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/fp16-promote.ll

976 lines
26 KiB
LLVM
Raw Normal View History

[FileCheck] Add -allow-deprecated-dag-overlap to failing llvm tests See https://reviews.llvm.org/D47106 for details. Reviewed By: probinson Differential Revision: https://reviews.llvm.org/D47171 This commit drops that patch's changes to: llvm/test/CodeGen/NVPTX/f16x2-instructions.ll llvm/test/CodeGen/NVPTX/param-load-store.ll For some reason, the dos line endings there prevent me from commiting via the monorepo. A follow-up commit (not via the monorepo) will finish the patch. llvm-svn: 336843
2018-07-12 04:25:49 +08:00
; RUN: llc -asm-verbose=false < %s -mattr=+vfp3,+fp16 | FileCheck -allow-deprecated-dag-overlap %s -check-prefix=CHECK-FP16 --check-prefix=CHECK-VFP -check-prefix=CHECK-ALL
; RUN: llc -asm-verbose=false < %s | FileCheck -allow-deprecated-dag-overlap %s -check-prefix=CHECK-LIBCALL --check-prefix=CHECK-VFP -check-prefix=CHECK-ALL --check-prefix=CHECK-LIBCALL-VFP
[ARM] Replace fp-only-sp and d16 with fp64 and d32. Those two subtarget features were awkward because their semantics are reversed: each one indicates the _lack_ of support for something in the architecture, rather than the presence. As a consequence, you don't get the behavior you want if you combine two sets of feature bits. Each SubtargetFeature for an FP architecture version now comes in four versions, one for each combination of those options. So you can still say (for example) '+vfp2' in a feature string and it will mean what it's always meant, but there's a new string '+vfp2d16sp' meaning the version without those extra options. A lot of this change is just mechanically replacing positive checks for the old features with negative checks for the new ones. But one more interesting change is that I've rearranged getFPUFeatures() so that the main FPU feature is appended to the output list *before* rather than after the features derived from the Restriction field, so that -fp64 and -d32 can override defaults added by the main feature. Reviewers: dmgreen, samparker, SjoerdMeijer Subscribers: srhines, javed.absar, eraman, kristof.beyls, hiraditya, zzheng, Petar.Avramovic, cfe-commits, llvm-commits Tags: #clang, #llvm Differential Revision: https://reviews.llvm.org/D60691 llvm-svn: 361845
2019-05-29 00:13:20 +08:00
; RUN: llc -asm-verbose=false < %s -mattr=-vfp2d16sp | FileCheck -allow-deprecated-dag-overlap %s --check-prefix=CHECK-LIBCALL -check-prefix=CHECK-NOVFP -check-prefix=CHECK-ALL
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
target datalayout = "e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:64:64-f32:32:32-f64:64:64-v64:64:64-v128:128:128-a0:0:64-n32"
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
target triple = "armv7---eabihf"
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-ALL-LABEL: test_fadd:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: vcvtb.f32.f16
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
; CHECK-LIBCALL: bl __aeabi_h2f
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-VFP: vadd.f32
; CHECK-NOVFP: bl __aeabi_fadd
; CHECK-FP16: vcvtb.f16.f32
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_fadd(half* %p, half* %q) #0 {
%a = load half, half* %p, align 2
%b = load half, half* %q, align 2
%r = fadd half %a, %b
store half %r, half* %p
ret void
}
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-ALL-LABEL: test_fsub:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: vcvtb.f32.f16
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
; CHECK-LIBCALL: bl __aeabi_h2f
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-VFP: vsub.f32
; CHECK-NOVFP: bl __aeabi_fsub
; CHECK-FP16: vcvtb.f16.f32
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_fsub(half* %p, half* %q) #0 {
%a = load half, half* %p, align 2
%b = load half, half* %q, align 2
%r = fsub half %a, %b
store half %r, half* %p
ret void
}
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-ALL-LABEL: test_fmul:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: vcvtb.f32.f16
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
; CHECK-LIBCALL: bl __aeabi_h2f
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-VFP: vmul.f32
; CHECK-NOVFP: bl __aeabi_fmul
; CHECK-FP16: vcvtb.f16.f32
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_fmul(half* %p, half* %q) #0 {
%a = load half, half* %p, align 2
%b = load half, half* %q, align 2
%r = fmul half %a, %b
store half %r, half* %p
ret void
}
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-ALL-LABEL: test_fdiv:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: vcvtb.f32.f16
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
; CHECK-LIBCALL: bl __aeabi_h2f
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-VFP: vdiv.f32
; CHECK-NOVFP: bl __aeabi_fdiv
; CHECK-FP16: vcvtb.f16.f32
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_fdiv(half* %p, half* %q) #0 {
%a = load half, half* %p, align 2
%b = load half, half* %q, align 2
%r = fdiv half %a, %b
store half %r, half* %p
ret void
}
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-ALL-LABEL: test_frem:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: vcvtb.f32.f16
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
; CHECK-LIBCALL: bl __aeabi_h2f
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL: bl fmodf
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-FP16: vcvtb.f16.f32
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_frem(half* %p, half* %q) #0 {
%a = load half, half* %p, align 2
%b = load half, half* %q, align 2
%r = frem half %a, %b
store half %r, half* %p
ret void
}
; CHECK-ALL-LABEL: test_load_store:
; CHECK-ALL-NEXT: .fnstart
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-ALL: ldrh {{r[0-9]+}}, [{{r[0-9]+}}]
; CHECK-ALL: strh {{r[0-9]+}}, [{{r[0-9]+}}]
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_load_store(half* %p, half* %q) #0 {
%a = load half, half* %p, align 2
store half %a, half* %q
ret void
}
; Testing only successfull compilation of function calls. In ARM ABI, half
; args and returns are handled as f32.
