All of lore.kernel.org
 help / color / mirror / Atom feed
From: Richard Purdie <richard.purdie@linuxfoundation.org>
To: Peter Kjellerstedt <peter.kjellerstedt@axis.com>,
	"bitbake-devel@lists.openembedded.org"
	<bitbake-devel@lists.openembedded.org>
Subject: Re: [PATCH] cooker: Drop package-depends.dot and pn-depends.dot generation
Date: Wed, 25 Jan 2017 11:12:57 +0000	[thread overview]
Message-ID: <1485342777.30673.79.camel@linuxfoundation.org> (raw)
In-Reply-To: <16e4f38d781445449f3290f035a655ed@XBOX02.axis.com>

On Wed, 2017-01-25 at 11:05 +0000, Peter Kjellerstedt wrote:
> > 
> > -----Original Message-----
> > From: bitbake-devel-bounces@lists.openembedded.org [mailto:bitbake-
> > devel-bounces@lists.openembedded.org] On Behalf Of Richard Purdie
> > Sent: den 23 januari 2017 23:38
> > To: bitbake-devel@lists.openembedded.org
> > Subject: [bitbake-devel] [PATCH] cooker: Drop package-depends.dot
> > and
> > pn-depends.dot generation
> > 
> > A long time ago when we switched to task basked execution we added
> > task-depends.dot and generated package-depends.dot and pn-
> > depends.dot
> > for compatibility as best we could.	
> > 
> > The problem is they contain partial data about the taskgraph, its
> > incomplete and tends to confuse users.
> > 
> > I propose we remove the two compatibilty outputs and just generate
> > the one which contains definitive data.
> > 
> > Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
> Is there any alternative to removing package-depends.dot and 
> especially pn-depends.dot?

We don't have to remove them but what worries me is that the data in
them is incomplete. Its some subset of what the real task graph looks
like and at some point someone is going to complain the data was
inaccurate/misleading (which it is).

>  The reason I ask is because we use 
> pn-dependes.dot when visualizing the dependencies between packages. 
> Even though task-depends.dot is more complete, trying to render it 
> is near impossible given the sheer number of nodes and edges it 
> contains. Already visualizing pn-depends.dot is hard but possible 
> with some gvpr and tred filtering. Here are some statistics from 
> one of my typical builds:
> 
>                         Nodes  Edges
>                         -----  ------
> pn-depends.dot           1183   12087
> package-depends.dot      5939   55615
> task-depends.dot        13756  125870
> 
> As can be seen, the number of nodes and edges in task-depends.dot 
> is a magnitude greater than in pn-depends.dot.

Its certainly simpler, but the data is just plain buggy. I'm not sure
I'd trust anything those files told me, even if they are easier to
view.

I appreciate task-depends is hard to load graphically, it is helpful
even as a text file though, I do use it a lot since its a near direct
dump of bitbake's internal parsed task dependencies.

Cheers,

Richard



  reply	other threads:[~2017-01-25 11:13 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-01-25 11:05 [PATCH] cooker: Drop package-depends.dot and pn-depends.dot generation Peter Kjellerstedt
2017-01-25 11:12 ` Richard Purdie [this message]
2017-01-25 19:17   ` Paul Eggleton
2017-01-30 10:18     ` Peter Kjellerstedt
2017-02-02 15:13       ` Peter Kjellerstedt
2017-02-03 13:18     ` Tobias Hagelborn
  -- strict thread matches above, loose matches on Subject: below --
2017-01-23 22:38 Richard Purdie

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=1485342777.30673.79.camel@linuxfoundation.org \
    --to=richard.purdie@linuxfoundation.org \
    --cc=bitbake-devel@lists.openembedded.org \
    --cc=peter.kjellerstedt@axis.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.