Commit d1ad09db authored by Heiko Carstens's avatar Heiko Carstens Committed by James Bottomley
Browse files

[SCSI] zfcp: fix use after free bug.

zfcp_erp_strategy_check_fsfreq() checks if it is safe to access the
fsf_req associated with the erp_action that gets passed. To test if
it is safe it accesses the fsf_req in order to get its index into
the hash list. This is broken since the fsf_req might be freed already
and the read index has no meaning. It could lead to memory corruption.
Fix this by introducing a new zfcp_reqlist_find_safe() method which
just checks if addresses are equal. This is slower, but only gets
called in case of error recovery.
Signed-off-by: default avatarHeiko Carstens <>
Signed-off-by: default avatarMartin Schwidefsky <>
Signed-off-by: default avatarChristof Schmitt <>
Signed-off-by: default avatarMartin Peschke <>
Signed-off-by: default avatarJames Bottomley <>
parent bfd90dce
......@@ -1123,6 +1123,20 @@ zfcp_reqlist_find(struct zfcp_adapter *adapter, unsigned long req_id)
return NULL;
static inline struct zfcp_fsf_req *
zfcp_reqlist_find_safe(struct zfcp_adapter *adapter, struct zfcp_fsf_req *req)
struct zfcp_fsf_req *request;
unsigned int idx;
for (idx = 0; idx < REQUEST_LIST_SIZE; idx++) {
list_for_each_entry(request, &adapter->req_list[idx], list)
if (request == req)
return request;
return NULL;
* functions needed for reference/usage counting
......@@ -846,7 +846,8 @@ zfcp_erp_strategy_check_fsfreq(struct zfcp_erp_action *erp_action)
if (erp_action->fsf_req) {
/* take lock to ensure that request is not deleted meanwhile */
if (zfcp_reqlist_find(adapter, erp_action->fsf_req->req_id)) {
if (zfcp_reqlist_find_safe(adapter, erp_action->fsf_req) &&
erp_action->fsf_req->erp_action == erp_action) {
/* fsf_req still exists */
debug_text_event(adapter->erp_dbf, 3, "a_ca_req");
debug_event(adapter->erp_dbf, 3, &erp_action->fsf_req,
Supports Markdown
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment