From: Uros Bizjak <ubizjak@gmail.com>
To: GCC Patches <gcc-patches@gcc.gnu.org>
Subject: [PATCH, i386]: Avoid store forwarding stalls in assign_i386_stack_local
Date: Sat, 22 Mar 2008 19:43:00 -0000 [thread overview]
Message-ID: <47E55F31.9020605@gmail.com> (raw)
[-- Attachment #1: Type: text/plain, Size: 560 bytes --]
Hello!
This patch avoids store forwarding stalls in stack slots, created by
assign_i386_stack_local. We use DImode slots to load DImode values into
x87 reg usign fildll insn through the stack. When stored value crosses
16byte boundary this causes store forwarding stall.
2008-03-22 Uros Bizjak <ubizjak@gmail.com>
* config/i386/i386.c (assign_386_stack_local): Align DImode slots
to their natural alignment to avoid store forwarding stalls.
Patch is bootstrapped and regression tested on x86_64-pc-linux-gnu,
committed to SVN.
Uros.
[-- Attachment #2: p.diff.txt --]
[-- Type: text/plain, Size: 1033 bytes --]
Index: i386.c
===================================================================
--- i386.c (revision 133445)
+++ i386.c (working copy)
@@ -16315,6 +16315,7 @@ rtx
assign_386_stack_local (enum machine_mode mode, enum ix86_stack_slot n)
{
struct stack_local_entry *s;
+ int align;
gcc_assert (n < MAX_386_STACK_LOCALS);
@@ -16325,11 +16326,19 @@ assign_386_stack_local (enum machine_mod
if (s->mode == mode && s->n == n)
return copy_rtx (s->rtl);
+ /* Align DImode slots to their natural alignment
+ to avoid store forwarding stalls. */
+ if (mode == DImode
+ && (GET_MODE_BITSIZE (mode) > GET_MODE_ALIGNMENT (mode)))
+ align = GET_MODE_BITSIZE (mode);
+ else
+ align = 0;
+
s = (struct stack_local_entry *)
ggc_alloc (sizeof (struct stack_local_entry));
s->n = n;
s->mode = mode;
- s->rtl = assign_stack_local (mode, GET_MODE_SIZE (mode), 0);
+ s->rtl = assign_stack_local (mode, GET_MODE_SIZE (mode), align);
s->next = ix86_stack_locals;
ix86_stack_locals = s;
reply other threads:[~2008-03-22 19:34 UTC|newest]
Thread overview: [no followups] expand[flat|nested] mbox.gz Atom feed
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=47E55F31.9020605@gmail.com \
--to=ubizjak@gmail.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).