This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [PATCH] Fix PR77826
On Thu, 13 Oct 2016, Richard Biener wrote:
The patch introduces '@@' captures which do two things - first they
change the comparison used for matching back to operand_equal_p
only (thus perform "value-matching"), second the @@ capture denotes
the specific operand that should be refered to in the result
section (ifs and result expressions).
When we face (plus @0 @1) (convert? @@0) for example this is lowered
to (plus @__2 @1) (convert? @__2@0) marking the @__2 match for
value handling.
Funny, I had the opposite convention in mind ;-)
On a completely artificial example:
(minus (plus (plus @0 @@0) @@0) @0)
would correspond to:
(minus (plus (plus @1 @2) @3) @4)
where either @1 or @4 is captured as @0, say @1 for example, @4 is
compared strictly to @1, while @2 and @3 are value-compared to @1 (lax).
This way, when we talk about @0 later in the transformation, we are indeed
talking about the thing that was called @0 in the input pattern, while @@0
is not-quite-@0.
But your version should be fine.
I modified the patterns you identified and the ones I did and
removed the operand_equal_p uses where possible.
Thanks. (some can be further generalized, as you explained in an earlier
message, but we can do that later as the need arises)
--
Marc Glisse