declare half @test_callee(half %a, half %b) #0
; CHECK-ALL-LABEL: test_call:
; CHECK-ALL-NEXT: .fnstart
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-ALL-NEXT: .save {r11, lr}
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-ALL-NEXT: push {r11, lr}
; CHECK-ALL-NEXT: bl test_callee
; CHECK-ALL-NEXT: pop {r11, pc}
define half @test_call(half %a, half %b) #0 {
%r = call half @test_callee(half %a, half %b)
ret half %r
}
; CHECK-ALL-LABEL: test_call_flipped:
; CHECK-ALL-NEXT: .fnstart
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-ALL-NEXT: .save {r11, lr}
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-ALL-NEXT: push {r11, lr}
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-VFP-NEXT: vmov.f32 s2, s0
; CHECK-VFP-NEXT: vmov.f32 s0, s1
; CHECK-VFP-NEXT: vmov.f32 s1, s2
; CHECK-NOVFP-NEXT: mov r2, r0
; CHECK-NOVFP-NEXT: mov r0, r1
; CHECK-NOVFP-NEXT: mov r1, r2
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-ALL-NEXT: bl test_callee
; CHECK-ALL-NEXT: pop {r11, pc}
define half @test_call_flipped(half %a, half %b) #0 {
%r = call half @test_callee(half %b, half %a)
ret half %r
}
; CHECK-ALL-LABEL: test_tailcall_flipped:
; CHECK-ALL-NEXT: .fnstart
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-VFP-NEXT: vmov.f32 s2, s0
; CHECK-VFP-NEXT: vmov.f32 s0, s1
; CHECK-VFP-NEXT: vmov.f32 s1, s2
; CHECK-NOVFP-NEXT: mov r2, r0
; CHECK-NOVFP-NEXT: mov r0, r1
; CHECK-NOVFP-NEXT: mov r1, r2
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-ALL-NEXT: b test_callee
define half @test_tailcall_flipped(half %a, half %b) #0 {
%r = tail call half @test_callee(half %b, half %a)
ret half %r
}
; Optimizer picks %p or %q based on %c and only loads that value
; No conversion is needed
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-ALL-LABEL: test_select:
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-ALL: cmp {{r[0-9]+}}, #0
; CHECK-ALL: movne {{r[0-9]+}}, {{r[0-9]+}}
; CHECK-ALL: ldrh {{r[0-9]+}}, [{{r[0-9]+}}]
; CHECK-ALL: strh {{r[0-9]+}}, [{{r[0-9]+}}]
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_select(half* %p, half* %q, i1 zeroext %c) #0 {
%a = load half, half* %p, align 2
%b = load half, half* %q, align 2
%r = select i1 %c, half %a, half %b
store half %r, half* %p
ret void
}
[ARM] Use VCMP, not VCMPE, for floating point equality comparisons When generating a floating point comparison we currently unconditionally generate VCMPE. This has the sideeffect of setting the cumulative Invalid bit in FPSCR if any of the operands are QNaN. It is expected that use of a relational predicate on a QNaN value should raise Invalid. Quoting from the C standard: The relational and equality operators support the usual mathematical relationships between numeric values. For any ordered pair of numeric values exactly one of relationships the less, greater, equal and is true. Relational operators may raise the floating-point exception when argument values are NaNs. The standard doesn't explicitly state the expectation for equality operators, but the implication and obvious expectation is that equality operators should not raise Invalid on a QNaN input, as those predicates are wholly defined on unordered inputs (to return not equal). Therefore, add a new operand to ARMISD::FPCMP and FPCMPZ indicating if QNaN should raise Invalid, and pipe that through to TableGen. llvm-svn: 294945
2017-02-13 20:32:47 +08:00
; Test only two variants of fcmp. These get translated to f32 vcmp
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; instructions anyway.
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-ALL-LABEL: test_fcmp_une:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: vcvtb.f32.f16
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
; CHECK-LIBCALL: bl __aeabi_h2f
[ARM] Use VCMP, not VCMPE, for floating point equality comparisons When generating a floating point comparison we currently unconditionally generate VCMPE. This has the sideeffect of setting the cumulative Invalid bit in FPSCR if any of the operands are QNaN. It is expected that use of a relational predicate on a QNaN value should raise Invalid. Quoting from the C standard: The relational and equality operators support the usual mathematical relationships between numeric values. For any ordered pair of numeric values exactly one of relationships the less, greater, equal and is true. Relational operators may raise the floating-point exception when argument values are NaNs. The standard doesn't explicitly state the expectation for equality operators, but the implication and obvious expectation is that equality operators should not raise Invalid on a QNaN input, as those predicates are wholly defined on unordered inputs (to return not equal). Therefore, add a new operand to ARMISD::FPCMP and FPCMPZ indicating if QNaN should raise Invalid, and pipe that through to TableGen. llvm-svn: 294945
2017-02-13 20:32:47 +08:00
; CHECK-VFP: vcmp.f32
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-NOVFP: bl __aeabi_fcmpeq
[ARM] Materialise some boolean values to avoid a branch This patch combines some cases of ARMISD::CMOV for integers that arise in comparisons of the form a != b ? x : 0 a == b ? 0 : x and that currently (e.g. in Thumb1) are emitted as branches. Differential Revision: https://reviews.llvm.org/D34515 llvm-svn: 325323
2018-02-16 17:23:59 +08:00
; CHECK-VFP-NEXT: vmrs APSR_nzcv, fpscr
; CHECK-VFP-NEXT: movwne
; CHECK-NOVFP-NEXT: clz r0, r0
; CHECK-NOVFP-NEXT: lsr r0, r0, #5
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define i1 @test_fcmp_une(half* %p, half* %q) #0 {
%a = load half, half* %p, align 2
%b = load half, half* %q, align 2
%r = fcmp une half %a, %b
ret i1 %r
}
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-ALL-LABEL: test_fcmp_ueq:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: vcvtb.f32.f16
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
; CHECK-LIBCALL: bl __aeabi_h2f
[ARM] Use VCMP, not VCMPE, for floating point equality comparisons When generating a floating point comparison we currently unconditionally generate VCMPE. This has the sideeffect of setting the cumulative Invalid bit in FPSCR if any of the operands are QNaN. It is expected that use of a relational predicate on a QNaN value should raise Invalid. Quoting from the C standard: The relational and equality operators support the usual mathematical relationships between numeric values. For any ordered pair of numeric values exactly one of relationships the less, greater, equal and is true. Relational operators may raise the floating-point exception when argument values are NaNs. The standard doesn't explicitly state the expectation for equality operators, but the implication and obvious expectation is that equality operators should not raise Invalid on a QNaN input, as those predicates are wholly defined on unordered inputs (to return not equal). Therefore, add a new operand to ARMISD::FPCMP and FPCMPZ indicating if QNaN should raise Invalid, and pipe that through to TableGen. llvm-svn: 294945
