| /dpdk/app/test/ |
| H A D | test_member_perf.c | 62 unsigned int cycle; member 105 hashtest_key_lens[params->cycle]); in shuffle_input_keys() 107 hashtest_key_lens[params->cycle]); in shuffle_input_keys() 140 params->key_size = hashtest_key_lens[cycle]; in setup_keys_and_data() 141 params->cycle = cycle; in setup_keys_and_data() 251 false_data[type][params->cycle] = 0; in timed_lookups() 270 false_data[type][params->cycle]++; in timed_lookups() 290 false_data_bulk[type][params->cycle] = 0; in timed_lookups_bulk() 335 false_data_multi[type][params->cycle] = 0; in timed_lookups_multimatch() 456 false_hit[type][params->cycle] = 0; in timed_miss_lookup() [all …]
|
| H A D | test_efd_perf.c | 67 unsigned int cycle; member 104 memcpy(temp_key, keys[i], hashtest_key_lens[params->cycle]); in shuffle_input_keys() 107 memcpy(keys[i], keys[swap_idx], hashtest_key_lens[params->cycle]); in shuffle_input_keys() 110 memcpy(keys[swap_idx], temp_key, hashtest_key_lens[params->cycle]); in shuffle_input_keys() 133 params->key_size = hashtest_key_lens[cycle]; in setup_keys_and_data() 134 params->cycle = cycle; in setup_keys_and_data() 201 cycles[params->cycle][ADD] = time_taken / KEYS_TO_ADD; in timed_adds() 234 cycles[params->cycle][LOOKUP] = time_taken / NUM_LOOKUPS; in timed_lookups() 278 cycles[params->cycle][LOOKUP_MULTI] = time_taken / NUM_LOOKUPS; in timed_lookups_multi() 307 cycles[params->cycle][DELETE] = time_taken / KEYS_TO_ADD; in timed_deletes() [all …]
|
| /dpdk/doc/guides/prog_guide/ |
| H A D | profile_app.rst | 74 High-resolution cycle counter 78 clock counter is through the ARMv8 PMU subsystem. The PMU cycle counter runs 79 at CPU frequency. However, access to the PMU cycle counter from user space is 81 cycle counter for user space access by configuring the PMU from the privileged 87 The example below shows the steps to configure the PMU based cycle counter on
|
| H A D | service_cores.rst | 53 cycle count collection is dynamically configurable, allowing any application to
|
| H A D | rawdev.rst | 39 has different semantics than a start-stop-start cycle.
|
| H A D | trace_lib.rst | 38 Typical trace overhead is ~20 cycles and instrumentation overhead is 1 cycle.
|
| /dpdk/doc/guides/sample_app_ug/ |
| H A D | rxtx_callbacks.rst | 107 The DPDK function ``rte_rdtsc()`` is used to add a cycle count timestamp to 125 transmitted the average cycle count per packet is printed out and the counters
|
| /dpdk/doc/guides/tools/ |
| H A D | testbbdev.rst | 111 (a) *SW Enq Offload Cost*: Software only enqueue offload cost, the cycle 114 (b) *Acc Enq Offload Cost*: The cycle count and time (us) from the 117 (c) *SW Deq Offload Cost*: Software dequeue cost, the cycle counts and 119 (d) *Empty Queue Enq Offload Cost*: The cycle count and time (us)
|
| H A D | cryptoperf.rst | 19 On hardware devices the cycle-count doesn't always represent the actual offload 20 cost. The cycle-count only represents the offload cost when the hardware 22 offload are still consumed by the test tool and included in the cycle-count. 25 in a larger cycle-count measurement and should not be interpreted as an offload
|
| H A D | comp_perf.rst | 26 the trade-off between throughput and cycle-count.
|
| /dpdk/doc/guides/contributing/ |
| H A D | stable.rst | 41 The duration of a stable is one complete release cycle (4 months). It can be 44 for one release cycle.
|
| H A D | abi_policy.rst | 295 is then dropped for the duration of this release cycle. 297 and some amended rules apply during this cycle:
|
| H A D | patches.rst | 619 patch accepted. The general cycle for patch review and acceptance is: 674 concrete definitions and expectations for a typical release cycle. 675 An average cycle lasts 3 months and have 4 release candidates in the last month.
|
| H A D | abi_versioning.rst | 123 at the start of each release cycle, and are managed at the project level.
|
| H A D | documentation.rst | 77 Issues that are introduced and then fixed within a release cycle do not have to be included here.
|
| /dpdk/doc/guides/vdpadevs/ |
| H A D | mlx5.rst | 71 A nonzero value defines the traffic off time, in polling cycle time units,
|
| /dpdk/doc/guides/rel_notes/ |
| H A D | release_18_08.rst | 40 * Add low cycle count Tx handler for no-offload Tx. 41 * Add low cycle count Rx handler for non-scattered Rx.
|
| H A D | release_20_02.rst | 197 * **Added cycle-count mode to the compression performance tool.** 199 Enhanced the compression performance tool by adding a cycle-count mode
|
| H A D | release_20_05.rst | 18 instrumentation overhead is 1 cycle. Added tracepoints in ``EAL``,
|
| H A D | release_16_04.rst | 270 * Added CPU utilization measurement and idle cycle rate computation.
|
| H A D | known_issues.rst | 101 but specific to an lcore and is a cycle reference, not a time reference.
|
| H A D | release_2_1.rst | 521 * **eal/ppc: Fix cpu cycle count for little endian.**
|
| /dpdk/doc/guides/nics/ |
| H A D | dpaa2.rst | 478 of the packet pull command which is issued in the previous cycle.
|
| /dpdk/doc/guides/testpmd_app_ug/ |
| H A D | testpmd_funcs.rst | 1305 Set the cycle to flush the GROed packets from reassembly tables::
|