From: Bruce Ashfield <bruce.ashfield@windriver.com>
To: Tom Zanussi <tom.zanussi@intel.com>
Cc: "yocto@yoctoproject.org" <yocto@yoctoproject.org>
Subject: Re: Current Build Failures for d82f205cfa790b40e132e11b7937050bc3b97ff3
Date: Fri, 14 Jan 2011 10:48:48 -0500 [thread overview]
Message-ID: <4D307060.4040802@windriver.com> (raw)
In-Reply-To: <1294986870.28996.10.camel@elmorro>
On 11-01-14 01:34 AM, Tom Zanussi wrote:
> On Thu, 2011-01-13 at 17:41 -0800, Bruce Ashfield wrote:
>> On 11-01-13 8:37 PM, João Henrique Freitas wrote:
>>> Hi,
>>>
>>> I have two build machines:
>>>
>>> x86 ubuntu 10.04 at work
>>> x86_64 ubuntu 10.04 at home
>>>
>>> At home all things goes ok. But at work perf has failed.
>>
>> Are your package lists identical ? Sounds like host contamination
>> to me and perf is deciding to build perl or not perl based on
>> what the host has installed.
>>
>>>
>>> The problem is LIBPERL. I can't understand why but at work the
>>> Makefile from perf does not have the flag NO_LIBPERL.
>>>
>>> To 'solve' it I put:
>>>
>>> do_compile_perf() {
>>> oe_runmake -C ${S}/tools/perf CC="${CC}" LD="${LD}" prefix=${prefix}
>>> NO_LIBPERL=1
>>
>> This works .. or to get the functionality back, we get perl
>> in the package list and kindly ask perf to look in our sysroot
>> for what it needs :)
>>
>
> One thing I notice is that recipes-devtools/perl/perl_5.12.2.bb doesn't
> install ExtUtils, which perl_5.8.8.bb did.
That would explain the "what changed under us?" if this
turns out to be the cause.
>
> The logic in perf that decides whether to define NO_LIBPERL depends on
> the output of ExtUtils::Embed - it looks like it may be picking up the
> host's ccopts instead, which might explain the /usr/local/include in the
> error message:
>
> cc1: error: include location "/usr/local/include" is unsafe for cross-compilation
Sounds like we are on the right track.
>
> Will look into it more tomorrow...
Many thanks!
Bruce
>
> Tom
>
>> Thanks for the digging on this.
>>
>> Bruce
>>
>>> }
>>>
>>> in linux-tools.inc
>>>
>>> Thanks.
>>>
>>> On Thu, Jan 13, 2011 at 2:47 PM, Tom Zanussi<tom.zanussi@intel.com> wrote:
>>>> On Thu, 2011-01-13 at 08:37 -0800, Bruce Ashfield wrote:
>>>>> On 11-01-13 11:33 AM, Elizabeth Flanagan wrote:
>>>>>> All,
>>>>>>
>>>>>> The current nightly based on d82f205cfa790b40e132e11b7937050bc3b97ff3 is
>>>>>> showing 3 build failures so far.
>>>>>>
>>>>>> Machine: qemux86
>>>>>> Failure: intermittent tasks_0.18 failure at do_compile
>>>>>> Description: Tasks_0.18 has failed during one buildset, however it's not
>>>>>> failing on the other.
>>>>>>
>>>>>>
>>>>>> Machine: arm
>>>>>> Failure: sanity tests time out and cause cascading failures.
>>>>>> Description: I'm patching the autobuilder today so that sanity tests
>>>>>> timing out shouldn't cause a cascading failure, however the arm sanity
>>>>>> tests are still timing out.
>>>>>>
>>>>>> Machine: qemux86-64
>>>>>> Failure: linux-yocto_git failing at do_compile_perf
>>>>>> Description:
>>>>>> | CC util/scripting-engines/trace-event-perl.o
>>>>>> | cc1: warnings being treated as errors
>>>>>> | cc1: error: include location "/usr/local/include" is unsafe for
>>>>>> cross-compilation
>>>>>
>>>>> I saw this one as well, but hadn't gone back to look at
>>>>> it yet. Did something change underneath us?
>>>>>
>>>>> I haven't updated any parts of the recipes or kernel that
>>>>> would have triggered the new failure. Or at least nothing
>>>>> that has ever showed up in my testing.
>>>>>
>>>>
>>>> Yeah, it's strange that this starts happening seemingly out of the blue
>>>> - nothing wrt this has changed recently AFAICS.
>>>>
>>>>> Darren/Tom: you guys have been into perf/trace recently,
>>>>> any ideas (or time) for a quick fix ? I'm going to be
>>>>> pegged for the day trying to sort out some other items.
>>>>>
>>>>
>>>> Coincidentally, I was going to look at perf scripting next week - guess
>>>> I'll start this week instead. Don't have a quick fix off the top of my
>>>> head - will start looking into it later today/tonight...
>>>>
>>>> Tom
>>>>
>>>>> Cheers,
>>>>>
>>>>> Bruce
>>>>>
>>>>>> | make: *** [util/scripting-engines/trace-event-perl.o] Error 1
>>>>>>
>>>>>> _______________________________________________
>>>>>> yocto mailing list
>>>>>> yocto@yoctoproject.org
>>>>>> https://lists.yoctoproject.org/listinfo/yocto
>>>>>
>>>>
>>>>
>>>> _______________________________________________
>>>> yocto mailing list
>>>> yocto@yoctoproject.org
>>>> https://lists.yoctoproject.org/listinfo/yocto
>>>>
>>>
>>>
>>>
>>
>
>
next prev parent reply other threads:[~2011-01-14 15:48 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-01-13 16:33 Current Build Failures for d82f205cfa790b40e132e11b7937050bc3b97ff3 Elizabeth Flanagan
2011-01-13 16:37 ` Bruce Ashfield
2011-01-13 16:47 ` Tom Zanussi
2011-01-14 1:37 ` João Henrique Freitas
2011-01-14 1:41 ` Bruce Ashfield
2011-01-14 6:34 ` Tom Zanussi
2011-01-14 15:48 ` Bruce Ashfield [this message]
2011-01-14 1:59 ` Xu, Jiajun
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=4D307060.4040802@windriver.com \
--to=bruce.ashfield@windriver.com \
--cc=tom.zanussi@intel.com \
--cc=yocto@yoctoproject.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.