summaryrefslogtreecommitdiffstats
path: root/src/compiler/nir
Commit message (Collapse)AuthorAgeFilesLines
* nir: Drop an unused program/hash_table.h include.Eric Anholt2016-08-101-1/+0
| | | | | Reviewed-by: Ian Romanick <[email protected]> Reviewed-by: Timothy Arceri <[email protected]>
* nir: make use of nir_cf_list_extract() helperTimothy Arceri2016-08-091-2/+1
| | | | Reviewed-by: Kenneth Graunke <[email protected]>
* nir: Always print non-identity swizzles.Matt Turner2016-08-081-1/+9
| | | | | | | | | | | | | | | | | | | | | | | Previously we would not print a swizzle on ssa_52 when only its .x component is used (as seen in the definition of ssa_53): vec3 ssa_52 = fadd ssa_51, ssa_51 vec1 ssa_53 = flog2 ssa_52 vec1 ssa_54 = flog2 ssa_52.y vec1 ssa_55 = flog2 ssa_52.z But this makes the interpretation of the RHS of the definition difficult to understand and dependent on the size of the LHS. Just print swizzles when they are not the identity swizzle, so the previous example is now printed as: vec3 ssa_52 = fadd ssa_51.xyz, ssa_51.xyz vec1 ssa_53 = flog2 ssa_52.x vec1 ssa_54 = flog2 ssa_52.y vec1 ssa_55 = flog2 ssa_52.z Reviewed-by: Timothy Arceri <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]>
* nir: Make nir_opt_remove_phis see through moves.Kenneth Graunke2016-08-041-1/+23
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | I found a shader in Tales of Maj'Eyal that contains: if ssa_21 { block block_1: /* preds: block_0 */ ...instructions that prevent the select peephole... vec1 32 ssa_23 = imov ssa_4 vec1 32 ssa_24 = imov ssa_4.y vec1 32 ssa_25 = imov ssa_4.z /* succs: block_3 */ } else { block block_2: /* preds: block_0 */ vec1 32 ssa_26 = imov ssa_4 vec1 32 ssa_27 = imov ssa_4.y vec1 32 ssa_28 = imov ssa_4.z /* succs: block_3 */ } block block_3: /* preds: block_1 block_2 */ vec1 32 ssa_29 = phi block_1: ssa_23, block_2: ssa_26 vec1 32 ssa_30 = phi block_1: ssa_24, block_2: ssa_27 vec1 32 ssa_31 = phi block_1: ssa_25, block_2: ssa_28 Here, copy propagation will bail because phis cannot perform swizzles, and CSE won't do anything because there is no dominance relationship between the imovs. By making nir_opt_remove_phis handle identical moves, we can eliminate the phis and rewrite everything to use ssa_4 directly, so all the moves become dead and get eliminated. I don't think we need to check "exact" - just the alu sources. Presumably phi sources should match in their exactness. On Broadwell: total instructions in shared programs: 11639872 -> 11638535 (-0.01%) instructions in affected programs: 134222 -> 132885 (-1.00%) helped: 338 HURT: 0 v2: Fix return value to be NULL, not false (caught by Iago). Signed-off-by: Kenneth Graunke <[email protected]> Reviewed-by: Iago Toral Quiroga <[email protected]>
* nir: Make nir_alu_srcs_equal non-static.Kenneth Graunke2016-08-042-1/+4
| | | | | Signed-off-by: Kenneth Graunke <[email protected]> Reviewed-by: Iago Toral Quiroga <[email protected]>
* nir: Turn imov/fmov of undef into undef.Kenneth Graunke2016-08-041-1/+3
| | | | | | | | | | | | | | | | | | | | | | On Broadwell: total instructions in shared programs: 11640214 -> 11639872 (-0.00%) instructions in affected programs: 17744 -> 17402 (-1.93%) helped: 78 HURT: 0 total spills in shared programs: 2924 -> 2922 (-0.07%) spills in affected programs: 104 -> 102 (-1.92%) helped: 1 HURT: 0 total fills in shared programs: 4394 -> 4389 (-0.11%) fills in affected programs: 237 -> 232 (-2.11%) helped: 1 HURT: 0 Signed-off-by: Kenneth Graunke <[email protected]> Reviewed-by: Iago Toral Quiroga <[email protected]>
* nir: Allow opt_peephole_select to work on empty blocks.Eric Anholt2016-08-031-7/+0
| | | | | | | | | | | | | | nir_opt_peephole_select has the job of removing IF statements with no side effects. However, if the IF statement's successor didn't have any instructions in it, we were skipping it, which occurred in mupen64 on vc4 with glsl_to_nir enabled: instructions in affected programs: 6134 -> 4120 (-32.83%) total uniforms in shared programs: 38268 -> 38219 (-0.13%) No changes on Haswell shader-db. Reviewed-by: Jason Ekstrand <[email protected]>
