maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
llvm-project/llvm/test/Transforms/FunctionAttrs/convergent.ll

111 lines
3.0 KiB
LLVM
Raw Normal View History

[PM][FunctionAttrs] add NoUnwind attribute inference to PostOrderFunctionAttrs pass Summary: This was motivated by absence of PrunEH functionality in new PM. It was decided that a proper way to do PruneEH is to add NoUnwind inference into PostOrderFunctionAttrs and then perform normal SimplifyCFG on top. This change generalizes attribute handling implemented for (a removal of) Convergent attribute, by introducing a generic builder-like class AttributeInferer It registers all the attribute inference requests, storing per-attribute predicates into a vector, and then goes through an SCC Node, scanning all the instructions for not breaking attribute assumptions. The main idea is that as soon all the instructions from all the functions of SCC Node conform to attribute assumptions then we are free to infer the attribute as set for all the functions of SCC Node. It handles two distinct cases of attributes: - those that might break due to derefinement of the function code for these attributes we are allowed to apply inference only if all the functions are "exact definitions". Example - NoUnwind. - those that do not care about derefinement for these attributes we are allowed to apply inference as soon as we see any function definition. Example - removal of Convergent attribute. Also in this commit: * Converted all the FunctionAttrs tests to use FileCheck and added new-PM invocations to them * FunctionAttrs/convergent.ll test demonstrates a difference in behavior between new and old PM implementations. Marked with FIXME. * PruneEH tests were converted to new-PM as well, using function-attrs+simplify-cfg combo as intended * some of "other" tests were updated since function-attrs now infers 'nounwind' even for old PM pipeline * -disable-nounwind-inference hidden option added as a possible workaround for a supposedly rare case when nounwind being inferred by default presents a problem Reviewers: chandlerc, jlebar Reviewed By: jlebar Subscribers: eraman, llvm-commits Differential Revision: https://reviews.llvm.org/D44415 llvm-svn: 328377
2018-03-24 05:46:16 +08:00
; FIXME: convert CHECK-INDIRECT into CHECK (and remove -check-prefixes) as soon
; FIXME: as new-pass-manager's handling of indirect_non_convergent_call is fixed
;
; RUN: opt -functionattrs -S < %s | FileCheck %s --check-prefixes=CHECK,CHECK-INDIRECT
; RUN: opt -passes=function-attrs -S < %s | FileCheck %s
Add convergent-removing bits to FunctionAttrs pass. Summary: Remove the convergent attribute on any functions which provably do not contain or invoke any convergent functions. After this change, we'll be able to modify clang to conservatively add 'convergent' to all functions when compiling CUDA. Reviewers: jingyue, joker.eph Subscribers: llvm-commits, tra, jhen, hfinkel, resistor, chandlerc, arsenm Differential Revision: http://reviews.llvm.org/D17013 llvm-svn: 260319
2016-02-10 07:03:22 +08:00
; CHECK: Function Attrs
; CHECK-NOT: convergent
; CHECK-NEXT: define i32 @nonleaf()
define i32 @nonleaf() convergent {
%a = call i32 @leaf()
ret i32 %a
}
; CHECK: Function Attrs
; CHECK-NOT: convergent
; CHECK-NEXT: define i32 @leaf()
define i32 @leaf() convergent {
ret i32 0
}
; CHECK: Function Attrs
; CHECK-SAME: convergent
; CHECK-NEXT: declare i32 @k()
declare i32 @k() convergent
; CHECK: Function Attrs
; CHECK-SAME: convergent
; CHECK-NEXT: define i32 @extern()
define i32 @extern() convergent {
[attrs] Handle convergent CallSites. Summary: Previously we had a notion of convergent functions but not of convergent calls. This is insufficient to correctly analyze calls where the target is unknown, e.g. indirect calls. Now a call is convergent if it targets a known-convergent function, or if it's explicitly marked as convergent. As usual, we can remove convergent where we can prove that no convergent operations are performed in the call. Originally landed as r261544, then reverted in r261544 for (incidental) build breakage. Re-landed here with no changes. Reviewers: chandlerc, jingyue Subscribers: llvm-commits, tra, jhen, hfinkel Differential Revision: http://reviews.llvm.org/D17739 llvm-svn: 263481
2016-03-15 04:18:54 +08:00
%a = call i32 @k() convergent
Add convergent-removing bits to FunctionAttrs pass. Summary: Remove the convergent attribute on any functions which provably do not contain or invoke any convergent functions. After this change, we'll be able to modify clang to conservatively add 'convergent' to all functions when compiling CUDA. Reviewers: jingyue, joker.eph Subscribers: llvm-commits, tra, jhen, hfinkel, resistor, chandlerc, arsenm Differential Revision: http://reviews.llvm.org/D17013 llvm-svn: 260319
