8e1aee62d0
## Motivation Resolved: SWDEV-566226 The current implementation of agents inside of rocprof-systems keeps just the minimal necessary set of information required for populating the `info_agent` table inside of rocpd database. There is a sufficient amount of data that is being left out from database, so this change should fix that and store the additional agent information as an `extdata` row inside of `info_agent` table. ## Technical Details This PR introduces additional filed inside of `agent` structure inside which is representing the JSON formatted string of all the additional information we can acquire about particular agent. This data is processed and added during the initial fetching of agents, and afterwards pushed inside of the database. --------- Co-authored-by: David Galiffi <David.Galiffi@amd.com> * SWDEV-557412 - Incorporate proper chunk offset when remapping virtual memory (#1848) * SWDEV-557412 - Incorporate proper offset when remapping virtual memory * Fix condition to check if VMHeap allocation address matches a chunk address * Move offset calculation outside if/else block --------- Co-authored-by: JeniferC99 <150404595+JeniferC99@users.noreply.github.com> * SWDEV-567852 - Clean-up hip::init() (#1948) * SWDEV-559267 - Use CLPrint to DevLogPrintf with Log Level - detail debug. (#1160) * SWDEV-548892 - Stop using ocml isinf wrapper (#1854) * SWDEV-562708 - change default maximum SVM size to 256GB (#1731) * SWDEV-503089 - Fix and enable disabled HIP tests from math group (#1319) * SWDEV-503089 - Fix and enable disabled HIP tests from math group * SWDEV-503089 - Move single precision reduced run to a common function * SWDEV-548892 - Stop using ockl steadyctr function (#1882) Directly use the builtin * Implement PTL support (#1957) * Implement PTL support Signed-off-by: adapryor <Adam.pryor@amd.com> (cherry picked from commit 45bc31292e7940a3b8fca044ef7df22047b95733) Signed-off-by: Maisam Arif <Maisam.Arif@amd.com> --------- Signed-off-by: adapryor <Adam.pryor@amd.com> Signed-off-by: Maisam Arif <Maisam.Arif@amd.com> Co-authored-by: Maisam Arif <Maisam.Arif@amd.com> * SWDEV-558080 - Add recommended granularity (#1176) * Add recommended granularity * Improve granularity testing * Update based on feedback * Fix and enable VMM tests on cuda (#1855) * Fix and enable VMM tests on cuda * Minor syntax fixes --------- Co-authored-by: Rahul Manocha <rmanocha@amd.com> * [rocprofiler-systems] Add support for ompt_callback_thread_begin (#1681) * Add thread_begin callback * Make OMPT callbacks that are instant have start_ts = end_ts * SWDEV-567514: Remove default stream wait (#1977) - when virtual map command is called - can create deadlock Signed-off-by: sdashmiz <shadi.dashmiz@amd.com> * Fix flaky test Unit_hipStreamAddCallback_StrmSyncTiming (#2022) * Review comments * skip the 3 failing tests to merge hip-tests rocm-systems PR --------- Signed-off-by: Bindhiya Kanangot Balakrishnan <Bindhiya.KanangotBalakrishnan@amd.com> Signed-off-by: adapryor <Adam.pryor@amd.com> Signed-off-by: Maisam Arif <Maisam.Arif@amd.com> Signed-off-by: sdashmiz <shadi.dashmiz@amd.com> Co-authored-by: GunaShekar <agunashe@amd.com> Co-authored-by: agunashe <ajay.gunashekar@amd.com> Co-authored-by: Ethan Trinh <Ethan.Trinh@amd.com> Co-authored-by: JeniferC99 <150404595+JeniferC99@users.noreply.github.com> Co-authored-by: Victor Zhang <111778801+victzhan@users.noreply.github.com> Co-authored-by: German Andryeyev <56892148+gandryey@users.noreply.github.com> Co-authored-by: usrihari123 <srihari.u@amd.com> Co-authored-by: Bindhiya Kanangot Balakrishnan <Bindhiya.KanangotBalakrishnan@amd.com> Co-authored-by: anujshuk-amd <anujshuk@amd.com> Co-authored-by: itrowbri <Ian.Trowbridge@amd.com> Co-authored-by: marantic-amd <marantic@amd.com> Co-authored-by: David Galiffi <David.Galiffi@amd.com> Co-authored-by: cadolphe-amd <chris.adolphe@amd.com> Co-authored-by: Karthik Jayaprakash <54370791+kjayapra-amd@users.noreply.github.com> Co-authored-by: Matt Arsenault <Matthew.Arsenault@amd.com> Co-authored-by: Todd tiantuo Li <88386084+lttamd@users.noreply.github.com> Co-authored-by: amilanov-amd <Aleksandar.Milanov@amd.com> Co-authored-by: Adam Pryor <61172547+adam360x@users.noreply.github.com> Co-authored-by: Maisam Arif <Maisam.Arif@amd.com> Co-authored-by: AidanBeltonS <abeltons@amd.com> Co-authored-by: Rahul Manocha <153310294+manocharahul@users.noreply.github.com> Co-authored-by: Rahul Manocha <rmanocha@amd.com> Co-authored-by: Kian Cossettini <Kian.Cossettini@amd.com> Co-authored-by: Shadi Dashmiz <94885391+shadidashmiz@users.noreply.github.com> Co-authored-by: Ioannis Assiouras <38722728+iassiour@users.noreply.github.com> Co-authored-by: Ajay GunaShekar <86270081+agunashe@users.noreply.github.com>
193 baris
7.5 KiB
C++
193 baris
7.5 KiB
C++
/*
|
|
Copyright (c) 2025 Advanced Micro Devices, Inc. All rights reserved.
|
|
Permission is hereby granted, free of charge, to any person obtaining a copy
|
|
of this software and associated documentation files (the "Software"), to deal
|
|
in the Software without restriction, including without limitation the rights
|
|
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
|
|
copies of the Software, and to permit persons to whom the Software is
|
|
furnished to do so, subject to the following conditions:
|
|
The above copyright notice and this permission notice shall be included in
|
|
all copies or substantial portions of the Software.
|
|
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
|
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
|
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
|
|
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
|
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
|
|
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN
|
|
THE SOFTWARE.
|
|
*/
|
|
#define HIP_ENABLE_WARP_SYNC_BUILTINS
|
|
#define HIP_ENABLE_EXTRA_WARP_SYNC_TYPES
|
|
|
|
#include "warp_common.hh"
|
|
#include <hip/hip_runtime.h>
|
|
#include <tuple>
|
|
#include <cmd_options.hh>
|
|
#include <functional>
|
|
|
|
#define NELEMS(array) (sizeof(array) / sizeof(array[0]))
|
|
|
|
// compiles the program, reusing the same compiling session for all the types
|
|
// (as opposed as calling the rtc compiler for each of the types)
|
|
