From: Peng Yuan <yuan.peng@intel.com>
To: dts@dpdk.org
Cc: pengyuan <yuan.peng@intel.com>
Subject: [dts] [PATCH v1]test_plans: remove per queue stats check
Date: Thu, 26 Nov 2020 14:46:38 +0000 [thread overview]
Message-ID: <1606401999-50212-1-git-send-email-yuan.peng@intel.com> (raw)
From: pengyuan <yuan.peng@intel.com>
FVL does not support hardware per queue stats,
so remove the per queue stats check from test plan.
Signed-off-by: pengyuan <yuan.peng@intel.com>
diff --git a/test_plans/stats_checks_test_plan.rst b/test_plans/stats_checks_test_plan.rst
index e571cb34..f314ba88 100644
--- a/test_plans/stats_checks_test_plan.rst
+++ b/test_plans/stats_checks_test_plan.rst
@@ -128,7 +128,8 @@ The fields checked are RX-packets and TX-packets of each queue stats,
RX-packets, RX-bytes, TX-packets and TX-bytes of each port stats,
rx_good_packets and rx_good_bytes of each port xstats,
tx_good_packets and tx_good_bytes of each port xstats,
-rx_qx_packets, rx_qx_bytes, tx_qx_packets and tx_qx_bytes of each port xstats.
+FVL does not support hardware per queue stats,
+so we won't check rx and tx per queue stats.
Test Case: PF xstatus Checks
============================
@@ -210,17 +211,17 @@ Test Case: PF xstatus Checks
rx_good_bytes: 6000
tx_good_bytes: 0
......
- rx_q0_packets: 29
- rx_q0_bytes: 1740
+ rx_q0_packets: 0
+ rx_q0_bytes: 0
rx_q0_errors: 0
- rx_q1_packets: 21
- rx_q1_bytes: 1260
+ rx_q1_packets: 0
+ rx_q1_bytes: 0
rx_q1_errors: 0
- rx_q2_packets: 24
- rx_q2_bytes: 1440
+ rx_q2_packets: 0
+ rx_q2_bytes: 0
rx_q2_errors: 0
- rx_q3_packets: 26
- rx_q3_bytes: 1560
+ rx_q3_packets: 0
+ rx_q3_bytes: 0
rx_q3_errors: 0
tx_q0_packets: 0
tx_q0_bytes: 0
@@ -248,21 +249,20 @@ Test Case: PF xstatus Checks
rx_q3_packets: 0
rx_q3_bytes: 0
rx_q3_errors: 0
- tx_q0_packets: 29
- tx_q0_bytes: 1740
- tx_q1_packets: 21
- tx_q1_bytes: 1260
- tx_q2_packets: 24
- tx_q2_bytes: 1440
- tx_q3_packets: 26
- tx_q3_bytes: 1560
+ tx_q0_packets: 0
+ tx_q0_bytes: 0
+ tx_q1_packets: 0
+ tx_q1_bytes: 0
+ tx_q2_packets: 0
+ tx_q2_bytes: 0
+ tx_q3_packets: 0
+ tx_q3_bytes: 0
verify rx_good_packets, RX-packets of port 0 and tx_good_packets, TX-packets of port 1 are both 100.
rx_good_bytes, RX-bytes of port 0 and tx_good_bytes, TX-bytes of port 1 are the same.
-RX-packets of each queue of port 0 are equal to rx_qx_packets of each queue of port 0.
-TX-packets of each queue of port 1 are equal to tx_qx_packets of each queue of port 1.
-The sum of rx_qx_bytes of each queue of port 0 is equal to the rx_good_packets of port 0.
-The sum of tx_qx_bytes of each queue of port 1 is equal to the tx_good_packets of port 1.
+FVL does not support hardware per queue stats,
+so rx_qx_packets and rx_qx_bytes are both 0.
+tx_qx_packets and tx_qx_bytes are both 0 too.
Test Case: VF xstats Checks
============================
--
2.17.1
next reply other threads:[~2020-11-26 7:43 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2020-11-26 14:46 Peng Yuan [this message]
2020-11-26 8:33 ` Tu, Lijuan
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1606401999-50212-1-git-send-email-yuan.peng@intel.com \
--to=yuan.peng@intel.com \
--cc=dts@dpdk.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
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).