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/WebAssembly/comparisons_f32.ll

172 lines
4.9 KiB
LLVM
Raw Normal View History

WebAssembly: implement comparison. Some of the FP comparisons (ueq, one, ult, ule, ugt, uge) are currently broken, I'll fix them in a follow-up. Reviewers: sunfish Subscribers: llvm-commits, jfb Differential Revision: http://reviews.llvm.org/D11924 llvm-svn: 244665
2015-08-12 05:02:46 +08:00
; RUN: llc < %s -asm-verbose=false | FileCheck %s
; Test that basic 32-bit floating-point comparison operations assemble as
; expected.
[WebAssembly] Use the default alignment for SIMD types. Previously WebAssembly's datalayout string had -v128:8:128. This had been an attempt to declare a certain level of support for unaligned SIMD accesses. However, clang makes its own determinations for SIMD alignment that are independent of the datalayout string, so this wasn't actually meaningful. llvm-svn: 245494
2015-08-20 04:30:20 +08:00
target datalayout = "e-p:32:32-i64:64-n32:64-S128"
WebAssembly: implement comparison. Some of the FP comparisons (ueq, one, ult, ule, ugt, uge) are currently broken, I'll fix them in a follow-up. Reviewers: sunfish Subscribers: llvm-commits, jfb Differential Revision: http://reviews.llvm.org/D11924 llvm-svn: 244665
2015-08-12 05:02:46 +08:00
target triple = "wasm32-unknown-unknown"
[WebAssembly] Switch to a more traditional assembly syntax This new syntax is built around putting each instruction on its own line in a "mnemonic op, op, op" like syntax. It also uses conventional data section directives like ".byte" and so on rather than requiring everything to be in hierarchical S-expression format. This is a more natural syntax for a ".s" file format from the perspective of LLVM MC and related tools, while remaining easy to translate into other forms as needed. llvm-svn: 249364
2015-10-06 08:27:55 +08:00
; CHECK-LABEL: ord_f32:
; CHECK-NEXT: .param f32{{$}}
; CHECK-NEXT: .param f32{{$}}
; CHECK-NEXT: .result i32{{$}}
WebAssembly: update syntax Summary: Follow the same syntax as for the spec repo. Both have evolved slightly independently and need to converge again. This, along with wasmate changes, allows me to do the following: echo "int add(int a, int b) { return a + b; }" > add.c ./out/bin/clang -O2 -S --target=wasm32-unknown-unknown add.c -o add.wack ./experimental/prototype-wasmate/wasmate.py add.wack > add.wast ./sexpr-wasm-prototype/out/sexpr-wasm add.wast -o add.wasm ./sexpr-wasm-prototype/third_party/v8-native-prototype/v8/v8/out/Release/d8 -e "print(WASM.instantiateModule(readbuffer('add.wasm'), {print:print}).add(42, 1337));" As you'd expect, the d8 shell prints out the right value. Reviewers: sunfish Subscribers: jfb, llvm-commits, dschuff Differential Revision: http://reviews.llvm.org/D13712 llvm-svn: 250480
2015-10-16 08:53:49 +08:00
; CHECK-NEXT: .local f32, f32, i32, i32, i32{{$}}
[WebAssembly] Reapply r252858, with svn add for the new file. Switch to MC for instruction printing. This encompasses several changes which are all interconnected: - Use the MC framework for printing almost all instructions. - AsmStrings are now live. - This introduces an indirection between LLVM vregs and WebAssembly registers, and a new pass, WebAssemblyRegNumbering, for computing a basic the mapping. This addresses some basic issues with argument registers and unused registers. - The way ARGUMENT instructions are handled no longer generates redundant get_local+set_local for every argument. This also changes the assembly syntax somewhat; most notably, MC's printing does not use sigils on label names, so those are no longer present, and push/pop now have a sigil to keep them unambiguous. The usage of set_local/get_local/$push/$pop will continue to evolve significantly. This patch is just one step of a larger change. llvm-svn: 252910
2015-11-13 01:04:33 +08:00
; CHECK-NEXT: f32.eq $push, (get_local 1), (get_local 1){{$}}
; CHECK-NEXT: set_local 2, $pop{{$}}
; CHECK-NEXT: f32.eq $push, (get_local 0), (get_local 0){{$}}
; CHECK-NEXT: set_local 3, $pop{{$}}
; CHECK-NEXT: i32.and $push, (get_local 3), (get_local 2){{$}}
; CHECK-NEXT: set_local 4, $pop{{$}}
; CHECK-NEXT: return (get_local 4){{$}}
WebAssembly: floating-point comparisons Summary: D11924 implemented part of the floating-point comparisons, this patch implements the rest: * Tell ISelLowering that all booleans are either 0 or 1. * Expand the eq/ne/lt/le/gt/ge floating-point comparisons to the canonical ones (similar to what Mips32r6InstrInfo.td does). * Add tests for ord/uno. * Add tests for ueq/one/ult/ule/ugt/uge. * Fix existing comparison tests to remove the (res & 1) code, which setBooleanContents stops from generating. Reviewers: sunfish Subscribers: llvm-commits, jfb Differential Revision: http://reviews.llvm.org/D11970 llvm-svn: 244779
