[PATCH] nfp: Fix memory leak in nfp_resource_acquire()

[PATCH] nfp: Fix memory leak in nfp_resource_acquire()

[Keita Suzuki]

This patch fixes a memory leak in nfp_resource_acquire(). res->mutex is
alllocated in nfp_resource_try_acquire(). However, when
msleep_interruptible() or time_is_before_eq_jiffies() fails, it falls
into err_fails path where res is freed, but res->mutex is not.

Fix this by changing call to free to nfp_resource_release().

Signed-off-by: Keita Suzuki <keitasuzuki.park@sslab.ics.keio.ac.jp>
---
 drivers/net/ethernet/netronome/nfp/nfpcore/nfp_resource.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/netronome/nfp/nfpcore/nfp_resource.c b/drivers/net/ethernet/netronome/nfp/nfpcore/nfp_resource.c
index ce7492a6a98f..95e7bdc95652 100644
--- a/drivers/net/ethernet/netronome/nfp/nfpcore/nfp_resource.c
+++ b/drivers/net/ethernet/netronome/nfp/nfpcore/nfp_resource.c
@@ -200,7 +200,7 @@ nfp_resource_acquire(struct nfp_cpp *cpp, const char *name)
 
 err_free:
 	nfp_cpp_mutex_free(dev_mutex);
-	kfree(res);
+	nfp_resource_relase(res);
 	return ERR_PTR(err);
 }
 
-- 
2.17.1

Re: [PATCH] nfp: Fix memory leak in nfp_resource_acquire()

[David Miller]

From: Keita Suzuki <keitasuzuki.park@sslab.ics.keio.ac.jp>
Date: Thu,  9 Apr 2020 15:02:07 +0000

> This patch fixes a memory leak in nfp_resource_acquire(). res->mutex is
> alllocated in nfp_resource_try_acquire(). However, when
> msleep_interruptible() or time_is_before_eq_jiffies() fails, it falls
> into err_fails path where res is freed, but res->mutex is not.
> 
> Fix this by changing call to free to nfp_resource_release().
> 
> Signed-off-by: Keita Suzuki <keitasuzuki.park@sslab.ics.keio.ac.jp>

Did you test compile this?

drivers/net/ethernet/netronome/nfp/nfpcore/nfp_resource.c: In function ‘nfp_resource_acquire’:
drivers/net/ethernet/netronome/nfp/nfpcore/nfp_resource.c:203:2: error: implicit declaration of function ‘nfp_resource_relase’; did you mean ‘nfp_resource_release’? [-Werror=implicit-function-declaration]
  nfp_resource_relase(res);
  ^~~~~~~~~~~~~~~~~~~
  nfp_resource_release

And this makes me feel like the test was not runtime tested either.

Re: [PATCH] nfp: Fix memory leak in nfp_resource_acquire()

[Keita Suzuki]

Hi,

So sorry about this. It seems like I accidentally touched the patch
file after generating / testing the patch.
I will resend the new patch immediately.

I have tested the patch using kmemleak.

Thanks.

2020年4月10日(金) 2:18 David Miller <davem@davemloft.net>:
>
> From: Keita Suzuki <keitasuzuki.park@sslab.ics.keio.ac.jp>
> Date: Thu,  9 Apr 2020 15:02:07 +0000
>
> > This patch fixes a memory leak in nfp_resource_acquire(). res->mutex is
> > alllocated in nfp_resource_try_acquire(). However, when
> > msleep_interruptible() or time_is_before_eq_jiffies() fails, it falls
> > into err_fails path where res is freed, but res->mutex is not.
> >
> > Fix this by changing call to free to nfp_resource_release().
> >
> > Signed-off-by: Keita Suzuki <keitasuzuki.park@sslab.ics.keio.ac.jp>
>
> Did you test compile this?
>
> drivers/net/ethernet/netronome/nfp/nfpcore/nfp_resource.c: In function ‘nfp_resource_acquire’:
> drivers/net/ethernet/netronome/nfp/nfpcore/nfp_resource.c:203:2: error: implicit declaration of function ‘nfp_resource_relase’; did you mean ‘nfp_resource_release’? [-Werror=implicit-function-declaration]
>   nfp_resource_relase(res);
>   ^~~~~~~~~~~~~~~~~~~
>   nfp_resource_release
>
> And this makes me feel like the test was not runtime tested either.

