maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
llvm-project/llvm/test/Transforms/InstCombine/redundant-left-shift-input-...

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

239 lines
9.4 KiB
LLVM
Raw Normal View History

[NFC][InstCombine] Tests for "fold variable mask before variable shift-of-trunc" (PR42563) https://bugs.llvm.org/show_bug.cgi?id=42563 llvm-svn: 375135
2019-10-18 01:20:12 +08:00
; NOTE: Assertions have been autogenerated by utils/update_test_checks.py
[InstCombine] use redirect of input file in regression tests; NFC This is a repeat of 1880092722 from 2009. We should have less risk of hitting bugs at this point because we auto-generate positive CHECK lines only, but this makes things consistent. Copying the original commit msg: "Change tests from "opt %s" to "opt < %s" so that opt doesn't see the input filename so that opt doesn't print the input filename in the output so that grep lines in the tests don't unintentionally match strings in the input filename."
2020-09-29 23:02:03 +08:00
; RUN: opt < %s -instcombine -S | FileCheck %s
[NFC][InstCombine] Tests for "fold variable mask before variable shift-of-trunc" (PR42563) https://bugs.llvm.org/show_bug.cgi?id=42563 llvm-svn: 375135
2019-10-18 01:20:12 +08:00
; If we have some pattern that leaves only some low bits set, and then performs
; left-shift of those bits, we can combine those two shifts into a shift+mask.
; There are many variants to this pattern:
; b) (trunc ((x & (~(-1 << maskNbits))))) << shiftNbits
; simplify to:
; ((trunc(x)) << shiftNbits) & (~(-1 << (maskNbits+shiftNbits)))
; Simple tests.
declare void @use32(i32)
declare void @use64(i64)
define i32 @t0_basic(i64 %x, i32 %nbits) {
; CHECK-LABEL: @t0_basic(
; CHECK-NEXT: [[T0:%.*]] = zext i32 [[NBITS:%.*]] to i64
; CHECK-NEXT: [[T1:%.*]] = shl i64 -1, [[T0]]
; CHECK-NEXT: [[T2:%.*]] = xor i64 [[T1]], -1
; CHECK-NEXT: [[T3:%.*]] = sub i32 32, [[NBITS]]
; CHECK-NEXT: [[T4:%.*]] = and i64 [[T2]], [[X:%.*]]
; CHECK-NEXT: call void @use32(i32 [[NBITS]])
; CHECK-NEXT: call void @use64(i64 [[T0]])
; CHECK-NEXT: call void @use64(i64 [[T1]])
; CHECK-NEXT: call void @use64(i64 [[T2]])
; CHECK-NEXT: call void @use32(i32 [[T3]])
; CHECK-NEXT: call void @use64(i64 [[T4]])
[InstCombine] dropRedundantMaskingOfLeftShiftInput(): truncation (PR42563) Summary: That fold keeps growing and growing :( I think this may be one of the last pieces for it. Since D67677/D67725, the fold knowns the general form of the pattern - where some masking is needed: https://rise4fun.com/Alive/F5R https://rise4fun.com/Alive/gslRa But there is one more huge piece missing - if you are extracting some bits, it is not impossible that the origin is wider than the extraction, i.e. there may be a truncation. And we don't deal with that yet. But we can, and the generalization remains fully identical: https://rise4fun.com/Alive/Uar https://rise4fun.com/Alive/5SW After a preparatory cleanup i think the diff looks rather clean. One missing piece is that in some patterns (especially pat. b), `-1` only needs to be `-1` in final type, but that is for later.. https://bugs.llvm.org/show_bug.cgi?id=42563 Reviewers: spatel, nikic Reviewed By: spatel Subscribers: hiraditya, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D69125
