DPDK patches and discussions
 help / color / mirror / Atom feed
From: Olivier Matz <olivier.matz@6wind.com>
To: Alexander Bechikov <asb.tyum@gmail.com>
Cc: dev@dpdk.org
Subject: Re: [dpdk-dev] [PATCH] mbuf: Improved error message. Added check if shared memory already allocated
Date: Fri, 29 Oct 2021 09:31:47 +0200	[thread overview]
Message-ID: <YXujY4CVgqG5C/15@platinum> (raw)
In-Reply-To: <20211027111118.395928-1-asb.tyum@gmail.com>

Hi Alexander,

Thanks for submitting this patch. Few comments below.

On Wed, Oct 27, 2021 at 07:11:18AM -0400, Alexander Bechikov wrote:
> mbuf: Improved error message. Added check if shared memory already allocated

I suggest another title:

  mbuf: fix dump of dynamic fields and flags

Indeed, it appears that the dump is currently broken. We can see it
in the unit tests:

  RTE>>mbuf_autotest
  ...
  MBUF: Failed to get mbuf dyn shared memory

As the function returns void, we failure is not detected by the test.

Can you add a small commit log?

I suppose this issue is introduced by d4902ed31c63 ("mbuf: check shared
memory before dumping dynamic space"). Do you confirm?

Can you please also add:

  Cc: stable@dpdk.org
  Fixes: d4902ed31c63 ("mbuf: check shared memory before dumping dynamic space")

> Signed-off-by: Alexander Bechikov <asb.tyum@gmail.com>
> ---
>  lib/mbuf/rte_mbuf_dyn.c | 5 +++--
>  1 file changed, 3 insertions(+), 2 deletions(-)
> 
> diff --git a/lib/mbuf/rte_mbuf_dyn.c b/lib/mbuf/rte_mbuf_dyn.c
> index db8e020665..a3bc9b66d2 100644
> --- a/lib/mbuf/rte_mbuf_dyn.c
> +++ b/lib/mbuf/rte_mbuf_dyn.c
> @@ -116,7 +116,8 @@ init_shared_mem(void)
>  		mz = rte_memzone_lookup(RTE_MBUF_DYN_MZNAME);
>  	}
>  	if (mz == NULL) {
> -		RTE_LOG(ERR, MBUF, "Failed to get mbuf dyn shared memory\n");
> +		RTE_LOG(ERR, MBUF, "Failed to get mbuf dyn shared memory: %s (%d)\n",
> +			rte_strerror(rte_errno), rte_errno);
>  		return -1;
>  	}
>  

I think rte_errno is not set when rte_memzone_lookup() returns NULL (this
happens when init_shared_mem() is called from a secondary process). You can
set it in this function to ENOENT, or just drop this change.


> @@ -531,7 +532,7 @@ void rte_mbuf_dyn_dump(FILE *out)
>  	size_t i;
>  
>  	rte_mcfg_tailq_write_lock();
> -	if (init_shared_mem() < 0) {
> +	if (shm == NULL && init_shared_mem() < 0) {
>  		rte_mcfg_tailq_write_unlock();
>  		return;
>  	}
> -- 
> 2.30.2
> 


Thanks,
Olivier

  reply	other threads:[~2021-10-29  7:31 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-10-27 11:11 Alexander Bechikov
2021-10-29  7:31 ` Olivier Matz [this message]
2021-10-29 19:35 ` [dpdk-dev] [PATCH v2 1/2] " Alexander Bechikov
2021-10-29 19:35   ` [dpdk-dev] [PATCH v2 2/2] mbuf: fix dump of dynamic fields and flags Alexander Bechikov
2021-11-24 12:57 ` [PATCH v3] " Olivier Matz
2021-11-24 14:08   ` Thomas Monjalon

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=YXujY4CVgqG5C/15@platinum \
    --to=olivier.matz@6wind.com \
    --cc=asb.tyum@gmail.com \
    --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 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).