maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
llvm-project/llvm/test/CodeGen/ARM/debug-frame.ll

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

572 lines
20 KiB
LLVM
Raw Normal View History

Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; ARM EHABI integrated test
; This test case checks whether the ARM DWARF stack frame directives
; are properly generated or not.
; We have to check several cases:
Replace "no-frame-pointer-*" function attributes with "frame-pointer" Part of the effort to refactoring frame pointer code generation. We used to use two function attributes "no-frame-pointer-elim" and "no-frame-pointer-elim-non-leaf" to represent three kinds of frame pointer usage: (all) frames use frame pointer, (non-leaf) frames use frame pointer, (none) frame use frame pointer. This CL makes the idea explicit by using only one enum function attribute "frame-pointer" Option "-frame-pointer=" replaces "-disable-fp-elim" for tools such as llc. "no-frame-pointer-elim" and "no-frame-pointer-elim-non-leaf" are still supported for easy migration to "frame-pointer". tests are mostly updated with // replace command line args ‘-disable-fp-elim=false’ with ‘-frame-pointer=none’ grep -iIrnl '\-disable-fp-elim=false' * | xargs sed -i '' -e "s/-disable-fp-elim=false/-frame-pointer=none/g" // replace command line args ‘-disable-fp-elim’ with ‘-frame-pointer=all’ grep -iIrnl '\-disable-fp-elim' * | xargs sed -i '' -e "s/-disable-fp-elim/-frame-pointer=all/g" Patch by Yuanfang Chen (tabloid.adroit)! Differential Revision: https://reviews.llvm.org/D56351 llvm-svn: 351049
2019-01-14 18:55:55 +08:00
; (1) arm with -frame-pointer=all
; (2) arm without -frame-pointer=all
; (3) armv7 with -frame-pointer=all
; (4) armv7 without -frame-pointer=all
; (5) thumb with -frame-pointer=all
; (6) thumb without -frame-pointer=all
; (7) thumbv7 with -frame-pointer=all
; (8) thumbv7 without -frame-pointer=all
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; (9) thumbv7 with -no-integrated-as
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; RUN: llc -mtriple arm-unknown-linux-gnueabi \
Replace "no-frame-pointer-*" function attributes with "frame-pointer" Part of the effort to refactoring frame pointer code generation. We used to use two function attributes "no-frame-pointer-elim" and "no-frame-pointer-elim-non-leaf" to represent three kinds of frame pointer usage: (all) frames use frame pointer, (non-leaf) frames use frame pointer, (none) frame use frame pointer. This CL makes the idea explicit by using only one enum function attribute "frame-pointer" Option "-frame-pointer=" replaces "-disable-fp-elim" for tools such as llc. "no-frame-pointer-elim" and "no-frame-pointer-elim-non-leaf" are still supported for easy migration to "frame-pointer". tests are mostly updated with // replace command line args ‘-disable-fp-elim=false’ with ‘-frame-pointer=none’ grep -iIrnl '\-disable-fp-elim=false' * | xargs sed -i '' -e "s/-disable-fp-elim=false/-frame-pointer=none/g" // replace command line args ‘-disable-fp-elim’ with ‘-frame-pointer=all’ grep -iIrnl '\-disable-fp-elim' * | xargs sed -i '' -e "s/-disable-fp-elim/-frame-pointer=all/g" Patch by Yuanfang Chen (tabloid.adroit)! Differential Revision: https://reviews.llvm.org/D56351 llvm-svn: 351049
2019-01-14 18:55:55 +08:00
; RUN: -frame-pointer=all -filetype=asm -o - %s \
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; RUN: | FileCheck %s --check-prefix=CHECK-FP
; RUN: llc -mtriple arm-unknown-linux-gnueabi \
; RUN: -filetype=asm -o - %s \
; RUN: | FileCheck %s --check-prefix=CHECK-FP-ELIM
; RUN: llc -mtriple armv7-unknown-linux-gnueabi \
Replace "no-frame-pointer-*" function attributes with "frame-pointer" Part of the effort to refactoring frame pointer code generation. We used to use two function attributes "no-frame-pointer-elim" and "no-frame-pointer-elim-non-leaf" to represent three kinds of frame pointer usage: (all) frames use frame pointer, (non-leaf) frames use frame pointer, (none) frame use frame pointer. This CL makes the idea explicit by using only one enum function attribute "frame-pointer" Option "-frame-pointer=" replaces "-disable-fp-elim" for tools such as llc. "no-frame-pointer-elim" and "no-frame-pointer-elim-non-leaf" are still supported for easy migration to "frame-pointer". tests are mostly updated with // replace command line args ‘-disable-fp-elim=false’ with ‘-frame-pointer=none’ grep -iIrnl '\-disable-fp-elim=false' * | xargs sed -i '' -e "s/-disable-fp-elim=false/-frame-pointer=none/g" // replace command line args ‘-disable-fp-elim’ with ‘-frame-pointer=all’ grep -iIrnl '\-disable-fp-elim' * | xargs sed -i '' -e "s/-disable-fp-elim/-frame-pointer=all/g" Patch by Yuanfang Chen (tabloid.adroit)! Differential Revision: https://reviews.llvm.org/D56351 llvm-svn: 351049
2019-01-14 18:55:55 +08:00
; RUN: -frame-pointer=all -filetype=asm -o - %s \
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; RUN: | FileCheck %s --check-prefix=CHECK-V7-FP
; RUN: llc -mtriple armv7-unknown-linux-gnueabi \
; RUN: -filetype=asm -o - %s \
; RUN: | FileCheck %s --check-prefix=CHECK-V7-FP-ELIM
Fix ARMv4T (Thumb1) epilogue generation Summary: Before ARMv5T, Thumb1 code could not pop PC, as described at D14357 and D14986; so we need the special fixup in the epilogue. Reviewers: jroelofs, qcolombet Subscribers: aemerson, llvm-commits, rengolin Differential Revision: http://reviews.llvm.org/D15126 llvm-svn: 255047
2015-12-09 03:59:01 +08:00
; RUN: llc -mtriple thumbv5-unknown-linux-gnueabi \
Replace "no-frame-pointer-*" function attributes with "frame-pointer" Part of the effort to refactoring frame pointer code generation. We used to use two function attributes "no-frame-pointer-elim" and "no-frame-pointer-elim-non-leaf" to represent three kinds of frame pointer usage: (all) frames use frame pointer, (non-leaf) frames use frame pointer, (none) frame use frame pointer. This CL makes the idea explicit by using only one enum function attribute "frame-pointer" Option "-frame-pointer=" replaces "-disable-fp-elim" for tools such as llc. "no-frame-pointer-elim" and "no-frame-pointer-elim-non-leaf" are still supported for easy migration to "frame-pointer". tests are mostly updated with // replace command line args ‘-disable-fp-elim=false’ with ‘-frame-pointer=none’ grep -iIrnl '\-disable-fp-elim=false' * | xargs sed -i '' -e "s/-disable-fp-elim=false/-frame-pointer=none/g" // replace command line args ‘-disable-fp-elim’ with ‘-frame-pointer=all’ grep -iIrnl '\-disable-fp-elim' * | xargs sed -i '' -e "s/-disable-fp-elim/-frame-pointer=all/g" Patch by Yuanfang Chen (tabloid.adroit)! Differential Revision: https://reviews.llvm.org/D56351 llvm-svn: 351049
2019-01-14 18:55:55 +08:00
; RUN: -frame-pointer=all -filetype=asm -o - %s \
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; RUN: | FileCheck %s --check-prefix=CHECK-THUMB-FP
Fix ARMv4T (Thumb1) epilogue generation Summary: Before ARMv5T, Thumb1 code could not pop PC, as described at D14357 and D14986; so we need the special fixup in the epilogue. Reviewers: jroelofs, qcolombet Subscribers: aemerson, llvm-commits, rengolin Differential Revision: http://reviews.llvm.org/D15126 llvm-svn: 255047
2015-12-09 03:59:01 +08:00
; RUN: llc -mtriple thumbv5-unknown-linux-gnueabi \
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; RUN: -filetype=asm -o - %s \
; RUN: | FileCheck %s --check-prefix=CHECK-THUMB-FP-ELIM
; RUN: llc -mtriple thumbv7-unknown-linux-gnueabi \