* nir: fix validation messageTimothy Arceri2016-08-031-2/+2
| | | | | | | Looks like a copy and paste error from f752effa087 Reviewed-by: Jason Ekstrand <[email protected]> Reviewed-by: Alejandro Piñeiro <[email protected]>
* glsl: Separate overlapping sentinel nodes in exec_list.Matt Turner2016-07-262-3/+3
| | | | | | | | | | | I do appreciate the cleverness, but unfortunately it prevents a lot more cleverness in the form of additional compiler optimizations brought on by -fstrict-aliasing. No difference in OglBatch7 (n=20). Co-authored-by: Davin McCall <[email protected]> Reviewed-by: Ian Romanick <[email protected]>
* nir: Lower interp_var_at_* like a normal load_var for flat inputs.Kenneth Graunke2016-07-221-0/+4
| | | | | | | | | | "flat centroid" and "flat sample" both just mean "flat", so we should ignore interpolateAtCentroid/Sample and just return the flat value. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=97032 Signed-off-by: Kenneth Graunke <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]> Reviewed-by: Anuj Phogat <[email protected]>
* nir/lower_tex: Add support for lowering coordinate offsetsJason Ekstrand2016-07-222-0/+64
| | | | | | | | | | On i965, we can't support coordinate offsets for texelFetch or rectangle textures. Previously, we were doing this with a GLSL pass but we need to do it in NIR if we want those workarounds for SPIR-V. Signed-off-by: Jason Ekstrand <[email protected]> Reviewed-by: Kenneth Graunke <[email protected]> Cc: "12.0" <[email protected]>
* nir/lower_tex: Add some helpers for working with tex sourcesJason Ekstrand2016-07-221-16/+30
| | | | | | Signed-off-by: Jason Ekstrand <[email protected]> Reviewed-by: Kenneth Graunke <[email protected]> Cc: "12.0" <[email protected]>
* nir: Add a helper for determining the type of a texture sourceJason Ekstrand2016-07-221-0/+44
| | | | | | Signed-off-by: Jason Ekstrand <[email protected]> Reviewed-by: Kenneth Graunke <[email protected]> Cc: "12.0" <[email protected]>
* nir: Add a base const_index to shared atomic intrinsics.Kenneth Graunke2016-07-211-10/+10
| | | | | | | | | | | | | | | | | | | Commit 52e75dcb8c04c0dde989970c4c587cbe8313f7cf made nir_lower_io start using nir_intrinsic_set_base instead of writing const_index[0] directly. However, those intrinsics apparently don't /have/ a base, so this caused assert failures. However, the old code was happily setting non-existent const_index fields, so it was pretty bogus too. Jason pointed out that load_shared and store_shared have a base, and that the i965 driver uses that field. So presumably atomics should have one as well, so that loads/stores/atomics all refer to variables with consistent addressing. Cc: "12.0" <[email protected]> Signed-off-by: Kenneth Graunke <[email protected]> Reviewed-by: Timothy Arceri <[email protected]>
* nir: add doubles component packing supportTimothy Arceri2016-07-211-0/+20
| | | | | | | | | | | This makes sure we give the correct driver location for doubles when using component packing. Specifically it handles packing a dvec3 with a double which is the only packing scenario allowed which spans across two locations. Acked-by: Kenneth Graunke <[email protected]> Reviewed-by: Alejandro Piñeiro <[email protected]>
* nir/inline: Constant-initialize local variables in the callee if neededJason Ekstrand2016-07-201-2/+40
| | | | | | Signed-off-by: Jason Ekstrand <[email protected]> Reviewed-by: Kenneth Graunke <[email protected]> Cc: "12.0" <[email protected]>
* nir: Add a nir_deref_foreach_leaf helperJason Ekstrand2016-07-202-0/+120
| | | | | | Signed-off-by: Jason Ekstrand <[email protected]> Reviewed-by: Kenneth Graunke <[email protected]> Cc: "12.0" <[email protected]>