[PATCH] nfp: Fix memory leak in nfp_resource_acquire()

[Keita Suzuki]

This patch fixes a memory leak in nfp_resource_acquire(). res->mutex is
alllocated in nfp_resource_try_acquire(). However, when
msleep_interruptible() or time_is_before_eq_jiffies() fails, it falls
into err_fails path where res is freed, but res->mutex is not.

Fix this by changing call to free to nfp_resource_release().

Signed-off-by: Keita Suzuki <keitasuzuki.park@sslab.ics.keio.ac.jp>
---
 drivers/net/ethernet/netronome/nfp/nfpcore/nfp_resource.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/netronome/nfp/nfpcore/nfp_resource.c b/drivers/net/ethernet/netronome/nfp/nfpcore/nfp_resource.c
index ce7492a6a98f..95e7bdc95652 100644
--- a/drivers/net/ethernet/netronome/nfp/nfpcore/nfp_resource.c
+++ b/drivers/net/ethernet/netronome/nfp/nfpcore/nfp_resource.c
@@ -200,7 +200,7 @@ nfp_resource_acquire(struct nfp_cpp *cpp, const char *name)
 
 err_free:
 	nfp_cpp_mutex_free(dev_mutex);
-	kfree(res);
+	nfp_resource_release(res);
 	return ERR_PTR(err);
 }
 
-- 
2.17.1

Re: [PATCH] nfp: Fix memory leak in nfp_resource_acquire()

[Jakub Kicinski]

On Thu,  9 Apr 2020 17:41:11 +0000 Keita Suzuki wrote:
> This patch fixes a memory leak in nfp_resource_acquire(). res->mutex is
> alllocated in nfp_resource_try_acquire(). However, when
> msleep_interruptible() or time_is_before_eq_jiffies() fails, it falls
> into err_fails path where res is freed, but res->mutex is not.
> 
> Fix this by changing call to free to nfp_resource_release().

I don't see a leak here. Maybe you could rephrase the description to
make things clearer?

AFAICS nfp_resource_try_acquire() calls nfp_cpp_mutex_free(res->mutex)
if it's about to return an error. We can only hit msleep or time check
if it returned an error.

Re: [PATCH] nfp: Fix memory leak in nfp_resource_acquire()

[Keita Suzuki]

Hi,

Thanks for reviewing.

I'll check back the runtime log and see what I can do.
Thanks.


2020年4月10日(金) 4:32 Jakub Kicinski <kuba@kernel.org>:
>
> On Thu,  9 Apr 2020 17:41:11 +0000 Keita Suzuki wrote:
> > This patch fixes a memory leak in nfp_resource_acquire(). res->mutex is
> > alllocated in nfp_resource_try_acquire(). However, when
> > msleep_interruptible() or time_is_before_eq_jiffies() fails, it falls
> > into err_fails path where res is freed, but res->mutex is not.
> >
> > Fix this by changing call to free to nfp_resource_release().
>
> I don't see a leak here. Maybe you could rephrase the description to
> make things clearer?
>
> AFAICS nfp_resource_try_acquire() calls nfp_cpp_mutex_free(res->mutex)
> if it's about to return an error. We can only hit msleep or time check
> if it returned an error.

