[daisy][PATCH 1/3] layer.conf: Add in useradd dependencies as ABISAFE dependencies

[daisy][PATCH 1/3] layer.conf: Add in useradd dependencies as ABISAFE dependencies

[Steffen Sledz]

Currently allarch recipes using useradd rebuild each time MACHINE
changes which is not desireable. Adding the useradd dependencies
to this list ensures they do not change under these circumstances.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Steffen Sledz <sledz@dresearch-fe.de>
---
 meta/conf/layer.conf | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/meta/conf/layer.conf b/meta/conf/layer.conf
index 18609dd..a764ee0 100644
--- a/meta/conf/layer.conf
+++ b/meta/conf/layer.conf
@@ -30,4 +30,8 @@ SIGGEN_EXCLUDERECIPES_ABISAFE += " \
   udev-extraconf \
   packagegroup-x11-xserver \
   systemd-serialgetty \
+  initscripts \                                                                                                                                                                                                                                    
+  shadow \
+  shadow-sysroot \
+  base-passwd \
 "
-- 
2.1.1

[daisy][PATCH 2/3] update-rc.d/systemd: Remove OVERRIDES dependency

[Steffen Sledz]

Taking run-postinsts and building for two machines which have different
OVERRIDES leads to two different sets of stamps for an allarch package.

We don't need to depend on OVERRIDES in these classes, the end resulting
variables are good enough. We can therefore exclude the dependency
and allow a single package to be generated for run-postinsts.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Steffen Sledz <sledz@dresearch-fe.de>
---
 meta/classes/systemd.bbclass     | 1 +
 meta/classes/update-rc.d.bbclass | 1 +
 2 files changed, 2 insertions(+)

diff --git a/meta/classes/systemd.bbclass b/meta/classes/systemd.bbclass
index 2bd63a4..1516cd9 100644
--- a/meta/classes/systemd.bbclass
+++ b/meta/classes/systemd.bbclass
@@ -55,6 +55,7 @@ fi
 
 
 systemd_populate_packages[vardeps] += "systemd_prerm systemd_postinst"
+systemd_populate_packages[vardepsexclude] += "OVERRIDES"
 
 python systemd_populate_packages() {
     if not oe.utils.contains('DISTRO_FEATURES', 'systemd', True, False, d):
diff --git a/meta/classes/update-rc.d.bbclass b/meta/classes/update-rc.d.bbclass
index 668ead8..e8debdf 100644
--- a/meta/classes/update-rc.d.bbclass
+++ b/meta/classes/update-rc.d.bbclass
@@ -71,6 +71,7 @@ PACKAGESPLITFUNCS_prepend = "populate_packages_updatercd "
 PACKAGESPLITFUNCS_remove_class-nativesdk = "populate_packages_updatercd "
 
 populate_packages_updatercd[vardeps] += "updatercd_prerm updatercd_postrm updatercd_preinst updatercd_postinst"
+populate_packages_updatercd[vardepsexclude] += "OVERRIDES"
 
 python populate_packages_updatercd () {
     def update_rcd_auto_depend(pkg):
-- 
2.1.1

[daisy][PATCH 3/3] layer.conf: Mark opkg-utils as ABISAFE for update-alternatives usage

[Steffen Sledz]

Currently linux-firmware rebuilds for each machine due to its usage of
update-alternatives which in turn means a dependency on opkg-utils.
Marking opkg-utils as ABISAFE is the only option we have right now
to avoid this.

Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Signed-off-by: Steffen Sledz <sledz@dresearch-fe.de>
---
 meta/conf/layer.conf | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/meta/conf/layer.conf b/meta/conf/layer.conf
index a764ee0..13814dc 100644
--- a/meta/conf/layer.conf
+++ b/meta/conf/layer.conf
@@ -14,6 +14,7 @@ LAYERVERSION_core = "4"
 # Set a variable to get to the top of the metadata location
 COREBASE = '${@os.path.normpath("${LAYERDIR}/../")}'
 
