Commit Graph

1348 Commits

Author SHA1 Message Date
greg-lunarg
92c41ff1e7 Remove Common Uniform Elimination Pass (#2731)
Remove Common Uniform Elimination Pass

Fixes #2520.
2019-07-12 11:02:10 -04:00
Ryan Harrison
55adf4cf70
Update execution scope rules for WebGPU (#2730)
Fixes #2722
2019-07-11 14:37:36 -04:00
alan-baker
1a2de48a12 Extra small storage validation (#2732)
Fixes #2729

* Check acceptable uses of small type generators
2019-07-11 13:05:14 -04:00
Jeff Bolz
327963765b Add validation for SPV_EXT_demote_to_helper_invocation (#2707) 2019-07-11 10:33:22 -04:00
Steven Perron
5ce8cf781f
Change the order branches are simplified in dead branch elim (#2728)
Dead branch elimination needs to know about the constructs that a block is contained it when determining what to do with its merge instruction.  We currently fold branches in block as we see them, which is parent constructs before their children.  This causes the struct cfg analysis to crash because it tries to get the parent construct for a block after the parent has been folded.

This can be fixed by folding the branch of the children before the parents.

Fixes #2667.
2019-07-10 14:59:44 -04:00
Thomas Roughton
cd153db8ed Add —preserve-bindings and —preserve-spec-constants (#2693)
Add optimizer options to for preservation of spec constants and variable with
binding decorations.  They are to be preserved even if they are unused.
2019-07-10 14:12:19 -04:00
Steven Perron
86e45efe15
Handle decorations better in some optimizations (#2716)
There are a couple spots where we are not looking at decorations when we should.

1. Value numbering is suppose to assign a different value number to ids if they have different decorations.  However that is not being done for OpCopyObject and OpPhi.

1. Instruction simplification is propagating OpCopyObject instruction without checking for decorations.  It should only do that if no decorations are being lost.

Add a new function to the decoration manager to check if the decorations of one id are a subset of the decorations of another.

Fixes #2715.
2019-07-10 11:37:16 -04:00
Ryan Harrison
3a252a267b
Update memory scope rules for WebGPU (#2725)
Fixes #2721
2019-07-10 10:34:50 -04:00
alan-baker
0c4feb643b
Remove extra semis (#2717)
* Remove extra semi-colons
* Update re2 dep
2019-07-08 15:07:36 -04:00
alan-baker
456cc598af
Validate usage of 8- and 16-bit types with only storage capabilities (#2704)
Fixes #2669

* Check capabilities when validating variables
* validate load and store types
* Constant check
* Don't checks pointers for stores, constants and loads
* Validate composite instructions
* Validate conversions for 8- and 16-bit limited types
* Unified tests and expanded them
* Disallow OpCopyMemory
* new tests and update old tests
2019-07-08 14:10:13 -04:00
Alastair Donaldson
b8ab80843f
Shrinker for spirv-fuzz (#2708)
Adds to spirv-fuzz the option to shrink a sequence of transformations
that lead to an interesting binary to be generated, to find a smaller
sub-sequence of transformations that still lead to an interesting (but
hopefully simpler) binary being generated. The notion of what counts
as "interesting" comes from a user-provided script, the
"interestingness function", similar to the way the spirv-reduce tool
works. The shrinking process will give up after a maximum number of
steps, which can be configured on the command line.

Tests for the combination of fuzzing and shrinking are included, using
a variety of interestingness functions.
2019-07-07 08:55:30 +01:00
Steven Perron
37e8f79946
Perform merge return with single return in loop. (#2714)
Inlining does not inline functions that have a single return that is in a loop.  This is because the return cannot be replaced by a branch outside of the loop easily.  Merge return knows how to rewrite the function so the return is replaced by a branch.

Fixes #2038.
2019-07-04 14:14:49 -04:00
Steven Perron
fe7cc9c612
Do not inline OpKill Instructions (#2713)
It is illegal to inline an OpKill instruction into a continue construct because the continue header will no longer dominate the backedge.

This commit adds a check for this, and does not inline.

If we still want to be able to inline a function that contains an OpKill, we can add a new pass that will wrap OpKill instructions into its own function with just the single instruction.

I do not believe that this is a common case right now, so I will not do that yet.