Replace "no-frame-pointer-*" function attributes with "frame-pointer" Part of the effort to refactoring frame pointer code generation. We used to use two function attributes "no-frame-pointer-elim" and "no-frame-pointer-elim-non-leaf" to represent three kinds of frame pointer usage: (all) frames use frame pointer, (non-leaf) frames use frame pointer, (none) frame use frame pointer. This CL makes the idea explicit by using only one enum function attribute "frame-pointer" Option "-frame-pointer=" replaces "-disable-fp-elim" for tools such as llc. "no-frame-pointer-elim" and "no-frame-pointer-elim-non-leaf" are still supported for easy migration to "frame-pointer". tests are mostly updated with // replace command line args ‘-disable-fp-elim=false’ with ‘-frame-pointer=none’ grep -iIrnl '\-disable-fp-elim=false' * | xargs sed -i '' -e "s/-disable-fp-elim=false/-frame-pointer=none/g" // replace command line args ‘-disable-fp-elim’ with ‘-frame-pointer=all’ grep -iIrnl '\-disable-fp-elim' * | xargs sed -i '' -e "s/-disable-fp-elim/-frame-pointer=all/g" Patch by Yuanfang Chen (tabloid.adroit)! Differential Revision: https://reviews.llvm.org/D56351 llvm-svn: 351049
2019-01-14 18:55:55 +08:00
; RUN: -frame-pointer=all -filetype=asm -o - %s \
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; RUN: | FileCheck %s --check-prefix=CHECK-THUMB-V7-FP
; RUN: llc -mtriple thumbv7-unknown-linux-gnueabi \
; RUN: -filetype=asm -o - %s \
; RUN: | FileCheck %s --check-prefix=CHECK-THUMB-V7-FP-ELIM
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; RUN: llc -mtriple thumbv7-unknown-linux-gnueabi \
Replace "no-frame-pointer-*" function attributes with "frame-pointer" Part of the effort to refactoring frame pointer code generation. We used to use two function attributes "no-frame-pointer-elim" and "no-frame-pointer-elim-non-leaf" to represent three kinds of frame pointer usage: (all) frames use frame pointer, (non-leaf) frames use frame pointer, (none) frame use frame pointer. This CL makes the idea explicit by using only one enum function attribute "frame-pointer" Option "-frame-pointer=" replaces "-disable-fp-elim" for tools such as llc. "no-frame-pointer-elim" and "no-frame-pointer-elim-non-leaf" are still supported for easy migration to "frame-pointer". tests are mostly updated with // replace command line args ‘-disable-fp-elim=false’ with ‘-frame-pointer=none’ grep -iIrnl '\-disable-fp-elim=false' * | xargs sed -i '' -e "s/-disable-fp-elim=false/-frame-pointer=none/g" // replace command line args ‘-disable-fp-elim’ with ‘-frame-pointer=all’ grep -iIrnl '\-disable-fp-elim' * | xargs sed -i '' -e "s/-disable-fp-elim/-frame-pointer=all/g" Patch by Yuanfang Chen (tabloid.adroit)! Differential Revision: https://reviews.llvm.org/D56351 llvm-svn: 351049
2019-01-14 18:55:55 +08:00
; RUN: -frame-pointer=all -no-integrated-as -filetype=asm -o - %s \
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; RUN: | FileCheck %s --check-prefix=CHECK-THUMB-V7-FP-NOIAS
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
;-------------------------------------------------------------------------------
; Test 1
;-------------------------------------------------------------------------------
; This is the LLVM assembly generated from following C++ code:
;
; extern void print(int, int, int, int, int);
; extern void print(double, double, double, double, double);
;
; void test(int a, int b, int c, int d, int e,
; double m, double n, double p, double q, double r) {
; try {
; print(a, b, c, d, e);
; } catch (...) {
; print(m, n, p, q, r);
; }
; }
declare void @_Z5printiiiii(i32, i32, i32, i32, i32)
declare void @_Z5printddddd(double, double, double, double, double)
define void @_Z4testiiiiiddddd(i32 %a, i32 %b, i32 %c, i32 %d, i32 %e,
double %m, double %n, double %p,
Move the personality function from LandingPadInst to Function The personality routine currently lives in the LandingPadInst. This isn't desirable because: - All LandingPadInsts in the same function must have the same personality routine. This means that each LandingPadInst beyond the first has an operand which produces no additional information. - There is ongoing work to introduce EH IR constructs other than LandingPadInst. Moving the personality routine off of any one particular Instruction and onto the parent function seems a lot better than have N different places a personality function can sneak onto an exceptional function. Differential Revision: http://reviews.llvm.org/D10429 llvm-svn: 239940
2015-06-18 04:52:32 +08:00
double %q, double %r) personality i8* bitcast (i32 (...)* @__gxx_personality_v0 to i8*) {
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
entry:
invoke void @_Z5printiiiii(i32 %a, i32 %b, i32 %c, i32 %d, i32 %e)
to label %try.cont unwind label %lpad
lpad:
%0 = landingpad { i8*, i32 }
catch i8* null
%1 = extractvalue { i8*, i32 } %0, 0
%2 = tail call i8* @__cxa_begin_catch(i8* %1)
invoke void @_Z5printddddd(double %m, double %n, double %p,
double %q, double %r)
to label %invoke.cont2 unwind label %lpad1
invoke.cont2:
tail call void @__cxa_end_catch()
br label %try.cont
try.cont:
ret void
lpad1:
%3 = landingpad { i8*, i32 }
cleanup
invoke void @__cxa_end_catch()
to label %eh.resume unwind label %terminate.lpad
eh.resume:
resume { i8*, i32 } %3
terminate.lpad:
%4 = landingpad { i8*, i32 }
catch i8* null
%5 = extractvalue { i8*, i32 } %4, 0
tail call void @__clang_call_terminate(i8* %5)
unreachable
}
declare void @__clang_call_terminate(i8*)
declare i32 @__gxx_personality_v0(...)
declare i8* @__cxa_begin_catch(i8*)
declare void @__cxa_end_catch()
declare void @_ZSt9terminatev()
!llvm.dbg.cu = !{!0}
!llvm.module.flags = !{!10, !11}
!llvm.ident = !{!12}
[PR27284] Reverse the ownership between DICompileUnit and DISubprogram. Currently each Function points to a DISubprogram and DISubprogram has a scope field. For member functions the scope is a DICompositeType. DIScopes point to the DICompileUnit to facilitate type uniquing. Distinct DISubprograms (with isDefinition: true) are not part of the type hierarchy and cannot be uniqued. This change removes the subprograms list from DICompileUnit and instead adds a pointer to the owning compile unit to distinct DISubprograms. This would make it easy for ThinLTO to strip unneeded DISubprograms and their transitively referenced debug info. Motivation ---------- Materializing DISubprograms is currently the most expensive operation when doing a ThinLTO build of clang. We want the DISubprogram to be stored in a separate Bitcode block (or the same block as the function body) so we can avoid having to expensively deserialize all DISubprograms together with the global metadata. If a function has been inlined into another subprogram we need to store a reference the block containing the inlined subprogram. Attached to https://llvm.org/bugs/show_bug.cgi?id=27284 is a python script that updates LLVM IR testcases to the new format. http://reviews.llvm.org/D19034 <rdar://problem/25256815> llvm-svn: 266446
2016-04-15 23:57:41 +08:00
!0 = distinct !DICompileUnit(language: DW_LANG_C_plus_plus, producer: "clang version 3.5 ", isOptimized: false, emissionKind: FullDebug, file: !1, enums: !2, retainedTypes: !2, globals: !2, imports: !2)
IR: Give 'DI' prefix to debug info metadata Finish off PR23080 by renaming the debug info IR constructs from `MD*` to `DI*`. The last of the `DIDescriptor` classes were deleted in r235356, and the last of the related typedefs removed in r235413, so this has all baked for about a week. Note: If you have out-of-tree code (like a frontend), I recommend that you get everything compiling and tests passing with the *previous* commit before updating to this one. It'll be easier to keep track of what code is using the `DIDescriptor` hierarchy and what you've already updated, and I think you're extremely unlikely to insert bugs. YMMV of course. Back to *this* commit: I did this using the rename-md-di-nodes.sh upgrade script I've attached to PR23080 (both code and testcases) and filtered through clang-format-diff.py. I edited the tests for test/Assembler/invalid-generic-debug-node-*.ll by hand since the columns were off-by-three. It should work on your out-of-tree testcases (and code, if you've followed the advice in the previous paragraph). Some of the tests are in badly named files now (e.g., test/Assembler/invalid-mdcompositetype-missing-tag.ll should be 'dicompositetype'); I'll come back and move the files in a follow-up commit. llvm-svn: 236120
