* [dpdk-users] Jumbo packets get segmented on RX
@ 2020-10-27 13:56 Moessbauer, Felix
0 siblings, 0 replies; only message in thread
From: Moessbauer, Felix @ 2020-10-27 13:56 UTC (permalink / raw)
To: users; +Cc: Schmidt, Adriaan, henning.schild
I'm trying to send and receive jumbo frame ethernet packets.
While sending works without any issues, I always receive fragmented / segmented packets.
This leads me to the following questions:
1. Is it possible to receive a frame without segmentation
2. How to properly size the mempool elements. Here, I tried the following (pkt_bytes is the maximum expected frame size):
buf_size = RTE_PKTMBUF_HEADROOM + pkt_bytes;
rte_mempool_create(mempool_name, NB_MBUF, buf_size,
MAX_PKT_BURST, sizeof(struct rte_pktmbuf_pool_private),
rte_pktmbuf_init, NULL, rte_socket_id(),
MEMPOOL_F_SP_PUT | MEMPOOL_F_SC_GET);
Some experiments with various buf_sizes showed, that if I make the buffer large enough (e.g. pkt_bytes * 2) it seems to work at least for some NICs (with IXGBE driver).
^ permalink raw reply [flat|nested] only message in thread
only message in thread, other threads:[~2020-10-27 13:56 UTC | newest]
Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-10-27 13:56 [dpdk-users] Jumbo packets get segmented on RX Moessbauer, Felix
DPDK usage discussions
This inbox may be cloned and mirrored by anyone:
git clone --mirror http://inbox.dpdk.org/users/0 users/git/0.git
# If you have public-inbox 1.1+ installed, you may
# initialize and index your mirror using the following commands:
public-inbox-init -V2 users users/ http://inbox.dpdk.org/users \
Example config snippet for mirrors.
Newsgroup available over NNTP:
AGPL code for this site: git clone https://public-inbox.org/public-inbox.git