The original solution replaced all overrides with the package name that
was being checked. This can have unforseen consequences where some
variable involved in defining the value for the PKG:<package> variable
may rely on some override which is not set as expected. It also meant
that any PKG variable set using an override would not be caught, e.g.,
PKG:${PN}:${MACHINE} = "${PN}-dev" (made up example that would always
fail with the old code).
Signed-off-by: Peter Kjellerstedt <peter.kjellerst...@axis.com>
---
meta/classes-global/package.bbclass | 8 +++-----
1 file changed, 3 insertions(+), 5 deletions(-)
diff --git a/meta/classes-global/package.bbclass
b/meta/classes-global/package.bbclass
index 0338a5c690..7787042e21 100644
--- a/meta/classes-global/package.bbclass
+++ b/meta/classes-global/package.bbclass
@@ -508,11 +508,9 @@ python do_package () {
# Check for conflict between renamed packages and existing ones
# for each package in PACKAGES, check if it will be renamed to an existing
one
for p in packages:
- localdata = bb.data.createCopy(d)
- localdata.setVar('OVERRIDES', p)
- rename = localdata.getVar('PKG')
- if (rename != None) and rename in packages:
- bb.fatal('package "%s" is renamed to "%s" using PKG:%s, but
package name already exists'%(p,rename,p))
+ rename = d.getVar('PKG:%s' % p)
+ if rename and rename in packages:
+ bb.fatal('package "%s" is renamed to "%s" using PKG:%s, but
package name already exists' % (p, rename, p))
###########################################################################
# Optimisations
-=-=-=-=-=-=-=-=-=-=-=-
Links: You receive all messages sent to this group.
View/Reply Online (#188646):
https://lists.openembedded.org/g/openembedded-core/message/188646
Mute This Topic: https://lists.openembedded.org/mt/101739066/21656
Group Owner: openembedded-core+ow...@lists.openembedded.org
Unsubscribe: https://lists.openembedded.org/g/openembedded-core/unsub
[arch...@mail-archive.com]
-=-=-=-=-=-=-=-=-=-=-=-
