From: Eric Botcazou <botcazou@adacore.com>
To: gcc-patches@gcc.gnu.org
Subject: [PATCH] Fix wrong array type conversion with different storage order
Date: Tue, 22 Nov 2022 12:05:16 +0100 [thread overview]
Message-ID: <4442231.LvFx2qVVIh@fomalhaut> (raw)
[-- Attachment #1: Type: text/plain, Size: 678 bytes --]
Hi,
when two arrays of scalars have a different storage order in Ada, the
front-end makes sure that the conversion is performed component-wise
so that each component can be reversed. So it's a little bit counter
productive that the ldist pass performs the opposite transformation
and synthesizes a memcpy/memmove in this case.
Tested on x86-64/Linux, OK for the mainline?
2022-11-22 Eric Botcazou <ebotcazou@adacore.com>
* tree-loop-distribution.cc (loop_distribution::classify_builtin_ldst):
Bail out if source and destination do not have the same storage order.
2022-11-22 Eric Botcazou <ebotcazou@adacore.com>
* gnat.dg/sso18.adb: New test.
--
Eric Botcazou
[-- Attachment #2: p.diff --]
[-- Type: text/x-patch, Size: 856 bytes --]
diff --git a/gcc/tree-loop-distribution.cc b/gcc/tree-loop-distribution.cc
index ed3dd73e1a9..15ae2410861 100644
--- a/gcc/tree-loop-distribution.cc
+++ b/gcc/tree-loop-distribution.cc
@@ -1790,10 +1790,15 @@ loop_distribution::classify_builtin_ldst (loop_p loop, struct graph *rdg,
if (res != 2)
return;
- /* They much have the same access size. */
+ /* They must have the same access size. */
if (!operand_equal_p (size, src_size, 0))
return;
+ /* They must have the same storage order. */
+ if (reverse_storage_order_for_component_p (DR_REF (dst_dr))
+ != reverse_storage_order_for_component_p (DR_REF (src_dr)))
+ return;
+
/* Load and store in loop nest must access memory in the same way, i.e,
their must have the same steps in each loop of the nest. */
if (dst_steps.length () != src_steps.length ())
[-- Attachment #3: sso18.adb --]
[-- Type: text/x-adasrc, Size: 419 bytes --]
-- { dg-do run }
-- { dg-options "-O2" }
with System;
procedure SSO18 is
type Arr is array (1..32) of Short_Integer;
type Rev_Arr is array (1..32) of Short_Integer
with Scalar_Storage_Order => System.High_Order_First;
C : constant Arr := (others => 16);
RA : Rev_Arr;
A : Arr;
begin
RA := Rev_Arr(C);
A := Arr (RA);
if A /= C or else RA(1) /= 16 then
raise Program_Error;
end if;
end;
next reply other threads:[~2022-11-22 11:05 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-11-22 11:05 Eric Botcazou [this message]
2022-11-22 11:22 ` Richard Biener
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=4442231.LvFx2qVVIh@fomalhaut \
--to=botcazou@adacore.com \
--cc=gcc-patches@gcc.gnu.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).