public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
* [Bug target/36473] New: Generate bit test (bt) instructions
@ 2008-06-09 9:54 ubizjak at gmail dot com
2008-06-09 11:40 ` [Bug target/36473] " ubizjak at gmail dot com
` (3 more replies)
0 siblings, 4 replies; 5+ messages in thread
From: ubizjak at gmail dot com @ 2008-06-09 9:54 UTC (permalink / raw)
To: gcc-bugs
[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain, Size: 1805 bytes --]
According to Intel Technology Journal [1], page 270, bt instruction runs 20%
faster on Core2 Duo than equivalent generic code.
---Qoute from p.270---
The bit test instruction bt was introduced in the i386
processor. In some implementations, including the Intel
NetBurst® micro-architecture, the instruction has a high
latency. The Intel Core micro-architecture executes bt in
a single cycle, when the bit base operand is a register.
Therefore, the Intel C++/Fortran compiler uses the bt
instruction to implement a common bit test idiom when
optimizing for the Intel Core micro-architecture. The
optimized code runs about 20% faster than the generic
version on an Intel Core 2 Duo processor. Both of these
versions are shown below:
C source code
int x, n;
...
if (x & (1 << n)) ...
Generic code generation
; edx contains x, ecx contains n.
mov eax, 1
shl eax, cl
test edx, eax
je taken
Intel Core micro-architecture code generation
; edx contains x, eax contains n.
bt edx, eax
jae taken
---/Quote---
I have a patch in testing that implements suggested optimization for
TARGET_USE_BT (including core2) targets.
[1] Inside the Intel® 10.1 Compilers: New Threadizer and New
Vectorizer for Intel® Core2 Processors, Intel Technology Journal, Vol. 11,
Issue 4, November 15, 2007,
http://download.intel.com/technology/itj/2007/v11i4/1-inside/1-Inside_the_Intel_Compilers.pdf
--
Summary: Generate bit test (bt) instructions
Product: gcc
Version: 4.3.0
Status: UNCONFIRMED
Severity: enhancement
Priority: P3
Component: target
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: ubizjak at gmail dot com
GCC target triplet: x86
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=36473
^ permalink raw reply [flat|nested] 5+ messages in thread
* [Bug target/36473] Generate bit test (bt) instructions
2008-06-09 9:54 [Bug target/36473] New: Generate bit test (bt) instructions ubizjak at gmail dot com
@ 2008-06-09 11:40 ` ubizjak at gmail dot com
2008-06-09 11:41 ` ubizjak at gmail dot com
` (2 subsequent siblings)
3 siblings, 0 replies; 5+ messages in thread
From: ubizjak at gmail dot com @ 2008-06-09 11:40 UTC (permalink / raw)
To: gcc-bugs
------- Comment #1 from ubizjak at gmail dot com 2008-06-09 11:39 -------
Created an attachment (id=15741)
--> (http://gcc.gnu.org/bugzilla/attachment.cgi?id=15741&action=view)
Proposed patch
Following code:
--cut here--
void foo (void);
int test (int x, int n)
{
if (x & (1 << n))
foo ();
return 0;
}
--cut here--
compiles using -O2 to:
test:
subl $12, %esp
movl 16(%esp), %eax
movl 20(%esp), %ecx
sarl %cl, %eax
testb $1, %al
je .L2
call foo
.L2:
xorl %eax, %eax
addl $12, %esp
ret
With attached patch, following code is produced:
test:
subl $12, %esp
movl 20(%esp), %edx
movl 16(%esp), %eax
btl %edx, %eax
jnc .L2
call foo
.L2:
xorl %eax, %eax
addl $12, %esp
ret
Attached patch doesn't have TARGET_USE_BT insn predicates, and it generates bt
instructions by default. It was used to bootstrap gcc on i686-pc-linux-gnu,
where it converts >1800 shift-and-test sequences into eqivalent bt instruction.
The patch as is was bootstrapped and regression tested on x86_64-pc-linux-gnu
as well as i686-pc-linux-gnu.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=36473
^ permalink raw reply [flat|nested] 5+ messages in thread
* [Bug target/36473] Generate bit test (bt) instructions
2008-06-09 9:54 [Bug target/36473] New: Generate bit test (bt) instructions ubizjak at gmail dot com
2008-06-09 11:40 ` [Bug target/36473] " ubizjak at gmail dot com
@ 2008-06-09 11:41 ` ubizjak at gmail dot com
2008-06-10 10:30 ` uros at gcc dot gnu dot org
2008-07-13 8:44 ` ubizjak at gmail dot com
3 siblings, 0 replies; 5+ messages in thread
From: ubizjak at gmail dot com @ 2008-06-09 11:41 UTC (permalink / raw)
To: gcc-bugs
------- Comment #2 from ubizjak at gmail dot com 2008-06-09 11:41 -------
Mine, the patch needs testcases.