2015-08-13 01:53:29 +08:00
define i32 @ord_f32(float %x, float %y) {
%a = fcmp ord float %x, %y
%b = zext i1 %a to i32
ret i32 %b
}
[WebAssembly] Switch to a more traditional assembly syntax This new syntax is built around putting each instruction on its own line in a "mnemonic op, op, op" like syntax. It also uses conventional data section directives like ".byte" and so on rather than requiring everything to be in hierarchical S-expression format. This is a more natural syntax for a ".s" file format from the perspective of LLVM MC and related tools, while remaining easy to translate into other forms as needed. llvm-svn: 249364
2015-10-06 08:27:55 +08:00
; CHECK-LABEL: uno_f32:
; CHECK-NEXT: .param f32{{$}}
; CHECK-NEXT: .param f32{{$}}
; CHECK-NEXT: .result i32{{$}}
WebAssembly: update syntax Summary: Follow the same syntax as for the spec repo. Both have evolved slightly independently and need to converge again. This, along with wasmate changes, allows me to do the following: echo "int add(int a, int b) { return a + b; }" > add.c ./out/bin/clang -O2 -S --target=wasm32-unknown-unknown add.c -o add.wack ./experimental/prototype-wasmate/wasmate.py add.wack > add.wast ./sexpr-wasm-prototype/out/sexpr-wasm add.wast -o add.wasm ./sexpr-wasm-prototype/third_party/v8-native-prototype/v8/v8/out/Release/d8 -e "print(WASM.instantiateModule(readbuffer('add.wasm'), {print:print}).add(42, 1337));" As you'd expect, the d8 shell prints out the right value. Reviewers: sunfish Subscribers: jfb, llvm-commits, dschuff Differential Revision: http://reviews.llvm.org/D13712 llvm-svn: 250480
2015-10-16 08:53:49 +08:00
; CHECK-NEXT: .local f32, f32, i32, i32, i32{{$}}
[WebAssembly] Reapply r252858, with svn add for the new file. Switch to MC for instruction printing. This encompasses several changes which are all interconnected: - Use the MC framework for printing almost all instructions. - AsmStrings are now live. - This introduces an indirection between LLVM vregs and WebAssembly registers, and a new pass, WebAssemblyRegNumbering, for computing a basic the mapping. This addresses some basic issues with argument registers and unused registers. - The way ARGUMENT instructions are handled no longer generates redundant get_local+set_local for every argument. This also changes the assembly syntax somewhat; most notably, MC's printing does not use sigils on label names, so those are no longer present, and push/pop now have a sigil to keep them unambiguous. The usage of set_local/get_local/$push/$pop will continue to evolve significantly. This patch is just one step of a larger change. llvm-svn: 252910
2015-11-13 01:04:33 +08:00
; CHECK-NEXT: f32.ne $push, (get_local 1), (get_local 1){{$}}
; CHECK-NEXT: set_local 2, $pop{{$}}
; CHECK-NEXT: f32.ne $push, (get_local 0), (get_local 0){{$}}
; CHECK-NEXT: set_local 3, $pop{{$}}
; CHECK-NEXT: i32.or $push, (get_local 3), (get_local 2){{$}}
; CHECK-NEXT: set_local 4, $pop{{$}}
; CHECK-NEXT: return (get_local 4){{$}}
WebAssembly: floating-point comparisons Summary: D11924 implemented part of the floating-point comparisons, this patch implements the rest: * Tell ISelLowering that all booleans are either 0 or 1. * Expand the eq/ne/lt/le/gt/ge floating-point comparisons to the canonical ones (similar to what Mips32r6InstrInfo.td does). * Add tests for ord/uno. * Add tests for ueq/one/ult/ule/ugt/uge. * Fix existing comparison tests to remove the (res & 1) code, which setBooleanContents stops from generating. Reviewers: sunfish Subscribers: llvm-commits, jfb Differential Revision: http://reviews.llvm.org/D11970 llvm-svn: 244779
2015-08-13 01:53:29 +08:00
define i32 @uno_f32(float %x, float %y) {
%a = fcmp uno float %x, %y
%b = zext i1 %a to i32
ret i32 %b
}
WebAssembly: implement comparison. Some of the FP comparisons (ueq, one, ult, ule, ugt, uge) are currently broken, I'll fix them in a follow-up. Reviewers: sunfish Subscribers: llvm-commits, jfb Differential Revision: http://reviews.llvm.org/D11924 llvm-svn: 244665
2015-08-12 05:02:46 +08:00
