Buildroot Archive on lore.kernel.org
 help / color / mirror / Atom feed
From: Arnout Vandecappelle via buildroot <buildroot@buildroot.org>
To: "Yann E. MORIN" <yann.morin.1998@free.fr>, buildroot@buildroot.org
Cc: Julien Olivain <ju.o@free.fr>,
	Fabrice Fontaine <fontaine.fabrice@gmail.com>,
	Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Subject: Re: [Buildroot] [PATCH] package/gdb: gdbserver does not need zlib
Date: Sun, 17 Sep 2023 21:32:41 +0200	[thread overview]
Message-ID: <23f8c528-9158-07ae-615e-e2ca44c842e8@mind.be> (raw)
In-Reply-To: <20230917152303.1300577-1-yann.morin.1998@free.fr>



On 17/09/2023 17:23, Yann E. MORIN wrote:
> Since 3341ceb1e585 (package/gdb: zlib is mandatory, not optional), zlib
> has become a mandatory dependencies of the gdb package.
> 
> However, zlib is only needed for the debugger, gdb itself, while the
> server, gdbserver, does not use it.
> 
> This means that, when building an SDK to be later reused as an external
> toolchain, the zlib headers and libraries are present in the sysroot of
> the toolchain, tainting the toolchain and making it unsuitable to be
> reused.
> 
> As Julien noticed, for example, tcl will try and link with zlib if
> available, and at build time it is. But at runtime, it is not, and thus
> tclsh fails to run; see 7af8dee3a8a0 (package/tcl: add mandatory
> dependency to zlib)

  So this commit should be reverted now?


> When we only need to build gdbserver, we still need to configure and
> build the whole gdb distribution, which means we call the top-level
> configure script; that script has no option to disable the detection
> of zlib: it wants to either use a system one, or it will build the
> bundled one.
> 
> So, when we onlyt build gdbserver, we tell configure to not use a system
> zlib, which triggers a build of the bundled one, which we do not use...

  I didn't think this paragraph was very clear, so I reformulated it a little, 
and applied to master, thanks.

  Regards,
  Arnout

> 
> Reported-by: Julien Olivain <ju.o@free.fr>
> Reported-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
> Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
> Cc: Fabrice Fontaine <fontaine.fabrice@gmail.com>
> ---
>   package/gdb/Config.in |  2 +-
>   package/gdb/gdb.mk    | 14 ++++++++++----
>   2 files changed, 11 insertions(+), 5 deletions(-)
> 
> diff --git a/package/gdb/Config.in b/package/gdb/Config.in
> index 20fd262c93..e4ab39a2bd 100644
> --- a/package/gdb/Config.in
> +++ b/package/gdb/Config.in
> @@ -24,7 +24,6 @@ config BR2_PACKAGE_GDB
>   	# The or1k musl port is incomplete, elf_gregset_t definition is missing:
>   	# https://git.musl-libc.org/cgit/musl/tree/arch/or1k/bits/user.h?h=v1.2.3
>   	depends on !BR2_or1k || !BR2_TOOLCHAIN_USES_MUSL
> -	select BR2_PACKAGE_ZLIB
>   	# When the external toolchain gdbserver is copied to the
>   	# target, we don't allow building a separate gdbserver. The
>   	# one from the external toolchain should be used.
> @@ -61,6 +60,7 @@ config BR2_PACKAGE_GDB_DEBUGGER
>   	depends on !BR2_sh
>   	select BR2_PACKAGE_GMP if !BR2_arc
>   	select BR2_PACKAGE_NCURSES
> +	select BR2_PACKAGE_ZLIB
>   
>   comment "full gdb on target needs a toolchain w/ wchar"
>   	depends on !BR2_sh
> diff --git a/package/gdb/gdb.mk b/package/gdb/gdb.mk
> index 0d41548ed0..070598b385 100644
> --- a/package/gdb/gdb.mk
> +++ b/package/gdb/gdb.mk
> @@ -32,7 +32,6 @@ GDB_PRE_CONFIGURE_HOOKS += GDB_CONFIGURE_SYMLINK
>   # also need ncurses.
>   # As for libiberty, gdb may use a system-installed one if present, so
>   # we must ensure ours is installed first.
> -GDB_DEPENDENCIES = zlib
>   HOST_GDB_DEPENDENCIES = host-expat host-libiberty host-ncurses host-zlib
>   
>   # Disable building documentation
> @@ -131,22 +130,29 @@ GDB_CONF_OPTS = \
>   	--disable-sim \
>   	$(GDB_DISABLE_BINUTILS_CONF_OPTS) \
>   	--without-included-gettext \
> -	--with-system-zlib \
>   	--disable-werror \
>   	--enable-static \
>   	--without-mpfr \
>   	--disable-source-highlight
>   
>   ifeq ($(BR2_PACKAGE_GDB_DEBUGGER),y)
> +GDB_DEPENDENCIES += zlib
>   GDB_CONF_OPTS += \
>   	--enable-gdb \
> -	--with-curses
> +	--with-curses \
> +	--with-system-zlib
>   GDB_DEPENDENCIES += ncurses \
>   	$(if $(BR2_PACKAGE_LIBICONV),libiconv)
>   else
> +# When only building gdbserver, we don't need zlib. But we have no way to
> +# tell the top-level configure that we don't need zlib: it either wants to
> +# build the bundled one, or use the system one.
> +# Since we're going to only install the gdbserver to the target, we don't
> +# care that the bundled zlib is built, as it is not used.
>   GDB_CONF_OPTS += \
>   	--disable-gdb \
> -	--without-curses
> +	--without-curses \
> +	--without-system-zlib
>   endif
>   
>   # Starting from GDB 11.x, gmp is needed as a dependency to build full
_______________________________________________
buildroot mailing list
buildroot@buildroot.org
https://lists.buildroot.org/mailman/listinfo/buildroot

  reply	other threads:[~2023-09-17 19:32 UTC|newest]

Thread overview: 4+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-09-17 15:23 [Buildroot] [PATCH] package/gdb: gdbserver does not need zlib Yann E. MORIN
2023-09-17 19:32 ` Arnout Vandecappelle via buildroot [this message]
2023-09-17 19:41   ` Yann E. MORIN
2023-09-25  5:33 ` Peter Korsgaard

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=23f8c528-9158-07ae-615e-e2ca44c842e8@mind.be \
    --to=buildroot@buildroot.org \
    --cc=arnout@mind.be \
    --cc=fontaine.fabrice@gmail.com \
    --cc=ju.o@free.fr \
    --cc=thomas.petazzoni@bootlin.com \
    --cc=yann.morin.1998@free.fr \
    /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