From: Chengwen Feng <fengchengwen@huawei.com>
To: <thomas@monjalon.net>, <kevin.laatz@intel.com>,
<bruce.richardson@intel.com>
Cc: <dev@dpdk.org>
Subject: [PATCH v2 4/4] examples/dma: add force minimal copy size parameter
Date: Mon, 11 Apr 2022 20:14:59 +0800 [thread overview]
Message-ID: <20220411121459.23898-5-fengchengwen@huawei.com> (raw)
In-Reply-To: <20220411121459.23898-1-fengchengwen@huawei.com>
This patch adds force minimal copy size parameter
(-m/--force-min-copy-size), so when do copy by CPU or DMA, the real copy
size will be the maximum of mbuf's data_len and this parameter.
This parameter was designed to compare the performance between CPU copy
and DMA copy. User could send small packets with a high rate to drive
the performance test.
Signed-off-by: Chengwen Feng <fengchengwen@huawei.com>
---
examples/dma/dmafwd.c | 30 +++++++++++++++++++++++++++---
1 file changed, 27 insertions(+), 3 deletions(-)
diff --git a/examples/dma/dmafwd.c b/examples/dma/dmafwd.c
index 6b1b777cb8..3bd86b98c1 100644
--- a/examples/dma/dmafwd.c
+++ b/examples/dma/dmafwd.c
@@ -25,6 +25,7 @@
#define CMD_LINE_OPT_RING_SIZE "ring-size"
#define CMD_LINE_OPT_BATCH_SIZE "dma-batch-size"
#define CMD_LINE_OPT_FRAME_SIZE "max-frame-size"
+#define CMD_LINE_OPT_FORCE_COPY_SIZE "force-min-copy-size"
#define CMD_LINE_OPT_STATS_INTERVAL "stats-interval"
/* configurable number of RX/TX ring descriptors */
@@ -119,6 +120,7 @@ static volatile bool force_quit;
static uint32_t dma_batch_sz = MAX_PKT_BURST;
static uint32_t max_frame_size;
+static uint32_t force_min_copy_size;
/* ethernet addresses of ports */
static struct rte_ether_addr dma_ports_eth_addr[RTE_MAX_ETHPORTS];
@@ -208,7 +210,13 @@ print_stats(char *prgname)
"Rx Queues = %d, ", nb_queues);
status_strlen += snprintf(status_string + status_strlen,
sizeof(status_string) - status_strlen,
- "Ring Size = %d", ring_size);
+ "Ring Size = %d\n", ring_size);
+ status_strlen += snprintf(status_string + status_strlen,
+ sizeof(status_string) - status_strlen,
+ "Force Min Copy Size = %u Packet Data Room Size = %u",
+ force_min_copy_size,
+ rte_pktmbuf_data_room_size(dma_pktmbuf_pool) -
+ RTE_PKTMBUF_HEADROOM);
memset(&ts, 0, sizeof(struct total_statistics));
@@ -307,7 +315,8 @@ static inline void
pktmbuf_sw_copy(struct rte_mbuf *src, struct rte_mbuf *dst)
{
rte_memcpy(rte_pktmbuf_mtod(dst, char *),
- rte_pktmbuf_mtod(src, char *), src->data_len);
+ rte_pktmbuf_mtod(src, char *),
+ RTE_MAX(src->data_len, force_min_copy_size));
}
/* >8 End of perform packet copy there is a user-defined function. */
@@ -324,7 +333,9 @@ dma_enqueue_packets(struct rte_mbuf *pkts[], struct rte_mbuf *pkts_copy[],
ret = rte_dma_copy(dev_id, 0,
rte_pktmbuf_iova(pkts[i]),
rte_pktmbuf_iova(pkts_copy[i]),
- rte_pktmbuf_data_len(pkts[i]), 0);
+ RTE_MAX(rte_pktmbuf_data_len(pkts[i]),
+ force_min_copy_size),
+ 0);
if (ret < 0)
break;
@@ -576,6 +587,7 @@ dma_usage(const char *prgname)
printf("%s [EAL options] -- -p PORTMASK [-q NQ]\n"
" -b --dma-batch-size: number of requests per DMA batch\n"
" -f --max-frame-size: max frame size\n"
+ " -m --force-min-copy-size: force a minimum copy length, even for smaller packets\n"
" -p --portmask: hexadecimal bitmask of ports to configure\n"
" -q NQ: number of RX queues per port (default is 1)\n"
" --[no-]mac-updating: Enable or disable MAC addresses updating (enabled by default)\n"
@@ -621,6 +633,7 @@ dma_parse_args(int argc, char **argv, unsigned int nb_ports)
