From mboxrd@z Thu Jan 1 00:00:00 1970
Return-Path: <aburakov@ecsmtp.ir.intel.com>
Received: from mga11.intel.com (mga11.intel.com [192.55.52.93])
by dpdk.org (Postfix) with ESMTP id D89DE1E496
for <dev@dpdk.org>; Tue, 12 Jun 2018 11:46:20 +0200 (CEST)
X-Amp-Result: SKIPPED(no attachment in message)
X-Amp-File-Uploaded: False
Received: from orsmga001.jf.intel.com ([10.7.209.18])
by fmsmga102.fm.intel.com with ESMTP/TLS/DHE-RSA-AES256-GCM-SHA384;
12 Jun 2018 02:46:18 -0700
X-ExtLoop1: 1
X-IronPort-AV: E=Sophos;i="5.51,214,1526367600"; d="scan'208";a="63861001"
Received: from irvmail001.ir.intel.com ([163.33.26.43])
by orsmga001.jf.intel.com with ESMTP; 12 Jun 2018 02:46:17 -0700
Received: from sivswdev01.ir.intel.com (sivswdev01.ir.intel.com
[10.237.217.45])
by irvmail001.ir.intel.com (8.14.3/8.13.6/MailSET/Hub) with ESMTP id
w5C9kGeM015123 for <dev@dpdk.org>; Tue, 12 Jun 2018 10:46:16 +0100
Received: from sivswdev01.ir.intel.com (localhost [127.0.0.1])
by sivswdev01.ir.intel.com with ESMTP id w5C9kGdU021134
for <dev@dpdk.org>; Tue, 12 Jun 2018 10:46:16 +0100
Received: (from aburakov@localhost)
by sivswdev01.ir.intel.com with LOCAL id w5C9kGlH021130
for dev@dpdk.org; Tue, 12 Jun 2018 10:46:16 +0100
From: Anatoly Burakov <anatoly.burakov@intel.com>
To: dev@dpdk.org
Date: Tue, 12 Jun 2018 10:46:16 +0100
Message-Id: <3d7db240745dd4ebb6dc254530dc476588f5981b.1528796062.git.anatoly.burakov@intel.com>
X-Mailer: git-send-email 1.7.0.7
In-Reply-To: <a033d2f5109529a159d85c58d7b101b5a6dadacf.1528796062.git.anatoly.burakov@intel.com>
References: <a033d2f5109529a159d85c58d7b101b5a6dadacf.1528796062.git.anatoly.burakov@intel.com>
In-Reply-To: <a033d2f5109529a159d85c58d7b101b5a6dadacf.1528796062.git.anatoly.burakov@intel.com>
References: <a033d2f5109529a159d85c58d7b101b5a6dadacf.1528796062.git.anatoly.burakov@intel.com>
Subject: [dpdk-dev] [PATCH 3/3] mem: provide thread-unsafe memseg list walk
variant
X-BeenThere: dev@dpdk.org
X-Mailman-Version: 2.1.15
Precedence: list
List-Id: DPDK patches and discussions <dev.dpdk.org>
List-Unsubscribe: <https://dpdk.org/ml/options/dev>,
<mailto:dev-request@dpdk.org?subject=unsubscribe>
List-Archive: <http://dpdk.org/ml/archives/dev/>
List-Post: <mailto:dev@dpdk.org>
List-Help: <mailto:dev-request@dpdk.org?subject=help>
List-Subscribe: <https://dpdk.org/ml/listinfo/dev>,
<mailto:dev-request@dpdk.org?subject=subscribe>
X-List-Received-Date: Tue, 12 Jun 2018 09:46:21 -0000
Sometimes, user code needs to walk memseg list while being inside
a memory-related callback. Rather than making everyone copy around
the same iteration code and depending on DPDK internals, provide an
official way to do memseg_list_walk() inside callbacks.
Also, remove existing reimplementation from memalloc code and use
the new API instead.
Signed-off-by: Anatoly Burakov <anatoly.burakov@intel.com>
---
lib/librte_eal/common/eal_common_memory.c | 29 +++++++++--------
lib/librte_eal/common/include/rte_memory.h | 18 +++++++++++
lib/librte_eal/linuxapp/eal/eal_memalloc.c | 37 +++++-----------------
lib/librte_eal/rte_eal_version.map | 1 +
4 files changed, 43 insertions(+), 42 deletions(-)
diff --git a/lib/librte_eal/common/eal_common_memory.c b/lib/librte_eal/common/eal_common_memory.c
index afe0d5b57..6c4a8d40b 100644
--- a/lib/librte_eal/common/eal_common_memory.c
+++ b/lib/librte_eal/common/eal_common_memory.c
@@ -883,14 +883,11 @@ rte_memseg_walk(rte_memseg_walk_t func, void *arg)
}
int __rte_experimental
-rte_memseg_list_walk(rte_memseg_list_walk_t func, void *arg)
+rte_memseg_list_walk_thread_unsafe(rte_memseg_list_walk_t func, void *arg)
{
struct rte_mem_config *mcfg = rte_eal_get_configuration()->mem_config;
int i, ret = 0;
- /* do not allow allocations/frees/init while we iterate */
- rte_rwlock_read_lock(&mcfg->memory_hotplug_lock);
-
for (i = 0; i < RTE_MAX_MEMSEG_LISTS; i++) {
struct rte_memseg_list *msl = &mcfg->memsegs[i];
@@ -898,17 +895,23 @@ rte_memseg_list_walk(rte_memseg_list_walk_t func, void *arg)
continue;
ret = func(msl, arg);
- if (ret < 0) {
- ret = -1;
- goto out;
- }
- if (ret > 0) {
- ret = 1;
- goto out;
- }
+ if (ret)
+ return ret;
}
-out:
+ return 0;
+}
+
+int __rte_experimental
+rte_memseg_list_walk(rte_memseg_list_walk_t func, void *arg)
+{
+ struct rte_mem_config *mcfg = rte_eal_get_configuration()->mem_config;
+ int ret = 0;
+
+ /* do not allow allocations/frees/init while we iterate */
+ rte_rwlock_read_lock(&mcfg->memory_hotplug_lock);
+ ret = rte_memseg_list_walk_thread_unsafe(func, arg);
rte_rwlock_read_unlock(&mcfg->memory_hotplug_lock);
+
return ret;
}
diff --git a/lib/librte_eal/common/include/rte_memory.h b/lib/librte_eal/common/include/rte_memory.h
index c5a84c333..c4b7f4cff 100644
--- a/lib/librte_eal/common/include/rte_memory.h
+++ b/lib/librte_eal/common/include/rte_memory.h
@@ -299,6 +299,24 @@ rte_memseg_walk_thread_unsafe(rte_memseg_walk_t func, void *arg);
int __rte_experimental
rte_memseg_contig_walk_thread_unsafe(rte_memseg_contig_walk_t func, void *arg);
+/**
+ * Walk each allocated memseg list without performing any locking.