Fixes #2433.
2019-07-04 12:08:23 -04:00
Jason Macnak
e6e3e2ccc6 Update type for loaded builtin GlobalInvocationID in pass instrumentation (#2705)
When working on descriptor indexing validation for compute shaders, the
gl_GlobalInvocationID builtin was being loaded as uint which would cause
compute shaders instrumented by the bindless check pass to have:

%83 = OpLoad %uint %gl_GlobalInvocationID
%84 = OpCompositeExtract %uint %83 0
%85 = OpCompositeExtract %uint %83 1
%86 = OpCompositeExtract %uint %83 2

which results in validation failures:

error: line 127: Reached non-composite type while indexes still remain
to be traversed.
%84 = OpCompositeExtract %uint %83 0

for trying to extract a uint from a uint.
2019-06-28 09:46:16 -04:00
Alastair Donaldson
efde682369
Disallow movement of unreachable blocks. (#2700)
Fixes #2695.  Allowing unreachable blocks to be moved can lead to an
unreachable block A getting placed after an unreachable successor B,
which is a problem if B uses ids that A generates.
2019-06-26 15:32:25 +01:00
Alastair Donaldson
dfcb5a1e10
Refactor fuzzer transformations (#2694)
Introduced abstract class for transformations, and refactored all transformations to inherit from this abstract class.
2019-06-25 20:49:46 +01:00
Józef Kucia
888aeef8a9 Fix Component decoration validation for arrays (#2697) 2019-06-25 13:28:16 -04:00
Józef Kucia
7c294608ca Basic validation for Component decorations (#2679)
* Add basic validation for Component decoration
* Add validator tests for Component decoration
2019-06-20 18:16:12 -04:00
alan-baker
2b84d25f10
Fix store to uniform Vulkan check (#2688)
* Wrong operands were used for pointer and array types
* added tests to catch the wierd number corner
2019-06-20 14:22:41 -04:00
Alastair Donaldson
51b0d5ce50
Represent uniform facts via descriptor set and binding. (#2681)
* Represent uniform facts via descriptor set and binding.

Previously uniform facts were expressed with resepect to the id of a
uniform variable.  Describing them with respect to a descriptor set
and binding is more convenient from the point of view of expressing
facts about a shader without requiring analysis of its SPIR-V.

* Fix equality testing for uniform buffer element descriptors.

The equality test now checks that the lengths of the index vectors
match.  Added a test that exposes the previous omission.
2019-06-19 20:45:14 +01:00
Alastair Donaldson
001e823b65
Add fuzzer pass to obfuscate constants. (#2671)
Adds a new transformation that can replace a constant with a uniform known to have the same value, and adds a fuzzer pass that (a) replaces a boolean with a comparison of literals (e.g. replacing "true" with "42 > 24"), and then (b) obfuscates the literals appearing in this comparison by replacing them with identically-valued uniforms, if available.

The fuzzer_replayer test file has also been updated to allow initial facts to be provided, and to do error checking of the status results returned by the fuzzer and replayer components.
2019-06-18 18:41:08 +01:00
alan-baker
2090d7a2d2
Handle volatile memory semantics in upgrade (#2674)
* If an atomic is decorated with volatile add the volatile bit to its
memory semantics
2019-06-17 16:01:37 -04:00
alan-baker
3d5fb7b908
Validate Volatile memory semantics bit (#2672)
* Can only be used with Vulkan memory model
* Can only be used with atomics
* Bit setting must match for compare exchange opcodes
* Updated memory semantics checks to allow constant instructions
generally with CooperativeMatrixNV
2019-06-17 13:35:40 -04:00
alan-baker
400dbde0ba
Disallow stores to UBOs (#2651)
Fixes #2638

* Adds a check that errors out if there is a store to a UBO in the
Vulkan environment
  * tests
* Function to trace pointers
2019-06-17 13:13:07 -04:00
alan-baker
59983a6010 Validate variable initializer type (#2668)
Fixes #249

* The pointed to type of Result Type must match the initializer type
* Had to update some opt tests to be valid
2019-06-15 00:34:18 -04:00
Alastair Donaldson
42830e5a68
Add replayer tool for spirv-fuzz. (#2664)
The replayer takes an existing sequence of transformations and applies
them to a module.  Replaying a sequence of transformations that were
obtained via fuzzing should lead to an identical module to the module
that was fuzzed.  Tests have been added to check for this.
2019-06-13 14:08:33 +01:00
alan-baker
b4bf7bcf0a
Add validation for Subgroup builtins (#2637)
Fixes #2611

* Validates builtins in the Vulkan environment:
  * NumSubgroups
  * SubgroupId
  * SubgroupEqMask
  * SubgroupGeMask
  * SubgroupGtMask
  * SubgroupLeMask
  * SubgroupLtMask
  * SubgroupLocalInvocationId
  * SubgroupSize
2019-06-13 08:47:05 -04:00
Alastair Donaldson
9c0830133b
Add constant == uniform facts. (#2660)
Adds a new (and first) kind of fact to the fact manager, which is that
a specific uniform value is guaranteed to be equal to a specific
constant.  The point of this is that such information (if known to be
true by some external source) can be used by spirv-fuzz to transform
the module in interesting ways that a static compiler cannot reverse
via compile-time analysis.

