Fix Metal invalid as_type cast for 64-bit RWByteAddressBuffer.Store values (#7843)

* Fix 64-bit val lowering for metal

* Add ByteAddressBuffer load/store 64-bit tests

* Handle Store/Load ptr types

* Use bitcast for non-pointer typers

* format code (#7966)

Co-authored-by: slangbot <186143334+slangbot@users.noreply.github.com>

---------

Co-authored-by: slangbot <ellieh+slangbot@nvidia.com>
Co-authored-by: slangbot <186143334+slangbot@users.noreply.github.com>

3 files changed, 60 insertions, 2 deletions

diff --git a/tests/compute/byte-address-buffer-64bit.slang b/tests/compute/byte-address-buffer-64bit.slang
new file mode 100644
index 000000000..e8d6a42f2
--- /dev/null
+++ b/tests/compute/byte-address-buffer-64bit.slang
@@ -0,0 +1,32 @@
+// byte-address-buffer.slang
+
+//TEST(compute):COMPARE_COMPUTE_EX:-cpu -compute -shaderobj
+//TEST(compute):COMPARE_COMPUTE_EX:-d3d12 -compute -shaderobj -profile cs_6_0 -use-dxil
+//TEST(compute, vulkan):COMPARE_COMPUTE_EX:-vk -compute -shaderobj
+
+//TEST_INPUT:ubuffer(data=[0 1 2 3]):name=inputBuffer
+RWByteAddressBuffer inputBuffer;
+
+//TEST_INPUT:ubuffer(data=[0 0 0 0 0 0 0 0]):out,name=outputBuffer
+RWByteAddressBuffer outputBuffer;
+
+void testInt64(uint val)
+{
+    // Load a 32-bit value from the input buffer
+    uint tmp = inputBuffer.Load(uint(val * 4));
+
+    // Cast to uint64_t
+    uint64_t tmp64 = uint64_t(tmp) + 1;
+
+    // Store the result back as uint64_t
+    outputBuffer.Store(uint(val * 8), tmp64);
+}
+
+[numthreads(4, 1, 1)]
+[shader("compute")]
+void computeMain(uint3 dispatchThreadID : SV_DispatchThreadID)
+{
+    uint tid = dispatchThreadID.x;
+
+    testInt64(tid);
+}
diff --git a/tests/compute/byte-address-buffer-64bit.slang.expected.txt b/tests/compute/byte-address-buffer-64bit.slang.expected.txt
new file mode 100644
index 000000000..ac5027985
--- /dev/null
+++ b/tests/compute/byte-address-buffer-64bit.slang.expected.txt
@@ -0,0 +1,8 @@
+1
+0
+2
+0
+3
+0
+4
+0
\ No newline at end of file
diff --git a/tests/metal/byte-address-buffer.slang b/tests/metal/byte-address-buffer.slang
index 53c0e0ac5..8514dafe6 100644
--- a/tests/metal/byte-address-buffer.slang
+++ b/tests/metal/byte-address-buffer.slang
@@ -1,8 +1,6 @@
 //TEST:SIMPLE(filecheck=CHECK): -target metal
 //TEST:SIMPLE(filecheck=CHECK-ASM): -target metallib
 
-uniform RWStructuredBuffer<float> outputBuffer;
-
 RWByteAddressBuffer buffer;
 
 // CHECK-ASM: define void @main_kernel
@@ -27,4 +25,24 @@ void main_kernel(uint3 tid: SV_DispatchThreadID)
     // CHECK: {{.*}}[(128U)>>2] = as_type<uint32_t>(({{.*}} & 4294967040U) | (uint([[A]]) << 0U));
     // CHECK: {{.*}}[(128U)>>2] = as_type<uint32_t>(({{.*}} & 65535U) | (uint(as_type<ushort>([[H]])) << 16U));
     buffer.Store(128, buffer.Load<TestStruct>(0));
+
+    // CHECK: {{.*}}[(256U)>>2] = as_type<uint32_t>(4294967295U);
+    // CHECK: {{.*}}[(260U)>>2] = as_type<uint32_t>(4294967295U);
+    int64_t i64 = -1;
+    buffer.Store(256, i64);
+
+    // CHECK: {{.*}}[(264U)>>2] = as_type<uint32_t>(123U);
+    // CHECK: {{.*}}[(268U)>>2] = as_type<uint32_t>(0U);
+    uint64_t u64 = 123;
+    buffer.Store(264, u64);
+
+    int64_t* ptr = Ptr<int64_t>(0xFF);
+    // CHECK: {{.*}}[(272U)>>2] = as_type<uint32_t>({{.*}});
+    // CHECK: {{.*}}[(276U)>>2] = as_type<uint32_t>({{.*}});
+    buffer.Store(272, ptr);
+
+    // CHECK: {{.*}}[(280U)>>2] = as_type<uint32_t>(4294967295U);
+    // CHECK: {{.*}}[(284U)>>2] = as_type<uint32_t>(4294967295U);
+    uintptr_t uintptr_val = (uintptr_t)-1;
+    buffer.Store(280, uintptr_val);
 }