2016-02-10 07:03:22 +08:00
ret i32 %a
}
[attrs] Handle convergent CallSites. Summary: Previously we had a notion of convergent functions but not of convergent calls. This is insufficient to correctly analyze calls where the target is unknown, e.g. indirect calls. Now a call is convergent if it targets a known-convergent function, or if it's explicitly marked as convergent. As usual, we can remove convergent where we can prove that no convergent operations are performed in the call. Originally landed as r261544, then reverted in r261544 for (incidental) build breakage. Re-landed here with no changes. Reviewers: chandlerc, jingyue Subscribers: llvm-commits, tra, jhen, hfinkel Differential Revision: http://reviews.llvm.org/D17739 llvm-svn: 263481
2016-03-15 04:18:54 +08:00
; Convergent should not be removed on the function here. Although the call is
; not explicitly convergent, it picks up the convergent attr from the callee.
;
[attrs] Handle convergent CallSites. Summary: Previously we had a notion of convergent functions but not of convergent calls. This is insufficient to correctly analyze calls where the target is unknown, e.g. indirect calls. Now a call is convergent if it targets a known-convergent function, or if it's explicitly marked as convergent. As usual, we can remove convergent where we can prove that no convergent operations are performed in the call. Reviewers: chandlerc, jingyue Subscribers: hfinkel, jhen, tra, llvm-commits Differential Revision: http://reviews.llvm.org/D17317 llvm-svn: 261544
2016-02-23 01:51:35 +08:00
; CHECK: Function Attrs
; CHECK-SAME: convergent
[attrs] Handle convergent CallSites. Summary: Previously we had a notion of convergent functions but not of convergent calls. This is insufficient to correctly analyze calls where the target is unknown, e.g. indirect calls. Now a call is convergent if it targets a known-convergent function, or if it's explicitly marked as convergent. As usual, we can remove convergent where we can prove that no convergent operations are performed in the call. Originally landed as r261544, then reverted in r261544 for (incidental) build breakage. Re-landed here with no changes. Reviewers: chandlerc, jingyue Subscribers: llvm-commits, tra, jhen, hfinkel Differential Revision: http://reviews.llvm.org/D17739 llvm-svn: 263481
2016-03-15 04:18:54 +08:00
; CHECK-NEXT: define i32 @extern_non_convergent_call()
define i32 @extern_non_convergent_call() convergent {
%a = call i32 @k()
Revert "[attrs] Handle convergent CallSites." This reverts r261544, which was causing a test failure in Transforms/FunctionAttrs/readattrs.ll. llvm-svn: 261549
2016-02-23 02:24:43 +08:00
ret i32 %a
[attrs] Handle convergent CallSites. Summary: Previously we had a notion of convergent functions but not of convergent calls. This is insufficient to correctly analyze calls where the target is unknown, e.g. indirect calls. Now a call is convergent if it targets a known-convergent function, or if it's explicitly marked as convergent. As usual, we can remove convergent where we can prove that no convergent operations are performed in the call. Reviewers: chandlerc, jingyue Subscribers: hfinkel, jhen, tra, llvm-commits Differential Revision: http://reviews.llvm.org/D17317 llvm-svn: 261544
2016-02-23 01:51:35 +08:00
}
[attrs] Handle convergent CallSites. Summary: Previously we had a notion of convergent functions but not of convergent calls. This is insufficient to correctly analyze calls where the target is unknown, e.g. indirect calls. Now a call is convergent if it targets a known-convergent function, or if it's explicitly marked as convergent. As usual, we can remove convergent where we can prove that no convergent operations are performed in the call. Originally landed as r261544, then reverted in r261544 for (incidental) build breakage. Re-landed here with no changes. Reviewers: chandlerc, jingyue Subscribers: llvm-commits, tra, jhen, hfinkel Differential Revision: http://reviews.llvm.org/D17739 llvm-svn: 263481
2016-03-15 04:18:54 +08:00
; CHECK: Function Attrs
; CHECK-SAME: convergent
; CHECK-NEXT: define i32 @indirect_convergent_call(
define i32 @indirect_convergent_call(i32 ()* %f) convergent {
%a = call i32 %f() convergent
ret i32 %a
}
; Give indirect_non_convergent_call the norecurse attribute so we get a
; "Function Attrs" comment in the output.