+ *
+ * @note This function does not perform any locking, and is only safe to call
+ * from within memory-related callback functions.
+ *
+ * @param func
+ * Iterator function
+ * @param arg
+ * Argument passed to iterator
+ * @return
+ * 0 if walked over the entire list
+ * 1 if stopped by the user
+ * -1 if user function reported error
+ */
+int __rte_experimental
+rte_memseg_list_walk_thread_unsafe(rte_memseg_list_walk_t func, void *arg);
+
/**
* Dump the physical memory layout to a file.
*
diff --git a/lib/librte_eal/linuxapp/eal/eal_memalloc.c b/lib/librte_eal/linuxapp/eal/eal_memalloc.c
index 8c11f98c9..1ebc4b571 100644
--- a/lib/librte_eal/linuxapp/eal/eal_memalloc.c
+++ b/lib/librte_eal/linuxapp/eal/eal_memalloc.c
@@ -171,32 +171,6 @@ get_file_size(int fd)
return st.st_size;
}
-/* we cannot use rte_memseg_list_walk() here because we will be holding a
- * write lock whenever we enter every function in this file, however copying
- * the same iteration code everywhere is not ideal as well. so, use a lockless
- * copy of memseg list walk here.
- */
-static int
-memseg_list_walk_thread_unsafe(rte_memseg_list_walk_t func, void *arg)
-{
- struct rte_mem_config *mcfg = rte_eal_get_configuration()->mem_config;
- int i, ret = 0;
-
- for (i = 0; i < RTE_MAX_MEMSEG_LISTS; i++) {
- struct rte_memseg_list *msl = &mcfg->memsegs[i];
-
- if (msl->base_va == NULL)
- continue;
-
- ret = func(msl, arg);
- if (ret < 0)
- return -1;
- if (ret > 0)
- return 1;
- }
- return 0;
-}
-
/* returns 1 on successful lock, 0 on unsuccessful lock, -1 on error */
static int lock(int fd, int type)
{
@@ -878,7 +852,8 @@ eal_memalloc_alloc_seg_bulk(struct rte_memseg **ms, int n_segs, size_t page_sz,
wa.socket = socket;
wa.segs_allocated = 0;
- ret = memseg_list_walk_thread_unsafe(alloc_seg_walk, &wa);
+ /* memalloc is locked, so it's safe to use thread-unsafe version */
+ ret = rte_memseg_list_walk_thread_unsafe(alloc_seg_walk, &wa);
if (ret == 0) {
RTE_LOG(ERR, EAL, "%s(): couldn't find suitable memseg_list\n",
__func__);
@@ -943,7 +918,10 @@ eal_memalloc_free_seg_bulk(struct rte_memseg **ms, int n_segs)
wa.ms = cur;
wa.hi = hi;
- walk_res = memseg_list_walk_thread_unsafe(free_seg_walk, &wa);
+ /* memalloc is locked, so it's safe to use thread-unsafe version
+ */
+ walk_res = rte_memseg_list_walk_thread_unsafe(free_seg_walk,
+ &wa);
if (walk_res == 1)
continue;
if (walk_res == 0)
@@ -1230,7 +1208,8 @@ eal_memalloc_sync_with_primary(void)
if (rte_eal_process_type() == RTE_PROC_PRIMARY)
return 0;
- if (memseg_list_walk_thread_unsafe(sync_walk, NULL))
+ /* memalloc is locked, so it's safe to call thread-unsafe version */
+ if (rte_memseg_list_walk_thread_unsafe(sync_walk, NULL))
return -1;
return 0;
}
diff --git a/lib/librte_eal/rte_eal_version.map b/lib/librte_eal/rte_eal_version.map
index 72d32fc39..592ffb867 100644
--- a/lib/librte_eal/rte_eal_version.map
+++ b/lib/librte_eal/rte_eal_version.map
@@ -288,6 +288,7 @@ EXPERIMENTAL {
rte_memseg_contig_walk;
rte_memseg_contig_walk_thread_unsafe;
rte_memseg_list_walk;
+ rte_memseg_list_walk_thread_unsafe;
rte_memseg_walk;
rte_memseg_walk_thread_unsafe;
rte_mp_action_register;
--
2.17.1