2017-02-13 20:32:47 +08:00
; CHECK-VFP: vcmp.f32
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-NOVFP: bl __aeabi_fcmpeq
; CHECK-FP16: vmrs APSR_nzcv, fpscr
; CHECK-LIBCALL: movw{{ne|eq}}
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define i1 @test_fcmp_ueq(half* %p, half* %q) #0 {
%a = load half, half* %p, align 2
%b = load half, half* %q, align 2
%r = fcmp ueq half %a, %b
ret i1 %r
}
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-ALL-LABEL: test_br_cc:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: vcvtb.f32.f16
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
; CHECK-LIBCALL: bl __aeabi_h2f
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-VFP: vcmpe.f32
; CHECK-NOVFP: bl __aeabi_fcmplt
; CHECK-FP16: vmrs APSR_nzcv, fpscr
; CHECK-VFP: strmi
; CHECK-VFP: strpl
; CHECK-NOVFP: strne
; CHECK-NOVFP: streq
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_br_cc(half* %p, half* %q, i32* %p1, i32* %p2) #0 {
%a = load half, half* %p, align 2
%b = load half, half* %q, align 2
%c = fcmp uge half %a, %b
br i1 %c, label %then, label %else
then:
store i32 0, i32* %p1
ret void
else:
store i32 0, i32* %p2
ret void
}
declare i1 @test_dummy(half* %p) #0
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-ALL-LABEL: test_phi:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: [[LOOP:.LBB[1-9_]+]]:
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: bl test_dummy
; CHECK-FP16: bne [[LOOP]]
; CHECK-FP16: vcvtb.f16.f32
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-LIBCALL-VFP: bl __aeabi_h2f
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL: [[LOOP:.LBB[1-9_]+]]:
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-LIBCALL-VFP: bl __aeabi_h2f
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL: bl test_dummy
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-LIBCALL: bne [[LOOP]]
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-LIBCALL-VFP: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_phi(half* %p) #0 {
entry:
%a = load half, half* %p
br label %loop
loop:
%r = phi half [%a, %entry], [%b, %loop]
%b = load half, half* %p
%c = call i1 @test_dummy(half* %p)
br i1 %c, label %loop, label %return
return:
store half %r, half* %p
ret void
}
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-ALL-LABEL: test_fptosi_i32:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-VFP: vcvt.s32.f32
; CHECK-NOVFP: bl __aeabi_f2iz
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define i32 @test_fptosi_i32(half* %p) #0 {
%a = load half, half* %p, align 2
%r = fptosi half %a to i32
ret i32 %r
}
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-ALL-LABEL: test_fptosi_i64:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-ALL: bl __aeabi_f2lz
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define i64 @test_fptosi_i64(half* %p) #0 {
%a = load half, half* %p, align 2
%r = fptosi half %a to i64
ret i64 %r
}
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-ALL-LABEL: test_fptoui_i32:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-VFP: vcvt.u32.f32
; CHECK-NOVFP: bl __aeabi_f2uiz
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define i32 @test_fptoui_i32(half* %p) #0 {
%a = load half, half* %p, align 2
%r = fptoui half %a to i32
ret i32 %r
}
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-ALL-LABEL: test_fptoui_i64:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-ALL: bl __aeabi_f2ulz
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define i64 @test_fptoui_i64(half* %p) #0 {
%a = load half, half* %p, align 2
%r = fptoui half %a to i64
ret i64 %r
}
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-ALL-LABEL: test_sitofp_i32:
; CHECK-VFP: vcvt.f32.s32
; CHECK-NOVFP: bl __aeabi_i2f
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f16.f32
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_sitofp_i32(i32 %a, half* %p) #0 {
%r = sitofp i32 %a to half
store half %r, half* %p
ret void
}
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-ALL-LABEL: test_uitofp_i32:
; CHECK-VFP: vcvt.f32.u32
; CHECK-NOVFP: bl __aeabi_ui2f
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f16.f32
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_uitofp_i32(i32 %a, half* %p) #0 {
%r = uitofp i32 %a to half
store half %r, half* %p
ret void
}
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-ALL-LABEL: test_sitofp_i64:
; CHECK-ALL: bl __aeabi_l2f
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f16.f32
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_sitofp_i64(i64 %a, half* %p) #0 {
%r = sitofp i64 %a to half
store half %r, half* %p
ret void
}
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-ALL-LABEL: test_uitofp_i64:
; CHECK-ALL: bl __aeabi_ul2f
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f16.f32
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_uitofp_i64(i64 %a, half* %p) #0 {
%r = uitofp i64 %a to half
store half %r, half* %p
ret void
}
; CHECK-FP16-LABEL: test_fptrunc_float:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f16.f32
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL-LABEL: test_fptrunc_float:
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_fptrunc_float(float %f, half* %p) #0 {
%a = fptrunc float %f to half
store half %a, half* %p
ret void
}
; CHECK-FP16-LABEL: test_fptrunc_double:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: bl __aeabi_d2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL-LABEL: test_fptrunc_double:
; CHECK-LIBCALL: bl __aeabi_d2h
define void @test_fptrunc_double(double %d, half* %p) #0 {
%a = fptrunc double %d to half
store half %a, half* %p
ret void
}
; CHECK-FP16-LABEL: test_fpextend_float:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL-LABEL: test_fpextend_float:
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define float @test_fpextend_float(half* %p) {
%a = load half, half* %p, align 2
%r = fpext half %a to float
ret float %r
}
; CHECK-FP16-LABEL: test_fpextend_double:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL-LABEL: test_fpextend_double:
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-VFP: vcvt.f64.f32
; CHECK-NOVFP: bl __aeabi_f2d
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define double @test_fpextend_double(half* %p) {
%a = load half, half* %p, align 2
%r = fpext half %a to double