2019-11-05 16:49:12 +08:00
; CHECK-NEXT: [[TMP1:%.*]] = trunc i64 [[X]] to i32
; CHECK-NEXT: [[T6:%.*]] = shl i32 [[TMP1]], [[T3]]
[NFC][InstCombine] Tests for "fold variable mask before variable shift-of-trunc" (PR42563) https://bugs.llvm.org/show_bug.cgi?id=42563 llvm-svn: 375135
2019-10-18 01:20:12 +08:00
; CHECK-NEXT: ret i32 [[T6]]
;
%t0 = zext i32 %nbits to i64
%t1 = shl i64 -1, %t0
%t2 = xor i64 %t1, -1
%t3 = sub i32 32, %nbits
%t4 = and i64 %t2, %x
call void @use32(i32 %nbits)
call void @use64(i64 %t0)
call void @use64(i64 %t1)
call void @use64(i64 %t2)
call void @use32(i32 %t3)
call void @use64(i64 %t4)
%t5 = trunc i64 %t4 to i32
%t6 = shl i32 %t5, %t3
ret i32 %t6
}
; Vectors
declare void @use8xi32(<8 x i32>)
declare void @use8xi64(<8 x i64>)
define <8 x i32> @t1_vec_splat(<8 x i64> %x, <8 x i32> %nbits) {
; CHECK-LABEL: @t1_vec_splat(
; CHECK-NEXT: [[T0:%.*]] = zext <8 x i32> [[NBITS:%.*]] to <8 x i64>
; CHECK-NEXT: [[T1:%.*]] = shl <8 x i64> <i64 -1, i64 -1, i64 -1, i64 -1, i64 -1, i64 -1, i64 -1, i64 -1>, [[T0]]
; CHECK-NEXT: [[T2:%.*]] = xor <8 x i64> [[T1]], <i64 -1, i64 -1, i64 -1, i64 -1, i64 -1, i64 -1, i64 -1, i64 -1>
; CHECK-NEXT: [[T3:%.*]] = sub <8 x i32> <i32 32, i32 32, i32 32, i32 32, i32 32, i32 32, i32 32, i32 32>, [[NBITS]]
; CHECK-NEXT: [[T4:%.*]] = and <8 x i64> [[T2]], [[X:%.*]]
; CHECK-NEXT: call void @use8xi32(<8 x i32> [[NBITS]])
; CHECK-NEXT: call void @use8xi64(<8 x i64> [[T0]])
; CHECK-NEXT: call void @use8xi64(<8 x i64> [[T1]])
; CHECK-NEXT: call void @use8xi64(<8 x i64> [[T2]])
; CHECK-NEXT: call void @use8xi32(<8 x i32> [[T3]])
; CHECK-NEXT: call void @use8xi64(<8 x i64> [[T4]])
[InstCombine] dropRedundantMaskingOfLeftShiftInput(): truncation (PR42563) Summary: That fold keeps growing and growing :( I think this may be one of the last pieces for it. Since D67677/D67725, the fold knowns the general form of the pattern - where some masking is needed: https://rise4fun.com/Alive/F5R https://rise4fun.com/Alive/gslRa But there is one more huge piece missing - if you are extracting some bits, it is not impossible that the origin is wider than the extraction, i.e. there may be a truncation. And we don't deal with that yet. But we can, and the generalization remains fully identical: https://rise4fun.com/Alive/Uar https://rise4fun.com/Alive/5SW After a preparatory cleanup i think the diff looks rather clean. One missing piece is that in some patterns (especially pat. b), `-1` only needs to be `-1` in final type, but that is for later.. https://bugs.llvm.org/show_bug.cgi?id=42563 Reviewers: spatel, nikic Reviewed By: spatel Subscribers: hiraditya, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D69125
2019-11-05 16:49:12 +08:00
; CHECK-NEXT: [[TMP1:%.*]] = trunc <8 x i64> [[X]] to <8 x i32>
; CHECK-NEXT: [[T6:%.*]] = shl <8 x i32> [[TMP1]], [[T3]]
[NFC][InstCombine] Tests for "fold variable mask before variable shift-of-trunc" (PR42563) https://bugs.llvm.org/show_bug.cgi?id=42563 llvm-svn: 375135
2019-10-18 01:20:12 +08:00
; CHECK-NEXT: ret <8 x i32> [[T6]]
;
%t0 = zext <8 x i32> %nbits to <8 x i64>
%t1 = shl <8 x i64> <i64 -1, i64 -1, i64 -1, i64 -1, i64 -1, i64 -1, i64 -1, i64 -1>, %t0
%t2 = xor <8 x i64> %t1, <i64 -1, i64 -1, i64 -1, i64 -1, i64 -1, i64 -1, i64 -1, i64 -1>
