test suite reviews and discussions
 help / color / mirror / Atom feed
From: Angela Czubak <aczubak@caviumnetworks.com>
To: dts@dpdk.org, "Liu, Yong" <yong.liu@intel.com>
Subject: [dts] Fwd: Re:  [PATCH v2 00/28] DTS support for Cavium NICs
Date: Tue, 18 Apr 2017 16:02:12 +0200	[thread overview]
Message-ID: <99d84330-e5ec-a924-84dd-d9a9fba7f972@caviumnetworks.com> (raw)
In-Reply-To: <88720c3f-fbde-f9b6-d97b-72a7607085df@caviumnetworks.com>

Hi Marvin,

please wait for another version, I would like to include another small
fix in the shutdown api test suite, but I need some more time to
investigate how should MTU/max rx pkt settings work in both DPDK and DTS
(it seems we have found a bug in thunder driver for DPDK and I want to
make sure the test expects the correct behaviour). I will merge the two
commits mentioned into one, so please wait for V3.

Regards,

Angela


On 17.04.2017 10:02, Liu, Yong wrote:
> Angela,
> I would like to merge your patches. There're two patches look like the latter one will replace the previous one.
> I can merge others first or waiting for next version with these two patches combined to one. Which choice is your preferred?
>
> [dts] [PATCH v2 02/28] framework dut and project_dpdk: using correct available binding scripts
> [dts] [PATCH v2 08/28] framework: getting DPDK binding script moved to a method in dut
>
>
> Thanks,
> Marvin
>
>> -----Original Message-----
>> From: dts [mailto:dts-bounces@dpdk.org] On Behalf Of Angela Czubak
>> Sent: Tuesday, April 11, 2017 1:09 AM
>> To:dts@dpdk.org
>> Cc: Angela Czubak<aczubak@caviumnetworks.com>
>> Subject: [dts] [PATCH v2 00/28] DTS support for Cavium NICs
>>
>> Changes done to run DTS for Cavium NICs
>>
>> 1) Cavium devices with 177d:a034 and 177d:0011 PCI IDs are added to NICs
>> list, both of which are supported by thunder-nicvf driver
>> 2) When appending pci devices info, the only Cavium NICs appended are the
>> ones with 10Gb/s linkspeed
>> 3) For arm64 architecture the hugepage size is detected and if it is
>> 524288, then lesser amount of hugepages is acquired
>> 4) Created a method in Dut class for getting the right binding script (if
>> someone chooses to use older DPDK version)
>> 5) Checking link is done via IPv4 ping alternately, so that DUT can be a
>> system without IPv6 (previously link was detected solely by checking if an
>> interface obtained an IPv6 address and then by using ping6). This also
>> makes force binding necessary, as those interfaces are detected as active.
>> 6) Checking whether vfio-pci was loaded using lsmod can be deceptive. If
>> someone decides to use DPDK on a system with modules built into the kernel,
>> then this approach will fail. Instead, DPDK binding script can be used and
>> if it shows that vfio-pci can be used, DTS proceeds, otherwise it send
>> with an error due to a failed assertion.
>> 7) Modifying tests and config for Cavium NICs
>> 8) Added support for Cavium NICs to choose only cards with specified
>> linkspeed
>> 9) Skipping recompling DPDK in TestSuite_pmd.py if the driver is not meant
>> for FVL
>>
>> Changes since initial version:
>> - Added support for Cavium NICs to choose only cards with specified
>> linkspeed
>> - Skipping recompling DPDK in TestSuite_pmd.py if the driver is not meant
>> for FVL
>> - Setting max-pkt-len before setting vlan stripping off in
>> TestSuite_shutdown_api.py when testing jumbo frames
>>
>>
>> Angela Czubak (28):
>>    framework settings: Adding cavium PCI IDs and drivers
>>    framework dut and project_dpdk: using correct available binding
>>      scripts
>>    framework dut and project_dpdk: binding smaller number of hugepages
>>      for     arm64
>>    framework crb: Appending only 10G devices for cavium
