From: Cliff Burdick <shaklee3@gmail.com>
To: users <users@dpdk.org>
Subject: [dpdk-users] rte_eth_tx_burst not freeing mbufs
Date: Sat, 8 Apr 2017 14:06:38 -0700 [thread overview]
Message-ID: <CA+Gp1nYA=NzUmNbaBv96umpG9_S6Z3Bc+U9ULh=en9wsyDLxzA@mail.gmail.com> (raw)
Hi, I have an application where I'm sending a very low rate (1pps) to
certain ports just for gratuitous arp. I allocate a pktmbuf on a different
core, send it through a ring to my tx core thread, and
use rte_eth_tx_burst() to send that packet, and hopefully free. I've
noticed that even though the packets are sent, the driver is never freeing
the packets back to the mempool, which just throws errors when I try to
allocate another one. I've tried playing with tx_conf.tx_free_thresh, but
that didn't seem to have any effect. If I called rte_pktmbuf_free() right
after sending, it appears to free just fine, although that's not a
solution. Does anyone know what might cause this?
reply other threads:[~2017-04-08 21:06 UTC|newest]
Thread overview: [no followups] expand[flat|nested] mbox.gz Atom feed
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='CA+Gp1nYA=NzUmNbaBv96umpG9_S6Z3Bc+U9ULh=en9wsyDLxzA@mail.gmail.com' \
--to=shaklee3@gmail.com \
--cc=users@dpdk.org \
/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).