%t3 = sub <8 x i32> <i32 32, i32 32, i32 32, i32 32, i32 32, i32 32, i32 32, i32 32>, %nbits
%t4 = and <8 x i64> %t2, %x
call void @use8xi32(<8 x i32> %nbits)
call void @use8xi64(<8 x i64> %t0)
call void @use8xi64(<8 x i64> %t1)
call void @use8xi64(<8 x i64> %t2)
call void @use8xi32(<8 x i32> %t3)
call void @use8xi64(<8 x i64> %t4)
%t5 = trunc <8 x i64> %t4 to <8 x i32>
%t6 = shl <8 x i32> %t5, %t3
ret <8 x i32> %t6
}
define <8 x i32> @t2_vec_splat_undef(<8 x i64> %x, <8 x i32> %nbits) {
; CHECK-LABEL: @t2_vec_splat_undef(
; CHECK-NEXT: [[T0:%.*]] = zext <8 x i32> [[NBITS:%.*]] to <8 x i64>
; CHECK-NEXT: [[T1:%.*]] = shl <8 x i64> <i64 -1, i64 -1, i64 -1, i64 -1, i64 -1, i64 -1, i64 undef, i64 -1>, [[T0]]
; CHECK-NEXT: [[T2:%.*]] = xor <8 x i64> [[T1]], <i64 -1, i64 -1, i64 -1, i64 -1, i64 -1, i64 -1, i64 undef, i64 -1>
; CHECK-NEXT: [[T3:%.*]] = sub <8 x i32> <i32 32, i32 32, i32 32, i32 32, i32 32, i32 32, i32 undef, i32 32>, [[NBITS]]
; CHECK-NEXT: [[T4:%.*]] = and <8 x i64> [[T2]], [[X:%.*]]
; CHECK-NEXT: call void @use8xi32(<8 x i32> [[NBITS]])
; CHECK-NEXT: call void @use8xi64(<8 x i64> [[T0]])
; CHECK-NEXT: call void @use8xi64(<8 x i64> [[T1]])
; CHECK-NEXT: call void @use8xi64(<8 x i64> [[T2]])
; CHECK-NEXT: call void @use8xi32(<8 x i32> [[T3]])
; CHECK-NEXT: call void @use8xi64(<8 x i64> [[T4]])
[InstCombine] dropRedundantMaskingOfLeftShiftInput(): truncation (PR42563) Summary: That fold keeps growing and growing :( I think this may be one of the last pieces for it. Since D67677/D67725, the fold knowns the general form of the pattern - where some masking is needed: https://rise4fun.com/Alive/F5R https://rise4fun.com/Alive/gslRa But there is one more huge piece missing - if you are extracting some bits, it is not impossible that the origin is wider than the extraction, i.e. there may be a truncation. And we don't deal with that yet. But we can, and the generalization remains fully identical: https://rise4fun.com/Alive/Uar https://rise4fun.com/Alive/5SW After a preparatory cleanup i think the diff looks rather clean. One missing piece is that in some patterns (especially pat. b), `-1` only needs to be `-1` in final type, but that is for later.. https://bugs.llvm.org/show_bug.cgi?id=42563 Reviewers: spatel, nikic Reviewed By: spatel Subscribers: hiraditya, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D69125
2019-11-05 16:49:12 +08:00
; CHECK-NEXT: [[TMP1:%.*]] = trunc <8 x i64> [[X]] to <8 x i32>
; CHECK-NEXT: [[T6:%.*]] = shl <8 x i32> [[TMP1]], [[T3]]
[NFC][InstCombine] Tests for "fold variable mask before variable shift-of-trunc" (PR42563) https://bugs.llvm.org/show_bug.cgi?id=42563 llvm-svn: 375135
2019-10-18 01:20:12 +08:00
; CHECK-NEXT: ret <8 x i32> [[T6]]
;
%t0 = zext <8 x i32> %nbits to <8 x i64>
%t1 = shl <8 x i64> <i64 -1, i64 -1, i64 -1, i64 -1, i64 -1, i64 -1, i64 undef, i64 -1>, %t0
%t2 = xor <8 x i64> %t1, <i64 -1, i64 -1, i64 -1, i64 -1, i64 -1, i64 -1, i64 undef, i64 -1>
%t3 = sub <8 x i32> <i32 32, i32 32, i32 32, i32 32, i32 32, i32 32, i32 undef, i32 32>, %nbits
%t4 = and <8 x i64> %t2, %x
call void @use8xi32(<8 x i32> %nbits)
call void @use8xi64(<8 x i64> %t0)
call void @use8xi64(<8 x i64> %t1)
call void @use8xi64(<8 x i64> %t2)
call void @use8xi32(<8 x i32> %t3)
call void @use8xi64(<8 x i64> %t4)