template <template <typename> class Op, class T, typename... Types>
|
|
void compileProgram(hiprtcProgram& prog, const std::tuple<T, Types...>&) {
|
|
std::string scalarName, intrinsicName, expression;
|
|
std::tuple<Types...> remainingTypes;
|
|
|
|
expression = std::string("reduceRtcKernel<") + typeToString<T>() + ", unsigned long long>";
|
|
HIPRTC_CHECK(hiprtcAddNameExpression(prog, expression.c_str()));
|
|
compileProgram<Op>(prog, remainingTypes);
|
|
}
|
|
|
|
template <class T, class MaskType, template <typename> class Op>
|
|
void runRtcReduceOp(hiprtcProgram& prog, T* output, const T* input, const MaskType* masks,
|
|
int numReduces, Op<T>) {
|
|
unsigned int wavefrontSize = getWarpSize();
|
|
const char* loweredName;
|
|
hipFunction_t kernel;
|
|
hipModule_t module;
|
|
struct {
|
|
const T* d_output;
|
|
const T* d_input;
|
|
const MaskType* d_masks;
|
|
int numReduces;
|
|
} args{output, input, masks, numReduces};
|
|
int size = 4;
|
|
void* config[] = {HIP_LAUNCH_PARAM_BUFFER_POINTER, &args, HIP_LAUNCH_PARAM_BUFFER_SIZE, &size,
|
|
HIP_LAUNCH_PARAM_END};
|
|
std::vector<char> code;
|
|
size_t codeSize;
|
|
std::string expression =
|
|
std::string("reduceRtcKernel<") + typeToString<T>() + ", unsigned long long>";
|
|
dim3 grdDim{1u};
|
|
dim3 blkDim{wavefrontSize};
|
|
|
|
HIPRTC_CHECK(hiprtcGetCodeSize(prog, &codeSize));
|
|
code.resize(codeSize);
|
|
HIPRTC_CHECK(hiprtcGetCode(prog, code.data()));
|
|
HIP_CHECK(hipModuleLoadData(&module, code.data()));
|
|
HIPRTC_CHECK(hiprtcGetLoweredName(prog, expression.c_str(), &loweredName));
|
|
HIP_CHECK(hipModuleGetFunction(&kernel, module, loweredName));
|
|
HIP_CHECK(hipModuleLaunchKernel(kernel, grdDim.x, grdDim.y, grdDim.z, blkDim.x, blkDim.y,
|
|
blkDim.z, 0, 0, nullptr, config));
|
|
HIP_CHECK(hipModuleUnload(module));
|
|
}
|
|
|
|
template <template <typename> class Op, class Type = void>
|
|
void runTestReduceForTypes(hiprtcProgram&, const std::tuple<>) {}
|
|
|
|
template <template <typename> class Op, class T, typename... Types>
|
|
void runTestReduceForTypes(hiprtcProgram& prog, const std::tuple<T, Types...>) {
|
|
std::tuple<Types...> remainingTypes;
|
|
int iteration = 0;
|
|
|
|
auto reduceFunc = [&prog](T* d_output, const T* d_input, const unsigned long long* d_masks,
|
|
int numReduces, Op<T> op) {
|
|
runRtcReduceOp(prog, d_output, d_input, d_masks, numReduces, op);
|
|
};
|
|
|
|
while (iteration < cmd_options.reduce_iterations) {
|
|
runTestReduce<T, decltype(reduceFunc), Op>(iteration, reduceFunc);
|
|
iteration++;
|
|
|
|
if (cmd_options.reduce_iterations != 1) {
|
|
std::cout << "\rIteration: " << iteration;
|
|
std::flush(std::cout);
|
|
}
|
|
}
|
|
|
|
runTestReduceForTypes<Op>(prog, remainingTypes);
|
|
}
|
|
|
|
template <class T, template <typename> class Op>
|
|
void opToString(std::string& scalarName, std::string& intrinsicName) {
|
|
if constexpr (std::is_same<Op<T>, std::plus<T>>::value) {
|
|
scalarName = "std::plus";
|
|
intrinsicName = "__reduce_add_sync";
|
|
} else if constexpr (std::is_same<Op<T>, MinOp<T>>::value) {
|
|
scalarName = "MinOp";
|
|
intrinsicName = "__reduce_min_sync";
|
|
} else if constexpr (std::is_same<Op<T>, MaxOp<T>>::value) {
