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/AArch64/arm64-movi.ll

203 lines
5.2 KiB
LLVM
Raw Normal View History

ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
; RUN: llc < %s -march=arm64 | FileCheck %s
;==--------------------------------------------------------------------------==
; Tests for MOV-immediate implemented with ORR-immediate.
;==--------------------------------------------------------------------------==
; 64-bit immed with 32-bit pattern size, rotated by 0.
define i64 @test64_32_rot0() nounwind {
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK-LABEL: test64_32_rot0:
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
; CHECK: orr x0, xzr, #0x700000007
ret i64 30064771079
}
; 64-bit immed with 32-bit pattern size, rotated by 2.
define i64 @test64_32_rot2() nounwind {
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK-LABEL: test64_32_rot2:
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
; CHECK: orr x0, xzr, #0xc0000003c0000003
ret i64 13835058071388291075
}
; 64-bit immed with 4-bit pattern size, rotated by 3.
define i64 @test64_4_rot3() nounwind {
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK-LABEL: test64_4_rot3:
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
; CHECK: orr x0, xzr, #0xeeeeeeeeeeeeeeee
ret i64 17216961135462248174
}
; 32-bit immed with 32-bit pattern size, rotated by 16.
define i32 @test32_32_rot16() nounwind {
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK-LABEL: test32_32_rot16:
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
; CHECK: orr w0, wzr, #0xff0000
ret i32 16711680
}
; 32-bit immed with 2-bit pattern size, rotated by 1.
define i32 @test32_2_rot1() nounwind {
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK-LABEL: test32_2_rot1:
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
; CHECK: orr w0, wzr, #0xaaaaaaaa
ret i32 2863311530
}
;==--------------------------------------------------------------------------==
; Tests for MOVZ with MOVK.
;==--------------------------------------------------------------------------==
define i32 @movz() nounwind {
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK-LABEL: movz:
; CHECK: movz w0, #0x5
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
ret i32 5
}
define i64 @movz_3movk() nounwind {
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK-LABEL: movz_3movk:
; CHECK: movz x0, #0x5, lsl #48
; CHECK-NEXT: movk x0, #0x1234, lsl #32
; CHECK-NEXT: movk x0, #0xabcd, lsl #16
; CHECK-NEXT: movk x0, #0x5678
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
ret i64 1427392313513592
}
define i64 @movz_movk_skip1() nounwind {
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK-LABEL: movz_movk_skip1:
; CHECK: movz x0, #0x5, lsl #32
; CHECK-NEXT: movk x0, #0x4321, lsl #16
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
ret i64 22601072640
}
define i64 @movz_skip1_movk() nounwind {
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK-LABEL: movz_skip1_movk:
; CHECK: movz x0, #0x8654, lsl #32
; CHECK-NEXT: movk x0, #0x1234
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
ret i64 147695335379508
}
;==--------------------------------------------------------------------------==
; Tests for MOVN with MOVK.
;==--------------------------------------------------------------------------==
define i64 @movn() nounwind {
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK-LABEL: movn:
; CHECK: movn x0, #0x29
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
ret i64 -42
}
define i64 @movn_skip1_movk() nounwind {
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK-LABEL: movn_skip1_movk:
; CHECK: movn x0, #0x29, lsl #32
; CHECK-NEXT: movk x0, #0x1234
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
ret i64 -176093720012
}
;==--------------------------------------------------------------------------==
; Tests for ORR with MOVK.