[WebAssembly] Switch to a more traditional assembly syntax This new syntax is built around putting each instruction on its own line in a "mnemonic op, op, op" like syntax. It also uses conventional data section directives like ".byte" and so on rather than requiring everything to be in hierarchical S-expression format. This is a more natural syntax for a ".s" file format from the perspective of LLVM MC and related tools, while remaining easy to translate into other forms as needed. llvm-svn: 249364
2015-10-06 08:27:55 +08:00
; CHECK-LABEL: oeq_f32:
; CHECK-NEXT: .param f32{{$}}
; CHECK-NEXT: .param f32{{$}}
; CHECK-NEXT: .result i32{{$}}
WebAssembly: update syntax Summary: Follow the same syntax as for the spec repo. Both have evolved slightly independently and need to converge again. This, along with wasmate changes, allows me to do the following: echo "int add(int a, int b) { return a + b; }" > add.c ./out/bin/clang -O2 -S --target=wasm32-unknown-unknown add.c -o add.wack ./experimental/prototype-wasmate/wasmate.py add.wack > add.wast ./sexpr-wasm-prototype/out/sexpr-wasm add.wast -o add.wasm ./sexpr-wasm-prototype/third_party/v8-native-prototype/v8/v8/out/Release/d8 -e "print(WASM.instantiateModule(readbuffer('add.wasm'), {print:print}).add(42, 1337));" As you'd expect, the d8 shell prints out the right value. Reviewers: sunfish Subscribers: jfb, llvm-commits, dschuff Differential Revision: http://reviews.llvm.org/D13712 llvm-svn: 250480
2015-10-16 08:53:49 +08:00
; CHECK-NEXT: .local f32, f32, i32{{$}}
[WebAssembly] Reapply r252858, with svn add for the new file. Switch to MC for instruction printing. This encompasses several changes which are all interconnected: - Use the MC framework for printing almost all instructions. - AsmStrings are now live. - This introduces an indirection between LLVM vregs and WebAssembly registers, and a new pass, WebAssemblyRegNumbering, for computing a basic the mapping. This addresses some basic issues with argument registers and unused registers. - The way ARGUMENT instructions are handled no longer generates redundant get_local+set_local for every argument. This also changes the assembly syntax somewhat; most notably, MC's printing does not use sigils on label names, so those are no longer present, and push/pop now have a sigil to keep them unambiguous. The usage of set_local/get_local/$push/$pop will continue to evolve significantly. This patch is just one step of a larger change. llvm-svn: 252910
2015-11-13 01:04:33 +08:00
; CHECK-NEXT: f32.eq $push, (get_local 0), (get_local 1){{$}}
; CHECK-NEXT: set_local 2, $pop{{$}}
; CHECK-NEXT: return (get_local 2){{$}}
WebAssembly: implement comparison. Some of the FP comparisons (ueq, one, ult, ule, ugt, uge) are currently broken, I'll fix them in a follow-up. Reviewers: sunfish Subscribers: llvm-commits, jfb Differential Revision: http://reviews.llvm.org/D11924 llvm-svn: 244665
2015-08-12 05:02:46 +08:00
define i32 @oeq_f32(float %x, float %y) {
%a = fcmp oeq float %x, %y
%b = zext i1 %a to i32
ret i32 %b
}
[WebAssembly] Switch to a more traditional assembly syntax This new syntax is built around putting each instruction on its own line in a "mnemonic op, op, op" like syntax. It also uses conventional data section directives like ".byte" and so on rather than requiring everything to be in hierarchical S-expression format. This is a more natural syntax for a ".s" file format from the perspective of LLVM MC and related tools, while remaining easy to translate into other forms as needed. llvm-svn: 249364
2015-10-06 08:27:55 +08:00
; CHECK-LABEL: une_f32:
[WebAssembly] Reapply r252858, with svn add for the new file. Switch to MC for instruction printing. This encompasses several changes which are all interconnected: - Use the MC framework for printing almost all instructions. - AsmStrings are now live. - This introduces an indirection between LLVM vregs and WebAssembly registers, and a new pass, WebAssemblyRegNumbering, for computing a basic the mapping. This addresses some basic issues with argument registers and unused registers. - The way ARGUMENT instructions are handled no longer generates redundant get_local+set_local for every argument. This also changes the assembly syntax somewhat; most notably, MC's printing does not use sigils on label names, so those are no longer present, and push/pop now have a sigil to keep them unambiguous. The usage of set_local/get_local/$push/$pop will continue to evolve significantly. This patch is just one step of a larger change. llvm-svn: 252910
2015-11-13 01:04:33 +08:00
; CHECK: f32.ne $push, (get_local 0), (get_local 1){{$}}
; CHECK-NEXT: set_local 2, $pop{{$}}
