refactor - update of clang-tidy project file. automatic clang-tidy refactors

master
Francisco 4 years ago committed by Francisco Paisana
parent fb48f4484c
commit d5750f6fe6

@ -7,6 +7,7 @@
# - init of static memory may cause an exception (cert-err58)
# - forbidden implicit conversion from pointer/int to bool
# - recommended auto
# - remove llvm-specific checks (header guard style, usage of llvm namespace, restriction of libc includes, etc.)
# Naming conventions set to snake_case
Checks: '*,-fuchsia-*,
-cppcoreguidelines-pro-type-vararg,-hicpp-vararg,
@ -14,12 +15,15 @@ Checks: '*,-fuchsia-*,
-cppcoreguidelines-pro-bounds-array-to-pointer-decay,-hicpp-no-array-decay,
-cppcoreguidelines-pro-bounds-constant-array-index,-cppcoreguidelines-pro-type-cstyle-cast,
-cppcoreguidelines-pro-type-union-access,
-cppcoreguidelines-pro-type-static-cast-downcast,
-modernize-use-using,-modernize-use-trailing-return-type,
-modernize-use-auto,-hicpp-use-auto,
-llvmlibc-callee-namespace,
-llvmlibc-callee-namespace,-llvmlibc-implementation-in-namespace,-llvmlibc-restrict-system-libc-headers,
-llvm-header-guard,
-google-runtime-references,-google-readability-casting,-google-build-using-namespace,
google-default-arguments,-cppcoreguidelines-pro-bounds-pointer-arithmetic,
-cert-err58-cpp,
-readability-function-cognitive-complexity,-readability-isolate-declaration,
-misc-non-private-member-variables-in-classes,-altera-struct-pack-align,-readability-uppercase-literal-suffix,
-cppcoreguidelines-non-private-member-variables-in-classes,
readability-identifier-naming'

