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/Attributor/nosync.ll

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

520 lines
23 KiB
LLVM
Raw Normal View History

[Attributor][NFC] Update tests after D94741 The update_test_checks script can now check for global symbols and is able to handle them properly when they differ across prefixes, e.g., attribute #0 might be different in different runs. This patch simply updates all the Attributor tests with the new script. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D97906
2021-03-12 13:21:32 +08:00
; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --function-signature --check-attributes --check-globals
[Attributor][FIX] Liveness handling in the isAssumedDead helpers This fixes a conceptual problem with our AAIsDead usage which conflated call site liveness with call site return value liveness. Without the fix tests would obviously miscompile as we make genericValueTraversal more powerful (in a follow up). The effects on the tests are mixed but mostly marginal. The most prominent one is the lack of `noreturn` for functions. The reason is that we make entire blocks live at the same time (for time reasons). Now that we actually look at the block liveness, which we need to do, the return instructions are live and will survive. As an example, `noreturn_async.ll` has been modified to retain the `noreturn` even with block granularity. We could address this easily but there is little need in practice.
2022-02-01 10:23:18 +08:00
; RUN: opt -attributor -enable-new-pm=0 -attributor-manifest-internal -attributor-max-iterations-verify -attributor-annotate-decl-cs -attributor-max-iterations=3 -S < %s | FileCheck %s --check-prefixes=CHECK,NOT_CGSCC_NPM,NOT_CGSCC_OPM,NOT_TUNIT_NPM,IS__TUNIT____,IS________OPM,IS__TUNIT_OPM
; RUN: opt -aa-pipeline=basic-aa -passes=attributor -attributor-manifest-internal -attributor-max-iterations-verify -attributor-annotate-decl-cs -attributor-max-iterations=3 -S < %s | FileCheck %s --check-prefixes=CHECK,NOT_CGSCC_OPM,NOT_CGSCC_NPM,NOT_TUNIT_OPM,IS__TUNIT____,IS________NPM,IS__TUNIT_NPM
[NewPM][Attributor] Pin tests with -attributor to legacy PM All these tests already explicitly test against both legacy PM and NPM. $ sed -i 's/ -attributor / -attributor -enable-new-pm=0 /g' $(rg --path-separator // -l -- -passes=) $ sed -i 's/ -attributor-cgscc / -attributor-cgscc -enable-new-pm=0 /g' $(rg --path-separator // -l -- -passes=) Now all tests in Transforms/Attributor/ pass under NPM. Reviewed By: jdoerfert Differential Revision: https://reviews.llvm.org/D84813
2020-07-29 08:57:21 +08:00
; RUN: opt -attributor-cgscc -enable-new-pm=0 -attributor-manifest-internal -attributor-annotate-decl-cs -S < %s | FileCheck %s --check-prefixes=CHECK,NOT_TUNIT_NPM,NOT_TUNIT_OPM,NOT_CGSCC_NPM,IS__CGSCC____,IS________OPM,IS__CGSCC_OPM
[Attributor][NFC] Remove obsolete option from tests Since D76871 it is sufficient to run `opt -atributor` or `-attributor-cgscc`.
2020-04-18 10:09:16 +08:00
; RUN: opt -aa-pipeline=basic-aa -passes=attributor-cgscc -attributor-manifest-internal -attributor-annotate-decl-cs -S < %s | FileCheck %s --check-prefixes=CHECK,NOT_TUNIT_NPM,NOT_TUNIT_OPM,NOT_CGSCC_OPM,IS__CGSCC____,IS________NPM,IS__CGSCC_NPM
[Attributor] Deduce "nosync" function attribute. Introduce and deduce "nosync" function attribute to indicate that a function does not synchronize with another thread in a way that other thread might free memory. Reviewers: jdoerfert, jfb, nhaehnle, arsenm Subscribers: wdng, hfinkel, nhaenhle, mehdi_amini, steven_wu, dexonsmith, arsenm, uenoku, hiraditya, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D62766 llvm-svn: 365830
2019-07-12 05:37:40 +08:00
target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
; Test cases designed for the nosync function attribute.
; FIXME's are used to indicate problems and missing attributes.
; struct RT {
; char A;
; int B[10][20];
; char C;
; };
; struct ST {
; int X;
; double Y;
; struct RT Z;
; };
;
; int *foo(struct ST *s) {
; return &s[1].Z.B[5][13];
; }
; TEST 1
; non-convergent and readnone implies nosync
%struct.RT = type { i8, [10 x [20 x i32]], i8 }
%struct.ST = type { i32, double, %struct.RT }
[Attributor][NFC] Update tests after D94741 The update_test_checks script can now check for global symbols and is able to handle them properly when they differ across prefixes, e.g., attribute #0 might be different in different runs. This patch simply updates all the Attributor tests with the new script. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D97906
2021-03-12 13:21:32 +08:00
;.
; CHECK: @[[A:[a-zA-Z0-9_$"\\.-]+]] = common global i32 0, align 4
;.
[Attributor] Deduce "nosync" function attribute. Introduce and deduce "nosync" function attribute to indicate that a function does not synchronize with another thread in a way that other thread might free memory. Reviewers: jdoerfert, jfb, nhaehnle, arsenm Subscribers: wdng, hfinkel, nhaenhle, mehdi_amini, steven_wu, dexonsmith, arsenm, uenoku, hiraditya, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D62766 llvm-svn: 365830
2019-07-12 05:37:40 +08:00
define i32* @foo(%struct.ST* %s) nounwind uwtable readnone optsize ssp {
Extend the `uwtable` attribute with unwind table kind We have the `clang -cc1` command-line option `-funwind-tables=1|2` and the codegen option `VALUE_CODEGENOPT(UnwindTables, 2, 0) ///< Unwind tables (1) or asynchronous unwind tables (2)`. However, this is encoded in LLVM IR by the presence or the absence of the `uwtable` attribute, i.e. we lose the information whether to generate want just some unwind tables or asynchronous unwind tables. Asynchronous unwind tables take more space in the runtime image, I'd estimate something like 80-90% more, as the difference is adding roughly the same number of CFI directives as for prologues, only a bit simpler (e.g. `.cfi_offset reg, off` vs. `.cfi_restore reg`). Or even more, if you consider tail duplication of epilogue blocks. Asynchronous unwind tables could also restrict code generation to having only a finite number of frame pointer adjustments (an example of *not* having a finite number of `SP` adjustments is on AArch64 when untagging the stack (MTE) in some cases the compiler can modify `SP` in a loop). Having the CFI precise up to an instruction generally also means one cannot bundle together CFI instructions once the prologue is done, they need to be interspersed with ordinary instructions, which means extra `DW_CFA_advance_loc` commands, further increasing the unwind tables size. That is to say, async unwind tables impose a non-negligible overhead, yet for the most common use cases (like C++ exceptions), they are not even needed. This patch extends the `uwtable` attribute with an optional value: - `uwtable` (default to `async`) - `uwtable(sync)`, synchronous unwind tables - `uwtable(async)`, asynchronous (instruction precise) unwind tables Reviewed By: MaskRay Differential Revision: https://reviews.llvm.org/D114543
2022-02-14 21:41:34 +08:00
; CHECK: Function Attrs: nofree norecurse nosync nounwind optsize readnone ssp willreturn uwtable
[Attributor] Use AAFunctionReachability to determine AANoRecurse We missed out on AANoRecurse in the module pass because we had no call graph. With AAFunctionReachability we can simply ask if the function may reach itself. Differential Revision: https://reviews.llvm.org/D110099
2021-09-21 03:41:52 +08:00
; CHECK-LABEL: define {{[^@]+}}@foo
; CHECK-SAME: (%struct.ST* nofree readnone "no-capture-maybe-returned" [[S:%.*]]) #[[ATTR0:[0-9]+]] {
; CHECK-NEXT: entry:
; CHECK-NEXT: [[ARRAYIDX:%.*]] = getelementptr inbounds [[STRUCT_ST:%.*]], %struct.ST* [[S]], i64 1, i32 2, i32 1, i64 5, i64 13
; CHECK-NEXT: ret i32* [[ARRAYIDX]]
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
;
[Attributor] Deduce "nosync" function attribute. Introduce and deduce "nosync" function attribute to indicate that a function does not synchronize with another thread in a way that other thread might free memory. Reviewers: jdoerfert, jfb, nhaehnle, arsenm Subscribers: wdng, hfinkel, nhaenhle, mehdi_amini, steven_wu, dexonsmith, arsenm, uenoku, hiraditya, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D62766 llvm-svn: 365830
2019-07-12 05:37:40 +08:00
entry:
%arrayidx = getelementptr inbounds %struct.ST, %struct.ST* %s, i64 1, i32 2, i32 1, i64 5, i64 13
ret i32* %arrayidx
}
; TEST 2
; atomic load with monotonic ordering
; int load_monotonic(_Atomic int *num) {
; int n = atomic_load_explicit(num, memory_order_relaxed);
; return n;
; }
IR: print value numbers for unnamed function arguments For consistency with normal instructions and clarity when reading IR, it's best to print the %0, %1, ... names of function arguments in definitions. Also modifies the parser to accept IR in that form for obvious reasons. llvm-svn: 367755
2019-08-03 22:28:34 +08:00
define i32 @load_monotonic(i32* nocapture readonly %0) norecurse nounwind uwtable {
Extend the `uwtable` attribute with unwind table kind We have the `clang -cc1` command-line option `-funwind-tables=1|2` and the codegen option `VALUE_CODEGENOPT(UnwindTables, 2, 0) ///< Unwind tables (1) or asynchronous unwind tables (2)`. However, this is encoded in LLVM IR by the presence or the absence of the `uwtable` attribute, i.e. we lose the information whether to generate want just some unwind tables or asynchronous unwind tables. Asynchronous unwind tables take more space in the runtime image, I'd estimate something like 80-90% more, as the difference is adding roughly the same number of CFI directives as for prologues, only a bit simpler (e.g. `.cfi_offset reg, off` vs. `.cfi_restore reg`). Or even more, if you consider tail duplication of epilogue blocks. Asynchronous unwind tables could also restrict code generation to having only a finite number of frame pointer adjustments (an example of *not* having a finite number of `SP` adjustments is on AArch64 when untagging the stack (MTE) in some cases the compiler can modify `SP` in a loop). Having the CFI precise up to an instruction generally also means one cannot bundle together CFI instructions once the prologue is done, they need to be interspersed with ordinary instructions, which means extra `DW_CFA_advance_loc` commands, further increasing the unwind tables size. That is to say, async unwind tables impose a non-negligible overhead, yet for the most common use cases (like C++ exceptions), they are not even needed. This patch extends the `uwtable` attribute with an optional value: - `uwtable` (default to `async`) - `uwtable(sync)`, synchronous unwind tables - `uwtable(async)`, asynchronous (instruction precise) unwind tables Reviewed By: MaskRay Differential Revision: https://reviews.llvm.org/D114543
2022-02-14 21:41:34 +08:00
; CHECK: Function Attrs: argmemonly nofree norecurse nosync nounwind willreturn uwtable
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
; CHECK-LABEL: define {{[^@]+}}@load_monotonic
[Attributor][NFC] Update tests after D94741 The update_test_checks script can now check for global symbols and is able to handle them properly when they differ across prefixes, e.g., attribute #0 might be different in different runs. This patch simply updates all the Attributor tests with the new script. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D97906
2021-03-12 13:21:32 +08:00
; CHECK-SAME: (i32* nocapture nofree noundef nonnull readonly align 4 dereferenceable(4) [[TMP0:%.*]]) #[[ATTR1:[0-9]+]] {
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
; CHECK-NEXT: [[TMP2:%.*]] = load atomic i32, i32* [[TMP0]] monotonic, align 4
; CHECK-NEXT: ret i32 [[TMP2]]
;
[Attributor] Deduce "nosync" function attribute. Introduce and deduce "nosync" function attribute to indicate that a function does not synchronize with another thread in a way that other thread might free memory. Reviewers: jdoerfert, jfb, nhaehnle, arsenm Subscribers: wdng, hfinkel, nhaenhle, mehdi_amini, steven_wu, dexonsmith, arsenm, uenoku, hiraditya, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D62766 llvm-svn: 365830
2019-07-12 05:37:40 +08:00
%2 = load atomic i32, i32* %0 monotonic, align 4
ret i32 %2
}
; TEST 3
; atomic store with monotonic ordering.
