Re: [PATCH] gcc-source: exlude from world

[Randy MacLeod <randy.macleod@windriver.com>]

On 2015-08-27 11:27 PM, Khem Raj wrote:
> On Thu, Aug 27, 2015 at 8:17 PM, Randy MacLeod
> <randy.macleod@windriver.com> wrote:
>>
>> I've sent the parent email to fix
>> $ bitbake -c <stage> world.
>> I've confirmed that my change works for:
>> $ bitbake -c configure world
>>
>> I wasn't sure if I should leave gcc-source as
>> is or change it to be similar to kernel-devsrc
>> because meta/recipes-devtools/gcc/gcc-source.inc does:
>>
>> deltask do_configure
>> deltask do_compile
>> deltask do_package
>
> if a task is deleted and you do -c <that-task> then you should get an
> error. deltask would create no stamps as opposed to noop tasks. May be
> it can be turned into warning ( not convinced for that too)
> but excluding it world doesn't sound a good solution here.

I'd be happy to change the deltask lines to do_foo[noexec] = "1"
if people agree on that. Would that work for you Khem?

The kernel-devsrc and gcc-source packages are really
a different class of recipes from -native or target and if
we had more than two of them, we should create a class to
encapsulate their behaviour and ensure that they don't
break bitbake -c <stage> world builds. I haven't tried that
yet but I expect the class would just have the do_foo[noexec] = "1"
so you don't have to type it in each of the two recipes of this
type that we currently have! :)

I'd like to do the:
    do_foo[noexec] = "1"
now and figure out how to handle foo-source in 2.1. Agreed?

I'll include the [YOCTO #7878] in a v2 once I get some
guidance on what is acceptable.

../Randy


>
>> ...
>>
>> whereas meta/recipes-kernel/linux/kernel-devsrc.bb does:
>> # There's nothing to do here, except install the source where we can package
>> it
>> do_fetch[noexec] = "1"
>> do_unpack[noexec] = "1"
>> do_patch[noexec] = "1"
>> ...
>>
>> The kernel-devsrc approach doesn't break world builds
>> but it's probably slightly less efficient.
>>
>> Also do people care about consistency in naming:
>>    gcc-source.inc vs
>>    kernel-devsrc.bb
>> ? There are only two such 'source' packages so it's not a big deal.
>>
>> ../Randy
>>
>>
>>
>> On 2015-08-27 11:09 PM, Randy MacLeod wrote:
>>>
>>> Exclude all versions of gcc-source from world builds so that:
>>>      bitbake -c <stage> world
>>> will work. gcc-source deletes most bitbake build stages
>>> since it is a source-only package.
>>>
>>> Signed-off-by: Randy MacLeod <Randy.MacLeod@windriver.com>
>>> ---
>>>    meta/recipes-devtools/gcc/gcc-source_4.8.bb | 2 ++
>>>    meta/recipes-devtools/gcc/gcc-source_4.9.bb | 2 ++
>>>    meta/recipes-devtools/gcc/gcc-source_5.2.bb | 2 ++
>>>    3 files changed, 6 insertions(+)
>>>
>>> diff --git a/meta/recipes-devtools/gcc/gcc-source_4.8.bb
>>> b/meta/recipes-devtools/gcc/gcc-source_4.8.bb
>>> index 234b82e..b890fa3 100644
>>> --- a/meta/recipes-devtools/gcc/gcc-source_4.8.bb
>>> +++ b/meta/recipes-devtools/gcc/gcc-source_4.8.bb
>>> @@ -1,2 +1,4 @@
>>>    require recipes-devtools/gcc/gcc-${PV}.inc
>>>    require recipes-devtools/gcc/gcc-source.inc
>>> +
>>> +EXCLUDE_FROM_WORLD = "1"
>>> diff --git a/meta/recipes-devtools/gcc/gcc-source_4.9.bb
>>> b/meta/recipes-devtools/gcc/gcc-source_4.9.bb
>>> index 234b82e..b890fa3 100644
>>> --- a/meta/recipes-devtools/gcc/gcc-source_4.9.bb
>>> +++ b/meta/recipes-devtools/gcc/gcc-source_4.9.bb
>>> @@ -1,2 +1,4 @@
>>>    require recipes-devtools/gcc/gcc-${PV}.inc
>>>    require recipes-devtools/gcc/gcc-source.inc
>>> +
>>> +EXCLUDE_FROM_WORLD = "1"
>>> diff --git a/meta/recipes-devtools/gcc/gcc-source_5.2.bb
>>> b/meta/recipes-devtools/gcc/gcc-source_5.2.bb
>>> index 234b82e..b890fa3 100644
>>> --- a/meta/recipes-devtools/gcc/gcc-source_5.2.bb
>>> +++ b/meta/recipes-devtools/gcc/gcc-source_5.2.bb
>>> @@ -1,2 +1,4 @@
>>>    require recipes-devtools/gcc/gcc-${PV}.inc
>>>    require recipes-devtools/gcc/gcc-source.inc
>>> +
>>> +EXCLUDE_FROM_WORLD = "1"
>>>
>>
>>
>> --
>> # Randy MacLeod. SMTS, Linux, Wind River
>> Direct: 613.963.1350 | 350 Terry Fox Drive, Suite 200, Ottawa, ON, Canada,
>> K2K 2W5
>>
>> --
>> _______________________________________________
>> Openembedded-core mailing list
>> Openembedded-core@lists.openembedded.org
>> http://lists.openembedded.org/mailman/listinfo/openembedded-core


-- 
# Randy MacLeod. SMTS, Linux, Wind River
Direct: 613.963.1350 | 350 Terry Fox Drive, Suite 200, Ottawa, ON, 
Canada, K2K 2W5