DMA Engine development
 help / color / mirror / Atom feed
From: Jerry Snitselaar <jsnitsel@redhat.com>
To: Fenghua Yu <fenghua.yu@intel.com>
Cc: Vinod Koul <vkoul@kernel.org>,
	Arjan Van De Ven <arjan.van.de.ven@intel.com>,
	Dave Hansen <dave.hansen@linux.intel.com>,
	Dave Jiang <dave.jiang@intel.com>,
	Lu Baolu <baolu.lu@linux.intel.com>,
	Jacob Pan <jacob.jun.pan@linux.intel.com>,
	dmaengine@vger.kernel.org,
	linux-kernel <linux-kernel@vger.kernel.org>,
	stable@vger.kernel.org
Subject: Re: [PATCH v2] dmaengine: idxd: Do not enable user type Work Queue without Shared Virtual Addressing
Date: Mon, 17 Oct 2022 08:56:57 -0700	[thread overview]
Message-ID: <20221017155657.kpwvx5jicitoxbzp@cantor> (raw)
In-Reply-To: <20221014222541.3912195-1-fenghua.yu@intel.com>

On Fri, Oct 14, 2022 at 03:25:41PM -0700, Fenghua Yu wrote:
> When the idxd_user_drv driver is bound to a Work Queue (WQ) device
> without IOMMU or with IOMMU Passthrough without Shared Virtual
> Addressing (SVA), the application gains direct access to physical
> memory via the device by programming physical address to a submitted
> descriptor. This allows direct userspace read and write access to
> arbitrary physical memory. This is inconsistent with the security
> goals of a good kernel API.
> 
> Unlike vfio_pci driver, the IDXD char device driver does not provide any
> ways to pin user pages and translate the address from user VA to IOVA or
> PA without IOMMU SVA. Therefore the application has no way to instruct the
> device to perform DMA function. This makes the char device not usable for
> normal application usage.
> 
> Since user type WQ without SVA cannot be used for normal application usage
> and presents the security issue, bind idxd_user_drv driver and enable user
> type WQ only when SVA is enabled (i.e. user PASID is enabled).
> 
> Fixes: 448c3de8ac83 ("dmaengine: idxd: create user driver for wq 'device'")
> Cc: stable@vger.kernel.org
> Suggested-by: Arjan Van De Ven <arjan.van.de.ven@intel.com>
> Signed-off-by: Fenghua Yu <fenghua.yu@intel.com>
> Reviewed-by: Dave Jiang <dave.jiang@intel.com>

Reviewed-by: Jerry Snitselaar <jsnitsel@redhat.com>

> ---
> v2:
> - Update changlog per Dave Hansen's comments
> 
>  drivers/dma/idxd/cdev.c   | 18 ++++++++++++++++++
>  include/uapi/linux/idxd.h |  1 +
>  2 files changed, 19 insertions(+)
> 
> diff --git a/drivers/dma/idxd/cdev.c b/drivers/dma/idxd/cdev.c
> index c2808fd081d6..a9b96b18772f 100644
> --- a/drivers/dma/idxd/cdev.c
> +++ b/drivers/dma/idxd/cdev.c
> @@ -312,6 +312,24 @@ static int idxd_user_drv_probe(struct idxd_dev *idxd_dev)
>  	if (idxd->state != IDXD_DEV_ENABLED)
>  		return -ENXIO;
>  
> +	/*
> +	 * User type WQ is enabled only when SVA is enabled for two reasons:
> +	 *   - If no IOMMU or IOMMU Passthrough without SVA, userspace
> +	 *     can directly access physical address through the WQ.
> +	 *   - The IDXD cdev driver does not provide any ways to pin
> +	 *     user pages and translate the address from user VA to IOVA or
> +	 *     PA without IOMMU SVA. Therefore the application has no way
> +	 *     to instruct the device to perform DMA function. This makes
> +	 *     the cdev not usable for normal application usage.
> +	 */
> +	if (!device_user_pasid_enabled(idxd)) {
> +		idxd->cmd_status = IDXD_SCMD_WQ_USER_NO_IOMMU;
> +		dev_dbg(&idxd->pdev->dev,
> +			"User type WQ cannot be enabled without SVA.\n");
> +
> +		return -EOPNOTSUPP;
> +	}
> +
>  	mutex_lock(&wq->wq_lock);
>  	wq->type = IDXD_WQT_USER;
>  	rc = drv_enable_wq(wq);
> diff --git a/include/uapi/linux/idxd.h b/include/uapi/linux/idxd.h
> index 095299c75828..2b9e7feba3f3 100644
> --- a/include/uapi/linux/idxd.h
> +++ b/include/uapi/linux/idxd.h
> @@ -29,6 +29,7 @@ enum idxd_scmd_stat {
>  	IDXD_SCMD_WQ_NO_SIZE = 0x800e0000,
>  	IDXD_SCMD_WQ_NO_PRIV = 0x800f0000,
>  	IDXD_SCMD_WQ_IRQ_ERR = 0x80100000,
> +	IDXD_SCMD_WQ_USER_NO_IOMMU = 0x80110000,
>  };
>  
>  #define IDXD_SCMD_SOFTERR_MASK	0x80000000
> -- 
> 2.32.0
> 


  reply	other threads:[~2022-10-17 15:57 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-10-14 22:25 [PATCH v2] dmaengine: idxd: Do not enable user type Work Queue without Shared Virtual Addressing Fenghua Yu
2022-10-17 15:56 ` Jerry Snitselaar [this message]
2022-10-19  9:46 ` Tian, Kevin
2022-10-19 13:32 ` Vinod Koul

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20221017155657.kpwvx5jicitoxbzp@cantor \
    --to=jsnitsel@redhat.com \
    --cc=arjan.van.de.ven@intel.com \
    --cc=baolu.lu@linux.intel.com \
    --cc=dave.hansen@linux.intel.com \
    --cc=dave.jiang@intel.com \
    --cc=dmaengine@vger.kernel.org \
    --cc=fenghua.yu@intel.com \
    --cc=jacob.jun.pan@linux.intel.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=stable@vger.kernel.org \
    --cc=vkoul@kernel.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox