DPDK patches and discussions
 help / color / mirror / Atom feed
From: Tyler Retzlaff <roretzla@linux.microsoft.com>
To: Jianyue Wu <wujianyue000@163.com>
Cc: stephen@networkplumber.org, ferruh.yigit@amd.com, dev@dpdk.org,
	Jianyue Wu <jianyue.wu@nokia-sbell.com>
Subject: Re: [PATCH v2] eal/linux: enhanced error handling for affinity
Date: Fri, 26 Apr 2024 08:47:37 -0700	[thread overview]
Message-ID: <20240426154737.GB27008@linuxonhyperv3.guj3yctzbm1etfxqx2vob5hsef.xx.internal.cloudapp.net> (raw)
In-Reply-To: <20240425111130.8306-1-wujianyue000@163.com>

On Thu, Apr 25, 2024 at 07:11:30PM +0800, Jianyue Wu wrote:
> From: Jianyue Wu <jianyue.wu@nokia-sbell.com>
> 
> Improve the robustness of setting thread affinity in DPDK
> by adding detailed error logging.
> 
> Changes:
> 1. Check the return value of pthread_setaffinity_np() and log an error
> if the call fails.
> 2. Include the current thread name, the intended CPU set, and a detailed
> error message in the log.
> 
> Sample prints:
> EAL: Cannot set affinity for thread dpdk-test with cpus 0,
> ret: 22, errno: 0, error description: Success
> EAL: Cannot set affinity for thread dpdk-worker1 with cpus 1,
> ret: 22, errno: 0, error description: Success
> 
> Signed-off-by: Jianyue Wu <jianyue.wu@nokia-sbell.com>
> ---
>  lib/eal/common/eal_common_thread.c |  2 +-
>  lib/eal/common/eal_thread.h        |  2 +-
>  lib/eal/unix/rte_thread.c          | 27 +++++++++++++++++++++++++--
>  3 files changed, 27 insertions(+), 4 deletions(-)
> 
> diff --git a/lib/eal/common/eal_common_thread.c b/lib/eal/common/eal_common_thread.c
> index a53bc639ae..31a2fab2a7 100644
> --- a/lib/eal/common/eal_common_thread.c
> +++ b/lib/eal/common/eal_common_thread.c
> @@ -103,7 +103,7 @@ rte_thread_get_affinity(rte_cpuset_t *cpusetp)
>  }
>  
>  int
> -eal_thread_dump_affinity(rte_cpuset_t *cpuset, char *str, unsigned int size)
> +eal_thread_dump_affinity(const rte_cpuset_t *cpuset, char *str, unsigned int size)
>  {
>  	unsigned cpu;
>  	int ret;
> diff --git a/lib/eal/common/eal_thread.h b/lib/eal/common/eal_thread.h
> index 1c3c3442d3..85ab84baa5 100644
> --- a/lib/eal/common/eal_thread.h
> +++ b/lib/eal/common/eal_thread.h
> @@ -50,7 +50,7 @@ unsigned eal_cpu_socket_id(unsigned cpu_id);
>   *   0 for success, -1 if truncation happens.
>   */
>  int
> -eal_thread_dump_affinity(rte_cpuset_t *cpuset, char *str, unsigned int size);
> +eal_thread_dump_affinity(const rte_cpuset_t *cpuset, char *str, unsigned int size);

no objection to adding const

>  
>  /**
>   * Dump the current thread cpuset.
> diff --git a/lib/eal/unix/rte_thread.c b/lib/eal/unix/rte_thread.c
> index 1b4c73f58e..34ac0eabbf 100644
> --- a/lib/eal/unix/rte_thread.c
> +++ b/lib/eal/unix/rte_thread.c
> @@ -369,8 +369,31 @@ int
>  rte_thread_set_affinity_by_id(rte_thread_t thread_id,
>  		const rte_cpuset_t *cpuset)
>  {
> -	return pthread_setaffinity_np((pthread_t)thread_id.opaque_id,
> -		sizeof(*cpuset), cpuset);
> +	int ret;
> +#if defined(__linux__) && defined(_GNU_SOURCE)
> +	char cpus_str[RTE_CPU_AFFINITY_STR_LEN] = {'\0'};
> +	char thread_name[RTE_MAX_THREAD_NAME_LEN] = {'\0'};
> +	errno = 0;
> +#endif
> +
> +	ret = pthread_setaffinity_np((pthread_t)thread_id.opaque_id,
> +				sizeof(*cpuset), cpuset);
> +
> +#if defined(__linux__) && defined(_GNU_SOURCE)
> +	if (ret != 0) {
> +		if (pthread_getname_np((pthread_t)thread_id.opaque_id,
> +					thread_name, sizeof(thread_name)) != 0)
> +			EAL_LOG(ERR, "pthread_getname_np failed!");
> +		if (eal_thread_dump_affinity(cpuset, cpus_str, RTE_CPU_AFFINITY_STR_LEN) != 0)
> +			EAL_LOG(ERR, "eal_thread_dump_affinity failed!");
> +		EAL_LOG(ERR, "Cannot set affinity for thread %s with cpus %s, "
> +			"ret: %d, errno: %d, error description: %s",
> +			thread_name, cpus_str,
> +			ret, errno, strerror(errno));
> +	}
> +#endif
> +
> +	return ret;
>  }
>  
>  int
> -- 

i do not think introducing os specific behavior/logging to the EAL
is a good idea. logging although not formally part of the api surface
should present the same experience for all platforms. the EAL should
have a higher standard here.

> 2.34.1

  reply	other threads:[~2024-04-26 15:47 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-04-25 11:11 Jianyue Wu
2024-04-26 15:47 ` Tyler Retzlaff [this message]
2024-04-27  0:18   ` Stephen Hemminger
2024-04-28 12:26     ` Jianyue Wu
  -- strict thread matches above, loose matches on Subject: below --
2024-04-23  5:20 Jianyue Wu
2021-06-09 11:25 [dpdk-dev] Does memif support primary<->secondary process communication? Wu, Jianyue (NSB - CN/Hangzhou)
2021-06-09 12:07 ` Wu, Jianyue (NSB - CN/Hangzhou)
2024-04-22 13:23   ` [PATCH] eal/linux: enhanced error handling for affinity Jianyue Wu (NSB)
2024-04-24  5:45     ` [PATCH v2] " Jianyue Wu (NSB)

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=20240426154737.GB27008@linuxonhyperv3.guj3yctzbm1etfxqx2vob5hsef.xx.internal.cloudapp.net \
    --to=roretzla@linux.microsoft.com \
    --cc=dev@dpdk.org \
    --cc=ferruh.yigit@amd.com \
    --cc=jianyue.wu@nokia-sbell.com \
    --cc=stephen@networkplumber.org \
    --cc=wujianyue000@163.com \
    /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).