All of lore.kernel.org
 help / color / mirror / Atom feed
From: Jason Gunthorpe <jgg@nvidia.com>
To: Dave Jiang <dave.jiang@intel.com>
Cc: <vkoul@kernel.org>, Dan Williams <dan.j.williams@intel.com>,
	<dmaengine@vger.kernel.org>
Subject: Re: [PATCH v5] dmaengine: idxd: Do not use devm for 'struct device' object allocation
Date: Thu, 4 Mar 2021 14:03:08 -0400	[thread overview]
Message-ID: <20210304180308.GH4247@nvidia.com> (raw)
In-Reply-To: <161478326635.3900104.2067961356060195664.stgit@djiang5-desk3.ch.intel.com>

On Wed, Mar 03, 2021 at 07:56:30AM -0700, Dave Jiang wrote:
> Remove devm_* allocation of memory of 'struct device' objects.
> The devm_* lifetime is incompatible with device->release() lifetime.
> Address issues flagged by CONFIG_DEBUG_KOBJECT_RELEASE. Add release
> functions for each component in order to free the allocated memory at
> the appropriate time. Each component such as wq, engine, and group now
> needs to be allocated individually in order to setup the lifetime properly.
> In the process also fix up issues from the fallout of the changes.
> 
> Reported-by: Jason Gunthorpe <jgg@nvidia.com>
> Fixes: bfe1d56091c1 ("dmaengine: idxd: Init and probe for Intel data accelerators")
> Signed-off-by: Dave Jiang <dave.jiang@intel.com>
> Reviewed-by: Dan Williams <dan.j.williams@intel.com>
> v5:
> - Rebased against 5.12-rc dmaengine/fixes
> v4:
> - fix up the life time of cdev creation/destruction (Jason)
> - Tested with KASAN and other memory allocation leak detections. (Jason)
> 
> v3:
> - Remove devm_* for irq request and cleanup related bits (Jason)
> v2:
> - Remove all devm_* alloc for idxd_device (Jason)
> - Add kref dep for dma_dev (Jason)
> 
>  drivers/dma/idxd/cdev.c   |   32 +++---
>  drivers/dma/idxd/device.c |   20 ++-
>  drivers/dma/idxd/dma.c    |   13 ++
>  drivers/dma/idxd/idxd.h   |    8 +
>  drivers/dma/idxd/init.c   |  261 +++++++++++++++++++++++++++++++++------------
>  drivers/dma/idxd/irq.c    |    6 +
>  drivers/dma/idxd/sysfs.c  |   77 +++++++++----
>  7 files changed, 290 insertions(+), 127 deletions(-)
> 
> diff --git a/drivers/dma/idxd/cdev.c b/drivers/dma/idxd/cdev.c
> index 0db9b82ed8cf..1b98e06fa228 100644
> +++ b/drivers/dma/idxd/cdev.c
> @@ -259,6 +259,7 @@ static int idxd_wq_cdev_dev_setup(struct idxd_wq *wq)
>  		return -ENOMEM;
>  
>  	dev = idxd_cdev->dev;
> +	device_initialize(dev);
>  	dev->parent = &idxd->pdev->dev;
>  	dev_set_name(dev, "%s/wq%u.%u", idxd_get_dev_name(idxd),
>  		     idxd->id, wq->id);

dev_set_name() can fail

> @@ -268,25 +269,17 @@ static int idxd_wq_cdev_dev_setup(struct idxd_wq *wq)
>  	minor = ida_simple_get(&cdev_ctx->minor_ida, 0, MINORMASK, GFP_KERNEL);
>  	if (minor < 0) {
>  		rc = minor;
> -		kfree(dev);
>  		goto ida_err;

This doesn't work

>  	}
>  
>  	dev->devt = MKDEV(MAJOR(cdev_ctx->devt), minor);
>  	dev->type = &idxd_cdev_device_type;

Because this hasn't been done yet and release is thus NULL, will leak memory.

Also the order here is wrong:

	rc = cdev_device_add(cdev, dev);
	 [..]
	init_waitqueue_head(&idxd_cdev->err_queue);

Userspace can race a call to poll() before err_queue is setup.

And probably more. Please check your code carefully!

Jason

  reply	other threads:[~2021-03-04 18:05 UTC|newest]

Thread overview: 19+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-03-03 14:56 [PATCH v5] dmaengine: idxd: Do not use devm for 'struct device' object allocation Dave Jiang
2021-03-04 18:03 ` Jason Gunthorpe [this message]
2021-03-04 18:20   ` Dave Jiang
2021-03-24  5:07   ` Dan Williams
2021-03-24 11:56     ` Jason Gunthorpe
2021-03-24 16:13       ` Dan Williams
2021-03-24 16:52         ` Jason Gunthorpe
2021-03-24 17:01           ` Dan Williams
2021-03-24 19:57             ` Dan Carpenter
2021-03-24 20:00               ` Dan Williams
2021-03-25 16:48                 ` Jason Gunthorpe
2021-03-25 18:02                   ` Dan Williams
2021-03-26 23:55                     ` Jason Gunthorpe
2021-03-24 20:07             ` Dan Carpenter
2021-03-24 19:52           ` Dan Carpenter
2021-03-24 20:31             ` Dave Jiang
2021-03-24 23:35             ` Jason Gunthorpe
2021-03-25  6:52               ` Dan Carpenter
2021-03-25 11:45                 ` Jason Gunthorpe

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=20210304180308.GH4247@nvidia.com \
    --to=jgg@nvidia.com \
    --cc=dan.j.williams@intel.com \
    --cc=dave.jiang@intel.com \
    --cc=dmaengine@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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.