* nir: Add nir_load_interpolated_input lowering code.Kenneth Graunke2016-07-202-5/+100
| | | | | | | | | | | | | | Now nir_lower_io can optionally produce load_interpolated_input and load_barycentric_* intrinsics for fragment shader inputs. flat inputs continue using regular load_input. v2: Use a nir_shader_compiler_options flag rather than ad-hoc boolean passing (in response to review feedback from Chris Forbes). Signed-off-by: Kenneth Graunke <[email protected]> Reviewed-by: Chris Forbes <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]>
* nir: Add new intrinsics for fragment shader input interpolation.Kenneth Graunke2016-07-205-0/+51
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Backends can normally handle shader inputs solely by looking at load_input intrinsics, and ignore the nir_variables in nir->inputs. One exception is fragment shader inputs. load_input doesn't capture the necessary interpolation information - flat, smooth, noperspective mode, and centroid, sample, or pixel for the location. This means that backends have to interpolate based on the nir_variables, then associate those with the load_input intrinsics (say, by storing a map of which variables are at which locations). With GL_ARB_enhanced_layouts, we're going to have multiple varyings packed into a single vec4 location. The intrinsics make this easy: simply load N components from location <loc, component>. However, working with variables and correlating the two is very awkward; we'd much rather have intrinsics capture all the necessary information. Fragment shader input interpolation typically works by producing a set of barycentric coordinates, then using those to do a linear interpolation between the values at the triangle's corners. We represent this by introducing five new load_barycentric_* intrinsics: - load_barycentric_pixel (ordinary variable) - load_barycentric_centroid (centroid qualified variable) - load_barycentric_sample (sample qualified variable) - load_barycentric_at_sample (ARB_gpu_shader5's interpolateAtSample()) - load_barycentric_at_offset (ARB_gpu_shader5's interpolateAtOffset()) Each of these take the interpolation mode (smooth or noperspective only) as a const_index, and produce a vec2. The last two also take a sample or offset source. We then introduce a new load_interpolated_input intrinsic, which is like a normal load_input intrinsic, but with an additional barycentric coordinate source. The intention is that flat inputs will still use regular load_input intrinsics. This makes them distinguishable from normal inputs that need fancy interpolation, while also providing all the necessary data. This nicely unifies regular inputs and interpolateAt functions. Qualifiers and variables become irrelevant; there are just load_barycentric intrinsics that determine the interpolation. v2: Document the interp_mode const_index value, define a new BARYCENTRIC() helper rather than using SYSTEM_VALUE() for some of them (requested by Jason Ekstrand). Signed-off-by: Kenneth Graunke <[email protected]> Reviewed-by: Chris Forbes <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]>
* nir: Fix uninitialized use of 'replacement'.Kenneth Graunke2016-07-191-1/+1
| | | | | | | | | | | For intrinsics we don't care about, just skip to the next loop iteration and process the next instruction. We don't want to execute the rest of the code. This was a bug in commit cdfc05ea6e8c87876cdbf588aa8e03d70f3da4bb. Signed-off-by: Kenneth Graunke <[email protected]> Reviewed-by: Anuj Phogat <[email protected]>
* nir/algebraic: Optimize fabs(u2f(x))Ian Romanick2016-07-191-0/+1
| | | | | | | | | | I noticed this when I tried to do frexp(float(some_unsigned)) in the ir_unop_find_lsb lowering pass. The code generated for frexp() uses fabs, and this resulted in an extra instruction. Ultimately I ended up not using frexp. Signed-off-by: Ian Romanick <[email protected]> Reviewed-by: Matt Turner <[email protected]>
