From: Aaron Conole <aconole@redhat.com>
To: "Dumitrescu\, Cristian" <cristian.dumitrescu@intel.com>
Cc: Thomas Monjalon <thomas@monjalon.net>, "dev\@dpdk.org" <dev@dpdk.org>
Subject: Re: [dpdk-dev] [PATCH 2/2] pipeline: autodetect endianness of action args
Date: Tue, 20 Apr 2021 10:48:05 -0400 [thread overview]
Message-ID: <f7ttuo1ui56.fsf@dhcp-25.97.bos.redhat.com> (raw)
In-Reply-To: <DM6PR11MB2796FAC61F5A4EFD9A569F92EB489@DM6PR11MB2796.namprd11.prod.outlook.com> (Cristian Dumitrescu's message of "Tue, 20 Apr 2021 10:20:28 +0000")
"Dumitrescu, Cristian" <cristian.dumitrescu@intel.com> writes:
>> -----Original Message-----
>> From: Thomas Monjalon <thomas@monjalon.net>
>> Sent: Tuesday, April 20, 2021 11:07 AM
>> To: Dumitrescu, Cristian <cristian.dumitrescu@intel.com>
>> Cc: dev@dpdk.org
>> Subject: Re: [dpdk-dev] [PATCH 2/2] pipeline: autodetect endianness of
>> action args
>>
>> 20/04/2021 12:04, Dumitrescu, Cristian:
>> > From: Thomas Monjalon <thomas@monjalon.net>
>> > > 12/04/2021 01:23, Cristian Dumitrescu:
>> > > > Each table entry is made up of match fields and action data, with the
>> > > > latter made up of the action ID and the action arguments. The approach
>> > > > of having the user specify explicitly the endianness of the action
>> > > > arguments is difficult to be picked up by P4 compilers, as the P4
>> > > > compiler is generally unaware about this aspect.
>> > > >
>> > > > This commit introduces the auto-detection of the endianness of the
>> > > > action arguments by examining the endianness of the their destination:
>> > > > network byte order (NBO) when they get copied to headers and host
>> byte
>> > > > order (HBO) when they get copied to packet meta-data or mailboxes.
>> > > >
>> > > > The endianness specification of each action argument as part of the
>> > > > rule specification, e.g. H(...) and N(...) is removed from the rule
>> > > > file and auto-detected based on their destination. The DMA instruction
>> > > > scope is made internal, so mov instructions need to be used. The
>> > > > pattern of transferring complete headers from table entry action args
>> > > > to headers is detected, and the associated set of mov instructions
>> > > > plus header validate is internally detected and replaced with the
>> > > > internal-only DMA instruction to preserve performance.
>> > > >
>> > > > Signed-off-by: Cristian Dumitrescu <cristian.dumitrescu@intel.com>
>> > > > ---
>> > > > examples/pipeline/examples/vxlan.spec | 41 ++-
>> > > > examples/pipeline/examples/vxlan_table.py | 44 ++--
>> > > > examples/pipeline/examples/vxlan_table.txt | 32 +--
>> > > > lib/librte_pipeline/rte_swx_ctl.c | 15 +-
>> > > > lib/librte_pipeline/rte_swx_ctl.h | 6 +
>> > > > lib/librte_pipeline/rte_swx_pipeline.c | 282 ++++++++++++++++++-
>> --
>> > > > lib/librte_pipeline/rte_swx_pipeline.h | 4 -
>> > >
>> > > There are compilation issues in the CI:
>> > > https://patches.dpdk.org/project/dpdk/patch/20210411232338.4005-2-
>> > > cristian.dumitrescu@intel.com/
>> > >
>> >
>> > Hi Thomas,
>> >
>> > The reason for the CI issues is because this patch set has dependencies on
>> the previous patches that were pending, but now already applied by you
>> (thank you!), so there should not be any real issues?
>>
>> Yes I didn't see any issue locally, just wanted to confirm.
>>
>
> Thanks, Thomas. I just double checked by applying this set on the main
> latest, and I could not see any apply or build issues.
I restarted the builds for this.
Currently, travis robot (but not github actions) reports an error even
after the restart of this build - this is in error. The correct build
should be:
https://travis-ci.com/github/ovsrobot/dpdk/builds/223614512
And that shows the build completed successfully. Looks like we pull the
build list incorrectly.
next prev parent reply other threads:[~2021-04-20 14:48 UTC|newest]
Thread overview: 15+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-04-11 23:23 [dpdk-dev] [PATCH 1/2] pipeline: modularize the instruction optimizer Cristian Dumitrescu
2021-04-11 23:23 ` [dpdk-dev] [PATCH 2/2] pipeline: autodetect endianness of action args Cristian Dumitrescu
2021-04-20 0:48 ` Thomas Monjalon
2021-04-20 10:04 ` Dumitrescu, Cristian
2021-04-20 10:06 ` Thomas Monjalon
2021-04-20 10:20 ` Dumitrescu, Cristian
2021-04-20 14:48 ` Aaron Conole [this message]
2021-04-20 19:57 ` Thomas Monjalon
2021-04-21 7:49 ` Ali Alnubani
2021-04-21 12:57 ` Aaron Conole
2021-04-21 13:03 ` Ali Alnubani
2021-04-21 13:21 ` Aaron Conole
2021-04-21 13:58 ` Dumitrescu, Cristian
2021-04-21 14:24 ` Ali Alnubani
2021-04-21 15:10 ` Dumitrescu, Cristian
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=f7ttuo1ui56.fsf@dhcp-25.97.bos.redhat.com \
--to=aconole@redhat.com \
--cc=cristian.dumitrescu@intel.com \
--cc=dev@dpdk.org \
--cc=thomas@monjalon.net \
/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).