On Fri, Sep 28, 2012 at 6:22 AM, Richard Purdie
<richard.pur...@linuxfoundation.org> wrote:
> On Fri, 2012-09-28 at 05:04 +0000, McClintock Matthew-B29882 wrote:
>> On Thu, Sep 27, 2012 at 11:38 PM, Khem Raj <raj.k...@gmail.com> wrote:
>> >
>> > On Sep 27, 2012, at 9:23 PM, McClintock Matthew-B29882 
>> > <b29...@freescale.com> wrote:
>> >
>> >> On Thu, Sep 27, 2012 at 11:18 PM, Khem Raj <raj.k...@gmail.com> wrote:
>> >>>
>> >>> On Sep 27, 2012, at 8:52 PM, McClintock Matthew-B29882 
>> >>> <b29...@freescale.com> wrote:
>> >>>
>> >>>> On Thu, Sep 27, 2012 at 10:43 PM, Khem Raj <raj.k...@gmail.com> wrote:
>> >>>>>
>> >>>>> On Sep 27, 2012, at 8:40 PM, McClintock Matthew-B29882 
>> >>>>> <b29...@freescale.com> wrote:
>> >>>>>
>> >>>>>> Khem,
>> >>>>>>
>> >>>>>> I'm just not looking at a similiar issue. Do this:
>> >>>>>>
>> >>>>>
>> >>>>>
>> >>>>> just 'not' or just 'now' ?
>> >>>>>
>> >>>>>> cat tmp/sstate-control/manifest-* | grep usr/lib/perl5
>> >>>>>>
>> >>>>>> And see which recipe is making this folder. It should be the perl
>> >>>>>> recipe, first not some other recipe…
>> >>>>>
>> >>>>>
>> >>>>> cat sstate-control/manifest-* | grep usr/lib/perl5
>> >>>>>
>> >>>>> return nothing
>> >>>>
>> >>>> Maybe this is not the same root cause then.
>> >>>
>> >>> reverting that commit fixed it. before that I tried to cleans state 
>> >>> clean all perl and perl-native
>> >>> to no avail.
>> >>
>> >> OK, well we got the same error and ours was from another sstate-cache
>> >> installing files in /usr/lib/perl5 before perl installed it's own
>> >> symlink. cleaning cache for perl did not help either.
>> >>
>> >> Reverting that patch could have causes rebuilds and change of ordering
>> >> (possibly) that might mask the error a little longer. If ours are
>> >> related.
>> >
>> >
>> > Does it mean that if I invalidate whole state it will work ?
>> > thats kind of gross
>>
>> I think just making sure perl is deployed to sysroot first fixes this
>> issue. For us we were simply missing a 'DEPENDS = perl' for a recipe
>> which installed stuff in /usr/lib/perl5 - if perl is deployed to
>> sysroot first the symlink is always in place properly.
>
> Please try the latest perl patch I've posted which should hopefully
> address this problem once and for all. The sstate business is a red
> herring as far as I can tell.

Yes my problem seems to be gone with

>
> Cheers,
>
> Richard
>

_______________________________________________
Openembedded-core mailing list
Openembedded-core@lists.openembedded.org
http://lists.linuxtogo.org/cgi-bin/mailman/listinfo/openembedded-core

Reply via email to