2014-04-12 17:40:18 -07:00
|
|
|
/*
|
|
|
|
|
* Copyright © 2014 Intel Corporation
|
|
|
|
|
*
|
|
|
|
|
* Permission is hereby granted, free of charge, to any person obtaining a
|
|
|
|
|
* copy of this software and associated documentation files (the "Software"),
|
|
|
|
|
* to deal in the Software without restriction, including without limitation
|
|
|
|
|
* the rights to use, copy, modify, merge, publish, distribute, sublicense,
|
|
|
|
|
* and/or sell copies of the Software, and to permit persons to whom the
|
|
|
|
|
* Software is furnished to do so, subject to the following conditions:
|
|
|
|
|
*
|
|
|
|
|
* The above copyright notice and this permission notice (including the next
|
|
|
|
|
* paragraph) shall be included in all copies or substantial portions of the
|
|
|
|
|
* Software.
|
|
|
|
|
*
|
|
|
|
|
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
|
|
|
|
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
|
|
|
|
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
|
|
|
|
|
* THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
|
|
|
|
* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
|
|
|
|
|
* FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
|
|
|
|
|
* IN THE SOFTWARE.
|
|
|
|
|
*/
|
|
|
|
|
|
|
|
|
|
#include "brw_fs.h"
|
|
|
|
|
#include "brw_fs_live_variables.h"
|
|
|
|
|
#include "brw_cfg.h"
|
|
|
|
|
|
|
|
|
|
/** @file brw_fs_dead_code_eliminate.cpp
|
|
|
|
|
*
|
|
|
|
|
* Dataflow-aware dead code elimination.
|
|
|
|
|
*
|
|
|
|
|
* Walks the instruction list from the bottom, removing instructions that
|
|
|
|
|
* have results that both aren't used in later blocks and haven't been read
|
|
|
|
|
* yet in the tail end of this block.
|
|
|
|
|
*/
|
|
|
|
|
|
2016-03-12 18:50:24 -08:00
|
|
|
using namespace brw;
|
|
|
|
|
|
2017-01-14 23:32:12 -08:00
|
|
|
/**
|
|
|
|
|
* Is it safe to eliminate the instruction?
|
|
|
|
|
*/
|
|
|
|
|
static bool
|
intel/fs: sel.cond writes the flags on Gfx4 and Gfx5
On Gfx4 and Gfx5, sel.l (for min) and sel.ge (for max) are implemented
using a separte cmpn and sel instruction. This lowering occurs in
fs_vistor::lower_minmax which is called very, very late... a long, long
time after the first calls to opt_cmod_propagation. As a result,
conditional modifiers can be incorrectly propagated across sel.cond on
those platforms.
No tests were affected by this change, and I find that quite shocking.
After just changing flags_written(), all of the atan tests started
failing on ILK. That required the change in cmod_propagatin (and the
addition of the prop_across_into_sel_gfx5 unit test).
Shader-db results for ILK and GM45 are below. I looked at a couple
before and after shaders... and every case that I looked at had
experienced incorrect cmod propagation. This affected a LOT of apps!
Euro Truck Simulator 2, The Talos Principle, Serious Sam 3, Sanctum 2,
Gang Beasts, and on and on... :(
I discovered this bug while working on a couple new optimization
passes. One of the passes attempts to remove condition modifiers that
are never used. The pass made no progress except on ILK and GM45.
After investigating a couple of the affected shaders, I noticed that
the code in those shaders looked wrong... investigation led to this
cause.
v2: Trivial changes in the unit tests.
v3: Fix type in comment in unit tests. Noticed by Jason and Priit.
v4: Tweak handling of BRW_OPCODE_SEL special case. Suggested by Jason.
Fixes: df1aec763eb ("i965/fs: Define methods to calculate the flag subset read or written by an fs_inst.")
