DPDK patches and discussions
 help / color / mirror / Atom feed
From: Thomas Monjalon <thomas@monjalon.net>
To: dev@dpdk.org
Cc: Matan Azrad <matan@mellanox.com>,
	Shahaf Shuler <shahafs@mellanox.com>,
	Viacheslav Ovsiienko <viacheslavo@mellanox.com>
Subject: [dpdk-dev] [PATCH 1/3] common/mlx5: split glue initialization
Date: Mon, 10 Feb 2020 11:32:14 +0100	[thread overview]
Message-ID: <20200210103216.1168439-2-thomas@monjalon.net> (raw)
In-Reply-To: <20200210103216.1168439-1-thomas@monjalon.net>

The function mlx5_glue_init was doing three things:
	- initialize logs
	- load glue library if in dlopen mode
	- initialize glue layer
They are split in three functions for clarity.

The config option RTE_IBVERBS_LINK_DLOPEN is not used anymore
outside of make and meson files. It is replaced with MLX5_GLUE,
which is defined in the same condition and is already used with dlopen.

Signed-off-by: Thomas Monjalon <thomas@monjalon.net>
---
 drivers/common/mlx5/mlx5_common.c | 77 +++++++++++++++++++------------
 1 file changed, 47 insertions(+), 30 deletions(-)

diff --git a/drivers/common/mlx5/mlx5_common.c b/drivers/common/mlx5/mlx5_common.c
index 610fb480b5..f96506c27a 100644
--- a/drivers/common/mlx5/mlx5_common.c
+++ b/drivers/common/mlx5/mlx5_common.c
@@ -160,7 +160,7 @@ mlx5_translate_port_name(const char *port_name_in,
 	return;
 }
 
-#ifdef RTE_IBVERBS_LINK_DLOPEN
+#ifdef MLX5_GLUE
 
 /**
  * Suffix RTE_EAL_PMD_PATH with "-glue".
@@ -202,36 +202,12 @@ mlx5_glue_path(char *buf, size_t size)
 		" re-configure DPDK");
 	return NULL;
 }
-#endif
 
-/**
- * Initialization routine for run-time dependency on rdma-core.
- */
-RTE_INIT_PRIO(mlx5_glue_init, CLASS)
+static int
+mlx5_glue_dlopen(void)
 {
 	void *handle = NULL;
 
-	/* Initialize common log type. */
-	mlx5_common_logtype = rte_log_register("pmd.common.mlx5");
-	if (mlx5_common_logtype >= 0)
-		rte_log_set_level(mlx5_common_logtype, RTE_LOG_NOTICE);
-	/*
-	 * RDMAV_HUGEPAGES_SAFE tells ibv_fork_init() we intend to use
-	 * huge pages. Calling ibv_fork_init() during init allows
-	 * applications to use fork() safely for purposes other than
-	 * using this PMD, which is not supported in forked processes.
-	 */
-	setenv("RDMAV_HUGEPAGES_SAFE", "1", 1);
-	/* Match the size of Rx completion entry to the size of a cacheline. */
-	if (RTE_CACHE_LINE_SIZE == 128)
-		setenv("MLX5_CQE_SIZE", "128", 0);
-	/*
-	 * MLX5_DEVICE_FATAL_CLEANUP tells ibv_destroy functions to
-	 * cleanup all the Verbs resources even when the device was removed.
-	 */
-	setenv("MLX5_DEVICE_FATAL_CLEANUP", "1", 1);
-	/* The glue initialization was done earlier by mlx5 common library. */
-#ifdef RTE_IBVERBS_LINK_DLOPEN
 	char glue_path[sizeof(RTE_EAL_PMD_PATH) - 1 + sizeof("-glue")];
 	const char *path[] = {
 		/*
@@ -301,7 +277,49 @@ RTE_INIT_PRIO(mlx5_glue_init, CLASS)
 		goto glue_error;
 	}
 	mlx5_glue = *sym;
-#endif /* RTE_IBVERBS_LINK_DLOPEN */
+	return 0;
+
+glue_error:
+	if (handle)
+		dlclose(handle);
+	return -1;
+}
+
+#endif
+
+RTE_INIT_PRIO(mlx5_log_init, LOG)
+{
+	mlx5_common_logtype = rte_log_register("pmd.common.mlx5");
+	if (mlx5_common_logtype >= 0)
+		rte_log_set_level(mlx5_common_logtype, RTE_LOG_NOTICE);
+}
+
+/**
+ * Initialization routine for run-time dependency on rdma-core.
+ */
+RTE_INIT_PRIO(mlx5_glue_init, CLASS)
+{
+	/*
+	 * RDMAV_HUGEPAGES_SAFE tells ibv_fork_init() we intend to use
+	 * huge pages. Calling ibv_fork_init() during init allows
+	 * applications to use fork() safely for purposes other than
+	 * using this PMD, which is not supported in forked processes.
+	 */
+	setenv("RDMAV_HUGEPAGES_SAFE", "1", 1);
+	/* Match the size of Rx completion entry to the size of a cacheline. */
+	if (RTE_CACHE_LINE_SIZE == 128)
+		setenv("MLX5_CQE_SIZE", "128", 0);
+	/*
+	 * MLX5_DEVICE_FATAL_CLEANUP tells ibv_destroy functions to
+	 * cleanup all the Verbs resources even when the device was removed.
+	 */
+	setenv("MLX5_DEVICE_FATAL_CLEANUP", "1", 1);
+
+#ifdef MLX5_GLUE
+	if (mlx5_glue_dlopen() != 0)
+		goto glue_error;
+#endif
+
 #ifdef RTE_LIBRTE_MLX5_DEBUG
 	/* Glue structure must not contain any NULL pointers. */
 	{
@@ -319,9 +337,8 @@ RTE_INIT_PRIO(mlx5_glue_init, CLASS)
 	}
 	mlx5_glue->fork_init();
 	return;
+
 glue_error:
-	if (handle)
-		dlclose(handle);
 	DRV_LOG(WARNING, "Cannot initialize MLX5 common due to missing"
 		" run-time dependency on rdma-core libraries (libibverbs,"
 		" libmlx5)");
-- 
2.25.0


