From: Lingli Chen <linglix.chen@intel.com>
To: dts@dpdk.org
Cc: Lingli Chen <linglix.chen@intel.com>
Subject: [dts][PATCH V1] tests/distributor: modify script add novector path expect
Date: Tue, 5 Jul 2022 04:16:47 -0400 [thread overview]
Message-ID: <20220705081647.35533-1-linglix.chen@intel.com> (raw)
[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: text/plain; charset=y, Size: 1993 bytes --]
The “--force-max-simd-bitwidth=64” is relatively new to DPDK (novector code path) and severely limits the performance of the burst mode of the distributor.
So modify script to add novector path expect
Signed-off-by: Lingli Chen <linglix.chen@intel.com>
---
test_plans/distributor_test_plan.rst | 4 ++++
tests/TestSuite_distributor.py | 12 +++++++++---
2 files changed, 13 insertions(+), 3 deletions(-)
diff --git a/test_plans/distributor_test_plan.rst b/test_plans/distributor_test_plan.rst
index 7e61bef4..12fd9e9f 100644
--- a/test_plans/distributor_test_plan.rst
+++ b/test_plans/distributor_test_plan.rst
@@ -40,6 +40,10 @@ Compared CPU cycles for normal distributor and burst API
Verify burst distributor API cost much less cycles then normal version
+The novector path severely limits the performance of the burst mode of the distributor.
+so novector: expect single mode > burst mode
+vector: expect single mode > burst mode * 2
+
Test case: Distribute with maximum workers
==========================================
Start distributor sample with 63(0xeffffffffffffffff0) workers
diff --git a/tests/TestSuite_distributor.py b/tests/TestSuite_distributor.py
index 7b41eb01..6878edbd 100644
--- a/tests/TestSuite_distributor.py
+++ b/tests/TestSuite_distributor.py
@@ -65,9 +65,15 @@ class TestDistributor(TestCase):
)
self.dut.send_expect("quit", "# ")
self.verify("Test OK" in out, "Test failed")
- self.verify(
- cycles_single > cycles_burst * 2, "Burst performance should be much better"
- )
+ if "force-max-simd-bitwidth=64" in eal_para:
+ self.verify(
+ cycles_single > cycles_burst, "Burst performance should be much better"
+ )
+ else:
+ self.verify(
+ cycles_single > cycles_burst * 2,
+ "Burst performance should be much better",
+ )
def test_perf_distributor(self):
"""
--
2.17.1
next reply other threads:[~2022-07-05 8:54 UTC|newest]
Thread overview: 2+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-07-05 8:16 Lingli Chen [this message]
2022-07-06 6:11 ` lijuan.tu
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=20220705081647.35533-1-linglix.chen@intel.com \
--to=linglix.chen@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).