc_api roc_dpi_configure(struct roc_dpi *dpi);
+int __roc_api roc_dpi_configure(struct roc_dpi *dpi, uint32_t chunk_sz,
uint64_t aura,
+ uint64_t chunk_base);
int __roc_api roc_dpi_enable(struct roc_dpi *dpi);
int __roc_api roc_dpi_disable(struct roc_dpi *dpi);
diff --git a/drivers/common/cnxk/roc_dpi_priv.h
b/drivers/common/cnxk/roc_dpi_priv.h
index 1fa1a715d3..518a3e7351 100644
--- a/drivers/common/cnxk/roc_dpi_priv.h
+++ b/drivers/common/cnxk/roc_dpi_priv.h
@@ -16,9 +16,6 @@
#define DPI_REG_DUMP 0x3
#define DPI_GET_REG_CFG 0x4
-#define DPI_CMD_QUEUE_SIZE 4096
-#define DPI_CMD_QUEUE_BUFS 1024
-
typedef union dpi_mbox_msg_t {
uint64_t u[2];
struct dpi_mbox_message_s {
diff --git a/drivers/common/cnxk/roc_platform.c
b/drivers/common/cnxk/roc_platform.c
index f91b95ceab..f8287bcf6b 100644
--- a/drivers/common/cnxk/roc_platform.c
+++ b/drivers/common/cnxk/roc_platform.c
@@ -70,4 +70,5 @@ RTE_LOG_REGISTER(cnxk_logtype_npc, pmd.net.cnxk.flow, NOTICE);
RTE_LOG_REGISTER(cnxk_logtype_sso, pmd.event.cnxk, NOTICE);
RTE_LOG_REGISTER(cnxk_logtype_tim, pmd.event.cnxk.timer, NOTICE);
RTE_LOG_REGISTER(cnxk_logtype_tm, pmd.net.cnxk.tm, NOTICE);
+RTE_LOG_REGISTER(cnxk_logtype_dpi, pmd.dma.cnxk.dpi, NOTICE);
RTE_LOG_REGISTER_DEFAULT(cnxk_logtype_ree, NOTICE);
diff --git a/drivers/common/cnxk/roc_platform.h
b/drivers/common/cnxk/roc_platform.h
index 08f83aba12..dfd4da21b6 100644
--- a/drivers/common/cnxk/roc_platform.h
+++ b/drivers/common/cnxk/roc_platform.h
@@ -242,6 +242,7 @@ extern int cnxk_logtype_sso;
extern int cnxk_logtype_tim;
extern int cnxk_logtype_tm;
extern int cnxk_logtype_ree;
+extern int cnxk_logtype_dpi;
#define plt_err(fmt, args...)
\
RTE_LOG(ERR, PMD, "%s():%u " fmt "\n", __func__, __LINE__, ##args)
@@ -270,6 +271,7 @@ extern int cnxk_logtype_ree;
#define plt_tim_dbg(fmt, ...) plt_dbg(tim, fmt, ##__VA_ARGS__)
#define plt_tm_dbg(fmt, ...) plt_dbg(tm, fmt, ##__VA_ARGS__)
#define plt_ree_dbg(fmt, ...) plt_dbg(ree, fmt, ##__VA_ARGS__)
+#define plt_dpi_dbg(fmt, ...) plt_dbg(dpi, fmt, ##__VA_ARGS__)
/* Datapath logs */
#define plt_dp_err(fmt, args...)
\
diff --git a/drivers/common/cnxk/version.map b/drivers/common/cnxk/version.map
index 8c71497df8..1540dfadf9 100644
--- a/drivers/common/cnxk/version.map
+++ b/drivers/common/cnxk/version.map
@@ -7,6 +7,7 @@ INTERNAL {
cnxk_ipsec_outb_roundup_byte;
cnxk_logtype_base;
cnxk_logtype_cpt;
+ cnxk_logtype_dpi;
cnxk_logtype_mbox;
cnxk_logtype_ml;
cnxk_logtype_nix;
diff --git a/drivers/dma/cnxk/cnxk_dmadev.c b/drivers/dma/cnxk/cnxk_dmadev.c
index eec6a897e2..35c2b79156 100644
--- a/drivers/dma/cnxk/cnxk_dmadev.c
+++ b/drivers/dma/cnxk/cnxk_dmadev.c
@@ -11,6 +11,7 @@
#include <rte_dmadev_pmd.h>
#include <rte_eal.h>
#include <rte_lcore.h>
+#include <rte_mbuf_pool_ops.h>
#include <rte_mempool.h>
#include <rte_pci.h>
@@ -70,10 +71,54 @@ cnxk_dmadev_vchan_free(struct cnxk_dpi_vf_s *dpivf,
uint16_t vchan)
return 0;
}
+static int
+cnxk_dmadev_chunk_pool_create(struct rte_dma_dev *dev)
+{
+ char pool_name[RTE_MEMPOOL_NAMESIZE];
+ struct cnxk_dpi_vf_s *dpivf = NULL;
+ uint64_t nb_chunks;
+ int rc;
+
+ dpivf = dev->fp_obj->dev_private;
+ /* Create chunk pool. */
