Re: [yocto] using Hob to run images on a custom simulator, other than qemu
On 24/12/2015 18:19, "Mirela Rabulea" wrote: >Hi Belén, >Thanks for the info on the tests, I'll check on that. >In the meantime, please find attached my proposal for the new GUI >options. Basically, if an image is qemu-compatible, already there was a >"Run image" button displayed. I let the button be named "Run qemu image", >while if a non qemu-compatible image is selected, the same button shows >up with the name "Run custom image" Hi Mirela, This looks ok to me. > and besides that, an option shows-up to allow the selection of the >custom script I guess the only thing I am missing is how do I select the custom script. Will you present the option in a dialog? >(by default it points out to runqemu script) to be used for launching >this custom image. >I also attached my original versus changed python scripts, for more >clarity. Can't comment on those, I'm afraid. I only take care of the pretty pictures ;) Cheers Belén >Looking forward for your feedback. > >And...Merry Christmas :) > >Regards, >Mirela > > >-Original Message- >From: Barros Pena, Belen [mailto:belen.barros.p...@intel.com] >Sent: Wednesday, December 23, 2015 12:21 PM >To: Mirela Rabulea ; Mirela Rabulea >; yocto@yoctoproject.org; >p...@yoctoproject.org >Cc: Nicolae Manescu ; Georgescu, Alexandru C > >Subject: Re: [yocto] using Hob to run images on a custom simulator, other >than qemu > > > >On 22/12/2015 11:35, "Mirela Rabulea" wrote: > >>Hi Belén, >>Thanks for the answer. Is there a set of tests that would ensure there >>are no regressions? Where can it be found? > >These are the test cases I can find in Testopia: > >https://bugzilla.yoctoproject.org/tr_list_cases.cgi?searchstr=hob¤t_ >t >ab=case&summary_type=anywordssubstr&summary=hob&field-1-0-0=summary&type-1 >- >0-0=anywordssubstr&value-1-0-0=hob > >Hopefully someone from QA can provide more details on how to run them. > >If you want any input on how to present the new options in the UI, let me >know. I'll be happy to help. > >Cheers > >Belén > >> >>Thanks, >>Mirela >> >>-Original Message- >>From: Barros Pena, Belen [mailto:belen.barros.p...@intel.com] >>Sent: Monday, December 21, 2015 1:26 PM >>To: Mirela Rabulea ; >>yocto@yoctoproject.org; p...@yoctoproject.org >>Cc: nicolae.mane...@nxp.com >>Subject: Re: [yocto] using Hob to run images on a custom simulator, >>other than qemu >> >> >> >>On 18/12/2015 16:05, "yocto-boun...@yoctoproject.org on behalf of >>Mirela Rabulea" >mirela.rabu...@freescale.com> wrote: >> >>> >>> >>>Hello yocto world >>>J >>> >>> >>> >>>This is my first message on any yocto list, I¹m not sure which list to >>>address, so I choose the ones which seemed the most generic and suited >>>to my purpose. Please redirect me if a more specific list is better >>>suited to ask this question. >>> >>> >>> >>>The problem I see at present is: there is one single button in Hob, >>>³Run Image², which is active only for qemu images, and it is >>>triggering the execution of the ³poky\scripts\runqemu² script. Please >>>scroll down to the picture at end of this email to see what I¹m talking >>>about. >>> >>> >>> >>>The question I want to ask is if it would be acceptable to add more >>>buttons to the ³Image details² dialog (or a new dialog to pop-up when >>>³Run Image² is clicked), in order to allow the customization of the >>>simulator to be launched, basically to specify the location & command >>>line of such custom simulator. Who may advise in this matter? I am >>>willing to contribute with changes. >> >>Hi Mirela, >> >>I