Quoting Eric Dumazet <eric.duma...@gmail.com>:
On 08/27/2018 09:16 AM, rpj...@crashcourse.ca wrote:
Quoting David Miller <da...@davemloft.net>:
From: "Robert P. J. Day" <rpj...@crashcourse.ca>
Date: Mon, 27 Aug 2018 04:55:29 -0400 (EDT)
another pedantic oddity -- is there a reason for these two double
negations in net/core/net-sysfs.c?
It turns an arbitrary integer into a boolean, this is a common
construct across the kernel tree so I'm surprised you've never seen
it before.
Although, I don't know how much more hand holding we're willing to
tolerate continuing to give to you at this point.
Thanks.
I mentioned in my earlier email that I know what that construct is
*typically* used for; I also pointed out that, AFAICT, it was totally
unnecessary in the context of the two routines I mentioned, which
would appear to ever return only one of two boolean values, 0 or 1.
My experience with kernel code is that one should not introduce
unnecessary complexities, which suggests (as I stated) that there
seems to be no value in the "!!" construct *in those particular
cases*, hence my curiosity.
Have you checked git history ?
My guess is that netif_carrier_ok() used to return an int, not a bool.
!!netif_carrier_ok() was not complexity, it was probably shorter than
the form used in
u32 ethtool_op_get_link(struct net_device *dev)
{
return netif_carrier_ok(dev) ? 1 : 0;
}
But really, this is really trivial stuff, we have more interesting stuff
to take care of these days.
Point taken -- for trivialities like this, I'll check Git history, and
apologies for the noise on that one.
rday
