841077a7e9
Summary: The bit order of the has_vec and longtbtable bits in the traceback table generated by the XL compiler flipped at some point after v12.1. This is different from the definition is the AIX header debug.h. The change in the XL compiler that caused the deviation from the OS header definition was unintentional. Since both orderings are extant and the XL compiler runtime also expects the ordering defined by the OS, we will correct the output from LLVM to match the defined ordering given by the OS (which is also consistent with the Assembler Language Reference). Mitigation for traceback tables encoded with the wrong ordering is required for either ordering. Reviewers: XingXue, HubertTong Differential Revision: https://reviews.llvm.org/D105487
130 lines
7.4 KiB
LLVM
130 lines
7.4 KiB
LLVM
; RUN: llc -verify-machineinstrs -mtriple powerpc-ibm-aix-xcoff -mcpu=pwr7 \
|
|
; RUN: -mattr=+altivec -vec-extabi -xcoff-traceback-table=true < %s | \
|
|
; RUN: FileCheck --check-prefixes=CHECK-ASM,COMMON %s
|
|
|
|
; RUN: llc -verify-machineinstrs -mtriple powerpc-ibm-aix-xcoff -function-sections \
|
|
; RUN: -mcpu=pwr7 -mattr=+altivec -vec-extabi < %s | \
|
|
; RUN: FileCheck --check-prefixes=CHECK-FUNC,COMMON %s
|
|
|
|
;; #include <altivec.h>
|
|
;; vector float f(vector int vi1, int i1, int i2, float f1, vector float vf,double d1, vector char vc1) {
|
|
;; return vec_abs(vf);
|
|
;; }
|
|
;; vector float fin(int x) {
|
|
;; vector float vf ={1.0,1.0,1.0,1.0};
|
|
;; if (x) return vf;
|
|
;; return vec_abs(vf);
|
|
;; }
|
|
|
|
define dso_local <4 x float> @f(<4 x i32> %vi1, i32 signext %i1, i32 signext %i2, float %f1, <4 x float> %vf, double %d1, <16 x i8> %vc1) #0 {
|
|
entry:
|
|
%__a.addr.i = alloca <4 x float>, align 16
|
|
%vi1.addr = alloca <4 x i32>, align 16
|
|
%i1.addr = alloca i32, align 4
|
|
%i2.addr = alloca i32, align 4
|
|
%f1.addr = alloca float, align 4
|
|
%vf.addr = alloca <4 x float>, align 16
|
|
%d1.addr = alloca double, align 8
|
|
%vc1.addr = alloca <16 x i8>, align 16
|
|
store <4 x i32> %vi1, <4 x i32>* %vi1.addr, align 16
|
|
store i32 %i1, i32* %i1.addr, align 4
|
|
store i32 %i2, i32* %i2.addr, align 4
|
|
store float %f1, float* %f1.addr, align 4
|
|
store <4 x float> %vf, <4 x float>* %vf.addr, align 16
|
|
store double %d1, double* %d1.addr, align 8
|
|
store <16 x i8> %vc1, <16 x i8>* %vc1.addr, align 16
|
|
%0 = load <4 x float>, <4 x float>* %vf.addr, align 16
|
|
store <4 x float> %0, <4 x float>* %__a.addr.i, align 16
|
|
%1 = load <4 x float>, <4 x float>* %__a.addr.i, align 16
|
|
%2 = load <4 x float>, <4 x float>* %__a.addr.i, align 16
|
|
%3 = call <4 x float> @llvm.fabs.v4f32(<4 x float> %2) #2
|
|
ret <4 x float> %3
|
|
}
|
|
|
|
define <4 x float> @fin(i32 %x) #0 {
|
|
entry:
|
|
%__a.addr.i = alloca <4 x float>, align 16
|
|
%__res.i = alloca <4 x i32>, align 16
|
|
%retval = alloca <4 x float>, align 16
|
|
%x.addr = alloca i32, align 4
|
|
%vf = alloca <4 x float>, align 16
|
|
store i32 %x, i32* %x.addr, align 4
|
|
store <4 x float> <float 1.000000e+00, float 1.000000e+00, float 1.000000e+00, float 1.000000e+00>, <4 x float>* %vf, align 16
|
|
%0 = load i32, i32* %x.addr, align 4
|
|
%tobool = icmp ne i32 %0, 0
|
|
br i1 %tobool, label %if.then, label %if.end
|
|
|
|
if.then: ; preds = %entry
|
|
%1 = load <4 x float>, <4 x float>* %vf, align 16
|
|
store <4 x float> %1, <4 x float>* %retval, align 16
|
|
br label %return
|
|
|
|
if.end: ; preds = %entry
|
|
%2 = load <4 x float>, <4 x float>* %vf, align 16
|
|
store <4 x float> %2, <4 x float>* %__a.addr.i, align 16
|
|
%3 = load <4 x float>, <4 x float>* %__a.addr.i, align 16