2015-04-30 00:38:44 +08:00
!1 = !DIFile(filename: "exp.cpp", directory: "/tmp")
IR: Make metadata typeless in assembly Now that `Metadata` is typeless, reflect that in the assembly. These are the matching assembly changes for the metadata/value split in r223802. - Only use the `metadata` type when referencing metadata from a call intrinsic -- i.e., only when it's used as a `Value`. - Stop pretending that `ValueAsMetadata` is wrapped in an `MDNode` when referencing it from call intrinsics. So, assembly like this: define @foo(i32 %v) { call void @llvm.foo(metadata !{i32 %v}, metadata !0) call void @llvm.foo(metadata !{i32 7}, metadata !0) call void @llvm.foo(metadata !1, metadata !0) call void @llvm.foo(metadata !3, metadata !0) call void @llvm.foo(metadata !{metadata !3}, metadata !0) ret void, !bar !2 } !0 = metadata !{metadata !2} !1 = metadata !{i32* @global} !2 = metadata !{metadata !3} !3 = metadata !{} turns into this: define @foo(i32 %v) { call void @llvm.foo(metadata i32 %v, metadata !0) call void @llvm.foo(metadata i32 7, metadata !0) call void @llvm.foo(metadata i32* @global, metadata !0) call void @llvm.foo(metadata !3, metadata !0) call void @llvm.foo(metadata !{!3}, metadata !0) ret void, !bar !2 } !0 = !{!2} !1 = !{i32* @global} !2 = !{!3} !3 = !{} I wrote an upgrade script that handled almost all of the tests in llvm and many of the tests in cfe (even handling many `CHECK` lines). I've attached it (or will attach it in a moment if you're speedy) to PR21532 to help everyone update their out-of-tree testcases. This is part of PR21532. llvm-svn: 224257
2014-12-16 03:07:53 +08:00
!2 = !{}
[DebugInfo] Add DILabel metadata and intrinsic llvm.dbg.label. In order to set breakpoints on labels and list source code around labels, we need collect debug information for labels, i.e., label name, the function label belong, line number in the file, and the address label located. In order to keep these information in LLVM IR and to allow backend to generate debug information correctly. We create a new kind of metadata for labels, DILabel. The format of DILabel is !DILabel(scope: !1, name: "foo", file: !2, line: 3) We hope to keep debug information as much as possible even the code is optimized. So, we create a new kind of intrinsic for label metadata to avoid the metadata is eliminated with basic block. The intrinsic will keep existing if we keep it from optimized out. The format of the intrinsic is llvm.dbg.label(metadata !1) It has only one argument, that is the DILabel metadata. The intrinsic will follow the label immediately. Backend could get the label metadata through the intrinsic's parameter. We also create DIBuilder API for labels to be used by Frontend. Frontend could use createLabel() to allocate DILabel objects, and use insertLabel() to insert llvm.dbg.label intrinsic in LLVM IR. Differential Revision: https://reviews.llvm.org/D45024 Patch by Hsiangkai Wang. llvm-svn: 331841
2018-05-09 10:40:45 +08:00
!4 = distinct !DISubprogram(name: "test", linkageName: "_Z4testiiiiiddddd", line: 4, isLocal: false, isDefinition: true, virtualIndex: 6, flags: DIFlagPrototyped, isOptimized: false, unit: !0, scopeLine: 5, file: !1, scope: !5, type: !6, retainedNodes: !2)
IR: Give 'DI' prefix to debug info metadata Finish off PR23080 by renaming the debug info IR constructs from `MD*` to `DI*`. The last of the `DIDescriptor` classes were deleted in r235356, and the last of the related typedefs removed in r235413, so this has all baked for about a week. Note: If you have out-of-tree code (like a frontend), I recommend that you get everything compiling and tests passing with the *previous* commit before updating to this one. It'll be easier to keep track of what code is using the `DIDescriptor` hierarchy and what you've already updated, and I think you're extremely unlikely to insert bugs. YMMV of course. Back to *this* commit: I did this using the rename-md-di-nodes.sh upgrade script I've attached to PR23080 (both code and testcases) and filtered through clang-format-diff.py. I edited the tests for test/Assembler/invalid-generic-debug-node-*.ll by hand since the columns were off-by-three. It should work on your out-of-tree testcases (and code, if you've followed the advice in the previous paragraph). Some of the tests are in badly named files now (e.g., test/Assembler/invalid-mdcompositetype-missing-tag.ll should be 'dicompositetype'); I'll come back and move the files in a follow-up commit. llvm-svn: 236120
2015-04-30 00:38:44 +08:00
!5 = !DIFile(filename: "exp.cpp", directory: "/tmp")
!6 = !DISubroutineType(types: !7)
IR: Make metadata typeless in assembly Now that `Metadata` is typeless, reflect that in the assembly. These are the matching assembly changes for the metadata/value split in r223802. - Only use the `metadata` type when referencing metadata from a call intrinsic -- i.e., only when it's used as a `Value`. - Stop pretending that `ValueAsMetadata` is wrapped in an `MDNode` when referencing it from call intrinsics. So, assembly like this: define @foo(i32 %v) { call void @llvm.foo(metadata !{i32 %v}, metadata !0) call void @llvm.foo(metadata !{i32 7}, metadata !0) call void @llvm.foo(metadata !1, metadata !0) call void @llvm.foo(metadata !3, metadata !0) call void @llvm.foo(metadata !{metadata !3}, metadata !0) ret void, !bar !2 } !0 = metadata !{metadata !2} !1 = metadata !{i32* @global} !2 = metadata !{metadata !3} !3 = metadata !{} turns into this: define @foo(i32 %v) { call void @llvm.foo(metadata i32 %v, metadata !0) call void @llvm.foo(metadata i32 7, metadata !0) call void @llvm.foo(metadata i32* @global, metadata !0) call void @llvm.foo(metadata !3, metadata !0) call void @llvm.foo(metadata !{!3}, metadata !0) ret void, !bar !2 } !0 = !{!2} !1 = !{i32* @global} !2 = !{!3} !3 = !{} I wrote an upgrade script that handled almost all of the tests in llvm and many of the tests in cfe (even handling many `CHECK` lines). I've attached it (or will attach it in a moment if you're speedy) to PR21532 to help everyone update their out-of-tree testcases. This is part of PR21532. llvm-svn: 224257
2014-12-16 03:07:53 +08:00
!7 = !{null, !8, !8, !8, !8, !8, !9, !9, !9, !9, !9}
IR: Give 'DI' prefix to debug info metadata Finish off PR23080 by renaming the debug info IR constructs from `MD*` to `DI*`. The last of the `DIDescriptor` classes were deleted in r235356, and the last of the related typedefs removed in r235413, so this has all baked for about a week. Note: If you have out-of-tree code (like a frontend), I recommend that you get everything compiling and tests passing with the *previous* commit before updating to this one. It'll be easier to keep track of what code is using the `DIDescriptor` hierarchy and what you've already updated, and I think you're extremely unlikely to insert bugs. YMMV of course. Back to *this* commit: I did this using the rename-md-di-nodes.sh upgrade script I've attached to PR23080 (both code and testcases) and filtered through clang-format-diff.py. I edited the tests for test/Assembler/invalid-generic-debug-node-*.ll by hand since the columns were off-by-three. It should work on your out-of-tree testcases (and code, if you've followed the advice in the previous paragraph). Some of the tests are in badly named files now (e.g., test/Assembler/invalid-mdcompositetype-missing-tag.ll should be 'dicompositetype'); I'll come back and move the files in a follow-up commit. llvm-svn: 236120