;
; CHECK: Function Attrs
[PM][FunctionAttrs] add NoUnwind attribute inference to PostOrderFunctionAttrs pass Summary: This was motivated by absence of PrunEH functionality in new PM. It was decided that a proper way to do PruneEH is to add NoUnwind inference into PostOrderFunctionAttrs and then perform normal SimplifyCFG on top. This change generalizes attribute handling implemented for (a removal of) Convergent attribute, by introducing a generic builder-like class AttributeInferer It registers all the attribute inference requests, storing per-attribute predicates into a vector, and then goes through an SCC Node, scanning all the instructions for not breaking attribute assumptions. The main idea is that as soon all the instructions from all the functions of SCC Node conform to attribute assumptions then we are free to infer the attribute as set for all the functions of SCC Node. It handles two distinct cases of attributes: - those that might break due to derefinement of the function code for these attributes we are allowed to apply inference only if all the functions are "exact definitions". Example - NoUnwind. - those that do not care about derefinement for these attributes we are allowed to apply inference as soon as we see any function definition. Example - removal of Convergent attribute. Also in this commit: * Converted all the FunctionAttrs tests to use FileCheck and added new-PM invocations to them * FunctionAttrs/convergent.ll test demonstrates a difference in behavior between new and old PM implementations. Marked with FIXME. * PruneEH tests were converted to new-PM as well, using function-attrs+simplify-cfg combo as intended * some of "other" tests were updated since function-attrs now infers 'nounwind' even for old PM pipeline * -disable-nounwind-inference hidden option added as a possible workaround for a supposedly rare case when nounwind being inferred by default presents a problem Reviewers: chandlerc, jlebar Reviewed By: jlebar Subscribers: eraman, llvm-commits Differential Revision: https://reviews.llvm.org/D44415 llvm-svn: 328377
2018-03-24 05:46:16 +08:00
; CHECK-INDIRECT-NOT: convergent
; CHECK-INDIRECT-NEXT: define i32 @indirect_non_convergent_call(
[attrs] Handle convergent CallSites. Summary: Previously we had a notion of convergent functions but not of convergent calls. This is insufficient to correctly analyze calls where the target is unknown, e.g. indirect calls. Now a call is convergent if it targets a known-convergent function, or if it's explicitly marked as convergent. As usual, we can remove convergent where we can prove that no convergent operations are performed in the call. Originally landed as r261544, then reverted in r261544 for (incidental) build breakage. Re-landed here with no changes. Reviewers: chandlerc, jingyue Subscribers: llvm-commits, tra, jhen, hfinkel Differential Revision: http://reviews.llvm.org/D17739 llvm-svn: 263481
2016-03-15 04:18:54 +08:00
define i32 @indirect_non_convergent_call(i32 ()* %f) convergent norecurse {
%a = call i32 %f()
ret i32 %a
}
Add convergent-removing bits to FunctionAttrs pass. Summary: Remove the convergent attribute on any functions which provably do not contain or invoke any convergent functions. After this change, we'll be able to modify clang to conservatively add 'convergent' to all functions when compiling CUDA. Reviewers: jingyue, joker.eph Subscribers: llvm-commits, tra, jhen, hfinkel, resistor, chandlerc, arsenm Differential Revision: http://reviews.llvm.org/D17013 llvm-svn: 260319
2016-02-10 07:03:22 +08:00
