site  contact  subhomenews

meta-qt5 layer now in oe-qky-src

July 15, 2018 — BarryK

A little over a week ago, I added the meta-qt5 layer to oe-qky-src, my fork of OpenEmbedded. However, it broke the build of some packages.

So, I went on a journey, exploring how to compile Qt5 without the meta-qt5 layer, from basic principles. A very intense several days, ended with frustration. Well, lots of frustration during those several days!

I did get it to build, using autotools only, but only for x86_64 host and x86_64 target. Kept getting errors when attempted cross-compile for aarch64 target.

Yesterday, gave up, and went back to the meta-qt5 layer, this time hunted down why it's introduction caused some other packages to fail.

Got it sorted, qt5 5.10.1 now builds, but have not yet ported any qt5-based apps into oe-qky-src. Did attempt Scribus, but got a compile error. For now, will build Scribus in the final running system.

This morning, tried to use Smartgit to commit the latest changes. Did something wrong... it seems, when committed a couple of files and a directory together, Smartgit got confused. Smartgit seems to be thinking that the 'meta-qt5' folder is a file!

Git is very powerful, but when something goes wrong, you can really get your knickers in a twist.

To get it uploaded, here it is as a tarball:

Expand somewhere, it will expand to folder 'oe-qky-src', and there is a readme inside.

Right, now to fix Smartgit...

OK, fixed. The problem was, folder meta-qt5 had a .git folder in it, which confused Smartgit. See commits for July 15, 2018:

EDIT 20180717:
I built EasyOS in woofQ from binary packages imported from oe-qky-src, then attempted to compile Scribus. Lots of configure errors. I found that .cmake files in /usr/lib/cmake are somewhat broken. They are setup to work inside OE. I have fixed them, the fixes are in the template files in woofQ, woof-code/packages-templates/qtbase, qttools, etc.

Interesting observation: cmake seems a lot more troublesome in a cross-compiling environment than autotools-based projects. This is an interesting observation due to the praise for cmake that I read everywhere, and some projects, such as Scribus going over to it.

Tags: oe

Busybox 1.28.4 compiled statically

June 29, 2018 — BarryK

I have created a new recipe "busybox-static" in my fork of OpenEmbedded (June 28, 2018):

Busybox is compiled statically for amd64 and aarch64, and I have created PETs.

The amd64 PET will be used in future builds of EasyOS and Quirky, and there is a aarch64 build planned sometime, after I get my hands on the Librem 5 phone dev kit.

Which I am hanging out for...

The Purism website has the occasional blog post to keep us updated. This latest is interesting:

...but, where does the trust start and end? We have always trusted the major hardware manufacturers such as Intel -- after all, their chips are just big "binary blobs" -- the hardware inside a chip could be doing anything, we have no way of knowing.

What is different about firmware blobs? Nothing really. But I suppose, if firmware blobs can be removed, it is reducing the attack-surface.

Tags: oe

Packages recompiled in oe-qky-src

June 23, 2018 — BarryK

Finalised, for now, the backporting of packages from the Sumo release of OpenEmbedded, in my fork 'oe-qky-src', see commits (June 23, 2018):

Did a build overnight, and have now imported the binary packages, all 724 of them, into woofQ, for future builds of EasyOS and Quirky. The packages will be available in the "oe-pyro" repository in the PPM (Package Manager), and so are uploaded.

Here they are:

Now, I am keen to get back to EasyPak, an exciting new idea for "universal packages"...

Tags: oe, easy, quirky

Backporting Sumo to Pyro-based OE-port

June 22, 2018 — BarryK

I forked OpenEmbedded in June 2017, when the Pyro version was released. My fork, named oe-qky-src, is on GitHub, and is very different from the mainline OE project, so I can't easily sync with it.

The latest release of OE is Sumo, and I am working around the sync problem by gradually backporting Sumo recipes. To that end, I created a "sumo" layer, and have made some commits (June 22, 2018):

In particular, meta/recipes-graphics and meta-oe/recipes-graphics have been backported, which includes Xorg and mesa.

Tags: oe

rox-filer, shared-mime-info tweaks

June 09, 2018 — BarryK

In ROX-Filer file manager, when you click on the little "home" icon at top of window, it will open a window at the user's home directory. In the case of Puppy that is /root, however with Quirky I want to open at /file, and for Easy at /mnt/wkg/home.

I have modified the rox-filer source so that it detects if running in Quirky or Easy, or if not then opens at ~ (/root for Puppy).