2015-04-30 00:38:44 +08:00
!8 = !DIBasicType(tag: DW_TAG_base_type, name: "int", size: 32, align: 32, encoding: DW_ATE_signed)
!9 = !DIBasicType(tag: DW_TAG_base_type, name: "double", size: 64, align: 64, encoding: DW_ATE_float)
IR: Make metadata typeless in assembly Now that `Metadata` is typeless, reflect that in the assembly. These are the matching assembly changes for the metadata/value split in r223802. - Only use the `metadata` type when referencing metadata from a call intrinsic -- i.e., only when it's used as a `Value`. - Stop pretending that `ValueAsMetadata` is wrapped in an `MDNode` when referencing it from call intrinsics. So, assembly like this: define @foo(i32 %v) { call void @llvm.foo(metadata !{i32 %v}, metadata !0) call void @llvm.foo(metadata !{i32 7}, metadata !0) call void @llvm.foo(metadata !1, metadata !0) call void @llvm.foo(metadata !3, metadata !0) call void @llvm.foo(metadata !{metadata !3}, metadata !0) ret void, !bar !2 } !0 = metadata !{metadata !2} !1 = metadata !{i32* @global} !2 = metadata !{metadata !3} !3 = metadata !{} turns into this: define @foo(i32 %v) { call void @llvm.foo(metadata i32 %v, metadata !0) call void @llvm.foo(metadata i32 7, metadata !0) call void @llvm.foo(metadata i32* @global, metadata !0) call void @llvm.foo(metadata !3, metadata !0) call void @llvm.foo(metadata !{!3}, metadata !0) ret void, !bar !2 } !0 = !{!2} !1 = !{i32* @global} !2 = !{!3} !3 = !{} I wrote an upgrade script that handled almost all of the tests in llvm and many of the tests in cfe (even handling many `CHECK` lines). I've attached it (or will attach it in a moment if you're speedy) to PR21532 to help everyone update their out-of-tree testcases. This is part of PR21532. llvm-svn: 224257
2014-12-16 03:07:53 +08:00
!10 = !{i32 2, !"Dwarf Version", i32 4}
DebugInfo: Move new hierarchy into place Move the specialized metadata nodes for the new debug info hierarchy into place, finishing off PR22464. I've done bootstraps (and all that) and I'm confident this commit is NFC as far as DWARF output is concerned. Let me know if I'm wrong :). The code changes are fairly mechanical: - Bumped the "Debug Info Version". - `DIBuilder` now creates the appropriate subclass of `MDNode`. - Subclasses of DIDescriptor now expect to hold their "MD" counterparts (e.g., `DIBasicType` expects `MDBasicType`). - Deleted a ton of dead code in `AsmWriter.cpp` and `DebugInfo.cpp` for printing comments. - Big update to LangRef to describe the nodes in the new hierarchy. Feel free to make it better. Testcase changes are enormous. There's an accompanying clang commit on its way. If you have out-of-tree debug info testcases, I just broke your build. - `upgrade-specialized-nodes.sh` is attached to PR22564. I used it to update all the IR testcases. - Unfortunately I failed to find way to script the updates to CHECK lines, so I updated all of these by hand. This was fairly painful, since the old CHECKs are difficult to reason about. That's one of the benefits of the new hierarchy. This work isn't quite finished, BTW. The `DIDescriptor` subclasses are almost empty wrappers, but not quite: they still have loose casting checks (see the `RETURN_FROM_RAW()` macro). Once they're completely gutted, I'll rename the "MD" classes to "DI" and kill the wrappers. I also expect to make a few schema changes now that it's easier to reason about everything. llvm-svn: 231082
2015-03-04 01:24:31 +08:00
!11 = !{i32 1, !"Debug Info Version", i32 3}
IR: Make metadata typeless in assembly Now that `Metadata` is typeless, reflect that in the assembly. These are the matching assembly changes for the metadata/value split in r223802. - Only use the `metadata` type when referencing metadata from a call intrinsic -- i.e., only when it's used as a `Value`. - Stop pretending that `ValueAsMetadata` is wrapped in an `MDNode` when referencing it from call intrinsics. So, assembly like this: define @foo(i32 %v) { call void @llvm.foo(metadata !{i32 %v}, metadata !0) call void @llvm.foo(metadata !{i32 7}, metadata !0) call void @llvm.foo(metadata !1, metadata !0) call void @llvm.foo(metadata !3, metadata !0) call void @llvm.foo(metadata !{metadata !3}, metadata !0) ret void, !bar !2 } !0 = metadata !{metadata !2} !1 = metadata !{i32* @global} !2 = metadata !{metadata !3} !3 = metadata !{} turns into this: define @foo(i32 %v) { call void @llvm.foo(metadata i32 %v, metadata !0) call void @llvm.foo(metadata i32 7, metadata !0) call void @llvm.foo(metadata i32* @global, metadata !0) call void @llvm.foo(metadata !3, metadata !0) call void @llvm.foo(metadata !{!3}, metadata !0) ret void, !bar !2 } !0 = !{!2} !1 = !{i32* @global} !2 = !{!3} !3 = !{} I wrote an upgrade script that handled almost all of the tests in llvm and many of the tests in cfe (even handling many `CHECK` lines). I've attached it (or will attach it in a moment if you're speedy) to PR21532 to help everyone update their out-of-tree testcases. This is part of PR21532. llvm-svn: 224257
2014-12-16 03:07:53 +08:00
!12 = !{!"clang version 3.5 "}
DI: Remove DW_TAG_arg_variable and DW_TAG_auto_variable Remove the fake `DW_TAG_auto_variable` and `DW_TAG_arg_variable` tags, using `DW_TAG_variable` in their place Stop exposing the `tag:` field at all in the assembly format for `DILocalVariable`. Most of the testcase updates were generated by the following sed script: find test/ -name "*.ll" -o -name "*.mir" | xargs grep -l 'DILocalVariable' | xargs sed -i '' \ -e 's/tag: DW_TAG_arg_variable, //' \ -e 's/tag: DW_TAG_auto_variable, //' There were only a handful of tests in `test/Assembly` that I needed to update by hand. (Note: a follow-up could change `DILocalVariable::DILocalVariable()` to set the tag to `DW_TAG_formal_parameter` instead of `DW_TAG_variable` (as appropriate), instead of having that logic magically in the backend in `DbgVariable`. I've added a FIXME to that effect.) llvm-svn: 243774
2015-08-01 02:58:39 +08:00
!13 = !DILocalVariable(name: "a", line: 4, arg: 1, scope: !4, file: !5, type: !8)
IR: Give 'DI' prefix to debug info metadata Finish off PR23080 by renaming the debug info IR constructs from `MD*` to `DI*`. The last of the `DIDescriptor` classes were deleted in r235356, and the last of the related typedefs removed in r235413, so this has all baked for about a week. Note: If you have out-of-tree code (like a frontend), I recommend that you get everything compiling and tests passing with the *previous* commit before updating to this one. It'll be easier to keep track of what code is using the `DIDescriptor` hierarchy and what you've already updated, and I think you're extremely unlikely to insert bugs. YMMV of course. Back to *this* commit: I did this using the rename-md-di-nodes.sh upgrade script I've attached to PR23080 (both code and testcases) and filtered through clang-format-diff.py. I edited the tests for test/Assembler/invalid-generic-debug-node-*.ll by hand since the columns were off-by-three. It should work on your out-of-tree testcases (and code, if you've followed the advice in the previous paragraph). Some of the tests are in badly named files now (e.g., test/Assembler/invalid-mdcompositetype-missing-tag.ll should be 'dicompositetype'); I'll come back and move the files in a follow-up commit. llvm-svn: 236120
2015-04-30 00:38:44 +08:00
!14 = !DILocation(line: 4, scope: !4)