; void store_monotonic(_Atomic int *num) {
; atomic_load_explicit(num, memory_order_relaxed);
; }
IR: print value numbers for unnamed function arguments For consistency with normal instructions and clarity when reading IR, it's best to print the %0, %1, ... names of function arguments in definitions. Also modifies the parser to accept IR in that form for obvious reasons. llvm-svn: 367755
2019-08-03 22:28:34 +08:00
define void @store_monotonic(i32* nocapture %0) norecurse nounwind uwtable {
Extend the `uwtable` attribute with unwind table kind We have the `clang -cc1` command-line option `-funwind-tables=1|2` and the codegen option `VALUE_CODEGENOPT(UnwindTables, 2, 0) ///< Unwind tables (1) or asynchronous unwind tables (2)`. However, this is encoded in LLVM IR by the presence or the absence of the `uwtable` attribute, i.e. we lose the information whether to generate want just some unwind tables or asynchronous unwind tables. Asynchronous unwind tables take more space in the runtime image, I'd estimate something like 80-90% more, as the difference is adding roughly the same number of CFI directives as for prologues, only a bit simpler (e.g. `.cfi_offset reg, off` vs. `.cfi_restore reg`). Or even more, if you consider tail duplication of epilogue blocks. Asynchronous unwind tables could also restrict code generation to having only a finite number of frame pointer adjustments (an example of *not* having a finite number of `SP` adjustments is on AArch64 when untagging the stack (MTE) in some cases the compiler can modify `SP` in a loop). Having the CFI precise up to an instruction generally also means one cannot bundle together CFI instructions once the prologue is done, they need to be interspersed with ordinary instructions, which means extra `DW_CFA_advance_loc` commands, further increasing the unwind tables size. That is to say, async unwind tables impose a non-negligible overhead, yet for the most common use cases (like C++ exceptions), they are not even needed. This patch extends the `uwtable` attribute with an optional value: - `uwtable` (default to `async`) - `uwtable(sync)`, synchronous unwind tables - `uwtable(async)`, asynchronous (instruction precise) unwind tables Reviewed By: MaskRay Differential Revision: https://reviews.llvm.org/D114543
2022-02-14 21:41:34 +08:00
; CHECK: Function Attrs: argmemonly nofree norecurse nosync nounwind willreturn uwtable
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
; CHECK-LABEL: define {{[^@]+}}@store_monotonic
[Attributor][NFC] Update tests after D94741 The update_test_checks script can now check for global symbols and is able to handle them properly when they differ across prefixes, e.g., attribute #0 might be different in different runs. This patch simply updates all the Attributor tests with the new script. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D97906
2021-03-12 13:21:32 +08:00
; CHECK-SAME: (i32* nocapture nofree noundef nonnull writeonly align 4 dereferenceable(4) [[TMP0:%.*]]) #[[ATTR1]] {
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
; CHECK-NEXT: store atomic i32 10, i32* [[TMP0]] monotonic, align 4
; CHECK-NEXT: ret void
;
[Attributor] Deduce "nosync" function attribute. Introduce and deduce "nosync" function attribute to indicate that a function does not synchronize with another thread in a way that other thread might free memory. Reviewers: jdoerfert, jfb, nhaehnle, arsenm Subscribers: wdng, hfinkel, nhaenhle, mehdi_amini, steven_wu, dexonsmith, arsenm, uenoku, hiraditya, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D62766 llvm-svn: 365830
2019-07-12 05:37:40 +08:00
store atomic i32 10, i32* %0 monotonic, align 4
ret void
}
; TEST 4 - negative, should not deduce nosync
; atomic load with acquire ordering.
; int load_acquire(_Atomic int *num) {
; int n = atomic_load_explicit(num, memory_order_acquire);
; return n;
; }
IR: print value numbers for unnamed function arguments For consistency with normal instructions and clarity when reading IR, it's best to print the %0, %1, ... names of function arguments in definitions. Also modifies the parser to accept IR in that form for obvious reasons. llvm-svn: 367755
2019-08-03 22:28:34 +08:00
define i32 @load_acquire(i32* nocapture readonly %0) norecurse nounwind uwtable {
Extend the `uwtable` attribute with unwind table kind We have the `clang -cc1` command-line option `-funwind-tables=1|2` and the codegen option `VALUE_CODEGENOPT(UnwindTables, 2, 0) ///< Unwind tables (1) or asynchronous unwind tables (2)`. However, this is encoded in LLVM IR by the presence or the absence of the `uwtable` attribute, i.e. we lose the information whether to generate want just some unwind tables or asynchronous unwind tables. Asynchronous unwind tables take more space in the runtime image, I'd estimate something like 80-90% more, as the difference is adding roughly the same number of CFI directives as for prologues, only a bit simpler (e.g. `.cfi_offset reg, off` vs. `.cfi_restore reg`). Or even more, if you consider tail duplication of epilogue blocks. Asynchronous unwind tables could also restrict code generation to having only a finite number of frame pointer adjustments (an example of *not* having a finite number of `SP` adjustments is on AArch64 when untagging the stack (MTE) in some cases the compiler can modify `SP` in a loop). Having the CFI precise up to an instruction generally also means one cannot bundle together CFI instructions once the prologue is done, they need to be interspersed with ordinary instructions, which means extra `DW_CFA_advance_loc` commands, further increasing the unwind tables size. That is to say, async unwind tables impose a non-negligible overhead, yet for the most common use cases (like C++ exceptions), they are not even needed. This patch extends the `uwtable` attribute with an optional value: - `uwtable` (default to `async`) - `uwtable(sync)`, synchronous unwind tables - `uwtable(async)`, asynchronous (instruction precise) unwind tables Reviewed By: MaskRay Differential Revision: https://reviews.llvm.org/D114543
2022-02-14 21:41:34 +08:00
; CHECK: Function Attrs: argmemonly nofree norecurse nounwind willreturn uwtable
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
; CHECK-LABEL: define {{[^@]+}}@load_acquire
[Attributor][NFC] Update tests after D94741 The update_test_checks script can now check for global symbols and is able to handle them properly when they differ across prefixes, e.g., attribute #0 might be different in different runs. This patch simply updates all the Attributor tests with the new script. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D97906
2021-03-12 13:21:32 +08:00
; CHECK-SAME: (i32* nocapture nofree noundef nonnull readonly align 4 dereferenceable(4) [[TMP0:%.*]]) #[[ATTR2:[0-9]+]] {
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
; CHECK-NEXT: [[TMP2:%.*]] = load atomic i32, i32* [[TMP0]] acquire, align 4
; CHECK-NEXT: ret i32 [[TMP2]]
;
[Attributor] Deduce "nosync" function attribute. Introduce and deduce "nosync" function attribute to indicate that a function does not synchronize with another thread in a way that other thread might free memory. Reviewers: jdoerfert, jfb, nhaehnle, arsenm Subscribers: wdng, hfinkel, nhaenhle, mehdi_amini, steven_wu, dexonsmith, arsenm, uenoku, hiraditya, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D62766 llvm-svn: 365830
2019-07-12 05:37:40 +08:00
%2 = load atomic i32, i32* %0 acquire, align 4
ret i32 %2
}
; TEST 5 - negative, should not deduce nosync
; atomic load with release ordering
; void load_release(_Atomic int *num) {
; atomic_store_explicit(num, 10, memory_order_release);
; }
IR: print value numbers for unnamed function arguments For consistency with normal instructions and clarity when reading IR, it's best to print the %0, %1, ... names of function arguments in definitions. Also modifies the parser to accept IR in that form for obvious reasons. llvm-svn: 367755
2019-08-03 22:28:34 +08:00
define void @load_release(i32* nocapture %0) norecurse nounwind uwtable {
Extend the `uwtable` attribute with unwind table kind We have the `clang -cc1` command-line option `-funwind-tables=1|2` and the codegen option `VALUE_CODEGENOPT(UnwindTables, 2, 0) ///< Unwind tables (1) or asynchronous unwind tables (2)`. However, this is encoded in LLVM IR by the presence or the absence of the `uwtable` attribute, i.e. we lose the information whether to generate want just some unwind tables or asynchronous unwind tables. Asynchronous unwind tables take more space in the runtime image, I'd estimate something like 80-90% more, as the difference is adding roughly the same number of CFI directives as for prologues, only a bit simpler (e.g. `.cfi_offset reg, off` vs. `.cfi_restore reg`). Or even more, if you consider tail duplication of epilogue blocks. Asynchronous unwind tables could also restrict code generation to having only a finite number of frame pointer adjustments (an example of *not* having a finite number of `SP` adjustments is on AArch64 when untagging the stack (MTE) in some cases the compiler can modify `SP` in a loop). Having the CFI precise up to an instruction generally also means one cannot bundle together CFI instructions once the prologue is done, they need to be interspersed with ordinary instructions, which means extra `DW_CFA_advance_loc` commands, further increasing the unwind tables size. That is to say, async unwind tables impose a non-negligible overhead, yet for the most common use cases (like C++ exceptions), they are not even needed. This patch extends the `uwtable` attribute with an optional value: - `uwtable` (default to `async`) - `uwtable(sync)`, synchronous unwind tables - `uwtable(async)`, asynchronous (instruction precise) unwind tables Reviewed By: MaskRay Differential Revision: https://reviews.llvm.org/D114543
2022-02-14 21:41:34 +08:00
; CHECK: Function Attrs: argmemonly nofree norecurse nounwind willreturn uwtable
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
; CHECK-LABEL: define {{[^@]+}}@load_release
[Attributor][NFC] Update tests after D94741 The update_test_checks script can now check for global symbols and is able to handle them properly when they differ across prefixes, e.g., attribute #0 might be different in different runs. This patch simply updates all the Attributor tests with the new script. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D97906
2021-03-12 13:21:32 +08:00
; CHECK-SAME: (i32* nocapture nofree noundef writeonly align 4 [[TMP0:%.*]]) #[[ATTR2]] {
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
; CHECK-NEXT: store atomic volatile i32 10, i32* [[TMP0]] release, align 4
; CHECK-NEXT: ret void
;
[Attributor] Deduce "nosync" function attribute. Introduce and deduce "nosync" function attribute to indicate that a function does not synchronize with another thread in a way that other thread might free memory. Reviewers: jdoerfert, jfb, nhaehnle, arsenm Subscribers: wdng, hfinkel, nhaenhle, mehdi_amini, steven_wu, dexonsmith, arsenm, uenoku, hiraditya, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D62766 llvm-svn: 365830
2019-07-12 05:37:40 +08:00
store atomic volatile i32 10, i32* %0 release, align 4
ret void
}
; TEST 6 - negative volatile, relaxed atomic
IR: print value numbers for unnamed function arguments For consistency with normal instructions and clarity when reading IR, it's best to print the %0, %1, ... names of function arguments in definitions. Also modifies the parser to accept IR in that form for obvious reasons. llvm-svn: 367755
2019-08-03 22:28:34 +08:00
define void @load_volatile_release(i32* nocapture %0) norecurse nounwind uwtable {
Extend the `uwtable` attribute with unwind table kind We have the `clang -cc1` command-line option `-funwind-tables=1|2` and the codegen option `VALUE_CODEGENOPT(UnwindTables, 2, 0) ///< Unwind tables (1) or asynchronous unwind tables (2)`. However, this is encoded in LLVM IR by the presence or the absence of the `uwtable` attribute, i.e. we lose the information whether to generate want just some unwind tables or asynchronous unwind tables. Asynchronous unwind tables take more space in the runtime image, I'd estimate something like 80-90% more, as the difference is adding roughly the same number of CFI directives as for prologues, only a bit simpler (e.g. `.cfi_offset reg, off` vs. `.cfi_restore reg`). Or even more, if you consider tail duplication of epilogue blocks. Asynchronous unwind tables could also restrict code generation to having only a finite number of frame pointer adjustments (an example of *not* having a finite number of `SP` adjustments is on AArch64 when untagging the stack (MTE) in some cases the compiler can modify `SP` in a loop). Having the CFI precise up to an instruction generally also means one cannot bundle together CFI instructions once the prologue is done, they need to be interspersed with ordinary instructions, which means extra `DW_CFA_advance_loc` commands, further increasing the unwind tables size. That is to say, async unwind tables impose a non-negligible overhead, yet for the most common use cases (like C++ exceptions), they are not even needed. This patch extends the `uwtable` attribute with an optional value: - `uwtable` (default to `async`) - `uwtable(sync)`, synchronous unwind tables - `uwtable(async)`, asynchronous (instruction precise) unwind tables Reviewed By: MaskRay Differential Revision: https://reviews.llvm.org/D114543
2022-02-14 21:41:34 +08:00
; CHECK: Function Attrs: argmemonly nofree norecurse nounwind willreturn uwtable
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
; CHECK-LABEL: define {{[^@]+}}@load_volatile_release
[Attributor][NFC] Update tests after D94741 The update_test_checks script can now check for global symbols and is able to handle them properly when they differ across prefixes, e.g., attribute #0 might be different in different runs. This patch simply updates all the Attributor tests with the new script. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D97906
2021-03-12 13:21:32 +08:00
; CHECK-SAME: (i32* nocapture nofree noundef writeonly align 4 [[TMP0:%.*]]) #[[ATTR2]] {
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
; CHECK-NEXT: store atomic volatile i32 10, i32* [[TMP0]] release, align 4
; CHECK-NEXT: ret void
;
[Attributor] Deduce "nosync" function attribute. Introduce and deduce "nosync" function attribute to indicate that a function does not synchronize with another thread in a way that other thread might free memory. Reviewers: jdoerfert, jfb, nhaehnle, arsenm Subscribers: wdng, hfinkel, nhaenhle, mehdi_amini, steven_wu, dexonsmith, arsenm, uenoku, hiraditya, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D62766 llvm-svn: 365830
2019-07-12 05:37:40 +08:00
store atomic volatile i32 10, i32* %0 release, align 4
ret void
}
; TEST 7 - negative, should not deduce nosync
; volatile store.