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Tested-by: Dave Airlie <airlied@redhat.com>
Iron Lake
total instructions in shared programs: 8180493 -> 8181781 (0.02%)
instructions in affected programs: 541796 -> 543084 (0.24%)
helped: 28
HURT: 1158
helped stats (abs) min: 1 max: 1 x̄: 1.00 x̃: 1
helped stats (rel) min: 0.35% max: 0.86% x̄: 0.53% x̃: 0.50%
HURT stats (abs) min: 1 max: 3 x̄: 1.14 x̃: 1
HURT stats (rel) min: 0.12% max: 4.00% x̄: 0.37% x̃: 0.23%
95% mean confidence interval for instructions value: 1.06 1.11
95% mean confidence interval for instructions %-change: 0.31% 0.38%
Instructions are HURT.
total cycles in shared programs: 239420470 -> 239421690 (<.01%)
cycles in affected programs: 2925992 -> 2927212 (0.04%)
helped: 49
HURT: 157
helped stats (abs) min: 2 max: 284 x̄: 62.69 x̃: 70
helped stats (rel) min: 0.04% max: 6.20% x̄: 1.68% x̃: 1.96%
HURT stats (abs) min: 2 max: 48 x̄: 27.34 x̃: 24
HURT stats (rel) min: 0.02% max: 2.91% x̄: 0.31% x̃: 0.20%
95% mean confidence interval for cycles value: -0.80 12.64
95% mean confidence interval for cycles %-change: -0.31% <.01%
Inconclusive result (value mean confidence interval includes 0).
GM45
total instructions in shared programs: 4985517 -> 4986207 (0.01%)
instructions in affected programs: 306935 -> 307625 (0.22%)
helped: 14
HURT: 625
helped stats (abs) min: 1 max: 1 x̄: 1.00 x̃: 1
helped stats (rel) min: 0.35% max: 0.82% x̄: 0.52% x̃: 0.49%
HURT stats (abs) min: 1 max: 3 x̄: 1.13 x̃: 1
HURT stats (rel) min: 0.12% max: 3.90% x̄: 0.34% x̃: 0.22%
95% mean confidence interval for instructions value: 1.04 1.12
95% mean confidence interval for instructions %-change: 0.29% 0.36%
Instructions are HURT.
total cycles in shared programs: 153827268 -> 153828052 (<.01%)
cycles in affected programs: 1669290 -> 1670074 (0.05%)
helped: 24
HURT: 84
helped stats (abs) min: 2 max: 232 x̄: 64.33 x̃: 67
helped stats (rel) min: 0.04% max: 4.62% x̄: 1.60% x̃: 1.94%
HURT stats (abs) min: 2 max: 48 x̄: 27.71 x̃: 24
HURT stats (rel) min: 0.02% max: 2.66% x̄: 0.34% x̃: 0.14%
95% mean confidence interval for cycles value: -1.94 16.46
95% mean confidence interval for cycles %-change: -0.29% 0.11%
Inconclusive result (value mean confidence interval includes 0).
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12191>
2021-08-02 21:33:17 -07:00
|
|
|
can_eliminate(const intel_device_info *devinfo, const fs_inst *inst,
|
|
|
|
|
BITSET_WORD *flag_live)
|
2017-01-14 23:32:12 -08:00
|
|
|
{
|
|
|
|
|
return !inst->is_control_flow() &&
|
|
|
|
|
!inst->has_side_effects() &&
|
intel/fs: sel.cond writes the flags on Gfx4 and Gfx5
On Gfx4 and Gfx5, sel.l (for min) and sel.ge (for max) are implemented
using a separte cmpn and sel instruction. This lowering occurs in
fs_vistor::lower_minmax which is called very, very late... a long, long
time after the first calls to opt_cmod_propagation. As a result,
conditional modifiers can be incorrectly propagated across sel.cond on
those platforms.
No tests were affected by this change, and I find that quite shocking.
After just changing flags_written(), all of the atan tests started
failing on ILK. That required the change in cmod_propagatin (and the
addition of the prop_across_into_sel_gfx5 unit test).
Shader-db results for ILK and GM45 are below. I looked at a couple
before and after shaders... and every case that I looked at had
experienced incorrect cmod propagation. This affected a LOT of apps!
Euro Truck Simulator 2, The Talos Principle, Serious Sam 3, Sanctum 2,
Gang Beasts, and on and on... :(
I discovered this bug while working on a couple new optimization
passes. One of the passes attempts to remove condition modifiers that
are never used. The pass made no progress except on ILK and GM45.