DI: Remove DW_TAG_arg_variable and DW_TAG_auto_variable Remove the fake `DW_TAG_auto_variable` and `DW_TAG_arg_variable` tags, using `DW_TAG_variable` in their place Stop exposing the `tag:` field at all in the assembly format for `DILocalVariable`. Most of the testcase updates were generated by the following sed script: find test/ -name "*.ll" -o -name "*.mir" | xargs grep -l 'DILocalVariable' | xargs sed -i '' \ -e 's/tag: DW_TAG_arg_variable, //' \ -e 's/tag: DW_TAG_auto_variable, //' There were only a handful of tests in `test/Assembly` that I needed to update by hand. (Note: a follow-up could change `DILocalVariable::DILocalVariable()` to set the tag to `DW_TAG_formal_parameter` instead of `DW_TAG_variable` (as appropriate), instead of having that logic magically in the backend in `DbgVariable`. I've added a FIXME to that effect.) llvm-svn: 243774
2015-08-01 02:58:39 +08:00
!15 = !DILocalVariable(name: "b", line: 4, arg: 2, scope: !4, file: !5, type: !8)
!16 = !DILocalVariable(name: "c", line: 4, arg: 3, scope: !4, file: !5, type: !8)
!17 = !DILocalVariable(name: "d", line: 4, arg: 4, scope: !4, file: !5, type: !8)
!18 = !DILocalVariable(name: "e", line: 4, arg: 5, scope: !4, file: !5, type: !8)
!19 = !DILocalVariable(name: "m", line: 5, arg: 6, scope: !4, file: !5, type: !9)
IR: Give 'DI' prefix to debug info metadata Finish off PR23080 by renaming the debug info IR constructs from `MD*` to `DI*`. The last of the `DIDescriptor` classes were deleted in r235356, and the last of the related typedefs removed in r235413, so this has all baked for about a week. Note: If you have out-of-tree code (like a frontend), I recommend that you get everything compiling and tests passing with the *previous* commit before updating to this one. It'll be easier to keep track of what code is using the `DIDescriptor` hierarchy and what you've already updated, and I think you're extremely unlikely to insert bugs. YMMV of course. Back to *this* commit: I did this using the rename-md-di-nodes.sh upgrade script I've attached to PR23080 (both code and testcases) and filtered through clang-format-diff.py. I edited the tests for test/Assembler/invalid-generic-debug-node-*.ll by hand since the columns were off-by-three. It should work on your out-of-tree testcases (and code, if you've followed the advice in the previous paragraph). Some of the tests are in badly named files now (e.g., test/Assembler/invalid-mdcompositetype-missing-tag.ll should be 'dicompositetype'); I'll come back and move the files in a follow-up commit. llvm-svn: 236120
2015-04-30 00:38:44 +08:00
!20 = !DILocation(line: 5, scope: !4)
DI: Remove DW_TAG_arg_variable and DW_TAG_auto_variable Remove the fake `DW_TAG_auto_variable` and `DW_TAG_arg_variable` tags, using `DW_TAG_variable` in their place Stop exposing the `tag:` field at all in the assembly format for `DILocalVariable`. Most of the testcase updates were generated by the following sed script: find test/ -name "*.ll" -o -name "*.mir" | xargs grep -l 'DILocalVariable' | xargs sed -i '' \ -e 's/tag: DW_TAG_arg_variable, //' \ -e 's/tag: DW_TAG_auto_variable, //' There were only a handful of tests in `test/Assembly` that I needed to update by hand. (Note: a follow-up could change `DILocalVariable::DILocalVariable()` to set the tag to `DW_TAG_formal_parameter` instead of `DW_TAG_variable` (as appropriate), instead of having that logic magically in the backend in `DbgVariable`. I've added a FIXME to that effect.) llvm-svn: 243774
2015-08-01 02:58:39 +08:00
!21 = !DILocalVariable(name: "n", line: 5, arg: 7, scope: !4, file: !5, type: !9)
!22 = !DILocalVariable(name: "p", line: 5, arg: 8, scope: !4, file: !5, type: !9)
!23 = !DILocalVariable(name: "q", line: 5, arg: 9, scope: !4, file: !5, type: !9)
!24 = !DILocalVariable(name: "r", line: 5, arg: 10, scope: !4, file: !5, type: !9)
IR: Give 'DI' prefix to debug info metadata Finish off PR23080 by renaming the debug info IR constructs from `MD*` to `DI*`. The last of the `DIDescriptor` classes were deleted in r235356, and the last of the related typedefs removed in r235413, so this has all baked for about a week. Note: If you have out-of-tree code (like a frontend), I recommend that you get everything compiling and tests passing with the *previous* commit before updating to this one. It'll be easier to keep track of what code is using the `DIDescriptor` hierarchy and what you've already updated, and I think you're extremely unlikely to insert bugs. YMMV of course. Back to *this* commit: I did this using the rename-md-di-nodes.sh upgrade script I've attached to PR23080 (both code and testcases) and filtered through clang-format-diff.py. I edited the tests for test/Assembler/invalid-generic-debug-node-*.ll by hand since the columns were off-by-three. It should work on your out-of-tree testcases (and code, if you've followed the advice in the previous paragraph). Some of the tests are in badly named files now (e.g., test/Assembler/invalid-mdcompositetype-missing-tag.ll should be 'dicompositetype'); I'll come back and move the files in a follow-up commit. llvm-svn: 236120
2015-04-30 00:38:44 +08:00
!25 = !DILocation(line: 7, scope: !26)
!26 = distinct !DILexicalBlock(line: 6, column: 0, file: !1, scope: !4)
!27 = !DILocation(line: 8, scope: !26)
!28 = !DILocation(line: 11, scope: !26)
!29 = !DILocation(line: 9, scope: !30)
!30 = distinct !DILexicalBlock(line: 8, column: 0, file: !1, scope: !4)
!31 = !DILocation(line: 10, scope: !30)
!32 = !DILocation(line: 10, scope: !4)
!33 = !DILocation(line: 11, scope: !4)
!34 = !DILocation(line: 11, scope: !30)
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-FP-LABEL: _Z4testiiiiiddddd:
; CHECK-FP: .cfi_startproc
; CHECK-FP: push {r4, r5, r6, r7, r8, r9, r10, r11, lr}
; CHECK-FP: .cfi_def_cfa_offset 36
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-FP: .cfi_offset lr, -4
; CHECK-FP: .cfi_offset r11, -8
; CHECK-FP: .cfi_offset r10, -12
; CHECK-FP: .cfi_offset r9, -16
; CHECK-FP: .cfi_offset r8, -20
; CHECK-FP: .cfi_offset r7, -24
; CHECK-FP: .cfi_offset r6, -28
; CHECK-FP: .cfi_offset r5, -32
; CHECK-FP: .cfi_offset r4, -36
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-FP: add r11, sp, #28
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-FP: .cfi_def_cfa r11, 8
ScheduleDAGInstrs: In functions with tail calls PseudoSourceValues are not non-aliasing distinct objects The code that builds the dependence graph assumes that two PseudoSourceValues don't alias. In a tail calling function two FixedStackObjects might refer to the same location. Worse 'immutable' fixed stack objects like function arguments are not immutable and will be clobbered. Change this so that a load from a FixedStackObject is not invariant in a tail calling function and don't return a PseudoSourceValue for an instruction in tail calling functions when building the dependence graph so that we handle function arguments conservatively. Fix for PR23459. rdar://20740035 llvm-svn: 236916
2015-05-09 07:52:00 +08:00
; CHECK-FP: sub sp, sp, #44
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-FP: .cfi_endproc
; CHECK-FP-ELIM-LABEL: _Z4testiiiiiddddd:
; CHECK-FP-ELIM: .cfi_startproc
; CHECK-FP-ELIM: push {r4, r5, r6, r7, r8, r9, r10, r11, lr}
; CHECK-FP-ELIM: .cfi_def_cfa_offset 36
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-FP-ELIM: .cfi_offset lr, -4
; CHECK-FP-ELIM: .cfi_offset r11, -8
; CHECK-FP-ELIM: .cfi_offset r10, -12
; CHECK-FP-ELIM: .cfi_offset r9, -16
; CHECK-FP-ELIM: .cfi_offset r8, -20
; CHECK-FP-ELIM: .cfi_offset r7, -24
; CHECK-FP-ELIM: .cfi_offset r6, -28
; CHECK-FP-ELIM: .cfi_offset r5, -32
; CHECK-FP-ELIM: .cfi_offset r4, -36
