prov/shm: fix shm multi recv setopt segfault

shm uses the util srx and sets the minimum multi receive size through the srx. However, the srx code doesn't get initialized until the endpoint gets enabled. So if the application calls setopt (before FI_ENABLE), this will segfault because the srx has not been initialized. Instead, we need to save the multi recv size in the shm endpoint to be valid during setopt and then pass that into the util_srx creation to set the multi recv size Signed-off-by: Alexia Ingerson <[email protected]>
ofiwg · Dec 9, 2024 · 2c53734 · 2c53734
1 parent 7953ad7
commit 2c53734
Show file tree

Hide file tree

Showing 2 changed files with 5 additions and 3 deletions.
diff --git a/prov/shm/src/smr.h b/prov/shm/src/smr.h
@@ -229,6 +229,7 @@ struct smr_ep {
 	struct smr_tx_fs	*tx_fs;
 	struct dlist_entry	sar_list;
 	struct dlist_entry	ipc_cpy_pend_list;
+	size_t			min_multi_recv_size;
 
 	int			ep_idx;
 	enum ofi_shm_p2p_type	p2p_type;

diff --git a/prov/shm/src/smr_ep.c b/prov/shm/src/smr_ep.c
@@ -134,8 +134,7 @@ int smr_ep_setopt(fid_t fid, int level, int optname, const void *optval,
 		return -FI_ENOPROTOOPT;
 
 	if (optname == FI_OPT_MIN_MULTI_RECV) {
-		srx = smr_ep->srx->ep_fid.fid.context;
-		srx->min_multi_recv_size = *(size_t *)optval;
+		smr_ep->min_multi_recv_size = *(size_t *)optval;
 		return FI_SUCCESS;
 	}
 
@@ -1146,7 +1145,7 @@ static int smr_ep_ctrl(struct fid *fid, int command, void *arg)
 					      util_domain.domain_fid);
 			ret = util_ep_srx_context(&domain->util_domain,
 					ep->rx_size, SMR_IOV_LIMIT,
-					SMR_INJECT_SIZE, &smr_update,
+					ep->min_multi_recv_size, &smr_update,
 					&ep->util_ep.lock, &srx);
 			if (ret)
 				return ret;
@@ -1308,6 +1307,8 @@ int smr_endpoint(struct fid_domain *domain, struct fi_info *info,
 	dlist_init(&ep->sar_list);
 	dlist_init(&ep->ipc_cpy_pend_list);
 
+	ep->min_multi_recv_size = SMR_INJECT_SIZE;
+
 	ep->util_ep.ep_fid.fid.ops = &smr_ep_fi_ops;
 	ep->util_ep.ep_fid.ops = &smr_ep_ops;
 	ep->util_ep.ep_fid.cm = &smr_cm_ops;