1
0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-10-30 15:32:52 +01:00
llvm-mirror/test/CodeGen/ARM/struct-byval-frame-index.ll
Andrew Trick b401fd4c9e Allocate local registers in order for optimal coloring.
Also avoid locals evicting locals just because they want a cheaper register.

Problem: MI Sched knows exactly how many registers we have and assumes
they can be colored. In cases where we have large blocks, usually from
unrolled loops, greedy coloring fails. This is a source of
"regressions" from the MI Scheduler on x86. I noticed this issue on
x86 where we have long chains of two-address defs in the same live
range. It's easy to see this in matrix multiplication benchmarks like
IRSmk and even the unit test misched-matmul.ll.

A fundamental difference between the LLVM register allocator and
conventional graph coloring is that in our model a live range can't
discover its neighbors, it can only verify its neighbors. That's why
we initially went for greedy coloring and added eviction to deal with
the hard cases. However, for singly defined and two-address live
ranges, we can optimally color without visiting neighbors simply by
processing the live ranges in instruction order.

Other beneficial side effects:

It is much easier to understand and debug regalloc for large blocks
when the live ranges are allocated in order. Yes, global allocation is
still very confusing, but it's nice to be able to comprehend what
happened locally.

Heuristics could be added to bias register assignment based on
instruction locality (think late register pairing, banks...).

Intuituvely this will make some test cases that are on the threshold
of register pressure more stable.

