[SYCL] Remove XFAIL for AMD from group collectives related tests

SYCL/GroupAlgorithm/SYCL2020/all_of.cpp

@@ -2,9 +2,6 @@
 // RUN: %CPU_RUN_PLACEHOLDER %t.out
 // RUN: %GPU_RUN_PLACEHOLDER %t.out
 // RUN: %ACC_RUN_PLACEHOLDER %t.out
-//
-// Missing __spirv_GroupAll on AMD:
-// XFAIL: hip_amd
 
 #include "support.h"
 #include <CL/sycl.hpp>

SYCL/GroupAlgorithm/SYCL2020/any_of.cpp

@@ -2,9 +2,6 @@
 // RUN: %CPU_RUN_PLACEHOLDER %t.out
 // RUN: %GPU_RUN_PLACEHOLDER %t.out
 // RUN: %ACC_RUN_PLACEHOLDER %t.out
-//
-// Missing __spirv_GroupAny on AMD:
-// XFAIL: hip_amd
 
 #include "support.h"
 #include <CL/sycl.hpp>

SYCL/GroupAlgorithm/SYCL2020/exclusive_scan.cpp

@@ -2,10 +2,6 @@
 // RUN: %CPU_RUN_PLACEHOLDER %t.out
 // RUN: %GPU_RUN_PLACEHOLDER %t.out
 // RUN: %ACC_RUN_PLACEHOLDER %t.out
-//
-// Missing __spirv_GroupIAdd, __spirv_GroupBroadcast, __spirv_GroupSMin and
-// __spirv_GroupSMax on AMD:
-// XFAIL: hip_amd
 
 // TODO: enable compile+runtime checks for operations defined in SPIR-V 1.3.
 // That requires either adding a switch to clang (-spirv-max-version=1.3) or

SYCL/GroupAlgorithm/SYCL2020/group_broadcast.cpp

@@ -2,9 +2,6 @@
 // RUN: %CPU_RUN_PLACEHOLDER %t.out
 // RUN: %GPU_RUN_PLACEHOLDER %t.out
 // RUN: %ACC_RUN_PLACEHOLDER %t.out
-//
-// Missing __spirv_GroupBroadcast on AMD:
-// XFAIL: hip_amd
 
 #include "support.h"
 #include <CL/sycl.hpp>

SYCL/GroupAlgorithm/SYCL2020/inclusive_scan.cpp

@@ -2,10 +2,6 @@
 // RUN: %CPU_RUN_PLACEHOLDER %t.out
 // RUN: %GPU_RUN_PLACEHOLDER %t.out
 // RUN: %ACC_RUN_PLACEHOLDER %t.out
-//
-// Missing __spirv_GroupIAdd, __spirv_GroupBroadcast, __spirv_GroupSMin and
-// __spirv_GroupSMax on AMD:
-// XFAIL: hip_amd
 
 // TODO: enable compile+runtime checks for operations defined in SPIR-V 1.3.
 // That requires either adding a switch to clang (-spirv-max-version=1.3) or

SYCL/GroupAlgorithm/SYCL2020/none_of.cpp

@@ -2,9 +2,6 @@
 // RUN: %CPU_RUN_PLACEHOLDER %t.out
 // RUN: %GPU_RUN_PLACEHOLDER %t.out
 // RUN: %ACC_RUN_PLACEHOLDER %t.out
-//
-// Missing __spirv_GroupAll and __spirv_GroupAny on AMD:
-// XFAIL: hip_amd
 
 #include "support.h"
 #include <CL/sycl.hpp>

SYCL/GroupAlgorithm/SYCL2020/reduce.cpp

@@ -2,9 +2,6 @@
 // RUN: %CPU_RUN_PLACEHOLDER %t.out
 // RUN: %GPU_RUN_PLACEHOLDER %t.out
 // RUN: %ACC_RUN_PLACEHOLDER %t.out
-//
-// Missinsg __spirv_GroupIAdd, __spirv_GroupSMin and __spirv_GroupSMax on AMD:
-// XFAIL: hip_amd
 
