DPDK-dev Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: "Morten Brørup" <mb@smartsharesystems.com>
To: "Stephen Hemminger" <stephen@networkplumber.org>, <dev@dpdk.org>
Subject: RE: [PATCH 13/20] mbuf: allow NULL array in rte_pktmbuf_free_bulk
Date: Sat, 9 May 2026 10:47:53 +0200	[thread overview]
Message-ID: <98CBD80474FA8B44BF855DF32C47DC35F65869@smartserver.smartshare.dk> (raw)
In-Reply-To: <20260508203607.1003036-14-stephen@networkplumber.org>

> From: Stephen Hemminger [mailto:stephen@networkplumber.org]
> Sent: Friday, 8 May 2026 22.34
> 
> This allows callers to avoid NULL checks and just call
> rte_pktmbuf_free_bulk, similar to rte_pktmbuf_free.
> 
> Signed-off-by: Stephen Hemminger <stephen@networkplumber.org>

I disagree with this patch.

The parameter is an array of (pointers to) mbufs.
We already accept that the array can contain NULL pointers (no mbuf present).
This is extremely forgiving, considering that other fast path functions don't allow NULL pointers in arrays;
e.g. rte_eth_tx_burst(), rte_mempool_put_bulk().
But since it's a "free()" class of function, I don't object to it.

However, this patch changes the parameter type from "array" to "array or NULL (no array present)".
And I don't think we should change the parameter type; it should remain "array" only.

If there are any scenarios where a non-present array (NULL) is passed to the function, the count should be zero too.
And when the count is zero, the function does not dereference the array, so explicitly checking for NULL is superfluous.

We have a convention of not checking parameter validity in fast path functions.
And I consider it invalid parameters passing NULL with a non-zero count.

You might argue that this is a "free()" class of function, which warrants checking for NULL; but since it already accepts NULL with zero count, it is already covered.

We could change the function declaration for clarity:

void rte_pktmbuf_free_bulk(
		unsigned int count;
		struct rte_mbuf *mbufs[count], unsigned int count);

Or add a debug assertion at the start of the function:
RTE_ASSERT(mbufs != NULL || count == 0);


> ---
>  devtools/cocci/nullfree.cocci          | 5 ++++-
>  doc/guides/rel_notes/release_26_07.rst | 5 +++++
>  lib/mbuf/rte_mbuf.c                    | 3 +++
>  lib/mbuf/rte_mbuf.h                    | 1 +
>  4 files changed, 13 insertions(+), 1 deletion(-)
> 
> diff --git a/devtools/cocci/nullfree.cocci
> b/devtools/cocci/nullfree.cocci
> index e7417b69ff..78e30f730a 100644
> --- a/devtools/cocci/nullfree.cocci
> +++ b/devtools/cocci/nullfree.cocci
> @@ -4,7 +4,7 @@
>  // free(NULL) as a no-op.
>  //
>  @@
> -expression E;
> +expression E, N;
>  @@
>  (
>  - if (E != NULL) cmdline_free(E);
> @@ -79,6 +79,9 @@ expression E;
>  - if (E != NULL) rte_pktmbuf_free(E);
>  + rte_pktmbuf_free(E);
>  |
> +- if (E != NULL) rte_pktmbuf_free_bulk(E, N);
> ++ rte_pktmbuf_free_bulk(E, N);
> +|
>  - if (E != NULL) rte_rib_free(E);
>  + rte_rib_free(E);
>  |
> diff --git a/doc/guides/rel_notes/release_26_07.rst
> b/doc/guides/rel_notes/release_26_07.rst
> index f012d47a4b..6f2f7f849f 100644
> --- a/doc/guides/rel_notes/release_26_07.rst
> +++ b/doc/guides/rel_notes/release_26_07.rst
> @@ -63,6 +63,11 @@ New Features
>      ``rte_eal_init`` and the application is responsible for probing
> each device,
>    * ``--auto-probing`` enables the initial bus probing, which is the
> current default behavior.
> 
> +* **mbuf: rte_pktmbuf_free_bulk now accepts NULL array pointer.**
> +
> +  Calling ``rte_pktmbuf_free_bulk`` with a NULL ``mbufs`` argument  is
> a no-op,
> +  matching the behavior of ``rte_pktmbuf_free()`` and ``rte_free()``.
> +
> 
>  Removed Items
>  -------------
> diff --git a/lib/mbuf/rte_mbuf.c b/lib/mbuf/rte_mbuf.c
> index c2476e7704..8ec3bb53bb 100644
> --- a/lib/mbuf/rte_mbuf.c
> +++ b/lib/mbuf/rte_mbuf.c
> @@ -562,6 +562,9 @@ void rte_pktmbuf_free_bulk(struct rte_mbuf **mbufs,
> unsigned int count)
>  	struct rte_mbuf *m, *m_next,
> *pending[RTE_PKTMBUF_FREE_PENDING_SZ];
>  	unsigned int idx, nb_pending = 0;
> 
> +	if (mbufs == NULL)
> +		return;
> +
>  	rte_mbuf_history_mark_bulk(mbufs, count,
> RTE_MBUF_HISTORY_OP_LIB_FREE);
> 
>  	for (idx = 0; idx < count; idx++) {
> diff --git a/lib/mbuf/rte_mbuf.h b/lib/mbuf/rte_mbuf.h
> index e7c3bbadd4..e9f8932e80 100644
> --- a/lib/mbuf/rte_mbuf.h
> +++ b/lib/mbuf/rte_mbuf.h
> @@ -1534,6 +1534,7 @@ static inline void rte_pktmbuf_free(struct
> rte_mbuf *m)
>   *  @param mbufs
>   *    Array of pointers to packet mbufs.
>   *    The array may contain NULL pointers.
> + *    The array pointer is NULL function does nothing.
>   *  @param count
>   *    Array size.
>   */
> --
> 2.53.0


