Buildroot Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: "Yann E. MORIN" <yann.morin.1998@free.fr>
To: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Cc: Jonathan Borne <jborne@kalray.eu>,
	jmaselbas@kalray.eu, ysionneaux@kalray.eu,
	Baruch Siach via buildroot <buildroot@buildroot.org>
Subject: Re: [Buildroot] [RFC PATCH 0/1] Fix binutils/libiberty target install directory
Date: Mon, 24 Jan 2022 22:21:49 +0100	[thread overview]
Message-ID: <20220124212149.GJ2502@scaer> (raw)
In-Reply-To: <20220124220238.1fa0568b@windsurf>

Thomas, All,

On 2022-01-24 22:02 +0100, Thomas Petazzoni spake thusly:
> On Mon, 24 Jan 2022 19:44:29 +0200
> Baruch Siach via buildroot <buildroot@buildroot.org> wrote:
> 
> > > Is there a reason behind this?  
> > 
> > As I understand that's because libiberty is static only. Buildroot does
> > not copy static libraries to $(TARGET_DIR) because they are only used at
> > build time, but not at run-time.
> 
> True, but it doesn't make sense to do that in
> BINUTILS_INSTALL_TARGET_CMDS. It should go in
> BINUTILS_INSTALL_STAGING_CMDS, where it is already present, as Romain
> pointed out.
> 
> So basically:
> 
>         $(TARGET_MAKE_ENV) $(MAKE) -C $(@D)/libiberty DESTDIR=$(STAGING_DIR) install
> 
> needs to be dropped from BINUTILS_INSTALL_TARGET_CMDS, if indeed
> libiberty is only installed as a static library.
> 
> However, taking a step back:
> 
> # We just want libbfd, libiberty and libopcodes,
> # not the full-blown binutils in staging
> define BINUTILS_INSTALL_STAGING_CMDS
>         $(TARGET_MAKE_ENV) $(MAKE) -C $(@D)/bfd DESTDIR=$(STAGING_DIR) install
>         $(TARGET_MAKE_ENV) $(MAKE) -C $(@D)/opcodes DESTDIR=$(STAGING_DIR) install
>         $(TARGET_MAKE_ENV) $(MAKE) -C $(@D)/libiberty DESTDIR=$(STAGING_DIR) install
> endef
> 
> # If we don't want full binutils on target
> ifneq ($(BR2_PACKAGE_BINUTILS_TARGET),y)
> define BINUTILS_INSTALL_TARGET_CMDS
>         $(TARGET_MAKE_ENV) $(MAKE) -C $(@D)/bfd DESTDIR=$(TARGET_DIR) install
>         $(TARGET_MAKE_ENV) $(MAKE) -C $(@D)/opcodes DESTDIR=$(TARGET_DIR) install
>         $(TARGET_MAKE_ENV) $(MAKE) -C $(@D)/libiberty DESTDIR=$(STAGING_DIR) install
> endef
> endif
> 
> This is *extremely* suspicious. Indeed, we get the libraries
> unconditionally installed in staging, so any other library or program
> that we build after binutils can potentially link against thosze
> libraries.
> 
> But we only install them conditionally in $(TARGET_DIR) if
> BR2_PACKAGE_BINUTILS_TARGET=y.
> 
> This seems very bad to me. Both target and staging installations should
> be done under the same condition.

Not really. You missed the fact that binutils is an autotools package.

As such, when BR2_PACKAGE_BINUTILS_TARGET=n, then the full binutils are
instaleld in target, so the libs will be installed in target.

So whatever gets in staging is present in target, but target can be a
superset of what is in staging, so whatever links to a binutils lib from
staging will find it in target.

Regards,
Yann E. MORIN.

-- 
.-----------------.--------------------.------------------.--------------------.
|  Yann E. MORIN  | Real-Time Embedded | /"\ ASCII RIBBON | Erics' conspiracy: |
| +33 662 376 056 | Software  Designer | \ / CAMPAIGN     |  ___               |
| +33 561 099 427 `------------.-------:  X  AGAINST      |  \e/  There is no  |
| http://ymorin.is-a-geek.org/ | _/*\_ | / \ HTML MAIL    |   v   conspiracy.  |
'------------------------------^-------^------------------^--------------------'
_______________________________________________
buildroot mailing list
buildroot@buildroot.org
https://lists.buildroot.org/mailman/listinfo/buildroot

  reply	other threads:[~2022-01-24 21:22 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-01-24 17:35 [Buildroot] [RFC PATCH 0/1] Fix binutils/libiberty target install directory Jonathan Borne via buildroot
2022-01-24 17:35 ` [Buildroot] [RFC PATCH 1/1] " Jonathan Borne via buildroot
2022-01-24 17:44 ` [Buildroot] [RFC PATCH 0/1] " Baruch Siach via buildroot
2022-01-24 18:12   ` Romain Naour
2022-01-24 21:02   ` Thomas Petazzoni
2022-01-24 21:21     ` Yann E. MORIN [this message]
2022-01-24 21:53       ` Thomas Petazzoni
2022-01-26 15:30         ` Jonathan Borne via buildroot
2022-01-26 15:52           ` Thomas Petazzoni
2022-01-26 17:14           ` Yann E. MORIN

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=20220124212149.GJ2502@scaer \
    --to=yann.morin.1998@free.fr \
    --cc=buildroot@buildroot.org \
    --cc=jborne@kalray.eu \
    --cc=jmaselbas@kalray.eu \
    --cc=thomas.petazzoni@bootlin.com \
    --cc=ysionneaux@kalray.eu \
    /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