procyberian/gcc
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions
https://gcc.gnu.org/git.html https://procyberian.xyz
208989 commits 1 branch 0 tags 2.9 GiB
Find a file
Juzhe-Zhong 6082024891 RISC-V: Suppress the vsetvl fusion for conflict successors 
Update in v2: Add dump information.

This patch fixes the following ineffective vsetvl insertion:

void f (int32_t * restrict in, int32_t * restrict out, size_t n, size_t cond, size_t cond2)
{
  for (size_t i = 0; i < n; i++)
    {
      if (i == cond) {
        vint8mf8_t v = *(vint8mf8_t*)(in + i + 100);
        *(vint8mf8_t*)(out + i + 100) = v;
      } else if (i == cond2) {
        vfloat32mf2_t v = *(vfloat32mf2_t*)(in + i + 200);
        *(vfloat32mf2_t*)(out + i + 200) = v;
      } else if (i == (cond2 - 1)) {
        vuint16mf2_t v = *(vuint16mf2_t*)(in + i + 300);
        *(vuint16mf2_t*)(out + i + 300) = v;
      } else {
        vint8mf4_t v = *(vint8mf4_t*)(in + i + 400);
        *(vint8mf4_t*)(out + i + 400) = v;
      }
    }
}

Before this patch:

f:
.LFB0:
        .cfi_startproc
        beq     a2,zero,.L12
        addi    a7,a0,400
        addi    a6,a1,400
        addi    a0,a0,1600
        addi    a1,a1,1600
        li      a5,0
        addi    t6,a4,-1
        vsetvli t3,zero,e8,mf8,ta,ma     ---> ineffective uplift
.L7:
        beq     a3,a5,.L15
        beq     a4,a5,.L16
        beq     t6,a5,.L17
        vsetvli t1,zero,e8,mf4,ta,ma
        vle8.v  v1,0(a0)
        vse8.v  v1,0(a1)
        vsetvli t3,zero,e8,mf8,ta,ma
.L4:
        addi    a5,a5,1
        addi    a7,a7,4
        addi    a6,a6,4
        addi    a0,a0,4
        addi    a1,a1,4
        bne     a2,a5,.L7
.L12:
        ret
.L15:
        vle8.v  v1,0(a7)
        vse8.v  v1,0(a6)
        j       .L4
.L17:
        vsetvli t1,zero,e8,mf4,ta,ma
        addi    t5,a0,-400
        addi    t4,a1,-400
        vle16.v v1,0(t5)
        vse16.v v1,0(t4)
        vsetvli t3,zero,e8,mf8,ta,ma
        j       .L4
.L16:
        addi    t5,a0,-800
        addi    t4,a1,-800
        vle32.v v1,0(t5)
        vse32.v v1,0(t4)
        j       .L4

It's obvious that we are hoisting the e8mf8 vsetvl to the top. It's ineffective since e8mf8 comes from
low probability block which is if (i == cond).

For this case, we disable such fusion.

After this patch:

f:
	beq	a2,zero,.L12
	addi	a7,a0,400
	addi	a6,a1,400
	addi	a0,a0,1600
	addi	a1,a1,1600
	li	a5,0
	addi	t6,a4,-1
.L7:
	beq	a3,a5,.L15
	beq	a4,a5,.L16
	beq	t6,a5,.L17
	vsetvli	t1,zero,e8,mf4,ta,ma
	vle8.v	v1,0(a0)
	vse8.v	v1,0(a1)
.L4:
	addi	a5,a5,1
	addi	a7,a7,4
	addi	a6,a6,4
	addi	a0,a0,4
	addi	a1,a1,4
	bne	a2,a5,.L7
.L12:
	ret
.L15:
	vsetvli	t3,zero,e8,mf8,ta,ma
	vle8.v	v1,0(a7)
	vse8.v	v1,0(a6)
	j	.L4
.L17:
	addi	t5,a0,-400
	addi	t4,a1,-400
	vsetvli	t1,zero,e8,mf4,ta,ma
	vle16.v	v1,0(t5)
	vse16.v	v1,0(t4)
	j	.L4
.L16:
	addi	t5,a0,-800
	addi	t4,a1,-800
	vsetvli	t3,zero,e32,mf2,ta,ma
	vle32.v	v1,0(t5)
	vse32.v	v1,0(t4)
	j	.L4

Tested on both RV32/RV64 no regression. Ok for trunk ?