@ -53,7 +53,7 @@ class sched_tester : public sched_sim_base
static std::vector<sched_interface::cell_cfg_t> get_cell_cfg(srslte::span<const sched_cell_params_t> cell_params)
{
std::vector<sched_interface::cell_cfg_t> cell_cfg_list;
for (auto& c : cell_params) {
for (const auto& c : cell_params) {
cell_cfg_list.push_back(c.cfg);
}
return cell_cfg_list;
@ -114,9 +114,9 @@ public:
sched_ptr->dl_rlc_buffer_state(ue_ctxt.rnti, 3, ul_bytes_per_tti, 0);
if (get_tti_rx().to_uint() % 5 == 0) {
for (uint32_t cc = 0; cc < pending_events.cc_list.size(); ++cc) {
pending_events.cc_list[cc].dl_cqi = current_run_params.cqi;
pending_events.cc_list[cc].ul_snr = 40;
for (auto& cc : pending_events.cc_list) {
cc.dl_cqi = current_run_params.cqi;
cc.ul_snr = 40;
}
}
}
@ -126,18 +126,18 @@ public:
{
for (uint32_t cc = 0; cc < get_cell_params().size(); ++cc) {
uint32_t dl_tbs = 0, ul_tbs = 0, dl_mcs = 0, ul_mcs = 0;
for (uint32_t i = 0; i < sf_out.dl_cc_result[cc].data.size(); ++i) {
dl_tbs += sf_out.dl_cc_result[cc].data[i].tbs[0];
dl_tbs += sf_out.dl_cc_result[cc].data[i].tbs[1];
dl_mcs = std::max(dl_mcs, sf_out.dl_cc_result[cc].data[i].dci.tb[0].mcs_idx);
for (const auto& data : sf_out.dl_cc_result[cc].data) {
dl_tbs += data.tbs[0];
dl_tbs += data.tbs[1];
dl_mcs = std::max(dl_mcs, data.dci.tb[0].mcs_idx);
}
total_stats.mean_dl_tbs.push(dl_tbs);
if (sf_out.dl_cc_result[cc].data.size() > 0) {
if (not sf_out.dl_cc_result[cc].data.empty()) {
total_stats.avg_dl_mcs.push(dl_mcs);
}
for (uint32_t i = 0; i < sf_out.ul_cc_result[cc].pusch.size(); ++i) {
ul_tbs += sf_out.ul_cc_result[cc].pusch[i].tbs;
ul_mcs = std::max(ul_mcs, sf_out.ul_cc_result[cc].pusch[i].dci.tb.mcs_idx);
for (const auto& pusch : sf_out.ul_cc_result[cc].pusch) {
ul_tbs += pusch.tbs;
ul_mcs = std::max(ul_mcs, pusch.dci.tb.mcs_idx);
}
total_stats.mean_ul_tbs.push(ul_tbs);
if (not sf_out.ul_cc_result[cc].pusch.empty()) {
@ -147,38 +147,6 @@ public:
}
};
int run_sched_new_ue(sched_tester& sched_tester,
const run_params& params,
uint16_t rnti,
const sched_interface::ue_cfg_t& ue_cfg)
{
const uint32_t ENB_CC_IDX = 0;
sched_tester.total_stats = {};
sched_tester.current_run_params = params;
// Add user (first need to advance to a PRACH TTI)
while (not srslte_prach_tti_opportunity_config_fdd(
sched_tester.get_cell_params()[ue_cfg.supported_cc_list[0].enb_cc_idx].cfg.prach_config,
sched_tester.get_tti_rx().to_uint(),
-1)) {
TESTASSERT(sched_tester.advance_tti() == SRSLTE_SUCCESS);
}
TESTASSERT(sched_tester.add_user(rnti, ue_cfg, 16) == SRSLTE_SUCCESS);
// Ignore stats of the first TTIs until UE DRB1 is added
while (not sched_tester.get_enb_ctxt().ue_db.at(rnti)->conres_rx) {
sched_tester.advance_tti();
}
sched_tester.total_stats = {};
for (uint32_t count = 0; count < params.nof_ttis; ++count) {
sched_tester.advance_tti();
}
return SRSLTE_SUCCESS;
}
struct run_data {
run_params params;
float avg_dl_throughput;
@ -233,8 +201,8 @@ int run_benchmark_scenario(run_params params, std::vector<run_data>& run_results
run_data run_result = {};
run_result.params = params;
run_result.avg_dl_throughput = tester.total_stats.mean_dl_tbs.value() * 8.0 / 1e-3;
run_result.avg_ul_throughput = tester.total_stats.mean_ul_tbs.value() * 8.0 / 1e-3;
run_result.avg_dl_throughput = tester.total_stats.mean_dl_tbs.value() * 8.0F / 1e-3F;
run_result.avg_ul_throughput = tester.total_stats.mean_ul_tbs.value() * 8.0F / 1e-3F;
run_result.avg_dl_mcs = tester.total_stats.avg_dl_mcs.value();
run_result.avg_ul_mcs = tester.total_stats.avg_ul_mcs.value();
run_result.avg_latency = std::chrono::microseconds(static_cast<int>(tester.total_stats.avg_latency.value() / 1000));
@ -249,12 +217,12 @@ run_data expected_run_result(run_params params)
run_data ret{};
int tbs_idx = srslte_ra_tbs_idx_from_mcs(28, false, false);
int tbs = srslte_ra_tbs_from_idx(tbs_idx, params.nof_prbs);
ret.avg_dl_throughput = tbs * 1e3; // bps
ret.avg_dl_throughput = static_cast<float>(tbs) * 1e3F; // bps
tbs_idx = srslte_ra_tbs_idx_from_mcs(28, false, true);
uint32_t nof_pusch_prbs = params.nof_prbs - (params.nof_prbs == 6 ? 2 : 4);
tbs = srslte_ra_tbs_from_idx(tbs_idx, nof_pusch_prbs);
ret.avg_ul_throughput = tbs * 1e3; // bps
ret.avg_ul_throughput = static_cast<float>(tbs) * 1e3F; // bps
ret.avg_dl_mcs = 27;
ret.avg_ul_mcs = 22;
@ -288,11 +256,11 @@ void print_benchmark_results(const std::vector<run_data>& run_results)
int tbs_idx = srslte_ra_tbs_idx_from_mcs(28, false, false);
int tbs = srslte_ra_tbs_from_idx(tbs_idx, r.params.nof_prbs);
float dl_rate_overhead = 1.0F - r.avg_dl_throughput / (tbs * 1e3);
float dl_rate_overhead = 1.0F - r.avg_dl_throughput / (static_cast<float>(tbs) * 1e3F);
tbs_idx = srslte_ra_tbs_idx_from_mcs(28, false, true);
uint32_t nof_pusch_prbs = r.params.nof_prbs - (r.params.nof_prbs == 6 ? 2 : 4);
tbs = srslte_ra_tbs_from_idx(tbs_idx, nof_pusch_prbs);
float ul_rate_overhead = 1.0F - r.avg_ul_throughput / (tbs * 1e3);
float ul_rate_overhead = 1.0F - r.avg_ul_throughput / (static_cast<float>(tbs) * 1e3F);
fmt::print("{:>3d}{:>6d}{:>6d}{:>12}{:>6d}{:>9.2}/{:>4.2}{:>9.1f}/{:>4.1f}{:9.1f}/{:>4.1f}{:12d}\n",
i,
@ -420,7 +388,7 @@ int main(int argc, char* argv[])
}
auto* spy = static_cast<srslte::log_sink_spy*>(srslog::find_sink(srslte::log_sink_spy::name()));
if (!spy) {
if (spy == nullptr) {
return SRSLTE_ERROR;
}

