Re: [PATCH] net: don't unnecessarily load kernel modules in dev_ioctl()

[David Miller <davem-fT/PcQaiUtIeIZ0/mPfg9Q@public.gmane.org>]

From: Paul Moore <paul-r2n+y4ga6xFZroRs9YW3xA@public.gmane.org>
Date: Tue, 6 Mar 2018 17:32:47 -0500

> On Tue, Mar 6, 2018 at 5:27 PM, Paul Moore <pmoore-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org> wrote:
>> From: Paul Moore <paul-r2n+y4ga6xFZroRs9YW3xA@public.gmane.org>
>>
>> Starting with v4.16-rc1 we've been seeing a higher than usual number
>> of requests for the kernel to load networking modules, even on events
>> which shouldn't trigger a module load (e.g. ioctl(TCGETS)).  Stephen
>> Smalley suggested the problem may lie in commit 44c02a2c3dc5
>> ("dev_ioctl(): move copyin/copyout to callers") which moves changes
>> the network dev_ioctl() function to always call dev_load(),
>> regardless of the requested ioctl.
>>
>> This patch moves the dev_load() calls back into the individual ioctls
>> while preserving the rest of the original patch.
>>
>> Reported-by: Dominick Grift <dac.override-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
>> Suggested-by: Stephen Smalley <sds-+05T5uksL2qpZYMLLGbcSA@public.gmane.org>
>> Signed-off-by: Paul Moore <paul-r2n+y4ga6xFZroRs9YW3xA@public.gmane.org>
>> ---
>>  net/core/dev_ioctl.c |    7 +++++--
>>  1 file changed, 5 insertions(+), 2 deletions(-)
> 
> In the interest of full disclosure, I've compiled this code but I
> haven't booted it yet (test kernel building now).  I just wanted to
> post this sooner rather than later in case the networking folks, or
> Al, had a different solution they would prefer.

This is definitely the right fix, so patch applied.

Taking the dev_load() out of that switch statement definitely has
side effects.