Re: [PATCH v1] Travis: also test on 32-bit Linux

From: Junio C Hamano <gitster@pobox.com>
Date: Mar 10, 2017, 23:33 UTC
Message-ID: <xmqqr324wvum.fsf@gitster.mtv.corp.google.com>
In-Reply-To: <c93a61a4-05d9-8e49-9734-12c81289517b@web.de>

René Scharfe <l.s.r@web.de> writes:

Show 5 quoted lines

> 	@ depends on r @
> 	expression E;
> 	@@
> 	- *&
> 	  E

I guess my source of the confusion is that the tool that understands the semantics of the C language still needs to be told about that.

I was hoping that something that understands C only needs to be told only a single rule:

	type T
	T src, dst

	-memcpy(&dst, &src, sizeof(dst));
	+dst = src;

and then can apply that rule to this code in four ways:

	struct foo A, *Bp;

	memcpy(Bp, &A, sizeof(*Bp));
	memcpy(Bp, &A, sizeof(A));
	memcpy(&src, dstp, sizeof(A));
	memcpy(&src, dstp, sizeof(*Bp));

to obtain its rewrite:

	struct foo A, *Bp;

	*Bp = A;
	*Bp = A;
	A = *Bp;
	A = *Bp;

by knowing that (*Bp) is of type "struct foo" (even though Bp is of type "struct foo *") and sizeof(dst) and sizeof(src) are the same thing in the rule because src and dst are both of type T.