After investigating a couple of the affected shaders, I noticed that
the code in those shaders looked wrong... investigation led to this
cause.
v2: Trivial changes in the unit tests.
v3: Fix type in comment in unit tests. Noticed by Jason and Priit.
v4: Tweak handling of BRW_OPCODE_SEL special case. Suggested by Jason.
Fixes: df1aec763eb ("i965/fs: Define methods to calculate the flag subset read or written by an fs_inst.")
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Tested-by: Dave Airlie <airlied@redhat.com>
Iron Lake
total instructions in shared programs: 8180493 -> 8181781 (0.02%)
instructions in affected programs: 541796 -> 543084 (0.24%)
helped: 28
HURT: 1158
helped stats (abs) min: 1 max: 1 x̄: 1.00 x̃: 1
helped stats (rel) min: 0.35% max: 0.86% x̄: 0.53% x̃: 0.50%
HURT stats (abs) min: 1 max: 3 x̄: 1.14 x̃: 1
HURT stats (rel) min: 0.12% max: 4.00% x̄: 0.37% x̃: 0.23%
95% mean confidence interval for instructions value: 1.06 1.11
95% mean confidence interval for instructions %-change: 0.31% 0.38%
Instructions are HURT.
total cycles in shared programs: 239420470 -> 239421690 (<.01%)
cycles in affected programs: 2925992 -> 2927212 (0.04%)
helped: 49
HURT: 157
helped stats (abs) min: 2 max: 284 x̄: 62.69 x̃: 70
helped stats (rel) min: 0.04% max: 6.20% x̄: 1.68% x̃: 1.96%
HURT stats (abs) min: 2 max: 48 x̄: 27.34 x̃: 24
HURT stats (rel) min: 0.02% max: 2.91% x̄: 0.31% x̃: 0.20%
95% mean confidence interval for cycles value: -0.80 12.64
95% mean confidence interval for cycles %-change: -0.31% <.01%
Inconclusive result (value mean confidence interval includes 0).
GM45
total instructions in shared programs: 4985517 -> 4986207 (0.01%)
instructions in affected programs: 306935 -> 307625 (0.22%)
helped: 14
HURT: 625
helped stats (abs) min: 1 max: 1 x̄: 1.00 x̃: 1
helped stats (rel) min: 0.35% max: 0.82% x̄: 0.52% x̃: 0.49%
HURT stats (abs) min: 1 max: 3 x̄: 1.13 x̃: 1
HURT stats (rel) min: 0.12% max: 3.90% x̄: 0.34% x̃: 0.22%
95% mean confidence interval for instructions value: 1.04 1.12
95% mean confidence interval for instructions %-change: 0.29% 0.36%
Instructions are HURT.
total cycles in shared programs: 153827268 -> 153828052 (<.01%)
cycles in affected programs: 1669290 -> 1670074 (0.05%)
helped: 24
HURT: 84
helped stats (abs) min: 2 max: 232 x̄: 64.33 x̃: 67
helped stats (rel) min: 0.04% max: 4.62% x̄: 1.60% x̃: 1.94%
HURT stats (abs) min: 2 max: 48 x̄: 27.71 x̃: 24
HURT stats (rel) min: 0.02% max: 2.66% x̄: 0.34% x̃: 0.14%
95% mean confidence interval for cycles value: -1.94 16.46
95% mean confidence interval for cycles %-change: -0.29% 0.11%
Inconclusive result (value mean confidence interval includes 0).
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12191>
2021-08-02 21:33:17 -07:00
|
|
|
!(flag_live[0] & inst->flags_written(devinfo)) &&
|
2017-01-14 23:32:12 -08:00
|
|
|
!inst->writes_accumulator;
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
/**
|
|
|
|
|
* Is it safe to omit the write, making the destination ARF null?
|
|
|
|
|
*/
|
|
|
|
|
static bool
|
|
|
|
|
can_omit_write(const fs_inst *inst)
|
|
|
|
|
{
|
|
|
|
|
switch (inst->opcode) {
|
2023-11-22 01:46:58 -08:00
|
|
|
case SHADER_OPCODE_A64_UNTYPED_ATOMIC_LOGICAL:
|
2017-01-14 23:32:12 -08:00
|
|
|
case SHADER_OPCODE_UNTYPED_ATOMIC_LOGICAL:
|
|
|
|
|
case SHADER_OPCODE_TYPED_ATOMIC_LOGICAL:
|
|
|
|
|
return true;
|
|
|
|
|
default:
|
|
|
|
|
/* We can eliminate the destination write for ordinary instructions,
|
|
|
|
|
* but not most SENDs.
