RE: [PATCH] examples/vhost: add option to control mbuf pool size

["Xia, Chenbo" <chenbo.xia@intel.com>]

Hi,

> -----Original Message-----
> From: Ma, WenwuX <wenwux.ma@intel.com>
> Sent: Thursday, February 17, 2022 11:17 PM
> To: maxime.coquelin@redhat.com; Xia, Chenbo <chenbo.xia@intel.com>
> Cc: dev@dpdk.org; Hu, Jiayu <jiayu.hu@intel.com>; Wang, Yinan
> <yinan.wang@intel.com>; He, Xingguang <xingguang.he@intel.com>; Ma, WenwuX
> <wenwux.ma@intel.com>
> Subject: [PATCH] examples/vhost: add option to control mbuf pool size
> 
> dpdk-vhost will fail to launch with a 40G i40e port because
> there are not enough mbufs. This patch adds a new option
> --total-num-mbufs, through which the user can set larger
> mbuf pool to avoid this problem.
> 
> Signed-off-by: Wenwu Ma <wenwux.ma@intel.com>
> ---
>  examples/vhost/main.c | 83 +++++++++++++++----------------------------
>  1 file changed, 29 insertions(+), 54 deletions(-)
> 
> diff --git a/examples/vhost/main.c b/examples/vhost/main.c
> index 3e784f5c6f..360f9f7f4d 100644
> --- a/examples/vhost/main.c
> +++ b/examples/vhost/main.c
> @@ -33,6 +33,8 @@
>  #define MAX_QUEUES 128
>  #endif
> 
> +#define NUM_MBUFS_DEFAULT 0x24000
> +
>  /* the maximum number of external ports supported */
>  #define MAX_SUP_PORTS 1
> 
> @@ -61,6 +63,9 @@
> 
>  #define DMA_RING_SIZE 4096
> 
> +/* number of mbufs in all pools - if specified on command-line. */
> +static int total_num_mbufs = NUM_MBUFS_DEFAULT;
> +
>  struct dma_for_vhost dma_bind[RTE_MAX_VHOST_DEVICE];
>  int16_t dmas_id[RTE_DMADEV_DEFAULT_MAX];
>  static int dma_count;
> @@ -609,7 +614,8 @@ us_vhost_usage(const char *prgname)
>  	"		--tso [0|1] disable/enable TCP segment offload.\n"
>  	"		--client register a vhost-user socket as client mode.\n"
>  	"		--dma-type register dma type for your vhost async driver.
> For example \"ioat\" for now.\n"
> -	"		--dmas register dma channel for specific vhost device.\n",
> +	"		--dmas register dma channel for specific vhost device.\n"
> +	"		--total-num-mbufs [0-N] set the number of mbufs to be
> allocated in mbuf pools.\n",

Let's tell the user about the default value here. And this patch can't apply
on current next-virtio tree now.

>  	       prgname);
>  }
> 
> @@ -638,6 +644,8 @@ enum {
>  	OPT_BUILTIN_NET_DRIVER_NUM,
>  #define OPT_DMAS                "dmas"
>  	OPT_DMAS_NUM,
> +#define OPT_NUM_MBUFS          "total-num-mbufs"

Align it with "dmas".

And as I said in the previous version, this fix is not good. But since we will move
all vhost test to vhost driver and give up this example soon. I will accept this
fix to help test all stuff.

