Commit 6aa7a103 authored by Ian Romanick's avatar Ian Romanick Committed by Dylan Baker

nir/range-analysis: Fix incorrect fadd range result for (ne_zero, ne_zero)

Found by inspection.  I tried really, really hard to make a test case
that would trigger this problem, but I was unsuccesful.  It's very hard
to get an instruction to produce a ne_zero result without ne_zero
sources.  The most plausible way is using bcsel.  That proves
problematic because bcsel interprets its sources as integers, so it
cannot currently be used to "clean" values for floating point
instructions.

No shader-db changes on any Intel platform.
Reviewed-by: Caio Marcelo de Oliveira Filho's avatarCaio Marcelo de Oliveira Filho <caio.oliveira@intel.com>
Fixes: 405de7cc ("nir/range-analysis: Rudimentary value range analysis pass")
(cherry picked from commit 0b4782fc)
parent 97e44d68
......@@ -247,9 +247,15 @@ analyze_expression(const nir_alu_instr *instr, unsigned src,
* | lt_zero + lt_zero
* ;
*
* ne_zero: eq_zero + ne_zero
* | ne_zero + eq_zero # Addition is commutative
* ;
*
* eq_zero: eq_zero + eq_zero
* ;
*
* All other cases are 'unknown'.
* All other cases are 'unknown'. The seeming odd entry is (ne_zero,
* ne_zero), but that could be (-5, +5) which is not ne_zero.
*/
static const enum ssa_ranges fadd_table[last_range + 1][last_range + 1] = {
/* left\right unknown lt_zero le_zero gt_zero ge_zero ne_zero eq_zero */
......@@ -258,12 +264,11 @@ analyze_expression(const nir_alu_instr *instr, unsigned src,
/* le_zero */ { _______, lt_zero, le_zero, _______, _______, _______, le_zero },
/* gt_zero */ { _______, _______, _______, gt_zero, gt_zero, _______, gt_zero },
/* ge_zero */ { _______, _______, _______, gt_zero, ge_zero, _______, ge_zero },
/* ne_zero */ { _______, _______, _______, _______, _______, ne_zero, ne_zero },
/* ne_zero */ { _______, _______, _______, _______, _______, _______, ne_zero },
/* eq_zero */ { _______, lt_zero, le_zero, gt_zero, ge_zero, ne_zero, eq_zero },
};
ASSERT_TABLE_IS_COMMUTATIVE(fadd_table);
ASSERT_TABLE_IS_DIAGONAL(fadd_table);
/* Due to flush-to-zero semanatics of floating-point numbers with very
* small mangnitudes, we can never really be sure a result will be
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment