[PATCH 0/4] Updated mc patch and enable udev-cache

[PATCH 0/4] Updated mc patch and enable udev-cache

[Saul Wold]

This patch set brings the mc helpers files back and packages
them.  It also enables the udev-cache code since there is a
boot time speed up.

Sau!


Saul Wold (4):
  mc: Don't remove libdir and split helpers into packages
  packagegroups: Update for split mc packages
  packagegroup-core-basic: remove hardcode udev
  packagegroup-core-boot: used udev-cache for sysvinit

 .../packagegroups/packagegroup-core-boot.bb        |  1 +
 .../packagegroups/packagegroup-self-hosted.bb      |  4 ++++
 meta/recipes-extended/mc/mc_4.8.8.bb               | 22 +++++++++++++++++++++-
 .../packagegroups/packagegroup-core-basic.bb       |  5 ++++-
 .../packagegroups/packagegroup-core-lsb.bb         |  4 ++++
 5 files changed, 34 insertions(+), 2 deletions(-)

-- 
1.8.1.4

[PATCH 1/4] mc: Don't remove libdir and split helpers into packages

[Saul Wold]

It contains helper programs that are needed to make mc do the right actions
for the various file formats it understands.

The helpers are perl, python and shell scripts, split them out so the core
mc does not try to pull in perl and python, it will still run without these
helpers.

[YOCTO #4432]

Signed-off-by: Saul Wold <sgw@linux.intel.com>
---
 meta/recipes-extended/mc/mc_4.8.8.bb | 22 +++++++++++++++++++++-
 1 file changed, 21 insertions(+), 1 deletion(-)

diff --git a/meta/recipes-extended/mc/mc_4.8.8.bb b/meta/recipes-extended/mc/mc_4.8.8.bb
index 20ae797..b2851bb 100644
--- a/meta/recipes-extended/mc/mc_4.8.8.bb
+++ b/meta/recipes-extended/mc/mc_4.8.8.bb
@@ -19,5 +19,25 @@ FILES_${PN}-dbg += "${libexecdir}/mc/.debug/"
 
 do_install_append () {
 	sed -i -e '1s,#!.*perl,#!${bindir}/env perl,' ${D}${libexecdir}/mc/extfs.d/*
-	rm -rf ${D}${libdir}
 }
+
+PACKAGES =+ "${BPN}-helpers-perl ${BPN}-helpers-python ${BPN}-helpers ${BPN}-fish"
+
+DESCRIPTION_${BPN}-helpers-perl = "Midnight Commander perl based helper scripts" 
+FILES_${BPN}-helpers-perl = "${libexecdir}/mc/extfs.d/a+ ${libexecdir}/mc/extfs.d/apt+ \
+                             ${libexecdir}/mc/extfs.d/deb ${libexecdir}/mc/extfs.d/deba \
+                             ${libexecdir}/mc/extfs.d/debd ${libexecdir}/mc/extfs.d/dpkg+ \
+                             ${libexecdir}/mc/extfs.d/mailfs ${libexecdir}/mc/extfs.d/patchfs \ 
+                             ${libexecdir}/mc/extfs.d/rpms+ ${libexecdir}/mc/extfs.d/ulib \ 
+                             ${libexecdir}/mc/extfs.d/uzip"
+RDEPENDS_${BPN}-helpers-perl = "perl"
+
+DESCRIPTION_${BPN}-helpers-python = "Midnight Commander python based helper scripts" 
+FILES_${BPN}-helpers-python = "${libexecdir}/mc/extfs.d/s3+ ${libexecdir}/mc/extfs.d/uc1541"
+RDEPENDS_${BPN}-helpers-python = "python"
+
+DESCRIPTION_${BPN}-helpers = "Midnight Commander shell helper scripts" 
+FILES_${BPN}-helpers = "${libexecdir}/mc/extfs.d/* ${libexecdir}/mc/ext.d/*"
+
+DESCRIPTION_${BPN}-fish = "Midnight Commander Fish scripts"
+FILES_${BPN}-fish = "${libexecdir}/mc/fish"
-- 
1.8.1.4

[PATCH 2/4] packagegroups: Update for split mc packages

[Saul Wold]

Signed-off-by: Saul Wold <sgw@linux.intel.com>
---
 meta/recipes-core/packagegroups/packagegroup-self-hosted.bb    | 4 ++++
 meta/recipes-extended/packagegroups/packagegroup-core-basic.bb | 4 ++++
 meta/recipes-extended/packagegroups/packagegroup-core-lsb.bb   | 4 ++++
 3 files changed, 12 insertions(+)

diff --git a/meta/recipes-core/packagegroups/packagegroup-self-hosted.bb b/meta/recipes-core/packagegroups/packagegroup-self-hosted.bb
index ed73617..d4e89a1 100644
--- a/meta/recipes-core/packagegroups/packagegroup-self-hosted.bb
+++ b/meta/recipes-core/packagegroups/packagegroup-self-hosted.bb
@@ -39,6 +39,10 @@ RDEPENDS_packagegroup-self-hosted-host-tools = "\
     iptables \
     lsb \
     mc \
+    mc-fish \
+    mc-helpers \
+    mc-helpers-perl \
+    mc-helpers-python \
     parted \
     pseudo \
     screen \
diff --git a/meta/recipes-extended/packagegroups/packagegroup-core-basic.bb b/meta/recipes-extended/packagegroups/packagegroup-core-basic.bb
index 00405d6..47b980e 100644
--- a/meta/recipes-extended/packagegroups/packagegroup-core-basic.bb
+++ b/meta/recipes-extended/packagegroups/packagegroup-core-basic.bb
@@ -59,6 +59,10 @@ RDEPENDS_packagegroup-core-basic-utils = "\
     grep \
     makedevs \
     mc \
+    mc-fish \
+    mc-helpers \
+    mc-helpers-perl \
+    mc-helpers-python \
     mktemp \
     ncurses \
     net-tools \
diff --git a/meta/recipes-extended/packagegroups/packagegroup-core-lsb.bb b/meta/recipes-extended/packagegroups/packagegroup-core-lsb.bb
index ed854b8..d71db26 100644
--- a/meta/recipes-extended/packagegroups/packagegroup-core-lsb.bb
+++ b/meta/recipes-extended/packagegroups/packagegroup-core-lsb.bb
@@ -50,6 +50,10 @@ RDEPENDS_packagegroup-core-sys-extended = "\
     lrzsz \
     lzo \
     mc \
+    mc-fish \
+    mc-helpers \
+    mc-helpers-perl \
+    mc-helpers-python \
     mdadm \
     minicom \
     neon \
-- 
1.8.1.4

[PATCH 3/4] packagegroup-core-basic: remove hardcode udev

[Saul Wold]

udev is brought in via packagegroup-core-boot

Signed-off-by: Saul Wold <sgw@linux.intel.com>
---
 meta/recipes-extended/packagegroups/packagegroup-core-basic.bb | 1 -
 1 file changed, 1 deletion(-)

diff --git a/meta/recipes-extended/packagegroups/packagegroup-core-basic.bb b/meta/recipes-extended/packagegroups/packagegroup-core-basic.bb
index 47b980e..e802a02 100644
--- a/meta/recipes-extended/packagegroups/packagegroup-core-basic.bb
+++ b/meta/recipes-extended/packagegroups/packagegroup-core-basic.bb
@@ -73,7 +73,6 @@ RDEPENDS_packagegroup-core-basic-utils = "\
     sed \
     tar \
     time \
-    udev \
     util-linux \
     zlib \
     "
-- 
1.8.1.4

[PATCH 4/4] packagegroup-core-boot: used udev-cache for sysvinit

[Saul Wold]

Enabling this by default since it does improve boot time when using sysvinit udev

Signed-off-by: Saul Wold <sgw@linux.intel.com>
---
 meta/recipes-core/packagegroups/packagegroup-core-boot.bb | 1 +
 1 file changed, 1 insertion(+)

diff --git a/meta/recipes-core/packagegroups/packagegroup-core-boot.bb b/meta/recipes-core/packagegroups/packagegroup-core-boot.bb
index d69c41d..1e1ee8e 100644
--- a/meta/recipes-core/packagegroups/packagegroup-core-boot.bb
+++ b/meta/recipes-core/packagegroups/packagegroup-core-boot.bb
@@ -43,6 +43,7 @@ RDEPENDS_${PN} = "\
     ${VIRTUAL-RUNTIME_init_manager} \
     ${VIRTUAL-RUNTIME_initscripts} \
     ${VIRTUAL-RUNTIME_dev_manager} \
+    ${@base_contains("DISTRO_FEATURES", "systemd", "", "udev-cache", d)} \
     ${VIRTUAL-RUNTIME_update-alternatives} \
     ${MACHINE_ESSENTIAL_EXTRA_RDEPENDS}"
 
-- 
1.8.1.4

Re: [PATCH 4/4] packagegroup-core-boot: used udev-cache for sysvinit

[Paul Eggleton]

On Thursday 27 June 2013 15:43:17 Saul Wold wrote:
> Enabling this by default since it does improve boot time when using sysvinit
> udev
> 
> Signed-off-by: Saul Wold <sgw@linux.intel.com>
> ---
>  meta/recipes-core/packagegroups/packagegroup-core-boot.bb | 1 +
>  1 file changed, 1 insertion(+)
> 
> diff --git a/meta/recipes-core/packagegroups/packagegroup-core-boot.bb
> b/meta/recipes-core/packagegroups/packagegroup-core-boot.bb index
> d69c41d..1e1ee8e 100644
> --- a/meta/recipes-core/packagegroups/packagegroup-core-boot.bb
> +++ b/meta/recipes-core/packagegroups/packagegroup-core-boot.bb
> @@ -43,6 +43,7 @@ RDEPENDS_${PN} = "\
>      ${VIRTUAL-RUNTIME_init_manager} \
>      ${VIRTUAL-RUNTIME_initscripts} \
>      ${VIRTUAL-RUNTIME_dev_manager} \
> +    ${@base_contains("DISTRO_FEATURES", "systemd", "", "udev-cache", d)} \
>      ${VIRTUAL-RUNTIME_update-alternatives} \
>      ${MACHINE_ESSENTIAL_EXTRA_RDEPENDS}"

I appreciate the intention here but I'm not sure if this is really the 
appropriate place to put this; packagegroup-core-boot is supposed to only 
directly pull in the essentials required for booting. This will cause problems 
for those people using alternative device managers (e.g. busybox mdev) as 
well.

I'm not sure where this should go, but adding it to VIRTUAL-
RUNTIME_dev_manager (in the distro config?) could work. Anyone else have any 
better suggestion?

Cheers,
Paul

-- 

Paul Eggleton
Intel Open Source Technology Centre

Re: [PATCH 4/4] packagegroup-core-boot: used udev-cache for sysvinit

[Burton, Ross]

On 28 June 2013 11:02, Paul Eggleton <paul.eggleton@linux.intel.com> wrote:
> I appreciate the intention here but I'm not sure if this is really the
> appropriate place to put this; packagegroup-core-boot is supposed to only
> directly pull in the essentials required for booting. This will cause problems
> for those people using alternative device managers (e.g. busybox mdev) as
> well.
>
> I'm not sure where this should go, but adding it to VIRTUAL-
> RUNTIME_dev_manager (in the distro config?) could work. Anyone else have any
> better suggestion?

How about making udev RRECOMMEND udev-cache?

