TDefaultIoResolverBase::reserveSlot and getFreeSlot now have a size
parameter to reserve a range of bindings. This is used by
TDefaultIoResolver::resolveBinding to reserve a continuous range when
the type is an array and the target API is GL.
When assigning a location to an interface whose stage automatically
converts the interfaces to an array, it now strips off the outermost
array from the type before calculating how many locations it consumes.
When assigning uniform locations it now takes into account the number
of locations occupied by the type. For uniforms, all types except
arrays and structs take up one location. For arrays the base location
count is multiplied by the array dimensions and for structs it is the
sum of the locations of each member.
This factored computeTypeLocationSize() out of needing the TIntermediate contents,
and uses it to show how to know how many locations an object needs.
However, it still does not do cross stage, or mixed location/no-location
analysis.
This PR adds the ability to provide per-descriptor-set IO mapping shift
values. If a particular binding does not land into a per-set value,
then it falls back to the prior behavior (global shifts per resource class).
Because there were already 6 copies of many different methods and internal
variables and functions, and this PR would have added 6 more, a new API is
introduced to cut down on replication and present a cleaner interface.
For the global (non-set-specific) API, the old entry points still exist
for backward compatibility, but are phrased internally in terms of the
following.
// Resource type for IO resolver
enum TResourceType {
EResSampler,
EResTexture,
EResImage,
EResUbo,
EResSsbo,
EResUav,
EResCount
};
Methods on TShader:
void setShiftBinding(TResourceType res, unsigned int base);
void setShiftBindingForSet(TResourceType res, unsigned int set, unsigned int base);
The first method replaces the 6 prior entry points of various spellings, which
exist now in depreciated form. The second provides per-resource-set functionality.
Both accept an enum from the list above.
From the command line, the existing options can accept either a single shift value as
before, or a series of 1 or more [set offset] pairs. Both can be provided, as in:
... --stb 20 --stb 2 25 3 30 ...
which will use the offset 20 for anything except descriptor set 2 (which uses 25) and
3 (which uses 30).
Add support for Subpass Input proposal of issue #1069.
Subpass input types are given as:
layout(input_attachment_index = 1) SubpassInput<float4> subpass_f;
layout(input_attachment_index = 2) SubpassInput<int4> subpass_i;
layout(input_attachment_index = 3) SubpassInput<uint4> subpass_u;
layout(input_attachment_index = 1) SubpassInputMS<float4> subpass_ms_f;
layout(input_attachment_index = 2) SubpassInputMS<int4> subpass_ms_i;
layout(input_attachment_index = 3) SubpassInputMS<uint4> subpass_ms_u;
The input attachment may also be specified using attribute syntax:
[[vk::input_attachment_index(7)]] SubpassInput subpass_2;
The template type may be a shorter-than-vec4 vector, but currently user
structs are not supported. (An unimplemented error will be issued).
The load operations are methods on objects of the above type:
float4 result = subpass_f.SubpassLoad();
int4 result = subpass_i.SubpassLoad();
uint4 result = subpass_u.SubpassLoad();
float4 result = subpass_ms_f.SubpassLoad(samp);
int4 result = subpass_ms_i.SubpassLoad(samp);
uint4 result = subpass_ms_u.SubpassLoad(samp);
Additionally, the AST printer could not print EOpSubpass* nodes. Now it can.
Fixes#1069
--resource-set-binding has a mode which allows per-register assignments of
bindings and descriptor sets on the command line, and another accepting a
single descriptor set value to assign to all variables.
The former worked, but the latter would crash when assigning the values.
This fixes it, and makes the former case a bit more robust against premature
termination of the pre-register values, which must come in (regname,set,binding)
triples.
This also allows the form "--resource-set-binding stage setnum", which was
mentioned in the usage message, but did not parse.
The operation of the per-register form of this option is unchanged.
Also, provides an option to auto-assign locations.
Existing tests use this option, to avoid the error message,
however, it is not fully implemented yet.
Adds a notification phase to the io remapper.
The idea behind this is to give the user a
chance to group uniforms and/or in/out variables
for a better pipeline layout sharing for vulkan.
Change-Id: I7492421085a4156ed3534f01d906ab390d73a623
glslang/MachineIndependent/iomapper.cpp:207:9: error: field 'resolver' will be initialized after field 'stage' [-Werror,-Wreorder]
: resolver(r)
^
glslang/MachineIndependent/iomapper.cpp:263:9: error: field 'resolver' will be initialized after field 'stage' [-Werror,-Wreorder]
: resolver(r)
^
hlsl/hlslParseHelper.cpp:70:5: error: field 'gsStreamOutput' will be initialized after field 'inputPatch' [-Werror,-Wreorder]
gsStreamOutput(nullptr),
^
Adds --hlsl-iomap option to perform IO mapping in HLSL register space.
--shift-cbuffer-binding is now a synonym for --shift-ubo-binding.
The idea way to do this seems to be passing in a dedicated IO resolver, but
that would require more intrusive restructuring, so maybe best for its
own PR.
The TDefaultHlslIoResolver class and the former TDefaultIoResolver class
share quite a bit of mechanism in a common base class.
TODO: tbuffers are landing in the wrong register class, which needs some
investigation. They're either wrong upstream, or the detection in the
resolver is wrong.
New command line option --shift-ssbo-binding mirrors --shift-ubo-binding, etc.
New reflection query getLocalSize(int dim) queries local size, e.g, CS threads.
This PR adds:
1. The "u" register class for RW* objects.
2. --shift-image-bindings (== --sib), analogous to --shift-texture-bindings etc.
3. Case insensitive reg classes.
4. Tests for above.
- add optional callback to handle mapping of uniform variables in linking phase
- if no resolver is provided, it uses the internal default resolver with all shifts and auto bind settings
Change-Id: Icfe38a9eabe8bfc8f8bb6d8150c06f7ed38bb762
Previously, the binding auto-mapping facility was free to use any unused
binding. This change makes auto-bindings use the same offset value as
explicit bindings.
Fix for two defects as follows:
- The IO mapping traverser was not setting inVisit, and would skip some AST nodes.
Depending on the order of nodes, this could have prevented the binding from
showing up in the generated SPIR-V.
- If a uniform array was flattened, each of the flattened scalars from the array
is still a (now-scalar) uniform. It was being converted to a temporary.
This PR adds the ability to offset sampler, texture, and UBO bindings
from provided base bindings, and to auto-number bindings that are not
provided with explicit register numbers. The mechanism works as
follows:
- Offsets may be given on the command line for all stages, or
individually for one or more single stages, in which case the
offset will be auto-selected according to the stage being
compiled. There is also an API to set them. The new command line
options are --shift-sampler-binding, --shift-texture-binding, and
--shift-UBO-binding.
- Uniforms which are not given explicit bindings in the source code
are auto-numbered if and only if they are in live code as
determined by the algorithm used to build the reflection
database, and the --auto-map-bindings option is given. This auto-numbering
avoids using any binding slots which were explicitly provided in
the code, whether or not that explicit use was live. E.g, "uniform
Texture1D foo : register(t3);" with --shift-texture-binding 10 will
reserve binding 13, whether or not foo is used in live code.
- Shorter synonyms for the command line options are available. See
the --help output.
The testing infrastructure is slightly extended to allow use of the
binding offset API, and two new tests spv.register.(no)autoassign.frag are
added for comparing the resulting SPIR-V.