* compiler: Rename INTERP_QUALIFIER_* to INTERP_MODE_*.Kenneth Graunke2016-07-174-4/+4
| | | | | | | | | | | | | | | | | Likewise, rename the enum type to glsl_interp_mode. Beyond the GLSL front-end, talking about "interpolation modes" seems more natural than "interpolation qualifiers" - in the IR, we're removed from how exactly the source language specifies how to interpolate an input. Also, SPIR-V calls these "decorations" rather than "qualifiers". Generated by: $ find . -regextype egrep -regex '.*\.(c|cpp|h)' -type f -exec sed -i \ -e 's/INTERP_QUALIFIER_/INTERP_MODE_/g' \ -e 's/glsl_interp_qualifier/glsl_interp_mode/g' {} \; Signed-off-by: Kenneth Graunke <[email protected]> Acked-by: Dave Airlie <[email protected]>
* nir: Use dest.ssa.num_components rather than intrin->num_components.Kenneth Graunke2016-07-151-1/+1
| | | | | | | | | I recently refactored this to share code between load and atomic lowering. loads used intrin->num_components, while atomics used intrin->dest.ssa.num_components. They should be equivalent, but Jason wanted me to use the latter. I missed applying his review. Signed-off-by: Kenneth Graunke <[email protected]>
* nir: Update outdated intrinsic const_index comments.Kenneth Graunke2016-07-151-7/+9
| | | | | Signed-off-by: Kenneth Graunke <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]>
* nir: Use nir_intrinsic_set_base in atomic lowering.Kenneth Graunke2016-07-151-2/+2
| | | | | | | | | | This is more readable and also offers assertions that protect against setting const_index fields on the wrong kind of intrinsic. Suggested by Jason. Signed-off-by: Kenneth Graunke <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]>
* nir: Split nir_lower_io's input/output/atomic handling into helpers.Kenneth Graunke2016-07-151-91/+87
| | | | | | | | | | | | | | | The original function was becoming a bit hard to read, with the details of creating and filling out load/store/atomic atomics all in one function. This patch makes helpers for creating each type of intrinsic, and also combines them with the *_op() helpers, as they're closely coupled and not too large. v2: Minor style nits from Jason. Signed-off-by: Kenneth Graunke <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]>
* nir: Drop bogus nir_var_shader_in case in nir_lower_io's store_op().Kenneth Graunke2016-07-151-1/+0
| | | | | | | This can't happen, the caller asserts that mode is shader_out or shared. Signed-off-by: Kenneth Graunke <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]>
* nir: Share destination rewriting and replacement code in IO lowering.Kenneth Graunke2016-07-151-25/+19
| | | | | | | | Both loads and atomics had identical code to rewrite destinations, and all cases had the same two lines to replace instructions. Signed-off-by: Kenneth Graunke <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]>
* nir: Share get_io_offset handling in nir_lower_io.Kenneth Graunke2016-07-151-24/+9
| | | | | | | | | | | | The load/store/atomic cases all duplicated the get_io_offset code, with a few tiny differences: stores didn't bother checking for per-vertex inputs, because they can't be stored to, and atomics didn't check at all, since shared variables aren't per-vertex. However, it's harmless to check, and allows us to share more code. Signed-off-by: Kenneth Graunke <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]>
* nir: Make a 'var' temporary in nir_lower_io.Kenneth Graunke2016-07-151-16/+12
| | | | | | | Less typing and word wrapping issues than intrin->variables[0]->var. Signed-off-by: Kenneth Graunke <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]>
* nir: Add optimization for (a || True == True)Eric Anholt2016-07-121-0/+1