	PR target/113696

gcc/ChangeLog:

	* config/riscv/riscv-vsetvl.cc (pre_vsetvl::earliest_fuse_vsetvl_info):
	Suppress vsetvl fusion.

gcc/testsuite/ChangeLog:

	* gcc.target/riscv/rvv/vsetvl/pr113696.c: New test.
2024-02-19 16:19:41 +08:00
.github Minor formatting fix for newly-added file from previous commit 2023-11-01 19:28:56 -04:00
c++tools Update copyright years. 2024-01-03 12:19:35 +01:00
config Daily bump. 2024-01-12 00:17:54 +00:00
contrib Daily bump. 2024-02-06 00:18:46 +00:00
fixincludes Daily bump. 2023-11-23 00:18:14 +00:00
gcc RISC-V: Suppress the vsetvl fusion for conflict successors 2024-02-19 16:19:41 +08:00
gnattools Update Copyright year in ChangeLog files 2024-01-03 11:35:18 +01:00
gotools Daily bump. 2023-11-04 00:16:45 +00:00
include Daily bump. 2024-01-14 00:17:47 +00:00
INSTALL
libada Update copyright years. 2024-01-03 12:19:35 +01:00
libatomic Daily bump. 2024-02-15 00:17:50 +00:00
libbacktrace Update copyright years. 2024-01-05 08:54:28 +01:00
libcc1 Daily bump. 2024-02-10 00:17:14 +00:00
libcody Update Copyright year in ChangeLog files 2024-01-03 11:35:18 +01:00
libcpp Daily bump. 2024-02-17 00:17:08 +00:00
libdecnumber Update copyright years. 2024-01-03 12:19:35 +01:00
libffi Daily bump. 2023-10-27 00:17:12 +00:00
libgcc Daily bump. 2024-02-17 00:17:08 +00:00
libgfortran Daily bump. 2024-02-18 00:17:05 +00:00
libgm2 Daily bump. 2024-02-14 00:17:32 +00:00
libgo libgo: bump libgo version for GCC 14 release 2024-02-05 11:28:30 -08:00
libgomp Daily bump. 2024-02-16 00:17:22 +00:00
libgrust Daily bump. 2024-01-17 00:21:29 +00:00
libiberty Daily bump. 2024-02-16 00:17:22 +00:00
libitm Daily bump. 2024-02-06 00:18:46 +00:00
libobjc Update copyright years. 2024-01-03 12:19:35 +01:00
libphobos Daily bump. 2024-02-18 00:17:05 +00:00
libquadmath Daily bump. 2024-01-04 00:18:45 +00:00
libsanitizer Daily bump. 2024-02-17 00:17:08 +00:00
libssp Daily bump. 2024-02-07 00:18:31 +00:00
libstdc++-v3 Daily bump. 2024-02-18 00:17:05 +00:00
libvtv Update copyright years. 2024-01-03 12:19:35 +01:00
lto-plugin Update copyright years. 2024-01-03 12:19:35 +01:00
maintainer-scripts Daily bump. 2023-11-14 12:23:39 +00:00
zlib Daily bump. 2023-10-23 00:16:43 +00:00
.dir-locals.el
.gitattributes
.gitignore *: add modern gettext 2023-11-14 00:47:11 +01:00
ABOUT-NLS
ar-lib
ChangeLog Daily bump. 2024-02-11 00:17:01 +00:00
ChangeLog.jit
ChangeLog.tree-ssa
compile
config-ml.in LoongArch: Reimplement multilib build option handling. 2023-09-15 10:42:12 +08:00
config.guess
config.rpath
config.sub
configure build: Add libgrust as compilation modules 2023-12-14 13:58:57 +01:00
configure.ac build: Add libgrust as compilation modules 2023-12-14 13:58:57 +01:00
COPYING
COPYING.LIB
COPYING.RUNTIME
COPYING3
COPYING3.LIB
depcomp
install-sh
libtool-ldflags
libtool.m4 Build: fix error in fixinclude configure 2023-11-22 11:54:33 +01:00
ltgcc.m4
ltmain.sh
ltoptions.m4
ltsugar.m4
ltversion.m4
lt~obsolete.m4
MAINTAINERS Add Myself to Write After Approval and DCO List 2024-02-10 12:49:29 -06:00
Makefile.def gccrs: Fix missing build dependency 2024-01-16 16:23:02 +01:00
Makefile.in gccrs: Fix missing build dependency 2024-01-16 16:23:02 +01:00
Makefile.tpl Pass GUILE down to subdirectories 2024-01-09 08:02:31 -07:00
missing
mkdep
mkinstalldirs
move-if-change
multilib.am
README
SECURITY.txt SECURITY.txt: Drop "exploitable" in reference to hardening issues 2024-01-09 10:49:01 -05:00
symlink-tree
test-driver
ylwrap

README 

This directory contains the GNU Compiler Collection (GCC).

The GNU Compiler Collection is free software.  See the files whose
names start with COPYING for copying permission.  The manuals, and
some of the runtime libraries, are under different terms; see the
individual source files for details.

The directory INSTALL contains copies of the installation information
as HTML and plain text.  The source of this information is
gcc/doc/install.texi.  The installation information includes details
of what is included in the GCC sources and what files GCC installs.

See the file gcc/doc/gcc.texi (together with other files that it
includes) for usage and porting information.  An online readable
version of the manual is in the files gcc/doc/gcc.info*.

See http://gcc.gnu.org/bugs/ for how to report bugs usefully.

Copyright years on GCC source files may be listed using range
notation, e.g., 1987-2012, indicating that every year in the range,
inclusive, is a copyrightable year that could otherwise be listed
individually.