* Hardware Timestamping
@ 2023-09-11 7:09 Arne Vogel
0 siblings, 0 replies; only message in thread
From: Arne Vogel @ 2023-09-11 7:09 UTC (permalink / raw)
[-- Attachment #1.1: Type: text/plain, Size: 1309 bytes --]
I am looking into getting the hardware timestamp of when a packet is
received by the NIC inside of my DPDK program. In a regular program I
can get the timestamp the packet is received by the NIC and the kernel
using SO_TIMESTAMPING . Is there a way to get this timestamp in dpdk?
What I found when looking for this, it looks like `mbuf->timestamp` was
deprecated  and was replaced in this  example by
`rte_mbuf_timestamp_t` (in hwts_field). When looking at the values I get
from `hwts_field` it cycles around 50244192 though, going up _and_ down?
Looking at the output I am unsure what is actually being measured by the
rxtx_callback example. With a hardware freq. of 156024300 and about 40
cycles the latency is about 262ns. But what is this latency? From the
time the packet is received by the NIC to the first time dpdk interacts
with it? Or when I get the packet in my dpdk program?
[-- Attachment #1.2: Type: text/html, Size: 3899 bytes --]
[-- Attachment #2: S/MIME Cryptographic Signature --]
[-- Type: application/pkcs7-signature, Size: 5945 bytes --]
^ permalink raw reply [flat|nested] only message in thread
only message in thread, other threads:[~2023-09-11 7:09 UTC | newest]
Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2023-09-11 7:09 Hardware Timestamping Arne Vogel
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).