| | | | | | | | | | | | This was appearing in vc4 VS/CS in mupen64, due to vertex attrib lowering producing some constants that were getting compared. total instructions in shared programs: 112276 -> 112198 (-0.07%) instructions in affected programs: 2239 -> 2161 (-3.48%) total estimated cycles in shared programs: 283102 -> 283038 (-0.02%) estimated cycles in affected programs: 2365 -> 2301 (-2.71%) Reviewed-by: Jason Ekstrand <[email protected]>
* nir: use the same driver location for packed varyingsTimothy Arceri2016-07-072-4/+28
| | | | Reviewed-by: Kenneth Graunke <[email protected]>
* nir: add new intrinsic field for storing component offsetTimothy Arceri2016-07-074-6/+23
| | | | | | This offset is used for packing. Reviewed-by: Kenneth Graunke <[email protected]>
* nir: Fix copy_prop_src when src is an indirect access on a reg.Eric Anholt2016-06-261-1/+1
| | | | | | | | | | | | The intent was to continue down the indirect chain, not to call ourselves with unchanged input arguments. Found by code inspection, and comparison to copy_prop_alu_src(). We haven't hit this because callers of NIR's copy prop are doing so in SSA, before indirect variable dereferences have been lowered to registers. Reviewed-by: Rob Clark <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]>
* nir: Add a NIR_VALIDATE environment variableJason Ekstrand2016-06-251-0/+6
| | | | | | | | | | | It defaults to true so default behavior doesn't change but it allows you to do NIR_VALIDATE=false if you don't want validation. Disabling validation can substantially speed up shader compiles so you frequently want to turn it off if compiler invariants aren't in question. Reviewed-by: Matt Turner <[email protected]> Reviewed-by: Matt Turner <[email protected]> Signed-off-by: Rob Clark <[email protected]>
* Remove wrongly repeated words in commentsGiuseppe Bilotta2016-06-236-7/+7
| | | | | | | | | | | | | | | | | Clean up misrepetitions ('if if', 'the the' etc) found throughout the comments. This has been done manually, after grepping case-insensitively for duplicate if, is, the, then, do, for, an, plus a few other typos corrected in fly-by v2: * proper commit message and non-joke title; * replace two 'as is' followed by 'is' to 'as-is'. v3: * 'a integer' => 'an integer' and similar (originally spotted by Jason Ekstrand, I fixed a few other similar ones while at it) Signed-off-by: Giuseppe Bilotta <[email protected]> Reviewed-by: Chad Versace <[email protected]>
* nir/alu_to_scalar: Respect the exact ALU operation qualifierJason Ekstrand2016-06-201-0/+2
| | | | | | | | | | Just setting builder->exact isn't sufficient because that only applies to instructions that are built with the builder but instructions created manually and only inserted using the builder are left alone. Signed-off-by: Jason Ekstrand <[email protected]> Reviewed-by: Kenneth Graunke <[email protected]> Cc: "12.0" <[email protected]>
* nir: Add a pass for propagating invariant decorationsJason Ekstrand2016-06-202-0/+198
| | | | | | | | | | This pass is similar to propagate_invariance in the GLSL compiler. The real "output" of this pass is that any algebraic operations which are eventually consumed by an invariant variable get marked as "exact". Signed-off-by: Jason Ekstrand <[email protected]> Reviewed-by: Kenneth Graunke <[email protected]> Cc: "12.0" <[email protected]>
* nir/algebraic: Remove imprecise flog2 optimizationsJason Ekstrand2016-06-201-2/+0
| | | | | | | | | | | | | | | | | | | | | | | | While mathematically correct, these two optimizations result in an expression with substantially lower precision than the original. For any positive finite floating-point value, log2(x) is well-defined and finite. More precisely, it is in the range [-150, 150] so any sum of logarithms log2(a) + log2(b) is also well-defined and finite as long as a and b are both positive and finite. However, if a and b are either very small or very large, their product may get flushed to infinity or zero causing log2(a * b) to be nowhere close to log2(a) + log2(b). This imprecision was causing incorrect rendering in Talos