+ snprintf(pool_name, sizeof(pool_name), "cnxk_dma_chunk_pool%d",
dev->data->dev_id);
+
+ nb_chunks = DPI_CMD_QUEUE_BUFS;
+ nb_chunks += (CNXK_DMA_POOL_MAX_CACHE_SZ * rte_lcore_count());
+ dpivf->chunk_pool =
+ rte_mempool_create_empty(pool_name, nb_chunks,
DPI_CMD_QUEUE_BUF_SIZE,
+ CNXK_DMA_POOL_MAX_CACHE_SZ, 0,
rte_socket_id(), 0);
+
+ if (dpivf->chunk_pool == NULL) {
+ plt_err("Unable to create chunkpool.");
+ return -ENOMEM;
+ }
+
+ rc = rte_mempool_set_ops_byname(dpivf->chunk_pool,
rte_mbuf_platform_mempool_ops(), NULL);
+ if (rc < 0) {
+ plt_err("Unable to set chunkpool ops");
+ goto free;
+ }
+
+ rc = rte_mempool_populate_default(dpivf->chunk_pool);
+ if (rc < 0) {
+ plt_err("Unable to set populate chunkpool.");
+ goto free;
+ }
+ dpivf->aura = roc_npa_aura_handle_to_aura(dpivf->chunk_pool->pool_id);
+
+ return 0;
+
+free:
+ rte_mempool_free(dpivf->chunk_pool);
+ return rc;
+}
+
static int
cnxk_dmadev_configure(struct rte_dma_dev *dev, const struct rte_dma_conf
*conf, uint32_t conf_sz)
{
struct cnxk_dpi_vf_s *dpivf = NULL;
+ void *chunk;
int rc = 0;
RTE_SET_USED(conf_sz);
@@ -92,12 +137,29 @@ cnxk_dmadev_configure(struct rte_dma_dev *dev, const
struct rte_dma_conf *conf,
if (dpivf->flag & CNXK_DPI_DEV_CONFIG)
return rc;
- rc = roc_dpi_configure(&dpivf->rdpi);
+ rc = cnxk_dmadev_chunk_pool_create(dev);
+ if (rc < 0) {
+ plt_err("DMA pool configure failed err = %d", rc);
+ goto done;
+ }
+
+ rc = rte_mempool_get(dpivf->chunk_pool, &chunk);
+ if (rc < 0) {
+ plt_err("DMA failed to get chunk pointer err = %d", rc);
+ rte_mempool_free(dpivf->chunk_pool);
+ goto done;
+ }
+
+ rc = roc_dpi_configure(&dpivf->rdpi, DPI_CMD_QUEUE_BUF_SIZE,
dpivf->aura, (uint64_t)chunk);
if (rc < 0) {
plt_err("DMA configure failed err = %d", rc);
+ rte_mempool_free(dpivf->chunk_pool);
goto done;
}
+ dpivf->chunk_base = chunk;
+ dpivf->chunk_head = 0;
+ dpivf->chunk_size_m1 = (DPI_CMD_QUEUE_BUF_SIZE >> 3) - 2;
dpivf->flag |= CNXK_DPI_DEV_CONFIG;
done:
@@ -335,7 +397,7 @@ cnxk_dmadev_close(struct rte_dma_dev *dev)
}
static inline int
-__dpi_queue_write(struct roc_dpi *dpi, uint64_t *cmds, int cmd_count)
+__dpi_queue_write(struct cnxk_dpi_vf_s *dpi, uint64_t *cmds, int cmd_count)
{
uint64_t *ptr = dpi->chunk_base;
@@ -346,31 +408,25 @@ __dpi_queue_write(struct roc_dpi *dpi, uint64_t *cmds,
int cmd_count)
* Normally there is plenty of room in the current buffer for the
* command
*/
- if (dpi->chunk_head + cmd_count < dpi->pool_size_m1) {
+ if (dpi->chunk_head + cmd_count < dpi->chunk_size_m1) {
ptr += dpi->chunk_head;
dpi->chunk_head += cmd_count;
while (cmd_count--)
*ptr++ = *cmds++;
} else {
+ uint64_t *new_buff = NULL;
int count;
- uint64_t *new_buff = dpi->chunk_next;
-
- dpi->chunk_next = (void
*)roc_npa_aura_op_alloc(dpi->aura_handle, 0);
- if (!dpi->chunk_next) {
- plt_dp_dbg("Failed to alloc next buffer from NPA");
- /* NPA failed to allocate a buffer. Restoring chunk_next
- * to its original address.
- */
- dpi->chunk_next = new_buff;
- return -ENOSPC;
+ if (rte_mempool_get(dpi->chunk_pool, (void **)&new_buff) < 0) {
+ plt_dpi_dbg("Failed to alloc next buffer from NPA");
+ return -ENOMEM;
}
/*
* Figure out how many cmd words will fit in this buffer.