|
|
|
|
|
*/
|
|
|
|
|
if (inst->opcode < 128 && inst->mlen == 0)
|
|
|
|
|
return true;
|
|
|
|
|
|
|
|
|
|
/* It might not be safe for other virtual opcodes. */
|
|
|
|
|
return false;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
2014-04-12 17:40:18 -07:00
|
|
|
bool
|
2024-01-03 11:08:36 -08:00
|
|
|
brw_fs_opt_dead_code_eliminate(fs_visitor &s)
|
2014-04-12 17:40:18 -07:00
|
|
|
{
|
2024-01-03 11:08:36 -08:00
|
|
|
const intel_device_info *devinfo = s.devinfo;
|
|
|
|
|
|
2014-04-12 17:40:18 -07:00
|
|
|
bool progress = false;
|
|
|
|
|
|
2024-01-03 11:08:36 -08:00
|
|
|
const fs_live_variables &live_vars = s.live_analysis.require();
|
2016-03-13 16:25:57 -07:00
|
|
|
int num_vars = live_vars.num_vars;
|
2016-10-11 13:48:16 +03:00
|
|
|
BITSET_WORD *live = rzalloc_array(NULL, BITSET_WORD, BITSET_WORDS(num_vars));
|
|
|
|
|
BITSET_WORD *flag_live = rzalloc_array(NULL, BITSET_WORD, 1);
|
2014-04-12 17:40:18 -07:00
|
|
|
|
2024-01-03 11:08:36 -08:00
|
|
|
foreach_block_reverse_safe(block, s.cfg) {
|
2016-03-13 16:25:57 -07:00
|
|
|
memcpy(live, live_vars.block_data[block->num].liveout,
|
2014-04-12 17:40:18 -07:00
|
|
|
sizeof(BITSET_WORD) * BITSET_WORDS(num_vars));
|
2016-03-13 16:25:57 -07:00
|
|
|
memcpy(flag_live, live_vars.block_data[block->num].flag_liveout,
|
2014-10-28 20:12:21 -07:00
|
|
|
sizeof(BITSET_WORD));
|
2014-04-12 17:40:18 -07:00
|
|
|
|
2015-11-30 09:25:19 -08:00
|
|
|
foreach_inst_in_block_reverse_safe(fs_inst, inst, block) {
|
2017-01-14 23:32:12 -08:00
|
|
|
if (inst->dst.file == VGRF) {
|
2016-03-13 16:25:57 -07:00
|
|
|
const unsigned var = live_vars.var_from_reg(inst->dst);
|
2014-04-12 17:40:18 -07:00
|
|
|
bool result_live = false;
|
|
|
|
|
|
2016-09-01 16:46:34 -07:00
|
|
|
for (unsigned i = 0; i < regs_written(inst); i++)
|
|
|
|
|
result_live |= BITSET_TEST(live, var + i);
|
2014-04-12 17:40:18 -07:00
|
|
|
|
2017-01-14 23:32:12 -08:00
|
|
|
if (!result_live &&
|
intel/fs: sel.cond writes the flags on Gfx4 and Gfx5
On Gfx4 and Gfx5, sel.l (for min) and sel.ge (for max) are implemented
using a separte cmpn and sel instruction. This lowering occurs in
fs_vistor::lower_minmax which is called very, very late... a long, long
time after the first calls to opt_cmod_propagation. As a result,
conditional modifiers can be incorrectly propagated across sel.cond on
those platforms.
No tests were affected by this change, and I find that quite shocking.
After just changing flags_written(), all of the atan tests started
failing on ILK. That required the change in cmod_propagatin (and the
addition of the prop_across_into_sel_gfx5 unit test).
Shader-db results for ILK and GM45 are below. I looked at a couple
before and after shaders... and every case that I looked at had
experienced incorrect cmod propagation. This affected a LOT of apps!
