linux-kernel - [PATCH 1/1] s390/virtio_ccw: fix dma

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-next>] [day] [month] [year] [list]

Message-ID: <20241007201030.204028-1-pasic@linux.ibm.com>
Date: Mon,  7 Oct 2024 22:10:30 +0200
From: Halil Pasic <pasic@...ux.ibm.com>
To: Cornelia Huck <cohuck@...hat.com>, Halil Pasic <pasic@...ux.ibm.com>,
        Eric Farman <farman@...ux.ibm.com>, Heiko Carstens <hca@...ux.ibm.com>,
        Vasily Gorbik <gor@...ux.ibm.com>,
        Alexander Gordeev <agordeev@...ux.ibm.com>,
        Christian Borntraeger <borntraeger@...ux.ibm.com>,
        Sven Schnelle <svens@...ux.ibm.com>,
        "Martin K. Petersen" <martin.petersen@...cle.com>,
        Robin Murphy <robin.murphy@....com>,
        Ulf Hansson <ulf.hansson@...aro.org>, linux-s390@...r.kernel.org,
        virtualization@...ts.linux.dev, kvm@...r.kernel.org,
        linux-kernel@...r.kernel.org
Cc: "Marc Hartmayer" <mhartmay@...ux.ibm.com>
Subject: [PATCH 1/1] s390/virtio_ccw: fix dma_parm pointer not set up

At least since commit 334304ac2bac ("dma-mapping: don't return errors
from dma_set_max_seg_size") setting up device.dma_parms is basically
mandated by the DMA API. As of now Channel (CCW) I/O in general does not
utilize the DMA API, except for virtio. For virtio-ccw however the
common virtio DMA infrastructure is such that most of the DMA stuff
hinges on the virtio parent device, which is a CCW device.

So lets set up the dma_parms pointer for the CCW parent device and hope
for the best!

Signed-off-by: Halil Pasic <pasic@...ux.ibm.com>
Fixes: 334304ac2bac ("dma-mapping: don't return errors from dma_set_max_seg_size")
Reported-by: "Marc Hartmayer" <mhartmay@...ux.ibm.com>
Closes: https://bugzilla.linux.ibm.com/show_bug.cgi?id=209131
Reviewed-by: Eric Farman <farman@...ux.ibm.com>
---

In the long run it may make sense to move dma_parms into struct
ccw_device, since layering-wise it is much cleaner. I decided
to put it in virtio_ccw_device because currently it is only used for
virtio.

---
 drivers/s390/virtio/virtio_ccw.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/drivers/s390/virtio/virtio_ccw.c b/drivers/s390/virtio/virtio_ccw.c
index 62eca9419ad7..21fa7ac849e5 100644
--- a/drivers/s390/virtio/virtio_ccw.c
+++ b/drivers/s390/virtio/virtio_ccw.c
@@ -58,6 +58,8 @@ struct virtio_ccw_device {
 	struct virtio_device vdev;
 	__u8 config[VIRTIO_CCW_CONFIG_SIZE];
 	struct ccw_device *cdev;
+	/* we make cdev->dev.dma_parms point to this */
+	struct device_dma_parameters dma_parms;
 	__u32 curr_io;
 	int err;
 	unsigned int revision; /* Transport revision */
@@ -1303,6 +1305,7 @@ static int virtio_ccw_offline(struct ccw_device *cdev)
 	unregister_virtio_device(&vcdev->vdev);
 	spin_lock_irqsave(get_ccwdev_lock(cdev), flags);
 	dev_set_drvdata(&cdev->dev, NULL);
+	cdev->dev.dma_parms = NULL;
 	spin_unlock_irqrestore(get_ccwdev_lock(cdev), flags);
 	return 0;
 }
@@ -1366,6 +1369,7 @@ static int virtio_ccw_online(struct ccw_device *cdev)
 	}
 	vcdev->vdev.dev.parent = &cdev->dev;
 	vcdev->cdev = cdev;
+	cdev->dev.dma_parms = &vcdev->dma_parms;
 	vcdev->dma_area = ccw_device_dma_zalloc(vcdev->cdev,
 						sizeof(*vcdev->dma_area),
 						&vcdev->dma_area_addr);

base-commit: 87d6aab2389e5ce0197d8257d5f8ee965a67c4cd
-- 
2.43.0