All of lore.kernel.org
 help / color / mirror / Atom feed
From: bugzilla@dpdk.org
To: dev@dpdk.org
Subject: [DPDK/ethdev Bug 2092] net/pcap: filename truncation in secondary process
Date: Wed, 04 Nov 2026 08:04:11 +0000	[thread overview]
Message-ID: <bug-2092-3@https.bugs.dpdk.org/> (raw)

https://bugs.dpdk.org/show_bug.cgi?id=2092

            Bug ID: 2092
           Summary: net/pcap: filename truncation in secondary process
           Product: DPDK
           Version: 26.11
          Hardware: All
                OS: All
            Status: UNCONFIRMED
          Severity: normal
          Priority: Normal
         Component: ethdev
          Assignee: dev@dpdk.org
          Reporter: stephen@networkplumber.org
  Target Milestone: ---

Summary: net/pcap is truncating silently and breaks secondary
The paths need to be larger and variable length

Longer worded AI version.

Component: drivers/net/pcap
Version: 26.11-rc2 (a13c105)

Description:
The vdev bus multi-process sync passes only the device name to a
secondary, not its arguments. To work around this, the pcap PMD
stores a copy of the devargs string in dev_private, which is
shared with secondaries:

    struct pmd_internals {
        ...
        char devargs[ETH_PCAP_ARG_MAXLEN];   /* 64 */

    strlcpy((*internals)->devargs, rte_vdev_device_args(vdev),
            ETH_PCAP_ARG_MAXLEN);

When a secondary probes, it parses internal->devargs and opens
its own pcap handles from that copy. The limit applies to the
whole argument string, not to each file name, so it is easy to
hit with one long path or with several short arguments.

The primary works normally because it parses the full string.
Only secondaries see the truncated copy, so the problem stays
hidden until something like dpdk-dumpcap, dpdk-pdump or
dpdk-proc-info attaches.

Effects in the secondary:
 - rx_pcap/tx_pcap open fails on the truncated path, so the
   attach fails.
 - Arguments after the cut point (snaplen, infinite_rx, ...)
   are dropped or mis-parsed.
 - If the cut lands on a path that exists, the secondary opens
   the wrong file. For tx_pcap, pcap_dump_open() creates or
   truncates that file.

Steps to reproduce:
 dpdk-testpmd --vdev 'net_pcap0,rx_pcap=<path, devargs 64+ bytes>' \
     -- -i
 dpdk-dumpcap -i net_pcap0 -w /tmp/x.pcapng

Works when the same file is under a short path, e.g. /tmp/a.pcap.

Expected: the secondary sees the same arguments as the primary.
Actual: the secondary probes with a truncated devargs string.

Suggested fix:
Copy the devargs with rte_strdup() so the full string sits in
shared memory, keep a pointer to it in pmd_internals, and free
it on close. Or at least fail the primary probe when the
string will not fit, instead of truncating silently.

-- 
You are receiving this mail because:
You are the assignee for the bug.

                 reply	other threads:[~2026-11-04  8:04 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=bug-2092-3@https.bugs.dpdk.org/ \
    --to=bugzilla@dpdk.org \
    --cc=dev@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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.