Test Failure: Transforms/InstSimplify/fast-math.ll

Test source: git

Log:

Source: <stdin>

----------------------------------------
define float @mul_zero_1(float %a) {
%0:
  %b = fmul nnan nsz float %a, 0.000000
  ret float %b
}
=>
define float @mul_zero_1(float %a) {
%0:
  ret float 0.000000
}
Transformation seems to be correct!


----------------------------------------
define float @mul_zero_2(float %a) {
%0:
  %b = fmul fast float 0.000000, %a
  ret float %b
}
=>
define float @mul_zero_2(float %a) {
%0:
  ret float 0.000000
}
Transformation doesn't verify!
ERROR: Invalid expr


----------------------------------------
define <2 x float> @mul_zero_nsz_nnan_vec_undef(<2 x float> %a) {
%0:
  %b = fmul nnan nsz <2 x float> %a, { 0.000000, undef }
  ret <2 x float> %b
}
=>
define <2 x float> @mul_zero_nsz_nnan_vec_undef(<2 x float> %a) {
%0:
  ret <2 x float> { 0.000000, 0.000000 }
}
Transformation seems to be correct!


----------------------------------------
define float @no_mul_zero_1(float %a) {
%0:
  %b = fmul nsz float %a, 0.000000
  ret float %b
}
=>
define float @no_mul_zero_1(float %a) {
%0:
  %b = fmul nsz float %a, 0.000000
  ret float %b
}
Transformation seems to be correct!


----------------------------------------
define float @no_mul_zero_2(float %a) {
%0:
  %b = fmul nnan float %a, 0.000000
  ret float %b
}
=>
define float @no_mul_zero_2(float %a) {
%0:
  %b = fmul nnan float %a, 0.000000
  ret float %b
}
Transformation seems to be correct!


----------------------------------------
define float @no_mul_zero_3(float %a) {
%0:
  %b = fmul float %a, 0.000000
  ret float %b
}
=>
define float @no_mul_zero_3(float %a) {
%0:
  %b = fmul float %a, 0.000000
  ret float %b
}
Transformation seems to be correct!


----------------------------------------
define float @fadd_binary_fnegx(float %x) {
%0:
  %negx = fsub float -0.000000, %x
  %r = fadd nnan float %negx, %x
  ret float %r
}
=>
define float @fadd_binary_fnegx(float %x) {
%0:
  ret float 0.000000
}
Transformation doesn't verify!
ERROR: Timeout


----------------------------------------
define float @fadd_unary_fnegx(float %x) {
%0:
  %negx = fneg float %x
  %r = fadd nnan float %negx, %x
  ret float %r
}
=>
define float @fadd_unary_fnegx(float %x) {
%0:
  ret float 0.000000
}
Transformation seems to be correct!


----------------------------------------
define <2 x float> @fadd_binary_fnegx_commute_vec(<2 x float> %x) {
%0:
  %negx = fsub <2 x float> { -0.000000, -0.000000 }, %x
  %r = fadd nnan <2 x float> %x, %negx
  ret <2 x float> %r
}
=>
define <2 x float> @fadd_binary_fnegx_commute_vec(<2 x float> %x) {
%0:
  ret <2 x float> { 0.000000, 0.000000 }
}
Transformation seems to be correct!


----------------------------------------
define <2 x float> @fadd_unary_fnegx_commute_vec(<2 x float> %x) {
%0:
  %negx = fneg <2 x float> %x
  %r = fadd nnan <2 x float> %x, %negx
  ret <2 x float> %r
}
=>
define <2 x float> @fadd_unary_fnegx_commute_vec(<2 x float> %x) {
%0:
  ret <2 x float> { 0.000000, 0.000000 }
}
Transformation seems to be correct!


----------------------------------------
define <2 x float> @fadd_fnegx_commute_vec_undef(<2 x float> %x) {
%0:
  %negx = fsub <2 x float> { undef, -0.000000 }, %x
  %r = fadd nnan <2 x float> %x, %negx
  ret <2 x float> %r
}
=>
define <2 x float> @fadd_fnegx_commute_vec_undef(<2 x float> %x) {
%0:
  ret <2 x float> { 0.000000, 0.000000 }
}
Transformation doesn't verify!
ERROR: Timeout


----------------------------------------
define float @fadd_binary_fneg_nan(float %x) {
%0:
  %t = fsub nnan float -0.000000, %x
  %could_be_nan = fadd ninf float %t, %x
  ret float %could_be_nan
}
=>
define float @fadd_binary_fneg_nan(float %x) {
%0:
  %t = fsub nnan float -0.000000, %x
  %could_be_nan = fadd ninf float %t, %x
  ret float %could_be_nan
}
Transformation seems to be correct!