This change introduces protobuf messages for the fact, and adds
capabilities to the fact manager to store this kind of fact and
provide information about it.
2019-06-11 15:56:08 +01:00
Alastair Donaldson
a8ae579f7a
Add transformation to replace a boolean constant with a numeric comparison (#2659)
The transformation can, for example, replace "true" with "12.0 > 6.0",
if constants for those floating-point values are available.

This introduces a new 'id use descriptor' structure, which provides a
way to describe a particular use of an id, and which will be heavily
used in future transformations.  Describing an id use is trivial if
the use occurs in an instruction that itself generates an id, but is
less straightforward if the id of interest is used by an instruction
such as OpStore that does not have a result id.  The 'id use
descriptor' structure caters for such cases.
2019-06-06 22:22:35 +01:00
Daniel Koch
0755d6ce82 Add builtin validation for SPV_NV_shader_sm_builtins (#2656)
Also add a Builtin test generator variant that takes
capabilities and extensions.

Tests
 - verify that the SMCountNV, SMIDNV, WarpsPerSMNV, and WarpIDNV Builtins are
   accepted as Inputs in Vertex, Fragment, TessControl, TessEval, Geometry,
   and Compute.
 - verify that the SMCountNV, SMIDNV, WarpsPerSMNV, and WarpIDNV Builtins are
   accepted as Inputs in MeshNV and TaskNV shaders.
 - verify that the SMCountNV, SMIDNV, WarpsPerSMNV, and WarpIDNV Builtins are
   accepted as Inputs in the 6 ray tracing stages
 - verify that the SMCountNV, SMIDNV, WarpsPerSMNV, and WarpIDNV Builtins are
   NOT accepted as Outputs.
 - verify that the SMCountNV, SMIDNV, WarpsPerSMNV, and WarpIDNV Builtins are
   NOT accepted as non-scalar integers (f32, uvec3)
 - verify that the SMCountNV, SMIDNV, WarpsPerSMNV, and WarpIDNV Builtins are
   NOT accepted as non-32-bit integers (u64)
2019-06-06 14:53:48 -04:00
greg-lunarg
43fb2403a6 Instrument: Fix code for version 2 output format. (#2655)
Correct record size. Also bring version 2 tests up to version 1
equivalence.
2019-06-06 11:35:34 -04:00
Alastair Donaldson
08cc49ec59
Fix bug in 'split blocks', and add tests for fuzzer. (#2658)
There turned out to be a bug in the 'split blocks' transformation due
to blocks being split while they were being iterated over.  This
change fixes that issue, and adds tests that were able to expose the
issue by running the fuzzer on some example shaders.
2019-06-05 21:54:47 +01:00
David Neto
d01a3c3b4b
Optimizer: Handle array type with OpSpecConstantOp length (#2652)
When it's an OpConstant or OpSpecConstant, then the literal
values are compared.  If the OpSpecConstant also has a SpecId
decoration, then that's also compared.

Otherwise, it's an OpSpecConstantOp and we only compare the
ID of the OpSpecConstantOp instruction itself.

Fixes #2649
2019-06-05 16:35:50 -04:00
Alastair Donaldson
4a00a80c40
Add fuzzer pass to add dead breaks. (#2654)
This pass randomly add breaks to the merge blocks of selection and
loop constructs, such that the breaking edges will not be dynamically
reachable.
2019-06-05 08:02:16 +01:00
Alastair Donaldson
620197bd65
Add fuzzer pass that adds useful constructs to a module (#2647)
This new pass adds some basic ingredients to a module on which future
passes are likely to depend, such as boolean constants and some
specfic integer and floating-point values.  This is not a fuzzer pass
in the true sense in that it does not employ randomization, but it
makes sense to define it as a fuzzer pass since it is the first of a
number of transformations passes that the fuzzer will run on a module.
2019-06-04 14:55:00 +01:00
Jeff Bolz
2c0111e6eb Add validation for SPV_EXT_fragment_shader_interlock (#2650) 2019-06-03 10:55:07 -04:00
Ryan Harrison
699e167d78
Remove asserts from GetUnderlyingType (#2646)
Fixes #2463
2019-05-31 08:57:41 -07:00
Kévin Petit
f99d7ad5c0 Validate OpenCL rules for ImageRead and OpImageSampleExplicitLod (#2643)
Fixes #2594.