ret double %r
}
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-ALL-LABEL: test_bitcast_halftoi16:
; CHECK-ALL-NEXT: .fnstart
; CHECK-ALL-NEXT: ldrh r0, [r0]
; CHECK-ALL-NEXT: bx lr
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define i16 @test_bitcast_halftoi16(half* %p) #0 {
%a = load half, half* %p, align 2
%r = bitcast half %a to i16
ret i16 %r
}
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-ALL-LABEL: test_bitcast_i16tohalf:
; CHECK-ALL-NEXT: .fnstart
; CHECK-ALL-NEXT: strh r0, [r1]
; CHECK-ALL-NEXT: bx lr
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_bitcast_i16tohalf(i16 %a, half* %p) #0 {
%r = bitcast i16 %a to half
store half %r, half* %p
ret void
}
declare half @llvm.sqrt.f16(half %a) #0
declare half @llvm.powi.f16(half %a, i32 %b) #0
declare half @llvm.sin.f16(half %a) #0
declare half @llvm.cos.f16(half %a) #0
declare half @llvm.pow.f16(half %a, half %b) #0
declare half @llvm.exp.f16(half %a) #0
declare half @llvm.exp2.f16(half %a) #0
declare half @llvm.log.f16(half %a) #0
declare half @llvm.log10.f16(half %a) #0
declare half @llvm.log2.f16(half %a) #0
declare half @llvm.fma.f16(half %a, half %b, half %c) #0
declare half @llvm.fabs.f16(half %a) #0
declare half @llvm.minnum.f16(half %a, half %b) #0
declare half @llvm.maxnum.f16(half %a, half %b) #0
declare half @llvm.copysign.f16(half %a, half %b) #0
declare half @llvm.floor.f16(half %a) #0
declare half @llvm.ceil.f16(half %a) #0
declare half @llvm.trunc.f16(half %a) #0
declare half @llvm.rint.f16(half %a) #0
declare half @llvm.nearbyint.f16(half %a) #0
declare half @llvm.round.f16(half %a) #0
declare half @llvm.fmuladd.f16(half %a, half %b, half %c) #0
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-ALL-LABEL: test_sqrt:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: vsqrt.f32
; CHECK-FP16: vcvtb.f16.f32
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-VFP-LIBCALL: vsqrt.f32
; CHECK-NOVFP: bl sqrtf
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_sqrt(half* %p) #0 {
%a = load half, half* %p, align 2
%r = call half @llvm.sqrt.f16(half %a)
store half %r, half* %p
ret void
}
; CHECK-FP16-LABEL: test_fpowi:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: bl __powisf2
; CHECK-FP16: vcvtb.f16.f32
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL-LABEL: test_fpowi:
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL: bl __powisf2
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_fpowi(half* %p, i32 %b) #0 {
%a = load half, half* %p, align 2
%r = call half @llvm.powi.f16(half %a, i32 %b)
store half %r, half* %p
ret void
}
; CHECK-FP16-LABEL: test_sin:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: bl sinf
; CHECK-FP16: vcvtb.f16.f32
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL-LABEL: test_sin:
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL: bl sinf
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_sin(half* %p) #0 {
%a = load half, half* %p, align 2
%r = call half @llvm.sin.f16(half %a)
store half %r, half* %p
ret void
}
; CHECK-FP16-LABEL: test_cos:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: bl cosf
; CHECK-FP16: vcvtb.f16.f32
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL-LABEL: test_cos:
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL: bl cosf
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_cos(half* %p) #0 {
%a = load half, half* %p, align 2
%r = call half @llvm.cos.f16(half %a)
store half %r, half* %p
ret void
}
; CHECK-FP16-LABEL: test_pow:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: bl powf
; CHECK-FP16: vcvtb.f16.f32
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL-LABEL: test_pow:
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
; CHECK-LIBCALL: bl __aeabi_h2f
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL: bl powf
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_pow(half* %p, half* %q) #0 {
%a = load half, half* %p, align 2
%b = load half, half* %q, align 2
%r = call half @llvm.pow.f16(half %a, half %b)
store half %r, half* %p
ret void
}
; CHECK-FP16-LABEL: test_exp:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: bl expf
; CHECK-FP16: vcvtb.f16.f32
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL-LABEL: test_exp:
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL: bl expf
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_exp(half* %p) #0 {
%a = load half, half* %p, align 2
%r = call half @llvm.exp.f16(half %a)
store half %r, half* %p
ret void
}
; CHECK-FP16-LABEL: test_exp2:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: bl exp2f
; CHECK-FP16: vcvtb.f16.f32
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL-LABEL: test_exp2:
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL: bl exp2f
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_exp2(half* %p) #0 {
%a = load half, half* %p, align 2
%r = call half @llvm.exp2.f16(half %a)
store half %r, half* %p
ret void
}
; CHECK-FP16-LABEL: test_log:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: bl logf
; CHECK-FP16: vcvtb.f16.f32
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL-LABEL: test_log:
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL: bl logf
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_log(half* %p) #0 {
%a = load half, half* %p, align 2
%r = call half @llvm.log.f16(half %a)
store half %r, half* %p
ret void
}
; CHECK-FP16-LABEL: test_log10:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: bl log10f
; CHECK-FP16: vcvtb.f16.f32
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL-LABEL: test_log10:
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL: bl log10f
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_log10(half* %p) #0 {
%a = load half, half* %p, align 2
%r = call half @llvm.log10.f16(half %a)
store half %r, half* %p
ret void
}
; CHECK-FP16-LABEL: test_log2:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: bl log2f
; CHECK-FP16: vcvtb.f16.f32
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL-LABEL: test_log2:
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL: bl log2f
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_log2(half* %p) #0 {
%a = load half, half* %p, align 2
%r = call half @llvm.log2.f16(half %a)
store half %r, half* %p
ret void
}
; CHECK-FP16-LABEL: test_fma:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: bl fmaf
; CHECK-FP16: vcvtb.f16.f32
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL-LABEL: test_fma:
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
; CHECK-LIBCALL: bl __aeabi_h2f
; CHECK-LIBCALL: bl __aeabi_h2f
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL: bl fmaf
