Skip to contentSkip to conversation

opensm/osm_ucast_lash: fix use after free bug

PATCH In review Sasha Khapyorsky 3 participants 3 messages Sep 25, 2009 5 messages not in the archive InfiniBand

lore
Share & tools
lore

A message that is not in the archive (sent off-list or not yet received): [email protected]

A message that is not in the archive (sent off-list or not yet received): [email protected]

A message that is not in the archive (sent off-list or not yet received): 20090922185014.GF24398@me

A message that is not in the archive (sent off-list or not yet received): [email protected]

A message that is not in the archive (sent off-list or not yet received): 20090925135256.GE26931@me

#5 lore raw

PATCH 1/1 opensm/osm_ucast_lash: fix use after free bug

When LASH runs its switch structures cleanup OpenSM can rediscover a subnet and 'p_sw' pointer may refer already freed memory, so don't touch it, just free our own stuff. (Note also that for valids OpenSM switches objects' 'priv' pointers are cleared on lash_cleanup()).

---
 opensm/opensm/osm_ucast_lash.c |    5 +----
 1 files changed, 1 insertions(+), 4 deletions(-)
opensm/opensm/osm_ucast_lash.c +1 -4
index dbc6bcc..3c424cb 100644
@@ -628,8 +628,7 @@ static switch_t *switch_create(lash_t * p_lash, unsigned id, osm_switch_t * p_sw
 	}
 
 	sw->p_sw = p_sw;
-	if (p_sw)
-		p_sw->priv = sw;
+	p_sw->priv = sw;
 
 	if (osm_mesh_node_create(p_lash, sw)) {
 		free(sw->dij_channels);
@@ -644,8 +643,6 @@ static void switch_delete(lash_t *p_lash, switch_t * sw)
 {
 	if (sw->dij_channels)
 		free(sw->dij_channels);
-	if (sw->p_sw)
-		sw->p_sw->priv = NULL;
 	free(sw);
 }
 
Signature · 1 line
1.6.5.rc1
#6 lore raw

Hi Sasha,

Quoted 4 lines · Sasha Khapyorsky

When LASH runs its switch structures cleanup OpenSM can rediscover a subnet and 'p_sw' pointer may refer already freed memory, so don't touch it, just free our own stuff. (Note also that for valids OpenSM switches objects' 'priv' pointers are cleared on lash_cleanup()).

This worked for me on the test case that generated my earlier report.

Thanks -- Jim

Quoted 21 lines
---
 opensm/opensm/osm_ucast_lash.c |    5 +----
 1 files changed, 1 insertions(+), 4 deletions(-)
opensm/opensm/osm_ucast_lash.c +1 -2
index dbc6bcc..3c424cb 100644
@@ -628,8 +628,7 @@ static switch_t *switch_create(lash_t * p_lash, unsigned id, osm_switch_t * p_sw
 	}
 
 	sw->p_sw = p_sw;
-	if (p_sw)
-		p_sw->priv = sw;
+	p_sw->priv = sw;
 
 	if (osm_mesh_node_create(p_lash, sw)) {
 		free(sw->dij_channels);

opensm/opensm/osm_ucast_lash.c static void switch_delete(lash_t *p_lash, switch_t * sw) · line 648

Quoted 8 lines
opensm/opensm/osm_ucast_lash.c +0 -2
@@ -644,8 +643,6 @@ static void switch_delete(lash_t *p_lash, switch_t * sw)
 {
 	if (sw->dij_channels)
 		free(sw->dij_channels);
-	if (sw->p_sw)
-		sw->p_sw->priv = NULL;
 	free(sw);
 }
 

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html

Tested-by on v1 1/1
#7 lore raw
In reply to Sasha Khapyorsky
Quoted 7 lines · Sasha Khapyorsky

When LASH runs its switch structures cleanup OpenSM can rediscover a subnet and 'p_sw' pointer may refer already freed memory, so don't touch it, just free our own stuff. (Note also that for valids OpenSM switches objects' 'priv' pointers are cleared on lash_cleanup()).

Quoted 29 lines
---
 opensm/opensm/osm_ucast_lash.c |    5 +----
 1 files changed, 1 insertions(+), 4 deletions(-)
opensm/opensm/osm_ucast_lash.c +0 -0
index dbc6bcc..3c424cb 100644
@@ -628,8 +628,7 @@ static switch_t *switch_create(lash_t * p_lash, unsigned

id, osm_switch_t * p_sw

opensm/opensm/osm_ucast_lash.c +1 -5
 	}
 
 	sw->p_sw = p_sw;
-	if (p_sw)
-		p_sw->priv = sw;
+	p_sw->priv = sw;
 
 	if (osm_mesh_node_create(p_lash, sw)) {
 		free(sw->dij_channels);
@@ -644,8 +643,6 @@ static void switch_delete(lash_t *p_lash, switch_t * sw)
 {
 	if (sw->dij_channels)
 		free(sw->dij_channels);
-	if (sw->p_sw)
-		sw->p_sw->priv = NULL;
 	free(sw);
 }
 
--
Quoted 3 lines

1.6.5.rc1

--
To unsubscribe from this list: send the line "unsubscribe linux-rdma" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html