diff options
| author | Richard Sandiford <richard.sandiford@linaro.org> | 2018-01-13 17:58:06 +0000 |
|---|---|---|
| committer | Richard Sandiford <rsandifo@gcc.gnu.org> | 2018-01-13 17:58:06 +0000 |
| commit | 3ea518f6f63e66e48f2d41cfa41e1efae653a484 (patch) | |
| tree | 7473d7808fc322d3f684496efb33cccb2be4dc00 /gcc | |
| parent | 7e11fc7f5cecffe650b672ac1af212d4bd9f1335 (diff) | |
| download | gcc-3ea518f6f63e66e48f2d41cfa41e1efae653a484.zip gcc-3ea518f6f63e66e48f2d41cfa41e1efae653a484.tar.gz gcc-3ea518f6f63e66e48f2d41cfa41e1efae653a484.tar.bz2 | |
Protect against min_profitable_iters going negative
We had:
if (vec_outside_cost <= 0)
min_profitable_iters = 0;
else
{
min_profitable_iters = ((vec_outside_cost - scalar_outside_cost)
* assumed_vf
- vec_inside_cost * peel_iters_prologue
- vec_inside_cost * peel_iters_epilogue)
/ ((scalar_single_iter_cost * assumed_vf)
- vec_inside_cost);
which can lead to negative min_profitable_iters when the *_outside_costs
are the same and peel_iters_epilogue is nonzero (e.g. if we're peeling
for gaps).
This is tested as part of the patch that adds support for fully-predicated
loops.
2018-01-13 Richard Sandiford <richard.sandiford@linaro.org>
Alan Hayward <alan.hayward@arm.com>
David Sherwood <david.sherwood@arm.com>
gcc/
* tree-vect-loop.c (vect_estimate_min_profitable_iters): Make sure
min_profitable_iters doesn't go negative.
Co-Authored-By: Alan Hayward <alan.hayward@arm.com>
Co-Authored-By: David Sherwood <david.sherwood@arm.com>
From-SVN: r256621
Diffstat (limited to 'gcc')
| -rw-r--r-- | gcc/ChangeLog | 7 | ||||
| -rw-r--r-- | gcc/tree-vect-loop.c | 19 |
2 files changed, 17 insertions, 9 deletions
diff --git a/gcc/ChangeLog b/gcc/ChangeLog index 454af82..242ecb5 100644 --- a/gcc/ChangeLog +++ b/gcc/ChangeLog @@ -2,6 +2,13 @@ Alan Hayward <alan.hayward@arm.com> David Sherwood <david.sherwood@arm.com> + * tree-vect-loop.c (vect_estimate_min_profitable_iters): Make sure + min_profitable_iters doesn't go negative. + +2018-01-13 Richard Sandiford <richard.sandiford@linaro.org> + Alan Hayward <alan.hayward@arm.com> + David Sherwood <david.sherwood@arm.com> + * doc/md.texi (vec_mask_load_lanes@var{m}@var{n}): Document. (vec_mask_store_lanes@var{m}@var{n}): Likewise. * optabs.def (vec_mask_load_lanes_optab): New optab. diff --git a/gcc/tree-vect-loop.c b/gcc/tree-vect-loop.c index 8d1e003..5d6f1ab 100644 --- a/gcc/tree-vect-loop.c +++ b/gcc/tree-vect-loop.c @@ -3663,23 +3663,24 @@ vect_estimate_min_profitable_iters (loop_vec_info loop_vinfo, if ((scalar_single_iter_cost * assumed_vf) > (int) vec_inside_cost) { - if (vec_outside_cost <= 0) + min_profitable_iters = ((vec_outside_cost - scalar_outside_cost) + * assumed_vf + - vec_inside_cost * peel_iters_prologue + - vec_inside_cost * peel_iters_epilogue); + + if (min_profitable_iters <= 0) min_profitable_iters = 0; else - { - min_profitable_iters = ((vec_outside_cost - scalar_outside_cost) - * assumed_vf - - vec_inside_cost * peel_iters_prologue - - vec_inside_cost * peel_iters_epilogue) - / ((scalar_single_iter_cost * assumed_vf) - - vec_inside_cost); + { + min_profitable_iters /= ((scalar_single_iter_cost * assumed_vf) + - vec_inside_cost); if ((scalar_single_iter_cost * assumed_vf * min_profitable_iters) <= (((int) vec_inside_cost * min_profitable_iters) + (((int) vec_outside_cost - scalar_outside_cost) * assumed_vf))) min_profitable_iters++; - } + } } /* vector version will never be profitable. */ else |