; void volatile_store(volatile int *num) {
; *num = 14;
; }
IR: print value numbers for unnamed function arguments For consistency with normal instructions and clarity when reading IR, it's best to print the %0, %1, ... names of function arguments in definitions. Also modifies the parser to accept IR in that form for obvious reasons. llvm-svn: 367755
2019-08-03 22:28:34 +08:00
define void @volatile_store(i32* %0) norecurse nounwind uwtable {
Extend the `uwtable` attribute with unwind table kind We have the `clang -cc1` command-line option `-funwind-tables=1|2` and the codegen option `VALUE_CODEGENOPT(UnwindTables, 2, 0) ///< Unwind tables (1) or asynchronous unwind tables (2)`. However, this is encoded in LLVM IR by the presence or the absence of the `uwtable` attribute, i.e. we lose the information whether to generate want just some unwind tables or asynchronous unwind tables. Asynchronous unwind tables take more space in the runtime image, I'd estimate something like 80-90% more, as the difference is adding roughly the same number of CFI directives as for prologues, only a bit simpler (e.g. `.cfi_offset reg, off` vs. `.cfi_restore reg`). Or even more, if you consider tail duplication of epilogue blocks. Asynchronous unwind tables could also restrict code generation to having only a finite number of frame pointer adjustments (an example of *not* having a finite number of `SP` adjustments is on AArch64 when untagging the stack (MTE) in some cases the compiler can modify `SP` in a loop). Having the CFI precise up to an instruction generally also means one cannot bundle together CFI instructions once the prologue is done, they need to be interspersed with ordinary instructions, which means extra `DW_CFA_advance_loc` commands, further increasing the unwind tables size. That is to say, async unwind tables impose a non-negligible overhead, yet for the most common use cases (like C++ exceptions), they are not even needed. This patch extends the `uwtable` attribute with an optional value: - `uwtable` (default to `async`) - `uwtable(sync)`, synchronous unwind tables - `uwtable(async)`, asynchronous (instruction precise) unwind tables Reviewed By: MaskRay Differential Revision: https://reviews.llvm.org/D114543
2022-02-14 21:41:34 +08:00
; CHECK: Function Attrs: argmemonly nofree norecurse nounwind willreturn uwtable
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
; CHECK-LABEL: define {{[^@]+}}@volatile_store
[Attributor][NFC] Update tests after D94741 The update_test_checks script can now check for global symbols and is able to handle them properly when they differ across prefixes, e.g., attribute #0 might be different in different runs. This patch simply updates all the Attributor tests with the new script. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D97906
2021-03-12 13:21:32 +08:00
; CHECK-SAME: (i32* nofree noundef align 4 [[TMP0:%.*]]) #[[ATTR2]] {
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
; CHECK-NEXT: store volatile i32 14, i32* [[TMP0]], align 4
; CHECK-NEXT: ret void
;
[Attributor] Deduce "nosync" function attribute. Introduce and deduce "nosync" function attribute to indicate that a function does not synchronize with another thread in a way that other thread might free memory. Reviewers: jdoerfert, jfb, nhaehnle, arsenm Subscribers: wdng, hfinkel, nhaenhle, mehdi_amini, steven_wu, dexonsmith, arsenm, uenoku, hiraditya, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D62766 llvm-svn: 365830
2019-07-12 05:37:40 +08:00
store volatile i32 14, i32* %0, align 4
ret void
}
; TEST 8 - negative, should not deduce nosync
; volatile load.
; int volatile_load(volatile int *num) {
; int n = *num;
; return n;
; }
IR: print value numbers for unnamed function arguments For consistency with normal instructions and clarity when reading IR, it's best to print the %0, %1, ... names of function arguments in definitions. Also modifies the parser to accept IR in that form for obvious reasons. llvm-svn: 367755
2019-08-03 22:28:34 +08:00
define i32 @volatile_load(i32* %0) norecurse nounwind uwtable {
Extend the `uwtable` attribute with unwind table kind We have the `clang -cc1` command-line option `-funwind-tables=1|2` and the codegen option `VALUE_CODEGENOPT(UnwindTables, 2, 0) ///< Unwind tables (1) or asynchronous unwind tables (2)`. However, this is encoded in LLVM IR by the presence or the absence of the `uwtable` attribute, i.e. we lose the information whether to generate want just some unwind tables or asynchronous unwind tables. Asynchronous unwind tables take more space in the runtime image, I'd estimate something like 80-90% more, as the difference is adding roughly the same number of CFI directives as for prologues, only a bit simpler (e.g. `.cfi_offset reg, off` vs. `.cfi_restore reg`). Or even more, if you consider tail duplication of epilogue blocks. Asynchronous unwind tables could also restrict code generation to having only a finite number of frame pointer adjustments (an example of *not* having a finite number of `SP` adjustments is on AArch64 when untagging the stack (MTE) in some cases the compiler can modify `SP` in a loop). Having the CFI precise up to an instruction generally also means one cannot bundle together CFI instructions once the prologue is done, they need to be interspersed with ordinary instructions, which means extra `DW_CFA_advance_loc` commands, further increasing the unwind tables size. That is to say, async unwind tables impose a non-negligible overhead, yet for the most common use cases (like C++ exceptions), they are not even needed. This patch extends the `uwtable` attribute with an optional value: - `uwtable` (default to `async`) - `uwtable(sync)`, synchronous unwind tables - `uwtable(async)`, asynchronous (instruction precise) unwind tables Reviewed By: MaskRay Differential Revision: https://reviews.llvm.org/D114543
2022-02-14 21:41:34 +08:00
; CHECK: Function Attrs: argmemonly nofree norecurse nounwind willreturn uwtable
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
; CHECK-LABEL: define {{[^@]+}}@volatile_load
[Attributor][FIX] Liveness handling in the isAssumedDead helpers This fixes a conceptual problem with our AAIsDead usage which conflated call site liveness with call site return value liveness. Without the fix tests would obviously miscompile as we make genericValueTraversal more powerful (in a follow up). The effects on the tests are mixed but mostly marginal. The most prominent one is the lack of `noreturn` for functions. The reason is that we make entire blocks live at the same time (for time reasons). Now that we actually look at the block liveness, which we need to do, the return instructions are live and will survive. As an example, `noreturn_async.ll` has been modified to retain the `noreturn` even with block granularity. We could address this easily but there is little need in practice.
2022-02-01 10:23:18 +08:00
; CHECK-SAME: (i32* nofree noundef align 4 [[TMP0:%.*]]) #[[ATTR2]] {
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
; CHECK-NEXT: [[TMP2:%.*]] = load volatile i32, i32* [[TMP0]], align 4
; CHECK-NEXT: ret i32 [[TMP2]]
;
[Attributor] Deduce "nosync" function attribute. Introduce and deduce "nosync" function attribute to indicate that a function does not synchronize with another thread in a way that other thread might free memory. Reviewers: jdoerfert, jfb, nhaehnle, arsenm Subscribers: wdng, hfinkel, nhaenhle, mehdi_amini, steven_wu, dexonsmith, arsenm, uenoku, hiraditya, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D62766 llvm-svn: 365830
2019-07-12 05:37:40 +08:00
%2 = load volatile i32, i32* %0, align 4
ret i32 %2
}
; TEST 9
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
; CHECK: Function Attrs: noinline nosync nounwind uwtable
; CHECK-NEXT: declare void @nosync_function()
[Attributor] Deduce "nosync" function attribute. Introduce and deduce "nosync" function attribute to indicate that a function does not synchronize with another thread in a way that other thread might free memory. Reviewers: jdoerfert, jfb, nhaehnle, arsenm Subscribers: wdng, hfinkel, nhaenhle, mehdi_amini, steven_wu, dexonsmith, arsenm, uenoku, hiraditya, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D62766 llvm-svn: 365830
2019-07-12 05:37:40 +08:00
declare void @nosync_function() noinline nounwind uwtable nosync
define void @call_nosync_function() nounwind uwtable noinline {
[Attributor][NFC] applying update_test_checks with --check-attributes Summary: All tests are updated, except wrapper.ll since it is not working nicely with newly created functions. Reviewers: jdoerfert, uenoku, baziotis, homerdin Subscribers: arphaman, jfb, kuter, bbn, okura, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D84130
2020-07-20 03:55:43 +08:00
; CHECK: Function Attrs: noinline nosync nounwind uwtable
[Attributor][NFC] rerun update_test_checks without --scrub-attributes
2020-08-30 01:31:13 +08:00
; CHECK-LABEL: define {{[^@]+}}@call_nosync_function
[Attributor][NFC] Update tests after D94741 The update_test_checks script can now check for global symbols and is able to handle them properly when they differ across prefixes, e.g., attribute #0 might be different in different runs. This patch simply updates all the Attributor tests with the new script. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D97906
2021-03-12 13:21:32 +08:00
; CHECK-SAME: () #[[ATTR3:[0-9]+]] {
; CHECK-NEXT: tail call void @nosync_function() #[[ATTR4:[0-9]+]]
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
; CHECK-NEXT: ret void
;
[Attributor] Deduce "nosync" function attribute. Introduce and deduce "nosync" function attribute to indicate that a function does not synchronize with another thread in a way that other thread might free memory. Reviewers: jdoerfert, jfb, nhaehnle, arsenm Subscribers: wdng, hfinkel, nhaenhle, mehdi_amini, steven_wu, dexonsmith, arsenm, uenoku, hiraditya, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D62766 llvm-svn: 365830
2019-07-12 05:37:40 +08:00
tail call void @nosync_function() noinline nounwind uwtable
ret void
}
; TEST 10 - negative, should not deduce nosync
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
; CHECK: Function Attrs: noinline nounwind uwtable
; CHECK-NEXT: declare void @might_sync()
[Attributor] Deduce "nosync" function attribute. Introduce and deduce "nosync" function attribute to indicate that a function does not synchronize with another thread in a way that other thread might free memory. Reviewers: jdoerfert, jfb, nhaehnle, arsenm Subscribers: wdng, hfinkel, nhaenhle, mehdi_amini, steven_wu, dexonsmith, arsenm, uenoku, hiraditya, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D62766 llvm-svn: 365830
2019-07-12 05:37:40 +08:00
declare void @might_sync() noinline nounwind uwtable
define void @call_might_sync() nounwind uwtable noinline {
[Attributor][NFC] applying update_test_checks with --check-attributes Summary: All tests are updated, except wrapper.ll since it is not working nicely with newly created functions. Reviewers: jdoerfert, uenoku, baziotis, homerdin Subscribers: arphaman, jfb, kuter, bbn, okura, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D84130
2020-07-20 03:55:43 +08:00
; CHECK: Function Attrs: noinline nounwind uwtable
[Attributor][NFC] rerun update_test_checks without --scrub-attributes
2020-08-30 01:31:13 +08:00
; CHECK-LABEL: define {{[^@]+}}@call_might_sync
[Attributor][NFC] Update tests after D94741 The update_test_checks script can now check for global symbols and is able to handle them properly when they differ across prefixes, e.g., attribute #0 might be different in different runs. This patch simply updates all the Attributor tests with the new script. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D97906
2021-03-12 13:21:32 +08:00
; CHECK-SAME: () #[[ATTR4]] {
; CHECK-NEXT: tail call void @might_sync() #[[ATTR4]]
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
; CHECK-NEXT: ret void
;
[Attributor] Deduce "nosync" function attribute. Introduce and deduce "nosync" function attribute to indicate that a function does not synchronize with another thread in a way that other thread might free memory. Reviewers: jdoerfert, jfb, nhaehnle, arsenm Subscribers: wdng, hfinkel, nhaenhle, mehdi_amini, steven_wu, dexonsmith, arsenm, uenoku, hiraditya, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D62766 llvm-svn: 365830
2019-07-12 05:37:40 +08:00
tail call void @might_sync() noinline nounwind uwtable
ret void
}
[Attributor] Deduce the "no-return" attribute for functions A function is "no-return" if we never reach a return instruction, either because there are none or the ones that exist are dead. Test have been adjusted: - either noreturn was added, or - noreturn was avoided by modifying the code. The new noreturn_{sync,async} test make sure we do handle invoke instructions with a noreturn (and potentially nowunwind) callee correctly, even in the presence of potential asynchronous exceptions. llvm-svn: 367948
2019-08-06 07:22:05 +08:00
; TEST 11 - positive, should deduce nosync
; volatile operation in same scc but dead. Call volatile_load defined in TEST 8.
[Attributor] Deduce "nosync" function attribute. Introduce and deduce "nosync" function attribute to indicate that a function does not synchronize with another thread in a way that other thread might free memory. Reviewers: jdoerfert, jfb, nhaehnle, arsenm Subscribers: wdng, hfinkel, nhaenhle, mehdi_amini, steven_wu, dexonsmith, arsenm, uenoku, hiraditya, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D62766 llvm-svn: 365830
2019-07-12 05:37:40 +08:00
IR: print value numbers for unnamed function arguments For consistency with normal instructions and clarity when reading IR, it's best to print the %0, %1, ... names of function arguments in definitions. Also modifies the parser to accept IR in that form for obvious reasons. llvm-svn: 367755
2019-08-03 22:28:34 +08:00
define i32 @scc1(i32* %0) noinline nounwind uwtable {
[Attributor][FIX] Liveness handling in the isAssumedDead helpers This fixes a conceptual problem with our AAIsDead usage which conflated call site liveness with call site return value liveness. Without the fix tests would obviously miscompile as we make genericValueTraversal more powerful (in a follow up). The effects on the tests are mixed but mostly marginal. The most prominent one is the lack of `noreturn` for functions. The reason is that we make entire blocks live at the same time (for time reasons). Now that we actually look at the block liveness, which we need to do, the return instructions are live and will survive. As an example, `noreturn_async.ll` has been modified to retain the `noreturn` even with block granularity. We could address this easily but there is little need in practice.