Signed-off-by: Kevin Petit <kevin.petit@arm.com>
2019-05-31 10:05:34 -04:00
Alastair Donaldson
209ff0ce90
Add spirv-fuzz pass to permute blocks. (#2642)
The blocks within each function in the module will be permuted in a
randomized manner that respects dominance.
2019-05-31 09:59:06 +01:00
Pierre Moreau
e7866de4b1 Linker: Better type comparison for OpTypeArray and OpTypeForwardPointer (#2580)
* Types: Avoid comparing IDs for in Type::IsSameImpl

When linking, we end up with duplicate types for imported and exported
types, that needs to be removed. The current code would reject valid
import/export pairs of symbols due to IDs mismatch, even if the types or
constants behind those ID were the same.

Enabled remaining type_match_test

Fixes #2442
2019-05-29 16:12:02 -04:00
Ryan Harrison
0125b28ed4
Add compact ids to WebGPU <-> Vulkan transformations (#2639)
Fixes #2634
2019-05-29 12:58:37 -07:00
greg-lunarg
3d62cb8148 Instrument: Add version 2 of record formats (#2630)
New version has additional word in stage-specific section. Also
some changes in content for tesselation and compute shaders. Either
version can be invoked at pass creation. This is done to ease integration
and updating of validation layers. Version 1 is deprecated and eventually
will go away.

Also sneaking in fix to version 1 compute shaders.
2019-05-29 15:08:21 -04:00
Alastair Donaldson
1b71e45338
Add "split block" transformation. (#2633)
With this pass, the fuzzer can split blocks in the input module.  This
is mainly useful in order to give other (future) transformations more
opportunities to apply.
2019-05-29 16:42:46 +01:00
Ryan Harrison
f051812343
Add WebGPU specific fuzzer for validation (#2628)
Fixes #2627
2019-05-28 11:51:52 -07:00
Ryan Harrison
5a06fa4661
Add fuzzer for Vulkan->WebGPU spirv-opt passes (#2626)
Fixes #2622
2019-05-28 10:11:43 -07:00
Ryan Harrison
78b2b18661
Add fuzzer for WebGPU->Vulkan spirv-opt passes (#2625)
Fixes #2623
2019-05-28 07:18:03 -07:00
Steven Perron
6c7db9c630
Handle nested breaks from switches. (#2624)
* Handle nested breaks from switches.

There was a recent decision made to allow branches to the merge node of
a switch even if the switch is not the first enclosing construct.  They
can be generated by glslang from break statements in switches.

Dead branch elimination seems to be the only optimization that will
break because of this change, so I will update that optimizations.

The change made are:

- Track switches in structured cfg analysis.
- In Dead branch elimination:
  - Look for nested breaks that will require a switch instruction.
  - Rewrite, but don't delete, switchs that are required even if it
    could be replaced by an unconditional branch.
  - When looking for the first break, consider the merge of a switch
    as well.

See #2612.

* Fix variable names and comments.

* Add tests for the struct cfg analysis and switches.

* Fix typos in comments.
2019-05-27 16:28:14 -04:00
dan sinclair
42abaa099a Remove MarkV and Stats code. (#2576)
* Remove MarkV and Stats code.

This Cl removes the MarkV and Stats code from SPIRV-Tools. This code was
unused and currently un-maintained.
2019-05-24 15:43:59 -04:00
Jonathon Anderson
3b5ab540ca linker: Add tests for various type comparisons (#2613)
This adds a number of tests that check that all types will match to
identically written clones during linking, including nearly every Type
and some combinations (e.g. Functions of Arrays of Floats). Intent is
for use with https://github.com/KhronosGroup/SPIRV-Tools/pull/2580,
however that PR focuses on issues with TypeArray whereas these tests are
(more) comprehensive and test more subtle (and possibly incorrect)
cases.

A number of these tests fail, many are fixed by the aforementioned PR.
Some additional tests involving TypeForwardPointer are currently
disabled as they cause assertion failures.
2019-05-24 15:40:28 -04:00
Sahil Parmar
b8fe7211c4 Allow arrays of out per-primitive builtins for mesh shaders (#2617)
- PrimitiveID, Layer, ViewportIndex
* Add validation tests for mesh builtins
2019-05-23 15:08:59 -04:00