On Wed, Jan 24, 2018 at 12:35:38PM -0600, Segher Boessenkool wrote:
> Hi!
>
> On Wed, Jan 24, 2018 at 12:27:55AM -0500, Michael Meissner wrote:
> >
> > As Segher and I were discussing over private IRC, the root cause of this
> > bug is
> > the compiler no long generates the BDNZ instruction for a count down loop,
> > instead it decrements the index in a GPR and does a branch/comparison on it.
>
> Yes, ivopts makes a bad decision (it uses stride 8 for all IVs, it should
> keep one with stride -1 for the loop counter, for optimal code; it also
> does three separate increments for the three memory accesses, which is
> a bit excessive here).
>
> > In doing so, it now unrolls the loop twice, and and the resulting loop is
> > too
> > big for the target hook TARGET_ASM_LOOP_ALIGN_MAX_SKIP. This means the loop
> > isn't aligned to a 32 byte boundary.
>
> It's not really unrolling, it is bb-reorder copying an RTL block. However,
> even if you disable it you still get 9 insns on some configurations, so
> your patch does not hide the problem :-(
>
> Although, hrm, in your patch you also change "int i" to "long i"; that
> alone seems to be enough to fix everything? Could you check that please?
Replacing 'int' with 'unsigned long' allows the test to succeed once again. I
have checked this on a big endian power8 (both 32-bit and 64-bit) and on a
little endian power8 (64-bit only), and it passes in all three environments.
Can I install this on the trunk?
[gcc/testsuite]
2018-01-24 Michael Meissner <meiss...@linux.vnet.ibm.com>
PR target/81550
* gcc.target/powerpc/loop_align.c: Use unsigned long for the loop
index instead of int, which allows IVOPTs to properly optimize the
loop.
Index: gcc/testsuite/gcc.target/powerpc/loop_align.c
===================================================================
--- gcc/testsuite/gcc.target/powerpc/loop_align.c (revision 256992)
+++ gcc/testsuite/gcc.target/powerpc/loop_align.c (working copy)
@@ -4,8 +4,8 @@
/* { dg-options "-O2 -mcpu=power7 -falign-functions=16" } */
/* { dg-final { scan-assembler ".p2align 5,,31" } } */
-void f(double *a, double *b, double *c, int n) {
- int i;
+void f(double *a, double *b, double *c, unsigned long n) {
+ unsigned long i;
for (i=0; i < n; i++)
a[i] = b[i] + c[i];
}
--
Michael Meissner, IBM
IBM, M/S 2506R, 550 King Street, Littleton, MA 01460-6245, USA
email: meiss...@linux.vnet.ibm.com, phone: +1 (978) 899-4797