2022-02-01 10:23:18 +08:00
; IS__TUNIT____: Function Attrs: argmemonly nofree noinline nounwind uwtable
; IS__TUNIT____-LABEL: define {{[^@]+}}@scc1
; IS__TUNIT____-SAME: (i32* nofree [[TMP0:%.*]]) #[[ATTR5:[0-9]+]] {
[Attributor][FIX] Reachability needs to account for readonly callees The oversight caused us to ignore call sites that are effectively dead when we computed reachability (or more precise the call edges of a function). The problem is that loads in the readonly callee might depend on stores prior to the callee. If we do not track the call edge we mistakenly assumed the store before the call cannot reach the load. The problem is nicely visible in: `llvm/test/Transforms/Attributor/ArgumentPromotion/basictest.ll` Caused by D118673. Fixes https://github.com/llvm/llvm-project/issues/53726
2022-02-08 13:37:02 +08:00
; IS__TUNIT____-NEXT: tail call void @scc2(i32* nofree [[TMP0]]) #[[ATTR18:[0-9]+]]
; IS__TUNIT____-NEXT: [[VAL:%.*]] = tail call i32 @volatile_load(i32* nofree align 4 [[TMP0]]) #[[ATTR18]]
[Attributor][FIX] Liveness handling in the isAssumedDead helpers This fixes a conceptual problem with our AAIsDead usage which conflated call site liveness with call site return value liveness. Without the fix tests would obviously miscompile as we make genericValueTraversal more powerful (in a follow up). The effects on the tests are mixed but mostly marginal. The most prominent one is the lack of `noreturn` for functions. The reason is that we make entire blocks live at the same time (for time reasons). Now that we actually look at the block liveness, which we need to do, the return instructions are live and will survive. As an example, `noreturn_async.ll` has been modified to retain the `noreturn` even with block granularity. We could address this easily but there is little need in practice.
2022-02-01 10:23:18 +08:00
; IS__TUNIT____-NEXT: ret i32 [[VAL]]
;
; IS__CGSCC_OPM: Function Attrs: argmemonly nofree noinline nounwind uwtable
; IS__CGSCC_OPM-LABEL: define {{[^@]+}}@scc1
; IS__CGSCC_OPM-SAME: (i32* nofree [[TMP0:%.*]]) #[[ATTR5:[0-9]+]] {
[Attributor][FIX] Reachability needs to account for readonly callees The oversight caused us to ignore call sites that are effectively dead when we computed reachability (or more precise the call edges of a function). The problem is that loads in the readonly callee might depend on stores prior to the callee. If we do not track the call edge we mistakenly assumed the store before the call cannot reach the load. The problem is nicely visible in: `llvm/test/Transforms/Attributor/ArgumentPromotion/basictest.ll` Caused by D118673. Fixes https://github.com/llvm/llvm-project/issues/53726
2022-02-08 13:37:02 +08:00
; IS__CGSCC_OPM-NEXT: tail call void @scc2(i32* nofree [[TMP0]]) #[[ATTR18:[0-9]+]]
; IS__CGSCC_OPM-NEXT: [[VAL:%.*]] = tail call i32 @volatile_load(i32* nofree noundef align 4 [[TMP0]]) #[[ATTR19:[0-9]+]]
[Attributor][FIX] Liveness handling in the isAssumedDead helpers This fixes a conceptual problem with our AAIsDead usage which conflated call site liveness with call site return value liveness. Without the fix tests would obviously miscompile as we make genericValueTraversal more powerful (in a follow up). The effects on the tests are mixed but mostly marginal. The most prominent one is the lack of `noreturn` for functions. The reason is that we make entire blocks live at the same time (for time reasons). Now that we actually look at the block liveness, which we need to do, the return instructions are live and will survive. As an example, `noreturn_async.ll` has been modified to retain the `noreturn` even with block granularity. We could address this easily but there is little need in practice.
2022-02-01 10:23:18 +08:00
; IS__CGSCC_OPM-NEXT: ret i32 [[VAL]]
;
; IS__CGSCC_NPM: Function Attrs: argmemonly nofree noinline nounwind uwtable
[Attributor][NFC] applying update_test_checks with --check-attributes Summary: All tests are updated, except wrapper.ll since it is not working nicely with newly created functions. Reviewers: jdoerfert, uenoku, baziotis, homerdin Subscribers: arphaman, jfb, kuter, bbn, okura, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D84130
2020-07-20 03:55:43 +08:00
; IS__CGSCC_NPM-LABEL: define {{[^@]+}}@scc1
[Attributor][FIX] Liveness handling in the isAssumedDead helpers This fixes a conceptual problem with our AAIsDead usage which conflated call site liveness with call site return value liveness. Without the fix tests would obviously miscompile as we make genericValueTraversal more powerful (in a follow up). The effects on the tests are mixed but mostly marginal. The most prominent one is the lack of `noreturn` for functions. The reason is that we make entire blocks live at the same time (for time reasons). Now that we actually look at the block liveness, which we need to do, the return instructions are live and will survive. As an example, `noreturn_async.ll` has been modified to retain the `noreturn` even with block granularity. We could address this easily but there is little need in practice.
2022-02-01 10:23:18 +08:00
; IS__CGSCC_NPM-SAME: (i32* nofree [[TMP0:%.*]]) #[[ATTR5:[0-9]+]] {
[Attributor][FIX] Reachability needs to account for readonly callees The oversight caused us to ignore call sites that are effectively dead when we computed reachability (or more precise the call edges of a function). The problem is that loads in the readonly callee might depend on stores prior to the callee. If we do not track the call edge we mistakenly assumed the store before the call cannot reach the load. The problem is nicely visible in: `llvm/test/Transforms/Attributor/ArgumentPromotion/basictest.ll` Caused by D118673. Fixes https://github.com/llvm/llvm-project/issues/53726
2022-02-08 13:37:02 +08:00
; IS__CGSCC_NPM-NEXT: tail call void @scc2(i32* nofree [[TMP0]]) #[[ATTR18:[0-9]+]]
[Attributor][FIX] Liveness handling in the isAssumedDead helpers This fixes a conceptual problem with our AAIsDead usage which conflated call site liveness with call site return value liveness. Without the fix tests would obviously miscompile as we make genericValueTraversal more powerful (in a follow up). The effects on the tests are mixed but mostly marginal. The most prominent one is the lack of `noreturn` for functions. The reason is that we make entire blocks live at the same time (for time reasons). Now that we actually look at the block liveness, which we need to do, the return instructions are live and will survive. As an example, `noreturn_async.ll` has been modified to retain the `noreturn` even with block granularity. We could address this easily but there is little need in practice.
2022-02-01 10:23:18 +08:00
; IS__CGSCC_NPM-NEXT: [[VAL:%.*]] = tail call i32 @volatile_load(i32* nofree noundef align 4 [[TMP0]]) #[[ATTR14:[0-9]+]]
; IS__CGSCC_NPM-NEXT: ret i32 [[VAL]]
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
;
[Attributor] Deduce "nosync" function attribute. Introduce and deduce "nosync" function attribute to indicate that a function does not synchronize with another thread in a way that other thread might free memory. Reviewers: jdoerfert, jfb, nhaehnle, arsenm Subscribers: wdng, hfinkel, nhaenhle, mehdi_amini, steven_wu, dexonsmith, arsenm, uenoku, hiraditya, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D62766 llvm-svn: 365830
2019-07-12 05:37:40 +08:00
tail call void @scc2(i32* %0);
%val = tail call i32 @volatile_load(i32* %0);
ret i32 %val;
}
IR: print value numbers for unnamed function arguments For consistency with normal instructions and clarity when reading IR, it's best to print the %0, %1, ... names of function arguments in definitions. Also modifies the parser to accept IR in that form for obvious reasons. llvm-svn: 367755
2019-08-03 22:28:34 +08:00
define void @scc2(i32* %0) noinline nounwind uwtable {
[Attributor] Introduce the concept of query AAs D106720 introduced features that did not work properly as we could add new queries after a fixpoint was reached and which could not be answered by the information gathered up to the fixpoint alone. As an alternative to D110078, which forced eager computation where we want to continue to be lazy, this patch fixes the problem. QueryAAs are AAs that allow lazy queries during their lifetime. They are never fixed if they have no outstanding dependences and always run as part of the updates in an iteration. To determine if we are done, all query AAs are asked if they received new queries, if not, we only need to consider updated AAs, as before. If new queries are present we go for another iteration. Differential Revision: https://reviews.llvm.org/D118669
2022-02-01 04:16:43 +08:00
; CHECK: Function Attrs: argmemonly nofree noinline nounwind uwtable
; CHECK-LABEL: define {{[^@]+}}@scc2
; CHECK-SAME: (i32* nofree [[TMP0:%.*]]) #[[ATTR5:[0-9]+]] {
[Attributor][FIX] Reachability needs to account for readonly callees The oversight caused us to ignore call sites that are effectively dead when we computed reachability (or more precise the call edges of a function). The problem is that loads in the readonly callee might depend on stores prior to the callee. If we do not track the call edge we mistakenly assumed the store before the call cannot reach the load. The problem is nicely visible in: `llvm/test/Transforms/Attributor/ArgumentPromotion/basictest.ll` Caused by D118673. Fixes https://github.com/llvm/llvm-project/issues/53726
2022-02-08 13:37:02 +08:00
; CHECK-NEXT: [[TMP2:%.*]] = tail call i32 @scc1(i32* nofree [[TMP0]]) #[[ATTR18:[0-9]+]]
[Attributor] Introduce the concept of query AAs D106720 introduced features that did not work properly as we could add new queries after a fixpoint was reached and which could not be answered by the information gathered up to the fixpoint alone. As an alternative to D110078, which forced eager computation where we want to continue to be lazy, this patch fixes the problem. QueryAAs are AAs that allow lazy queries during their lifetime. They are never fixed if they have no outstanding dependences and always run as part of the updates in an iteration. To determine if we are done, all query AAs are asked if they received new queries, if not, we only need to consider updated AAs, as before. If new queries are present we go for another iteration. Differential Revision: https://reviews.llvm.org/D118669
2022-02-01 04:16:43 +08:00
; CHECK-NEXT: ret void
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
;
[Attributor] Deduce "nosync" function attribute. Introduce and deduce "nosync" function attribute to indicate that a function does not synchronize with another thread in a way that other thread might free memory. Reviewers: jdoerfert, jfb, nhaehnle, arsenm Subscribers: wdng, hfinkel, nhaenhle, mehdi_amini, steven_wu, dexonsmith, arsenm, uenoku, hiraditya, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D62766 llvm-svn: 365830
2019-07-12 05:37:40 +08:00
tail call i32 @scc1(i32* %0);
ret void;
}
; TEST 12 - fences, negative
;
; void foo1(int *a, std::atomic<bool> flag){
; *a = 100;
; atomic_thread_fence(std::memory_order_release);
; flag.store(true, std::memory_order_relaxed);
; }
;
; void bar(int *a, std::atomic<bool> flag){
; while(!flag.load(std::memory_order_relaxed))
; ;
;
; atomic_thread_fence(std::memory_order_acquire);
; int b = *a;
; }
%"struct.std::atomic" = type { %"struct.std::__atomic_base" }
%"struct.std::__atomic_base" = type { i8 }
IR: print value numbers for unnamed function arguments For consistency with normal instructions and clarity when reading IR, it's best to print the %0, %1, ... names of function arguments in definitions. Also modifies the parser to accept IR in that form for obvious reasons. llvm-svn: 367755
2019-08-03 22:28:34 +08:00
define void @foo1(i32* %0, %"struct.std::atomic"* %1) {
[Attributor] Introduce the concept of query AAs D106720 introduced features that did not work properly as we could add new queries after a fixpoint was reached and which could not be answered by the information gathered up to the fixpoint alone. As an alternative to D110078, which forced eager computation where we want to continue to be lazy, this patch fixes the problem. QueryAAs are AAs that allow lazy queries during their lifetime. They are never fixed if they have no outstanding dependences and always run as part of the updates in an iteration. To determine if we are done, all query AAs are asked if they received new queries, if not, we only need to consider updated AAs, as before. If new queries are present we go for another iteration. Differential Revision: https://reviews.llvm.org/D118669
2022-02-01 04:16:43 +08:00
; CHECK: Function Attrs: nofree norecurse nounwind willreturn
; CHECK-LABEL: define {{[^@]+}}@foo1