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_fma(half* %p, half* %q, half* %r) #0 {
%a = load half, half* %p, align 2
%b = load half, half* %q, align 2
%c = load half, half* %r, align 2
%v = call half @llvm.fma.f16(half %a, half %b, half %c)
store half %v, half* %p
ret void
}
; CHECK-FP16-LABEL: test_fabs:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: vabs.f32
; CHECK-FP16: vcvtb.f16.f32
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL-LABEL: test_fabs:
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
[ARM] Prefer BIC over BFC in ARM mode. BIC is generally faster, and it can put the output in a different register from the input. We already do this in Thumb2 mode; not sure why the equivalent fix never got applied to ARM mode. Differential Revision: https://reviews.llvm.org/D31797 llvm-svn: 299803
2017-04-08 06:01:23 +08:00
; CHECK-LIBCALL: bic
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_fabs(half* %p) {
%a = load half, half* %p, align 2
%r = call half @llvm.fabs.f16(half %a)
store half %r, half* %p
ret void
}
; CHECK-FP16-LABEL: test_minnum:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: bl fminf
; CHECK-FP16: vcvtb.f16.f32
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL-LABEL: test_minnum:
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
; CHECK-LIBCALL: bl __aeabi_h2f
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL: bl fminf
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_minnum(half* %p, half* %q) #0 {
%a = load half, half* %p, align 2
%b = load half, half* %q, align 2
%r = call half @llvm.minnum.f16(half %a, half %b)
store half %r, half* %p
ret void
}
; CHECK-FP16-LABEL: test_maxnum:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: bl fmaxf
; CHECK-FP16: vcvtb.f16.f32
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL-LABEL: test_maxnum:
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
; CHECK-LIBCALL: bl __aeabi_h2f
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL: bl fmaxf
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_maxnum(half* %p, half* %q) #0 {
%a = load half, half* %p, align 2
%b = load half, half* %q, align 2
%r = call half @llvm.maxnum.f16(half %a, half %b)
store half %r, half* %p
ret void
}
[NFC] Rename minnan and maxnan to minimum and maximum Summary: Changes all uses of minnan/maxnan to minimum/maximum globally. These names emphasize that the semantic difference between these operations is more than just NaN-propagation. Reviewers: arsenm, aheejin, dschuff, javed.absar Subscribers: jholewinski, sdardis, wdng, sbc100, jgravelle-google, jrtc27, atanasyan, llvm-commits Differential Revision: https://reviews.llvm.org/D53112 llvm-svn: 345218
2018-10-25 06:49:55 +08:00
; CHECK-ALL-LABEL: test_minimum:
DAG: Constant fold fp16_to_fp/fp16_to_fp This fixes emitting conversions of constants on targets without legal f16 that need to use these for legalization. llvm-svn: 293499
2017-01-31 00:57:41 +08:00
; CHECK-FP16: vmov.f32 s0, #1.000000e+00
[CodeGen] Promote FMINNAN/FMAXNAN like other binops. We think it's OK to generate half fminnan because it's legal for the transform-to type (f32; r245196). However, PromoteFloatRes was missing the case; simply promote like the other binops, including minnum. llvm-svn: 271317
2016-06-01 02:50:25 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-LIBCALL: bl __aeabi_h2f
DAG: Constant fold fp16_to_fp/fp16_to_fp This fixes emitting conversions of constants on targets without legal f16 that need to use these for legalization. llvm-svn: 293499
2017-01-31 00:57:41 +08:00
; CHECK-LIBCALL-VFP: vmov.f32 s{{[0-9]+}}, #1.000000e+00
; CHECK-NOVFP: mov r{{[0-9]+}}, #1065353216
[CodeGen] Promote FMINNAN/FMAXNAN like other binops. We think it's OK to generate half fminnan because it's legal for the transform-to type (f32; r245196). However, PromoteFloatRes was missing the case; simply promote like the other binops, including minnum. llvm-svn: 271317
2016-06-01 02:50:25 +08:00
; CHECK-VFP: vmin.f32
; CHECK-NOVFP: bl __aeabi_fcmpge
; CHECK-FP16: vcvtb.f16.f32
; CHECK-LIBCALL: bl __aeabi_f2h
[NFC] Rename minnan and maxnan to minimum and maximum Summary: Changes all uses of minnan/maxnan to minimum/maximum globally. These names emphasize that the semantic difference between these operations is more than just NaN-propagation. Reviewers: arsenm, aheejin, dschuff, javed.absar Subscribers: jholewinski, sdardis, wdng, sbc100, jgravelle-google, jrtc27, atanasyan, llvm-commits Differential Revision: https://reviews.llvm.org/D53112 llvm-svn: 345218
2018-10-25 06:49:55 +08:00
define void @test_minimum(half* %p) #0 {
[CodeGen] Promote FMINNAN/FMAXNAN like other binops. We think it's OK to generate half fminnan because it's legal for the transform-to type (f32; r245196). However, PromoteFloatRes was missing the case; simply promote like the other binops, including minnum. llvm-svn: 271317
2016-06-01 02:50:25 +08:00
%a = load half, half* %p, align 2
%c = fcmp ult half %a, 1.0
%r = select i1 %c, half %a, half 1.0
store half %r, half* %p
ret void
}
[NFC] Rename minnan and maxnan to minimum and maximum Summary: Changes all uses of minnan/maxnan to minimum/maximum globally. These names emphasize that the semantic difference between these operations is more than just NaN-propagation. Reviewers: arsenm, aheejin, dschuff, javed.absar Subscribers: jholewinski, sdardis, wdng, sbc100, jgravelle-google, jrtc27, atanasyan, llvm-commits Differential Revision: https://reviews.llvm.org/D53112 llvm-svn: 345218
2018-10-25 06:49:55 +08:00
; CHECK-ALL-LABEL: test_maximum:
DAG: Constant fold fp16_to_fp/fp16_to_fp This fixes emitting conversions of constants on targets without legal f16 that need to use these for legalization. llvm-svn: 293499
2017-01-31 00:57:41 +08:00
; CHECK-FP16: vmov.f32 s0, #1.000000e+00
[CodeGen] Promote FMINNAN/FMAXNAN like other binops. We think it's OK to generate half fminnan because it's legal for the transform-to type (f32; r245196). However, PromoteFloatRes was missing the case; simply promote like the other binops, including minnum. llvm-svn: 271317
2016-06-01 02:50:25 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-LIBCALL: bl __aeabi_h2f
DAG: Constant fold fp16_to_fp/fp16_to_fp This fixes emitting conversions of constants on targets without legal f16 that need to use these for legalization. llvm-svn: 293499
2017-01-31 00:57:41 +08:00
; CHECK-LIBCALL-VFP: vmov.f32 s0, #1.000000e+00
; CHECK-NOVFP: mov r{{[0-9]+}}, #1065353216
[CodeGen] Promote FMINNAN/FMAXNAN like other binops. We think it's OK to generate half fminnan because it's legal for the transform-to type (f32; r245196). However, PromoteFloatRes was missing the case; simply promote like the other binops, including minnum. llvm-svn: 271317