ScheduleDAGInstrs: In functions with tail calls PseudoSourceValues are not non-aliasing distinct objects The code that builds the dependence graph assumes that two PseudoSourceValues don't alias. In a tail calling function two FixedStackObjects might refer to the same location. Worse 'immutable' fixed stack objects like function arguments are not immutable and will be clobbered. Change this so that a load from a FixedStackObject is not invariant in a tail calling function and don't return a PseudoSourceValue for an instruction in tail calling functions when building the dependence graph so that we handle function arguments conservatively. Fix for PR23459. rdar://20740035 llvm-svn: 236916
2015-05-09 07:52:00 +08:00
; CHECK-FP-ELIM: sub sp, sp, #36
; CHECK-FP-ELIM: .cfi_def_cfa_offset 72
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-FP-ELIM: .cfi_endproc
; CHECK-V7-FP-LABEL: _Z4testiiiiiddddd:
; CHECK-V7-FP: .cfi_startproc
If a function needs a frame pointer, but r11 (aka fp) has not been used, remove it from the list of unspilled registers. Otherwise the following attempt to keep the stack aligned by picking an extra GPR register to spill will not work as it picks up r11. llvm-svn: 208129
2014-05-07 04:43:01 +08:00
; CHECK-V7-FP: push {r4, r10, r11, lr}
; CHECK-V7-FP: .cfi_def_cfa_offset 16
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-V7-FP: .cfi_offset lr, -4
; CHECK-V7-FP: .cfi_offset r11, -8
If a function needs a frame pointer, but r11 (aka fp) has not been used, remove it from the list of unspilled registers. Otherwise the following attempt to keep the stack aligned by picking an extra GPR register to spill will not work as it picks up r11. llvm-svn: 208129
2014-05-07 04:43:01 +08:00
; CHECK-V7-FP: .cfi_offset r10, -12
; CHECK-V7-FP: .cfi_offset r4, -16
; CHECK-V7-FP: add r11, sp, #8
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-V7-FP: .cfi_def_cfa r11, 8
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-V7-FP: vpush {d8, d9, d10, d11, d12}
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-V7-FP: .cfi_offset d12, -24
; CHECK-V7-FP: .cfi_offset d11, -32
; CHECK-V7-FP: .cfi_offset d10, -40
; CHECK-V7-FP: .cfi_offset d9, -48
; CHECK-V7-FP: .cfi_offset d8, -56
If a function needs a frame pointer, but r11 (aka fp) has not been used, remove it from the list of unspilled registers. Otherwise the following attempt to keep the stack aligned by picking an extra GPR register to spill will not work as it picks up r11. llvm-svn: 208129
2014-05-07 04:43:01 +08:00
; CHECK-V7-FP: sub sp, sp, #24
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-V7-FP: .cfi_endproc
; CHECK-V7-FP-ELIM-LABEL: _Z4testiiiiiddddd:
; CHECK-V7-FP-ELIM: .cfi_startproc
; CHECK-V7-FP-ELIM: push {r4, lr}
; CHECK-V7-FP-ELIM: .cfi_def_cfa_offset 8
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-V7-FP-ELIM: .cfi_offset lr, -4
; CHECK-V7-FP-ELIM: .cfi_offset r4, -8
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-V7-FP-ELIM: vpush {d8, d9, d10, d11, d12}
; CHECK-V7-FP-ELIM: .cfi_def_cfa_offset 48
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-V7-FP-ELIM: .cfi_offset d12, -16
; CHECK-V7-FP-ELIM: .cfi_offset d11, -24
; CHECK-V7-FP-ELIM: .cfi_offset d10, -32
; CHECK-V7-FP-ELIM: .cfi_offset d9, -40
; CHECK-V7-FP-ELIM: .cfi_offset d8, -48
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-V7-FP-ELIM: sub sp, sp, #24
; CHECK-V7-FP-ELIM: .cfi_def_cfa_offset 72
; CHECK-V7-FP-ELIM: .cfi_endproc
; CHECK-THUMB-FP-LABEL: _Z4testiiiiiddddd:
; CHECK-THUMB-FP: .cfi_startproc
; CHECK-THUMB-FP: push {r4, r5, r6, r7, lr}
; CHECK-THUMB-FP: .cfi_def_cfa_offset 20
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-THUMB-FP: .cfi_offset lr, -4
; CHECK-THUMB-FP: .cfi_offset r7, -8
; CHECK-THUMB-FP: .cfi_offset r6, -12
; CHECK-THUMB-FP: .cfi_offset r5, -16
; CHECK-THUMB-FP: .cfi_offset r4, -20
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-THUMB-FP: add r7, sp, #12
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-THUMB-FP: .cfi_def_cfa r7, 8
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-THUMB-FP: sub sp, #60
; CHECK-THUMB-FP: .cfi_endproc
; CHECK-THUMB-FP-ELIM-LABEL: _Z4testiiiiiddddd:
; CHECK-THUMB-FP-ELIM: .cfi_startproc
; CHECK-THUMB-FP-ELIM: push {r4, r5, r6, r7, lr}
; CHECK-THUMB-FP-ELIM: .cfi_def_cfa_offset 20
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-THUMB-FP-ELIM: .cfi_offset lr, -4
; CHECK-THUMB-FP-ELIM: .cfi_offset r7, -8
; CHECK-THUMB-FP-ELIM: .cfi_offset r6, -12
; CHECK-THUMB-FP-ELIM: .cfi_offset r5, -16
; CHECK-THUMB-FP-ELIM: .cfi_offset r4, -20
[ARM] Eliminate redundant "mov rN, sp" instructions in Thumb1. This takes sequences like "mov r4, sp; str r0, [r4]", and optimizes them to something like "str r0, [sp]". For regular stack variables, this optimization was already implemented: we lower loads and stores using frame indexes, which are expanded later. However, when constructing a call frame for a call with more than four arguments, the existing optimization doesn't apply. We need to use stores which are actually relative to the current value of sp, and don't have an associated frame index. This patch adds a special case to handle that construct. At the DAG level, this is an ISD::STORE where the address is a CopyFromReg from SP (plus a small constant offset). This applies only to Thumb1: in Thumb2 or ARM mode, a regular store instruction can access SP directly, so the COPY gets eliminated by existing code. The change to ARMDAGToDAGISel::SelectThumbAddrModeSP is a related cleanup: we shouldn't pretend that it can select anything other than frame indexes. Differential Revision: https://reviews.llvm.org/D59568 llvm-svn: 356601
2019-03-21 03:40:45 +08:00
; CHECK-THUMB-FP-ELIM: sub sp, #52
; CHECK-THUMB-FP-ELIM: .cfi_def_cfa_offset 72
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-THUMB-FP-ELIM: .cfi_endproc
; CHECK-THUMB-V7-FP-LABEL: _Z4testiiiiiddddd:
; CHECK-THUMB-V7-FP: .cfi_startproc
ARM: When spilling extra registers for alignment, prefer low registers on all Thumb targets. This makes it more likely that we can use the 16-bit push and pop instructions on Thumb-2, saving around 4 bytes per function. Differential Revision: http://reviews.llvm.org/D9165 llvm-svn: 235637
2015-04-24 04:31:26 +08:00
; CHECK-THUMB-V7-FP: push {r4, r6, r7, lr}
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-THUMB-V7-FP: .cfi_def_cfa_offset 16
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-THUMB-V7-FP: .cfi_offset lr, -4
ARM: When spilling extra registers for alignment, prefer low registers on all Thumb targets. This makes it more likely that we can use the 16-bit push and pop instructions on Thumb-2, saving around 4 bytes per function. Differential Revision: http://reviews.llvm.org/D9165 llvm-svn: 235637
2015-04-24 04:31:26 +08:00
; CHECK-THUMB-V7-FP: .cfi_offset r7, -8
; CHECK-THUMB-V7-FP: .cfi_offset r6, -12
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-THUMB-V7-FP: .cfi_offset r4, -16
ARM: When spilling extra registers for alignment, prefer low registers on all Thumb targets. This makes it more likely that we can use the 16-bit push and pop instructions on Thumb-2, saving around 4 bytes per function. Differential Revision: http://reviews.llvm.org/D9165 llvm-svn: 235637
2015-04-24 04:31:26 +08:00
; CHECK-THUMB-V7-FP: add r7, sp, #8
; CHECK-THUMB-V7-FP: .cfi_def_cfa r7, 8
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-THUMB-V7-FP: vpush {d8, d9, d10, d11, d12}
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-THUMB-V7-FP: .cfi_offset d12, -24
; CHECK-THUMB-V7-FP: .cfi_offset d11, -32
; CHECK-THUMB-V7-FP: .cfi_offset d10, -40
; CHECK-THUMB-V7-FP: .cfi_offset d9, -48
; CHECK-THUMB-V7-FP: .cfi_offset d8, -56
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-THUMB-V7-FP: sub sp, #24
; CHECK-THUMB-V7-FP: .cfi_endproc
; CHECK-THUMB-V7-FP-ELIM-LABEL: _Z4testiiiiiddddd:
; CHECK-THUMB-V7-FP-ELIM: .cfi_startproc
; CHECK-THUMB-V7-FP-ELIM: push {r4, lr}
; CHECK-THUMB-V7-FP-ELIM: .cfi_def_cfa_offset 8