; CHECK-SAME: (i32* nocapture nofree noundef nonnull writeonly align 4 dereferenceable(4) [[TMP0:%.*]], %"struct.std::atomic"* nocapture nofree nonnull writeonly dereferenceable(1) [[TMP1:%.*]]) #[[ATTR6:[0-9]+]] {
; CHECK-NEXT: store i32 100, i32* [[TMP0]], align 4
; CHECK-NEXT: fence release
; CHECK-NEXT: [[TMP3:%.*]] = getelementptr inbounds %"struct.std::atomic", %"struct.std::atomic"* [[TMP1]], i64 0, i32 0, i32 0
; CHECK-NEXT: store atomic i8 1, i8* [[TMP3]] monotonic, align 1
; CHECK-NEXT: ret void
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
;
[Attributor] Deduce "nosync" function attribute. Introduce and deduce "nosync" function attribute to indicate that a function does not synchronize with another thread in a way that other thread might free memory. Reviewers: jdoerfert, jfb, nhaehnle, arsenm Subscribers: wdng, hfinkel, nhaenhle, mehdi_amini, steven_wu, dexonsmith, arsenm, uenoku, hiraditya, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D62766 llvm-svn: 365830
2019-07-12 05:37:40 +08:00
store i32 100, i32* %0, align 4
fence release
%3 = getelementptr inbounds %"struct.std::atomic", %"struct.std::atomic"* %1, i64 0, i32 0, i32 0
store atomic i8 1, i8* %3 monotonic, align 1
ret void
}
IR: print value numbers for unnamed function arguments For consistency with normal instructions and clarity when reading IR, it's best to print the %0, %1, ... names of function arguments in definitions. Also modifies the parser to accept IR in that form for obvious reasons. llvm-svn: 367755
2019-08-03 22:28:34 +08:00
define void @bar(i32* %0, %"struct.std::atomic"* %1) {
[Attributor] Introduce the concept of query AAs D106720 introduced features that did not work properly as we could add new queries after a fixpoint was reached and which could not be answered by the information gathered up to the fixpoint alone. As an alternative to D110078, which forced eager computation where we want to continue to be lazy, this patch fixes the problem. QueryAAs are AAs that allow lazy queries during their lifetime. They are never fixed if they have no outstanding dependences and always run as part of the updates in an iteration. To determine if we are done, all query AAs are asked if they received new queries, if not, we only need to consider updated AAs, as before. If new queries are present we go for another iteration. Differential Revision: https://reviews.llvm.org/D118669
2022-02-01 04:16:43 +08:00
; CHECK: Function Attrs: nofree norecurse nounwind
; CHECK-LABEL: define {{[^@]+}}@bar
; CHECK-SAME: (i32* nocapture nofree readnone [[TMP0:%.*]], %"struct.std::atomic"* nocapture nofree nonnull readonly dereferenceable(1) [[TMP1:%.*]]) #[[ATTR7:[0-9]+]] {
; CHECK-NEXT: [[TMP3:%.*]] = getelementptr inbounds %"struct.std::atomic", %"struct.std::atomic"* [[TMP1]], i64 0, i32 0, i32 0
; CHECK-NEXT: br label [[TMP4:%.*]]
; CHECK: 4:
; CHECK-NEXT: [[TMP5:%.*]] = load atomic i8, i8* [[TMP3]] monotonic, align 1
; CHECK-NEXT: [[TMP6:%.*]] = and i8 [[TMP5]], 1
; CHECK-NEXT: [[TMP7:%.*]] = icmp eq i8 [[TMP6]], 0
; CHECK-NEXT: br i1 [[TMP7]], label [[TMP4]], label [[TMP8:%.*]]
; CHECK: 8:
; CHECK-NEXT: fence acquire
; CHECK-NEXT: ret void
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
;
[Attributor] Deduce "nosync" function attribute. Introduce and deduce "nosync" function attribute to indicate that a function does not synchronize with another thread in a way that other thread might free memory. Reviewers: jdoerfert, jfb, nhaehnle, arsenm Subscribers: wdng, hfinkel, nhaenhle, mehdi_amini, steven_wu, dexonsmith, arsenm, uenoku, hiraditya, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D62766 llvm-svn: 365830
2019-07-12 05:37:40 +08:00
%3 = getelementptr inbounds %"struct.std::atomic", %"struct.std::atomic"* %1, i64 0, i32 0, i32 0
br label %4
4: ; preds = %4, %2
%5 = load atomic i8, i8* %3 monotonic, align 1
%6 = and i8 %5, 1
%7 = icmp eq i8 %6, 0
br i1 %7, label %4, label %8
8: ; preds = %4
fence acquire
ret void
}
; TEST 13 - Fence syncscope("singlethread") seq_cst
[Attributor][NFC] applying update_test_checks with --check-attributes Summary: All tests are updated, except wrapper.ll since it is not working nicely with newly created functions. Reviewers: jdoerfert, uenoku, baziotis, homerdin Subscribers: arphaman, jfb, kuter, bbn, okura, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D84130
2020-07-20 03:55:43 +08:00
define void @foo1_singlethread(i32* %0, %"struct.std::atomic"* %1) {
[Attributor] Introduce the concept of query AAs D106720 introduced features that did not work properly as we could add new queries after a fixpoint was reached and which could not be answered by the information gathered up to the fixpoint alone. As an alternative to D110078, which forced eager computation where we want to continue to be lazy, this patch fixes the problem. QueryAAs are AAs that allow lazy queries during their lifetime. They are never fixed if they have no outstanding dependences and always run as part of the updates in an iteration. To determine if we are done, all query AAs are asked if they received new queries, if not, we only need to consider updated AAs, as before. If new queries are present we go for another iteration. Differential Revision: https://reviews.llvm.org/D118669
2022-02-01 04:16:43 +08:00
; CHECK: Function Attrs: nofree norecurse nosync nounwind willreturn
; CHECK-LABEL: define {{[^@]+}}@foo1_singlethread
; CHECK-SAME: (i32* nocapture nofree noundef nonnull writeonly align 4 dereferenceable(4) [[TMP0:%.*]], %"struct.std::atomic"* nocapture nofree nonnull writeonly dereferenceable(1) [[TMP1:%.*]]) #[[ATTR8:[0-9]+]] {
; CHECK-NEXT: store i32 100, i32* [[TMP0]], align 4
; CHECK-NEXT: fence syncscope("singlethread") release
; CHECK-NEXT: [[TMP3:%.*]] = getelementptr inbounds %"struct.std::atomic", %"struct.std::atomic"* [[TMP1]], i64 0, i32 0, i32 0
; CHECK-NEXT: store atomic i8 1, i8* [[TMP3]] monotonic, align 1
; CHECK-NEXT: ret void
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
;
[Attributor] Deduce "nosync" function attribute. Introduce and deduce "nosync" function attribute to indicate that a function does not synchronize with another thread in a way that other thread might free memory. Reviewers: jdoerfert, jfb, nhaehnle, arsenm Subscribers: wdng, hfinkel, nhaenhle, mehdi_amini, steven_wu, dexonsmith, arsenm, uenoku, hiraditya, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D62766 llvm-svn: 365830
2019-07-12 05:37:40 +08:00
store i32 100, i32* %0, align 4
fence syncscope("singlethread") release
%3 = getelementptr inbounds %"struct.std::atomic", %"struct.std::atomic"* %1, i64 0, i32 0, i32 0
store atomic i8 1, i8* %3 monotonic, align 1
ret void
}
[Attributor][NFC] applying update_test_checks with --check-attributes Summary: All tests are updated, except wrapper.ll since it is not working nicely with newly created functions. Reviewers: jdoerfert, uenoku, baziotis, homerdin Subscribers: arphaman, jfb, kuter, bbn, okura, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D84130
2020-07-20 03:55:43 +08:00
define void @bar_singlethread(i32* %0, %"struct.std::atomic"* %1) {
[Attributor] Introduce the concept of query AAs D106720 introduced features that did not work properly as we could add new queries after a fixpoint was reached and which could not be answered by the information gathered up to the fixpoint alone. As an alternative to D110078, which forced eager computation where we want to continue to be lazy, this patch fixes the problem. QueryAAs are AAs that allow lazy queries during their lifetime. They are never fixed if they have no outstanding dependences and always run as part of the updates in an iteration. To determine if we are done, all query AAs are asked if they received new queries, if not, we only need to consider updated AAs, as before. If new queries are present we go for another iteration. Differential Revision: https://reviews.llvm.org/D118669
2022-02-01 04:16:43 +08:00
; CHECK: Function Attrs: nofree norecurse nosync nounwind
; CHECK-LABEL: define {{[^@]+}}@bar_singlethread
; CHECK-SAME: (i32* nocapture nofree readnone [[TMP0:%.*]], %"struct.std::atomic"* nocapture nofree nonnull readonly dereferenceable(1) [[TMP1:%.*]]) #[[ATTR9:[0-9]+]] {
; CHECK-NEXT: [[TMP3:%.*]] = getelementptr inbounds %"struct.std::atomic", %"struct.std::atomic"* [[TMP1]], i64 0, i32 0, i32 0
; CHECK-NEXT: br label [[TMP4:%.*]]
; CHECK: 4:
; CHECK-NEXT: [[TMP5:%.*]] = load atomic i8, i8* [[TMP3]] monotonic, align 1
; CHECK-NEXT: [[TMP6:%.*]] = and i8 [[TMP5]], 1
; CHECK-NEXT: [[TMP7:%.*]] = icmp eq i8 [[TMP6]], 0
; CHECK-NEXT: br i1 [[TMP7]], label [[TMP4]], label [[TMP8:%.*]]
; CHECK: 8:
; CHECK-NEXT: fence syncscope("singlethread") acquire
; CHECK-NEXT: ret void
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
;
[Attributor] Deduce "nosync" function attribute. Introduce and deduce "nosync" function attribute to indicate that a function does not synchronize with another thread in a way that other thread might free memory. Reviewers: jdoerfert, jfb, nhaehnle, arsenm Subscribers: wdng, hfinkel, nhaenhle, mehdi_amini, steven_wu, dexonsmith, arsenm, uenoku, hiraditya, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D62766 llvm-svn: 365830
2019-07-12 05:37:40 +08:00
%3 = getelementptr inbounds %"struct.std::atomic", %"struct.std::atomic"* %1, i64 0, i32 0, i32 0
br label %4
4: ; preds = %4, %2
%5 = load atomic i8, i8* %3 monotonic, align 1
%6 = and i8 %5, 1
%7 = icmp eq i8 %6, 0
br i1 %7, label %4, label %8
8: ; preds = %4
fence syncscope("singlethread") acquire
ret void
}
declare void @llvm.memcpy(i8* %dest, i8* %src, i32 %len, i1 %isvolatile)
declare void @llvm.memset(i8* %dest, i8 %val, i32 %len, i1 %isvolatile)
; TEST 14 - negative, checking volatile intrinsics.
[Attributor] Deduce "no-capture" argument attribute Add the no-capture argument attribute deduction to the Attributor fixpoint framework. The new string attributed "no-capture-maybe-returned" is introduced to allow deduction of no-capture through functions that "capture" an argument but only by "returning" it. It is only used by the Attributor for testing. Differential Revision: https://reviews.llvm.org/D59922 llvm-svn: 370817
2019-09-04 04:37:24 +08:00
; It is odd to add nocapture but a result of the llvm.memcpy nocapture.
;
[Attributor] Deduce "nosync" function attribute. Introduce and deduce "nosync" function attribute to indicate that a function does not synchronize with another thread in a way that other thread might free memory. Reviewers: jdoerfert, jfb, nhaehnle, arsenm Subscribers: wdng, hfinkel, nhaenhle, mehdi_amini, steven_wu, dexonsmith, arsenm, uenoku, hiraditya, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D62766 llvm-svn: 365830
2019-07-12 05:37:40 +08:00
define i32 @memcpy_volatile(i8* %ptr1, i8* %ptr2) {
[Attributor] Use AAFunctionReachability to determine AANoRecurse We missed out on AANoRecurse in the module pass because we had no call graph. With AAFunctionReachability we can simply ask if the function may reach itself. Differential Revision: https://reviews.llvm.org/D110099
2021-09-21 03:41:52 +08:00
; NOT_CGSCC_OPM: Function Attrs: argmemonly nofree nounwind willreturn
; NOT_CGSCC_OPM-LABEL: define {{[^@]+}}@memcpy_volatile
; NOT_CGSCC_OPM-SAME: (i8* nocapture nofree writeonly [[PTR1:%.*]], i8* nocapture nofree readonly [[PTR2:%.*]]) #[[ATTR10:[0-9]+]] {
[Attributor][FIX] Reachability needs to account for readonly callees The oversight caused us to ignore call sites that are effectively dead when we computed reachability (or more precise the call edges of a function). The problem is that loads in the readonly callee might depend on stores prior to the callee. If we do not track the call edge we mistakenly assumed the store before the call cannot reach the load. The problem is nicely visible in: `llvm/test/Transforms/Attributor/ArgumentPromotion/basictest.ll` Caused by D118673. Fixes https://github.com/llvm/llvm-project/issues/53726
2022-02-08 13:37:02 +08:00
; NOT_CGSCC_OPM-NEXT: call void @llvm.memcpy.p0i8.p0i8.i32(i8* noalias nocapture nofree writeonly [[PTR1]], i8* noalias nocapture nofree readonly [[PTR2]], i32 noundef 8, i1 noundef true) #[[ATTR19:[0-9]+]]
[Attributor] Use AAFunctionReachability to determine AANoRecurse We missed out on AANoRecurse in the module pass because we had no call graph. With AAFunctionReachability we can simply ask if the function may reach itself. Differential Revision: https://reviews.llvm.org/D110099
2021-09-21 03:41:52 +08:00
; NOT_CGSCC_OPM-NEXT: ret i32 4
[IR] Make nosync, nofree and willreturn default for intrinsics. D70365 allows us to make attributes default. This is a follow up to actually make nosync, nofree and willreturn default. The approach we chose, for now, is to opt-in to default attributes to avoid introducing problems to target specific intrinsics. Intrinsics with default attributes can be created using `DefaultAttrsIntrinsic` class.