Thanks,
Chenbo

> +	OPT_NUM_MBUFS_NUM,
>  };
> 
>  /*
> @@ -675,6 +683,8 @@ us_vhost_parse_args(int argc, char **argv)
>  				NULL, OPT_BUILTIN_NET_DRIVER_NUM},
>  		{OPT_DMAS, required_argument,
>  				NULL, OPT_DMAS_NUM},
> +		{OPT_NUM_MBUFS, required_argument,
> +				NULL, OPT_NUM_MBUFS_NUM},
>  		{NULL, 0, 0, 0},
>  	};
> 
> @@ -802,6 +812,19 @@ us_vhost_parse_args(int argc, char **argv)
>  			}
>  			break;
> 
> +		case OPT_NUM_MBUFS_NUM:
> +			ret = parse_num_opt(optarg, INT32_MAX);
> +			if (ret == -1) {
> +				RTE_LOG(INFO, VHOST_CONFIG,
> +					"Invalid argument for total-num-mbufs [0..N]\n");
> +				us_vhost_usage(prgname);
> +				return -1;
> +			}
> +
> +			if (total_num_mbufs < ret)
> +				total_num_mbufs = ret;
> +			break;
> +
>  		case OPT_CLIENT_NUM:
>  			client_mode = 1;
>  			break;
> @@ -1731,57 +1754,6 @@ sigint_handler(__rte_unused int signum)
>  	exit(0);
>  }
> 
> -/*
> - * While creating an mbuf pool, one key thing is to figure out how
> - * many mbuf entries is enough for our use. FYI, here are some
> - * guidelines:
> - *
> - * - Each rx queue would reserve @nr_rx_desc mbufs at queue setup stage
> - *
> - * - For each switch core (A CPU core does the packet switch), we need
> - *   also make some reservation for receiving the packets from virtio
> - *   Tx queue. How many is enough depends on the usage. It's normally
> - *   a simple calculation like following:
> - *
> - *       MAX_PKT_BURST * max packet size / mbuf size
> - *
> - *   So, we definitely need allocate more mbufs when TSO is enabled.
> - *
> - * - Similarly, for each switching core, we should serve @nr_rx_desc
> - *   mbufs for receiving the packets from physical NIC device.
> - *
> - * - We also need make sure, for each switch core, we have allocated
> - *   enough mbufs to fill up the mbuf cache.
> - */
> -static void
> -create_mbuf_pool(uint16_t nr_port, uint32_t nr_switch_core, uint32_t
> mbuf_size,
> -	uint32_t nr_queues, uint32_t nr_rx_desc, uint32_t nr_mbuf_cache)
> -{
> -	uint32_t nr_mbufs;
> -	uint32_t nr_mbufs_per_core;
> -	uint32_t mtu = 1500;
> -
> -	if (mergeable)
> -		mtu = 9000;
> -	if (enable_tso)
> -		mtu = 64 * 1024;
> -
> -	nr_mbufs_per_core  = (mtu + mbuf_size) * MAX_PKT_BURST /
> -			(mbuf_size - RTE_PKTMBUF_HEADROOM);
> -	nr_mbufs_per_core += nr_rx_desc;
> -	nr_mbufs_per_core  = RTE_MAX(nr_mbufs_per_core, nr_mbuf_cache);
> -
> -	nr_mbufs  = nr_queues * nr_rx_desc;
> -	nr_mbufs += nr_mbufs_per_core * nr_switch_core;
> -	nr_mbufs *= nr_port;
> -
> -	mbuf_pool = rte_pktmbuf_pool_create("MBUF_POOL", nr_mbufs,
> -					    nr_mbuf_cache, 0, mbuf_size,
> -					    rte_socket_id());
> -	if (mbuf_pool == NULL)
> -		rte_exit(EXIT_FAILURE, "Cannot create mbuf pool\n");
> -}
> -
>  static void
>  reset_dma(void)
>  {
> @@ -1861,8 +1833,11 @@ main(int argc, char *argv[])
>  	 * many queues here. We probably should only do allocation for
>  	 * those queues we are going to use.
>  	 */
> -	create_mbuf_pool(valid_num_ports, rte_lcore_count() - 1, MBUF_DATA_SIZE,
> -			 MAX_QUEUES, RTE_TEST_RX_DESC_DEFAULT, MBUF_CACHE_SIZE);
> +	mbuf_pool = rte_pktmbuf_pool_create("MBUF_POOL", total_num_mbufs,
> +					    MBUF_CACHE_SIZE, 0, MBUF_DATA_SIZE,
> +					    rte_socket_id());
> +	if (mbuf_pool == NULL)
> +		rte_exit(EXIT_FAILURE, "Cannot create mbuf pool\n");
> 
>  	if (vm2vm_mode == VM2VM_HARDWARE) {
>  		/* Enable VT loop back to let L2 switch to do it. */
> --
> 2.25.1