IB/hfi1: Fix leak of rcvhdrtail_dummy_kvaddr

This buffer is currently allocated in hfi1_init():

	if (reinit)
		ret = init_after_reset(dd);
	else
		ret = loadtime_init(dd);
	if (ret)
		goto done;

	/* allocate dummy tail memory for all receive contexts */
	dd->rcvhdrtail_dummy_kvaddr = dma_alloc_coherent(&dd->pcidev->dev,
							 sizeof(u64),
							 &dd->rcvhdrtail_dummy_dma,
							 GFP_KERNEL);

	if (!dd->rcvhdrtail_dummy_kvaddr) {
		dd_dev_err(dd, "cannot allocate dummy tail memory\n");
		ret = -ENOMEM;
		goto done;
	}

The reinit triggered path will overwrite the old allocation and leak it.

Fix by moving the allocation to hfi1_alloc_devdata() and the deallocation
to hfi1_free_devdata().

Link: https://lore.kernel.org/r/20211129192008.101968.91302.stgit@awfm-01.cornelisnetworks.com
Cc: stable@vger.kernel.org
Fixes: 46b010d3ee ("staging/rdma/hfi1: Workaround to prevent corruption during packet delivery")
Signed-off-by: Mike Marciniszyn <mike.marciniszyn@cornelisnetworks.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@cornelisnetworks.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>

drivers/infiniband/hw/hfi1/init.c

@@ -875,18 +875,6 @@ int hfi1_init(struct hfi1_devdata *dd, int reinit)
 	if (ret)
 		goto done;
 
-	/* allocate dummy tail memory for all receive contexts */
-	dd->rcvhdrtail_dummy_kvaddr = dma_alloc_coherent(&dd->pcidev->dev,
-							 sizeof(u64),
-							 &dd->rcvhdrtail_dummy_dma,
-							 GFP_KERNEL);
-
-	if (!dd->rcvhdrtail_dummy_kvaddr) {
-		dd_dev_err(dd, "cannot allocate dummy tail memory\n");
-		ret = -ENOMEM;
-		goto done;
-	}
-
 	/* dd->rcd can be NULL if early initialization failed */
 	for (i = 0; dd->rcd && i < dd->first_dyn_alloc_ctxt; ++i) {
 		/*
@@ -1200,6 +1188,11 @@ void hfi1_free_devdata(struct hfi1_devdata *dd)
 	dd->tx_opstats    = NULL;
 	kfree(dd->comp_vect);
 	dd->comp_vect = NULL;
+	if (dd->rcvhdrtail_dummy_kvaddr)
+		dma_free_coherent(&dd->pcidev->dev, sizeof(u64),
+				  (void *)dd->rcvhdrtail_dummy_kvaddr,
+				  dd->rcvhdrtail_dummy_dma);
+	dd->rcvhdrtail_dummy_kvaddr = NULL;
 	sdma_clean(dd, dd->num_sdma);
 	rvt_dealloc_device(&dd->verbs_dev.rdi);
 }
@@ -1297,6 +1290,15 @@ static struct hfi1_devdata *hfi1_alloc_devdata(struct pci_dev *pdev,
 		goto bail;
 	}
 
+	/* allocate dummy tail memory for all receive contexts */
+	dd->rcvhdrtail_dummy_kvaddr =
+		dma_alloc_coherent(&dd->pcidev->dev, sizeof(u64),
+				   &dd->rcvhdrtail_dummy_dma, GFP_KERNEL);
+	if (!dd->rcvhdrtail_dummy_kvaddr) {
+		ret = -ENOMEM;
+		goto bail;
+	}
+
 	atomic_set(&dd->ipoib_rsm_usr_num, 0);
 	return dd;
 
@@ -1504,13 +1506,6 @@ static void cleanup_device_data(struct hfi1_devdata *dd)
 
 	free_credit_return(dd);
 
-	if (dd->rcvhdrtail_dummy_kvaddr) {
-		dma_free_coherent(&dd->pcidev->dev, sizeof(u64),
-				  (void *)dd->rcvhdrtail_dummy_kvaddr,
-				  dd->rcvhdrtail_dummy_dma);
-		dd->rcvhdrtail_dummy_kvaddr = NULL;
-	}
-
 	/*
 	 * Free any resources still in use (usually just kernel contexts)
 	 * at unload; we do for ctxtcnt, because that's what we allocate.