2020-10-15 19:04:36 +08:00
;
[Attributor][FIX] Liveness handling in the isAssumedDead helpers This fixes a conceptual problem with our AAIsDead usage which conflated call site liveness with call site return value liveness. Without the fix tests would obviously miscompile as we make genericValueTraversal more powerful (in a follow up). The effects on the tests are mixed but mostly marginal. The most prominent one is the lack of `noreturn` for functions. The reason is that we make entire blocks live at the same time (for time reasons). Now that we actually look at the block liveness, which we need to do, the return instructions are live and will survive. As an example, `noreturn_async.ll` has been modified to retain the `noreturn` even with block granularity. We could address this easily but there is little need in practice.
2022-02-01 10:23:18 +08:00
; IS__CGSCC_OPM: Function Attrs: argmemonly nofree nounwind willreturn
; IS__CGSCC_OPM-LABEL: define {{[^@]+}}@memcpy_volatile
[Attributor] Introduce the concept of query AAs D106720 introduced features that did not work properly as we could add new queries after a fixpoint was reached and which could not be answered by the information gathered up to the fixpoint alone. As an alternative to D110078, which forced eager computation where we want to continue to be lazy, this patch fixes the problem. QueryAAs are AAs that allow lazy queries during their lifetime. They are never fixed if they have no outstanding dependences and always run as part of the updates in an iteration. To determine if we are done, all query AAs are asked if they received new queries, if not, we only need to consider updated AAs, as before. If new queries are present we go for another iteration. Differential Revision: https://reviews.llvm.org/D118669
2022-02-01 04:16:43 +08:00
; IS__CGSCC_OPM-SAME: (i8* nocapture nofree writeonly [[PTR1:%.*]], i8* nocapture nofree readonly [[PTR2:%.*]]) #[[ATTR10:[0-9]+]] {
[Attributor][FIX] Reachability needs to account for readonly callees The oversight caused us to ignore call sites that are effectively dead when we computed reachability (or more precise the call edges of a function). The problem is that loads in the readonly callee might depend on stores prior to the callee. If we do not track the call edge we mistakenly assumed the store before the call cannot reach the load. The problem is nicely visible in: `llvm/test/Transforms/Attributor/ArgumentPromotion/basictest.ll` Caused by D118673. Fixes https://github.com/llvm/llvm-project/issues/53726
2022-02-08 13:37:02 +08:00
; IS__CGSCC_OPM-NEXT: call void @llvm.memcpy.p0i8.p0i8.i32(i8* noalias nocapture nofree writeonly [[PTR1]], i8* noalias nocapture nofree readonly [[PTR2]], i32 noundef 8, i1 noundef true) #[[ATTR20:[0-9]+]]
[Attributor][FIX] Liveness handling in the isAssumedDead helpers This fixes a conceptual problem with our AAIsDead usage which conflated call site liveness with call site return value liveness. Without the fix tests would obviously miscompile as we make genericValueTraversal more powerful (in a follow up). The effects on the tests are mixed but mostly marginal. The most prominent one is the lack of `noreturn` for functions. The reason is that we make entire blocks live at the same time (for time reasons). Now that we actually look at the block liveness, which we need to do, the return instructions are live and will survive. As an example, `noreturn_async.ll` has been modified to retain the `noreturn` even with block granularity. We could address this easily but there is little need in practice.
2022-02-01 10:23:18 +08:00
; IS__CGSCC_OPM-NEXT: ret i32 4
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
;
[Attributor] Deduce "nosync" function attribute. Introduce and deduce "nosync" function attribute to indicate that a function does not synchronize with another thread in a way that other thread might free memory. Reviewers: jdoerfert, jfb, nhaehnle, arsenm Subscribers: wdng, hfinkel, nhaenhle, mehdi_amini, steven_wu, dexonsmith, arsenm, uenoku, hiraditya, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D62766 llvm-svn: 365830
2019-07-12 05:37:40 +08:00
call void @llvm.memcpy(i8* %ptr1, i8* %ptr2, i32 8, i1 1)
ret i32 4
}
; TEST 15 - positive, non-volatile intrinsic.
[Attributor] Deduce "no-capture" argument attribute Add the no-capture argument attribute deduction to the Attributor fixpoint framework. The new string attributed "no-capture-maybe-returned" is introduced to allow deduction of no-capture through functions that "capture" an argument but only by "returning" it. It is only used by the Attributor for testing. Differential Revision: https://reviews.llvm.org/D59922 llvm-svn: 370817
2019-09-04 04:37:24 +08:00
; It is odd to add nocapture but a result of the llvm.memset nocapture.
;
[Attributor] Deduce "nosync" function attribute. Introduce and deduce "nosync" function attribute to indicate that a function does not synchronize with another thread in a way that other thread might free memory. Reviewers: jdoerfert, jfb, nhaehnle, arsenm Subscribers: wdng, hfinkel, nhaenhle, mehdi_amini, steven_wu, dexonsmith, arsenm, uenoku, hiraditya, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D62766 llvm-svn: 365830
2019-07-12 05:37:40 +08:00
define i32 @memset_non_volatile(i8* %ptr1, i8 %val) {
[Attributor] Use AAFunctionReachability to determine AANoRecurse We missed out on AANoRecurse in the module pass because we had no call graph. With AAFunctionReachability we can simply ask if the function may reach itself. Differential Revision: https://reviews.llvm.org/D110099
2021-09-21 03:41:52 +08:00
; NOT_CGSCC_OPM: Function Attrs: argmemonly nofree nosync nounwind willreturn writeonly
; NOT_CGSCC_OPM-LABEL: define {{[^@]+}}@memset_non_volatile
; NOT_CGSCC_OPM-SAME: (i8* nocapture nofree writeonly [[PTR1:%.*]], i8 [[VAL:%.*]]) #[[ATTR11:[0-9]+]] {
[Attributor][FIX] Reachability needs to account for readonly callees The oversight caused us to ignore call sites that are effectively dead when we computed reachability (or more precise the call edges of a function). The problem is that loads in the readonly callee might depend on stores prior to the callee. If we do not track the call edge we mistakenly assumed the store before the call cannot reach the load. The problem is nicely visible in: `llvm/test/Transforms/Attributor/ArgumentPromotion/basictest.ll` Caused by D118673. Fixes https://github.com/llvm/llvm-project/issues/53726
2022-02-08 13:37:02 +08:00
; NOT_CGSCC_OPM-NEXT: call void @llvm.memset.p0i8.i32(i8* nocapture nofree writeonly [[PTR1]], i8 [[VAL]], i32 noundef 8, i1 noundef false) #[[ATTR20:[0-9]+]]
[Attributor] Use AAFunctionReachability to determine AANoRecurse We missed out on AANoRecurse in the module pass because we had no call graph. With AAFunctionReachability we can simply ask if the function may reach itself. Differential Revision: https://reviews.llvm.org/D110099
2021-09-21 03:41:52 +08:00
; NOT_CGSCC_OPM-NEXT: ret i32 4
[IR] Make nosync, nofree and willreturn default for intrinsics. D70365 allows us to make attributes default. This is a follow up to actually make nosync, nofree and willreturn default. The approach we chose, for now, is to opt-in to default attributes to avoid introducing problems to target specific intrinsics. Intrinsics with default attributes can be created using `DefaultAttrsIntrinsic` class.
2020-10-15 19:04:36 +08:00
;
[Attributor][FIX] Liveness handling in the isAssumedDead helpers This fixes a conceptual problem with our AAIsDead usage which conflated call site liveness with call site return value liveness. Without the fix tests would obviously miscompile as we make genericValueTraversal more powerful (in a follow up). The effects on the tests are mixed but mostly marginal. The most prominent one is the lack of `noreturn` for functions. The reason is that we make entire blocks live at the same time (for time reasons). Now that we actually look at the block liveness, which we need to do, the return instructions are live and will survive. As an example, `noreturn_async.ll` has been modified to retain the `noreturn` even with block granularity. We could address this easily but there is little need in practice.
2022-02-01 10:23:18 +08:00
; IS__CGSCC_OPM: Function Attrs: argmemonly nofree nosync nounwind willreturn writeonly
; IS__CGSCC_OPM-LABEL: define {{[^@]+}}@memset_non_volatile
[Attributor] Introduce the concept of query AAs D106720 introduced features that did not work properly as we could add new queries after a fixpoint was reached and which could not be answered by the information gathered up to the fixpoint alone. As an alternative to D110078, which forced eager computation where we want to continue to be lazy, this patch fixes the problem. QueryAAs are AAs that allow lazy queries during their lifetime. They are never fixed if they have no outstanding dependences and always run as part of the updates in an iteration. To determine if we are done, all query AAs are asked if they received new queries, if not, we only need to consider updated AAs, as before. If new queries are present we go for another iteration. Differential Revision: https://reviews.llvm.org/D118669
2022-02-01 04:16:43 +08:00
; IS__CGSCC_OPM-SAME: (i8* nocapture nofree writeonly [[PTR1:%.*]], i8 [[VAL:%.*]]) #[[ATTR11:[0-9]+]] {
[Attributor][FIX] Reachability needs to account for readonly callees The oversight caused us to ignore call sites that are effectively dead when we computed reachability (or more precise the call edges of a function). The problem is that loads in the readonly callee might depend on stores prior to the callee. If we do not track the call edge we mistakenly assumed the store before the call cannot reach the load. The problem is nicely visible in: `llvm/test/Transforms/Attributor/ArgumentPromotion/basictest.ll` Caused by D118673. Fixes https://github.com/llvm/llvm-project/issues/53726
2022-02-08 13:37:02 +08:00
; IS__CGSCC_OPM-NEXT: call void @llvm.memset.p0i8.i32(i8* nocapture nofree writeonly [[PTR1]], i8 [[VAL]], i32 noundef 8, i1 noundef false) #[[ATTR21:[0-9]+]]
[Attributor][FIX] Liveness handling in the isAssumedDead helpers This fixes a conceptual problem with our AAIsDead usage which conflated call site liveness with call site return value liveness. Without the fix tests would obviously miscompile as we make genericValueTraversal more powerful (in a follow up). The effects on the tests are mixed but mostly marginal. The most prominent one is the lack of `noreturn` for functions. The reason is that we make entire blocks live at the same time (for time reasons). Now that we actually look at the block liveness, which we need to do, the return instructions are live and will survive. As an example, `noreturn_async.ll` has been modified to retain the `noreturn` even with block granularity. We could address this easily but there is little need in practice.
2022-02-01 10:23:18 +08:00
; IS__CGSCC_OPM-NEXT: ret i32 4
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
;
[Attributor] Deduce "nosync" function attribute. Introduce and deduce "nosync" function attribute to indicate that a function does not synchronize with another thread in a way that other thread might free memory. Reviewers: jdoerfert, jfb, nhaehnle, arsenm Subscribers: wdng, hfinkel, nhaenhle, mehdi_amini, steven_wu, dexonsmith, arsenm, uenoku, hiraditya, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D62766 llvm-svn: 365830
2019-07-12 05:37:40 +08:00
call void @llvm.memset(i8* %ptr1, i8 %val, i32 8, i1 0)
ret i32 4
}
; TEST 16 - negative, inline assembly.
