Marcin Juszkiewicz

I do not remember exactly when I started working on ARMv8 stuff. Checked old emails from Linaro times and found that we discussed AArch64 bootstrap using OpenEmbedded during Linaro Connect Asia (June 2012). But it had to wait a bit…

First we took OpenEmbedded and created all tasks/images we needed but built them for 32-bit ARM. But during September we had all toolchain parts available: binutils was public, gcc was public, glibc was on a way to be released. I remember that moment when built first “helloworld” — probably as one of first people outside ARM and their hardware partners.

At first week of October we had ARMv8 sprint in Cambridge, UK (in Linaro and ARM offices). When I arrived and took a seat I got information that glibc just went public. Fetched, rebased my OpenEmbedded tree to drop traces of “private” patches and started build. Once finished all went public at git.linaro.org repository.

But we still lacked hardware… The only thing available was Versatile Express emulator which required license from ARM Ltd. But then free (but proprietary) emulator was released so everyone was able to boot our images. OMG it was so slow…

Then fun porting work started. Patched this, that, sent patches to OpenEmbedded and to upstream projects and time was going. And going.

In January 2013 I started X11 on emulated AArch64. Had to wait few months before other distributions went to that point.

February 2013 was nice moment as Debian/Ubuntu team presented their AArch64 port. It was their first architecture bootstrapped without using external toolchains. Work was done in Ubuntu due to different approach to development than Debian has. All work was merged back so some time later Debian also had AArch64 port.

It was March or April when OpenSUSE did mass build of whole distribution for AArch64. They had biggest amount of packages built for quite long time. But I did not tracked their progress too much.

And then 31st May came. A day when I left Linaro. But I was already after call with Red Hat so future looked quite bright ;D

June was month when first silicon was publicly presented. I do not know what Jon Masters was showing but it probably was some prototype from Applied Micro.

On 1st August I got officially hired by Red Hat and started month later. My wife was joking that next step would be Retired Software Engineer ;D

So I moved from OpenEmbedded to Fedora with my AArch64 work. Lot of work here was already done as Fedora developers were planning 64-bit ARM port few years before — when it was at design phase. So when Fedora 15 was bootstrapped for “armhf” it was done as preparation for AArch64. 64-bit ARM port was started in October 2012 with Fedora 17 packages (and switched to Fedora 19 during work).

My first task at Red Hat was getting Qt4 working properly. That beast took few days in foundation model… Good that we got first hardware then so it went faster. 1-2 months later and I had remote APM Mustang available for my porting work.

In January 2014 QEmu got AArch64 system emulation. People started migrating from foundation model.

Next months were full of hardware announcements. AMD, Cavium, Freescale, Marvell, Mediatek, NVidia, Qualcomm and others.

In meantime I decided to make crazy experiment with OpenEmbedded. I was first to use it to build for AArch64 so why not be first to build OE on 64-bit ARM?

And then June came. With APM Mustang for use at home. Finally X11 forwarding started to be useful. One of first things to do was running firefox on AArch64 just to make fun of running software which porting/upstreaming took me biggest amount of time.

Did not took me long to get idea of transforming APM Mustang (which I named “pinkiepie” as all machines at my home are named after cartoon characters) into ARMv8 desktop. Still waiting for PCI Express riser and USB host support.

Now we have October. Soon will be 2 years since people got foundation model available. And there are rumors about AArch64 development boards in production with prices below 100 USD. Will do what needed to get one of them on my desk ;)

My daughter is in 1st class of elementary school. One of things she has there is English. For which they use a “Super Sparks Student 1” book from Oxford University Press. Book came with CDROM disk (marked as DVD Video). There are 73 audio files (mp3) and one short movie on it.

But why I write about it? Because I think that we live in an era when CSS means Cascading Style Shits rather than Content Scramble System. But not everyone thinks like that.

So back to DVD^wCDROM. Let’s copy data from it. Audio tracks went fine but problems started when tried to copy video:

[ 3701.096102] sr 5:0:0:0: [sr1]
[ 3701.096105] Sense Key : Illegal Request [current]
[ 3701.096109] sr 5:0:0:0: [sr1]
[ 3701.096114] Add. Sense: Read of scrambled sector without authentication
[ 3701.096117] sr 5:0:0:0: [sr1] CDB:
[ 3701.096119] Read(10): 28 00 00 01 8e e3 00 00 01 00
[ 3701.116089] sr 5:0:0:0: [sr1]
[ 3701.116096] Result: hostbyte=DID_OK driverbyte=DRIVER_SENSE

Took me a while to remind myself what it means. Scrambled disk! When it became useless? 2001? 1996? No — 1999 (according to Wikipedia). So just 15 years ago…

I think that Oxford is quite orthodox/conservative but IT goes on faster than they expect.

