* [PATCH] Avoid peeling in cunrolli
@ 2017-11-29 14:34 Richard Biener
2017-11-30 9:18 ` Richard Biener
0 siblings, 1 reply; 2+ messages in thread
From: Richard Biener @ 2017-11-29 14:34 UTC (permalink / raw)
To: gcc-patches; +Cc: Jan Hubicka
It turns out that we don't vectorize the 2nd testcase in PR83202
(or rather we do that in weird ways during BB vectorization) because
cunrolli decides to peel the inner loop completely based on
the size of the accessed arrays. That unfortunately leaves exit
tests in the outer loop body which in turn makes us not vectorize
the loop.
We have a late unrolling pass for these kind of unrollings so this
patch simply avoids doing this during cunrolli.
Bootstrapped on x86_64-unknown-linux-gnu, testing in progress.
Richard.
2017-11-29 Richard Biener <rguenther@suse.de>
PR tree-optimization/83202
* tree-ssa-loop-ivcanon.c (try_unroll_loop_completely): Add
allow_peel argument and guard peeling.
(canonicalize_loop_induction_variables): Likewise.
(canonicalize_induction_variables): Pass false.
(tree_unroll_loops_completely_1): Pass unroll_outer to disallow
peeling from cunrolli.
* gcc.dg/vect/pr83202-1.c: New testcase.
Index: gcc/tree-ssa-loop-ivcanon.c
===================================================================
--- gcc/tree-ssa-loop-ivcanon.c (revision 255201)
+++ gcc/tree-ssa-loop-ivcanon.c (working copy)
@@ -679,7 +679,7 @@ try_unroll_loop_completely (struct loop
edge exit, tree niter,
enum unroll_level ul,
HOST_WIDE_INT maxiter,
- location_t locus)
+ location_t locus, bool allow_peel)
{
unsigned HOST_WIDE_INT n_unroll = 0;
bool n_unroll_found = false;
@@ -711,7 +711,8 @@ try_unroll_loop_completely (struct loop
exit = NULL;
/* See if we can improve our estimate by using recorded loop bounds. */
- if (maxiter >= 0
+ if (allow_peel
+ && maxiter >= 0
&& (!n_unroll_found || (unsigned HOST_WIDE_INT)maxiter < n_unroll))
{
n_unroll = maxiter;
@@ -1139,7 +1140,7 @@ try_peel_loop (struct loop *loop,
static bool
canonicalize_loop_induction_variables (struct loop *loop,
bool create_iv, enum unroll_level ul,
- bool try_eval)
+ bool try_eval, bool allow_peel)
{
edge exit = NULL;
tree niter;
@@ -1207,7 +1208,8 @@ canonicalize_loop_induction_variables (s
populates the loop bounds. */
modified |= remove_redundant_iv_tests (loop);
- if (try_unroll_loop_completely (loop, exit, niter, ul, maxiter, locus))
+ if (try_unroll_loop_completely (loop, exit, niter, ul, maxiter, locus,
+ allow_peel))
return true;
if (create_iv
@@ -1238,7 +1240,7 @@ canonicalize_induction_variables (void)
{
changed |= canonicalize_loop_induction_variables (loop,
true, UL_SINGLE_ITER,
- true);
+ true, false);
}
gcc_assert (!need_ssa_update_p (cfun));
@@ -1353,7 +1355,7 @@ tree_unroll_loops_completely_1 (bool may
ul = UL_NO_GROWTH;
if (canonicalize_loop_induction_variables
- (loop, false, ul, !flag_tree_loop_ivcanon))
+ (loop, false, ul, !flag_tree_loop_ivcanon, unroll_outer))
{
/* If we'll continue unrolling, we need to propagate constants
within the new basic blocks to fold away induction variable
Index: gcc/testsuite/gcc.dg/vect/pr83202-1.c
===================================================================
--- gcc/testsuite/gcc.dg/vect/pr83202-1.c (nonexistent)
+++ gcc/testsuite/gcc.dg/vect/pr83202-1.c (working copy)
@@ -0,0 +1,19 @@
+/* { dg-do compile } */
+/* { dg-require-effective-target vect_double } */
+
+void test(double data[8][8])
+{
+ for (int i = 0; i < 8; i++)
+ {
+ for (int j = 0; j < i; j+=4)
+ {
+ data[i][j] *= data[i][j];
+ data[i][j+1] *= data[i][j+1];
+ data[i][j+2] *= data[i][j+2];
+ data[i][j+3] *= data[i][j+3];
+ }
+ }
+}
+
+/* { dg-final { scan-tree-dump "Loop contains only SLP stmts" "vect" } } */
+/* { dg-final { scan-tree-dump "ectorized 1 loops" "vect" } } */
^ permalink raw reply [flat|nested] 2+ messages in thread
* Re: [PATCH] Avoid peeling in cunrolli
2017-11-29 14:34 [PATCH] Avoid peeling in cunrolli Richard Biener
@ 2017-11-30 9:18 ` Richard Biener
0 siblings, 0 replies; 2+ messages in thread
From: Richard Biener @ 2017-11-30 9:18 UTC (permalink / raw)
To: gcc-patches; +Cc: Jan Hubicka
On Wed, 29 Nov 2017, Richard Biener wrote:
>
> It turns out that we don't vectorize the 2nd testcase in PR83202
> (or rather we do that in weird ways during BB vectorization) because
> cunrolli decides to peel the inner loop completely based on
> the size of the accessed arrays. That unfortunately leaves exit
> tests in the outer loop body which in turn makes us not vectorize
> the loop.
