[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20190820055623.GC27501@lst.de>
Date: Tue, 20 Aug 2019 07:56:23 +0200
From: Christoph Hellwig <hch@....de>
To: Sergey Senozhatsky <sergey.senozhatsky.work@...il.com>
Cc: Jani Nikula <jani.nikula@...ux.intel.com>,
Joonas Lahtinen <joonas.lahtinen@...ux.intel.com>,
Rodrigo Vivi <rodrigo.vivi@...el.com>,
Chris Wilson <chris@...is-wilson.co.uk>,
Alexander Viro <viro@...iv.linux.org.uk>,
David Howells <dhowells@...hat.com>,
Christoph Hellwig <hch@....de>,
intel-gfx@...ts.freedesktop.org, dri-devel@...ts.freedesktop.org,
linux-kernel@...r.kernel.org, linux-fsdevel@...r.kernel.org,
Sergey Senozhatsky <sergey.senozhatsky@...il.com>
Subject: Re: [PATCHv2 2/2] i915: do not leak module ref counter
On Tue, Aug 20, 2019 at 12:13:59PM +0900, Sergey Senozhatsky wrote:
> Always put_filesystem() in i915_gemfs_init().
>
> Signed-off-by: Sergey Senozhatsky <sergey.senozhatsky@...il.com>
> ---
> - v2: rebased (i915 does not remount gemfs anymore)
Which means it real doesn't need its mount anyore, and thus can use
plain old shmem_file_setup and doesn't need to mess with file system
types at all.
Assuming we find a legitimate rason for why a driver should be able
to create a kernel mount or a file system type where it doesn't have
access to the struct file_system_type an API that mount by file system
name and thus hides the get_fs_type and put_filesystem would be a much
better API than adding this random export.
Powered by blists - more mailing lists