Principal because part of its HDR rendering process involves doing 8 texture operations, clamping the result to [0, 65000], taking a dot-product with a constant, and then taking the log2. This is done 6 or 8 times and summed to produce the final result which is written to a red texture. In cases where you have a region of the screen that is very dark, it can end up getting a result value of -inf which is not what is intended. Signed-off-by: Jason Ekstrand <[email protected]> Reviewed-by: Matt Turner <[email protected]> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=96425 Cc: "11.1 11.2 12.0" <[email protected]>
* nir/info: Get rid of uses_interp_var_at_offsetJason Ekstrand2016-06-032-7/+0
| | | | | | | | | We were using this briefly in the i965 driver to trigger recompiles but we haven't been using it since we switched to the NIR y-transform lowering pass. Signed-off-by: Jason Ekstrand <[email protected]> Reviewed-by: Kenneth Graunke <[email protected]>
* nir/lower_indirect_derefs: Use the direct array deref for recursionJason Ekstrand2016-06-031-1/+1
| | | | | | | | This fixes about 100 of the new Vulkan CTS tests. Signed-off-by: Jason Ekstrand <[email protected]> Reviewed-by: Kenneth Graunke <[email protected]> Cc: "12.0" <[email protected]>
* nir/algebraic: support for power-of-two optimizationsRob Clark2016-06-036-5/+128
| | | | | | | | | | | | Some optimizations, like converting integer multiply/divide into left/ right shifts, have additional constraints on the search expression. Like requiring that a variable is a constant power of two. Support these cases by allowing a fxn name to be appended to the search var expression (ie. "a#32(is_power_of_two)"). Signed-off-by: Rob Clark <[email protected]> Reviewed-by: Kenneth Graunke <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]>
* i965: Add nir channel_num system valueJordan Justen2016-06-011-0/+1
| | | | | | | | | v2: * simd16/32 fixes (curro) Cc: "12.0" <[email protected]> Signed-off-by: Jordan Justen <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]>
* nir: Make lowering gl_LocalInvocationIndex optionalJordan Justen2016-06-015-4/+20
| | | | | | Cc: "12.0" <[email protected]> Signed-off-by: Jordan Justen <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]>
* nir: allow sat on all float destination typesIlia Mirkin2016-06-011-1/+3
| | | | | | | | | | | With the introduction of fp64 and fp16 to nir, there are now a bunch of float types running around. A F1 2015 shader ends up with an i2f.sat operation, which has a nir_type_float32 destination. Allow sat on all the float destination types. Signed-off-by: Ilia Mirkin <[email protected]> Cc: "12.0" <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]>
* nir/inline: Also rewrite param derefs for texture instructionsJason Ekstrand2016-05-271-6/+20
| | | | | | | Without this, samplers get left hanging as derefs to variables that don't actually exist. Reviewed-by: Connor Abbott <[email protected]>
* nir/inline: Break the guts of rewrite_param-derefs into a helperJason Ekstrand2016-05-271-19/+30
| | | | Reviewed-by: Connor Abbott <[email protected]>
* nir/inline: Make the rewrite_param_derefs helper work on instructionsJason Ekstrand2016-05-271-28/+25
| | | | | | | Now that we have the better nir_foreach_block macro, there's no reason to use the archaic block version for everything. Reviewed-by: Connor Abbott <[email protected]>
* nir/inline: Don't use foreach_instr_safe unless we need toJason Ekstrand2016-05-271-2/+2
| | | | Suggested-by: Connor Abbott <[email protected]>
* nir: Make nir_const_value a unionJason Ekstrand2016-05-262-10/+8
| | | | | | | | There's no good reason for it to be a struct of an anonymous union. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=96221 Tested-by: Vinson Lee <[email protected]> Reviewed-by: Matt Turner <[email protected]>
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-26 15:59:35 +0000