linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Dan Carpenter <dan.carpenter@oracle.com>
To: kbusch@kernel.org
Cc: linux-kernel@vger.kernel.org, kernel-janitors@vger.kernel.org,
	Jason Gunthorpe <jgg@nvidia.com>
Subject: [bug report] node: Add memory-side caching attributes
Date: Thu, 1 Apr 2021 12:00:39 +0300	[thread overview]
Message-ID: <YGWLtzMLqSW4cxma@mwanda> (raw)

Hi Keith,

I've been trying to figure out ways Smatch can check for device managed
resources.  Like adding rules that if we call dev_set_name(&foo->bar)
then it's device managaged and if there is a kfree(foo) without calling
device_put(&foo->bar) then that's a resource leak.

Of course one of the rules is that if you call device_register(dev) then
you can't kfree(dev), it has to released with device_put(dev) and that's
true even if the register fails.  But this code here feels very
intentional so maybe there is an exception to the rule?

The patch acc02a109b04: "node: Add memory-side caching attributes"
from Mar 11, 2019, leads to the following static checker warning:

	drivers/base/node.c:285 node_init_cache_dev()
	error: kfree after device_register(): 'dev'

drivers/base/node.c
   263  static void node_init_cache_dev(struct node *node)
   264  {
   265          struct device *dev;
   266  
   267          dev = kzalloc(sizeof(*dev), GFP_KERNEL);
   268          if (!dev)
   269                  return;
   270  
   271          dev->parent = &node->dev;
   272          dev->release = node_cache_release;
   273          if (dev_set_name(dev, "memory_side_cache"))
   274                  goto free_dev;
   275  
   276          if (device_register(dev))
                    ^^^^^^^^^^^^^^^^^^^
   277                  goto free_name;
   278  
   279          pm_runtime_no_callbacks(dev);
   280          node->cache_dev = dev;
   281          return;
   282  free_name:
   283          kfree_const(dev->kobj.name);
   284  free_dev:
   285          kfree(dev);
                ^^^^^^^^^^
   286  }

regards,
dan carpenter

             reply	other threads:[~2021-04-01  9:01 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-04-01  9:00 Dan Carpenter [this message]
2021-04-01 11:25 ` [bug report] node: Add memory-side caching attributes Jason Gunthorpe
2021-04-01 14:06   ` Dan Carpenter
2021-04-01 15:22     ` Jason Gunthorpe
2021-04-01 21:27 ` Keith Busch

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=YGWLtzMLqSW4cxma@mwanda \
    --to=dan.carpenter@oracle.com \
    --cc=jgg@nvidia.com \
    --cc=kbusch@kernel.org \
    --cc=kernel-janitors@vger.kernel.org \
    --cc=linux-kernel@vger.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;
as well as URLs for NNTP newsgroup(s).