have asked a few people about this, because Hob is pretty much >>deprecated at this point (it's being replaced by a web interface called >>Toaster). They replied they would be willing to merge changes to Hob, >>as long as they can be sure they won't cause any regressions. >> >>I hope this helps. >> >>Cheers >> >>Belén >> >>> >>> >>> >>>The changes I anticipate would be necessary are: >>> >>>1. >>>Changes in the GUI, somewhere in the python scripts from >>>³poky\bitbake\lib\bb\ui\crumbs\², especially in imagedetailspage.py, >>>maybe in builder.py >>> >>>2. >>>Be more permissive with the types of images that are allowed to run. >>>Currently, there are these restrictions related to the name and >>>extension of the image in ³poky\meta\classes\ image_types.bbclass², >>>these need to be extended/customized: >>> >>>RUNNABLE_IMAGE_TYPES ?= "ext2 ext3" >>> >>>RUNNABLE_MACHINE_PATTERNS ?= "qemu" >>> >>> >>> >>> >>>BTW, here is how Hob ³Image details² dialog currently looks like, see >>>attached picture, note that the highlighted ³Run Image² button shows >>>up only for qemu images. >>> >>> >>>Regards, >>> >>>Mirela >>> >>> >>> >>> >> > -- ___ yocto mailing list yocto@yoctoproject.org https://lists.yoctoproject.org/listinfo/yocto
Re: [yocto] BeagleBone...bitbake -v core-image-sato stops
On 28.12.2015 23:00, Khem Raj wrote:
On Dec 28, 2015, at 8:56 AM, kaffeesurrogat wrote:
Hello again,
well I'm really a noob and guessing what to do. Well I've interrupted bitbake
cause
nothing was happing and the log.do_compile of attr-native didn't show any error
message. I've started it all over again an now I do get an error message: yes
;-)
it says:
gmake[2]: ../-libtool: Command not found
This looks a little bit weird to me, especially the "minus" in front of the
"libtool" .
Where does it come from? A missing or empty variable? Something like
${MYNAMEIS}-libtool an ${MYNAMEIS} not known?
what is your build host ?.
I would suggest to setup a clean build system ( may be a VM ) and try it out on
that
Thanks Khem Raj,
my build host is a gentoo system. I know it is not officially supported,
but I'm using it for ages. Thus I know my way around. At least a little
bit. Would be nice not to switch.
I've removed my yocto-dir several times and made a fresh git-clone. I've
tried with the daisy and the jethro branch. I've tried to build
core-image-minimal and core-image-sato for qemu and the beaglebone
black. I did that on two different gentoo-machines, on one of them
things compiled fine. In the beginning I've had the impression, the
faults on the failing-gentoo system could be due to the missing multilib
support. Thus I replaced it with a fresh and clean install supporting
gcc-multilib stuff.
The error stays - with small variation - more or less the same:
attr-native does not finish its build-process.
1) It stops with the message similiar to "o pl.mo pl.poto" after gmake
left the directory getfattr or setfattr
or
2) It fails with "gmake[2]: o: Command not found"
or
3) It fails with "gmake[2]: ../-libtool: Command not found "
This is really weird. I'm not able to figure the small but important
differences on my two different gentoo machines.
Any help is really appreciated,
Kaffesurrogat
Sorry guys .
I thought I was done. Actually I'm not a bit smarter then before ;-)
I'm still trying to build a yocto-image for a beaglebone. I believed, that the
reason why yocto stopped compiling at attr-native-2.4.47-r0 was the missing
gcc-multilib support. I've tried the same build on a gentoo machine at work,
which had a multilib environment, and it finished.
