From: Shiraz Saleem <shiraz.saleem@intel.com>
Date: Thu, 28 Mar 2019 11:49:47 -0500
Subject: RDMA/rdmavt: Use correct sizing on buffers holding page DMA addresses
Patch-mainline: v5.2-rc1
Git-commit: 629e6f9db6bf4c5702212dd77da534b838f14859
References: jsc#SLE-4925

The buffer that holds the page DMA addresses is sized off umem->nmap.
This can potentially cause out of bound accesses on the PBL array when
iterating the umem DMA-mapped SGL. This is because if umem pages are
combined, umem->nmap can be much lower than the number of system pages
in umem.

Use ib_umem_num_pages() to size this buffer.

Cc: Dennis Dalessandro <dennis.dalessandro@intel.com>
Cc: Mike Marciniszyn <mike.marciniszyn@intel.com>
Cc: Michael J. Ruhl <michael.j.ruhl@intel.com>
Signed-off-by: Shiraz Saleem <shiraz.saleem@intel.com>
Signed-off-by: Jason Gunthorpe <jgg@mellanox.com>
Acked-by: Thomas Bogendoerfer <tbogendoerfer@suse.de>
---
 drivers/infiniband/sw/rdmavt/mr.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

--- a/drivers/infiniband/sw/rdmavt/mr.c
+++ b/drivers/infiniband/sw/rdmavt/mr.c
@@ -393,7 +393,7 @@ struct ib_mr *rvt_reg_user_mr(struct ib_
 	if (IS_ERR(umem))
 		return (void *)umem;
 
-	n = umem->nmap;
+	n = ib_umem_num_pages(umem);
 
 	mr = __rvt_alloc_mr(n, pd);
 	if (IS_ERR(mr)) {