[CIR][Lowering] Add the concept of simple lowering and use it to implement FP intrincis #434

philnik777 · 2024-01-31T13:12:42Z

No description provided.

Lowers the particular case where an array of structs is constant-initialized. ghstack-source-id: f8389d8 Pull Request resolved: llvm#202

Updates the lowering of cir.const to support #cir.const_struct as a initializer, allowing the initialization of local structs with said attribute. ghstack-source-id: cc6c137 Pull Request resolved: llvm#203

Global addresses are constant, so we can initialize them at compile time using CIR's global_view attribute. This patch adds codegen support for the initialization of variables with constant global addresses. Since a builder method was added for global_view, the patch also updates the codegen of global variables to use it wherever possible. ghstack-source-id: 513365c Pull Request resolved: llvm#204

Adds lowering logic for CIR's global_view attributes with no indexes. This is done by converting the global to a region-initialized LLVM global operation, where the region returns the address of the global used in the gloval_view initializer attribute. ghstack-source-id: a9452ed Pull Request resolved: llvm#205

Constant initialization of static local arrays would fail due to a mismatch between the variable and the initializer type size. This patch fixes the data layout interface implementation for the cir.array type. A complete array in C/C++ should have its type size in bits equal to the size of the array times the size of the element type. ghstack-source-id: 56f3f29 Pull Request resolved: llvm#206

ghstack-source-id: 1f793b2 Pull Request resolved: llvm#207

…es instead of region initializers.

PR llvm#200 broke the vtableAttr usage, as it was not properly refactored to use ArrayAttr instead of ConstStructAttr to store its members.

This change does the CIR generation for globals initialized by a constructor call. It currently only covers C++ to CIR generation. The corresponding LLVM lowering will be in a follow-up commit. A motivating example is ``` class Init { friend class ios_base; public: Init(bool); ~Init(); private: static bool _S_synced_with_stdio; }; static Init ioinit(true); ``` Unlike what the default Clang codegen generates LLVM that detaches the initialization code from the global var definition (like below), we are taking a different approach that keeps them together, which we think will make the later dataflow analysis/transform easier. ``` @_ZL8ioinit = internal global %class.Init zeroinitializer, align 1, !dbg !0 define internal void @cxx_global_var_init() #0 section ".text.startup" !dbg !23 { entry: call void @_ZN4InitC2Ev(ptr noundef nonnull align 1 dereferenceable(1) @_ZL8ioinit), !dbg !27 %0 = call i32 @cxa_atexit(ptr @_ZN4InitD1Ev, ptr @_ZL8ioinit, ptr @dso_handle) llvm#3, !dbg !29 ret void, !dbg !27 } ``` So on CIR, we have something like: ``` cir.global "private" internal @_ZL8__ioinit = ctor : !ty_22class2EInit22 { %0 = cir.get_global @_ZL8__ioinit : cir.ptr <!ty_22class2EInit22> loc(#loc8) %1 = cir.const(#true) : !cir.bool loc(#loc5) cir.call @_ZN4InitC1Eb(%0, %1) : (!cir.ptr<!ty_22class2EInit22>, !cir.bool) -> () loc(#loc6) } ``` The destructor support will also be in a separate change.

…eLocation

This PR fixes lowering for `break/continue` in loops. The idea is to replace `cir.yield break` and `cir.yield continue` with the branch operations to the corresponding blocks. Note, that we need to ignore nesting loops and don't touch `break` in switch operations. Also, `yield` from `if` need to be considered only when it's not the loop `yield` and `continue` in switch is ignored since it's processed in the loops lowering. Fixes llvm#160

…cations are the same

Essentially emits an LValue for the struct and then passes it as a call argument.

Enabling IR printing with --mlir-print-ir-after=passName1, passName2. This requires all CIR passes to be registered at startup time.

Summary: Setting the Optnone attribute for CIR functions and progating it all the way down to LLVM IR for those not supposed to be optimized.

This change is a prerequisite of llvm#235

…zers (llvm#235) As a follow up to llvm#197, this change pre-lowers high-level CIR for global initializers. High-level CIR: ``` cir.global "private" internal @_ZL8__ioinit = ctor : !ty_22class2EInit22 { %0 = cir.get_global @_ZL8__ioinit : cir.ptr <!ty_22class2EInit22> loc(#loc8) %1 = cir.const(#true) : !cir.bool loc(#loc5) cir.call @_ZN4InitC1Eb(%0, %1) : (!cir.ptr<!ty_22class2EInit22>, !cir.bool) -> () loc(#loc6) } ``` After pre-lowering: ``` cir.global "private" internal @_ZL8__ioinit = #cir.zero : !ty_22class2EInit22 {ast = #cir.vardecl.ast} cir.func internal private @__cxx_global_var_init() { %0 = cir.get_global @_ZL8__ioinit : cir.ptr <!ty_22class2EInit22> %1 = cir.const(#true) : !cir.bool cir.call @_ZN4InitC1Eb(%0, %1) : (!cir.ptr<!ty_22class2EInit22>, !cir.bool) -> () cir.return } cir.func private @_GLOBAL__sub_I_static.cpp() { cir.call @__cxx_global_var_init() : () -> () cir.return } ``` There is still work to be done to fully lower to LLVM. E.g, add `llvm.global_ctors` global to list all module initializers like `_GLOBAL__sub_I_static`. This will be handled in a separate change.