----------------------------------------
define float @fadd_unary_fneg_nan(float %x) {
%0:
  %t = fneg nnan float %x
  %could_be_nan = fadd ninf float %t, %x
  ret float %could_be_nan
}
=>
define float @fadd_unary_fneg_nan(float %x) {
%0:
  %t = fneg nnan float %x
  %could_be_nan = fadd ninf float %t, %x
  ret float %could_be_nan
}
Transformation doesn't verify!
ERROR: Invalid expr


----------------------------------------
define float @fadd_binary_fneg_nan_commute(float %x) {
%0:
  %t = fsub nnan ninf float -0.000000, %x
  %could_be_nan = fadd float %x, %t
  ret float %could_be_nan
}
=>
define float @fadd_binary_fneg_nan_commute(float %x) {
%0:
  %t = fsub nnan ninf float -0.000000, %x
  %could_be_nan = fadd float %x, %t
  ret float %could_be_nan
}
Transformation seems to be correct!


----------------------------------------
define float @fadd_unary_fneg_nan_commute(float %x) {
%0:
  %t = fneg nnan ninf float %x
  %could_be_nan = fadd float %x, %t
  ret float %could_be_nan
}
=>
define float @fadd_unary_fneg_nan_commute(float %x) {
%0:
  %t = fneg nnan ninf float %x
  %could_be_nan = fadd float %x, %t
  ret float %could_be_nan
}
Transformation doesn't verify!
ERROR: Invalid expr


----------------------------------------
define float @fadd_fsub_nnan_ninf(float %x) {
%0:
  %sub = fsub float 0.000000, %x
  %zero = fadd nnan ninf float %x, %sub
  ret float %zero
}
=>
define float @fadd_fsub_nnan_ninf(float %x) {
%0:
  ret float 0.000000
}
Transformation seems to be correct!


----------------------------------------
define <2 x float> @fadd_fsub_nnan_ninf_commute_vec(<2 x float> %x) {
%0:
  %sub = fsub <2 x float> { 0.000000, 0.000000 }, %x
  %zero = fadd nnan ninf <2 x float> %sub, %x
  ret <2 x float> %zero
}
=>
define <2 x float> @fadd_fsub_nnan_ninf_commute_vec(<2 x float> %x) {
%0:
  ret <2 x float> { 0.000000, 0.000000 }
}
Transformation doesn't verify!
ERROR: Timeout


----------------------------------------
define float @fadd_fsub_nnan(float %x) {
%0:
  %sub = fsub float 0.000000, %x
  %zero = fadd nnan float %sub, %x
  ret float %zero
}
=>
define float @fadd_fsub_nnan(float %x) {
%0:
  ret float 0.000000
}
Transformation seems to be correct!


----------------------------------------
define float @fsub_x_x(float %a) {
%0:
  %zero1 = fsub nnan float %a, %a
  %no_zero1 = fsub ninf float %a, %a
  %no_zero2 = fsub float %a, %a
  %no_zero = fadd float %no_zero1, %no_zero2
  %ret = fadd nsz float %no_zero, %zero1
  ret float %ret
}
=>
define float @fsub_x_x(float %a) {
%0:
  %no_zero1 = fsub ninf float %a, %a
  %no_zero2 = fsub float %a, %a
  %no_zero = fadd float %no_zero1, %no_zero2
  ret float %no_zero
}
Transformation doesn't verify!
ERROR: Timeout


----------------------------------------
define float @fsub_0_0_x(float %a) {
%0:
  %t1 = fsub float 0.000000, %a
  %ret = fsub nsz float 0.000000, %t1
  ret float %ret
}
=>
define float @fsub_0_0_x(float %a) {
%0:
  ret float %a
}
Transformation seems to be correct!


----------------------------------------
define float @fneg_x(float %a) {
%0:
  %t1 = fneg float %a
  %ret = fsub nsz float 0.000000, %t1
  ret float %ret
}
=>
define float @fneg_x(float %a) {
%0:
  ret float %a
}
Transformation seems to be correct!


----------------------------------------
define <2 x float> @fsub_0_0_x_vec_undef1(<2 x float> %a) {
%0:
  %t1 = fsub <2 x float> { 0.000000, undef }, %a
  %ret = fsub nsz <2 x float> { 0.000000, 0.000000 }, %t1
  ret <2 x float> %ret
}
=>
define <2 x float> @fsub_0_0_x_vec_undef1(<2 x float> %a) {
%0:
  ret <2 x float> %a
}
Transformation seems to be correct!