"b:" /* dma batch size */
"c:" /* copy type (sw|hw) */
"f:" /* max frame size */
+ "m:" /* force min copy size */
"p:" /* portmask */
"q:" /* number of RX queues per port */
"s:" /* ring size */
@@ -636,6 +649,7 @@ dma_parse_args(int argc, char **argv, unsigned int nb_ports)
{CMD_LINE_OPT_RING_SIZE, required_argument, NULL, 's'},
{CMD_LINE_OPT_BATCH_SIZE, required_argument, NULL, 'b'},
{CMD_LINE_OPT_FRAME_SIZE, required_argument, NULL, 'f'},
+ {CMD_LINE_OPT_FORCE_COPY_SIZE, required_argument, NULL, 'm'},
{CMD_LINE_OPT_STATS_INTERVAL, required_argument, NULL, 'i'},
{NULL, 0, 0, 0}
};
@@ -670,6 +684,10 @@ dma_parse_args(int argc, char **argv, unsigned int nb_ports)
}
break;
+ case 'm':
+ force_min_copy_size = atoi(optarg);
+ break;
+
/* portmask */
case 'p':
dma_enabled_port_mask = dma_parse_portmask(optarg);
@@ -1068,6 +1086,12 @@ main(int argc, char **argv)
rte_exit(EXIT_FAILURE, "Cannot init mbuf pool\n");
/* >8 End of allocates mempool to hold the mbufs. */
+ if (force_min_copy_size >
+ (uint32_t)(rte_pktmbuf_data_room_size(dma_pktmbuf_pool) -
+ RTE_PKTMBUF_HEADROOM))
+ rte_exit(EXIT_FAILURE,
+ "Force min copy size > packet mbuf size\n");
+
/* Initialize each port. 8< */
cfg.nb_ports = 0;
RTE_ETH_FOREACH_DEV(portid)
--
2.33.0
next prev parent reply other threads:[~2022-04-11 12:21 UTC|newest]
Thread overview: 27+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-04-11 2:56 [PATCH 0/4] bugfix and enhance features for DMA example Chengwen Feng
2022-04-11 2:56 ` [PATCH 1/4] examples/dma: fix MTU configuration Chengwen Feng
2022-04-11 2:56 ` [PATCH 2/4] examples/dma: fix Tx drop statistic is not collected Chengwen Feng
2022-04-11 2:56 ` [PATCH 3/4] examples/dma: support enqueue drop statistic Chengwen Feng
2022-04-11 2:56 ` [PATCH 4/4] examples/dma: add minimal copy size parameter Chengwen Feng
2022-04-11 9:27 ` Bruce Richardson
2022-04-11 12:23 ` fengchengwen
2022-04-11 12:14 ` [PATCH v2 0/4] bugfix and enhance features for DMA example Chengwen Feng
2022-04-11 12:14 ` [PATCH v2 1/4] examples/dma: fix MTU configuration Chengwen Feng
2022-04-11 12:14 ` [PATCH v2 2/4] examples/dma: fix Tx drop statistic is not collected Chengwen Feng
2022-04-13 14:57 ` Bruce Richardson
2022-04-11 12:14 ` [PATCH v2 3/4] examples/dma: support enqueue drop statistic Chengwen Feng
2022-04-13 15:01 ` Bruce Richardson
2022-04-16 6:19 ` fengchengwen
2022-04-19 8:45 ` Bruce Richardson
2022-04-19 12:09 ` fengchengwen
2022-04-24 3:55 ` fengchengwen
2022-04-11 12:14 ` Chengwen Feng [this message]
2022-04-13 15:04 ` [PATCH v2 4/4] examples/dma: add force minimal copy size parameter Bruce Richardson
2022-04-24 6:07 ` [PATCH v3 0/3] bugfix and enhance features for DMA example Chengwen Feng
2022-04-24 6:07 ` [PATCH v3 1/3] examples/dma: fix MTU configuration Chengwen Feng
2022-05-11 2:26 ` fengchengwen
2022-04-24 6:07 ` [PATCH v3 2/3] examples/dma: fix Tx drop statistic is not collected Chengwen Feng
2022-04-27 10:54 ` Kevin Laatz
2022-04-24 6:07 ` [PATCH v3 3/3] examples/dma: add force minimal copy size parameter Chengwen Feng
2022-04-27 10:54 ` Kevin Laatz
2022-06-06 21:35 ` [PATCH v3 0/3] bugfix and enhance features for DMA example Thomas Monjalon
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20220411121459.23898-5-fengchengwen@huawei.com \
--to=fengchengwen@huawei.com \
--cc=bruce.richardson@intel.com \
--cc=dev@dpdk.org \
--cc=kevin.laatz@intel.com \
--cc=thomas@monjalon.net \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).