2008-06-09 Uros Bizjak <ubizjak@gmail.com>
* config/i386/predicates.md (bt_comparison_operator): New predicate.
* config/i386/i386.md (*btdi_rex64): New instruction pattern.
(*btsi): Ditto.
(*jcc_btdi_rex64): New instruction and split pattern.
(*jcc_btsi): Ditto.
(*jcc_btsi_1): Ditto.
--
ubizjak at gmail dot com changed:
What |Removed |Added
----------------------------------------------------------------------------
AssignedTo|unassigned at gcc dot gnu |ubizjak at gmail dot com
|dot org |
Status|UNCONFIRMED |ASSIGNED
Ever Confirmed|0 |1
Last reconfirmed|0000-00-00 00:00:00 |2008-06-09 11:41:11
date| |
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=36473
^ permalink raw reply [flat|nested] 5+ messages in thread
* [Bug target/36473] Generate bit test (bt) instructions
2008-06-09 9:54 [Bug target/36473] New: Generate bit test (bt) instructions ubizjak at gmail dot com
2008-06-09 11:40 ` [Bug target/36473] " ubizjak at gmail dot com
2008-06-09 11:41 ` ubizjak at gmail dot com
@ 2008-06-10 10:30 ` uros at gcc dot gnu dot org
2008-07-13 8:44 ` ubizjak at gmail dot com
3 siblings, 0 replies; 5+ messages in thread
From: uros at gcc dot gnu dot org @ 2008-06-10 10:30 UTC (permalink / raw)
To: gcc-bugs
------- Comment #3 from uros at gcc dot gnu dot org 2008-06-10 10:30 -------
Subject: Bug 36473
Author: uros
Date: Tue Jun 10 10:29:36 2008
New Revision: 136615
URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=136615
Log:
PR target/36473
* config/i386/i386.c (ix86_tune_features) [TUNE_USE_BT]:
Add m_CORE2 and m_GENERIC.
* config/i386/predicates.md (bt_comparison_operator): New predicate.
* config/i386/i386.md (*btdi_rex64): New instruction pattern.
(*btsi): Ditto.
(*jcc_btdi_rex64): New instruction and split pattern.
(*jcc_btsi): Ditto.
(*jcc_btsi_1): Ditto.
(*btsq): Fix Intel asm dialect operand order.
(*btrq): Ditto.
(*btcq): Ditto.
testsuite/ChangeLog:
PR target/36473
* testsuite/gcc.target/i386/bt-1.c: New test.
* testsuite/gcc.target/i386/bt-2.c: Ditto.
Added:
trunk/gcc/testsuite/gcc.target/i386/bt-1.c
trunk/gcc/testsuite/gcc.target/i386/bt-2.c
Modified:
trunk/gcc/ChangeLog
trunk/gcc/config/i386/i386.c
trunk/gcc/config/i386/i386.md
trunk/gcc/config/i386/predicates.md
trunk/gcc/testsuite/ChangeLog
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=36473
^ permalink raw reply [flat|nested] 5+ messages in thread
* [Bug target/36473] Generate bit test (bt) instructions
2008-06-09 9:54 [Bug target/36473] New: Generate bit test (bt) instructions ubizjak at gmail dot com
` (2 preceding siblings ...)
2008-06-10 10:30 ` uros at gcc dot gnu dot org
@ 2008-07-13 8:44 ` ubizjak at gmail dot com
3 siblings, 0 replies; 5+ messages in thread
From: ubizjak at gmail dot com @ 2008-07-13 8:44 UTC (permalink / raw)
To: gcc-bugs
------- Comment #4 from ubizjak at gmail dot com 2008-07-13 08:44 -------
This is now implemented in mainline.
--
ubizjak at gmail dot com changed:
What |Removed |Added
----------------------------------------------------------------------------
URL| |http://gcc.gnu.org/ml/gcc-
| |patches/2008-
| |06/msg00503.html
Status|ASSIGNED |RESOLVED
Resolution| |FIXED
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=36473
^ permalink raw reply [flat|nested] 5+ messages in thread
end of thread, other threads:[~2008-07-13 8:44 UTC | newest]
Thread overview: 5+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2008-06-09 9:54 [Bug target/36473] New: Generate bit test (bt) instructions ubizjak at gmail dot com
2008-06-09 11:40 ` [Bug target/36473] " ubizjak at gmail dot com
2008-06-09 11:41 ` ubizjak at gmail dot com
2008-06-10 10:30 ` uros at gcc dot gnu dot org
2008-07-13 8:44 ` ubizjak at gmail dot com
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).