WebAssembly: implement comparison. Some of the FP comparisons (ueq, one, ult, ule, ugt, uge) are currently broken, I'll fix them in a follow-up. Reviewers: sunfish Subscribers: llvm-commits, jfb Differential Revision: http://reviews.llvm.org/D11924 llvm-svn: 244665
2015-08-12 05:02:46 +08:00
define i32 @une_f32(float %x, float %y) {
%a = fcmp une float %x, %y
%b = zext i1 %a to i32
ret i32 %b
}
[WebAssembly] Switch to a more traditional assembly syntax This new syntax is built around putting each instruction on its own line in a "mnemonic op, op, op" like syntax. It also uses conventional data section directives like ".byte" and so on rather than requiring everything to be in hierarchical S-expression format. This is a more natural syntax for a ".s" file format from the perspective of LLVM MC and related tools, while remaining easy to translate into other forms as needed. llvm-svn: 249364
2015-10-06 08:27:55 +08:00
; CHECK-LABEL: olt_f32:
[WebAssembly] Reapply r252858, with svn add for the new file. Switch to MC for instruction printing. This encompasses several changes which are all interconnected: - Use the MC framework for printing almost all instructions. - AsmStrings are now live. - This introduces an indirection between LLVM vregs and WebAssembly registers, and a new pass, WebAssemblyRegNumbering, for computing a basic the mapping. This addresses some basic issues with argument registers and unused registers. - The way ARGUMENT instructions are handled no longer generates redundant get_local+set_local for every argument. This also changes the assembly syntax somewhat; most notably, MC's printing does not use sigils on label names, so those are no longer present, and push/pop now have a sigil to keep them unambiguous. The usage of set_local/get_local/$push/$pop will continue to evolve significantly. This patch is just one step of a larger change. llvm-svn: 252910
2015-11-13 01:04:33 +08:00
; CHECK: f32.lt $push, (get_local 0), (get_local 1){{$}}
; CHECK-NEXT: set_local 2, $pop{{$}}
WebAssembly: implement comparison. Some of the FP comparisons (ueq, one, ult, ule, ugt, uge) are currently broken, I'll fix them in a follow-up. Reviewers: sunfish Subscribers: llvm-commits, jfb Differential Revision: http://reviews.llvm.org/D11924 llvm-svn: 244665
2015-08-12 05:02:46 +08:00
define i32 @olt_f32(float %x, float %y) {
%a = fcmp olt float %x, %y
%b = zext i1 %a to i32
ret i32 %b
}
[WebAssembly] Switch to a more traditional assembly syntax This new syntax is built around putting each instruction on its own line in a "mnemonic op, op, op" like syntax. It also uses conventional data section directives like ".byte" and so on rather than requiring everything to be in hierarchical S-expression format. This is a more natural syntax for a ".s" file format from the perspective of LLVM MC and related tools, while remaining easy to translate into other forms as needed. llvm-svn: 249364
2015-10-06 08:27:55 +08:00
; CHECK-LABEL: ole_f32:
[WebAssembly] Reapply r252858, with svn add for the new file. Switch to MC for instruction printing. This encompasses several changes which are all interconnected: - Use the MC framework for printing almost all instructions. - AsmStrings are now live. - This introduces an indirection between LLVM vregs and WebAssembly registers, and a new pass, WebAssemblyRegNumbering, for computing a basic the mapping. This addresses some basic issues with argument registers and unused registers. - The way ARGUMENT instructions are handled no longer generates redundant get_local+set_local for every argument. This also changes the assembly syntax somewhat; most notably, MC's printing does not use sigils on label names, so those are no longer present, and push/pop now have a sigil to keep them unambiguous. The usage of set_local/get_local/$push/$pop will continue to evolve significantly. This patch is just one step of a larger change. llvm-svn: 252910
2015-11-13 01:04:33 +08:00
; CHECK: f32.le $push, (get_local 0), (get_local 1){{$}}
; CHECK-NEXT: set_local 2, $pop{{$}}
WebAssembly: implement comparison. Some of the FP comparisons (ueq, one, ult, ule, ugt, uge) are currently broken, I'll fix them in a follow-up. Reviewers: sunfish Subscribers: llvm-commits, jfb Differential Revision: http://reviews.llvm.org/D11924 llvm-svn: 244665
2015-08-12 05:02:46 +08:00
define i32 @ole_f32(float %x, float %y) {
%a = fcmp ole float %x, %y
%b = zext i1 %a to i32
ret i32 %b
}