+# opkg-utils is for update-altnernatives :(
 SIGGEN_EXCLUDERECIPES_ABISAFE += " \
   sysvinit-inittab \
   shadow-securetty \
@@ -34,4 +35,5 @@ SIGGEN_EXCLUDERECIPES_ABISAFE += " \
   shadow \
   shadow-sysroot \
   base-passwd \
+  opkg-utils \
 "
-- 
2.1.1

Re: [daisy][PATCH 1/3] layer.conf: Add in useradd dependencies as ABISAFE dependencies

[Denys Dmytriyenko]

On Thu, Sep 25, 2014 at 02:17:55PM +0200, Steffen Sledz wrote:
> Currently allarch recipes using useradd rebuild each time MACHINE
> changes which is not desireable. Adding the useradd dependencies
> to this list ensures they do not change under these circumstances.
> 
> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
> Signed-off-by: Steffen Sledz <sledz@dresearch-fe.de>
> ---
>  meta/conf/layer.conf | 4 ++++
>  1 file changed, 4 insertions(+)
> 
> diff --git a/meta/conf/layer.conf b/meta/conf/layer.conf
> index 18609dd..a764ee0 100644
> --- a/meta/conf/layer.conf
> +++ b/meta/conf/layer.conf
> @@ -30,4 +30,8 @@ SIGGEN_EXCLUDERECIPES_ABISAFE += " \
>    udev-extraconf \
>    packagegroup-x11-xserver \
>    systemd-serialgetty \
> +  initscripts \                                                                                                                                                                                                                                    

This line above ^^^^ has a very long trail of whitespaces and it got commited 
that way, unfortunately :(


> +  shadow \
> +  shadow-sysroot \
> +  base-passwd \
>  "
> -- 
> 2.1.1
> 
> -- 
> _______________________________________________
> Openembedded-core mailing list
> Openembedded-core@lists.openembedded.org
> http://lists.openembedded.org/mailman/listinfo/openembedded-core

Re: [daisy][PATCH 1/3] layer.conf: Add in useradd dependencies as ABISAFE dependencies

[Khem Raj]

On Tue, Sep 30, 2014 at 5:55 PM, Denys Dmytriyenko <denis@denix.org> wrote:
> This line above ^^^^ has a very long trail of whitespaces and it got commited
> that way, unfortunately :(

Better was if you commented before it went in.

Re: [daisy][PATCH 1/3] layer.conf: Add in useradd dependencies as ABISAFE dependencies

[Denys Dmytriyenko]

On Tue, Sep 30, 2014 at 06:25:57PM -0700, Khem Raj wrote:
> On Tue, Sep 30, 2014 at 5:55 PM, Denys Dmytriyenko <denis@denix.org> wrote:
> > This line above ^^^^ has a very long trail of whitespaces and it got commited
> > that way, unfortunately :(
> 
> Better was if you commented before it went in.

And how would you suggest I was supposed to see it?? Nobody saw it. My email 
client is no different than others - it doesn't mark trailing whitespaces. The 
reason I saw it is because I ran "git log -p" on the file and that explicitly 
highlights trailing whitespaces in BRIGHT RED... :)

-- 
Denys

Re: [daisy][PATCH 1/3] layer.conf: Add in useradd dependencies as ABISAFE dependencies

[Khem Raj]

On Tue, Sep 30, 2014 at 6:32 PM, Denys Dmytriyenko <denis@denix.org> wrote:
> On Tue, Sep 30, 2014 at 06:25:57PM -0700, Khem Raj wrote:
>> On Tue, Sep 30, 2014 at 5:55 PM, Denys Dmytriyenko <denis@denix.org> wrote:
>> > This line above ^^^^ has a very long trail of whitespaces and it got commited
>> > that way, unfortunately :(
>>
>> Better was if you commented before it went in.
>
> And how would you suggest I was supposed to see it?? Nobody saw it. My email
> client is no different than others - it doesn't mark trailing whitespaces. The
> reason I saw it is because I ran "git log -p" on the file and that explicitly
> highlights trailing whitespaces in BRIGHT RED... :)

just for that may be we should start using gerrit ;)

>
> --
> Denys

Re: [daisy][PATCH 1/3] layer.conf: Add in useradd dependencies as ABISAFE dependencies

[Martin Jansa]

[-- Attachment #1: Type: text/plain, Size: 1012 bytes --]