%t5 = trunc <8 x i64> %t4 to <8 x i32>
%t6 = shl <8 x i32> %t5, %t3
ret <8 x i32> %t6
}
define <8 x i32> @t3_vec_nonsplat(<8 x i64> %x, <8 x i32> %nbits) {
; CHECK-LABEL: @t3_vec_nonsplat(
; CHECK-NEXT: [[T0:%.*]] = add <8 x i32> [[NBITS:%.*]], <i32 -1, i32 0, i32 0, i32 1, i32 0, i32 0, i32 0, i32 0>
; CHECK-NEXT: [[T1:%.*]] = zext <8 x i32> [[T0]] to <8 x i64>
; CHECK-NEXT: [[T2:%.*]] = shl <8 x i64> <i64 -1, i64 -1, i64 -1, i64 -1, i64 -1, i64 -1, i64 -1, i64 -1>, [[T1]]
; CHECK-NEXT: [[T3:%.*]] = xor <8 x i64> [[T2]], <i64 -1, i64 -1, i64 -1, i64 -1, i64 -1, i64 -1, i64 -1, i64 -1>
; CHECK-NEXT: [[T4:%.*]] = sub <8 x i32> <i32 33, i32 32, i32 33, i32 32, i32 32, i32 32, i32 32, i32 32>, [[NBITS]]
; CHECK-NEXT: [[T5:%.*]] = and <8 x i64> [[T3]], [[X:%.*]]
; CHECK-NEXT: call void @use8xi32(<8 x i32> [[T0]])
; CHECK-NEXT: call void @use8xi64(<8 x i64> [[T1]])
; CHECK-NEXT: call void @use8xi64(<8 x i64> [[T2]])
; CHECK-NEXT: call void @use8xi64(<8 x i64> [[T3]])
; CHECK-NEXT: call void @use8xi32(<8 x i32> [[T4]])
; CHECK-NEXT: call void @use8xi64(<8 x i64> [[T5]])
[InstCombine] dropRedundantMaskingOfLeftShiftInput(): truncation (PR42563) Summary: That fold keeps growing and growing :( I think this may be one of the last pieces for it. Since D67677/D67725, the fold knowns the general form of the pattern - where some masking is needed: https://rise4fun.com/Alive/F5R https://rise4fun.com/Alive/gslRa But there is one more huge piece missing - if you are extracting some bits, it is not impossible that the origin is wider than the extraction, i.e. there may be a truncation. And we don't deal with that yet. But we can, and the generalization remains fully identical: https://rise4fun.com/Alive/Uar https://rise4fun.com/Alive/5SW After a preparatory cleanup i think the diff looks rather clean. One missing piece is that in some patterns (especially pat. b), `-1` only needs to be `-1` in final type, but that is for later.. https://bugs.llvm.org/show_bug.cgi?id=42563 Reviewers: spatel, nikic Reviewed By: spatel Subscribers: hiraditya, llvm-commits Tags: #llvm Differential Revision: https://reviews.llvm.org/D69125
2019-11-05 16:49:12 +08:00
; CHECK-NEXT: [[TMP1:%.*]] = trunc <8 x i64> [[X]] to <8 x i32>
; CHECK-NEXT: [[T7:%.*]] = shl <8 x i32> [[TMP1]], [[T4]]
[NFC][InstCombine] Tests for "fold variable mask before variable shift-of-trunc" (PR42563) https://bugs.llvm.org/show_bug.cgi?id=42563 llvm-svn: 375135
2019-10-18 01:20:12 +08:00
; CHECK-NEXT: ret <8 x i32> [[T7]]
;
%t0 = add <8 x i32> %nbits, <i32 -1, i32 0, i32 0, i32 1, i32 0, i32 0, i32 0, i32 0>
%t1 = zext <8 x i32> %t0 to <8 x i64>
%t2 = shl <8 x i64> <i64 -1, i64 -1, i64 -1, i64 -1, i64 -1, i64 -1, i64 -1, i64 -1>, %t1 ; shifting by nbits-1
%t3 = xor <8 x i64> %t2, <i64 -1, i64 -1, i64 -1, i64 -1, i64 -1, i64 -1, i64 -1, i64 -1>
%t4 = sub <8 x i32> <i32 33, i32 32, i32 33, i32 32, i32 32, i32 32, i32 32, i32 32>, %nbits
%t5 = and <8 x i64> %t3, %x
call void @use8xi32(<8 x i32> %t0)
call void @use8xi64(<8 x i64> %t1)
call void @use8xi64(<8 x i64> %t2)
call void @use8xi64(<8 x i64> %t3)
call void @use8xi32(<8 x i32> %t4)
call void @use8xi64(<8 x i64> %t5)
%t6 = trunc <8 x i64> %t5 to <8 x i32>
%t7 = shl <8 x i32> %t6, %t4
ret <8 x i32> %t7
}
; -1 can be truncated.
