* [dpdk-dev] rte_get_tsc_cycles function slowly increase ?
@ 2021-07-02 9:36 =?gb18030?B?0KHP6dfT?=
0 siblings, 0 replies; only message in thread
From: =?gb18030?B?0KHP6dfT?= @ 2021-07-02 9:36 UTC (permalink / raw)
To: =?gb18030?B?ZGV2?=
[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain; charset="gb18030", Size: 1007 bytes --]
hi all,
rte_get_tsc_cycles function slowly increase, gettimeofday function is normally increase.
The following pseudo-code is my test example:
``` uint64_t hz = rte_get_tsc_hz(); uint64_t start_cycles = rte_get_tsc_cycles(); // running on first cpuuint64_t start_time = gettimeofday(); // running on first cpu <perform operations> // running for 1 day or 2 day uint64_t end_cycles = rte_get_tsc_cycles(); // running on second cpuuint64_t end_time = gettimeofday(); // running on first cpu
elapsed1_s = (end_cycles - start_cycles) / hz;elapsed2_s = (end_time - start_time);
```
I found a phenomenon that elapsed1_s is inaccurate, elapsed2_s is accurate. and the gap between elapsed2_s and elapsed1_s has increasingly widened.
why? how to explain the phenomenon or fix the problem.
Hope that helps! Regards,
wang.
^ permalink raw reply [flat|nested] only message in thread
only message in thread, other threads:[~2021-07-05 6:58 UTC | newest]
Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-07-02 9:36 [dpdk-dev] rte_get_tsc_cycles function slowly increase ? =?gb18030?B?0KHP6dfT?=
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).