2016-06-01 02:50:25 +08:00
; CHECK-VFP: vmax.f32
; CHECK-NOVFP: bl __aeabi_fcmple
; CHECK-FP16: vcvtb.f16.f32
; CHECK-LIBCALL: bl __aeabi_f2h
[NFC] Rename minnan and maxnan to minimum and maximum Summary: Changes all uses of minnan/maxnan to minimum/maximum globally. These names emphasize that the semantic difference between these operations is more than just NaN-propagation. Reviewers: arsenm, aheejin, dschuff, javed.absar Subscribers: jholewinski, sdardis, wdng, sbc100, jgravelle-google, jrtc27, atanasyan, llvm-commits Differential Revision: https://reviews.llvm.org/D53112 llvm-svn: 345218
2018-10-25 06:49:55 +08:00
define void @test_maximum(half* %p) #0 {
[CodeGen] Promote FMINNAN/FMAXNAN like other binops. We think it's OK to generate half fminnan because it's legal for the transform-to type (f32; r245196). However, PromoteFloatRes was missing the case; simply promote like the other binops, including minnum. llvm-svn: 271317
2016-06-01 02:50:25 +08:00
%a = load half, half* %p, align 2
%c = fcmp ugt half %a, 1.0
%r = select i1 %c, half %a, half 1.0
store half %r, half* %p
ret void
}
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-FP16-LABEL: test_copysign:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: vbsl
; CHECK-FP16: vcvtb.f16.f32
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL-LABEL: test_copysign:
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
; CHECK-LIBCALL: bl __aeabi_h2f
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-VFP-LIBCALL: vbsl
; CHECK-NOVFP: and
[ARM] Make -mcpu=generic schedule for an in-order core (Cortex-A8). The benchmarking summarized in http://lists.llvm.org/pipermail/llvm-dev/2017-May/113525.html showed this is beneficial for a wide range of cores. As is to be expected, quite a few small adaptations are needed to the regressions tests, as the difference in scheduling results in: - Quite a few small instruction schedule differences. - A few changes in register allocation decisions caused by different instruction schedules. - A few changes in IfConversion decisions, due to a difference in instruction schedule and/or the estimated cost of a branch mispredict. llvm-svn: 306514
2017-06-28 15:07:03 +08:00
; CHECK-NOVFP: bic
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-NOVFP: orr
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_copysign(half* %p, half* %q) #0 {
%a = load half, half* %p, align 2
%b = load half, half* %q, align 2
%r = call half @llvm.copysign.f16(half %a, half %b)
store half %r, half* %p
ret void
}
; CHECK-FP16-LABEL: test_floor:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: bl floorf
; CHECK-FP16: vcvtb.f16.f32
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL-LABEL: test_floor:
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL: bl floorf
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_floor(half* %p) {
%a = load half, half* %p, align 2
%r = call half @llvm.floor.f16(half %a)
store half %r, half* %p
ret void
}
; CHECK-FP16-LABEL: test_ceil:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: bl ceilf
; CHECK-FP16: vcvtb.f16.f32
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL-LABEL: test_ceil:
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL: bl ceilf
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_ceil(half* %p) {
%a = load half, half* %p, align 2
%r = call half @llvm.ceil.f16(half %a)
store half %r, half* %p
ret void
}
; CHECK-FP16-LABEL: test_trunc:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: bl truncf
; CHECK-FP16: vcvtb.f16.f32
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL-LABEL: test_trunc:
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL: bl truncf
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_trunc(half* %p) {
%a = load half, half* %p, align 2
%r = call half @llvm.trunc.f16(half %a)
store half %r, half* %p
ret void
}
; CHECK-FP16-LABEL: test_rint:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: bl rintf
; CHECK-FP16: vcvtb.f16.f32
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL-LABEL: test_rint:
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL: bl rintf
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_rint(half* %p) {
%a = load half, half* %p, align 2
%r = call half @llvm.rint.f16(half %a)
store half %r, half* %p
ret void
}
; CHECK-FP16-LABEL: test_nearbyint:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: bl nearbyintf
; CHECK-FP16: vcvtb.f16.f32
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL-LABEL: test_nearbyint:
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL: bl nearbyintf
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_nearbyint(half* %p) {
%a = load half, half* %p, align 2
%r = call half @llvm.nearbyint.f16(half %a)
store half %r, half* %p
ret void
}
; CHECK-FP16-LABEL: test_round:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: bl roundf
; CHECK-FP16: vcvtb.f16.f32
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL-LABEL: test_round:
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL: bl roundf
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_round(half* %p) {
%a = load half, half* %p, align 2
%r = call half @llvm.round.f16(half %a)
store half %r, half* %p
ret void
}
; CHECK-FP16-LABEL: test_fmuladd:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: vcvtb.f32.f16
; CHECK-FP16: vmla.f32
; CHECK-FP16: vcvtb.f16.f32
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-LIBCALL-LABEL: test_fmuladd:
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_h2f
; CHECK-LIBCALL: bl __aeabi_h2f
; CHECK-LIBCALL: bl __aeabi_h2f
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-VFP-LIBCALL: vmla.f32
; CHECK-NOVFP: bl __aeabi_fmul
Fix incorrect target triple in fp16-promote.ll Summary: Hyphens were missing from the triple, causing it to be parsed incorrectly. This patch updates the triple and makes necessary changes to the expected output. Patch is from Vinicius Tinti. Reviewers: ab, tinti Subscribers: srhines, llvm-commits Differential Revision: http://reviews.llvm.org/D13792 llvm-svn: 251020
2015-10-22 22:15:00 +08:00
; CHECK-LIBCALL: bl __aeabi_f2h
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_fmuladd(half* %p, half* %q, half* %r) #0 {
%a = load half, half* %p, align 2
%b = load half, half* %q, align 2
%c = load half, half* %r, align 2
%v = call half @llvm.fmuladd.f16(half %a, half %b, half %c)
store half %v, half* %p
ret void
}
; f16 vectors are not legal in the backend. Vector elements are not assigned
; to the register, but are stored in the stack instead. Hence insertelement
; and extractelement have these extra loads and stores.