cir-tidy has been our custom hack for using clang-tidy. There's no reason it needs to be that way - add a CIRModule to clang-tidy and make the lifetime checker its first check. Some of the way this was built was inspired in CLANG_TIDY_ENABLE_STATIC_ANALYZER Once we can fully migrate deps from cir-tidy, we should remove it in favor of this.

…s::options This unbreaks Linux builds, which points to cycles while using utils::options from ClangTidy.cpp. We need to find a better way to convey this.

This PR fixes CIR generation for the `switch-case` cases like the following: ``` case 'a': default: ... ``` or ``` default: case 'a': ... ``` i.e. when the `default` clause is sub-statement of the `case` one and vice versa.

Whenever a global array is declared and initialized with fewer elements than its size, the remaining elements are implicitly initialized with zero. For aggregates types, such as structs, the initialization is done through the #cir.zero attribute. ghstack-source-id: b3d172c Pull Request resolved: llvm#216

This small improvement allows for a more compact representation of structs with zero-initialized fields in the C++ codegen. ghstack-source-id: a3fe0dc Pull Request resolved: llvm#217

Due to the lack of zeroinitializer support in LLVM, some cases are tricky to lower #cir.zero. An example is when an array is only partially initialize with #cir.zero attributes. Since we can't just zeroinitialize the whole array, the current #cir.zero attribute amend does not suffice. To simplify the lowering, this patch introduces a new operation that is solely used to generate zeroinitialize LLVM IR constants. ghstack-source-id: a3fd40e Pull Request resolved: llvm#218

Supports lowering for #cir.zero attributes when it appears in aggregate attributes such as #const.array and #cir.struct. ghstack-source-id: d26a506 Pull Request resolved: llvm#219

Lancern · 2024-03-06T13:47:19Z

This is an awesome feature! This is not a blocker for the PR, just some ideas for further development. This feature may be generalized in the future to declare lowering rules more flexibly. Consider the cir.bit family of operations, which has an LLVMIR lowering scheme like the following:

%1 = cir.bit.clz(%0 : !u32i) : !s32i
// %1 = llvm.call_intrinsic "llvm.ctlz.i32"(%0 : i32) : i32 -> i32

%2 = cir.bit.ctz(%0 : !u32i) : !s32i
// %2 = llvm.call_intrinsic "llvm.cttz.i32"(%0 : i32) : i32 -> i32

%3 = cir.bit.popcount(%0 : !u32i) : !s32i
// %3 = llvm.call_intrinsic "llvm.ctpop.i32"(%0 : i32) : i32 -> i32

Basically all cir.bit.* operation will be lowered to an llvm.call_intrinsic operation whose intrin parameter is determined by the CIR operation type.

Stuff in this PR cannot be used to achieve the lowering scheme above because we cannot specify the parameters to the LLVMIR operations. It would be nice if this is added in the future; it would be very useful for lowering compiler intrinsics as shown above.

bcardosolopes · 2024-03-06T20:05:08Z

Good point @Lancern, we can definitely extend the mechanism later to also support things that get lowered to llvm.call_intrinsic.

Lancern · 2024-08-23T04:11:18Z

Any progress on this PR?

bcardosolopes · 2024-08-23T17:05:19Z

@philnik777 is probably busy with libc++ work.

@Lancern if you are interested in making progress here, my suggestion:

Cherry-pick his work locally (so as to preserve original authorship) and fix some of the conflicts.
Add your changes on top + previous reviewer feedback.
Take over this PR (assuming that's possible)

I'll then land without squashing so that we can preserve your work and his (assuming we get a final total of 2 commits, his + yours).

philnik777 · 2024-08-24T09:57:35Z

Sorry for dropping this. I completely forgot that I had a patch open here. @Lancern if you want to, feel free to take the patch and just add a Co-authored-by: Nikolas Klauser <nikolasklauser@berlin.de> in the commit message.