 // TODO: enable compile+runtime checks for operations defined in SPIR-V 1.3.
 // That requires either adding a switch to clang (-spirv-max-version=1.3) or

SYCL/GroupAlgorithm/all_of.cpp

@@ -3,9 +3,6 @@
 // RUN: %CPU_RUN_PLACEHOLDER %t.out
 // RUN: %GPU_RUN_PLACEHOLDER %t.out
 // RUN: %ACC_RUN_PLACEHOLDER %t.out
-//
-// Missing __spirv_GroupAll on AMD
-// XFAIL: hip_amd
 
 #include "support.h"
 #include <CL/sycl.hpp>

SYCL/GroupAlgorithm/any_of.cpp

@@ -3,9 +3,6 @@
 // RUN: %CPU_RUN_PLACEHOLDER %t.out
 // RUN: %GPU_RUN_PLACEHOLDER %t.out
 // RUN: %ACC_RUN_PLACEHOLDER %t.out
-//
-// Missing __spirv_GroupAny on AMD
-// XFAIL: hip_amd
 
 #include "support.h"
 #include <CL/sycl.hpp>

SYCL/GroupAlgorithm/back_to_back_collectives.cpp

@@ -3,9 +3,6 @@
 // RUN: %CPU_RUN_PLACEHOLDER %t.out
 // RUN: %GPU_RUN_PLACEHOLDER %t.out
 // RUN: %ACC_RUN_PLACEHOLDER %t.out
-//
-// Missing __spirv_GroupIAdd on AMD:
-// XFAIL: hip_amd
 
 #include <CL/sycl.hpp>
 #include <numeric>

SYCL/GroupAlgorithm/broadcast.cpp

@@ -3,9 +3,6 @@
 // RUN: %CPU_RUN_PLACEHOLDER %t.out
 // RUN: %GPU_RUN_PLACEHOLDER %t.out
 // RUN: %ACC_RUN_PLACEHOLDER %t.out
-//
-// Missing __spirv_GroupBroadcast on AMD
-// XFAIL: hip_amd
 
 #include "support.h"
 #include <CL/sycl.hpp>

SYCL/GroupAlgorithm/exclusive_scan.cpp

@@ -3,10 +3,6 @@
 // RUN: %CPU_RUN_PLACEHOLDER %t.out
 // RUN: %GPU_RUN_PLACEHOLDER %t.out
 // RUN: %ACC_RUN_PLACEHOLDER %t.out
-//
-// Missing __spirv_GroupIAdd, __spirv_GroupBroadcast, __spirv_GroupSMin,
-// __spirv_GroupSMax on AMD
-// XFAIL: hip_amd
 
 // TODO: enable compile+runtime checks for operations defined in SPIR-V 1.3.
 // That requires either adding a switch to clang (-spirv-max-version=1.3) or

SYCL/GroupAlgorithm/inclusive_scan.cpp

@@ -3,10 +3,6 @@
 // RUN: %CPU_RUN_PLACEHOLDER %t.out
 // RUN: %GPU_RUN_PLACEHOLDER %t.out
 // RUN: %ACC_RUN_PLACEHOLDER %t.out
-//
-// Missing __spirv_GroupIAdd, __spirv_GroupBroadcast, __spirv_GroupSMin,
-// __spirv_GroupSMax on AMD
-// XFAIL: hip_amd
 
 // TODO: enable compile+runtime checks for operations defined in SPIR-V 1.3.
 // That requires either adding a switch to clang (-spirv-max-version=1.3) or

SYCL/GroupAlgorithm/none_of.cpp

@@ -3,9 +3,6 @@
 // RUN: %CPU_RUN_PLACEHOLDER %t.out
 // RUN: %GPU_RUN_PLACEHOLDER %t.out
 // RUN: %ACC_RUN_PLACEHOLDER %t.out
-//
-// Missing __spirv_GroupAll, __spirv_GroupAny on AMD
-// XFAIL: hip_amd
 