; CHECK-ALL-LABEL: test_insertelement:
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-ALL: sub sp, sp, #8
[ARM][AArch64][DAG] Reenable post-legalize store merge Reenable post-legalize stores with constant merging computation and corresponding test case. * Properly truncate store merge constants * Disable merging of truncated stores floating points * Ensure merges of constant stores into a single vector are constructed from legal elements. Reviewers: eastig, efriedma Reviewed By: eastig Subscribers: spatel, rengolin, aemerson, javed.absar, kristof.beyls, hiraditya, llvm-commits Differential Revision: https://reviews.llvm.org/D40701 llvm-svn: 319899
2017-12-06 23:30:13 +08:00
; CHECK-VFP: and
; CHECK-VFP: mov
; CHECK-VFP: ldrd
; CHECK-VFP: orr
; CHECK-VFP: ldrh
; CHECK-VFP: stm
; CHECK-VFP: strh
; CHECK-VFP: ldm
; CHECK-VFP: stm
; CHECK-NOVFP: ldrh
; CHECK-NOVFP: ldrh
; CHECK-NOVFP: ldrh
; CHECK-NOVFP: ldrh
; CHECK-NOVFP-DAG: strh
; CHECK-NOVFP-DAG: strh
; CHECK-NOVFP-DAG: mov
; CHECK-NOVFP-DAG: ldrh
; CHECK-NOVFP-DAG: orr
; CHECK-NOVFP-DAG: strh
; CHECK-NOVFP-DAG: strh
; CHECK-NOVFP-DAG: strh
; CHECK-NOVFP-DAG: ldrh
; CHECK-NOVFP-DAG: ldrh
; CHECK-NOVFP-DAG: ldrh
; CHECK-NOVFP-DAG: strh
; CHECK-NOVFP-DAG: strh
; CHECK-NOVFP-DAG: strh
; CHECK-NOVFP-DAG: strh
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-ALL: add sp, sp, #8
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_insertelement(half* %p, <4 x half>* %q, i32 %i) #0 {
%a = load half, half* %p, align 2
%b = load <4 x half>, <4 x half>* %q, align 8
%c = insertelement <4 x half> %b, half %a, i32 %i
store <4 x half> %c, <4 x half>* %q
ret void
}
; CHECK-ALL-LABEL: test_extractelement:
[DAGCombiner] Support {a|s}ext, {a|z|s}ext load nodes in load combine Resubmit r295336 after the bug with non-zero offset patterns on BE targets is fixed (r296336). Support {a|s}ext, {a|z|s}ext load nodes as a part of load combine patters. Reviewed By: filcab Differential Revision: https://reviews.llvm.org/D29591 llvm-svn: 296651
2017-03-02 02:12:29 +08:00
; CHECK-VFP: push {{{.*}}, lr}
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-VFP: sub sp, sp, #8
[DAGCombiner] Support {a|s}ext, {a|z|s}ext load nodes in load combine Resubmit r295336 after the bug with non-zero offset patterns on BE targets is fixed (r296336). Support {a|s}ext, {a|z|s}ext load nodes as a part of load combine patters. Reviewed By: filcab Differential Revision: https://reviews.llvm.org/D29591 llvm-svn: 296651
2017-03-02 02:12:29 +08:00
; CHECK-VFP: ldrd
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-VFP: mov
DAG: Avoid OOB when legalizing vector indexing If a vector index is out of bounds, the result is supposed to be undefined but is not undefined behavior. Change the legalization for indexing the vector on the stack so that an out of bounds index does not create an out of bounds memory access. llvm-svn: 291604
2017-01-11 06:02:30 +08:00
; CHECK-VFP: orr
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-VFP: ldrh
; CHECK-VFP: strh
; CHECK-VFP: add sp, sp, #8
[DAGCombiner] Support {a|s}ext, {a|z|s}ext load nodes in load combine Resubmit r295336 after the bug with non-zero offset patterns on BE targets is fixed (r296336). Support {a|s}ext, {a|z|s}ext load nodes as a part of load combine patters. Reviewed By: filcab Differential Revision: https://reviews.llvm.org/D29591 llvm-svn: 296651
2017-03-02 02:12:29 +08:00
; CHECK-VFP: pop {{{.*}}, pc}
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-NOVFP: ldrh
; CHECK-NOVFP: strh
; CHECK-NOVFP: ldrh
; CHECK-NOVFP: strh
; CHECK-NOVFP: ldrh
; CHECK-NOVFP: strh
; CHECK-NOVFP: ldrh
; CHECK-NOVFP: strh
; CHECK-NOVFP: ldrh
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_extractelement(half* %p, <4 x half>* %q, i32 %i) #0 {
%a = load <4 x half>, <4 x half>* %q, align 8
%b = extractelement <4 x half> %a, i32 %i
store half %b, half* %p
ret void
}
; test struct operations
%struct.dummy = type { i32, half }
; CHECK-ALL-LABEL: test_insertvalue:
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-ALL-DAG: ldr
; CHECK-ALL-DAG: ldrh
; CHECK-ALL-DAG: strh
; CHECK-ALL-DAG: str
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_insertvalue(%struct.dummy* %p, half* %q) {
%a = load %struct.dummy, %struct.dummy* %p
