From: Richard Purdie <richard.purdie@linuxfoundation.org>
To: Dongxiao Xu <dongxiao.xu@intel.com>
Cc: poky@yoctoproject.org
Subject: Re: [PATCH 1/5] staging: Use relative path in sysroot-destdir for target recipes
Date: Mon, 17 Jan 2011 22:42:13 +0000 [thread overview]
Message-ID: <1295304133.14388.15857.camel@rex> (raw)
In-Reply-To: <a84ff52c94c4299e9c6093539ba25bc5df7769e6.1295042140.git.dongxiao.xu@intel.com>
Hi Dongxaio,
I agree with this patch as you know and I want to merge it but I'm not
entirely sure its 100% right yet.
On Sat, 2011-01-15 at 06:14 +0800, Dongxiao Xu wrote:
> diff --git a/meta/classes/staging.bbclass b/meta/classes/staging.bbclass
> index 8432565..4d2991b 100644
> --- a/meta/classes/staging.bbclass
> +++ b/meta/classes/staging.bbclass
> @@ -45,25 +45,25 @@ sysroot_stage_dirs() {
> from="$1"
> to="$2"
>
> - sysroot_stage_dir $from${includedir} $to${STAGING_INCDIR}
> + sysroot_stage_dir $from${includedir} $to${includedir}
> if [ "${BUILD_SYS}" = "${HOST_SYS}" ]; then
> - sysroot_stage_dir $from${bindir} $to${STAGING_DIR_HOST}${bindir}
> - sysroot_stage_dir $from${sbindir} $to${STAGING_DIR_HOST}${sbindir}
> - sysroot_stage_dir $from${base_bindir} $to${STAGING_DIR_HOST}${base_bindir}
> - sysroot_stage_dir $from${base_sbindir} $to${STAGING_DIR_HOST}${base_sbindir}
> - sysroot_stage_dir $from${libexecdir} $to${STAGING_DIR_HOST}${libexecdir}
> - sysroot_stage_dir $from${sysconfdir} $to${STAGING_DIR_HOST}${sysconfdir}
> - sysroot_stage_dir $from${localstatedir} $to${STAGING_DIR_HOST}${localstatedir}
> + sysroot_stage_dir $from${bindir} $to${bindir}
> + sysroot_stage_dir $from${sbindir} $to${sbindir}
> + sysroot_stage_dir $from${base_bindir} $to${base_bindir}
> + sysroot_stage_dir $from${base_sbindir} $to${base_sbindir}
> + sysroot_stage_dir $from${libexecdir} $to${libexecdir}
> + sysroot_stage_dir $from${sysconfdir} $to${sysconfdir}
> + sysroot_stage_dir $from${localstatedir} $to${localstatedir}
> fi
> if [ -d $from${libdir} ]
> then
> - sysroot_stage_libdir $from/${libdir} $to${STAGING_LIBDIR}
> + sysroot_stage_libdir $from/${libdir} $to${libdir}
> fi
> if [ -d $from${base_libdir} ]
> then
> - sysroot_stage_libdir $from${base_libdir} $to${STAGING_DIR_HOST}${base_libdir}
> + sysroot_stage_libdir $from${base_libdir} $to${base_libdir}
> fi
> - sysroot_stage_dir $from${datadir} $to${STAGING_DATADIR}
> + sysroot_stage_dir $from${datadir} $to${datadir}
> }
>
> sysroot_stage_all() {
> @@ -76,6 +76,7 @@ do_populate_sysroot[dirs] = "${STAGING_DIR_TARGET}/${bindir} ${STAGING_DIR_TARGE
> ${STAGING_INCDIR_NATIVE} \
> ${STAGING_DATADIR} \
> ${SYSROOT_DESTDIR}${STAGING_DIR_TARGET} \
> + ${SYSROOT_DESTDIR}${base_prefix} \
> ${S} ${B}"
>
> # Could be compile but populate_sysroot and do_install shouldn't run at the same time
> @@ -103,8 +104,8 @@ python do_populate_sysroot () {
>
> SSTATETASKS += "do_populate_sysroot"
> do_populate_sysroot[sstate-name] = "populate-sysroot"
> -do_populate_sysroot[sstate-inputdirs] = "${SYSROOT_DESTDIR}/${STAGING_DIR}"
> -do_populate_sysroot[sstate-outputdirs] = "${TMPDIR}/sysroots"
> +do_populate_sysroot[sstate-inputdirs] = "${SYSROOT_DESTDIR}/${base_prefix}"
> +do_populate_sysroot[sstate-outputdirs] = "${STAGING_DIR_HOST}/"
The reason why is I'm still not 100% comfortable with the two lines
above. They look unbalanced and should really be paired to match such
that they'd either be:
do_populate_sysroot[sstate-inputdirs] = "${SYSROOT_DESTDIR}/${base_prefix}"
do_populate_sysroot[sstate-outputdirs] = "${STAGING_DIR_HOST}/${base_prefix}"
or
do_populate_sysroot[sstate-inputdirs] = "${SYSROOT_DESTDIR}"
do_populate_sysroot[sstate-outputdirs] = "${STAGING_DIR_HOST}"
but just looking at code I'm having trouble working out whether either,
both or neither of these would work and in which cases.
The cross/crosssdk recipes are a mess as they install into two different
sysroots. Dexuan is working on some changes to improve that so I'm least
worried about those two classes and the patch I just merged for libgcc
is going to help. I'm therefore aiming for this to work for the
target/native/nativesdk/cross-canadian cases and the cross/crosssdk one
is secondary and can be messy for now if needed.
I feel most comfortable aiming for the latter of my two options above,
does that work assuming we hack cross/crosssdk to do whatever is needed
to make them work with it? Can the others then remain unaltered?
The other minor thing I'm thinking about is adding a definition of:
bindir_crossscripts = "${bindir}/crossscripts"
to bitbake.conf (near the top next to bindir_cross) and using that in
the sysroot functions.
Cheers,
Richard
next prev parent reply other threads:[~2011-01-17 22:42 UTC|newest]
Thread overview: 11+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-01-14 22:14 [PATCH 0/5][RFC v4] Machine specific sysroot implementation Dongxiao Xu
2011-01-14 22:14 ` [PATCH 1/5] staging: Use relative path in sysroot-destdir for target recipes Dongxiao Xu
2011-01-17 22:42 ` Richard Purdie [this message]
2011-01-18 0:45 ` Xu, Dongxiao
2011-01-18 1:33 ` Xu, Dongxiao
2011-01-14 22:14 ` [PATCH 2/5] staging: relocate *.la paths in destination dirs Dongxiao Xu
2011-01-17 22:10 ` Richard Purdie
2011-01-14 22:14 ` [PATCH 3/5] bitbake: Introduce stamp-extra-info into build stamp file Dongxiao Xu
2011-01-18 12:53 ` Richard Purdie
2011-01-14 22:14 ` [PATCH 4/5] bitbake: machine specific sysroots implementation Dongxiao Xu
2011-01-14 22:14 ` [PATCH 5/5] emenlow: Change PACKAGE_EXTRA_ARCHS and BASE_PACKAGE_ARCH to core2 Dongxiao Xu
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=1295304133.14388.15857.camel@rex \
--to=richard.purdie@linuxfoundation.org \
--cc=dongxiao.xu@intel.com \
--cc=poky@yoctoproject.org \
/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