DPDK patches and discussions
 help / color / mirror / Atom feed
* [dpdk-dev] [PATCH] mem: fix rte_malloc(SOCKET_ID_ANY), try to allocate on other nodes
@ 2013-07-26 14:55 Thomas Monjalon
  2013-07-26 15:08 ` Thomas Monjalon
  0 siblings, 1 reply; 2+ messages in thread
From: Thomas Monjalon @ 2013-07-26 14:55 UTC (permalink / raw)
  To: dev

From: Olivier Matz <olivier.matz@6wind.com>

Before this patch, rte_malloc(SOCKET_ID_ANY) was equivalent to
rte_malloc(this_socket). If the user specifies SOCKET_ID_ANY, it means that
memory can be allocated on any socket. So fix the behavior of rte_malloc() in
order to do that. The current CPU socket is still the default, but if it fails,
other sockets are tested.

Signed-off-by: Olivier Matz <olivier.matz@6wind.com>
Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>
---
 lib/librte_malloc/rte_malloc.c |   28 ++++++++++++++++++++++++----
 1 file changed, 24 insertions(+), 4 deletions(-)

diff --git a/lib/librte_malloc/rte_malloc.c b/lib/librte_malloc/rte_malloc.c
index a70a6b0..3a9edb1 100644
--- a/lib/librte_malloc/rte_malloc.c
+++ b/lib/librte_malloc/rte_malloc.c
@@ -69,23 +69,43 @@ void rte_free(void *addr)
  * Allocate memory on specified heap.
  */
 void *
-rte_malloc_socket(const char *type, size_t size, unsigned align, int socket)
+rte_malloc_socket(const char *type, size_t size, unsigned align, int socket_arg)
 {
 	struct rte_mem_config *mcfg = rte_eal_get_configuration()->mem_config;
+	int socket, i;
+	void *ret;
 
 	/* return NULL if size is 0 or alignment is not power-of-2 */
 	if (size == 0 || !rte_is_power_of_2(align))
 		return NULL;
 
-	if (socket == SOCKET_ID_ANY)
+	if (socket_arg == SOCKET_ID_ANY)
 		socket = malloc_get_numa_socket();
+	else
+		socket = socket_arg;
 
 	/* Check socket parameter */
 	if (socket >= RTE_MAX_NUMA_NODES)
 		return NULL;
 
-	return malloc_heap_alloc(&mcfg->malloc_heaps[socket], type,
-			size, align == 0 ? 1 : align);
+	ret = malloc_heap_alloc(&mcfg->malloc_heaps[socket], type,
+				size, align == 0 ? 1 : align);
+	if (ret != NULL || socket_arg != SOCKET_ID_ANY)
+		return ret;
+
+	/* try other heaps */
+	for (i = 0; i < RTE_MAX_NUMA_NODES; i++) {
+		/* we already tried this one */
+		if (i == socket)
+			continue;
+
+		ret = malloc_heap_alloc(&mcfg->malloc_heaps[i], type,
+					size, align == 0 ? 1 : align);
+		if (ret != NULL)
+			return ret;
+	}
+
+	return NULL;
 }
 
 /*
-- 
1.7.10.4

^ permalink raw reply	[flat|nested] 2+ messages in thread

* Re: [dpdk-dev] [PATCH] mem: fix rte_malloc(SOCKET_ID_ANY), try to allocate on other nodes
  2013-07-26 14:55 [dpdk-dev] [PATCH] mem: fix rte_malloc(SOCKET_ID_ANY), try to allocate on other nodes Thomas Monjalon
@ 2013-07-26 15:08 ` Thomas Monjalon
  0 siblings, 0 replies; 2+ messages in thread
From: Thomas Monjalon @ 2013-07-26 15:08 UTC (permalink / raw)
  To: dev

26/07/2013 16:55, Thomas Monjalon :
> From: Olivier Matz <olivier.matz@6wind.com>
> 
> Before this patch, rte_malloc(SOCKET_ID_ANY) was equivalent to
> rte_malloc(this_socket). If the user specifies SOCKET_ID_ANY, it means that
> memory can be allocated on any socket. So fix the behavior of rte_malloc()
> in order to do that. The current CPU socket is still the default, but if
> it fails, other sockets are tested.
> 
> Signed-off-by: Olivier Matz <olivier.matz@6wind.com>
> Acked-by: Thomas Monjalon <thomas.monjalon@6wind.com>

pushed

-- 
Thomas

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2013-07-26 15:08 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2013-07-26 14:55 [dpdk-dev] [PATCH] mem: fix rte_malloc(SOCKET_ID_ANY), try to allocate on other nodes Thomas Monjalon
2013-07-26 15:08 ` Thomas Monjalon

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).