RE: [GIT PULL] percpu fix for v5.9-rc6

From: David Laight <David.Laight@ACULAB.COM>
To: 'Arvind Sankar' <nivedita@alum.mit.edu>,
	Linus Torvalds <torvalds@linux-foundation.org>
Cc: Matthew Wilcox <willy@infradead.org>,
	"Gustavo A. R. Silva" <gustavoars@kernel.org>,
	Dennis Zhou <dennis@kernel.org>, Tejun Heo <tj@kernel.org>,
	Christoph Lameter <cl@linux.com>, Linux-MM <linux-mm@kvack.org>,
	Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
	Kees Cook <keescook@chromium.org>
Subject: RE: [GIT PULL] percpu fix for v5.9-rc6
Date: Sat, 19 Sep 2020 15:15:08 +0000	[thread overview]
Message-ID: <776bcc3e3814433a8ed2c2027a2cf7c1@AcuMS.aculab.com> (raw)
In-Reply-To: <20200918223957.GA2964553@rani.riverdale.lan>

From: Arvind Sankar
> Sent: 18 September 2020 23:40
..
> Ouch, offsetof() and sizeof() will give different results in the
> presence of alignment padding.
> 
> https://godbolt.org/z/rqnxTK
> 
> I think, grepping at random, that at least struct scsi_vpd is like this,
> size is 24 but data[] starts at offset 20.
> 
> 	struct scsi_vpd {
> 		struct rcu_head	rcu;
> 		int		len;
> 		unsigned char	data[];
> 	};

For another standards 'brain-fart' consider:
	x = malloc(offsetof(struct scsi_vpd, data[count]));

Since offsetof() is defined to return a compile-time constant
(hi Microsoft) this is illegal unless 'count' is also a
compile-time constant.
(It ought to be defined to be constant if the field is constant.)

If count < 4 then *x = *y will also write past the end of x.
Such structure assignments should be compile-time errors.

	David

-
Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK
Registration No: 1397386 (Wales)