initscripts: remove -i from halt/reboot arguments and allow override
cover initscripts: remove -i from halt/reboot arguments and allow override
From: Paul Eggleton <[email protected]>
Removes -i from halt/reboot arguments to fix shutdown/reboot with NFS root, and allows halt/reboot arguments to be overridden using a new HALTARGS variable. Refer to bug [1] and discussion [2].
[1] http://bugzilla.pokylinux.org/show_bug.cgi?id=997
[2] https://lists.yoctoproject.org/pipermail/poky/2011-April/005395.html
Pull URL: git://git.pokylinux.org/poky-contrib.git
Branch: paule/haltargs
Browse: http://git.pokylinux.org/cgit.cgi/poky-contrib/log/?h=paule/haltargs
Thanks,
Paul Eggleton <[email protected]>
---
Paul Eggleton (1):
initscripts: remove -i from halt/reboot arguments and allow override
meta/recipes-core/initscripts/initscripts-1.0/halt | 2 +- .../initscripts/initscripts-1.0/reboot | 2 +- meta/recipes-core/initscripts/initscripts_1.0.bb | 9 ++++++++- 3 files changed, 10 insertions(+), 3 deletions(-)
PATCH 1/1 initscripts: remove -i from halt/reboot arguments and allow override
From: Paul Eggleton <[email protected]>
Introduces a variable HALTARGS which specifies the arguments sent to halt and reboot, and sets the default value to "-d -f", dropping the previous -i (shut down all network interfaces before halt/reboot, which causes a freeze with NFS root.)
Fixes [YOCTO #997].
--- meta/recipes-core/initscripts/initscripts-1.0/halt | 2 +- .../initscripts/initscripts-1.0/reboot | 2 +- meta/recipes-core/initscripts/initscripts_1.0.bb | 9 ++++++++- 3 files changed, 10 insertions(+), 3 deletions(-)
meta/recipes-core/initscripts/initscripts-1.0/halt +1 -1
@@ -24,6 +24,6 @@ then hddown="" fi -halt -d -f -i -p $hddown +halt SED_HALTARGS -p $hddown : exit 0
meta/recipes-core/initscripts/initscripts-1.0/reboot +1 -1
@@ -12,4 +12,4 @@ PATH=/sbin:/bin:/usr/sbin:/usr/bin echo -n "Rebooting... " -reboot -d -f -i +reboot SED_HALTARGS
meta/recipes-core/initscripts/initscripts_1.0.bb +8 -1
@@ -6,7 +6,7 @@ DEPENDS = "makedevs" RDEPENDS_${PN} = "makedevs" LICENSE = "GPLv2" LIC_FILES_CHKSUM = "file://COPYING;md5=751419260aa954499f7abaabaa882bbe" -PR = "r124" +PR = "r125" SRC_URI = "file://functions \ file://halt \
@@ -38,6 +38,13 @@ SRC_URI_append_arm = " file://alignment.sh" KERNEL_VERSION = "" +HALTARGS ?= "-d -f" + +do_configure() { + sed -i -e "s:SED_HALTARGS:${HALTARGS}:g" ${WORKDIR}/halt + sed -i -e "s:SED_HALTARGS:${HALTARGS}:g" ${WORKDIR}/reboot +} + do_install () { # # Create directories and install device independent scripts
Signature · 1 line
1.7.1
Quoted 8 lines · Paul Eggleton
From: Paul Eggleton <[email protected]>
Introduces a variable HALTARGS which specifies the arguments sent to halt and reboot, and sets the default value to "-d -f", dropping the previous -i (shut down all network interfaces before halt/reboot, which causes a freeze with NFS root.)
Fixes [YOCTO #997].
Is the SED_* tag a common search/replace mechanism in OE recipes? It's fine, I'm just wondering if there is an accepted best practice for this sort of thing. I know we do some sed replacements of /usr/bin and /etc.
