[amdgpu][nfc] Separate processUsedLDS into independent pieces, rename it

[amdgpu][nfc] Extract kernel annotation from processUsedLDS

[amdgpu][nfc] Separate LDS struct creation from RAUW

[amdgpu][nfc] Skip operations on padding fields in LDS struct

[nfc][amdgpu] LDS. Move selection logic up the stack.

[nfc][amdgpu] Remove dead variable and function

[amdgpu] Elide module lds allocation in kernels with no callees

Introduces a string attribute, amdgpu-requires-module-lds, to allow
eliding the module.lds block from kernels. Will allocate the block

[amdgpu] Elide module lds allocation in kernels with no callees

Introduces a string attribute, amdgpu-requires-module-lds, to allow
eliding the module.lds block from kernels. Will allocate the block as before
if the attribute is missing or has its default value of true.

Patch uses the new attribute to detect the simplest possible instance of this,
where a kernel makes no calls and thus cannot call any functions that use LDS.

Tests updated to match, coverage was already good. Interesting cases is in
lower-module-lds-offsets where annotating the kernel allows the backend to pick
a different (in this case better) variable ordering than previously. A later
patch will avoid moving kernel variables into module.lds when the kernel can
have this attribute, allowing optimal ordering and locally unused variable
elimination.

Reviewed By: arsenm

Differential Revision: https://reviews.llvm.org/D122091

show more ...

[AMDGPU][NFC] Fix typos

Fix some typos in the amdgpu backend.

Differential Revision: https://reviews.llvm.org/D119235

[AMDGPU] Merge AMDGPULDSUtils into AMDGPUMemoryUtils

Differential Revision: https://reviews.llvm.org/D119502

[amdgpu] Increase alignment of all LDS variables

Currently the superalign option only increases the alignment of
variables that are moved into the module.lds block. Change that to all LDS
variables.

[amdgpu] Increase alignment of all LDS variables

Currently the superalign option only increases the alignment of
variables that are moved into the module.lds block. Change that to all LDS
variables. Also only increase the alignment once, instead of once per function.

Reviewed By: rampitec

Differential Revision: https://reviews.llvm.org/D115488

show more ...

[llvm] Use range-based for loops (NFC)

Revert "[amdgpu][nfc] Delete dead code in LowerModuleLDS"

This reverts commit 7b9ab06d10a6a989f76e6c5ecf89d906f838fe7d.
Said code is better removed as part of a larger change.

[amdgpu][nfc] Delete dead code in LowerModuleLDS

[amdgpu][nfc] Drop dead PtrSet, fix a comment

[amdgpu][nfc] Move non-shared code out of LDSUtils

[AMDGPU] Correctly merge alias.scope and noalias metadata for memops

When adding alias.scope and noalias metadata to a memcpy function,
the alias.scope and noalias metadata from the operands are mer

[AMDGPU] Correctly merge alias.scope and noalias metadata for memops

When adding alias.scope and noalias metadata to a memcpy function,
the alias.scope and noalias metadata from the operands are merged.
The rule for merging alias.scope is to take the intersection of
the domains and the union of the scopes within those domains.
The rule for merging noalias is to take the intersection.

The bug is that AMDGPULowerModuleLDS was using concatenation for
both alias.scope and noalias. For example, when f1 and f2 are added
to the LDS structure and there is a memcpy(f2, f1, sizeof(f1)).
Then, concatenation creates noalias metadata for the memcpy that
includes both {f1, f2}. That means that the memcpy is assumed
not to alias a prior load of f2, which enables the optimizer to
remove a load of f2 that occurs after mempcy.

The function MDNode::getmostGenericAliasScope defines the semantics
for alias.scope. There is a function, combineMetadata in Local.cpp,
that uses intersect for noalias.

Differential Revision: https://reviews.llvm.org/D110049

show more ...

[AMDGPU][NFC] Correct typos in lib/Target/AMDGPU/AMDGPU*.cpp files. Test commit for new contributor.

AMDGPU: Fix crashing on kernel declarations when lowering LDS

This was trying to insert the used marker into a declaration.

[AMDGPU] Add alias.scope metadata to lowered LDS struct

Alias analysis is unable to disambiguate accesses to the structure
fields without it unlike distinct variables. As a result we cannot
combine

[AMDGPU] Add alias.scope metadata to lowered LDS struct

Alias analysis is unable to disambiguate accesses to the structure
fields without it unlike distinct variables. As a result we cannot
combine ds_read and ds_write operations in a case of any store in
between which always considered clobbering.

Differential Revision: https://reviews.llvm.org/D108315

show more ...

[AMDGPU] Disable LDS lowering for GFX shaders

Apparently these need external LDS symbols to remain.

Fixes: SC1-3279

Differential Revision: https://reviews.llvm.org/D106288

[AMDGPU] Check for pointer operand while refining LDS align

Also skips the propagation if alignment is 1.

Differential Revision: https://reviews.llvm.org/D104796

[AMDGPU] Propagate LDS align into to instructions

Differential Revision: https://reviews.llvm.org/D104316

[AMDGPU] Use performOptimizedStructLayout for LDS sort

This gives better packing.

Differential Revision: https://reviews.llvm.org/D104331

[AMDGPU] Replace non-kernel function uses of LDS globals by pointers.

The main motivation behind pointer replacement of LDS use within non-kernel
functions is - to *avoid* subsequent LDS lowering pa

[AMDGPU] Replace non-kernel function uses of LDS globals by pointers.

The main motivation behind pointer replacement of LDS use within non-kernel
functions is - to *avoid* subsequent LDS lowering pass from directly packing
LDS (assume large LDS) into a struct type which would otherwise cause allocating
huge memory for struct instance within every kernel.

Reviewed By: rampitec

Differential Revision: https://reviews.llvm.org/D103225

show more ...

[AMDGPU] Fix missing lowering of LDS used in global scope.

Reviewed By: rampitec

Differential Revision: https://reviews.llvm.org/D103431