define i32 @inline_asm_test(i32 %x) {
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
; CHECK-LABEL: define {{[^@]+}}@inline_asm_test
[Attributor][NFC] rerun update_test_checks without --scrub-attributes
2020-08-30 01:31:13 +08:00
; CHECK-SAME: (i32 [[X:%.*]]) {
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
; CHECK-NEXT: [[TMP1:%.*]] = call i32 asm "bswap $0", "=r,r"(i32 [[X]])
; CHECK-NEXT: ret i32 4
;
[Attributor] Deduce "nosync" function attribute. Introduce and deduce "nosync" function attribute to indicate that a function does not synchronize with another thread in a way that other thread might free memory. Reviewers: jdoerfert, jfb, nhaehnle, arsenm Subscribers: wdng, hfinkel, nhaenhle, mehdi_amini, steven_wu, dexonsmith, arsenm, uenoku, hiraditya, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D62766 llvm-svn: 365830
2019-07-12 05:37:40 +08:00
call i32 asm "bswap $0", "=r,r"(i32 %x)
ret i32 4
}
declare void @readnone_test() convergent readnone
; TEST 17 - negative. Convergent
define void @convergent_readnone(){
[Attributor] Introduce the concept of query AAs D106720 introduced features that did not work properly as we could add new queries after a fixpoint was reached and which could not be answered by the information gathered up to the fixpoint alone. As an alternative to D110078, which forced eager computation where we want to continue to be lazy, this patch fixes the problem. QueryAAs are AAs that allow lazy queries during their lifetime. They are never fixed if they have no outstanding dependences and always run as part of the updates in an iteration. To determine if we are done, all query AAs are asked if they received new queries, if not, we only need to consider updated AAs, as before. If new queries are present we go for another iteration. Differential Revision: https://reviews.llvm.org/D118669
2022-02-01 04:16:43 +08:00
; CHECK: Function Attrs: readnone
; CHECK-LABEL: define {{[^@]+}}@convergent_readnone
; CHECK-SAME: () #[[ATTR13:[0-9]+]] {
; CHECK-NEXT: call void @readnone_test()
; CHECK-NEXT: ret void
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
;
call void @readnone_test()
ret void
[Attributor] Deduce "nosync" function attribute. Introduce and deduce "nosync" function attribute to indicate that a function does not synchronize with another thread in a way that other thread might free memory. Reviewers: jdoerfert, jfb, nhaehnle, arsenm Subscribers: wdng, hfinkel, nhaenhle, mehdi_amini, steven_wu, dexonsmith, arsenm, uenoku, hiraditya, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D62766 llvm-svn: 365830
2019-07-12 05:37:40 +08:00
}
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
; CHECK: Function Attrs: nounwind
; CHECK-NEXT: declare void @llvm.x86.sse2.clflush(i8*)
[Attributor] Deduce "nosync" function attribute. Introduce and deduce "nosync" function attribute to indicate that a function does not synchronize with another thread in a way that other thread might free memory. Reviewers: jdoerfert, jfb, nhaehnle, arsenm Subscribers: wdng, hfinkel, nhaenhle, mehdi_amini, steven_wu, dexonsmith, arsenm, uenoku, hiraditya, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D62766 llvm-svn: 365830
2019-07-12 05:37:40 +08:00
declare void @llvm.x86.sse2.clflush(i8*)
@a = common global i32 0, align 4
; TEST 18 - negative. Synchronizing intrinsic
define void @i_totally_sync() {
[Attributor] Introduce the concept of query AAs D106720 introduced features that did not work properly as we could add new queries after a fixpoint was reached and which could not be answered by the information gathered up to the fixpoint alone. As an alternative to D110078, which forced eager computation where we want to continue to be lazy, this patch fixes the problem. QueryAAs are AAs that allow lazy queries during their lifetime. They are never fixed if they have no outstanding dependences and always run as part of the updates in an iteration. To determine if we are done, all query AAs are asked if they received new queries, if not, we only need to consider updated AAs, as before. If new queries are present we go for another iteration. Differential Revision: https://reviews.llvm.org/D118669
2022-02-01 04:16:43 +08:00
; CHECK: Function Attrs: nounwind
; CHECK-LABEL: define {{[^@]+}}@i_totally_sync
; CHECK-SAME: () #[[ATTR14:[0-9]+]] {
; CHECK-NEXT: tail call void @llvm.x86.sse2.clflush(i8* noundef nonnull align 4 dereferenceable(4) bitcast (i32* @a to i8*))
; CHECK-NEXT: ret void
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
;
[Attributor] Deduce "nosync" function attribute. Introduce and deduce "nosync" function attribute to indicate that a function does not synchronize with another thread in a way that other thread might free memory. Reviewers: jdoerfert, jfb, nhaehnle, arsenm Subscribers: wdng, hfinkel, nhaenhle, mehdi_amini, steven_wu, dexonsmith, arsenm, uenoku, hiraditya, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D62766 llvm-svn: 365830
2019-07-12 05:37:40 +08:00
tail call void @llvm.x86.sse2.clflush(i8* bitcast (i32* @a to i8*))
ret void
}
declare float @llvm.cos(float %val) readnone
; TEST 19 - positive, readnone & non-convergent intrinsic.
[Attributor][NFC] applying update_test_checks with --check-attributes Summary: All tests are updated, except wrapper.ll since it is not working nicely with newly created functions. Reviewers: jdoerfert, uenoku, baziotis, homerdin Subscribers: arphaman, jfb, kuter, bbn, okura, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D84130
2020-07-20 03:55:43 +08:00
define i32 @cos_test(float %x) {
Reapply "[Intrinsics] Add `nocallback` to the default intrinsic attributes" This reverts commit c5f789050daab25aad6770790987e2b7c0395936 and reapplies 7aea3ea8c3b33c9bb338d5d6c0e4832be1d09ac3 with additional test changes.
2022-03-25 04:08:53 +08:00
; CHECK: Function Attrs: nofree norecurse nosync nounwind readnone willreturn
[Attributor] Introduce the concept of query AAs D106720 introduced features that did not work properly as we could add new queries after a fixpoint was reached and which could not be answered by the information gathered up to the fixpoint alone. As an alternative to D110078, which forced eager computation where we want to continue to be lazy, this patch fixes the problem. QueryAAs are AAs that allow lazy queries during their lifetime. They are never fixed if they have no outstanding dependences and always run as part of the updates in an iteration. To determine if we are done, all query AAs are asked if they received new queries, if not, we only need to consider updated AAs, as before. If new queries are present we go for another iteration. Differential Revision: https://reviews.llvm.org/D118669
2022-02-01 04:16:43 +08:00
; CHECK-LABEL: define {{[^@]+}}@cos_test
; CHECK-SAME: (float [[X:%.*]]) #[[ATTR15:[0-9]+]] {
; CHECK-NEXT: ret i32 4
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
;
[Attributor] Deduce "nosync" function attribute. Introduce and deduce "nosync" function attribute to indicate that a function does not synchronize with another thread in a way that other thread might free memory. Reviewers: jdoerfert, jfb, nhaehnle, arsenm Subscribers: wdng, hfinkel, nhaenhle, mehdi_amini, steven_wu, dexonsmith, arsenm, uenoku, hiraditya, jfb, llvm-commits Differential Revision: https://reviews.llvm.org/D62766 llvm-svn: 365830
2019-07-12 05:37:40 +08:00
call float @llvm.cos(float %x)
ret i32 4
}
[Attributor] Use fine-grained liveness in all helpers We used coarse-grained liveness before, thus we looked if the instruction was executed, but we did not use fine-grained liveness, hence if the instruction was needed or could be deleted even if the surrounding ones are live. This patches introduces this level of liveness checks together with other liveness queries, e.g., for uses. For more control we enforce that all liveness queries go through the Attributor. Test have been adjusted to reflect the changes or augmented to prevent deletion of the parts we want to check. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D73313
2020-01-12 15:09:22 +08:00
define float @cos_test2(float %x) {
Reapply "[Intrinsics] Add `nocallback` to the default intrinsic attributes" This reverts commit c5f789050daab25aad6770790987e2b7c0395936 and reapplies 7aea3ea8c3b33c9bb338d5d6c0e4832be1d09ac3 with additional test changes.
2022-03-25 04:08:53 +08:00
; NOT_CGSCC_OPM: Function Attrs: nofree norecurse nosync nounwind readnone willreturn
[Attributor] Use AAFunctionReachability to determine AANoRecurse We missed out on AANoRecurse in the module pass because we had no call graph. With AAFunctionReachability we can simply ask if the function may reach itself. Differential Revision: https://reviews.llvm.org/D110099
2021-09-21 03:41:52 +08:00
; NOT_CGSCC_OPM-LABEL: define {{[^@]+}}@cos_test2
[Attributor][FIX] Reachability needs to account for readonly callees The oversight caused us to ignore call sites that are effectively dead when we computed reachability (or more precise the call edges of a function). The problem is that loads in the readonly callee might depend on stores prior to the callee. If we do not track the call edge we mistakenly assumed the store before the call cannot reach the load. The problem is nicely visible in: `llvm/test/Transforms/Attributor/ArgumentPromotion/basictest.ll` Caused by D118673. Fixes https://github.com/llvm/llvm-project/issues/53726
2022-02-08 13:37:02 +08:00
; NOT_CGSCC_OPM-SAME: (float [[X:%.*]]) #[[ATTR15]] {
; NOT_CGSCC_OPM-NEXT: [[C:%.*]] = call float @llvm.cos.f32(float [[X]]) #[[ATTR21:[0-9]+]]
[Attributor] Use AAFunctionReachability to determine AANoRecurse We missed out on AANoRecurse in the module pass because we had no call graph. With AAFunctionReachability we can simply ask if the function may reach itself. Differential Revision: https://reviews.llvm.org/D110099
2021-09-21 03:41:52 +08:00
; NOT_CGSCC_OPM-NEXT: ret float [[C]]
[IR] Make nosync, nofree and willreturn default for intrinsics. D70365 allows us to make attributes default. This is a follow up to actually make nosync, nofree and willreturn default. The approach we chose, for now, is to opt-in to default attributes to avoid introducing problems to target specific intrinsics. Intrinsics with default attributes can be created using `DefaultAttrsIntrinsic` class.
2020-10-15 19:04:36 +08:00
;
Reapply "[Intrinsics] Add `nocallback` to the default intrinsic attributes" This reverts commit c5f789050daab25aad6770790987e2b7c0395936 and reapplies 7aea3ea8c3b33c9bb338d5d6c0e4832be1d09ac3 with additional test changes.
2022-03-25 04:08:53 +08:00
; IS__CGSCC_OPM: Function Attrs: nofree norecurse nosync nounwind readnone willreturn
[Attributor][FIX] Liveness handling in the isAssumedDead helpers This fixes a conceptual problem with our AAIsDead usage which conflated call site liveness with call site return value liveness. Without the fix tests would obviously miscompile as we make genericValueTraversal more powerful (in a follow up). The effects on the tests are mixed but mostly marginal. The most prominent one is the lack of `noreturn` for functions. The reason is that we make entire blocks live at the same time (for time reasons). Now that we actually look at the block liveness, which we need to do, the return instructions are live and will survive. As an example, `noreturn_async.ll` has been modified to retain the `noreturn` even with block granularity. We could address this easily but there is little need in practice.
2022-02-01 10:23:18 +08:00
; IS__CGSCC_OPM-LABEL: define {{[^@]+}}@cos_test2
[Attributor][FIX] Reachability needs to account for readonly callees The oversight caused us to ignore call sites that are effectively dead when we computed reachability (or more precise the call edges of a function). The problem is that loads in the readonly callee might depend on stores prior to the callee. If we do not track the call edge we mistakenly assumed the store before the call cannot reach the load. The problem is nicely visible in: `llvm/test/Transforms/Attributor/ArgumentPromotion/basictest.ll` Caused by D118673. Fixes https://github.com/llvm/llvm-project/issues/53726
2022-02-08 13:37:02 +08:00
; IS__CGSCC_OPM-SAME: (float [[X:%.*]]) #[[ATTR15]] {
; IS__CGSCC_OPM-NEXT: [[C:%.*]] = call float @llvm.cos.f32(float [[X]]) #[[ATTR22:[0-9]+]]
[Attributor][FIX] Liveness handling in the isAssumedDead helpers This fixes a conceptual problem with our AAIsDead usage which conflated call site liveness with call site return value liveness. Without the fix tests would obviously miscompile as we make genericValueTraversal more powerful (in a follow up). The effects on the tests are mixed but mostly marginal. The most prominent one is the lack of `noreturn` for functions. The reason is that we make entire blocks live at the same time (for time reasons). Now that we actually look at the block liveness, which we need to do, the return instructions are live and will survive. As an example, `noreturn_async.ll` has been modified to retain the `noreturn` even with block granularity. We could address this easily but there is little need in practice.
2022-02-01 10:23:18 +08:00
; IS__CGSCC_OPM-NEXT: ret float [[C]]
[Attributor] Unify testing (=updates,prefixes,run configurations,...) When the Attributor was created the test update scripts were not well suited to deal with the challenges of IR attribute checking. This partially improved. Since then we also added three additional configurations that need testing; in total we now have the following four: { TUNIT, CGSCC } x { old pass manager (OPM), new pass manager (NPM) } Finally, the number of developers and tests grew rapidly (partially due to the addition of ArgumentPromotion and IPConstantProp tests), which resulted in tests only being run in some configurations, different prefixes being used, and different "styles" of checks being used. Due to the above reasons I believed we needed to take another look at the test update scripts. While we started to use them, via UTC_ARGS: --enable/disable, the other problems remained. To improve the testing situation for *all* configurations, to simplify future updates to the test, and to help identify subtle effects of future changes, we now use the test update scripts for (almost) all Attributor tests. An exhaustive prefix list minimizes the number of check lines and makes it easy to identify and compare configurations. Tests have been adjusted in the process but we tried to keep their intend unchanged. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D76588
2020-04-09 08:04:57 +08:00
;
[Attributor] Use fine-grained liveness in all helpers We used coarse-grained liveness before, thus we looked if the instruction was executed, but we did not use fine-grained liveness, hence if the instruction was needed or could be deleted even if the surrounding ones are live. This patches introduces this level of liveness checks together with other liveness queries, e.g., for uses. For more control we enforce that all liveness queries go through the Attributor. Test have been adjusted to reflect the changes or augmented to prevent deletion of the parts we want to check. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D73313
2020-01-12 15:09:22 +08:00
%c = call float @llvm.cos(float %x)
ret float %c
}
[Attributor][NFC] Update tests after D94741 The update_test_checks script can now check for global symbols and is able to handle them properly when they differ across prefixes, e.g., attribute #0 might be different in different runs. This patch simply updates all the Attributor tests with the new script. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D97906
2021-03-12 13:21:32 +08:00
;.