Some of my shell scripts have a shebang line "#!/bin/ash" and ROX-Filer does not recognise this as an executable. The reason for this is the shared-mime-info package, which recognises such a file as plain-text only.

I have patched shared-mime-info to recognise "ash" and "dash" scripts as shell executables.

These fixes are to be found in my online "oe-qky-src" fork of OpenEmbedded ( June 8 and 9):

Note, I deliberately use an old version of shared-mime-info, 0.90, as I found that recent versions have stuffed up recognition of text files. Many text files in woofQ and Easy/Quirky/Puppy are mis-identified, which results in Geany and other text editors refusing to open them, and in the case of ROX-Filer not offering to open in a text-editor in the right-click "Open With..." menu.

Tags: oe

First OE aarch64 build, updates

April 17, 2018 — BarryK

I have done a first build for aarch64 (64-bit arm) on my fork of OpenEmbedded. project page:

Modifications for first aarch64 build:

Some packages failed, so I need to fix them. However, most of the big apps, such as Firefox and LibreOffice, compiled.

Have also started to update some packages:

jwm 1679:

bacon 3.7.2:

pup-tools 20180417:


Plowing ahead, fixing packages for aarch64, and updating some, adding some new packages.

Won't itemise here, look at the repo.

Tags: oe

No CUPS support in gtk+ in OpenEmbedded

November 29, 2017 — BarryK

Right now, I am in a state of great surprise. Testing Pyro64 0.6, printing does not work. In fact, it never has, for any Quirky builds from OpenEmbedded packages.

Despite the fact that 'cups' is available as a package in OE, gtk+ is hard-coded not to use it. Not only is cups missing from the DEPENDS variable in the gtk+ recipe, it is explicitly disabled with "--disable-cups".

There is isn't even a conditional test for existence of cups.

I kept looking at that, because I couldn't believe what I was seeing.

This situation is in the "pyro" release of OE/Yocto, same in the latest in master branch in github.

I have created gtk+_2.24.31.bbappend, with this in it:

PR = "r1"

# added the last line...
DEPENDS = "glib-2.0 pango atk jpeg libpng gdk-pixbuf-native \
cairo gdk-pixbuf \
cups libxinerama xinput pixman freetype fontconfig"

# original...
#EXTRA_OECONF = "--enable-xkb --disable-glibtest --disable-cups --disable-xinerama"

EXTRA_OECONF = "--enable-xkb --disable-glibtest --enable-cups --enable-xinerama \
--with-xinput=yes --enable-debug=minimum"

Added some more things that thought should be there. It compiles, not yet tested in a build.


Although OE/Yocto has the 'cups' package, it does not have 'cups-filters', which is required for printing. 'cups-filters' got split out of 'cups' sometime ago. OE actually has a recent 'cups', in which the 'cups-filters' is split out, yet there is no 'cups-filters', and I had to add it in my "meta-quirky" layer. Which shows just how much attention printing has been getting in OE/Yocto -- along with non-printing gtk+, I would say none at all.

i recompiled 'gtk+', no joy. Compiled 'cups', 'cups-filters' and 'poppler' manually in Pyro64, still no joy.

I examined /var/log/cups/error.log, and did a search with google. Found the same error was fixed here: rolling back 'cups-filters' to 1.11.4. I am using 1.13.5, so I tried it, compiled 1.11.4, and hey, printing to my USB printer works!

Oh, what a dark art this CUPS printing is!

Now, need to put the fix into OE.


These are the fixes in OE, I confirmed they have fixed printing (Nov 30, 2017):

Note, I am now doing a complete x86 32-bit (i686) recompile in OE.

Tags: oe

OE recompile, SM 2.49.1, kernel 4.14.1

November 23, 2017 — BarryK

I mentioned recently that some library packages compiled in OpenEmbedded were too old. I ended up only bumping 'icu', 'nss' and 'nspr', to versions 58.2, 3.31.1 and 4.16. Then did a complete rebuild in OE.

Built Pyro64 0.5.2, and compiled SeaMonkey 2.49.1, this time using the system icu, nss and nspr, making the package considerably smaller. The PET is here:

I also compiled the 4.14.1 kernel, and was most surprised when "make modules_install" did not create /lib/firmware. A quick google, and it seems the firmware is no longer in the kernel source tree.

I didn't know how to respond to this situation, but decided to take the /lib/firmware from the compile of kernel 4.13.11, for inclusion in the kernel PET. The PET is here:

The source, patches, and build scripts are here:

Tags: oe, quirky