[WebAssembly] Switch to a more traditional assembly syntax This new syntax is built around putting each instruction on its own line in a "mnemonic op, op, op" like syntax. It also uses conventional data section directives like ".byte" and so on rather than requiring everything to be in hierarchical S-expression format. This is a more natural syntax for a ".s" file format from the perspective of LLVM MC and related tools, while remaining easy to translate into other forms as needed. llvm-svn: 249364
2015-10-06 08:27:55 +08:00
; CHECK-LABEL: ogt_f32:
[WebAssembly] Reapply r252858, with svn add for the new file. Switch to MC for instruction printing. This encompasses several changes which are all interconnected: - Use the MC framework for printing almost all instructions. - AsmStrings are now live. - This introduces an indirection between LLVM vregs and WebAssembly registers, and a new pass, WebAssemblyRegNumbering, for computing a basic the mapping. This addresses some basic issues with argument registers and unused registers. - The way ARGUMENT instructions are handled no longer generates redundant get_local+set_local for every argument. This also changes the assembly syntax somewhat; most notably, MC's printing does not use sigils on label names, so those are no longer present, and push/pop now have a sigil to keep them unambiguous. The usage of set_local/get_local/$push/$pop will continue to evolve significantly. This patch is just one step of a larger change. llvm-svn: 252910
2015-11-13 01:04:33 +08:00
; CHECK: f32.gt $push, (get_local 0), (get_local 1){{$}}
; CHECK-NEXT: set_local 2, $pop{{$}}
WebAssembly: implement comparison. Some of the FP comparisons (ueq, one, ult, ule, ugt, uge) are currently broken, I'll fix them in a follow-up. Reviewers: sunfish Subscribers: llvm-commits, jfb Differential Revision: http://reviews.llvm.org/D11924 llvm-svn: 244665
2015-08-12 05:02:46 +08:00
define i32 @ogt_f32(float %x, float %y) {
%a = fcmp ogt float %x, %y
%b = zext i1 %a to i32
ret i32 %b
}
[WebAssembly] Switch to a more traditional assembly syntax This new syntax is built around putting each instruction on its own line in a "mnemonic op, op, op" like syntax. It also uses conventional data section directives like ".byte" and so on rather than requiring everything to be in hierarchical S-expression format. This is a more natural syntax for a ".s" file format from the perspective of LLVM MC and related tools, while remaining easy to translate into other forms as needed. llvm-svn: 249364
2015-10-06 08:27:55 +08:00
; CHECK-LABEL: oge_f32:
[WebAssembly] Reapply r252858, with svn add for the new file. Switch to MC for instruction printing. This encompasses several changes which are all interconnected: - Use the MC framework for printing almost all instructions. - AsmStrings are now live. - This introduces an indirection between LLVM vregs and WebAssembly registers, and a new pass, WebAssemblyRegNumbering, for computing a basic the mapping. This addresses some basic issues with argument registers and unused registers. - The way ARGUMENT instructions are handled no longer generates redundant get_local+set_local for every argument. This also changes the assembly syntax somewhat; most notably, MC's printing does not use sigils on label names, so those are no longer present, and push/pop now have a sigil to keep them unambiguous. The usage of set_local/get_local/$push/$pop will continue to evolve significantly. This patch is just one step of a larger change. llvm-svn: 252910
2015-11-13 01:04:33 +08:00
; CHECK: f32.ge $push, (get_local 0), (get_local 1){{$}}
; CHECK-NEXT: set_local 2, $pop{{$}}
WebAssembly: implement comparison. Some of the FP comparisons (ueq, one, ult, ule, ugt, uge) are currently broken, I'll fix them in a follow-up. Reviewers: sunfish Subscribers: llvm-commits, jfb Differential Revision: http://reviews.llvm.org/D11924 llvm-svn: 244665
2015-08-12 05:02:46 +08:00
define i32 @oge_f32(float %x, float %y) {
%a = fcmp oge float %x, %y
%b = zext i1 %a to i32
ret i32 %b
}
WebAssembly: floating-point comparisons Summary: D11924 implemented part of the floating-point comparisons, this patch implements the rest: * Tell ISelLowering that all booleans are either 0 or 1. * Expand the eq/ne/lt/le/gt/ge floating-point comparisons to the canonical ones (similar to what Mips32r6InstrInfo.td does). * Add tests for ord/uno. * Add tests for ueq/one/ult/ule/ugt/uge. * Fix existing comparison tests to remove the (res & 1) code, which setBooleanContents stops from generating. Reviewers: sunfish Subscribers: llvm-commits, jfb Differential Revision: http://reviews.llvm.org/D11970 llvm-svn: 244779
2015-08-13 01:53:29 +08:00
; Expanded comparisons, which also check for NaN.