define i32 @t4_allones_trunc(i64 %x, i32 %nbits) {
; CHECK-LABEL: @t4_allones_trunc(
; CHECK-NEXT: [[T0:%.*]] = zext i32 [[NBITS:%.*]] to i64
; CHECK-NEXT: [[T1:%.*]] = shl i64 -1, [[T0]]
; CHECK-NEXT: [[T2:%.*]] = xor i64 [[T1]], 4294967295
; CHECK-NEXT: [[T3:%.*]] = sub i32 32, [[NBITS]]
; CHECK-NEXT: [[T4:%.*]] = and i64 [[T2]], [[X:%.*]]
; CHECK-NEXT: call void @use32(i32 [[NBITS]])
; CHECK-NEXT: call void @use64(i64 [[T0]])
; CHECK-NEXT: call void @use64(i64 [[T1]])
; CHECK-NEXT: call void @use64(i64 [[T2]])
; CHECK-NEXT: call void @use32(i32 [[T3]])
; CHECK-NEXT: call void @use64(i64 [[T4]])
; CHECK-NEXT: [[T5:%.*]] = trunc i64 [[T4]] to i32
; CHECK-NEXT: [[T6:%.*]] = shl i32 [[T5]], [[T3]]
; CHECK-NEXT: ret i32 [[T6]]
;
%t0 = zext i32 %nbits to i64
%t1 = shl i64 -1, %t0
%t2 = xor i64 %t1, 4294967295
%t3 = sub i32 32, %nbits
%t4 = and i64 %t2, %x
call void @use32(i32 %nbits)
call void @use64(i64 %t0)
call void @use64(i64 %t1)
call void @use64(i64 %t2)
call void @use32(i32 %t3)
call void @use64(i64 %t4)
%t5 = trunc i64 %t4 to i32
%t6 = shl i32 %t5, %t3
ret i32 %t6
}
; Extra uses
define i32 @n5_extrause(i64 %x, i32 %nbits) {
; CHECK-LABEL: @n5_extrause(
; CHECK-NEXT: [[T0:%.*]] = zext i32 [[NBITS:%.*]] to i64
; CHECK-NEXT: [[T1:%.*]] = shl i64 -1, [[T0]]
; CHECK-NEXT: [[T2:%.*]] = xor i64 [[T1]], -1
; CHECK-NEXT: [[T3:%.*]] = sub i32 32, [[NBITS]]
; CHECK-NEXT: [[T4:%.*]] = and i64 [[T2]], [[X:%.*]]
; CHECK-NEXT: call void @use32(i32 [[NBITS]])
; CHECK-NEXT: call void @use64(i64 [[T0]])
; CHECK-NEXT: call void @use64(i64 [[T1]])
; CHECK-NEXT: call void @use64(i64 [[T2]])
; CHECK-NEXT: call void @use32(i32 [[T3]])
; CHECK-NEXT: call void @use64(i64 [[T4]])
; CHECK-NEXT: [[T5:%.*]] = trunc i64 [[T4]] to i32
; CHECK-NEXT: call void @use32(i32 [[T5]])
; CHECK-NEXT: [[T6:%.*]] = shl i32 [[T5]], [[T3]]
; CHECK-NEXT: ret i32 [[T6]]
;
%t0 = zext i32 %nbits to i64
%t1 = shl i64 -1, %t0
%t2 = xor i64 %t1, -1
%t3 = sub i32 32, %nbits
%t4 = and i64 %t2, %x
call void @use32(i32 %nbits)
call void @use64(i64 %t0)
call void @use64(i64 %t1)
call void @use64(i64 %t2)
call void @use32(i32 %t3)
call void @use64(i64 %t4)
%t5 = trunc i64 %t4 to i32
call void @use32(i32 %t5)
%t6 = shl i32 %t5, %t3
ret i32 %t6
}