Euro Truck Simulator 2, The Talos Principle, Serious Sam 3, Sanctum 2,
Gang Beasts, and on and on... :(
I discovered this bug while working on a couple new optimization
passes. One of the passes attempts to remove condition modifiers that
are never used. The pass made no progress except on ILK and GM45.
After investigating a couple of the affected shaders, I noticed that
the code in those shaders looked wrong... investigation led to this
cause.
v2: Trivial changes in the unit tests.
v3: Fix type in comment in unit tests. Noticed by Jason and Priit.
v4: Tweak handling of BRW_OPCODE_SEL special case. Suggested by Jason.
Fixes: df1aec763eb ("i965/fs: Define methods to calculate the flag subset read or written by an fs_inst.")
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Tested-by: Dave Airlie <airlied@redhat.com>
Iron Lake
total instructions in shared programs: 8180493 -> 8181781 (0.02%)
instructions in affected programs: 541796 -> 543084 (0.24%)
helped: 28
HURT: 1158
helped stats (abs) min: 1 max: 1 x̄: 1.00 x̃: 1
helped stats (rel) min: 0.35% max: 0.86% x̄: 0.53% x̃: 0.50%
HURT stats (abs) min: 1 max: 3 x̄: 1.14 x̃: 1
HURT stats (rel) min: 0.12% max: 4.00% x̄: 0.37% x̃: 0.23%
95% mean confidence interval for instructions value: 1.06 1.11
95% mean confidence interval for instructions %-change: 0.31% 0.38%
Instructions are HURT.
total cycles in shared programs: 239420470 -> 239421690 (<.01%)
cycles in affected programs: 2925992 -> 2927212 (0.04%)
helped: 49
HURT: 157
helped stats (abs) min: 2 max: 284 x̄: 62.69 x̃: 70
helped stats (rel) min: 0.04% max: 6.20% x̄: 1.68% x̃: 1.96%
HURT stats (abs) min: 2 max: 48 x̄: 27.34 x̃: 24
HURT stats (rel) min: 0.02% max: 2.91% x̄: 0.31% x̃: 0.20%
95% mean confidence interval for cycles value: -0.80 12.64
95% mean confidence interval for cycles %-change: -0.31% <.01%
Inconclusive result (value mean confidence interval includes 0).
GM45
total instructions in shared programs: 4985517 -> 4986207 (0.01%)
instructions in affected programs: 306935 -> 307625 (0.22%)
helped: 14
HURT: 625
helped stats (abs) min: 1 max: 1 x̄: 1.00 x̃: 1
helped stats (rel) min: 0.35% max: 0.82% x̄: 0.52% x̃: 0.49%
HURT stats (abs) min: 1 max: 3 x̄: 1.13 x̃: 1
HURT stats (rel) min: 0.12% max: 3.90% x̄: 0.34% x̃: 0.22%
95% mean confidence interval for instructions value: 1.04 1.12
95% mean confidence interval for instructions %-change: 0.29% 0.36%
Instructions are HURT.
total cycles in shared programs: 153827268 -> 153828052 (<.01%)
cycles in affected programs: 1669290 -> 1670074 (0.05%)
helped: 24
HURT: 84
helped stats (abs) min: 2 max: 232 x̄: 64.33 x̃: 67
helped stats (rel) min: 0.04% max: 4.62% x̄: 1.60% x̃: 1.94%
HURT stats (abs) min: 2 max: 48 x̄: 27.71 x̃: 24
HURT stats (rel) min: 0.02% max: 2.66% x̄: 0.34% x̃: 0.14%
95% mean confidence interval for cycles value: -1.94 16.46
95% mean confidence interval for cycles %-change: -0.29% 0.11%
Inconclusive result (value mean confidence interval includes 0).
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12191>
2021-08-02 21:33:17 -07:00
|
|
|
(can_omit_write(inst) || can_eliminate(devinfo, inst, flag_live))) {
|
2019-07-13 12:06:00 -05:00
|
|
|
inst->dst = fs_reg(spread(retype(brw_null_reg(), inst->dst.type),
|
|
|
|
|
inst->dst.stride));
|
2014-04-12 17:40:18 -07:00
|
|
|
progress = true;
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
intel/fs: sel.cond writes the flags on Gfx4 and Gfx5
On Gfx4 and Gfx5, sel.l (for min) and sel.ge (for max) are implemented
using a separte cmpn and sel instruction. This lowering occurs in
fs_vistor::lower_minmax which is called very, very late... a long, long
time after the first calls to opt_cmod_propagation. As a result,
conditional modifiers can be incorrectly propagated across sel.cond on
those platforms.