[WebAssembly] Switch to a more traditional assembly syntax This new syntax is built around putting each instruction on its own line in a "mnemonic op, op, op" like syntax. It also uses conventional data section directives like ".byte" and so on rather than requiring everything to be in hierarchical S-expression format. This is a more natural syntax for a ".s" file format from the perspective of LLVM MC and related tools, while remaining easy to translate into other forms as needed. llvm-svn: 249364
2015-10-06 08:27:55 +08:00
; CHECK-LABEL: ueq_f32:
; CHECK-NEXT: .param f32{{$}}
; CHECK-NEXT: .param f32{{$}}
; CHECK-NEXT: .result i32{{$}}
WebAssembly: update syntax Summary: Follow the same syntax as for the spec repo. Both have evolved slightly independently and need to converge again. This, along with wasmate changes, allows me to do the following: echo "int add(int a, int b) { return a + b; }" > add.c ./out/bin/clang -O2 -S --target=wasm32-unknown-unknown add.c -o add.wack ./experimental/prototype-wasmate/wasmate.py add.wack > add.wast ./sexpr-wasm-prototype/out/sexpr-wasm add.wast -o add.wasm ./sexpr-wasm-prototype/third_party/v8-native-prototype/v8/v8/out/Release/d8 -e "print(WASM.instantiateModule(readbuffer('add.wasm'), {print:print}).add(42, 1337));" As you'd expect, the d8 shell prints out the right value. Reviewers: sunfish Subscribers: jfb, llvm-commits, dschuff Differential Revision: http://reviews.llvm.org/D13712 llvm-svn: 250480
2015-10-16 08:53:49 +08:00
; CHECK-NEXT: .local f32, f32, i32, i32, i32, i32, i32{{$}}
[WebAssembly] Reapply r252858, with svn add for the new file. Switch to MC for instruction printing. This encompasses several changes which are all interconnected: - Use the MC framework for printing almost all instructions. - AsmStrings are now live. - This introduces an indirection between LLVM vregs and WebAssembly registers, and a new pass, WebAssemblyRegNumbering, for computing a basic the mapping. This addresses some basic issues with argument registers and unused registers. - The way ARGUMENT instructions are handled no longer generates redundant get_local+set_local for every argument. This also changes the assembly syntax somewhat; most notably, MC's printing does not use sigils on label names, so those are no longer present, and push/pop now have a sigil to keep them unambiguous. The usage of set_local/get_local/$push/$pop will continue to evolve significantly. This patch is just one step of a larger change. llvm-svn: 252910
2015-11-13 01:04:33 +08:00
; CHECK-NEXT: f32.eq $push, (get_local 0), (get_local 1){{$}}
; CHECK-NEXT: set_local 2, $pop{{$}}
; CHECK-NEXT: f32.ne $push, (get_local 1), (get_local 1){{$}}
; CHECK-NEXT: set_local 3, $pop{{$}}
; CHECK-NEXT: f32.ne $push, (get_local 0), (get_local 0){{$}}
; CHECK-NEXT: set_local 4, $pop{{$}}
; CHECK-NEXT: i32.or $push, (get_local 4), (get_local 3){{$}}
; CHECK-NEXT: set_local 5, $pop{{$}}
; CHECK-NEXT: i32.or $push, (get_local 2), (get_local 5){{$}}
; CHECK-NEXT: set_local 6, $pop{{$}}
; CHECK-NEXT: return (get_local 6){{$}}
WebAssembly: floating-point comparisons Summary: D11924 implemented part of the floating-point comparisons, this patch implements the rest: * Tell ISelLowering that all booleans are either 0 or 1. * Expand the eq/ne/lt/le/gt/ge floating-point comparisons to the canonical ones (similar to what Mips32r6InstrInfo.td does). * Add tests for ord/uno. * Add tests for ueq/one/ult/ule/ugt/uge. * Fix existing comparison tests to remove the (res & 1) code, which setBooleanContents stops from generating. Reviewers: sunfish Subscribers: llvm-commits, jfb Differential Revision: http://reviews.llvm.org/D11970 llvm-svn: 244779
2015-08-13 01:53:29 +08:00
define i32 @ueq_f32(float %x, float %y) {
%a = fcmp ueq float %x, %y
%b = zext i1 %a to i32
ret i32 %b
}
[WebAssembly] Switch to a more traditional assembly syntax This new syntax is built around putting each instruction on its own line in a "mnemonic op, op, op" like syntax. It also uses conventional data section directives like ".byte" and so on rather than requiring everything to be in hierarchical S-expression format. This is a more natural syntax for a ".s" file format from the perspective of LLVM MC and related tools, while remaining easy to translate into other forms as needed. llvm-svn: 249364
2015-10-06 08:27:55 +08:00
; CHECK-LABEL: one_f32:
[WebAssembly] Reapply r252858, with svn add for the new file. Switch to MC for instruction printing. This encompasses several changes which are all interconnected: - Use the MC framework for printing almost all instructions. - AsmStrings are now live. - This introduces an indirection between LLVM vregs and WebAssembly registers, and a new pass, WebAssemblyRegNumbering, for computing a basic the mapping. This addresses some basic issues with argument registers and unused registers. - The way ARGUMENT instructions are handled no longer generates redundant get_local+set_local for every argument. This also changes the assembly syntax somewhat; most notably, MC's printing does not use sigils on label names, so those are no longer present, and push/pop now have a sigil to keep them unambiguous. The usage of set_local/get_local/$push/$pop will continue to evolve significantly. This patch is just one step of a larger change. llvm-svn: 252910