I've setup a new gentoo with multilib environment on my laptop but ..
bitbake stops again at attr-native. So this was not the trick. The
log.do_compile in
/home/nico/yocto_bbb_daisy/poky/build_bbb/tmp/work/x86_64-linux/attr-native/2.4.47-r0/temp
says nothing special, at least for me. Here is the quote:
-
x86_64-linux-libtool: link: gcc -o .libs/setfattr -Wl,-rpath-link
-Wl,/home/nico/yocto_bbb_daisy/poky/build_bbb/tmp/sysroots/x86_64-linux/usr/lib
-Wl,-rpath-link
-Wl,/home/nico/yocto_bbb_daisy/poky/build_bbb/tmp/sysroots/x86_64-linux/lib
-Wl,-rpath
-Wl,/home/nico/yocto_bbb_daisy/poky/build_bbb/tmp/sysroots/x86_64-linux/usr/lib
-Wl,-rpath
-Wl,/home/nico/yocto_bbb_daisy/poky/build_bbb/tmp/sysroots/x86_64-linux/lib
-Wl,-O1 -Wl,-rpath-link
-Wl,/home/nico/yocto_bbb_daisy/poky/build_bbb/tmp/sysroots/x86_64-linux/usr/lib
-Wl,-rpath-link
-Wl,/home/nico/yocto_bbb_daisy/poky/build_bbb/tmp/sysroots/x86_64-linux/lib
-Wl,-rpath
-Wl,/home/nico/yocto_bbb_daisy/poky/build_bbb/tmp/sysroots/x86_64-linux/usr/lib
-Wl,-rpath
-Wl,/home/nico/yocto_bbb_daisy/poky/build_bbb/tmp/sysroots/x86_64-linux/lib
-Wl,-O1 -Wl,-rpath-link
-Wl,/home/nico/yocto_bbb_daisy/poky/build_bbb/tmp/sysroots/x86_64-linux/usr/lib
-Wl,-rpath-link
-Wl,/home/nico/yocto_bbb_daisy/poky/build_bbb/tmp/sysroots/x86_64-linux/lib
-Wl,-rpath
-Wl,/home/nico/yocto_bbb_daisy/poky/build_bbb/tmp/sysroots/x86_64-linux/usr/lib
-Wl,-rpath -Wl,/home/nico/yocto_bbb_daisy/poky/build_bbb/tmp/sysrootgmake[2]:
Leaving directory
'/home/nico/yocto_bbb_daisy/poky/build_bbb/tmp/work/x86_64-linux/attr-native/2.4.47-r0/attr-2.4.47/setfattr'
o fr.mo fr.po
---
ADD-ON:
I've had a look in the log.do-compile-file on my machine at work. It
says .
---
o fr.mo fr.po
gmake[3]: Nothing to be done for 'default'.
gmake[3]: Leaving directory
'/home/nico/yocto_bbb_daisy/poky/build_bbb/tmp/work/x86_64-linux/attr-native/2.4.47-r0/attr-2.4.47/man/man1'
/usr/bin/gmake -C man2
gmake[2]: o: Command not found
../include/buildrules:66: recipe for target 'fr.mo' failed
gmake[2]: [fr.mo] Error 127 (ignored)
---
Well .
It will stay at this point forever.
The last output messages of bitbake:
--
Re: [yocto] plan about YP 2.0.1
Hi Yang > We are targeting the build and ship to QA of YP 2.0.1 rc1 for 1/11. > > The merge window should be around Jan 6th. I got it. Thank you very much. Lei > -Original Message- > From: Robert Yang [mailto:liezhi.y...@windriver.com] > Sent: Tuesday, December 29, 2015 3:16 PM > To: Lei, Maohui > Cc: yocto@yoctoproject.org > Subject: Re: plan about YP 2.0.1 > > > On 12/21/2015 02:00 PM, Lei, Maohui wrote: > > Hello Yang > > > >Can you tell me the schedule about the YP 2.0.1? > >I want to know when YP 2.0.1 will be released and when the merge window > will be closed? > > We are targeting the build and ship to QA of YP 2.0.1 rc1 for 1/11. > > The merge window should be around Jan 6th. > > // Robert > > > > > > > Best regards > > Lei > > > > > > > > > -- ___ yocto mailing list yocto@yoctoproject.org https://lists.yoctoproject.org/listinfo/yocto