llvm-svn: 187139
2013-07-25 18:35:14 +00:00

226 lines
12 KiB
LLVM

; RUN: llc < %s -mcpu=cortex-a15 -verify-machineinstrs | FileCheck %s
; Check a spill right after a function call with large struct byval is correctly
; generated.
; PR16393
; CHECK: set_stored_macroblock_parameters
; CHECK: str r{{.*}}, [sp, [[SLOT:#[0-9]+]]] @ 4-byte Spill
; CHECK: bl RestoreMVBlock8x8
; CHECK: bl RestoreMVBlock8x8
; CHECK: bl RestoreMVBlock8x8
; CHECK: ldr r{{.*}}, [sp, [[SLOT]]] @ 4-byte Reload
target triple = "armv7l-unknown-linux-gnueabihf"
%structA = type { double, [16 x [16 x i16]], [16 x [16 x i16]], [16 x [16 x i16]], i32****, i32***, i32, i16, [4 x i32], [4 x i32], i8**, [16 x i8], [16 x i8], i32, i64, i32, i16******, i16******, [2 x [4 x [4 x i8]]], i32, i32, i32, i32, i32, i32, i32, i32, i32 }
%structB = type { i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, float, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i8**, i8**, i32, i32***, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, [9 x [16 x [16 x i16]]], [5 x [16 x [16 x i16]]], [9 x [8 x [8 x i16]]], [2 x [4 x [16 x [16 x i16]]]], [16 x [16 x i16]], [16 x [16 x i32]], i32****, i32***, i32***, i32***, i32****, i32****, %structC*, %structD*, %structK*, i32*, i32*, i32, i32, i32, i32, [4 x [4 x i32]], i32, i32, i32, i32, i32, double, i32, i32, i32, i32, i16******, i16******, i16******, i16******, [15 x i16], i32, i32, i32, i32, i32, i32, i32, i32, [6 x [32 x i32]], i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, [1 x i32], i32, i32, [2 x i32], i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, %structL*, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, double**, double***, i32***, double**, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, [3 x [2 x i32]], [2 x i32], i32, i32, i16, i32, i32, i32, i32, i32 }
%structC = type { i32, i32, [100 x %structD*], i32, float, float, float }
%structD = type { i32, i32, i32, i32, i32, i32, %structE*, %structH*, %structJ*, i32, i32*, i32*, i32*, i32, i32*, i32*, i32*, i32 (i32)*, [3 x [2 x i32]] }
%structE = type { %structF*, %structG, %structG }
%structF = type { i32, i32, i8, i32, i32, i8, i8, i32, i32, i8*, i32 }
%structG = type { i32, i32, i32, i32, i32, i8*, i32*, i32, i32 }
%structH = type { [3 x [11 x %structI]], [2 x [9 x %structI]], [2 x [10 x %structI]], [2 x [6 x %structI]], [4 x %structI], [4 x %structI], [3 x %structI] }
%structI = type { i16, i8, i32 }
%structJ = type { [2 x %structI], [4 x %structI], [3 x [4 x %structI]], [10 x [4 x %structI]], [10 x [15 x %structI]], [10 x [15 x %structI]], [10 x [5 x %structI]], [10 x [5 x %structI]], [10 x [15 x %structI]], [10 x [15 x %structI]] }
%structK = type { i32, i32, i32, [2 x i32], i32, [8 x i32], %structK*, %structK*, i32, [2 x [4 x [4 x [2 x i32]]]], [16 x i8], [16 x i8], i32, i64, [4 x i32], [4 x i32], i64, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i16, double, i32, i32, i32, i32, i32, i32, i32, i32, i32 }
%structL = type { i32, i32, i32, i32, i32, %structL* }
%structM = type { i32, i32, i32, i32, i32, i32, [6 x [33 x i64]], [6 x [33 x i64]], [6 x [33 x i64]], [6 x [33 x i64]], i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i32, i16**, i16****, i16****, i16*****, i16***, i8*, i8***, i64***, i64***, i16****, i8**, i8**, %structM*, %structM*, %structM*, i32, i32, i32, i32, i32, i32, i32 }
%structN = type { i32, [16 x [16 x i32]], [16 x [16 x i32]], [16 x [16 x i32]], [3 x [16 x [16 x i32]]], [4 x i16], [4 x i8], [4 x i8], [4 x i8], [16 x [16 x i16]], [16 x [16 x i16]], [16 x [16 x i32]] }
@cofAC = external global i32****, align 4
@cofDC = external global i32***, align 4
@rdopt = external global %structA*, align 4
@img = external global %structB*
@enc_picture = external global %structM*
@si_frame_indicator = external global i32, align 4
@sp2_frame_indicator = external global i32, align 4
@lrec = external global i32**, align 4
@tr8x8 = external global %structN, align 4
@best_mode = external global i16, align 2
@best_c_imode = external global i32, align 4
@best_i16offset = external global i32, align 4
@bi_pred_me = external global i16, align 2
@b8mode = external global [4 x i32], align 4
@b8pdir = external global [4 x i32], align 4
@b4_intra_pred_modes = external global [16 x i8], align 1
@b8_intra_pred_modes8x8 = external global [16 x i8], align 1
@b4_ipredmode = external global [16 x i8], align 1
@b8_ipredmode8x8 = external global [4 x [4 x i8]], align 1
@rec_mbY = external global [16 x [16 x i16]], align 2
@lrec_rec = external global [16 x [16 x i32]], align 4
@rec_mbU = external global [16 x [16 x i16]], align 2
@rec_mbV = external global [16 x [16 x i16]], align 2
@lrec_rec_U = external global [16 x [16 x i32]], align 4
@lrec_uv = external global i32***, align 4
@lrec_rec_V = external global [16 x [16 x i32]], align 4
@cbp = external global i32, align 4
@cbp_blk = external global i64, align 8
@luma_transform_size_8x8_flag = external global i32, align 4
@frefframe = external global [4 x [4 x i8]], align 1
@brefframe = external global [4 x [4 x i8]], align 1
; Function Attrs: nounwind
declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) #0
; Function Attrs: nounwind
declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i32, i1) #0
; Function Attrs: nounwind
declare void @SetMotionVectorsMB(%structK* nocapture, i32) #1
; Function Attrs: nounwind
define void @set_stored_macroblock_parameters() #1 {
entry:
%0 = load %structB** @img, align 4, !tbaa !0
%1 = load i32* undef, align 4, !tbaa !3
%mb_data = getelementptr inbounds %structB* %0, i32 0, i32 61
%2 = load %structK** %mb_data, align 4, !tbaa !0