Extend the `uwtable` attribute with unwind table kind We have the `clang -cc1` command-line option `-funwind-tables=1|2` and the codegen option `VALUE_CODEGENOPT(UnwindTables, 2, 0) ///< Unwind tables (1) or asynchronous unwind tables (2)`. However, this is encoded in LLVM IR by the presence or the absence of the `uwtable` attribute, i.e. we lose the information whether to generate want just some unwind tables or asynchronous unwind tables. Asynchronous unwind tables take more space in the runtime image, I'd estimate something like 80-90% more, as the difference is adding roughly the same number of CFI directives as for prologues, only a bit simpler (e.g. `.cfi_offset reg, off` vs. `.cfi_restore reg`). Or even more, if you consider tail duplication of epilogue blocks. Asynchronous unwind tables could also restrict code generation to having only a finite number of frame pointer adjustments (an example of *not* having a finite number of `SP` adjustments is on AArch64 when untagging the stack (MTE) in some cases the compiler can modify `SP` in a loop). Having the CFI precise up to an instruction generally also means one cannot bundle together CFI instructions once the prologue is done, they need to be interspersed with ordinary instructions, which means extra `DW_CFA_advance_loc` commands, further increasing the unwind tables size. That is to say, async unwind tables impose a non-negligible overhead, yet for the most common use cases (like C++ exceptions), they are not even needed. This patch extends the `uwtable` attribute with an optional value: - `uwtable` (default to `async`) - `uwtable(sync)`, synchronous unwind tables - `uwtable(async)`, asynchronous (instruction precise) unwind tables Reviewed By: MaskRay Differential Revision: https://reviews.llvm.org/D114543
2022-02-14 21:41:34 +08:00
; NOT_CGSCC_OPM: attributes #[[ATTR0]] = { nofree norecurse nosync nounwind optsize readnone ssp willreturn uwtable }
; NOT_CGSCC_OPM: attributes #[[ATTR1]] = { argmemonly nofree norecurse nosync nounwind willreturn uwtable }
; NOT_CGSCC_OPM: attributes #[[ATTR2]] = { argmemonly nofree norecurse nounwind willreturn uwtable }
[Attributor] Use AAFunctionReachability to determine AANoRecurse We missed out on AANoRecurse in the module pass because we had no call graph. With AAFunctionReachability we can simply ask if the function may reach itself. Differential Revision: https://reviews.llvm.org/D110099
2021-09-21 03:41:52 +08:00
; NOT_CGSCC_OPM: attributes #[[ATTR3]] = { noinline nosync nounwind uwtable }
; NOT_CGSCC_OPM: attributes #[[ATTR4]] = { noinline nounwind uwtable }
; NOT_CGSCC_OPM: attributes #[[ATTR5]] = { argmemonly nofree noinline nounwind uwtable }
; NOT_CGSCC_OPM: attributes #[[ATTR6]] = { nofree norecurse nounwind willreturn }
; NOT_CGSCC_OPM: attributes #[[ATTR7]] = { nofree norecurse nounwind }
; NOT_CGSCC_OPM: attributes #[[ATTR8]] = { nofree norecurse nosync nounwind willreturn }
; NOT_CGSCC_OPM: attributes #[[ATTR9]] = { nofree norecurse nosync nounwind }
; NOT_CGSCC_OPM: attributes #[[ATTR10]] = { argmemonly nofree nounwind willreturn }
; NOT_CGSCC_OPM: attributes #[[ATTR11]] = { argmemonly nofree nosync nounwind willreturn writeonly }
; NOT_CGSCC_OPM: attributes #[[ATTR12:[0-9]+]] = { convergent readnone }
; NOT_CGSCC_OPM: attributes #[[ATTR13]] = { readnone }
; NOT_CGSCC_OPM: attributes #[[ATTR14]] = { nounwind }
Reapply "[Intrinsics] Add `nocallback` to the default intrinsic attributes" This reverts commit c5f789050daab25aad6770790987e2b7c0395936 and reapplies 7aea3ea8c3b33c9bb338d5d6c0e4832be1d09ac3 with additional test changes.
2022-03-25 04:08:53 +08:00
; NOT_CGSCC_OPM: attributes #[[ATTR15]] = { nofree norecurse nosync nounwind readnone willreturn }
[Attributor][FIX] Reachability needs to account for readonly callees The oversight caused us to ignore call sites that are effectively dead when we computed reachability (or more precise the call edges of a function). The problem is that loads in the readonly callee might depend on stores prior to the callee. If we do not track the call edge we mistakenly assumed the store before the call cannot reach the load. The problem is nicely visible in: `llvm/test/Transforms/Attributor/ArgumentPromotion/basictest.ll` Caused by D118673. Fixes https://github.com/llvm/llvm-project/issues/53726
2022-02-08 13:37:02 +08:00
; NOT_CGSCC_OPM: attributes #[[ATTR16:[0-9]+]] = { argmemonly nofree nounwind willreturn writeonly }
Reapply "[Intrinsics] Add `nocallback` to the default intrinsic attributes" This reverts commit c5f789050daab25aad6770790987e2b7c0395936 and reapplies 7aea3ea8c3b33c9bb338d5d6c0e4832be1d09ac3 with additional test changes.
2022-03-25 04:08:53 +08:00
; NOT_CGSCC_OPM: attributes #[[ATTR17:[0-9]+]] = { nocallback nofree nosync nounwind readnone speculatable willreturn }
[Attributor][FIX] Reachability needs to account for readonly callees The oversight caused us to ignore call sites that are effectively dead when we computed reachability (or more precise the call edges of a function). The problem is that loads in the readonly callee might depend on stores prior to the callee. If we do not track the call edge we mistakenly assumed the store before the call cannot reach the load. The problem is nicely visible in: `llvm/test/Transforms/Attributor/ArgumentPromotion/basictest.ll` Caused by D118673. Fixes https://github.com/llvm/llvm-project/issues/53726
2022-02-08 13:37:02 +08:00
; NOT_CGSCC_OPM: attributes #[[ATTR18]] = { nofree nounwind }
; NOT_CGSCC_OPM: attributes #[[ATTR19]] = { willreturn }
; NOT_CGSCC_OPM: attributes #[[ATTR20]] = { willreturn writeonly }
; NOT_CGSCC_OPM: attributes #[[ATTR21]] = { readnone willreturn }
[Attributor][NFC] Update tests after D94741 The update_test_checks script can now check for global symbols and is able to handle them properly when they differ across prefixes, e.g., attribute #0 might be different in different runs. This patch simply updates all the Attributor tests with the new script. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D97906
2021-03-12 13:21:32 +08:00
;.
Extend the `uwtable` attribute with unwind table kind We have the `clang -cc1` command-line option `-funwind-tables=1|2` and the codegen option `VALUE_CODEGENOPT(UnwindTables, 2, 0) ///< Unwind tables (1) or asynchronous unwind tables (2)`. However, this is encoded in LLVM IR by the presence or the absence of the `uwtable` attribute, i.e. we lose the information whether to generate want just some unwind tables or asynchronous unwind tables. Asynchronous unwind tables take more space in the runtime image, I'd estimate something like 80-90% more, as the difference is adding roughly the same number of CFI directives as for prologues, only a bit simpler (e.g. `.cfi_offset reg, off` vs. `.cfi_restore reg`). Or even more, if you consider tail duplication of epilogue blocks. Asynchronous unwind tables could also restrict code generation to having only a finite number of frame pointer adjustments (an example of *not* having a finite number of `SP` adjustments is on AArch64 when untagging the stack (MTE) in some cases the compiler can modify `SP` in a loop). Having the CFI precise up to an instruction generally also means one cannot bundle together CFI instructions once the prologue is done, they need to be interspersed with ordinary instructions, which means extra `DW_CFA_advance_loc` commands, further increasing the unwind tables size. That is to say, async unwind tables impose a non-negligible overhead, yet for the most common use cases (like C++ exceptions), they are not even needed. This patch extends the `uwtable` attribute with an optional value: - `uwtable` (default to `async`) - `uwtable(sync)`, synchronous unwind tables - `uwtable(async)`, asynchronous (instruction precise) unwind tables Reviewed By: MaskRay Differential Revision: https://reviews.llvm.org/D114543
2022-02-14 21:41:34 +08:00
; IS__CGSCC_OPM: attributes #[[ATTR0]] = { nofree norecurse nosync nounwind optsize readnone ssp willreturn uwtable }
; IS__CGSCC_OPM: attributes #[[ATTR1]] = { argmemonly nofree norecurse nosync nounwind willreturn uwtable }
; IS__CGSCC_OPM: attributes #[[ATTR2]] = { argmemonly nofree norecurse nounwind willreturn uwtable }
[Attributor][NFC] Update tests after D94741 The update_test_checks script can now check for global symbols and is able to handle them properly when they differ across prefixes, e.g., attribute #0 might be different in different runs. This patch simply updates all the Attributor tests with the new script. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D97906
2021-03-12 13:21:32 +08:00
; IS__CGSCC_OPM: attributes #[[ATTR3]] = { noinline nosync nounwind uwtable }
; IS__CGSCC_OPM: attributes #[[ATTR4]] = { noinline nounwind uwtable }
[Attributor][FIX] Liveness handling in the isAssumedDead helpers This fixes a conceptual problem with our AAIsDead usage which conflated call site liveness with call site return value liveness. Without the fix tests would obviously miscompile as we make genericValueTraversal more powerful (in a follow up). The effects on the tests are mixed but mostly marginal. The most prominent one is the lack of `noreturn` for functions. The reason is that we make entire blocks live at the same time (for time reasons). Now that we actually look at the block liveness, which we need to do, the return instructions are live and will survive. As an example, `noreturn_async.ll` has been modified to retain the `noreturn` even with block granularity. We could address this easily but there is little need in practice.
2022-02-01 10:23:18 +08:00
; IS__CGSCC_OPM: attributes #[[ATTR5]] = { argmemonly nofree noinline nounwind uwtable }
[Attributor] Introduce the concept of query AAs D106720 introduced features that did not work properly as we could add new queries after a fixpoint was reached and which could not be answered by the information gathered up to the fixpoint alone. As an alternative to D110078, which forced eager computation where we want to continue to be lazy, this patch fixes the problem. QueryAAs are AAs that allow lazy queries during their lifetime. They are never fixed if they have no outstanding dependences and always run as part of the updates in an iteration. To determine if we are done, all query AAs are asked if they received new queries, if not, we only need to consider updated AAs, as before. If new queries are present we go for another iteration. Differential Revision: https://reviews.llvm.org/D118669
2022-02-01 04:16:43 +08:00
; IS__CGSCC_OPM: attributes #[[ATTR6]] = { nofree norecurse nounwind willreturn }
; IS__CGSCC_OPM: attributes #[[ATTR7]] = { nofree norecurse nounwind }
; IS__CGSCC_OPM: attributes #[[ATTR8]] = { nofree norecurse nosync nounwind willreturn }
; IS__CGSCC_OPM: attributes #[[ATTR9]] = { nofree norecurse nosync nounwind }
; IS__CGSCC_OPM: attributes #[[ATTR10]] = { argmemonly nofree nounwind willreturn }
; IS__CGSCC_OPM: attributes #[[ATTR11]] = { argmemonly nofree nosync nounwind willreturn writeonly }
; IS__CGSCC_OPM: attributes #[[ATTR12:[0-9]+]] = { convergent readnone }
; IS__CGSCC_OPM: attributes #[[ATTR13]] = { readnone }
; IS__CGSCC_OPM: attributes #[[ATTR14]] = { nounwind }
Reapply "[Intrinsics] Add `nocallback` to the default intrinsic attributes" This reverts commit c5f789050daab25aad6770790987e2b7c0395936 and reapplies 7aea3ea8c3b33c9bb338d5d6c0e4832be1d09ac3 with additional test changes.
2022-03-25 04:08:53 +08:00
; IS__CGSCC_OPM: attributes #[[ATTR15]] = { nofree norecurse nosync nounwind readnone willreturn }
[Attributor][FIX] Reachability needs to account for readonly callees The oversight caused us to ignore call sites that are effectively dead when we computed reachability (or more precise the call edges of a function). The problem is that loads in the readonly callee might depend on stores prior to the callee. If we do not track the call edge we mistakenly assumed the store before the call cannot reach the load. The problem is nicely visible in: `llvm/test/Transforms/Attributor/ArgumentPromotion/basictest.ll` Caused by D118673. Fixes https://github.com/llvm/llvm-project/issues/53726
2022-02-08 13:37:02 +08:00
; IS__CGSCC_OPM: attributes #[[ATTR16:[0-9]+]] = { argmemonly nofree nounwind willreturn writeonly }
Reapply "[Intrinsics] Add `nocallback` to the default intrinsic attributes" This reverts commit c5f789050daab25aad6770790987e2b7c0395936 and reapplies 7aea3ea8c3b33c9bb338d5d6c0e4832be1d09ac3 with additional test changes.
2022-03-25 04:08:53 +08:00
; IS__CGSCC_OPM: attributes #[[ATTR17:[0-9]+]] = { nocallback nofree nosync nounwind readnone speculatable willreturn }
[Attributor][FIX] Reachability needs to account for readonly callees The oversight caused us to ignore call sites that are effectively dead when we computed reachability (or more precise the call edges of a function). The problem is that loads in the readonly callee might depend on stores prior to the callee. If we do not track the call edge we mistakenly assumed the store before the call cannot reach the load. The problem is nicely visible in: `llvm/test/Transforms/Attributor/ArgumentPromotion/basictest.ll` Caused by D118673. Fixes https://github.com/llvm/llvm-project/issues/53726
2022-02-08 13:37:02 +08:00
; IS__CGSCC_OPM: attributes #[[ATTR18]] = { nofree nounwind }
; IS__CGSCC_OPM: attributes #[[ATTR19]] = { nounwind willreturn }
; IS__CGSCC_OPM: attributes #[[ATTR20]] = { willreturn }
; IS__CGSCC_OPM: attributes #[[ATTR21]] = { willreturn writeonly }
; IS__CGSCC_OPM: attributes #[[ATTR22]] = { readnone willreturn }
[Attributor][NFC] Update tests after D94741 The update_test_checks script can now check for global symbols and is able to handle them properly when they differ across prefixes, e.g., attribute #0 might be different in different runs. This patch simply updates all the Attributor tests with the new script. Reviewed By: sstefan1 Differential Revision: https://reviews.llvm.org/D97906
2021-03-12 13:21:32 +08:00
;.