Fix optimization pass not converging. (#2725)

* Fix optimization pass not converging.

* Fix.

* Fix tests.

---------

Co-authored-by: Yong He <yhe@nvidia.com>

1 files changed, 7 insertions, 9 deletions

diff --git a/tests/slang-extension/atomic-float-byte-address-buffer-cross.slang.glsl b/tests/slang-extension/atomic-float-byte-address-buffer-cross.slang.glsl
index fca1fc1fa..4fb647199 100644
--- a/tests/slang-extension/atomic-float-byte-address-buffer-cross.slang.glsl
+++ b/tests/slang-extension/atomic-float-byte-address-buffer-cross.slang.glsl
@@ -10,16 +10,14 @@ layout(std430, binding = 0) buffer _S2 {
 } _S3;
 void RWByteAddressBuffer_InterlockedAddF32_0(uint _S4, float _S5, out float _S6)
 {
-    uint _S7 = _S4 / 4U;
-    float _S8 = (atomicAdd((((_S3)._data[(_S7)])), (_S5)));
-    _S6 = _S8;
+    float _S7 = (atomicAdd((((_S3)._data[(_S4 / 4U)])), (_S5)));
+    _S6 = _S7;
     return;
 }
 
-void RWByteAddressBuffer_InterlockedAddF32_1(uint _S9, float _S10)
+void RWByteAddressBuffer_InterlockedAddF32_1(uint _S8, float _S9)
 {
-    uint _S11 = _S9 / 4U;
-    float _S12 = (atomicAdd((((_S3)._data[(_S11)])), (_S10)));
+    float _S10 = (atomicAdd((((_S3)._data[(_S8 / 4U)])), (_S9)));
     return;
 }
 
@@ -27,11 +25,11 @@ layout(local_size_x = 16, local_size_y = 1, local_size_z = 1) in;
 void main()
 {
     uint tid_0 = gl_GlobalInvocationID.x;
-    uint _S13 = tid_0 >> 2;
-    int idx_0 = int(tid_0 & 3U ^ _S13);
+    uint _S11 = tid_0 >> 2;
+    int idx_0 = int(tid_0 & 3U ^ _S11);
     float delta_0 = ((anotherBuffer_0)._data[(uint(idx_0 & 3))]);
     float previousValue_0 = 0.0;
     RWByteAddressBuffer_InterlockedAddF32_0(uint(idx_0 << 2), 1.0, previousValue_0);
-    RWByteAddressBuffer_InterlockedAddF32_1(uint(int(_S13) << 2), delta_0);
+    RWByteAddressBuffer_InterlockedAddF32_1(uint(int(_S11) << 2), delta_0);
     return;
 }