On 4/21/25 20:26, Richard Purdie wrote:
CAUTION: This email comes from a non Wind River email account!
Do not click links or open attachments unless you recognize the sender and know the content is safe.

On Fri, 2025-04-18 at 16:19 +0800, Changqing Li via lists.openembedded.org wrote:
From: Changqing Li <changqing.li@windriver.com>

icu will check program install path during do_configure, eg:
checking for a BSD-compatible install... /path/to/install -c

And this path will be writen into pkgdata.inc:
INSTALL_CMD=$(INSTALL-L)

Decided by if install is installed into recipe-sysroot-native during
do_configure stage, the INSTALL_CMD could be
/build/tmp/work/corei7-64-wrs-linux/icu/76-1/recipe-sysroot-native/usr/bin/install
or /build/tmp/hosttools/install

Add coreutils-native as DEPENDS to make a determined result of
INSTALL_CMD, avoid vary caused by the execute sequence of another task
which DEPENDS on coreutils-native and also independent with do_configure

Signed-off-by: Changqing Li <changqing.li@windriver.com>
---
 meta/recipes-support/icu/icu_76-1.bb | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
This needs to be fixed differently, we should be forcing it to use PATH
and set the value to just "install". We don't want to add dependencies
unless we really need them, they slow the builds down and add
inefficiencies.

Thanks Richard. Currently, it will find "install" in PATH,  but STAGING_BINDIR_NATIVE is before HOSTTOOLS_DIR,

so when there is install under STAGING_BINDIR_NATIVE, it will use this one,  if not,  it will use the one under

HOSTTOOLS_DIR.  According to your comments,  I send a V3,  force it to use HOSTTOOLS_DIR/install directly,

please help to review.

Regards

//Changqing


Cheers,

Richard