2015-11-13 01:04:33 +08:00
; CHECK: f32.ne $push, (get_local 0), (get_local 1){{$}}
; CHECK-NEXT: set_local 2, $pop{{$}}
WebAssembly: floating-point comparisons Summary: D11924 implemented part of the floating-point comparisons, this patch implements the rest: * Tell ISelLowering that all booleans are either 0 or 1. * Expand the eq/ne/lt/le/gt/ge floating-point comparisons to the canonical ones (similar to what Mips32r6InstrInfo.td does). * Add tests for ord/uno. * Add tests for ueq/one/ult/ule/ugt/uge. * Fix existing comparison tests to remove the (res & 1) code, which setBooleanContents stops from generating. Reviewers: sunfish Subscribers: llvm-commits, jfb Differential Revision: http://reviews.llvm.org/D11970 llvm-svn: 244779
2015-08-13 01:53:29 +08:00
define i32 @one_f32(float %x, float %y) {
%a = fcmp one float %x, %y
%b = zext i1 %a to i32
ret i32 %b
}
[WebAssembly] Switch to a more traditional assembly syntax This new syntax is built around putting each instruction on its own line in a "mnemonic op, op, op" like syntax. It also uses conventional data section directives like ".byte" and so on rather than requiring everything to be in hierarchical S-expression format. This is a more natural syntax for a ".s" file format from the perspective of LLVM MC and related tools, while remaining easy to translate into other forms as needed. llvm-svn: 249364
2015-10-06 08:27:55 +08:00
; CHECK-LABEL: ult_f32:
[WebAssembly] Reapply r252858, with svn add for the new file. Switch to MC for instruction printing. This encompasses several changes which are all interconnected: - Use the MC framework for printing almost all instructions. - AsmStrings are now live. - This introduces an indirection between LLVM vregs and WebAssembly registers, and a new pass, WebAssemblyRegNumbering, for computing a basic the mapping. This addresses some basic issues with argument registers and unused registers. - The way ARGUMENT instructions are handled no longer generates redundant get_local+set_local for every argument. This also changes the assembly syntax somewhat; most notably, MC's printing does not use sigils on label names, so those are no longer present, and push/pop now have a sigil to keep them unambiguous. The usage of set_local/get_local/$push/$pop will continue to evolve significantly. This patch is just one step of a larger change. llvm-svn: 252910
2015-11-13 01:04:33 +08:00
; CHECK: f32.lt $push, (get_local 0), (get_local 1){{$}}
; CHECK-NEXT: set_local 2, $pop{{$}}
WebAssembly: floating-point comparisons Summary: D11924 implemented part of the floating-point comparisons, this patch implements the rest: * Tell ISelLowering that all booleans are either 0 or 1. * Expand the eq/ne/lt/le/gt/ge floating-point comparisons to the canonical ones (similar to what Mips32r6InstrInfo.td does). * Add tests for ord/uno. * Add tests for ueq/one/ult/ule/ugt/uge. * Fix existing comparison tests to remove the (res & 1) code, which setBooleanContents stops from generating. Reviewers: sunfish Subscribers: llvm-commits, jfb Differential Revision: http://reviews.llvm.org/D11970 llvm-svn: 244779
2015-08-13 01:53:29 +08:00
define i32 @ult_f32(float %x, float %y) {
%a = fcmp ult float %x, %y
%b = zext i1 %a to i32
ret i32 %b
}
[WebAssembly] Switch to a more traditional assembly syntax This new syntax is built around putting each instruction on its own line in a "mnemonic op, op, op" like syntax. It also uses conventional data section directives like ".byte" and so on rather than requiring everything to be in hierarchical S-expression format. This is a more natural syntax for a ".s" file format from the perspective of LLVM MC and related tools, while remaining easy to translate into other forms as needed. llvm-svn: 249364
2015-10-06 08:27:55 +08:00
; CHECK-LABEL: ule_f32:
[WebAssembly] Reapply r252858, with svn add for the new file. Switch to MC for instruction printing. This encompasses several changes which are all interconnected: - Use the MC framework for printing almost all instructions. - AsmStrings are now live. - This introduces an indirection between LLVM vregs and WebAssembly registers, and a new pass, WebAssemblyRegNumbering, for computing a basic the mapping. This addresses some basic issues with argument registers and unused registers. - The way ARGUMENT instructions are handled no longer generates redundant get_local+set_local for every argument. This also changes the assembly syntax somewhat; most notably, MC's printing does not use sigils on label names, so those are no longer present, and push/pop now have a sigil to keep them unambiguous. The usage of set_local/get_local/$push/$pop will continue to evolve significantly. This patch is just one step of a larger change. llvm-svn: 252910
2015-11-13 01:04:33 +08:00
; CHECK: f32.le $push, (get_local 0), (get_local 1){{$}}
; CHECK-NEXT: set_local 2, $pop{{$}}