 #include "support.h"
 #include <CL/sycl.hpp>

SYCL/GroupAlgorithm/reduce.cpp

@@ -3,9 +3,6 @@
 // RUN: %CPU_RUN_PLACEHOLDER %t.out
 // RUN: %GPU_RUN_PLACEHOLDER %t.out
 // RUN: %ACC_RUN_PLACEHOLDER %t.out
-//
-// Missing __spirv_GroupIAdd, __spirv_GroupSMin, __spirv_GroupSMax on AMD
-// XFAIL: hip_amd
 
 // TODO: enable compile+runtime checks for operations defined in SPIR-V 1.3.
 // That requires either adding a switch to clang (-spirv-max-version=1.3) or

SYCL/Reduction/reduction_nd_conditional.cpp

@@ -4,10 +4,10 @@
 // RUN: %GPU_RUN_PLACEHOLDER %t.out
 // RUN: %ACC_RUN_PLACEHOLDER %t.out
 //
-// Missing __spirv_GroupIAdd on AMD, error message `The implementation handling
+// Error message `The implementation handling
 // parallel_for with reduction requires work group size not bigger than 1` on
 // Nvidia.
-// XFAIL: hip_amd || hip_nvidia
+// XFAIL: hip_nvidia
 
 // This test performs basic checks of parallel_for(nd_range, reduction, func)
 // with reduction and conditional increment of the reduction variable.

SYCL/Reduction/reduction_nd_ext_double.cpp

@@ -3,10 +3,10 @@
 // RUN: %GPU_RUN_PLACEHOLDER %t.out
 // RUN: %ACC_RUN_PLACEHOLDER %t.out
 //
-// Missing __spirv_GroupFAdd, __spirv_GroupFMin, __spirv_GroupFMax on AMD, error
+// Error
 // message `The implementation handling parallel_for with reduction requires
 // work group size not bigger than 1` on Nvidia.
-// XFAIL: hip_amd || hip_nvidia
+// XFAIL: hip_nvidia
 
 // TODO: Enable the test for HOST when it supports intel::reduce() and barrier()
 // RUNx: %HOST_RUN_PLACEHOLDER %t.out

SYCL/Reduction/reduction_range_usm_dw.cpp

@@ -3,9 +3,9 @@
 // RUN: %GPU_RUN_PLACEHOLDER %t.out
 // RUN: %ACC_RUN_PLACEHOLDER %t.out
 //
-// Missing __spirv_GroupFAdd on AMD, error message `Group algorithms are not
+// Error message `Group algorithms are not
 // supported on host device.` on Nvidia.
-// XFAIL: hip_amd || hip_nvidia
+// XFAIL: hip_nvidia
 
 // TODO: test disabled due to sporadic fails in level_zero:gpu RT.
 // UNSUPPORTED: linux && level_zero

SYCL/SubGroup/broadcast.cpp

@@ -3,9 +3,6 @@
 // RUN: %CPU_RUN_PLACEHOLDER %t.out
 // RUN: %GPU_RUN_PLACEHOLDER %t.out
 // RUN: %ACC_RUN_PLACEHOLDER %t.out
-//
-// Missing __spirv_GroupBroadcast on AMD
-// XFAIL: hip_amd
 
 //==--------- broadcast.cpp - SYCL sub_group broadcast test ----*- C++ -*---==//
 //

SYCL/SubGroup/broadcast_fp16.cpp

@@ -1,8 +1,5 @@
 // RUN: %clangxx -fsycl -fsycl-targets=%sycl_triple %s -o %t.out
 // RUN: %GPU_RUN_PLACEHOLDER %t.out
-//
-// Missing __spirv_GroupBroadcast on AMD
-// XFAIL: hip_amd
 
 //==---- broadcast_fp16.cpp - SYCL sub_group broadcast test ----*- C++ -*---==//
 //