On Tue, Sep 30, 2014 at 08:55:53PM -0700, Khem Raj wrote:
> On Tue, Sep 30, 2014 at 6:32 PM, Denys Dmytriyenko <denis@denix.org> wrote:
> > On Tue, Sep 30, 2014 at 06:25:57PM -0700, Khem Raj wrote:
> >> On Tue, Sep 30, 2014 at 5:55 PM, Denys Dmytriyenko <denis@denix.org> wrote:
> >> > This line above ^^^^ has a very long trail of whitespaces and it got commited
> >> > that way, unfortunately :(
> >>
> >> Better was if you commented before it went in.
> >
> > And how would you suggest I was supposed to see it?? Nobody saw it. My email
> > client is no different than others - it doesn't mark trailing whitespaces. The
> > reason I saw it is because I ran "git log -p" on the file and that explicitly
> > highlights trailing whitespaces in BRIGHT RED... :)
> 
> just for that may be we should start using gerrit ;)

I'm all for it, it would make my e-mail - review - patchwork -
master-next workflow a lot easier and reliable.

-- 
Martin 'JaMa' Jansa     jabber: Martin.Jansa@gmail.com

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 188 bytes --]

Re: [daisy][PATCH 1/3] layer.conf: Add in useradd dependencies as ABISAFE dependencies

[Richard Purdie]

On Wed, 2014-10-01 at 06:10 +0200, Martin Jansa wrote:
> On Tue, Sep 30, 2014 at 08:55:53PM -0700, Khem Raj wrote:
> > On Tue, Sep 30, 2014 at 6:32 PM, Denys Dmytriyenko <denis@denix.org> wrote:
> > > On Tue, Sep 30, 2014 at 06:25:57PM -0700, Khem Raj wrote:
> > >> On Tue, Sep 30, 2014 at 5:55 PM, Denys Dmytriyenko <denis@denix.org> wrote:
> > >> > This line above ^^^^ has a very long trail of whitespaces and it got commited
> > >> > that way, unfortunately :(
> > >>
> > >> Better was if you commented before it went in.
> > >
> > > And how would you suggest I was supposed to see it?? Nobody saw it. My email
> > > client is no different than others - it doesn't mark trailing whitespaces. The
> > > reason I saw it is because I ran "git log -p" on the file and that explicitly
> > > highlights trailing whitespaces in BRIGHT RED... :)
> > 
> > just for that may be we should start using gerrit ;)
> 
> I'm all for it, it would make my e-mail - review - patchwork -
> master-next workflow a lot easier and reliable.

And I am not. Why? I've actually tried using the thing and I absolutely
*despise* its interface. This isn't just a mild dislike, I simply cannot
do the things I want/need to do with it.

I believe I should have some kind of a say in this since it will
directly affect my workflow and general life (since I spend a lot of my
time dealing with this). Its appears to be very easy for other people to
tell me how I should work though :(.

Cheers,

Richard

Re: [daisy][PATCH 1/3] layer.conf: Add in useradd dependencies as ABISAFE dependencies

[Koen Kooi]

Op 1 okt. 2014, om 10:12 heeft Richard Purdie <richard.purdie@linuxfoundation.org> het volgende geschreven:

> On Wed, 2014-10-01 at 06:10 +0200, Martin Jansa wrote:
>> On Tue, Sep 30, 2014 at 08:55:53PM -0700, Khem Raj wrote:
>>> On Tue, Sep 30, 2014 at 6:32 PM, Denys Dmytriyenko <denis@denix.org> wrote:
>>>> On Tue, Sep 30, 2014 at 06:25:57PM -0700, Khem Raj wrote:
>>>>> On Tue, Sep 30, 2014 at 5:55 PM, Denys Dmytriyenko <denis@denix.org> wrote:
>>>>>> This line above ^^^^ has a very long trail of whitespaces and it got commited
>>>>>> that way, unfortunately :(
>>>>> 
>>>>> Better was if you commented before it went in.
>>>> 
>>>> And how would you suggest I was supposed to see it?? Nobody saw it. My email
>>>> client is no different than others - it doesn't mark trailing whitespaces. The
>>>> reason I saw it is because I ran "git log -p" on the file and that explicitly
>>>> highlights trailing whitespaces in BRIGHT RED... :)
>>> 
>>> just for that may be we should start using gerrit ;)
>> 
>> I'm all for it, it would make my e-mail - review - patchwork -
>> master-next workflow a lot easier and reliable.
> 
> And I am not. Why? I've actually tried using the thing and I absolutely
> *despise* its interface. This isn't just a mild dislike, I simply cannot
> do the things I want/need to do with it.