Ross

Re: [PATCH 4/4] packagegroup-core-boot: used udev-cache for sysvinit

[Paul Eggleton]

On Monday 01 July 2013 15:49:00 Burton, Ross wrote:
> On 28 June 2013 11:02, Paul Eggleton <paul.eggleton@linux.intel.com> wrote:
> > I appreciate the intention here but I'm not sure if this is really the
> > appropriate place to put this; packagegroup-core-boot is supposed to only
> > directly pull in the essentials required for booting. This will cause
> > problems for those people using alternative device managers (e.g. busybox
> > mdev) as well.
> > 
> > I'm not sure where this should go, but adding it to VIRTUAL-
> > RUNTIME_dev_manager (in the distro config?) could work. Anyone else have
> > any better suggestion?
> 
> How about making udev RRECOMMEND udev-cache?

It's not that udev-cache might not be available, rather that packagegroup-
core-boot shouldn't have any kind of reference to it. In any case unless udev 
is skipped somehow, an RRECOMMENDS on udev-cache will end up building udev 
even if you haven't selected it in VIRTUAL-RUNTIME_dev_manager so that won't 
help.

Cheers,
Paul

-- 

Paul Eggleton
Intel Open Source Technology Centre

Re: [PATCH 4/4] packagegroup-core-boot: used udev-cache for sysvinit

[Burton, Ross]

On 1 July 2013 15:58, Paul Eggleton <paul.eggleton@linux.intel.com> wrote:
> On Monday 01 July 2013 15:49:00 Burton, Ross wrote:
>> On 28 June 2013 11:02, Paul Eggleton <paul.eggleton@linux.intel.com> wrote:
>> > I appreciate the intention here but I'm not sure if this is really the
>> > appropriate place to put this; packagegroup-core-boot is supposed to only
>> > directly pull in the essentials required for booting. This will cause
>> > problems for those people using alternative device managers (e.g. busybox
>> > mdev) as well.
>> >
>> > I'm not sure where this should go, but adding it to VIRTUAL-
>> > RUNTIME_dev_manager (in the distro config?) could work. Anyone else have
>> > any better suggestion?
>>
>> How about making udev RRECOMMEND udev-cache?
>
> It's not that udev-cache might not be available, rather that packagegroup-
> core-boot shouldn't have any kind of reference to it. In any case unless udev
> is skipped somehow, an RRECOMMENDS on udev-cache will end up building udev
> even if you haven't selected it in VIRTUAL-RUNTIME_dev_manager so that won't
> help.

I mean add a recommends to the udev binary package itself, so you'll
get udev-cache if you build udev (and keep the existing udev-selection
logic as-is).

Ross

Re: [PATCH 4/4] packagegroup-core-boot: used udev-cache for sysvinit

[Paul Eggleton]

On Monday 01 July 2013 16:02:56 Burton, Ross wrote:
> On 1 July 2013 15:58, Paul Eggleton <paul.eggleton@linux.intel.com> wrote:
> > On Monday 01 July 2013 15:49:00 Burton, Ross wrote:
> >> On 28 June 2013 11:02, Paul Eggleton <paul.eggleton@linux.intel.com> 
wrote:
> >> > I appreciate the intention here but I'm not sure if this is really the
> >> > appropriate place to put this; packagegroup-core-boot is supposed to
> >> > only
> >> > directly pull in the essentials required for booting. This will cause
> >> > problems for those people using alternative device managers (e.g.
> >> > busybox
> >> > mdev) as well.
> >> > 
> >> > I'm not sure where this should go, but adding it to VIRTUAL-
> >> > RUNTIME_dev_manager (in the distro config?) could work. Anyone else
> >> > have any better suggestion?
> >> 
> >> How about making udev RRECOMMEND udev-cache?
> > 
> > It's not that udev-cache might not be available, rather that packagegroup-
> > core-boot shouldn't have any kind of reference to it. In any case unless
> > udev is skipped somehow, an RRECOMMENDS on udev-cache will end up
> > building udev even if you haven't selected it in
> > VIRTUAL-RUNTIME_dev_manager so that won't help.
> 
> I mean add a recommends to the udev binary package itself, so you'll
> get udev-cache if you build udev (and keep the existing udev-selection
> logic as-is).

Right, that could work yes.

Cheers,
Paul

-- 

Paul Eggleton
Intel Open Source Technology Centre