Took me few minutes to compile libdvdcss, libdvdread and dvdread then it was just simple “sudo dvdread /dev/sr1 >english.iso” plus extracting files from disk image. Would be faster but you know: I use Fedora (patents, mp3, decss are not accepted in repo).

As my wife’s laptop got SSD upgrade I had another spare HDD. But what to do with 320GB hdd? Decided to add it into “pinkiepie” (APM Mustang) and test Fedora installation.

There are probably many ways on how you can install Fedora on x86-64 machine. But on Mustang there are only three:

• DVD boot
• PXE boot
• use any method to load kernel and initrd of net installer

My DVD-RW drive decided to not read any discs so I bought new one. Put fresh Fedora 21 AArch64 installation DVD into it and… still something wrong. Not with contents of disk but with drive or who knows.

Decided to skip PXE as it always involves creation of TFTP server, setting DHCP server to direct into proper one etc. Too much work just to deliver two files.

So I went with last method and used GRUB which I already had on /dev/sda in EFI partition.

Fetched vmlinuz and initrd.img to /boot/pxeboot/ directory and rebooted machine.

Then into GRUB and it’s command line:

linux (hd0,gpt2)/pxeboot/linux console=ttyS0,115200 inst.repo=http://armpkgs.fedoraproject.org/compose/21_Alpha_TC1/Server/aarch64/os/
initrd (hd0,gpt2)/pxeboot/initrd.img

Then just Ctrl-x and it goes. As it has to fetch 258MB squashfs.img with installer I used that time to install VNC viewer (KRDC) on my desktop — it will be used to control installer as text mode over serial one is very limited.

Image fetched, booted. Then something like “Do you want to use VNC (recommended) or go with text mode (limited)?” appeared so I went for remote one, used no password and then installer appeared and I was able to select hard drive and what exactly I wanted to install. Too bad that it had to end due to some dependency issues:

So I went for a fix. Took me some time but next version of “abrt” should be fine.

Anyway why stop? I decided to alter a way of installation a bit. Let’s use Polish Fedora mirror and local copy of “squashfs.img” file:

linux (hd0,gpt2)/pxeboot/linux console=ttyS0,115200 inst.repo=http://ftp.icm.edu.pl/pub/Linux/dist/fedora-secondary/development/21/aarch64/os/ inst.stage2=http://192.168.0.13
initrd (hd0,gpt2)/pxeboot/initrd.img

This time I got wide choice of installation targets (it was only Fedora Server before). Wanted to install XFCE but chosen “Minimal system” instead as installer was unable to tell me what is wrong with my selection.

Few minutes later (after setting root password and creating normal user) I had installed system. Rebooted, went to UEFI shell, selected FS1: and booted GRUB which loaded fresh Fedora 21 for me.

Final words? It went much, much better then my previous installation of F21 on same hardware (I used chroot or other crazy stuff). Not checked iSCSI or other exotic (to me) ways of doing storage but that’s maybe for other day.

When GNU project was announced over 30 years ago it was something great. But time passed and I have a feeling that it is more and more politics instead of coding.

I do builds. For over 10 years now. It was ARM all the time with some bits of AVR32, MIPS, x86. During last two years it is nearly 100% AArch64. And during last months my dislike to GNU project grows.

Why? several reasons.

Commit messages

There is a lot of articles about “how to write good commit messages”. I can tell you where to look for bad ones: gcc, binutils, glibc — base of most of GNU/Linux distributions. All of them can be fetched from GIT repositories but look like deep in CVS era. Want to find what was changed? You will get it in commit message. Why it was changed? Forget.

GNU forks

Do you know what IceCat is? Or GNUzilla? Let me quote official homepage:

GNUzilla is the GNU version of the Mozilla suite, and GNU IceCat is the GNU version of the Firefox browser. Its main advantage is an ethical one: it is entirely free software. While the Firefox source code from the Mozilla project is free software, they distribute and recommend non-free software as plug-ins and addons.