|
|
%4 = bitcast <4 x float> %3 to <4 x i32>
|
|
%and.i = and <4 x i32> %4, <i32 2147483647, i32 2147483647, i32 2147483647, i32 2147483647>
|
|
store <4 x i32> %and.i, <4 x i32>* %__res.i, align 16
|
|
%5 = load <4 x i32>, <4 x i32>* %__res.i, align 16
|
|
%6 = bitcast <4 x i32> %5 to <4 x float>
|
|
store <4 x float> %6, <4 x float>* %retval, align 16
|
|
br label %return
|
|
|
|
return: ; preds = %if.end, %if.then
|
|
%7 = load <4 x float>, <4 x float>* %retval, align 16
|
|
ret <4 x float> %7
|
|
}
|
|
|
|
declare <4 x float> @llvm.fabs.v4f32(<4 x float>) #1
|
|
|
|
; COMMON: L..f0:
|
|
; COMMON-NEXT: .vbyte 4, 0x00000000 # Traceback table begin
|
|
; COMMON-NEXT: .byte 0x00 # Version = 0
|
|
; COMMON-NEXT: .byte 0x09 # Language = CPlusPlus
|
|
; COMMON-NEXT: .byte 0x22 # -IsGlobaLinkage, -IsOutOfLineEpilogOrPrologue
|
|
; COMMON-NEXT: # +HasTraceBackTableOffset, -IsInternalProcedure
|
|
; COMMON-NEXT: # -HasControlledStorage, -IsTOCless
|
|
; COMMON-NEXT: # +IsFloatingPointPresent
|
|
; COMMON-NEXT: # -IsFloatingPointOperationLogOrAbortEnabled
|
|
; COMMON-NEXT: .byte 0x40 # -IsInterruptHandler, +IsFunctionNamePresent, -IsAllocaUsed
|
|
; COMMON-NEXT: # OnConditionDirective = 0, -IsCRSaved, -IsLRSaved
|
|
; COMMON-NEXT: .byte 0x80 # +IsBackChainStored, -IsFixup, NumOfFPRsSaved = 0
|
|
; COMMON-NEXT: .byte 0x40 # -HasExtensionTable, +HasVectorInfo, NumOfGPRsSaved = 0
|
|
; COMMON-NEXT: .byte 0x02 # NumberOfFixedParms = 2
|
|
; COMMON-NEXT: .byte 0x05 # NumberOfFPParms = 2, +HasParmsOnStack
|
|
; COMMON-NEXT: .vbyte 4, 0x42740000 # Parameter type = v, i, i, f, v, d, v
|
|
; CHECK-ASM-NEXT: .vbyte 4, L..f0-.f # Function size
|
|
; CHECK-FUNC-NEXT: .vbyte 4, L..f0-.f[PR] # Function size
|
|
; COMMON-NEXT: .vbyte 2, 0x0001 # Function name len = 1
|
|
; COMMON-NEXT: .byte 102 # Function Name
|
|
; COMMON-NEXT: .byte 0x00 # NumOfVRsSaved = 0, -IsVRSavedOnStack, -HasVarArgs
|
|
; COMMON-NEXT: .byte 0x07 # NumOfVectorParams = 3, +HasVMXInstruction
|
|
; COMMON-NEXT: .vbyte 4, 0xb0000000 # Vector Parameter type = vi, vf, vc
|
|
; COMMON-NEXT: .vbyte 2, 0x0000 # Padding
|
|
|
|
; COMMON: L..fin0:
|
|
; COMMON-NEXT: .vbyte 4, 0x00000000 # Traceback table begin
|
|
; COMMON-NEXT: .byte 0x00 # Version = 0
|
|
; COMMON-NEXT: .byte 0x09 # Language = CPlusPlus
|
|
; COMMON-NEXT: .byte 0x22 # -IsGlobaLinkage, -IsOutOfLineEpilogOrPrologue
|
|
; COMMON-NEXT: # +HasTraceBackTableOffset, -IsInternalProcedure
|
|
; COMMON-NEXT: # -HasControlledStorage, -IsTOCless
|
|
; COMMON-NEXT: # +IsFloatingPointPresent
|
|
; COMMON-NEXT: # -IsFloatingPointOperationLogOrAbortEnabled
|
|
; COMMON-NEXT: .byte 0x40 # -IsInterruptHandler, +IsFunctionNamePresent, -IsAllocaUsed
|
|
; COMMON-NEXT: # OnConditionDirective = 0, -IsCRSaved, -IsLRSaved
|
|
; COMMON-NEXT: .byte 0x80 # +IsBackChainStored, -IsFixup, NumOfFPRsSaved = 0
|
|
; COMMON-NEXT: .byte 0x40 # -HasExtensionTable, +HasVectorInfo, NumOfGPRsSaved = 0
|
|
; COMMON-NEXT: .byte 0x01 # NumberOfFixedParms = 1
|
|
; COMMON-NEXT: .byte 0x01 # NumberOfFPParms = 0, +HasParmsOnStack
|
|
; COMMON-NEXT: .vbyte 4, 0x00000000 # Parameter type = i
|
|
; CHECK-ASM-NEXT: .vbyte 4, L..fin0-.fin # Function size
|
|
; CHECK-FUNC-NEXT: .vbyte 4, L..fin0-.fin[PR] # Function size
|
|
; COMMON-NEXT: .vbyte 2, 0x0003 # Function name len = 3
|
|
; COMMON-NEXT: .byte "fin" # Function Name
|
|
; COMMON-NEXT: .byte 0x00 # NumOfVRsSaved = 0, -IsVRSavedOnStack, -HasVarArgs
|
|
; COMMON-NEXT: .byte 0x01 # NumOfVectorParams = 0, +HasVMXInstruction
|
|
; COMMON-NEXT: .vbyte 4, 0x00000000 # Vector Parameter type =
|
|
; COMMON-NEXT: .vbyte 2, 0x0000 # Padding
|