Lancern · 2024-08-24T12:39:15Z

Let me work on this PR.

bcardosolopes · 2024-08-26T17:28:21Z

@philnik777 np, thanks for the clangir contribs so far!

bcardosolopes · 2024-08-26T21:14:08Z

This is now tracked in #806, closing

@philnik777

…ry fp2fp operations (#806) This PR is the continuation and refinement of PR #434 which is originally authored by @philnik777 . Does not update it in-place since I don't have commit access to Nikolas' repo. This PR basically just rebases #434 onto the latest `main`. I also updated some naming used in the original PR to keep naming styles consistent. Co-authored-by: Nikolas Klauser <nikolasklauser@berlin.de>

@philnik777

…ry fp2fp operations (llvm#806) This PR is the continuation and refinement of PR llvm#434 which is originally authored by @philnik777 . Does not update it in-place since I don't have commit access to Nikolas' repo. This PR basically just rebases llvm#434 onto the latest `main`. I also updated some naming used in the original PR to keep naming styles consistent. Co-authored-by: Nikolas Klauser <nikolasklauser@berlin.de>

@philnik777

…ry fp2fp operations (llvm#806) This PR is the continuation and refinement of PR llvm#434 which is originally authored by @philnik777 . Does not update it in-place since I don't have commit access to Nikolas' repo. This PR basically just rebases llvm#434 onto the latest `main`. I also updated some naming used in the original PR to keep naming styles consistent. Co-authored-by: Nikolas Klauser <nikolasklauser@berlin.de>

@philnik777

…ry fp2fp operations (llvm#806) This PR is the continuation and refinement of PR llvm#434 which is originally authored by @philnik777 . Does not update it in-place since I don't have commit access to Nikolas' repo. This PR basically just rebases llvm#434 onto the latest `main`. I also updated some naming used in the original PR to keep naming styles consistent. Co-authored-by: Nikolas Klauser <nikolasklauser@berlin.de>

@philnik777

…ry fp2fp operations (llvm#806) This PR is the continuation and refinement of PR llvm#434 which is originally authored by @philnik777 . Does not update it in-place since I don't have commit access to Nikolas' repo. This PR basically just rebases llvm#434 onto the latest `main`. I also updated some naming used in the original PR to keep naming styles consistent. Co-authored-by: Nikolas Klauser <nikolasklauser@berlin.de>

@philnik777

…ry fp2fp operations (#806) This PR is the continuation and refinement of PR #434 which is originally authored by @philnik777 . Does not update it in-place since I don't have commit access to Nikolas' repo. This PR basically just rebases #434 onto the latest `main`. I also updated some naming used in the original PR to keep naming styles consistent. Co-authored-by: Nikolas Klauser <nikolasklauser@berlin.de>

@philnik777

…ry fp2fp operations (#806) This PR is the continuation and refinement of PR #434 which is originally authored by @philnik777 . Does not update it in-place since I don't have commit access to Nikolas' repo. This PR basically just rebases #434 onto the latest `main`. I also updated some naming used in the original PR to keep naming styles consistent. Co-authored-by: Nikolas Klauser <nikolasklauser@berlin.de>

@philnik777

…ry fp2fp operations (llvm#806) This PR is the continuation and refinement of PR llvm#434 which is originally authored by @philnik777 . Does not update it in-place since I don't have commit access to Nikolas' repo. This PR basically just rebases llvm#434 onto the latest `main`. I also updated some naming used in the original PR to keep naming styles consistent. Co-authored-by: Nikolas Klauser <nikolasklauser@berlin.de>

sitio-couto and others added 30 commits January 28, 2024 20:44

[CIR][Lowering] Lower constant arrays of structs

b139b4d

Lowers the particular case where an array of structs is constant-initialized. ghstack-source-id: f8389d8 Pull Request resolved: llvm#202

[CIR][Lowering] Partially lower local #cir.const_struct initializers

4786f2b

Updates the lowering of cir.const to support #cir.const_struct as a initializer, allowing the initialization of local structs with said attribute. ghstack-source-id: cc6c137 Pull Request resolved: llvm#203

[CIR][CIRTidy] Add config options for codegen manipulation

ebcfb07

[CIR][CIRGen] Use #cir.zero on zero-initialized global arrays

e9b8cfe

ghstack-source-id: 1f793b2 Pull Request resolved: llvm#207

[CIR][Lowering] Lower global multidimensional array initializers.

894334a

[CIR][Lowering] Lower nested arrays of simple types to dense attribut…

20f4e86

…es instead of region initializers.

[CIR][Bugfix] Fix vtableAttr const struct usage

230a6b8

PR llvm#200 broke the vtableAttr usage, as it was not properly refactored to use ArrayAttr instead of ConstStructAttr to store its members.

[CIR][CIRTidy] Handle mlir::FusedLoc when translating to clang::Sourc…

8f2070e

…eLocation

[CIR][CodeGen] Fix generation of 'if statement' when begin and end lo…

7cd4eb5

…cations are the same

[CIR][CIRGen] Fix heap-use-after-free in recordDeclTypes, found by ASAN

376a938

[CIR][CIRGen] Support for struct call arguments

e91e0f9

Essentially emits an LValue for the struct and then passes it as a call argument.