I use gerrit at work and I despise the interface as well, but it does address a number is issues plaguing OE-core:

1) pull-request doesn't match emailed patches
2) Patches are forgotten
3) Wrong version of patch gets applied
4) Review comments aren't tracked

But patchwork can be used to address the above problems as well. Anyway, this isn't a purely technical problem, see below.

> I believe I should have some kind of a say in this since it will
> directly affect my workflow and general life (since I spend a lot of my
> time dealing with this). Its appears to be very easy for other people to
> tell me how I should work though :(.

I think the real solution doesn't involve extra layers of tools, but a move to a kernel style environment where subsystem maintainers collect and vet patches and send pull requests to RP. But that would require formalizing the commit and review rules which I suspect will be controversial.

--
Koen

Re: [daisy][PATCH 1/3] layer.conf: Add in useradd dependencies as ABISAFE dependencies

[Martin Jansa]

[-- Attachment #1: Type: text/plain, Size: 2285 bytes --]

On Wed, Oct 01, 2014 at 09:12:16AM +0100, Richard Purdie wrote:
> On Wed, 2014-10-01 at 06:10 +0200, Martin Jansa wrote:
> > On Tue, Sep 30, 2014 at 08:55:53PM -0700, Khem Raj wrote:
> > > On Tue, Sep 30, 2014 at 6:32 PM, Denys Dmytriyenko <denis@denix.org> wrote:
> > > > On Tue, Sep 30, 2014 at 06:25:57PM -0700, Khem Raj wrote:
> > > >> On Tue, Sep 30, 2014 at 5:55 PM, Denys Dmytriyenko <denis@denix.org> wrote:
> > > >> > This line above ^^^^ has a very long trail of whitespaces and it got commited
> > > >> > that way, unfortunately :(
> > > >>
> > > >> Better was if you commented before it went in.
> > > >
> > > > And how would you suggest I was supposed to see it?? Nobody saw it. My email
> > > > client is no different than others - it doesn't mark trailing whitespaces. The
> > > > reason I saw it is because I ran "git log -p" on the file and that explicitly
> > > > highlights trailing whitespaces in BRIGHT RED... :)
> > > 
> > > just for that may be we should start using gerrit ;)
> > 
> > I'm all for it, it would make my e-mail - review - patchwork -
> > master-next workflow a lot easier and reliable.
> 
> And I am not. Why? I've actually tried using the thing and I absolutely
> *despise* its interface. This isn't just a mild dislike, I simply cannot
> do the things I want/need to do with it.
> 
> I believe I should have some kind of a say in this since it will
> directly affect my workflow and general life (since I spend a lot of my
> time dealing with this).

> Its appears to be very easy for other people to tell me how I should work though :(.

Please calm down, we just said that we like it, we're not forcing it on
anyone and I don't think khem would start installing it on OE servers
without much longer and broader discussion between all people who should
have some kind of say in this..

FWIW: I despised the UI at first as well, then I got used to it (after a month
or so), now I really like it especially with the "New screen" UI
introduced IIRC in 2.8 version. Only annoying part for me is more
complicated permissions model (which I luckily don't need to
change so often) and missing option to use "simple" server side hooks.