%b = load half, half* %q
%c = insertvalue %struct.dummy %a, half %b, 1
store %struct.dummy %c, %struct.dummy* %p
ret void
}
; CHECK-ALL-LABEL: test_extractvalue:
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-ALL: .fnstart
; CHECK-ALL: ldrh
; CHECK-ALL: strh
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define void @test_extractvalue(%struct.dummy* %p, half* %q) {
%a = load %struct.dummy, %struct.dummy* %p
%b = extractvalue %struct.dummy %a, 1
store half %b, half* %q
ret void
}
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-ALL-LABEL: test_struct_return:
Fix flakiness in fp16-promote.ll Summary: In the f16-promote test, make the checks for native conversion instructions similar to the libcall checks: - Remove hard coded register names - Do not check exact instruction sequences. This fixes test flakiness due to non-determinism in instruction scheduling and register allocation. I also fixed a few minor things in the CHECK-LIBCALL checks. I'll try to find a way to check that unnecessary loads, stores, or conversions don't happen. Reviewers: mzolotukhin, srhines, ab Subscribers: llvm-commits Differential Revision: http://reviews.llvm.org/D9112 llvm-svn: 235363
2015-04-21 07:54:41 +08:00
; CHECK-FP16: vcvtb.f32.f16
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-VFP-LIBCALL: bl __aeabi_h2f
; CHECK-NOVFP-DAG: ldr
; CHECK-NOVFP-DAG: ldrh
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
define %struct.dummy @test_struct_return(%struct.dummy* %p) {
%a = load %struct.dummy, %struct.dummy* %p
ret %struct.dummy %a
}
; CHECK-ALL-LABEL: test_struct_arg:
; CHECK-ALL-NEXT: .fnstart
[CodeGen] Always promote f16 if not legal We don't currently have any runtime library functions for operations on f16 values (other than conversions to and from f32 and f64), so we should always promote it to f32, even if that is not a legal type. In that case, the f32 values would be softened to f32 library calls. SoftenFloatRes_FP_EXTEND now needs to check the promoted operand's type, as it may ne a no-op or require a different library call. getCopyFromParts and getCopyToParts now need to cope with a floating-point value stored in a larger integer part, as is the case for any target that needs to store an f16 value in a 32-bit integer register. Differential Revision: http://reviews.llvm.org/D12856 llvm-svn: 252459
2015-11-09 19:03:18 +08:00
; CHECK-NOVFP-NEXT: mov r0, r1
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
; CHECK-ALL-NEXT: bx lr
define half @test_struct_arg(%struct.dummy %p) {
%a = extractvalue %struct.dummy %p, 1
ret half %a
}
[CodeGen] Round [SU]INT_TO_FP result when promoting from f16. If we don't, values that aren't precisely representable in f16 could be used as-is in a promoted f32 operation, which would produce incorrect results. AArch64 had the correct behavior; add a focused test. Fixes http://llvm.org/PR26871 llvm-svn: 268700
2016-05-06 08:58:00 +08:00
; CHECK-LABEL: test_uitofp_i32_fadd:
; CHECK-VFP-DAG: vcvt.f32.u32
; CHECK-NOVFP-DAG: bl __aeabi_ui2f
; CHECK-FP16-DAG: vcvtb.f16.f32
; CHECK-FP16-DAG: vcvtb.f32.f16
; CHECK-LIBCALL-DAG: bl __aeabi_h2f
; CHECK-LIBCALL-DAG: bl __aeabi_h2f
; CHECK-VFP-DAG: vadd.f32
; CHECK-NOVFP-DAG: bl __aeabi_fadd
; CHECK-FP16-DAG: vcvtb.f16.f32
; CHECK-LIBCALL-DAG: bl __aeabi_f2h
define half @test_uitofp_i32_fadd(i32 %a, half %b) #0 {
%c = uitofp i32 %a to half
%r = fadd half %b, %c
ret half %r
}
; CHECK-LABEL: test_sitofp_i32_fadd:
; CHECK-VFP-DAG: vcvt.f32.s32
; CHECK-NOVFP-DAG: bl __aeabi_i2f
; CHECK-FP16-DAG: vcvtb.f16.f32
; CHECK-FP16-DAG: vcvtb.f32.f16
; CHECK-LIBCALL-DAG: bl __aeabi_h2f
; CHECK-LIBCALL-DAG: bl __aeabi_h2f
; CHECK-VFP-DAG: vadd.f32
; CHECK-NOVFP-DAG: bl __aeabi_fadd
; CHECK-FP16-DAG: vcvtb.f16.f32
; CHECK-LIBCALL-DAG: bl __aeabi_f2h
define half @test_sitofp_i32_fadd(i32 %a, half %b) #0 {
%c = sitofp i32 %a to half
%r = fadd half %b, %c
ret half %r
}
Add support to promote f16 to f32 Summary: This patch adds legalization support to operate on FP16 as a load/store type and do operations on it as floats. Tests for ARM are added to test/CodeGen/ARM/fp16-promote.ll Reviewers: srhines, t.p.northover Differential Revision: http://reviews.llvm.org/D8755 llvm-svn: 235215
2015-04-18 02:36:25 +08:00
attributes #0 = { nounwind }