WebAssembly: floating-point comparisons Summary: D11924 implemented part of the floating-point comparisons, this patch implements the rest: * Tell ISelLowering that all booleans are either 0 or 1. * Expand the eq/ne/lt/le/gt/ge floating-point comparisons to the canonical ones (similar to what Mips32r6InstrInfo.td does). * Add tests for ord/uno. * Add tests for ueq/one/ult/ule/ugt/uge. * Fix existing comparison tests to remove the (res & 1) code, which setBooleanContents stops from generating. Reviewers: sunfish Subscribers: llvm-commits, jfb Differential Revision: http://reviews.llvm.org/D11970 llvm-svn: 244779
2015-08-13 01:53:29 +08:00
define i32 @ule_f32(float %x, float %y) {
%a = fcmp ule float %x, %y
%b = zext i1 %a to i32
ret i32 %b
}
[WebAssembly] Switch to a more traditional assembly syntax This new syntax is built around putting each instruction on its own line in a "mnemonic op, op, op" like syntax. It also uses conventional data section directives like ".byte" and so on rather than requiring everything to be in hierarchical S-expression format. This is a more natural syntax for a ".s" file format from the perspective of LLVM MC and related tools, while remaining easy to translate into other forms as needed. llvm-svn: 249364
2015-10-06 08:27:55 +08:00
; CHECK-LABEL: ugt_f32:
[WebAssembly] Reapply r252858, with svn add for the new file. Switch to MC for instruction printing. This encompasses several changes which are all interconnected: - Use the MC framework for printing almost all instructions. - AsmStrings are now live. - This introduces an indirection between LLVM vregs and WebAssembly registers, and a new pass, WebAssemblyRegNumbering, for computing a basic the mapping. This addresses some basic issues with argument registers and unused registers. - The way ARGUMENT instructions are handled no longer generates redundant get_local+set_local for every argument. This also changes the assembly syntax somewhat; most notably, MC's printing does not use sigils on label names, so those are no longer present, and push/pop now have a sigil to keep them unambiguous. The usage of set_local/get_local/$push/$pop will continue to evolve significantly. This patch is just one step of a larger change. llvm-svn: 252910
2015-11-13 01:04:33 +08:00
; CHECK: f32.gt $push, (get_local 0), (get_local 1){{$}}
; CHECK-NEXT: set_local 2, $pop{{$}}
WebAssembly: floating-point comparisons Summary: D11924 implemented part of the floating-point comparisons, this patch implements the rest: * Tell ISelLowering that all booleans are either 0 or 1. * Expand the eq/ne/lt/le/gt/ge floating-point comparisons to the canonical ones (similar to what Mips32r6InstrInfo.td does). * Add tests for ord/uno. * Add tests for ueq/one/ult/ule/ugt/uge. * Fix existing comparison tests to remove the (res & 1) code, which setBooleanContents stops from generating. Reviewers: sunfish Subscribers: llvm-commits, jfb Differential Revision: http://reviews.llvm.org/D11970 llvm-svn: 244779
2015-08-13 01:53:29 +08:00
define i32 @ugt_f32(float %x, float %y) {
%a = fcmp ugt float %x, %y
%b = zext i1 %a to i32
ret i32 %b
}
[WebAssembly] Switch to a more traditional assembly syntax This new syntax is built around putting each instruction on its own line in a "mnemonic op, op, op" like syntax. It also uses conventional data section directives like ".byte" and so on rather than requiring everything to be in hierarchical S-expression format. This is a more natural syntax for a ".s" file format from the perspective of LLVM MC and related tools, while remaining easy to translate into other forms as needed. llvm-svn: 249364
2015-10-06 08:27:55 +08:00
; CHECK-LABEL: uge_f32:
[WebAssembly] Reapply r252858, with svn add for the new file. Switch to MC for instruction printing. This encompasses several changes which are all interconnected: - Use the MC framework for printing almost all instructions. - AsmStrings are now live. - This introduces an indirection between LLVM vregs and WebAssembly registers, and a new pass, WebAssemblyRegNumbering, for computing a basic the mapping. This addresses some basic issues with argument registers and unused registers. - The way ARGUMENT instructions are handled no longer generates redundant get_local+set_local for every argument. This also changes the assembly syntax somewhat; most notably, MC's printing does not use sigils on label names, so those are no longer present, and push/pop now have a sigil to keep them unambiguous. The usage of set_local/get_local/$push/$pop will continue to evolve significantly. This patch is just one step of a larger change. llvm-svn: 252910
2015-11-13 01:04:33 +08:00
; CHECK: f32.ge $push, (get_local 0), (get_local 1){{$}}
; CHECK-NEXT: set_local 2, $pop{{$}}
WebAssembly: floating-point comparisons Summary: D11924 implemented part of the floating-point comparisons, this patch implements the rest: * Tell ISelLowering that all booleans are either 0 or 1. * Expand the eq/ne/lt/le/gt/ge floating-point comparisons to the canonical ones (similar to what Mips32r6InstrInfo.td does). * Add tests for ord/uno. * Add tests for ueq/one/ult/ule/ugt/uge. * Fix existing comparison tests to remove the (res & 1) code, which setBooleanContents stops from generating. Reviewers: sunfish Subscribers: llvm-commits, jfb Differential Revision: http://reviews.llvm.org/D11970 llvm-svn: 244779
2015-08-13 01:53:29 +08:00
define i32 @uge_f32(float %x, float %y) {
%a = fcmp uge float %x, %y
%b = zext i1 %a to i32
ret i32 %b
}