Where is source? Somewhere in GNU cvs probably. I failed to find it. OK, there is a link to something which is probably source tarball but we have XXI century — developers take source control systems got granted.

Of course IceCat fails to build on AArch64. Why? Because it is based on already obsoleted by upstream version 24 of Firefox. Support for 64-bit ARM platform was merged around Firefox 30 and is complete in version 31. Sure, I could dig for patches for IceCat version but no. This time I refuse.

I do not know, but maybe GNU project needs some fresh blood which will make them more developer friendly?

In the morning I got an email:

Dear Marcin Juszkiewicz,

Congratulations on your one-year anniversary with Red Hat! Thank you for your commitment and work over the past year. We hope that it has been everything you expected it to be and look forward to celebrating your future success with the company.

Yes, already year passed since I joined ARM team at Red Hat. It was a good time and I do not plan to change it ;)

What I did during that time? Managed to get several packages built for AArch64, sent many patches upstream (some were easy, other required several updates) and even got one machine to use at home. It was not an easy ride but I am glad that I went that way.

I had some ARMv7a work done but over 80% of time spent with AArch64. First in simulators but then hardware started coming. First shared one with other developers (timezone differences helped a lot), then got remote one for own development use and finally one machine landed under my desk (the only one in Poland at that time). Do I have to add how it simplified work? GVim over X11 just works so the only difference is colorscheme and font used ;D

What next? More AArch64 work. There are still packages which fail to build ;D

I have a bunch of ARMv7a boards at home. They are from different years, have misc CPUs and GPUs. All I think that some of them suck for some reasons.

Pandaboard was great board when got released. Then Texas Instruments fired everyone involved so now it is crap. Sure, mainline kernel works fine but no audio/video decoding in hardware, no OpenGLES due to PowerVR stuff which no one cares about because it is proprietary.

Wandboard Quad. Cool, fast, 2GB of memory, SATA. And hardcoded XGA (1024x768) resolution which can not be changed. Awesome? Not so much when you connect it to FullHD monitor. And forget about community — they are mostly stuck on 3.0 and 3.10 kernels based on Freescale code drops. I should dig deeper when looked at i.mx6 hardware ;(

Looks like it is time to check other boards. Minnowboard Max probably — x86 will fully open drivers.

Lot was written about why upstreaming matters. And why it should be done earlier than later too. And I found out that I have perfect example.

Few days ago Linus Walleij contacted me with a question:

Hi Marcin, do you still have the Nomadik NHK8815 board? I’m contemplating fixing it to work some time, but haven’t got hold of one. I think it only needs a device tree actually, so we could hack it up at some conference if you’re at the same some time.

And then I realized that indeed — I still have it somewhere in my basement gathering dust. We quickly agreed on shipping it so Linus will be able to finish work on adding support for already obsolete board in mainline kernel.

For me NHK-15 is a perfect example of device which was supported in wrong way. When I got it 5 years ago it was already several kernel releases behind with 2.6.16.2 kernel (no stable updates even). I managed to get Poky Linux running on it but told clearly that it will gather dust after finishing task because for me board with no upstream support is worthless.

So board spent few weeks on my desk and then went back to the box and on shelf in basement. Then became forgotten. For 5 years. Today I opened box, checked contents, added some filling so it is ready for courier. Then I will forget about it again.

UPDATE: My Sim.One board joined NHK-15 as it is hard to find working device with EP93xx processor in it.

I am starting to dislike current situation of handling sound under Linux.

On my desktop I am using Fedora which is using PulseAudio to handle sound. And recently something changed cause I have to reach mute button too often.

Normally I have Deezer in Chrome playing some background music with volume in web app set to the highest value. I can use KDE mixer to control how loud it is and it works fine. When I run YouTube, Vimeo it is fine. Adobe Flash videos? Fine as well. And according to PulseAudio mixers it keeps stream value at same level every time.

But I do not like that browser. I use Firefox for most of my web stuff. But every time I play YouTube video it jumps with PA stream volume to the highest possible setting so even my next street neighbors are probably able to hear it. Sure, I can change YT volume to nearly zero but when will have to do it with all other web applications as well (and many of them do not have loudness control).

MPlayer (which I use for movies) plays nicely with volume - alters only own PA stream volume.

I think that it is a time for me to report bug against Firefox.