br label %for.body
for.body: ; preds = %for.body, %entry
br i1 undef, label %for.end, label %for.body
for.end: ; preds = %for.body
br i1 undef, label %for.body20, label %if.end
for.body20: ; preds = %for.end
unreachable
if.end: ; preds = %for.end
br i1 undef, label %if.end40, label %for.cond31.preheader
for.cond31.preheader: ; preds = %if.end
unreachable
if.end40: ; preds = %if.end
br i1 undef, label %if.end43, label %if.then42
if.then42: ; preds = %if.end40
br label %if.end43
if.end43: ; preds = %if.then42, %if.end40
br i1 undef, label %if.end164, label %for.cond47.preheader
for.cond47.preheader: ; preds = %if.end43
br i1 undef, label %for.body119, label %if.end164
for.body119: ; preds = %for.body119, %for.cond47.preheader
br i1 undef, label %for.body119, label %if.end164
if.end164: ; preds = %for.body119, %for.cond47.preheader, %if.end43
store i32*** null, i32**** @cofDC, align 4, !tbaa !0
%mb_type = getelementptr inbounds %structK* %2, i32 %1, i32 8
br i1 undef, label %if.end230, label %if.then169
if.then169: ; preds = %if.end164
br i1 undef, label %for.cond185.preheader, label %for.cond210.preheader
for.cond185.preheader: ; preds = %if.then169
unreachable
for.cond210.preheader: ; preds = %if.then169
unreachable
if.end230: ; preds = %if.end164
tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* undef, i8* bitcast ([4 x i32]* @b8mode to i8*), i32 16, i32 4, i1 false)
%b8pdir = getelementptr inbounds %structK* %2, i32 %1, i32 15
%3 = bitcast [4 x i32]* %b8pdir to i8*
tail call void @llvm.memcpy.p0i8.p0i8.i32(i8* %3, i8* bitcast ([4 x i32]* @b8pdir to i8*), i32 16, i32 4, i1 false)
br i1 undef, label %if.end236, label %if.then233
if.then233: ; preds = %if.end230
unreachable
if.end236: ; preds = %if.end230
%cmp242 = icmp ne i16 undef, 8
%4 = load i32* @luma_transform_size_8x8_flag, align 4, !tbaa !3
%tobool245 = icmp ne i32 %4, 0
%or.cond812 = or i1 %cmp242, %tobool245
br i1 %or.cond812, label %if.end249, label %land.lhs.true246
land.lhs.true246: ; preds = %if.end236
br i1 undef, label %if.end249, label %if.then248
if.then248: ; preds = %land.lhs.true246
tail call void asm sideeffect "", "~{r1},~{r2},~{r3},~{r4},~{r5},~{r6},~{r7},~{r8},~{r9},~{r10},~{r11}"() nounwind
tail call void @RestoreMVBlock8x8(i32 1, i32 0, %structN* byval @tr8x8, i32 0) #0
tail call void @RestoreMVBlock8x8(i32 1, i32 2, %structN* byval @tr8x8, i32 0) #0
tail call void @RestoreMVBlock8x8(i32 1, i32 3, %structN* byval @tr8x8, i32 0) #0
br label %if.end249
if.end249: ; preds = %if.then248, %land.lhs.true246, %if.end236
%5 = load i32* @luma_transform_size_8x8_flag, align 4, !tbaa !3
%6 = load %structA** @rdopt, align 4, !tbaa !0
%luma_transform_size_8x8_flag264 = getelementptr inbounds %structA* %6, i32 0, i32 21
store i32 %5, i32* %luma_transform_size_8x8_flag264, align 4, !tbaa !3
%7 = load i32* undef, align 4, !tbaa !3
%add281 = add nsw i32 %7, 0
br label %for.body285
for.body285: ; preds = %for.inc503, %if.end249
%8 = phi %structB* [ undef, %if.end249 ], [ %.pre1155, %for.inc503 ]
%i.21103 = phi i32 [ 0, %if.end249 ], [ %inc504, %for.inc503 ]
%block_x286 = getelementptr inbounds %structB* %8, i32 0, i32 37
%9 = load i32* %block_x286, align 4, !tbaa !3
%add287 = add nsw i32 %9, %i.21103
%shr289 = ashr i32 %i.21103, 1
%add290 = add nsw i32 %shr289, 0
%arrayidx292 = getelementptr inbounds %structK* %2, i32 %1, i32 15, i32 %add290
%10 = load %structM** @enc_picture, align 4, !tbaa !0
%ref_idx = getelementptr inbounds %structM* %10, i32 0, i32 35
%11 = load i8**** %ref_idx, align 4, !tbaa !0
%12 = load i8*** %11, align 4, !tbaa !0
%arrayidx313 = getelementptr inbounds i8** %12, i32 %add281
%13 = load i8** %arrayidx313, align 4, !tbaa !0
%arrayidx314 = getelementptr inbounds i8* %13, i32 %add287
store i8 -1, i8* %arrayidx314, align 1, !tbaa !1
%14 = load %structB** @img, align 4, !tbaa !0
%MbaffFrameFlag327 = getelementptr inbounds %structB* %14, i32 0, i32 100
%15 = load i32* %MbaffFrameFlag327, align 4, !tbaa !3
%tobool328 = icmp eq i32 %15, 0
br i1 %tobool328, label %if.end454, label %if.then329
if.then329: ; preds = %for.body285
%16 = load %structA** @rdopt, align 4, !tbaa !0
br label %if.end454
if.end454: ; preds = %if.then329, %for.body285
%17 = load i32* %arrayidx292, align 4, !tbaa !3
%cmp457 = icmp eq i32 %17, 0
br i1 %cmp457, label %if.then475, label %lor.lhs.false459
lor.lhs.false459: ; preds = %if.end454
%18 = load i32* %mb_type, align 4, !tbaa !3
switch i32 %18, label %for.inc503 [
i32 9, label %if.then475
i32 10, label %if.then475
i32 13, label %if.then475
i32 14, label %if.then475
]
if.then475: ; preds = %lor.lhs.false459, %lor.lhs.false459, %lor.lhs.false459, %lor.lhs.false459, %if.end454
store i16 0, i16* undef, align 2, !tbaa !4
br label %for.inc503
for.inc503: ; preds = %if.then475, %lor.lhs.false459
%inc504 = add nsw i32 %i.21103, 1
%.pre1155 = load %structB** @img, align 4, !tbaa !0
br label %for.body285
}
; Function Attrs: nounwind
declare void @update_offset_params(i32, i32) #1
; Function Attrs: nounwind
declare void @RestoreMVBlock8x8(i32, i32, %structN* byval nocapture, i32) #1
attributes #0 = { nounwind }
attributes #1 = { nounwind "less-precise-fpmad"="false" "no-frame-pointer-elim"="false" "no-frame-pointer-elim-non-leaf"="false" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "unsafe-fp-math"="false" "use-soft-float"="false" }
!0 = metadata !{metadata !"any pointer", metadata !1}
!1 = metadata !{metadata !"omnipotent char", metadata !2}
!2 = metadata !{metadata !"Simple C/C++ TBAA"}
!3 = metadata !{metadata !"int", metadata !1}
!4 = metadata !{metadata !"short", metadata !1}