No tests were affected by this change, and I find that quite shocking.
After just changing flags_written(), all of the atan tests started
failing on ILK. That required the change in cmod_propagatin (and the
addition of the prop_across_into_sel_gfx5 unit test).
Shader-db results for ILK and GM45 are below. I looked at a couple
before and after shaders... and every case that I looked at had
experienced incorrect cmod propagation. This affected a LOT of apps!
Euro Truck Simulator 2, The Talos Principle, Serious Sam 3, Sanctum 2,
Gang Beasts, and on and on... :(
I discovered this bug while working on a couple new optimization
passes. One of the passes attempts to remove condition modifiers that
are never used. The pass made no progress except on ILK and GM45.
After investigating a couple of the affected shaders, I noticed that
the code in those shaders looked wrong... investigation led to this
cause.
v2: Trivial changes in the unit tests.
v3: Fix type in comment in unit tests. Noticed by Jason and Priit.
v4: Tweak handling of BRW_OPCODE_SEL special case. Suggested by Jason.
Fixes: df1aec763eb ("i965/fs: Define methods to calculate the flag subset read or written by an fs_inst.")
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Tested-by: Dave Airlie <airlied@redhat.com>
Iron Lake
total instructions in shared programs: 8180493 -> 8181781 (0.02%)
instructions in affected programs: 541796 -> 543084 (0.24%)
helped: 28
HURT: 1158
helped stats (abs) min: 1 max: 1 x̄: 1.00 x̃: 1
helped stats (rel) min: 0.35% max: 0.86% x̄: 0.53% x̃: 0.50%
HURT stats (abs) min: 1 max: 3 x̄: 1.14 x̃: 1
HURT stats (rel) min: 0.12% max: 4.00% x̄: 0.37% x̃: 0.23%
95% mean confidence interval for instructions value: 1.06 1.11
95% mean confidence interval for instructions %-change: 0.31% 0.38%
Instructions are HURT.
total cycles in shared programs: 239420470 -> 239421690 (<.01%)
cycles in affected programs: 2925992 -> 2927212 (0.04%)
helped: 49
HURT: 157
helped stats (abs) min: 2 max: 284 x̄: 62.69 x̃: 70
helped stats (rel) min: 0.04% max: 6.20% x̄: 1.68% x̃: 1.96%
HURT stats (abs) min: 2 max: 48 x̄: 27.34 x̃: 24
HURT stats (rel) min: 0.02% max: 2.91% x̄: 0.31% x̃: 0.20%
95% mean confidence interval for cycles value: -0.80 12.64
95% mean confidence interval for cycles %-change: -0.31% <.01%
Inconclusive result (value mean confidence interval includes 0).
GM45
total instructions in shared programs: 4985517 -> 4986207 (0.01%)
instructions in affected programs: 306935 -> 307625 (0.22%)
helped: 14
HURT: 625
helped stats (abs) min: 1 max: 1 x̄: 1.00 x̃: 1
helped stats (rel) min: 0.35% max: 0.82% x̄: 0.52% x̃: 0.49%
HURT stats (abs) min: 1 max: 3 x̄: 1.13 x̃: 1
HURT stats (rel) min: 0.12% max: 3.90% x̄: 0.34% x̃: 0.22%
95% mean confidence interval for instructions value: 1.04 1.12
95% mean confidence interval for instructions %-change: 0.29% 0.36%
Instructions are HURT.
total cycles in shared programs: 153827268 -> 153828052 (<.01%)
cycles in affected programs: 1669290 -> 1670074 (0.05%)
helped: 24
HURT: 84
helped stats (abs) min: 2 max: 232 x̄: 64.33 x̃: 67
helped stats (rel) min: 0.04% max: 4.62% x̄: 1.60% x̃: 1.94%
HURT stats (abs) min: 2 max: 48 x̄: 27.71 x̃: 24
HURT stats (rel) min: 0.02% max: 2.66% x̄: 0.34% x̃: 0.14%
95% mean confidence interval for cycles value: -1.94 16.46
95% mean confidence interval for cycles %-change: -0.29% 0.11%
Inconclusive result (value mean confidence interval includes 0).