Regards,

-- 
Martin 'JaMa' Jansa     jabber: Martin.Jansa@gmail.com

[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 188 bytes --]

Re: [daisy][PATCH 1/3] layer.conf: Add in useradd dependencies as ABISAFE dependencies

[Koen Kooi]

Op 1 okt. 2014, om 17:34 heeft Martin Jansa <martin.jansa@gmail.com> het volgende geschreven:

> On Wed, Oct 01, 2014 at 09:12:16AM +0100, Richard Purdie wrote:
>> On Wed, 2014-10-01 at 06:10 +0200, Martin Jansa wrote:
>>> On Tue, Sep 30, 2014 at 08:55:53PM -0700, Khem Raj wrote:
>>>> On Tue, Sep 30, 2014 at 6:32 PM, Denys Dmytriyenko <denis@denix.org> wrote:
>>>>> On Tue, Sep 30, 2014 at 06:25:57PM -0700, Khem Raj wrote:
>>>>>> On Tue, Sep 30, 2014 at 5:55 PM, Denys Dmytriyenko <denis@denix.org> wrote:
>>>>>>> This line above ^^^^ has a very long trail of whitespaces and it got commited
>>>>>>> that way, unfortunately :(
>>>>>> 
>>>>>> Better was if you commented before it went in.
>>>>> 
>>>>> And how would you suggest I was supposed to see it?? Nobody saw it. My email
>>>>> client is no different than others - it doesn't mark trailing whitespaces. The
>>>>> reason I saw it is because I ran "git log -p" on the file and that explicitly
>>>>> highlights trailing whitespaces in BRIGHT RED... :)
>>>> 
>>>> just for that may be we should start using gerrit ;)
>>> 
>>> I'm all for it, it would make my e-mail - review - patchwork -
>>> master-next workflow a lot easier and reliable.
>> 
>> And I am not. Why? I've actually tried using the thing and I absolutely
>> *despise* its interface. This isn't just a mild dislike, I simply cannot
>> do the things I want/need to do with it.
>> 
>> I believe I should have some kind of a say in this since it will
>> directly affect my workflow and general life (since I spend a lot of my
>> time dealing with this).
> 
>> Its appears to be very easy for other people to tell me how I should work though :(.
> 
> Please calm down, we just said that we like it, we're not forcing it on
> anyone and I don't think khem would start installing it on OE servers
> without much longer and broader discussion between all people who should
> have some kind of say in this..
> 
> FWIW: I despised the UI at first as well, then I got used to it (after a month
> or so), now I really like it especially with the "New screen" UI
> introduced IIRC in 2.8 version.

That 'new screen' is indeed an improvement and I also found the switch to make it show a unidiff instead of side-by-side. I still despise the UI, though, but not as much as before :)

regards,

Koen

> Only annoying part for me is more
> complicated permissions model (which I luckily don't need to
> change so often) and missing option to use "simple" server side hooks.
> 
> Regards,
> 
> -- 
> Martin 'JaMa' Jansa     jabber: Martin.Jansa@gmail.com
> -- 
> _______________________________________________
> Openembedded-core mailing list
> Openembedded-core@lists.openembedded.org
> http://lists.openembedded.org/mailman/listinfo/openembedded-core

Re: [daisy][PATCH 1/3] layer.conf: Add in useradd dependencies as ABISAFE dependencies

[Trevor Woerner]

On 10/03/14 04:24, Koen Kooi wrote:
> Op 1 okt. 2014, om 17:34 heeft Martin Jansa <martin.jansa@gmail.com> het volgende geschreven:
>
>> On Wed, Oct 01, 2014 at 09:12:16AM +0100, Richard Purdie wrote:
>>> On Wed, 2014-10-01 at 06:10 +0200, Martin Jansa wrote:
>>>> On Tue, Sep 30, 2014 at 08:55:53PM -0700, Khem Raj wrote:
>>>>> On Tue, Sep 30, 2014 at 6:32 PM, Denys Dmytriyenko <denis@denix.org> wrote:
>>>>>> On Tue, Sep 30, 2014 at 06:25:57PM -0700, Khem Raj wrote:
>>>>>>> On Tue, Sep 30, 2014 at 5:55 PM, Denys Dmytriyenko <denis@denix.org> wrote:
>>>>>>>> This line above ^^^^ has a very long trail of whitespaces and it got commited
>>>>>>>> that way, unfortunately :(
>>>>>>> Better was if you commented before it went in.
>>>>>> And how would you suggest I was supposed to see it?? Nobody saw it. My email
>>>>>> client is no different than others - it doesn't mark trailing whitespaces. The
>>>>>> reason I saw it is because I ran "git log -p" on the file and that explicitly
>>>>>> highlights trailing whitespaces in BRIGHT RED... :)
>>>>> just for that may be we should start using gerrit ;)
>>>> I'm all for it, it would make my e-mail - review - patchwork -
>>>> master-next workflow a lot easier and reliable.
>>> And I am not. Why? I've actually tried using the thing and I absolutely
>>> *despise* its interface. This isn't just a mild dislike, I simply cannot
>>> do the things I want/need to do with it.
>>>
>>> I believe I should have some kind of a say in this since it will
>>> directly affect my workflow and general life (since I spend a lot of my
>>> time dealing with this).
>>> Its appears to be very easy for other people to tell me how I should work though :(.
>> Please calm down, we just said that we like it, we're not forcing it on
>> anyone and I don't think khem would start installing it on OE servers
>> without much longer and broader discussion between all people who should
>> have some kind of say in this..
>>
>> FWIW: I despised the UI at first as well, then I got used to it (after a month
>> or so), now I really like it especially with the "New screen" UI
>> introduced IIRC in 2.8 version.
> That 'new screen' is indeed an improvement and I also found the switch to make it show a unidiff instead of side-by-side. I still despise the UI, though, but not as much as before :)

