1; RUN: llvm-split -o %t %s -j 3 -mtriple amdgcn-amd-amdhsa -amdgpu-module-splitting-max-depth=0 -amdgpu-module-splitting-large-threshold=1.2 -amdgpu-module-splitting-merge-threshold=0.5 2; RUN: llvm-dis -o - %t0 | FileCheck --check-prefix=CHECK0 --implicit-check-not=define %s 3; RUN: llvm-dis -o - %t1 | FileCheck --check-prefix=CHECK1 --implicit-check-not=define %s 4; RUN: llvm-dis -o - %t2 | FileCheck --check-prefix=CHECK2 --implicit-check-not=define %s 5 6; RUN: llvm-split -o %t.nolarge %s -j 3 -mtriple amdgcn-amd-amdhsa -amdgpu-module-splitting-large-threshold=0 -amdgpu-module-splitting-max-depth=0 7; RUN: llvm-dis -o - %t.nolarge0 | FileCheck --check-prefix=NOLARGEKERNELS-CHECK0 --implicit-check-not=define %s 8; RUN: llvm-dis -o - %t.nolarge1 | FileCheck --check-prefix=NOLARGEKERNELS-CHECK1 --implicit-check-not=define %s 9; RUN: llvm-dis -o - %t.nolarge2 | FileCheck --check-prefix=NOLARGEKERNELS-CHECK2 --implicit-check-not=define %s 10 11; 2 kernels (A/B) are large and share all their dependencies. 12; They should go in the same partition, the remaining kernel should 13; go somewhere else, and one partition should be empty. 14; 15; Also check w/o large kernels processing to verify they are indeed handled 16; differently. 17 18; P0 is empty 19; CHECK0: declare 20 21; CHECK1: define internal void @HelperC() 22; CHECK1: define weak_odr amdgpu_kernel void @C 23 24; CHECK2: define internal void @large2() 25; CHECK2: define internal void @large1() 26; CHECK2: define internal void @large0() 27; CHECK2: define internal void @HelperA() 28; CHECK2: define internal void @HelperB() 29; CHECK2: define amdgpu_kernel void @A 30; CHECK2: define weak_odr amdgpu_kernel void @B 31 32; NOLARGEKERNELS-CHECK0: define internal void @HelperC() 33; NOLARGEKERNELS-CHECK0: define weak_odr amdgpu_kernel void @C 34 35; NOLARGEKERNELS-CHECK1: define internal void @large2() 36; NOLARGEKERNELS-CHECK1: define internal void @large1() 37; NOLARGEKERNELS-CHECK1: define internal void @large0() 38; NOLARGEKERNELS-CHECK1: define internal void @HelperB() 39; NOLARGEKERNELS-CHECK1: define weak_odr amdgpu_kernel void @B 40 41; NOLARGEKERNELS-CHECK2: define internal void @large2() 42; NOLARGEKERNELS-CHECK2: define internal void @large1() 43; NOLARGEKERNELS-CHECK2: define internal void @large0() 44; NOLARGEKERNELS-CHECK2: define internal void @HelperA() 45; NOLARGEKERNELS-CHECK2: define amdgpu_kernel void @A 46 47 48define internal void @large2() { 49 store volatile i32 42, ptr null 50 call void @large2() 51 ret void 52} 53 54define internal void @large1() { 55 call void @large1() 56 call void @large2() 57 ret void 58} 59 60define internal void @large0() { 61 call void @large0() 62 call void @large1() 63 call void @large2() 64 ret void 65} 66 67define internal void @HelperA() { 68 call void @large0() 69 ret void 70} 71 72define internal void @HelperB() { 73 call void @large0() 74 ret void 75} 76 77define amdgpu_kernel void @A() { 78 call void @HelperA() 79 ret void 80} 81 82define weak_odr amdgpu_kernel void @B() { 83 call void @HelperB() 84 ret void 85} 86 87define internal void @HelperC() { 88 ret void 89} 90 91define weak_odr amdgpu_kernel void @C() { 92 call void @HelperC() 93 ret void 94} 95