[CIR] Enable per-pass IR printing (llvm#234)

d929c69

Enabling IR printing with --mlir-print-ir-after=passName1, passName2. This requires all CIR passes to be registered at startup time.

[ClR] Set optnone attribute for functions. (llvm#115)

5c960ff

Summary: Setting the Optnone attribute for CIR functions and progating it all the way down to LLVM IR for those not supposed to be optimized.

[CIR][Lowering] Add an empty LoweringPrepare pass (llvm#236)

5baedbc

This change is a prerequisite of llvm#235

[CIR][CIRTidy] Break CIRASTConsumer dep from clang::tidy::utils

a92849b

[CIR][ClangTidy] Move dependence up and remove CIRASTConsumer files

54f525c

[CIR][ClangTidy] Copy parseStringList function from clang::tidy::util…

14f9a0d

…s::options This unbreaks Linux builds, which points to cycles while using utils::options from ClangTidy.cpp. We need to find a better way to convey this.

[CIR][ClangTidy] Fix cmake remaining issues

296ad65

[CIR][CIRGen] Update C++ codegen to use #cir.zero on structs

929969d

This small improvement allows for a more compact representation of structs with zero-initialized fields in the C++ codegen. ghstack-source-id: a3fe0dc Pull Request resolved: llvm#217

[CIR][Lowering] Lower #cir.zero nested in initializer attributes

5d5d2c3

Supports lowering for #cir.zero attributes when it appears in aggregate attributes such as #const.array and #cir.struct. ghstack-source-id: d26a506 Pull Request resolved: llvm#219

philnik777 changed the title ~~[CIR][Lowering] Add the concept of simple lowering and useit to implement FP intrincis~~ [CIR][Lowering] Add the concept of simple lowering and use it to implement FP intrincis Feb 20, 2024

bcardosolopes mentioned this pull request Feb 21, 2024

[CIR][CIRGen] Add support for builtin bit operations #474

Merged

lanza force-pushed the main branch 2 times, most recently from ed3955a to 8b7417c Compare March 23, 2024 05:07

lanza force-pushed the main branch from 8b7417c to 330c983 Compare March 30, 2024 04:29

lanza force-pushed the main branch from 93715bd to a8b0ecd Compare April 17, 2024 05:15

lanza force-pushed the main branch 2 times, most recently from c4db6d0 to e197d4e Compare April 29, 2024 20:11

This was referenced May 8, 2024

[CIR][Lowering] Add MLIR lowering support for CIR sin operations #586

Merged

[CIR][Lowering] Add MLIR lowering support for CIR math operations #592

Merged

lanza force-pushed the main branch from 8829868 to c290c04 Compare June 21, 2024 19:50

lanza self-requested a review as a code owner June 21, 2024 19:50

Lancern mentioned this pull request Aug 24, 2024

[CIR][Lowering] Add the concept of simple lowering and use it for unary fp2fp operations #806

Merged

bcardosolopes closed this Aug 26, 2024

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

[CIR][Lowering] Add the concept of simple lowering and use it to implement FP intrincis #434

[CIR][Lowering] Add the concept of simple lowering and use it to implement FP intrincis #434

Uh oh!

philnik777 commented Jan 31, 2024

Uh oh!

Lancern commented Mar 6, 2024 •

edited

Loading

Uh oh!

bcardosolopes commented Mar 6, 2024

Uh oh!

Lancern commented Aug 23, 2024

Uh oh!

bcardosolopes commented Aug 23, 2024 •

edited

Loading

Uh oh!

philnik777 commented Aug 24, 2024

Uh oh!

Lancern commented Aug 24, 2024

Uh oh!

bcardosolopes commented Aug 26, 2024

Uh oh!

bcardosolopes commented Aug 26, 2024

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

14 participants

[CIR][Lowering] Add the concept of simple lowering and use it to implement FP intrincis #434

[CIR][Lowering] Add the concept of simple lowering and use it to implement FP intrincis #434

Uh oh!

Conversation

philnik777 commented Jan 31, 2024

Uh oh!

Lancern commented Mar 6, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

bcardosolopes commented Mar 6, 2024

Uh oh!

Lancern commented Aug 23, 2024

Uh oh!

bcardosolopes commented Aug 23, 2024 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

philnik777 commented Aug 24, 2024

Uh oh!

Lancern commented Aug 24, 2024

Uh oh!

bcardosolopes commented Aug 26, 2024

Uh oh!

bcardosolopes commented Aug 26, 2024

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

14 participants

Lancern commented Mar 6, 2024 •

edited

Loading

bcardosolopes commented Aug 23, 2024 •

edited

Loading