@ -92,7 +92,7 @@ int test_pdcch_one_ue()
TESTASSERT(pdcch_result[0]->rnti == sched_ue.get_rnti());
TESTASSERT(pdcch_result[0]->total_mask.size() == cell_params[ENB_CC_IDX].nof_cce_table[pdcch.get_cfi() - 1]);
TESTASSERT(pdcch_result[0]->current_mask == pdcch_result[0]->total_mask);
TESTASSERT(pdcch_result[0]->current_mask.count() == 1u << aggr_idx);
TESTASSERT(pdcch_result[0]->current_mask.count() == 1U << aggr_idx);
TESTASSERT(std::count(dci_locs.begin(), dci_locs.end(), pdcch_result[0]->dci_pos.ncce) > 0);
// allocate UL user
@ -120,7 +120,7 @@ int test_pdcch_one_ue()
TESTASSERT(pdcch_result[1]->rnti == sched_ue.get_rnti());
TESTASSERT(pdcch_result[1]->total_mask.size() == cell_params[ENB_CC_IDX].nof_cce_table[pdcch.get_cfi() - 1]);
TESTASSERT((pdcch_result[1]->current_mask & pdcch_result[0]->current_mask).none());
TESTASSERT(pdcch_result[1]->current_mask.count() == 1u << aggr_idx);
TESTASSERT(pdcch_result[1]->current_mask.count() == 1U << aggr_idx);
TESTASSERT(pdcch_result[1]->total_mask == (pdcch_result[0]->current_mask | pdcch_result[1]->current_mask));
TESTASSERT(std::count(dci_locs2.begin(), dci_locs2.end(), pdcch_result[0]->dci_pos.ncce) > 0);
@ -133,7 +133,6 @@ int test_pdcch_one_ue()
int test_pdcch_ue_and_sibs()
{
const uint32_t ENB_CC_IDX = 0;
// Params
uint32_t nof_prb = 100;

@ -45,7 +45,7 @@ int test_pdu_alloc_successful(srsenb::lch_ue_manager& lch_handler,
int test_retx_until_empty(srsenb::lch_ue_manager& lch_handler, int lcid, uint32_t rlc_payload_size)
{
int start_rlc_bytes = lch_handler.get_dl_retx(lcid);
int nof_pdus = ceil(start_rlc_bytes / (float)rlc_payload_size);
int nof_pdus = ceil(static_cast<float>(start_rlc_bytes) / static_cast<float>(rlc_payload_size));
int rem_rlc_bytes = start_rlc_bytes;
sched_interface::dl_sched_pdu_t pdu;
@ -61,7 +61,7 @@ int test_retx_until_empty(srsenb::lch_ue_manager& lch_handler, int lcid, uint32_
int test_newtx_until_empty(srsenb::lch_ue_manager& lch_handler, int lcid, uint32_t rlc_payload_size)
{
int start_rlc_bytes = lch_handler.get_dl_tx(lcid);
int nof_pdus = ceil(start_rlc_bytes / (float)rlc_payload_size);
int nof_pdus = ceil(static_cast<float>(start_rlc_bytes) / (float)rlc_payload_size);
int rem_rlc_bytes = start_rlc_bytes;
sched_interface::dl_sched_pdu_t pdu;

Loading…
Cancel
Save