>
> We have a late unrolling pass for these kind of unrollings so this
> patch simply avoids doing this during cunrolli.
>
> Bootstrapped on x86_64-unknown-linux-gnu, testing in progress.
And this is what I applied after reviewing testsuite regressions of
the first.
Bootstrapped and tested on x86_64-unknown-linux-gnu.
Richard.
2017-11-30 Richard Biener <rguenther@suse.de>
PR tree-optimization/83202
* tree-ssa-loop-ivcanon.c (try_unroll_loop_completely): Add
allow_peel argument and guard peeling.
(canonicalize_loop_induction_variables): Likewise.
(canonicalize_induction_variables): Pass false.
(tree_unroll_loops_completely_1): Pass unroll_outer to disallow
peeling from cunrolli.
* gcc.dg/vect/pr83202-1.c: New testcase.
* gcc.dg/tree-ssa/pr61743-1.c: Adjust.
Index: gcc/tree-ssa-loop-ivcanon.c
===================================================================
--- gcc/tree-ssa-loop-ivcanon.c (revision 255201)
+++ gcc/tree-ssa-loop-ivcanon.c (working copy)
@@ -679,7 +679,7 @@ try_unroll_loop_completely (struct loop
edge exit, tree niter,
enum unroll_level ul,
HOST_WIDE_INT maxiter,
- location_t locus)
+ location_t locus, bool allow_peel)
{
unsigned HOST_WIDE_INT n_unroll = 0;
bool n_unroll_found = false;
@@ -711,7 +711,8 @@ try_unroll_loop_completely (struct loop
exit = NULL;
/* See if we can improve our estimate by using recorded loop bounds. */
- if (maxiter >= 0
+ if ((allow_peel || maxiter == 0 || ul == UL_NO_GROWTH)
+ && maxiter >= 0
&& (!n_unroll_found || (unsigned HOST_WIDE_INT)maxiter < n_unroll))
{
n_unroll = maxiter;
@@ -1139,7 +1140,7 @@ try_peel_loop (struct loop *loop,
static bool
canonicalize_loop_induction_variables (struct loop *loop,
bool create_iv, enum unroll_level ul,
- bool try_eval)
+ bool try_eval, bool allow_peel)
{
edge exit = NULL;
tree niter;
@@ -1207,7 +1208,8 @@ canonicalize_loop_induction_variables (s
populates the loop bounds. */
modified |= remove_redundant_iv_tests (loop);
- if (try_unroll_loop_completely (loop, exit, niter, ul, maxiter, locus))
+ if (try_unroll_loop_completely (loop, exit, niter, ul, maxiter, locus,
+ allow_peel))
return true;
if (create_iv
@@ -1238,7 +1240,7 @@ canonicalize_induction_variables (void)
{
changed |= canonicalize_loop_induction_variables (loop,
true, UL_SINGLE_ITER,
- true);
+ true, false);
}
gcc_assert (!need_ssa_update_p (cfun));
@@ -1353,7 +1355,7 @@ tree_unroll_loops_completely_1 (bool may
ul = UL_NO_GROWTH;
if (canonicalize_loop_induction_variables
- (loop, false, ul, !flag_tree_loop_ivcanon))
+ (loop, false, ul, !flag_tree_loop_ivcanon, unroll_outer))
{
/* If we'll continue unrolling, we need to propagate constants
within the new basic blocks to fold away induction variable
Index: gcc/testsuite/gcc.dg/vect/pr83202-1.c
===================================================================
--- gcc/testsuite/gcc.dg/vect/pr83202-1.c (nonexistent)
+++ gcc/testsuite/gcc.dg/vect/pr83202-1.c (working copy)
@@ -0,0 +1,19 @@
+/* { dg-do compile } */
+/* { dg-require-effective-target vect_double } */
+
+void test(double data[8][8])
+{
+ for (int i = 0; i < 8; i++)
+ {
+ for (int j = 0; j < i; j+=4)
+ {
+ data[i][j] *= data[i][j];
+ data[i][j+1] *= data[i][j+1];
+ data[i][j+2] *= data[i][j+2];
+ data[i][j+3] *= data[i][j+3];
+ }
+ }
+}
+
+/* { dg-final { scan-tree-dump "Loop contains only SLP stmts" "vect" } } */
+/* { dg-final { scan-tree-dump "ectorized 1 loops" "vect" } } */
Index: gcc/testsuite/gcc.dg/tree-ssa/pr61743-1.c
===================================================================
--- gcc/testsuite/gcc.dg/tree-ssa/pr61743-1.c (revision 255201)
+++ gcc/testsuite/gcc.dg/tree-ssa/pr61743-1.c (working copy)
@@ -48,5 +48,6 @@ int foo1 (e_u8 a[4][N], int b1, int b2,
return 0;
}
-/* { dg-final { scan-tree-dump-times "loop with 3 iterations completely unrolled" 8 "cunroll" } } */
-/* { dg-final { scan-tree-dump-times "loop with 8 iterations completely unrolled" 2 "cunrolli" } } */
+/* { dg-final { scan-tree-dump-times "loop with 3 iterations completely unrolled" 2 "cunroll" } } */
+/* { dg-final { scan-tree-dump-times "loop with 7 iterations completely unrolled" 2 "cunroll" } } */
+/* { dg-final { scan-tree-dump-not "completely unrolled" "cunrolli" } } */
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2017-11-30 7:52 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2017-11-29 14:34 [PATCH] Avoid peeling in cunrolli Richard Biener
2017-11-30 9:18 ` Richard Biener
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).