>>    framework project_dpdk: asserting loading vfio-pci by using DPDK
>>      binding script
>>    framework: checking link with IPv4 ping
>>    tests: ipv4_reassembly - added splitting result in tcpdump_command
>>    framework: getting DPDK binding script moved to a method in dut
>>    tests: Modified tests to take cavium nics into account
>>    tests: checksum_offload - fixes
>>    tests: dynamic_config - taking into account cavium NICs
>>    tests: fdir - skipping fdir tests for Cavium NICs
>>    tests: ieee1588 - skipping ieee1588_enable test for Cavium NICs
>>    tests: jumboframes - skipping jumboframes tests testing bigger than
>>       allowed frames for Cavium NICs
>>    tests: l2fwd - fixes to run on two cards (same NIC) and expecting
>>      "L2FWD: entering main loop" instead of "memory mapped"
>>    tests: link_status_interrupt - skipping
>>      test_link_status_interrupt_port_available for Cavium NICs
>>    tests: link_status_interrupt - skipping tests verifying adding MAC
>>      addresses for Cavium NICs
>>    tests: pmdpcap - modifying the right config file + importing utils
>>      after scapy
>>    tests: pmdrssreta - skipping tests for Cavium NICs
>>    tests: queue_start_stop - taking into account cavium NICs
>>    tests: quota_watermark - skipping tests for Cavium NICs
>>    tests: short_live - taking into account cavium NICs + fixes
>>    tests: tso - taking into account cavium NICs
>>    tests: uni_pkt - taking into account cavium NICs
>>    tests: unit_tests_dump - taking into account cavium NICs
>>    tests: unit_tests_lpm - asking for lesser amount of hugepages if they
>>         are 524288kB
>>    framework: Adding linkspeed parameter for Cavium NICs
>>    tests: pmd - Recompiling DPDK only if that is relevant
>>
>>   conf/dpdk_support_test_case.xls     | Bin 25600 -> 8704 bytes
>>   conf/dpdk_test_case_checklist.xls   | Bin 31232 -> 21504 bytes
>>   doc/dts_gsg/config.rst              |   4 +
>>   framework/crb.py                    |  12 +-
>>   framework/dts.py                    |   5 +
>>   framework/dut.py                    |  52 ++++++---
>>   framework/pmd_output.py             |   4 +
>>   framework/project_dpdk.py           |  18 ++-
>>   framework/settings.py               |   5 +
>>   framework/test_capabilities.py      |   9 ++
>>   framework/test_case.py              |   7 +-
>>   framework/tester.py                 |  11 ++
>>   nics/net_device.py                  |   8 +-
>>   tests/TestSuite_checksum_offload.py |  38 +++----
>>   tests/TestSuite_dynamic_config.py   |  10 +-
>>   tests/TestSuite_ipv4_reassembly.py  |   2 +-
>>   tests/TestSuite_l2fwd.py            |   4 +-
>>   tests/TestSuite_pmd.py              |   5 +-
>>   tests/TestSuite_pmdpcap.py          |  13 ++-
>>   tests/TestSuite_queue_start_stop.py |   5 +-
>>   tests/TestSuite_short_live.py       |  44 ++++++--
>>   tests/TestSuite_shutdown_api.py     |  31 ++---
>>   tests/TestSuite_tso.py              |  10 +-
>>   tests/TestSuite_uni_pkt.py          | 219 ++++++++++++++++++++++++-------
>> -----
>>   tests/TestSuite_unit_tests_dump.py  |  17 ++-
>>   tests/TestSuite_unit_tests_lpm.py   |   9 +-
>>   26 files changed, 383 insertions(+), 159 deletions(-)
>>   create mode 100644 framework/test_capabilities.py
>>
>> --
>> 2.7.4

           reply	other threads:[~2017-04-18 14:02 UTC|newest]

Thread overview: expand[flat|nested]  mbox.gz  Atom feed
 [parent not found: <88720c3f-fbde-f9b6-d97b-72a7607085df@caviumnetworks.com>]

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=99d84330-e5ec-a924-84dd-d9a9fba7f972@caviumnetworks.com \
    --to=aczubak@caviumnetworks.com \
    --cc=dts@dpdk.org \
    --cc=yong.liu@intel.com \
    /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).