diff --git a/drivers/tee/optee/ffa_abi.c b/drivers/tee/optee/ffa_abi.c
index 633715b98625..d027fe7ce86f 100644
--- a/drivers/tee/optee/ffa_abi.c
+++ b/drivers/tee/optee/ffa_abi.c
@@ -198,7 +198,8 @@ static int to_msg_param_ffa_mem(struct optee_msg_param *mp,
 	if (shm) {
 		u64 shm_offs = p->u.memref.shm_offs;
 
-		mp->u.fmem.internal_offs = shm->offset;
+		mp->u.fmem.internal_offs = tee_shm_get_page_offset(shm) &
+					   (FFA_PAGE_SIZE - 1);
 
 		mp->u.fmem.offs_low = shm_offs;
 		mp->u.fmem.offs_high = shm_offs >> 32;
@@ -284,14 +285,26 @@ static int optee_ffa_shm_register(struct tee_context *ctx, struct tee_shm *shm,
 		.nattrs = 1,
 	};
 	struct sg_table sgt;
+	size_t ffa_offs;
 	int rc;
 
+	if (!num_pages)
+		return -EINVAL;
+
 	rc = optee_check_mem_type(start, num_pages);
 	if (rc)
 		return rc;
 
-	rc = sg_alloc_table_from_pages(&sgt, pages, num_pages, 0,
-				       num_pages * PAGE_SIZE, GFP_KERNEL);
+	/*
+	 * Start the FF-A descriptor at the 4 KiB page containing the shared
+	 * buffer, skipping unused leading 4 KiB pages when PAGE_SIZE is
+	 * larger. Same approach as optee_fill_pages_list() in the SMC ABI.
+	 * This leaves only the offset into that 4 KiB page for internal_offs.
+	 */
+	ffa_offs = round_down(tee_shm_get_page_offset(shm), FFA_PAGE_SIZE);
+	rc = sg_alloc_table_from_pages(&sgt, pages, num_pages, ffa_offs,
+				       num_pages * PAGE_SIZE - ffa_offs,
+				       GFP_KERNEL);
 	if (rc)
 		return rc;
 	args.sg = sgt.sgl;
@@ -458,7 +471,8 @@ static void handle_ffa_rpc_func_cmd_shm_alloc(struct tee_context *ctx,
 		.attr = OPTEE_MSG_ATTR_TYPE_FMEM_OUTPUT,
 		.u.fmem.size = tee_shm_get_size(shm),
 		.u.fmem.global_id = shm->sec_world_id,
-		.u.fmem.internal_offs = shm->offset,
+		.u.fmem.internal_offs = tee_shm_get_page_offset(shm) &
+					(FFA_PAGE_SIZE - 1),
 	};
 
 	arg->ret = TEEC_SUCCESS;
@@ -1123,14 +1137,6 @@ static int optee_ffa_probe(struct ffa_device *ffa_dev)
 
 	optee_set_dev_group(optee);
 
-	rc = tee_device_register(optee->teedev);
-	if (rc)
-		goto err_unreg_supp_teedev;
-
-	rc = tee_device_register(optee->supp_teedev);
-	if (rc)
-		goto err_unreg_supp_teedev;
-
 	rc = rhashtable_init(&optee->ffa.global_ids, &shm_rhash_params);
 	if (rc)
 		goto err_unreg_supp_teedev;
@@ -1159,6 +1165,14 @@ static int optee_ffa_probe(struct ffa_device *ffa_dev)
 	if (optee_ffa_protmem_pool_init(optee, sec_caps))
 		pr_info("Protected memory service not available\n");
 
+	rc = tee_device_register(optee->teedev);
+	if (rc)
+		goto err_unregister_devices;
+
+	rc = tee_device_register(optee->supp_teedev);
+	if (rc)
+		goto err_unregister_devices;
+
 	rc = optee_enumerate_devices(PTA_CMD_GET_DEVICES);
 	if (rc)
 		goto err_unregister_devices;
diff --git a/drivers/tee/optee/optee_msg.h b/drivers/tee/optee/optee_msg.h
index 7d9b12e71c03..6c3043f8da33 100644
--- a/drivers/tee/optee/optee_msg.h
+++ b/drivers/tee/optee/optee_msg.h
@@ -136,8 +136,8 @@ struct optee_msg_param_rmem {
  * struct optee_msg_param_fmem - FF-A memory reference parameter
  * @offs_low:		lower bits of offset into shared memory reference
  * @offs_high:		higher bits of offset into shared memory reference
- * @internal_offs:	internal offset into the first page of shared memory
- *			reference
+ * @internal_offs:	offset into the first 4 KiB page of the FF-A shared
+ *			memory region
  * @size:		size of the buffer
  * @global_id:		global identifier of the shared memory
  */
diff --git a/drivers/tee/optee/smc_abi.c b/drivers/tee/optee/smc_abi.c
index b8a2bdac3208..51624443359b 100644
--- a/drivers/tee/optee/smc_abi.c
+++ b/drivers/tee/optee/smc_abi.c
@@ -1849,14 +1849,6 @@ static int optee_probe(struct platform_device *pdev)
 
 	optee_set_dev_group(optee);
 
-	rc = tee_device_register(optee->teedev);
-	if (rc)
-		goto err_unreg_supp_teedev;
-
-	rc = tee_device_register(optee->supp_teedev);
-	if (rc)
-		goto err_unreg_supp_teedev;
-
 	optee_cq_init(&optee->call_queue, thread_count);
 	optee_supp_init(&optee->supp);
 	optee->smc.memremaped_shm = memremaped_shm;
@@ -1916,6 +1908,14 @@ static int optee_probe(struct platform_device *pdev)
 	if (optee->smc.sec_caps & OPTEE_SMC_SEC_CAP_DYNAMIC_SHM)
 		pr_info("dynamic shared memory is enabled\n");
 
+	rc = tee_device_register(optee->teedev);
+	if (rc)
+		goto err_disable_shm_cache;
+
+	rc = tee_device_register(optee->supp_teedev);
+	if (rc)
+		goto err_disable_shm_cache;
+
 	rc = optee_enumerate_devices(PTA_CMD_GET_DEVICES);
 	if (rc)
 		goto err_disable_shm_cache;
@@ -1942,7 +1942,6 @@ static int optee_probe(struct platform_device *pdev)
 	optee_shm_arg_cache_uninit(optee);
 	optee_supp_uninit(&optee->supp);
 	mutex_destroy(&optee->call_queue.mutex);
-err_unreg_supp_teedev:
 	tee_device_unregister(optee->supp_teedev);
 err_unreg_teedev:
 	tee_device_unregister(optee->teedev);