gerrit does provide a rather complete REST api:
https://gerrit-review.googlesource.com/Documentation/rest-api.html So if
someone wanted to "go wild" they could get quite a bit done just using
curl and some simple scripts.

E.g. to get the current list of changes:
$ curl -X GET -H "Accept: application/xml"
https://review.linaro.org/changes/

- to get the details of a specific change:
$ curl -X GET -H "Accept: application/xml"
https://review.linaro.org/changes/If7f0ead3471a08f7b8fe9151c6c31cdfbaef1c4f/

- to retrieve a patch as zip (converted from base64):
$ curl -X GET -H "Accept: application/xml"
https://review.linaro.org/changes/If7f0ead3471a08f7b8fe9151c6c31cdfbaef1c4f/revisions/current/patch?zip

If you need to provide credentials, use curl's "-u" option.

The communication is done with JSON, and there are some JSON cmdline
parsers available too.

Re: [daisy][PATCH 1/3] layer.conf: Add in useradd dependencies as ABISAFE dependencies

[Khem Raj]

On Fri, Oct 3, 2014 at 1:24 AM, Koen Kooi <koen@dominion.thruhere.net> wrote:
> That 'new screen' is indeed an improvement and I also found the switch to make it show a unidiff instead of side-by-side. I still despise the UI, though, but not as much as before :)

all tools suck, some suck less some suck more.

Re: [daisy][PATCH 1/3] layer.conf: Add in useradd dependencies as ABISAFE dependencies

[Peter A. Bigot]

On 09/30/2014 07:55 PM, Denys Dmytriyenko wrote:
> On Thu, Sep 25, 2014 at 02:17:55PM +0200, Steffen Sledz wrote:
>> Currently allarch recipes using useradd rebuild each time MACHINE
>> changes which is not desireable. Adding the useradd dependencies
>> to this list ensures they do not change under these circumstances.
>>
>> Signed-off-by: Richard Purdie <richard.purdie@linuxfoundation.org>
>> Signed-off-by: Steffen Sledz <sledz@dresearch-fe.de>
>> ---
>>   meta/conf/layer.conf | 4 ++++
>>   1 file changed, 4 insertions(+)
>>
>> diff --git a/meta/conf/layer.conf b/meta/conf/layer.conf
>> index 18609dd..a764ee0 100644
>> --- a/meta/conf/layer.conf
>> +++ b/meta/conf/layer.conf
>> @@ -30,4 +30,8 @@ SIGGEN_EXCLUDERECIPES_ABISAFE += " \
>>     udev-extraconf \
>>     packagegroup-x11-xserver \
>>     systemd-serialgetty \
>> +  initscripts \
> This line above ^^^^ has a very long trail of whitespaces and it got commited
> that way, unfortunately :(

If OE developers and patch submitters do:

   cd .git/hooks
   mv pre-commit.sample pre-commit

in each layer, such errors will be detected at commit time. Wouldn't fix 
issues with whitespace in pulled material or git am but it could reduce 
introducing the problem.

Peter