; CHECK: Function Attrs
; CHECK-SAME: convergent
NVPTX: Replace uses of cuda.syncthreads with nvvm.barrier0 Everywhere where cuda.syncthreads or __syncthreads is used, use the properly namespaced nvvm.barrier0 instead. llvm-svn: 274664
2016-07-07 04:02:45 +08:00
; CHECK-NEXT: declare void @llvm.nvvm.barrier0()
declare void @llvm.nvvm.barrier0() convergent
Add convergent-removing bits to FunctionAttrs pass. Summary: Remove the convergent attribute on any functions which provably do not contain or invoke any convergent functions. After this change, we'll be able to modify clang to conservatively add 'convergent' to all functions when compiling CUDA. Reviewers: jingyue, joker.eph Subscribers: llvm-commits, tra, jhen, hfinkel, resistor, chandlerc, arsenm Differential Revision: http://reviews.llvm.org/D17013 llvm-svn: 260319
2016-02-10 07:03:22 +08:00
; CHECK: Function Attrs
; CHECK-SAME: convergent
; CHECK-NEXT: define i32 @intrinsic()
define i32 @intrinsic() convergent {
[attrs] Handle convergent CallSites. Summary: Previously we had a notion of convergent functions but not of convergent calls. This is insufficient to correctly analyze calls where the target is unknown, e.g. indirect calls. Now a call is convergent if it targets a known-convergent function, or if it's explicitly marked as convergent. As usual, we can remove convergent where we can prove that no convergent operations are performed in the call. Originally landed as r261544, then reverted in r261544 for (incidental) build breakage. Re-landed here with no changes. Reviewers: chandlerc, jingyue Subscribers: llvm-commits, tra, jhen, hfinkel Differential Revision: http://reviews.llvm.org/D17739 llvm-svn: 263481
2016-03-15 04:18:54 +08:00
; Implicitly convergent, because the intrinsic is convergent.
NVPTX: Replace uses of cuda.syncthreads with nvvm.barrier0 Everywhere where cuda.syncthreads or __syncthreads is used, use the properly namespaced nvvm.barrier0 instead. llvm-svn: 274664
2016-07-07 04:02:45 +08:00
call void @llvm.nvvm.barrier0()
Add convergent-removing bits to FunctionAttrs pass. Summary: Remove the convergent attribute on any functions which provably do not contain or invoke any convergent functions. After this change, we'll be able to modify clang to conservatively add 'convergent' to all functions when compiling CUDA. Reviewers: jingyue, joker.eph Subscribers: llvm-commits, tra, jhen, hfinkel, resistor, chandlerc, arsenm Differential Revision: http://reviews.llvm.org/D17013 llvm-svn: 260319
2016-02-10 07:03:22 +08:00
ret i32 0
}
; CHECK: Function Attrs
; CHECK-NOT: convergent
; CHECK-NEXT: define i32 @recursive1()
define i32 @recursive1() convergent {
[attrs] Handle convergent CallSites. Summary: Previously we had a notion of convergent functions but not of convergent calls. This is insufficient to correctly analyze calls where the target is unknown, e.g. indirect calls. Now a call is convergent if it targets a known-convergent function, or if it's explicitly marked as convergent. As usual, we can remove convergent where we can prove that no convergent operations are performed in the call. Originally landed as r261544, then reverted in r261544 for (incidental) build breakage. Re-landed here with no changes. Reviewers: chandlerc, jingyue Subscribers: llvm-commits, tra, jhen, hfinkel Differential Revision: http://reviews.llvm.org/D17739 llvm-svn: 263481
2016-03-15 04:18:54 +08:00
%a = call i32 @recursive2() convergent
