[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <88c3e5b4-8ac2-42fc-acf0-2aec2406a7ef@intel.com>
Date: Fri, 5 Dec 2025 13:31:04 +0100
From: Michal Wajdeczko <michal.wajdeczko@...el.com>
To: Dan Carpenter <dan.carpenter@...aro.org>,
MichaĆ Winiarski <michal.winiarski@...el.com>
CC: Jason Gunthorpe <jgg@...pe.ca>, Yishai Hadas <yishaih@...dia.com>,
"Shameer Kolothum" <skolothumtho@...dia.com>, Kevin Tian
<kevin.tian@...el.com>, "Alex Williamson" <alex@...zbot.org>, Rodrigo Vivi
<rodrigo.vivi@...el.com>, <kvm@...r.kernel.org>,
<intel-xe@...ts.freedesktop.org>, <linux-kernel@...r.kernel.org>,
<kernel-janitors@...r.kernel.org>
Subject: Re: [PATCH next] vfio/xe: Fix use after free in
xe_vfio_pci_alloc_file()
On 12/5/2025 12:39 PM, Dan Carpenter wrote:
> This code frees "migf" and then dereferences it on the next line to get
> the error code. Preserve the error code before freeing the pointer.
>
> Fixes: 2e38c50ae492 ("vfio/xe: Add device specific vfio_pci driver variant for Intel graphics")
> Signed-off-by: Dan Carpenter <dan.carpenter@...aro.org>
Reviewed-by: Michal Wajdeczko <michal.wajdeczko@...el.com>
> ---
> drivers/vfio/pci/xe/main.c | 4 +++-
> 1 file changed, 3 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/vfio/pci/xe/main.c b/drivers/vfio/pci/xe/main.c
> index 0156b53c678b..8e1595e00e18 100644
> --- a/drivers/vfio/pci/xe/main.c
> +++ b/drivers/vfio/pci/xe/main.c
> @@ -250,6 +250,7 @@ xe_vfio_pci_alloc_file(struct xe_vfio_pci_core_device *xe_vdev,
> struct xe_vfio_pci_migration_file *migf;
> const struct file_operations *fops;
> int flags;
> + int ret;
>
> migf = kzalloc(sizeof(*migf), GFP_KERNEL_ACCOUNT);
> if (!migf)
> @@ -259,8 +260,9 @@ xe_vfio_pci_alloc_file(struct xe_vfio_pci_core_device *xe_vdev,
> flags = type == XE_VFIO_FILE_SAVE ? O_RDONLY : O_WRONLY;
> migf->filp = anon_inode_getfile("xe_vfio_mig", fops, migf, flags);
> if (IS_ERR(migf->filp)) {
> + ret = PTR_ERR(migf->filp);
> kfree(migf);
> - return ERR_CAST(migf->filp);
> + return ERR_PTR(ret);
> }
>
> mutex_init(&migf->lock);
Powered by blists - more mailing lists