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12191>
2021-08-02 21:33:17 -07:00
|
|
|
if (inst->dst.is_null() && can_eliminate(devinfo, inst, flag_live)) {
|
2014-08-22 11:01:13 -07:00
|
|
|
inst->opcode = BRW_OPCODE_NOP;
|
|
|
|
|
progress = true;
|
|
|
|
|
}
|
|
|
|
|
|
2015-10-26 17:09:25 -07:00
|
|
|
if (inst->dst.file == VGRF) {
|
2019-04-24 12:38:28 +02:00
|
|
|
if (!inst->is_partial_write()) {
|
2016-03-13 16:25:57 -07:00
|
|
|
const unsigned var = live_vars.var_from_reg(inst->dst);
|
2016-09-07 16:59:35 -07:00
|
|
|
for (unsigned i = 0; i < regs_written(inst); i++) {
|
2014-09-08 18:34:28 -07:00
|
|
|
BITSET_CLEAR(live, var + i);
|
2014-04-14 11:17:13 -07:00
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
2016-05-18 21:34:27 -07:00
|
|
|
if (!inst->predicate && inst->exec_size >= 8)
|
intel/fs: sel.cond writes the flags on Gfx4 and Gfx5
On Gfx4 and Gfx5, sel.l (for min) and sel.ge (for max) are implemented
using a separte cmpn and sel instruction. This lowering occurs in
fs_vistor::lower_minmax which is called very, very late... a long, long
time after the first calls to opt_cmod_propagation. As a result,
conditional modifiers can be incorrectly propagated across sel.cond on
those platforms.
No tests were affected by this change, and I find that quite shocking.
After just changing flags_written(), all of the atan tests started
failing on ILK. That required the change in cmod_propagatin (and the
addition of the prop_across_into_sel_gfx5 unit test).
Shader-db results for ILK and GM45 are below. I looked at a couple
before and after shaders... and every case that I looked at had
experienced incorrect cmod propagation. This affected a LOT of apps!
Euro Truck Simulator 2, The Talos Principle, Serious Sam 3, Sanctum 2,
Gang Beasts, and on and on... :(
I discovered this bug while working on a couple new optimization
passes. One of the passes attempts to remove condition modifiers that
are never used. The pass made no progress except on ILK and GM45.
After investigating a couple of the affected shaders, I noticed that
the code in those shaders looked wrong... investigation led to this
cause.
v2: Trivial changes in the unit tests.
v3: Fix type in comment in unit tests. Noticed by Jason and Priit.
v4: Tweak handling of BRW_OPCODE_SEL special case. Suggested by Jason.
Fixes: df1aec763eb ("i965/fs: Define methods to calculate the flag subset read or written by an fs_inst.")
Reviewed-by: Jason Ekstrand <jason@jlekstrand.net>
Tested-by: Dave Airlie <airlied@redhat.com>
Iron Lake
total instructions in shared programs: 8180493 -> 8181781 (0.02%)
instructions in affected programs: 541796 -> 543084 (0.24%)
helped: 28
HURT: 1158
helped stats (abs) min: 1 max: 1 x̄: 1.00 x̃: 1
helped stats (rel) min: 0.35% max: 0.86% x̄: 0.53% x̃: 0.50%
HURT stats (abs) min: 1 max: 3 x̄: 1.14 x̃: 1
HURT stats (rel) min: 0.12% max: 4.00% x̄: 0.37% x̃: 0.23%
95% mean confidence interval for instructions value: 1.06 1.11
95% mean confidence interval for instructions %-change: 0.31% 0.38%
Instructions are HURT.
total cycles in shared programs: 239420470 -> 239421690 (<.01%)
cycles in affected programs: 2925992 -> 2927212 (0.04%)
helped: 49
HURT: 157
helped stats (abs) min: 2 max: 284 x̄: 62.69 x̃: 70
helped stats (rel) min: 0.04% max: 6.20% x̄: 1.68% x̃: 1.96%
HURT stats (abs) min: 2 max: 48 x̄: 27.34 x̃: 24
HURT stats (rel) min: 0.02% max: 2.91% x̄: 0.31% x̃: 0.20%
95% mean confidence interval for cycles value: -0.80 12.64
95% mean confidence interval for cycles %-change: -0.31% <.01%
Inconclusive result (value mean confidence interval includes 0).