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-THUMB-V7-FP-ELIM: .cfi_offset lr, -4
; CHECK-THUMB-V7-FP-ELIM: .cfi_offset r4, -8
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-THUMB-V7-FP-ELIM: vpush {d8, d9, d10, d11, d12}
; CHECK-THUMB-V7-FP-ELIM: .cfi_def_cfa_offset 48
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-THUMB-V7-FP-ELIM: .cfi_offset d12, -16
; CHECK-THUMB-V7-FP-ELIM: .cfi_offset d11, -24
; CHECK-THUMB-V7-FP-ELIM: .cfi_offset d10, -32
; CHECK-THUMB-V7-FP-ELIM: .cfi_offset d9, -40
; CHECK-THUMB-V7-FP-ELIM: .cfi_offset d8, -48
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-THUMB-V7-FP-ELIM: sub sp, #24
; CHECK-THUMB-V7-FP-ELIM: .cfi_def_cfa_offset 72
; CHECK-THUMB-V7-FP-ELIM: .cfi_endproc
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-THUMB-V7-FP-NOIAS-LABEL: _Z4testiiiiiddddd:
; CHECK-THUMB-V7-FP-NOIAS: .cfi_startproc
ARM: When spilling extra registers for alignment, prefer low registers on all Thumb targets. This makes it more likely that we can use the 16-bit push and pop instructions on Thumb-2, saving around 4 bytes per function. Differential Revision: http://reviews.llvm.org/D9165 llvm-svn: 235637
2015-04-24 04:31:26 +08:00
; CHECK-THUMB-V7-FP-NOIAS: push {r4, r6, r7, lr}
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-THUMB-V7-FP-NOIAS: .cfi_def_cfa_offset 16
; CHECK-THUMB-V7-FP-NOIAS: .cfi_offset 14, -4
ARM: When spilling extra registers for alignment, prefer low registers on all Thumb targets. This makes it more likely that we can use the 16-bit push and pop instructions on Thumb-2, saving around 4 bytes per function. Differential Revision: http://reviews.llvm.org/D9165 llvm-svn: 235637
2015-04-24 04:31:26 +08:00
; CHECK-THUMB-V7-FP-NOIAS: .cfi_offset 7, -8
; CHECK-THUMB-V7-FP-NOIAS: .cfi_offset 6, -12
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-THUMB-V7-FP-NOIAS: .cfi_offset 4, -16
ARM: When spilling extra registers for alignment, prefer low registers on all Thumb targets. This makes it more likely that we can use the 16-bit push and pop instructions on Thumb-2, saving around 4 bytes per function. Differential Revision: http://reviews.llvm.org/D9165 llvm-svn: 235637
2015-04-24 04:31:26 +08:00
; CHECK-THUMB-V7-FP-NOIAS: add r7, sp, #8
; CHECK-THUMB-V7-FP-NOIAS: .cfi_def_cfa 7, 8
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-THUMB-V7-FP-NOIAS: vpush {d8, d9, d10, d11, d12}
; CHECK-THUMB-V7-FP-NOIAS: .cfi_offset 268, -24
; CHECK-THUMB-V7-FP-NOIAS: .cfi_offset 267, -32
; CHECK-THUMB-V7-FP-NOIAS: .cfi_offset 266, -40
; CHECK-THUMB-V7-FP-NOIAS: .cfi_offset 265, -48
; CHECK-THUMB-V7-FP-NOIAS: .cfi_offset 264, -56
; CHECK-THUMB-V7-FP-NOIAS: sub sp, #24
; CHECK-THUMB-V7-FP-NOIAS: .cfi_endproc
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
;-------------------------------------------------------------------------------
; Test 2
;-------------------------------------------------------------------------------
declare void @throw_exception_2()
define void @test2() {
entry:
ARM: enable tail call optimisation on Thumb 2 Tail call optimisation was previously disabled on all targets other than iOS5.0+. This enables the tail call optimisation on all Thumb 2 capable platforms. The test adjustments are to remove the IR hint "tail" to function invocation. The tests were designed assuming that tail call optimisations would not kick in which no longer holds true. llvm-svn: 203575
2014-03-11 23:09:44 +08:00
call void @throw_exception_2()
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
ret void
}
; CHECK-FP-LABEL: test2:
; CHECK-FP: .cfi_startproc
; CHECK-FP: push {r11, lr}
; CHECK-FP: .cfi_def_cfa_offset 8
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-FP: .cfi_offset lr, -4
; CHECK-FP: .cfi_offset r11, -8
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-FP: mov r11, sp
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-FP: .cfi_def_cfa_register r11
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-FP: pop {r11, lr}
; CHECK-FP: mov pc, lr
; CHECK-FP: .cfi_endproc
; CHECK-FP-ELIM-LABEL: test2:
; CHECK-FP-ELIM: .cfi_startproc
; CHECK-FP-ELIM: push {r11, lr}
; CHECK-FP-ELIM: .cfi_def_cfa_offset 8
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-FP-ELIM: .cfi_offset lr, -4
; CHECK-FP-ELIM: .cfi_offset r11, -8
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-FP-ELIM: pop {r11, lr}
; CHECK-FP-ELIM: mov pc, lr
; CHECK-FP-ELIM: .cfi_endproc
; CHECK-V7-FP-LABEL: test2:
; CHECK-V7-FP: .cfi_startproc
; CHECK-V7-FP: push {r11, lr}
; CHECK-V7-FP: .cfi_def_cfa_offset 8
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-V7-FP: .cfi_offset lr, -4
; CHECK-V7-FP: .cfi_offset r11, -8
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-V7-FP: mov r11, sp
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-V7-FP: .cfi_def_cfa_register r11
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-V7-FP: pop {r11, pc}
; CHECK-V7-FP: .cfi_endproc
; CHECK-V7-FP-ELIM-LABEL: test2:
; CHECK-V7-FP-ELIM: .cfi_startproc
; CHECK-V7-FP-ELIM: push {r11, lr}
; CHECK-V7-FP-ELIM: .cfi_def_cfa_offset 8
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-V7-FP-ELIM: .cfi_offset lr, -4
; CHECK-V7-FP-ELIM: .cfi_offset r11, -8
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-V7-FP-ELIM: pop {r11, pc}
; CHECK-V7-FP-ELIM: .cfi_endproc
; CHECK-THUMB-FP-LABEL: test2:
; CHECK-THUMB-FP: .cfi_startproc
; CHECK-THUMB-FP: push {r7, lr}
; CHECK-THUMB-FP: .cfi_def_cfa_offset 8
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-THUMB-FP: .cfi_offset lr, -4
; CHECK-THUMB-FP: .cfi_offset r7, -8
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-THUMB-FP: add r7, sp, #0
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-THUMB-FP: .cfi_def_cfa_register r7
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-THUMB-FP: pop {r7, pc}
; CHECK-THUMB-FP: .cfi_endproc
; CHECK-THUMB-FP-ELIM-LABEL: test2:
; CHECK-THUMB-FP-ELIM: .cfi_startproc
; CHECK-THUMB-FP-ELIM: push {r7, lr}
; CHECK-THUMB-FP-ELIM: .cfi_def_cfa_offset 8
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-THUMB-FP-ELIM: .cfi_offset lr, -4
; CHECK-THUMB-FP-ELIM: .cfi_offset r7, -8
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-THUMB-FP-ELIM: pop {r7, pc}
; CHECK-THUMB-FP-ELIM: .cfi_endproc
; CHECK-THUMB-V7-FP-LABEL: test2:
; CHECK-THUMB-V7-FP: .cfi_startproc
; CHECK-THUMB-V7-FP: push {r7, lr}
; CHECK-THUMB-V7-FP: .cfi_def_cfa_offset 8
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-THUMB-V7-FP: .cfi_offset lr, -4
; CHECK-THUMB-V7-FP: .cfi_offset r7, -8
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-THUMB-V7-FP: mov r7, sp
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-THUMB-V7-FP: .cfi_def_cfa_register r7
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-THUMB-V7-FP: pop {r7, pc}
; CHECK-THUMB-V7-FP: .cfi_endproc
; CHECK-THUMB-V7-FP-ELIM-LABEL: test2:
; CHECK-THUMB-V7-FP-ELIM: .cfi_startproc
ARM: When spilling extra registers for alignment, prefer low registers on all Thumb targets. This makes it more likely that we can use the 16-bit push and pop instructions on Thumb-2, saving around 4 bytes per function. Differential Revision: http://reviews.llvm.org/D9165 llvm-svn: 235637
2015-04-24 04:31:26 +08:00
; CHECK-THUMB-V7-FP-ELIM: push {r7, lr}
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-THUMB-V7-FP-ELIM: .cfi_def_cfa_offset 8
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-THUMB-V7-FP-ELIM: .cfi_offset lr, -4
ARM: When spilling extra registers for alignment, prefer low registers on all Thumb targets. This makes it more likely that we can use the 16-bit push and pop instructions on Thumb-2, saving around 4 bytes per function. Differential Revision: http://reviews.llvm.org/D9165 llvm-svn: 235637
2015-04-24 04:31:26 +08:00
; CHECK-THUMB-V7-FP-ELIM: .cfi_offset r7, -8