  reply	other threads:[~2020-02-10 10:32 UTC|newest]

Thread overview: 15+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-02-10 10:32 [dpdk-dev] [PATCH 0/3] mlx: fix build with -fno-common (gcc 10) Thomas Monjalon
2020-02-10 10:32 ` Thomas Monjalon [this message]
2020-02-10 10:32 ` [dpdk-dev] [PATCH 2/3] common/mlx5: fix build with -fno-common Thomas Monjalon
2020-02-10 10:32 ` [dpdk-dev] [PATCH 3/3] net/mlx4: " Thomas Monjalon
2020-03-17  8:34 ` [dpdk-dev] [PATCH 0/3] mlx: fix build with -fno-common (gcc 10) Matan Azrad
2020-04-07  0:19   ` Thomas Monjalon
2020-04-07 23:06 ` [dpdk-dev] [PATCH v2 " Thomas Monjalon
2020-04-07 23:06   ` [dpdk-dev] [PATCH v2 1/3] common/mlx5: split glue initialization Thomas Monjalon
2020-04-07 23:06   ` [dpdk-dev] [PATCH v2 2/3] common/mlx5: fix build with -fno-common Thomas Monjalon
2020-04-07 23:06   ` [dpdk-dev] [PATCH v2 3/3] net/mlx4: " Thomas Monjalon
2020-04-08  0:08 ` [dpdk-dev] [PATCH v3 0/3] mlx: fix build with -fno-common (gcc 10) Thomas Monjalon
2020-04-08  0:08   ` [dpdk-dev] [PATCH v3 1/3] common/mlx5: split glue initialization Thomas Monjalon
2020-04-08  0:08   ` [dpdk-dev] [PATCH v3 2/3] common/mlx5: fix build with -fno-common Thomas Monjalon
2020-04-08  0:09   ` [dpdk-dev] [PATCH v3 3/3] net/mlx4: " Thomas Monjalon
2020-04-12 15:51   ` [dpdk-dev] [PATCH v3 0/3] mlx: fix build with -fno-common (gcc 10) Raslan Darawsheh

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=20200210103216.1168439-2-thomas@monjalon.net \
    --to=thomas@monjalon.net \
    --cc=dev@dpdk.org \
    --cc=matan@mellanox.com \
    --cc=shahafs@mellanox.com \
    --cc=viacheslavo@mellanox.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).