[llvm] [AMDGPU] gfx1250 has fixed GETPC bug and also extended VA to 57 bits (PR #152373)

[via llvm-commits]

llvmbot wrote:


<!--LLVM PR SUMMARY COMMENT-->

@llvm/pr-subscribers-backend-amdgpu

Author: Stanislav Mekhanoshin (rampitec)

<details>
<summary>Changes</summary>



---
Full diff: https://github.com/llvm/llvm-project/pull/152373.diff


1 Files Affected:

- (modified) llvm/lib/Target/AMDGPU/GCNSubtarget.h (+3-2) 


``````````diff
diff --git a/llvm/lib/Target/AMDGPU/GCNSubtarget.h b/llvm/lib/Target/AMDGPU/GCNSubtarget.h
index 5530886831cae..75069ef41971e 100644
--- a/llvm/lib/Target/AMDGPU/GCNSubtarget.h
+++ b/llvm/lib/Target/AMDGPU/GCNSubtarget.h
@@ -1559,8 +1559,9 @@ class GCNSubtarget final : public AMDGPUGenSubtargetInfo,
   bool hasSetPrioIncWgInst() const { return HasSetPrioIncWgInst; }
 
   // \returns true if S_GETPC_B64 zero-extends the result from 48 bits instead
-  // of sign-extending.
-  bool hasGetPCZeroExtension() const { return GFX12Insts; }
+  // of sign-extending. Note that GFX1250 has not only fixed the bug but also
+  // extended VA to 57 bits.
+  bool hasGetPCZeroExtension() const { return GFX12Insts && !GFX1250Insts; }
 
   /// \returns SGPR allocation granularity supported by the subtarget.
   unsigned getSGPRAllocGranule() const {

``````````

</details>


https://github.com/llvm/llvm-project/pull/152373