SYCL/SubGroup/broadcast_fp64.cpp

@@ -3,9 +3,6 @@
 // RUN: %CPU_RUN_PLACEHOLDER %t.out
 // RUN: %GPU_RUN_PLACEHOLDER %t.out
 // RUN: %ACC_RUN_PLACEHOLDER %t.out
-//
-// Missing __spirv_GroupBroadcast on AMD
-// XFAIL: hip_amd
 
 //==----- broadcast_fp64.cpp - SYCL sub_group broadcast test ----*- C++ -*--==//
 //

SYCL/SubGroup/reduce.cpp

@@ -7,12 +7,7 @@
 // RUN: %CPU_RUN_PLACEHOLDER %t.out
 // RUN: %GPU_RUN_PLACEHOLDER %t.out
 // RUN: %ACC_RUN_PLACEHOLDER %t.out
-//
-// Missing __spirv_GroupIAdd, __spirv_GroupSMin, __spirv_GroupSMax,
-// __spirv_GroupUMin, __spirv_GroupUMax, __spirv_GroupFAdd, __spirv_GroupFMin,
-// __spirv_GroupFMax on AMD
-// XFAIL: hip_amd
-//
+
 //==--------------- reduce.cpp - SYCL sub_group reduce test ----*- C++ -*---==//
 //
 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.

SYCL/SubGroup/reduce_fp64.cpp

@@ -7,9 +7,6 @@
 // RUN: %CPU_RUN_PLACEHOLDER %t.out
 // RUN: %GPU_RUN_PLACEHOLDER %t.out
 // RUN: %ACC_RUN_PLACEHOLDER %t.out
-//
-// Missing __spirv_GroupFAdd, __spirv_GroupFMin, __spirv_GroupFMax on AMD
-// XFAIL: hip_amd
 
 // This test verifies the correct work of the sub-group algorithm reduce().
 

SYCL/SubGroup/scan.cpp

@@ -7,11 +7,6 @@
 // RUN: %CPU_RUN_PLACEHOLDER %t.out
 // RUN: %GPU_RUN_PLACEHOLDER %t.out
 // RUN: %ACC_RUN_PLACEHOLDER %t.out
-//
-// Missing __spirv_SubgroupLocalInvocationId, __spirv_GroupIAdd,
-// __spirv_GroupSMin, __spirv_GroupSMax, __spirv_GroupUMin, __spirv_groupUMax,
-// __spirv_GroupFAdd, __spirv_GroupFMin, __spirv_GroupFMax on AMD
-// XFAIL: hip_amd
 
 //==--------------- scan.cpp - SYCL sub_group scan test --------*- C++ -*---==//
 //

SYCL/SubGroup/scan_fp64.cpp

@@ -7,10 +7,6 @@
 // RUN: %CPU_RUN_PLACEHOLDER %t.out
 // RUN: %GPU_RUN_PLACEHOLDER %t.out
 // RUN: %ACC_RUN_PLACEHOLDER %t.out
-//
-// Missing __spirv_SubgroupLocalInvocationId, __spirv_GroupFAdd,
-// __spirv_GroupFMin, __spirv_GroupFMax on AMD
-// XFAIL: hip_amd
 
 // This test verifies the correct work of the sub-group algorithms
 // exclusive_scan() and inclusive_scan().

SYCL/SubGroup/vote.cpp

@@ -3,10 +3,6 @@
 // RUN: %CPU_RUN_PLACEHOLDER %t.out
 // RUN: %GPU_RUN_PLACEHOLDER %t.out
 // RUN: %ACC_RUN_PLACEHOLDER %t.out
-//
-// Missing __spirv_SubgroupLocalInvocationId, __spirv_GroupAny, __spirv_GroupAll
-// on AMD
-// XFAIL: hip_amd
 
 //==--------------- vote.cpp - SYCL sub_group vote test --*- C++ -*---------==//
 //