; CHECK-THUMB-V7-FP-ELIM: pop {r7, pc}
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-THUMB-V7-FP-ELIM: .cfi_endproc
;-------------------------------------------------------------------------------
; Test 3
;-------------------------------------------------------------------------------
declare void @throw_exception_3(i32)
define i32 @test3(i32 %a, i32 %b, i32 %c, i32 %d,
i32 %e, i32 %f, i32 %g, i32 %h) {
entry:
%add = add nsw i32 %b, %a
%add1 = add nsw i32 %add, %c
%add2 = add nsw i32 %add1, %d
tail call void @throw_exception_3(i32 %add2)
%add3 = add nsw i32 %f, %e
%add4 = add nsw i32 %add3, %g
%add5 = add nsw i32 %add4, %h
tail call void @throw_exception_3(i32 %add5)
%add6 = add nsw i32 %add5, %add2
ret i32 %add6
}
; CHECK-FP-LABEL: test3:
; CHECK-FP: .cfi_startproc
; CHECK-FP: push {r4, r5, r11, lr}
; CHECK-FP: .cfi_def_cfa_offset 16
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-FP: .cfi_offset lr, -4
; CHECK-FP: .cfi_offset r11, -8
; CHECK-FP: .cfi_offset r5, -12
; CHECK-FP: .cfi_offset r4, -16
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-FP: add r11, sp, #8
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-FP: .cfi_def_cfa r11, 8
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-FP: pop {r4, r5, r11, lr}
; CHECK-FP: mov pc, lr
; CHECK-FP: .cfi_endproc
; CHECK-FP-ELIM-LABEL: test3:
; CHECK-FP-ELIM: .cfi_startproc
; CHECK-FP-ELIM: push {r4, r5, r11, lr}
; CHECK-FP-ELIM: .cfi_def_cfa_offset 16
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-FP-ELIM: .cfi_offset lr, -4
; CHECK-FP-ELIM: .cfi_offset r11, -8
; CHECK-FP-ELIM: .cfi_offset r5, -12
; CHECK-FP-ELIM: .cfi_offset r4, -16
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-FP-ELIM: pop {r4, r5, r11, lr}
; CHECK-FP-ELIM: mov pc, lr
; CHECK-FP-ELIM: .cfi_endproc
; CHECK-V7-FP-LABEL: test3:
; CHECK-V7-FP: .cfi_startproc
; CHECK-V7-FP: push {r4, r5, r11, lr}
; CHECK-V7-FP: .cfi_def_cfa_offset 16
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-V7-FP: .cfi_offset lr, -4
; CHECK-V7-FP: .cfi_offset r11, -8
; CHECK-V7-FP: .cfi_offset r5, -12
; CHECK-V7-FP: .cfi_offset r4, -16
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-V7-FP: add r11, sp, #8
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-V7-FP: .cfi_def_cfa r11, 8
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-V7-FP: pop {r4, r5, r11, pc}
; CHECK-V7-FP: .cfi_endproc
; CHECK-V7-FP-ELIM-LABEL: test3:
; CHECK-V7-FP-ELIM: .cfi_startproc
; CHECK-V7-FP-ELIM: push {r4, r5, r11, lr}
; CHECK-V7-FP-ELIM: .cfi_def_cfa_offset 16
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-V7-FP-ELIM: .cfi_offset lr, -4
; CHECK-V7-FP-ELIM: .cfi_offset r11, -8
; CHECK-V7-FP-ELIM: .cfi_offset r5, -12
; CHECK-V7-FP-ELIM: .cfi_offset r4, -16
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-V7-FP-ELIM: pop {r4, r5, r11, pc}
; CHECK-V7-FP-ELIM: .cfi_endproc
; CHECK-THUMB-FP-LABEL: test3:
; CHECK-THUMB-FP: .cfi_startproc
; CHECK-THUMB-FP: push {r4, r5, r7, lr}
; CHECK-THUMB-FP: .cfi_def_cfa_offset 16
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-THUMB-FP: .cfi_offset lr, -4
; CHECK-THUMB-FP: .cfi_offset r7, -8
; CHECK-THUMB-FP: .cfi_offset r5, -12
; CHECK-THUMB-FP: .cfi_offset r4, -16
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-THUMB-FP: add r7, sp, #8
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-THUMB-FP: .cfi_def_cfa r7, 8
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-THUMB-FP: pop {r4, r5, r7, pc}
; CHECK-THUMB-FP: .cfi_endproc
; CHECK-THUMB-FP-ELIM-LABEL: test3:
; CHECK-THUMB-FP-ELIM: .cfi_startproc
; CHECK-THUMB-FP-ELIM: push {r4, r5, r7, lr}
; CHECK-THUMB-FP-ELIM: .cfi_def_cfa_offset 16
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-THUMB-FP-ELIM: .cfi_offset lr, -4
; CHECK-THUMB-FP-ELIM: .cfi_offset r7, -8
; CHECK-THUMB-FP-ELIM: .cfi_offset r5, -12
; CHECK-THUMB-FP-ELIM: .cfi_offset r4, -16
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-THUMB-FP-ELIM: pop {r4, r5, r7, pc}
; CHECK-THUMB-FP-ELIM: .cfi_endproc
; CHECK-THUMB-V7-FP-LABEL: test3:
; CHECK-THUMB-V7-FP: .cfi_startproc
; CHECK-THUMB-V7-FP: push {r4, r5, r7, lr}
; CHECK-THUMB-V7-FP: .cfi_def_cfa_offset 16
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-THUMB-V7-FP: .cfi_offset lr, -4
; CHECK-THUMB-V7-FP: .cfi_offset r7, -8
; CHECK-THUMB-V7-FP: .cfi_offset r5, -12
; CHECK-THUMB-V7-FP: .cfi_offset r4, -16
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-THUMB-V7-FP: add r7, sp, #8
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-THUMB-V7-FP: .cfi_def_cfa r7, 8
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-THUMB-V7-FP: pop {r4, r5, r7, pc}
; CHECK-THUMB-V7-FP: .cfi_endproc
; CHECK-THUMB-V7-FP-ELIM-LABEL: test3:
; CHECK-THUMB-V7-FP-ELIM: .cfi_startproc
ARM: When spilling extra registers for alignment, prefer low registers on all Thumb targets. This makes it more likely that we can use the 16-bit push and pop instructions on Thumb-2, saving around 4 bytes per function. Differential Revision: http://reviews.llvm.org/D9165 llvm-svn: 235637
2015-04-24 04:31:26 +08:00
; CHECK-THUMB-V7-FP-ELIM: push {r4, r5, r7, lr}
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-THUMB-V7-FP-ELIM: .cfi_def_cfa_offset 16
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-THUMB-V7-FP-ELIM: .cfi_offset lr, -4
ARM: When spilling extra registers for alignment, prefer low registers on all Thumb targets. This makes it more likely that we can use the 16-bit push and pop instructions on Thumb-2, saving around 4 bytes per function. Differential Revision: http://reviews.llvm.org/D9165 llvm-svn: 235637
2015-04-24 04:31:26 +08:00
; CHECK-THUMB-V7-FP-ELIM: .cfi_offset r7, -8
[ARM] Use symbolic register names in .cfi directives only with IAS (PR19110) This is a follow-up to r203635. Saleem pointed out that since symbolic register names are much easier to read, it would be good if we could turn them off only when we really need to because we're using an external assembler. Differential Revision: http://llvm-reviews.chandlerc.com/D3056 llvm-svn: 203806
2014-03-13 23:56:41 +08:00
; CHECK-THUMB-V7-FP-ELIM: .cfi_offset r5, -12
; CHECK-THUMB-V7-FP-ELIM: .cfi_offset r4, -16
ARM: When spilling extra registers for alignment, prefer low registers on all Thumb targets. This makes it more likely that we can use the 16-bit push and pop instructions on Thumb-2, saving around 4 bytes per function. Differential Revision: http://reviews.llvm.org/D9165 llvm-svn: 235637
2015-04-24 04:31:26 +08:00
; CHECK-THUMB-V7-FP-ELIM: pop {r4, r5, r7, pc}
Generate the DWARF stack frame decode operations in the function prologue for ARM/Thumb functions. Patch by Keith Walker! llvm-svn: 201423
2014-02-15 01:19:07 +08:00
; CHECK-THUMB-V7-FP-ELIM: .cfi_endproc
;-------------------------------------------------------------------------------
; Test 4
;-------------------------------------------------------------------------------
define void @test4() nounwind {
entry:
ret void
}
; CHECK-FP-LABEL: test4:
; CHECK-FP: mov pc, lr
; CHECK-FP-NOT: .cfi_def_cfa_offset
; CHECK-FP-ELIM-LABEL: test4:
; CHECK-FP-ELIM: mov pc, lr
; CHECK-FP-ELIM-NOT: .cfi_def_cfa_offset
; CHECK-V7-FP-LABEL: test4:
; CHECK-V7-FP: bx lr
; CHECK-V7-FP-NOT: .cfi_def_cfa_offset
; CHECK-V7-FP-ELIM-LABEL: test4:
; CHECK-V7-FP-ELIM: bx lr
; CHECK-V7-FP-ELIM-NOT: .cfi_def_cfa_offset
; CHECK-THUMB-FP-LABEL: test4:
; CHECK-THUMB-FP: bx lr
; CHECK-THUMB-FP-NOT: .cfi_def_cfa_offset
; CHECK-THUMB-FP-ELIM-LABEL: test4:
; CHECK-THUMB-FP-ELIM: bx lr
; CHECK-THUMB-FP-ELIM-NOT: .cfi_def_cfa_offset
; CHECK-THUMB-V7-FP-LABEL: test4:
; CHECK-THUMB-V7-FP: bx lr
; CHECK-THUMB-V7-FP-NOT: .cfi_def_cfa_offset
; CHECK-THUMB-V7-FP-ELIM-LABEL: test4:
; CHECK-THUMB-V7-FP-ELIM: bx lr
; CHECK-THUMB-V7-FP-ELIM-NOT: .cfi_def_cfa_offset