DPDK patches and discussions
 help / color / mirror / Atom feed
From: Bruce Richardson <bruce.richardson@intel.com>
To: Ferruh Yigit <ferruh.yigit@amd.com>
Cc: <dev@dpdk.org>, <anatoly.burakov@intel.com>
Subject: Re: [PATCH] build: output a dependency log in build directory
Date: Wed, 31 Jul 2024 11:27:57 +0100	[thread overview]
Message-ID: <ZqoRrc57123L1THi@bricha3-mobl1.ger.corp.intel.com> (raw)
In-Reply-To: <f3deb013-bb35-495b-a3ac-ebc1fd7764b5@amd.com>

On Wed, Jul 31, 2024 at 11:17:54AM +0100, Ferruh Yigit wrote:
> On 7/30/2024 3:55 PM, Bruce Richardson wrote:
> > As meson processes our DPDK source tree it manages dependencies
> > specified by each individual driver. To enable future analysis of the
> > dependency links between components, log the dependencies of each DPDK
> > component as it gets processed. This could potentially allow other tools
> > to automatically enable or disable components based on the desired end
> > components to be built, e.g. if the user requests net/ice, ensure that
> > common/iavf is also enabled in the drivers.
> > 
> > The output file produced is in "dot" or "graphviz" format, which allows
> > producing a graphical representation of the DPDK dependency tree if so
> > desired. For example: "dot -Tpng -O build/deps.dot" to produce the
> > image file "build/deps.dot.png".
> > 
> > Signed-off-by: Bruce Richardson <bruce.richardson@intel.com>
> >
> 
> I tested it quickly, good to have this dependency list, at least it
> makes some duplicated dependencies obvious from .dot file.
> 
> But generated dependency graph is too large to be useful, does it make
> sense to have a new meson option that control this dependency generation:
> -Dgenerate_deps=apps
> -Dgenerate_deps=libs
> -Dgenerate_deps=drivers
> -Dgenerate_deps=all
> 

I had indeed noticed that. Right now I'm focused on cutting down our
dependency lists to make things more manageable - I have a script to
automate the identification of excess dependencies and I already have a
patch for cleaning up the libraries to a minimal dependency set.

I also think having separate drivers and libraries charts could be useful I
wouldn't want to add a build option for it though - it's not something that
should be necessary. Instead, I will look at other options, perhaps even
doing multiple dependency files - they don't seem to slow meson down, so
why not always do them all.

> Not sure about what should be the default option,
> as this is not always required/used, disabling this option by default
> saves unnecessary work,
> but disabling it by default may cause not noticing when it is broken,
> perhaps this can be addressed by adding this option to
> 'devtools/test-meson-builds.sh'.
> 
> btw, it generates following warning:

Oops - will look to fix in V2.

/Bruce

  reply	other threads:[~2024-07-31 10:30 UTC|newest]

Thread overview: 32+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-07-30 14:55 Bruce Richardson
2024-07-31  9:07 ` Konstantin Ananyev
2024-07-31 10:17 ` Ferruh Yigit
2024-07-31 10:27   ` Bruce Richardson [this message]
2024-08-02 12:44 ` [PATCH v2 0/7] record and rework component dependencies Bruce Richardson
2024-08-02 12:44   ` [PATCH v2 1/7] build: output a dependency log in build directory Bruce Richardson
2024-09-02 14:34     ` Burakov, Anatoly
2024-09-03  8:31       ` Bruce Richardson
2024-08-02 12:44   ` [PATCH v2 2/7] devtools: add script to flag unneeded dependencies Bruce Richardson
2024-08-02 12:44   ` [PATCH v2 3/7] build: remove kvargs from driver class dependencies Bruce Richardson
2024-08-02 12:44   ` [PATCH v2 4/7] build: reduce library dependencies Bruce Richardson
2024-08-02 12:44   ` [PATCH v2 5/7] build: reduce driver dependencies Bruce Richardson
2024-08-02 12:44   ` [PATCH v2 6/7] build: reduce app dependencies Bruce Richardson
2024-08-02 12:44   ` [PATCH v2 7/7] devtools: add script to generate DPDK dependency graphs Bruce Richardson
2024-08-02 13:29   ` [PATCH v2 0/7] record and rework component dependencies Morten Brørup
2024-08-02 15:05   ` Patrick Robb
2024-08-02 15:11     ` Bruce Richardson
2024-08-02 17:18   ` Ferruh Yigit
2024-08-06  8:35     ` Bruce Richardson
2024-09-04 15:08 ` [PATCH v3 0/8] " Anatoly Burakov
2024-09-04 15:08   ` [PATCH v3 1/8] build: split dependencies into mandatory and optional Anatoly Burakov
2024-09-06 14:51     ` Bruce Richardson
2024-09-09  8:41       ` Burakov, Anatoly
2024-09-09  9:01         ` Bruce Richardson
2024-09-04 15:08   ` [PATCH v3 2/8] build: output a dependency log in build directory Anatoly Burakov
2024-09-04 15:08   ` [PATCH v3 3/8] devtools: add script to flag unneeded dependencies Anatoly Burakov
2024-09-04 15:08   ` [PATCH v3 4/8] build: remove kvargs from driver class dependencies Anatoly Burakov
2024-09-04 15:08   ` [PATCH v3 5/8] build: reduce library dependencies Anatoly Burakov
2024-09-04 15:08   ` [PATCH v3 6/8] build: reduce driver dependencies Anatoly Burakov
2024-09-04 15:08   ` [PATCH v3 7/8] build: reduce app dependencies Anatoly Burakov
2024-09-04 15:08   ` [PATCH v3 8/8] devtools: add script to generate DPDK dependency graphs Anatoly Burakov
2024-09-05  6:05   ` [PATCH v3 0/8] record and rework component dependencies Morten Brørup

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=ZqoRrc57123L1THi@bricha3-mobl1.ger.corp.intel.com \
    --to=bruce.richardson@intel.com \
    --cc=anatoly.burakov@intel.com \
    --cc=dev@dpdk.org \
    --cc=ferruh.yigit@amd.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).