;==--------------------------------------------------------------------------==
; rdar://14987673
define i64 @orr_movk1() nounwind {
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK-LABEL: orr_movk1:
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
; CHECK: orr x0, xzr, #0xffff0000ffff0
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK: movk x0, #0xdead, lsl #16
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
ret i64 72056498262245120
}
define i64 @orr_movk2() nounwind {
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK-LABEL: orr_movk2:
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
; CHECK: orr x0, xzr, #0xffff0000ffff0
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK: movk x0, #0xdead, lsl #48
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
ret i64 -2400982650836746496
}
define i64 @orr_movk3() nounwind {
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK-LABEL: orr_movk3:
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
; CHECK: orr x0, xzr, #0xffff0000ffff0
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK: movk x0, #0xdead, lsl #32
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
ret i64 72020953688702720
}
define i64 @orr_movk4() nounwind {
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK-LABEL: orr_movk4:
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
; CHECK: orr x0, xzr, #0xffff0000ffff0
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK: movk x0, #0xdead
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
ret i64 72056494543068845
}
; rdar://14987618
define i64 @orr_movk5() nounwind {
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK-LABEL: orr_movk5:
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
; CHECK: orr x0, xzr, #0xff00ff00ff00ff00
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK: movk x0, #0xdead, lsl #16
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
ret i64 -71777214836900096
}
define i64 @orr_movk6() nounwind {
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK-LABEL: orr_movk6:
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
; CHECK: orr x0, xzr, #0xff00ff00ff00ff00
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK: movk x0, #0xdead, lsl #16
; CHECK: movk x0, #0xdead, lsl #48
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
ret i64 -2400982647117578496
}
define i64 @orr_movk7() nounwind {
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK-LABEL: orr_movk7:
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
; CHECK: orr x0, xzr, #0xff00ff00ff00ff00
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK: movk x0, #0xdead, lsl #48
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
ret i64 -2400982646575268096
}
define i64 @orr_movk8() nounwind {
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK-LABEL: orr_movk8:
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
; CHECK: orr x0, xzr, #0xff00ff00ff00ff00
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK: movk x0, #0xdead
; CHECK: movk x0, #0xdead, lsl #48
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
ret i64 -2400982646575276371
}
; rdar://14987715
define i64 @orr_movk9() nounwind {
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK-LABEL: orr_movk9:
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
; CHECK: orr x0, xzr, #0xffffff000000000
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK: movk x0, #0xff00
; CHECK: movk x0, #0xdead, lsl #16
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
ret i64 1152921439623315200
}
define i64 @orr_movk10() nounwind {
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK-LABEL: orr_movk10:
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
; CHECK: orr x0, xzr, #0xfffffffffffff00
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK: movk x0, #0xdead, lsl #16
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
ret i64 1152921504047824640
}
define i64 @orr_movk11() nounwind {
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK-LABEL: orr_movk11:
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
; CHECK: orr x0, xzr, #0xfff00000000000ff
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK: movk x0, #0xdead, lsl #16
; CHECK: movk x0, #0xffff, lsl #32
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
ret i64 -4222125209747201
}
define i64 @orr_movk12() nounwind {
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK-LABEL: orr_movk12:
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
; CHECK: orr x0, xzr, #0xfff00000000000ff
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK: movk x0, #0xdead, lsl #32
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
ret i64 -4258765016661761
}
define i64 @orr_movk13() nounwind {
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK-LABEL: orr_movk13:
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
; CHECK: orr x0, xzr, #0xfffff000000
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK: movk x0, #0xdead
; CHECK: movk x0, #0xdead, lsl #48
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
ret i64 -2401245434149282131
}
; rdar://13944082
define i64 @g() nounwind {
ARM64: use hex immediates for movz/movk instructions Since these are mostly used in "lsl #16", "lsl #32", "lsl #48" combinations to piece together an immediate in 16-bit chunks, hex is probably the most appropriate format. llvm-svn: 207635
2014-04-30 19:19:40 +08:00
; CHECK-LABEL: g:
; CHECK: movz x0, #0xffff, lsl #48
; CHECK: movk x0, #0x2
ARM64: initial backend import This adds a second implementation of the AArch64 architecture to LLVM, accessible in parallel via the "arm64" triple. The plan over the coming weeks & months is to merge the two into a single backend, during which time thorough code review should naturally occur. Everything will be easier with the target in-tree though, hence this commit. llvm-svn: 205090
2014-03-29 18:18:08 +08:00
entry:
ret i64 -281474976710654
}