d/numam-spdk
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

nvme: take the lock when disconnecting qpairs.

Browse Source 
If we disconnect qpairs without taking the lock, we run the risk of
trying to double free qpair resources before they have been marked as
NULL.
For example, polling on one thread and calling
nvme_rdma_qpair_disconnect from one thread while doing an
nvme_ctrlr_reset on another thread. nvme_ctrlr_reset will call down to
nvme_rdma_qpair_disconnect on the same qpair and without any locking it
can result in trying to destroy the qpair resources multiple times.

Change-Id: I9eef6f2f92961ef8e3f8ece0e4a3d54f3434cff8
Signed-off-by: Seth Howell <seth.howell@intel.com>
Reviewed-on: https://review.gerrithub.io/c/spdk/spdk/+/472413
Tested-by: SPDK CI Jenkins <sys_sgci@intel.com>
Reviewed-by: Jim Harris <james.r.harris@intel.com>
Reviewed-by: Shuhei Matsumoto <shuhei.matsumoto.xt@hitachi.com>
This commit is contained in:
Seth Howell 2019-10-25 09:24:27 -07:00 committed by Jim Harris
parent 4121477d91
commit a4925ba744
3 changed files with 33 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

18
lib/nvme/nvme_ctrlr.c
View File

@ -425,6 +425,24 @@ out:
return rc;
}
/*
* This internal function will attempt to take the controller
* lock before calling disconnect on a controller qpair.
* Functions already holding the controller lock should
* call nvme_transport_ctrlr_disconnect_qpair directly.
*/
void
nvme_ctrlr_disconnect_qpair(struct spdk_nvme_qpair *qpair)
{
struct spdk_nvme_ctrlr *ctrlr = qpair->ctrlr;
assert(ctrlr != NULL);
nvme_robust_mutex_lock(&ctrlr->ctrlr_lock);
nvme_transport_ctrlr_disconnect_qpair(ctrlr, qpair);
nvme_robust_mutex_unlock(&ctrlr->ctrlr_lock);
}
int
spdk_nvme_ctrlr_free_io_qpair(struct spdk_nvme_qpair *qpair)
{

9
lib/nvme/nvme_internal.h
View File

@ -858,10 +858,11 @@ int nvme_ctrlr_get_vs(struct spdk_nvme_ctrlr *ctrlr, union spdk_nvme_vs_register
int nvme_ctrlr_get_cmbsz(struct spdk_nvme_ctrlr *ctrlr, union spdk_nvme_cmbsz_register *cmbsz);
void nvme_ctrlr_init_cap(struct spdk_nvme_ctrlr *ctrlr, const union spdk_nvme_cap_register *cap,
const union spdk_nvme_vs_register *vs);
int nvme_qpair_init(struct spdk_nvme_qpair *qpair, uint16_t id,
struct spdk_nvme_ctrlr *ctrlr,
enum spdk_nvme_qprio qprio,
uint32_t num_requests);
void nvme_ctrlr_disconnect_qpair(struct spdk_nvme_qpair *qpair);
int nvme_qpair_init(struct spdk_nvme_qpair *qpair, uint16_t id,
struct spdk_nvme_ctrlr *ctrlr,
enum spdk_nvme_qprio qprio,
uint32_t num_requests);
void nvme_qpair_deinit(struct spdk_nvme_qpair *qpair);
void nvme_qpair_complete_error_reqs(struct spdk_nvme_qpair *qpair);
int nvme_qpair_submit_request(struct spdk_nvme_qpair *qpair,

11
lib/nvme/nvme_rdma.c
View File

@ -1911,7 +1911,16 @@ nvme_rdma_qpair_process_completions(struct spdk_nvme_qpair *qpair,
return reaped;
fail:
nvme_rdma_qpair_disconnect(qpair);
/*
* Since admin queues take the ctrlr_lock before entering this function,
* we can call nvme_rdma_qpair_disconnect. For other qpairs we need
* to call the generic function which will take the lock for us.
*/
if (nvme_qpair_is_admin_queue(qpair)) {
nvme_rdma_qpair_disconnect(qpair);
} else {
nvme_ctrlr_disconnect_qpair(qpair);
}
return -ENXIO;
}