|
|
scalarName = "MaxOp";
|
|
intrinsicName = "__reduce_max_sync";
|
|
} else if constexpr (std::is_same<Op<T>, std::logical_and<T>>::value) {
|
|
scalarName = "std::logical_and";
|
|
intrinsicName = "__reduce_and_sync";
|
|
} else if constexpr (std::is_same<Op<T>, std::logical_or<T>>::value) {
|
|
scalarName = "std::logical_or";
|
|
intrinsicName = "__reduce_or_sync";
|
|
} else if constexpr (std::is_same<Op<T>, XorOp<T>>::value) {
|
|
scalarName = "LogicalXor";
|
|
intrinsicName = "__reduce_xor_sync";
|
|
} else
|
|
static_assert(std::is_void<T>::value, "Unexpected operator");
|
|
}
|
|
|
|
template <template <typename> class Op, class T = void>
|
|
void compileProgram(hiprtcProgram& prog, const std::tuple<>&) {
|
|
size_t logSize;
|
|
std::string scalarName, intrinsicName;
|
|
hiprtcResult compileResult;
|
|
const char* options[] = {"-DHIP_ENABLE_WARP_SYNC_BUILTINS", "-DHIP_ENABLE_EXTRA_WARP_SYNC_TYPES"};
|
|
|
|
opToString<int, Op>(scalarName, intrinsicName);
|
|
compileResult = hiprtcResult{hiprtcCompileProgram(prog, NELEMS(options), options)};
|
|
HIPRTC_CHECK(hiprtcGetProgramLogSize(prog, &logSize));
|
|
|
|
if (compileResult != HIPRTC_SUCCESS || logSize > 0) {
|
|
std::string log(logSize, '\0');
|
|
|
|
HIPRTC_CHECK(hiprtcGetProgramLog(prog, &log[0]));
|
|
std::cerr << "Runtime compilation failed or contained warnings for operator: " << scalarName
|
|
<< " associated reduce function: " << intrinsicName << "\n";
|
|
std::cerr << log << '\n';
|
|
REQUIRE(false);
|
|
}
|
|
}
|
|
|
|
template <template <typename> class Op, typename... Types>
|
|
void runAndCompileTest(const std::tuple<Types...> types) {
|
|
std::string scalarName, intrinsicName, kernelStr;
|
|
hiprtcProgram prog;
|
|
|
|
opToString<int, Op>(scalarName, intrinsicName);
|
|
kernelStr = R"(
|
|
template <class T, class MaskType>
|
|
__global__ void reduceRtcKernel(T* output, const T* input, const MaskType* masks, int numReduces)
|
|
{
|
|
int tid = threadIdx.x;
|
|
|
|
for (int i = 0; i < numReduces; i++) {
|
|
if (masks[i] & (1ul << tid)) {
|
|
// call the operator only if the lane is mentioned in the mask
|
|
T& result = output[warpSize * i + tid];
|
|
result = )" +
|
|
intrinsicName + R"((masks[i], input[tid]);
|
|
}
|
|
}
|
|
})";
|
|
|
|
HIPRTC_CHECK(
|
|
hiprtcCreateProgram(&prog, kernelStr.c_str(), "warp_reduce.hip", 0, nullptr, nullptr));
|
|
compileProgram<Op>(prog, types);
|
|
runTestReduceForTypes<Op>(prog, types);
|
|
HIPRTC_CHECK(hiprtcDestroyProgram(&prog));
|
|
}
|
|
|
|
TEST_CASE("Unit_Rtc_ReduceRandom") {
|
|
const std::tuple<int, unsigned int, long long, unsigned long long, float, half, double> allTypes;
|
|
const std::tuple<int, unsigned int, long long, unsigned long long> integralTypes;
|
|
|
|
SECTION("add") { runAndCompileTest<std::plus>(allTypes); }
|
|
|
|
SECTION("min") { runAndCompileTest<MinOp>(allTypes); }
|
|
|
|
SECTION("max") { runAndCompileTest<MaxOp>(allTypes); }
|
|
|
|
SECTION("and") { runAndCompileTest<std::logical_and>(integralTypes); }
|
|
|
|
SECTION("or") { runAndCompileTest<std::logical_or>(integralTypes); }
|
|
|
|
SECTION("xor") { runAndCompileTest<XorOp>(integralTypes); }
|
|
}
|