Re: [PATCH] nfp: Fix memory leak in nfp_resource_acquire()

[kbuild test robot]

[-- Attachment #1: Type: text/plain, Size: 3565 bytes --]

Hi Keita,

Thank you for the patch! Yet something to improve:

[auto build test ERROR on linus/master]
[also build test ERROR on v5.7-rc1 next-20200416]
[if your patch is applied to the wrong git tree, please drop us a note to help
improve the system. BTW, we also suggest to use '--base' option to specify the
base tree in git format-patch, please see https://stackoverflow.com/a/37406982]

url:    https://github.com/0day-ci/linux/commits/Keita-Suzuki/nfp-Fix-memory-leak-in-nfp_resource_acquire/20200410-032846
base:   https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git 5d30bcacd91af6874481129797af364a53cd9b46
config: x86_64-rhel-7.6 (attached as .config)
compiler: gcc-7 (Ubuntu 7.5.0-6ubuntu2) 7.5.0
reproduce:
        # save the attached .config to linux build tree
        make ARCH=x86_64 

If you fix the issue, kindly add following tag as appropriate
Reported-by: kbuild test robot <lkp@intel.com>

All errors (new ones prefixed by >>):

   drivers/net/ethernet/netronome/nfp/nfpcore/nfp_resource.c: In function 'nfp_resource_acquire':
>> drivers/net/ethernet/netronome/nfp/nfpcore/nfp_resource.c:203:2: error: implicit declaration of function 'nfp_resource_relase'; did you mean 'nfp_resource_release'? [-Werror=implicit-function-declaration]
     nfp_resource_relase(res);
     ^~~~~~~~~~~~~~~~~~~
     nfp_resource_release
   cc1: some warnings being treated as errors

vim +203 drivers/net/ethernet/netronome/nfp/nfpcore/nfp_resource.c

   139	
   140	/**
   141	 * nfp_resource_acquire() - Acquire a resource handle
   142	 * @cpp:	NFP CPP handle
   143	 * @name:	Name of the resource
   144	 *
   145	 * NOTE: This function locks the acquired resource
   146	 *
   147	 * Return: NFP Resource handle, or ERR_PTR()
   148	 */
   149	struct nfp_resource *
   150	nfp_resource_acquire(struct nfp_cpp *cpp, const char *name)
   151	{
   152		unsigned long warn_at = jiffies + NFP_MUTEX_WAIT_FIRST_WARN * HZ;
   153		unsigned long err_at = jiffies + NFP_MUTEX_WAIT_ERROR * HZ;
   154		struct nfp_cpp_mutex *dev_mutex;
   155		struct nfp_resource *res;
   156		int err;
   157	
   158		res = kzalloc(sizeof(*res), GFP_KERNEL);
   159		if (!res)
   160			return ERR_PTR(-ENOMEM);
   161	
   162		strncpy(res->name, name, NFP_RESOURCE_ENTRY_NAME_SZ);
   163	
   164		dev_mutex = nfp_cpp_mutex_alloc(cpp, NFP_RESOURCE_TBL_TARGET,
   165						NFP_RESOURCE_TBL_BASE,
   166						NFP_RESOURCE_TBL_KEY);
   167		if (!dev_mutex) {
   168			kfree(res);
   169			return ERR_PTR(-ENOMEM);
   170		}
   171	
   172		for (;;) {
   173			err = nfp_resource_try_acquire(cpp, res, dev_mutex);
   174			if (!err)
   175				break;
   176			if (err != -EBUSY)
   177				goto err_free;
   178	
   179			err = msleep_interruptible(1);
   180			if (err != 0) {
   181				err = -ERESTARTSYS;
   182				goto err_free;
   183			}
   184	
   185			if (time_is_before_eq_jiffies(warn_at)) {
   186				warn_at = jiffies + NFP_MUTEX_WAIT_NEXT_WARN * HZ;
   187				nfp_warn(cpp, "Warning: waiting for NFP resource %s\n",
   188					 name);
   189			}
   190			if (time_is_before_eq_jiffies(err_at)) {
   191				nfp_err(cpp, "Error: resource %s timed out\n", name);
   192				err = -EBUSY;
   193				goto err_free;
   194			}
   195		}
   196	
   197		nfp_cpp_mutex_free(dev_mutex);
   198	
   199		return res;
   200	
   201	err_free:
   202		nfp_cpp_mutex_free(dev_mutex);
 > 203		nfp_resource_relase(res);
   204		return ERR_PTR(err);
   205	}
   206	

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org

[-- Attachment #2: .config.gz --]
[-- Type: application/gzip, Size: 49541 bytes --]