GM45
total instructions in shared programs: 4985517 -> 4986207 (0.01%)
instructions in affected programs: 306935 -> 307625 (0.22%)
helped: 14
HURT: 625
helped stats (abs) min: 1 max: 1 x̄: 1.00 x̃: 1
helped stats (rel) min: 0.35% max: 0.82% x̄: 0.52% x̃: 0.49%
HURT stats (abs) min: 1 max: 3 x̄: 1.13 x̃: 1
HURT stats (rel) min: 0.12% max: 3.90% x̄: 0.34% x̃: 0.22%
95% mean confidence interval for instructions value: 1.04 1.12
95% mean confidence interval for instructions %-change: 0.29% 0.36%
Instructions are HURT.
total cycles in shared programs: 153827268 -> 153828052 (<.01%)
cycles in affected programs: 1669290 -> 1670074 (0.05%)
helped: 24
HURT: 84
helped stats (abs) min: 2 max: 232 x̄: 64.33 x̃: 67
helped stats (rel) min: 0.04% max: 4.62% x̄: 1.60% x̃: 1.94%
HURT stats (abs) min: 2 max: 48 x̄: 27.71 x̃: 24
HURT stats (rel) min: 0.02% max: 2.66% x̄: 0.34% x̃: 0.14%
95% mean confidence interval for cycles value: -1.94 16.46
95% mean confidence interval for cycles %-change: -0.29% 0.11%
Inconclusive result (value mean confidence interval includes 0).
Part-of: <https://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/12191>
2021-08-02 21:33:17 -07:00
|
|
|
flag_live[0] &= ~inst->flags_written(devinfo);
|
2014-10-28 20:12:21 -07:00
|
|
|
|
2015-11-30 09:25:19 -08:00
|
|
|
if (inst->opcode == BRW_OPCODE_NOP) {
|
2021-06-28 19:02:40 -07:00
|
|
|
inst->remove(block, true);
|
2015-11-25 16:19:06 -08:00
|
|
|
continue;
|
2015-11-30 09:25:19 -08:00
|
|
|
}
|
2015-11-25 16:19:06 -08:00
|
|
|
|
2014-03-17 10:39:43 -07:00
|
|
|
for (int i = 0; i < inst->sources; i++) {
|
2015-10-26 17:09:25 -07:00
|
|
|
if (inst->src[i].file == VGRF) {
|
2016-03-13 16:25:57 -07:00
|
|
|
int var = live_vars.var_from_reg(inst->src[i]);
|
2014-04-12 17:40:18 -07:00
|
|
|
|
2016-09-07 16:59:35 -07:00
|
|
|
for (unsigned j = 0; j < regs_read(inst, i); j++) {
|
2014-09-08 18:34:28 -07:00
|
|
|
BITSET_SET(live, var + j);
|
2014-04-12 17:40:18 -07:00
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
}
|
2014-10-28 20:12:21 -07:00
|
|
|
|
2016-05-18 21:34:27 -07:00
|
|
|
flag_live[0] |= inst->flags_read(devinfo);
|
2014-04-12 17:40:18 -07:00
|
|
|
}
|
|
|
|
|
}
|
|
|
|
|
|
2024-01-03 11:08:36 -08:00
|
|
|
s.cfg->adjust_block_ips();
|
2021-06-28 19:02:40 -07:00
|
|
|
|
2014-04-12 17:40:18 -07:00
|
|
|
ralloc_free(live);
|
2014-10-28 20:12:21 -07:00
|
|
|
ralloc_free(flag_live);
|
2014-04-12 17:40:18 -07:00
|
|
|
|
2015-11-30 09:25:19 -08:00
|
|
|
if (progress)
|
2024-01-03 11:08:36 -08:00
|
|
|
s.invalidate_analysis(DEPENDENCY_INSTRUCTIONS);
|
2014-04-12 17:40:18 -07:00
|
|
|
|
|
|
|
|
return progress;
|
|
|
|
|
}
|