linux-kernel - Re: [PATCH] misc/genwqe: ensure zero initialization

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-Id: <alpine.LFD.2.20.1609231155030.1897@schleppi>
Date:   Fri, 23 Sep 2016 12:06:31 +0200 (CEST)
From:   Sebastian Ott <sebott@...ux.vnet.ibm.com>
To:     Frank Haverkamp <haver@...ux.vnet.ibm.com>
cc:     Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH] misc/genwqe: ensure zero initialization

On Thu, 22 Sep 2016, Frank Haverkamp wrote:
> > On 12.09.2016, at 20:09, Sebastian Ott <sebott@...ux.vnet.ibm.com> wrote:
> > Genwqe uses dma_alloc_coherent and depends on zero initialized memory. On
> > one occasion it ueses an explicit memset on others it uses un-initialized
> > memory.
> > 
> > This bug was covered because some archs actually return zero initialized
> > memory when using dma_alloc_coherent but this is by no means guaranteed.
> > Simply switch to dma_zalloc_coherent.
> 
> Thanks for figuring this out. I hope you found all spots where this misassumption is done.

Genwqe only uses the dma alloc API via a wrapper
(__genwqe_alloc_consistent) and I changed it there.

> I think this should be fine.
> 
> Signed-off-by: Frank Haverkamp <haver@...ux.vnet.ibm.com>

I changed that into an Ack. Seems more fitting here. Also I fixed a typo
in the patch description.

Greg, would you please take this one via your drivers/misc tree.

Thanks,
Sebastian

-------->8
misc/genwqe: ensure zero initialization

Genwqe uses dma_alloc_coherent and depends on zero initialized memory. On
one occasion it uses an explicit memset on others it uses un-initialized
memory.

This bug was covered because some archs actually return zero initialized
memory when using dma_alloc_coherent but this is by no means guaranteed.
Simply switch to dma_zalloc_coherent.

Signed-off-by: Sebastian Ott <sebott@...ux.vnet.ibm.com>
Acked-by: Frank Haverkamp <haver@...ux.vnet.ibm.com>
---
 drivers/misc/genwqe/card_ddcb.c  | 2 --
 drivers/misc/genwqe/card_utils.c | 4 ++--
 2 files changed, 2 insertions(+), 4 deletions(-)

diff --git a/drivers/misc/genwqe/card_ddcb.c b/drivers/misc/genwqe/card_ddcb.c
index 353ee0c..ddfeefe 100644
--- a/drivers/misc/genwqe/card_ddcb.c
+++ b/drivers/misc/genwqe/card_ddcb.c
@@ -1048,8 +1048,6 @@ static int setup_ddcb_queue(struct genwqe_dev *cd, struct ddcb_queue *queue)
 			"[%s] **err: could not allocate DDCB **\n", __func__);
 		return -ENOMEM;
 	}
-	memset(queue->ddcb_vaddr, 0, queue_size);
-
 	queue->ddcb_req = kzalloc(sizeof(struct ddcb_requ *) *
 				  queue->ddcb_max, GFP_KERNEL);
 	if (!queue->ddcb_req) {
diff --git a/drivers/misc/genwqe/card_utils.c b/drivers/misc/genwqe/card_utils.c
index 222367c..8a679ec 100644
--- a/drivers/misc/genwqe/card_utils.c
+++ b/drivers/misc/genwqe/card_utils.c
@@ -220,8 +220,8 @@ void *__genwqe_alloc_consistent(struct genwqe_dev *cd, size_t size,
 	if (get_order(size) > MAX_ORDER)
 		return NULL;
 
-	return dma_alloc_coherent(&cd->pci_dev->dev, size, dma_handle,
-				  GFP_KERNEL);
+	return dma_zalloc_coherent(&cd->pci_dev->dev, size, dma_handle,
+				   GFP_KERNEL);
 }
 
 void __genwqe_free_consistent(struct genwqe_dev *cd, size_t size,
-- 
2.5.5