  reply	other threads:[~2026-05-09  8:47 UTC|newest]

Thread overview: 26+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-05-08 20:33 [PATCH 00/20] pktmbuf free bulk cleanups Stephen Hemminger
2026-05-08 20:33 ` [PATCH 01/20] devtools/cocci: add transform for rte_pktmbuf_free_bulk Stephen Hemminger
2026-05-08 20:33 ` [PATCH 02/20] eventdev: use rte_pktmbuf_free_bulk Stephen Hemminger
2026-05-08 20:33 ` [PATCH 03/20] gso: " Stephen Hemminger
2026-05-08 20:33 ` [PATCH 04/20] ip_frag: " Stephen Hemminger
2026-05-08 20:33 ` [PATCH 05/20] pipeline: " Stephen Hemminger
2026-05-08 20:33 ` [PATCH 06/20] port: " Stephen Hemminger
2026-05-08 20:33 ` [PATCH 07/20] net/af_xdp: " Stephen Hemminger
2026-05-08 20:33 ` [PATCH 08/20] net/cnxk: " Stephen Hemminger
2026-05-08 20:33 ` [PATCH 09/20] net/pfe: " Stephen Hemminger
2026-05-08 20:33 ` [PATCH 10/20] net/virtio: " Stephen Hemminger
2026-05-08 20:33 ` [PATCH 11/20] net/zxdh: " Stephen Hemminger
2026-05-08 20:33 ` [PATCH 12/20] app/compress-perf: " Stephen Hemminger
2026-05-08 20:33 ` [PATCH 13/20] mbuf: allow NULL array in rte_pktmbuf_free_bulk Stephen Hemminger
2026-05-09  8:47   ` Morten Brørup [this message]
2026-05-09 15:46     ` Stephen Hemminger
2026-05-10 12:31       ` Morten Brørup
2026-05-10 15:21         ` Stephen Hemminger
2026-05-08 20:33 ` [PATCH 14/20] net/zxdh: remove unnecessary null check Stephen Hemminger
2026-05-08 20:33 ` [PATCH 15/20] net/ice: " Stephen Hemminger
2026-05-08 20:33 ` [PATCH 16/20] net/bnxt: " Stephen Hemminger
2026-05-08 20:33 ` [PATCH 17/20] test: use rte_pktmbuf_free_bulk Stephen Hemminger
2026-05-08 20:33 ` [PATCH 18/20] app/test-dma-perf: remove unnecessary null check Stephen Hemminger
2026-05-11  1:16   ` fengchengwen
2026-05-08 20:33 ` [PATCH 19/20] app/test-compress-perf: " Stephen Hemminger
2026-05-08 20:33 ` [PATCH 20/20] examples: use rte_pktmbuf_free_bulk Stephen Hemminger

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=98CBD80474FA8B44BF855DF32C47DC35F65869@smartserver.smartshare.dk \
    --to=mb@smartsharesystems.com \
    --cc=dev@dpdk.org \
    --cc=stephen@networkplumber.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