public inbox for devel@edk2.groups.io
 help / color / mirror / Atom feed
From: "Gao, Liming" <liming.gao@intel.com>
To: Michael Zimmermann <sigmaepsilon92@gmail.com>,
	edk2-devel-01 <edk2-devel@lists.01.org>,
	"Zhu, Yonghong" <yonghong.zhu@intel.com>
Subject: Re: Questions about the build process
Date: Mon, 2 Jul 2018 09:11:49 +0000	[thread overview]
Message-ID: <4A89E2EF3DFEDB4C8BFDE51014F606A14E2A0D2C@SHSMSX104.ccr.corp.intel.com> (raw)
In-Reply-To: <CAN9vWDJ3tpTBcnjTdWA2Po+O+Up0x=rXK6EiVbtZCLfR8XX0kQ@mail.gmail.com>

Michael:
  There is one bugzilia https://bugzilla.tianocore.org/show_bug.cgi?id=972 on top level makefile generation. Its purpose is to let the make utility figure out how to multi-thread the build and see whether there is the build performance improvement.
  
  Edk2 BaseTools generates AutoGen and Makefile for every module, then build module one by one. Every module has its own compiler flag, its included path, its AutoGen.h and AutoGen.c. So, the module level makefile is generated. 

  I don't use ninja before. Edk2 build has supported multiple thread build to trig Make in parallel. Does ninja improve build performance? 

Thanks
Liming
>-----Original Message-----
>From: Michael Zimmermann [mailto:sigmaepsilon92@gmail.com]
>Sent: Saturday, June 30, 2018 2:47 PM
>To: edk2-devel-01 <edk2-devel@lists.01.org>; Zhu, Yonghong
><yonghong.zhu@intel.com>; Gao, Liming <liming.gao@intel.com>
>Subject: [edk2] Questions about the build process
>
>Out of curiosity I took a closer look at how the build system works
>internall(especially the Makefile generation and parallel builds) and
>I got some questions which I hope you can answer:
>
>Why are we using make at all?
>As far as I can see we're not using any special make functionality and
>most dependency calculation is done in the python build tool anyway.
>So why don't we just run the build commands ourselves from within the
>build tool?
>
>Why is every module being build separately / Why don't we generate one
>makefile which does everything ?
>To give an opposite view on the previous question, why don't we
>generate makefiles which contain the whole internal dependency graph
>so the build tool can then call a top level 'make' and not worry about
>anything. We wouldn't even have to implement any kind of parallel
>build support(which we currently have a lot of code for) because make
>would do that for us.
>
>If scrambled log output during parallel builds is a concern, then we
>could switch to the "ninja" build system which would probably be way
>more suited for edk2's usecase because it's a pretty fast low level
>language which also gives you proper logs when building in parallel.
>ninja is available for linux, mac and windows.
>
>Thanks
>Michael

  reply	other threads:[~2018-07-02  9:11 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-06-30  6:46 Questions about the build process Michael Zimmermann
2018-07-02  9:11 ` Gao, Liming [this message]
2018-07-02 11:15   ` Michael Zimmermann

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-list 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=4A89E2EF3DFEDB4C8BFDE51014F606A14E2A0D2C@SHSMSX104.ccr.corp.intel.com \
    --to=devel@edk2.groups.io \
    /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