Add convergent-removing bits to FunctionAttrs pass. Summary: Remove the convergent attribute on any functions which provably do not contain or invoke any convergent functions. After this change, we'll be able to modify clang to conservatively add 'convergent' to all functions when compiling CUDA. Reviewers: jingyue, joker.eph Subscribers: llvm-commits, tra, jhen, hfinkel, resistor, chandlerc, arsenm Differential Revision: http://reviews.llvm.org/D17013 llvm-svn: 260319
2016-02-10 07:03:22 +08:00
ret i32 %a
}
; CHECK: Function Attrs
; CHECK-NOT: convergent
; CHECK-NEXT: define i32 @recursive2()
define i32 @recursive2() convergent {
[attrs] Handle convergent CallSites. Summary: Previously we had a notion of convergent functions but not of convergent calls. This is insufficient to correctly analyze calls where the target is unknown, e.g. indirect calls. Now a call is convergent if it targets a known-convergent function, or if it's explicitly marked as convergent. As usual, we can remove convergent where we can prove that no convergent operations are performed in the call. Originally landed as r261544, then reverted in r261544 for (incidental) build breakage. Re-landed here with no changes. Reviewers: chandlerc, jingyue Subscribers: llvm-commits, tra, jhen, hfinkel Differential Revision: http://reviews.llvm.org/D17739 llvm-svn: 263481
2016-03-15 04:18:54 +08:00
%a = call i32 @recursive1() convergent
Add convergent-removing bits to FunctionAttrs pass. Summary: Remove the convergent attribute on any functions which provably do not contain or invoke any convergent functions. After this change, we'll be able to modify clang to conservatively add 'convergent' to all functions when compiling CUDA. Reviewers: jingyue, joker.eph Subscribers: llvm-commits, tra, jhen, hfinkel, resistor, chandlerc, arsenm Differential Revision: http://reviews.llvm.org/D17013 llvm-svn: 260319
2016-02-10 07:03:22 +08:00
ret i32 %a
}
; CHECK: Function Attrs
; CHECK-SAME: convergent
; CHECK-NEXT: define i32 @noopt()
define i32 @noopt() convergent optnone noinline {
[attrs] Handle convergent CallSites. Summary: Previously we had a notion of convergent functions but not of convergent calls. This is insufficient to correctly analyze calls where the target is unknown, e.g. indirect calls. Now a call is convergent if it targets a known-convergent function, or if it's explicitly marked as convergent. As usual, we can remove convergent where we can prove that no convergent operations are performed in the call. Originally landed as r261544, then reverted in r261544 for (incidental) build breakage. Re-landed here with no changes. Reviewers: chandlerc, jingyue Subscribers: llvm-commits, tra, jhen, hfinkel Differential Revision: http://reviews.llvm.org/D17739 llvm-svn: 263481
2016-03-15 04:18:54 +08:00
%a = call i32 @noopt_friend() convergent
Add convergent-removing bits to FunctionAttrs pass. Summary: Remove the convergent attribute on any functions which provably do not contain or invoke any convergent functions. After this change, we'll be able to modify clang to conservatively add 'convergent' to all functions when compiling CUDA. Reviewers: jingyue, joker.eph Subscribers: llvm-commits, tra, jhen, hfinkel, resistor, chandlerc, arsenm Differential Revision: http://reviews.llvm.org/D17013 llvm-svn: 260319
2016-02-10 07:03:22 +08:00
ret i32 0
}
; A function which is mutually-recursive with a convergent, optnone function
; shouldn't have its convergent attribute stripped.
; CHECK: Function Attrs
; CHECK-SAME: convergent
; CHECK-NEXT: define i32 @noopt_friend()
define i32 @noopt_friend() convergent {
%a = call i32 @noopt()
ret i32 0
}