Quoted 2 lines
Signed-off-by Paul Eggleton <[email protected]>
Quoted 55 lines
--- meta/recipes-core/initscripts/initscripts-1.0/halt | 2 +- .../initscripts/initscripts-1.0/reboot | 2 +- meta/recipes-core/initscripts/initscripts_1.0.bb | 9 ++++++++- 3 files changed, 10 insertions(+), 3 deletions(-)meta/recipes-core/initscripts/initscripts-1.0/halt +1
-1@@ -24,6 +24,6 @@ then hddown="" fi -halt -d -f -i -p $hddown +halt SED_HALTARGS -p $hddown : exit 0meta/recipes-core/initscripts/initscripts-1.0/reboot +1
-1@@ -12,4 +12,4 @@ PATH=/sbin:/bin:/usr/sbin:/usr/bin echo -n "Rebooting... " -reboot -d -f -i +reboot SED_HALTARGSmeta/recipes-core/initscripts/initscripts_1.0.bb +8
-1@@ -6,7 +6,7 @@ DEPENDS = "makedevs" RDEPENDS_${PN} = "makedevs" LICENSE = "GPLv2" LIC_FILES_CHKSUM = "file://COPYING;md5=751419260aa954499f7abaabaa882bbe" -PR = "r124" +PR = "r125" SRC_URI = "file://functions \ file://halt \@@ -38,6 +38,13 @@ SRC_URI_append_arm = " file://alignment.sh" KERNEL_VERSION = "" +HALTARGS ?= "-d -f" + +do_configure() { + sed -i -e "s:SED_HALTARGS:${HALTARGS}:g" ${WORKDIR}/halt + sed -i -e "s:SED_HALTARGS:${HALTARGS}:g" ${WORKDIR}/reboot +} + do_install () { # # Create directories and install device independent scripts
Signature · 3 lines
Darren Hart Intel Open Source Technology Center Yocto Project - Linux Kernel
Quoted 3 lines · On Wednesday 27 April 2011 18:56:09 Darren Hart wrote:
Is the SED_* tag a common search/replace mechanism in OE recipes? It's fine, I'm just wondering if there is an accepted best practice for this sort of thing. I know we do some sed replacements of
/usr/binand /etc.
Is there a more appropriate mechanism for substituting values of variables into files (genuinely asking)? If this is the best mechanism it's possible we might want to have a standard naming scheme though, I just made this one up on the spot - I see in some other places we've used FIXMExxxx, so at least it's not the worst :)
Cheers,
Paul
Signature · 2 lines
Paul Eggleton Intel Open Source Technology Centre
Op 28 apr 2011, om 12:13 heeft Paul Eggleton het volgende geschreven:
Quoted 7 lines · On Wednesday 27 April 2011 18:56:09 Darren Hart wrote:
Quoted 3 lines
Is the SED_* tag a common search/replace mechanism in OE recipes? It's fine, I'm just wondering if there is an accepted best practice for this sort of thing. I know we do some sed replacements of
/usr/binand /etc.Is there a more appropriate mechanism for substituting values of variables into files (genuinely asking)?
It's the one I use :)
regards,
Koen
Quoted 10 lines · Paul Eggleton
Quoted 3 lines · On Wednesday 27 April 2011 18:56:09 Darren Hart wrote:
Is the SED_* tag a common search/replace mechanism in OE recipes? It's fine, I'm just wondering if there is an accepted best practice for this sort of thing. I know we do some sed replacements of
/usr/binand /etc.Is there a more appropriate mechanism for substituting values of variables into files (genuinely asking)? If this is the best mechanism it's possible we might want to have a standard naming scheme though, I just made this one up on the spot - I see in some other places we've used FIXMExxxx, so at least it's not the worst :)
There is an effort to document how to do initscripts, so including something along these lines would be appropriate there. I have no objection to SED_.... FIXMExxx that's a different story ;-)
Quoted 4 lines
Cheers,
Paul
Signature · 3 lines
Darren Hart Intel Open Source Technology Center Yocto Project - Linux Kernel