Re: [dpdk-dev] [PATCH] vhost: Fix `struct file' leakage in `eventfd_link'

["Xie, Huawei" <huawei.xie@intel.com>]

On 3/25/2015 4:08 AM, Pavel Boldin wrote:


On Tue, Mar 24, 2015 at 6:20 PM, Xie, Huawei <huawei.xie@intel.com<mailto:huawei.xie@intel.com>> wrote:
On 3/24/2015 7:10 PM, Pavel Boldin wrote:


On Tue, Mar 24, 2015 at 8:28 AM, Xie, Huawei <huawei.xie@intel.com<mailto:huawei.xie@intel.com><mailto:huawei.xie@intel.com<mailto:huawei.xie@intel.com>>> wrote:
On 3/23/2015 8:54 PM, Pavel Boldin wrote:
> Due to increased `struct file's reference counter subsequent call
> to `filp_close' does not free the `struct file'. Prepend `fput' call
> to decrease the reference counter.
>
> Signed-off-by: Pavel Boldin <pboldin@mirantis.com<mailto:pboldin@mirantis.com><mailto:pboldin@mirantis.com<mailto:pboldin@mirantis.com>>>
> ---
>  lib/librte_vhost/eventfd_link/eventfd_link.c | 1 +
>  1 file changed, 1 insertion(+)
>
> diff --git a/lib/librte_vhost/eventfd_link/eventfd_link.c b/lib/librte_vhost/eventfd_link/eventfd_link.c
> index 7755dd6..62c45c8 100644
> --- a/lib/librte_vhost/eventfd_link/eventfd_link.c
> +++ b/lib/librte_vhost/eventfd_link/eventfd_link.c
> @@ -117,6 +117,7 @@ eventfd_link_ioctl(struct file *f, unsigned int ioctl, unsigned long arg)
>                * Release the existing eventfd in the source process
>                */
>               spin_lock(&files->file_lock);
> +             fput(file);
>               filp_close(file, files);
>               fdt = files_fdtable(files);
>               fdt->fd[eventfd_copy.source_fd] = NULL;
Acked-by Huawei Xie <huawei.xie@intel.com<mailto:huawei.xie@intel.com><mailto:huawei.xie@intel.com<mailto:huawei.xie@intel.com>>>

In future, we should remove the allocation of src eventfd, allocate a
free fd from kernel, and install it with target fd.

Well, I don't think this is correct, because this will put too much job for the kernel module making it a combiner.

At the moment I propose to accept module refactoring patch to the upstream.

After that the module can be reworked along with the application code. The reason is I can't work on DPDK since I have no hardware to test application at so I can't help with patch that touches application code.

Pavel
Pavel:
I am not asking to do it right now but in future, and i agree we accept the refactoring patch now, so i ack the patch.
Huawei, this is the refactoring patch [1]. This patch is merely a bugfix.

[1] http://dpdk.org/dev/patchwork/patch/4113/

Huawei


Thanks for your patch and fd checking tool. I will re-ack the patch as i see quoting issue in my previous ack.