* One location will be needed for the next buffer pointer.
*/
- count = dpi->pool_size_m1 - dpi->chunk_head;
+ count = dpi->chunk_size_m1 - dpi->chunk_head;
ptr += dpi->chunk_head;
cmd_count -= count;
while (count--)
@@ -395,19 +451,11 @@ __dpi_queue_write(struct roc_dpi *dpi, uint64_t *cmds,
int cmd_count)
*ptr++ = *cmds++;
/* queue index may be greater than pool size */
- if (dpi->chunk_head >= dpi->pool_size_m1) {
- new_buff = dpi->chunk_next;
- dpi->chunk_next = (void
*)roc_npa_aura_op_alloc(dpi->aura_handle, 0);
- if (!dpi->chunk_next) {
- plt_dp_dbg("Failed to alloc next buffer from
NPA");
-
- /* NPA failed to allocate a buffer. Restoring
chunk_next
- * to its original address.
- */
- dpi->chunk_next = new_buff;
- return -ENOSPC;
+ if (dpi->chunk_head == dpi->chunk_size_m1) {
+ if (rte_mempool_get(dpi->chunk_pool, (void
**)&new_buff) < 0) {
+ plt_dpi_dbg("Failed to alloc next buffer from
NPA");
+ return -ENOMEM;
}
-
/* Write next buffer address */
*ptr = (uint64_t)new_buff;
dpi->chunk_base = new_buff;
@@ -465,7 +513,7 @@ cnxk_dmadev_copy(void *dev_private, uint16_t vchan,
rte_iova_t src, rte_iova_t d
cmd[num_words++] = length;
cmd[num_words++] = lptr;
- rc = __dpi_queue_write(&dpivf->rdpi, cmd, num_words);
+ rc = __dpi_queue_write(dpivf, cmd, num_words);
if (unlikely(rc)) {
STRM_DEC(dpi_conf->c_desc, tail);
return rc;
@@ -537,7 +585,7 @@ cnxk_dmadev_copy_sg(void *dev_private, uint16_t vchan,
const struct rte_dma_sge
lptr++;
}
- rc = __dpi_queue_write(&dpivf->rdpi, cmd, num_words);
+ rc = __dpi_queue_write(dpivf, cmd, num_words);
if (unlikely(rc)) {
STRM_DEC(dpi_conf->c_desc, tail);
return rc;
@@ -593,7 +641,7 @@ cn10k_dmadev_copy(void *dev_private, uint16_t vchan,
rte_iova_t src, rte_iova_t
cmd[num_words++] = length;
cmd[num_words++] = lptr;
- rc = __dpi_queue_write(&dpivf->rdpi, cmd, num_words);
+ rc = __dpi_queue_write(dpivf, cmd, num_words);
if (unlikely(rc)) {
STRM_DEC(dpi_conf->c_desc, tail);
return rc;
@@ -656,7 +704,7 @@ cn10k_dmadev_copy_sg(void *dev_private, uint16_t vchan,
const struct rte_dma_sge
lptr++;
}
- rc = __dpi_queue_write(&dpivf->rdpi, cmd, num_words);
+ rc = __dpi_queue_write(dpivf, cmd, num_words);
if (unlikely(rc)) {
STRM_DEC(dpi_conf->c_desc, tail);
return rc;
diff --git a/drivers/dma/cnxk/cnxk_dmadev.h b/drivers/dma/cnxk/cnxk_dmadev.h
index 254e7fea20..65f12d844d 100644
--- a/drivers/dma/cnxk/cnxk_dmadev.h
+++ b/drivers/dma/cnxk/cnxk_dmadev.h
@@ -12,12 +12,15 @@
#define DPI_MAX_DESC 2048
#define DPI_MIN_DESC 2
#define MAX_VCHANS_PER_QUEUE 4
+#define DPI_CMD_QUEUE_BUF_SIZE 4096
+#define DPI_CMD_QUEUE_BUFS 1024
/* Set Completion data to 0xFF when request submitted,
* upon successful request completion engine reset to completion status
*/
#define DPI_REQ_CDATA 0xFF
+#define CNXK_DMA_POOL_MAX_CACHE_SZ (16)
#define CNXK_DPI_DEV_CONFIG (1ULL << 0)
#define CNXK_DPI_DEV_START (1ULL << 1)
@@ -45,8 +48,13 @@ struct cnxk_dpi_conf {
};
struct cnxk_dpi_vf_s {
- struct roc_dpi rdpi;
+ uint64_t *chunk_base;
+ uint16_t chunk_head;
+ uint16_t chunk_size_m1;
+ struct rte_mempool *chunk_pool;
struct cnxk_dpi_conf conf[MAX_VCHANS_PER_QUEUE];
+ struct roc_dpi rdpi;
+ uint32_t aura;
uint16_t num_vchans;
uint16_t flag;
} __plt_cache_aligned;
--
2.25.1