----------------------------------------
define <2 x float> @fneg_x_vec_undef1(<2 x float> %a) {
%0:
  %t1 = fneg <2 x float> %a
  %ret = fsub nsz <2 x float> { 0.000000, undef }, %t1
  ret <2 x float> %ret
}
=>
define <2 x float> @fneg_x_vec_undef1(<2 x float> %a) {
%0:
  ret <2 x float> %a
}
Transformation doesn't verify!
ERROR: Timeout


----------------------------------------
define <2 x float> @fsub_0_0_x_vec_undef2(<2 x float> %a) {
%0:
  %t1 = fsub <2 x float> { 0.000000, 0.000000 }, %a
  %ret = fsub nsz <2 x float> { undef, -0.000000 }, %t1
  ret <2 x float> %ret
}
=>
define <2 x float> @fsub_0_0_x_vec_undef2(<2 x float> %a) {
%0:
  ret <2 x float> %a
}
Transformation doesn't verify!
ERROR: Timeout


----------------------------------------
define <2 x float> @fadd_zero_nsz_vec(<2 x float> %x) {
%0:
  %r = fadd nsz <2 x float> %x, { 0.000000, 0.000000 }
  ret <2 x float> %r
}
=>
define <2 x float> @fadd_zero_nsz_vec(<2 x float> %x) {
%0:
  ret <2 x float> %x
}
Transformation seems to be correct!


----------------------------------------
define <2 x float> @fadd_zero_nsz_vec_undef(<2 x float> %x) {
%0:
  %r = fadd nsz <2 x float> %x, { 0.000000, undef }
  ret <2 x float> %r
}
=>
define <2 x float> @fadd_zero_nsz_vec_undef(<2 x float> %x) {
%0:
  ret <2 x float> %x
}
Transformation doesn't verify!
ERROR: Timeout


----------------------------------------
define float @nofold_fadd_x_0(float %a) {
%0:
  %no_zero1 = fadd ninf float %a, 0.000000
  %no_zero2 = fadd nnan float %a, 0.000000
  %no_zero = fadd float %no_zero1, %no_zero2
  ret float %no_zero
}
=>
define float @nofold_fadd_x_0(float %a) {
%0:
  %no_zero1 = fadd ninf float %a, 0.000000
  %no_zero2 = fadd nnan float %a, 0.000000
  %no_zero = fadd float %no_zero1, %no_zero2
  ret float %no_zero
}
Transformation seems to be correct!


----------------------------------------
define float @fold_fadd_nsz_x_0(float %a) {
%0:
  %add = fadd nsz float %a, 0.000000
  ret float %add
}
=>
define float @fold_fadd_nsz_x_0(float %a) {
%0:
  ret float %a
}
Transformation seems to be correct!


----------------------------------------
define float @fold_fadd_cannot_be_neg0_nsz_src_x_0(float %a, float %b) {
%0:
  %nsz = fmul nsz float %a, %b
  %add = fadd float %nsz, 0.000000
  ret float %add
}
=>
define float @fold_fadd_cannot_be_neg0_nsz_src_x_0(float %a, float %b) {
%0:
  %nsz = fmul nsz float %a, %b
  ret float %nsz
}
Transformation doesn't verify!
ERROR: Value mismatch

Example:
float %a = undef
float %b = undef

Source:
float %nsz = NaN	[based on undef value]
float %add = NaN

Target:
float %nsz = #x80000000 (-0.0)
Source value: NaN
Target value: #x80000000 (-0.0)


------------------- SMT STATS -------------------
Num queries: 61
Num invalid: 3
Num skips:   0
Num trivial: 68 (52.7%)
Num timeout: 8 (13.1%)
Num errors:  0 (0.0%)
Num SAT:     30 (49.2%)
Num UNSAT:   23 (37.7%)

stderr:

+ : 'RUN: at line 2'
+ /home/nlopes/alive2/scripts/opt-alive.sh -instsimplify -S
+ /home/nlopes/llvm/build/bin/FileCheck /home/nlopes/llvm/llvm/test/Transforms/InstSimplify/fast-math.ll

Alive2: Transform doesn't verify; aborting!
FileCheck error: '<stdin>' is empty.
FileCheck command line:  /home/nlopes/llvm/build/bin/FileCheck /home/nlopes/llvm/llvm/test/Transforms/InstSimplify/fast-math.ll

<-- Back