XGene1: cursed processor?

Years ago Applied Micro (APM) released XGene processor. It went to APM BlackBird, APM Mustang, HPe M400 and several other systems. For some time there was no other AArch64 cpu available on market so those machines got popular as distribution builders, developer machines etc…

Then APM got aquired by someone, CPU part got bought by someone else and any support just vanished. Their developers moved to work on XGene2/XGene3 cpus (APM Merlin etc systems). And people woke up with not-supported hardware.

For some time it was not an issue – Linux boots, system works. Some companies got rid of their XGene systems by sending them to Linaro lab, some moved them to ‘internal use only, no external support’ queue etc.

Each mainline kernel release was “let us check what is broken on XGene this time” time. No serial console output again? Ok, we have that ugly patch for it (got cleaned and upstreamed). Now we have kernel 4.16 and guess what? Yes, it broke. Turned out that 4.15 was already faulty (we skipped it at Linaro).

Red Hat bugzilla has a Fedora bug for it. Turns out that firmware has wrong ACPI tables. Nothing new, right? We already know that it lacks PPTT for example (but it is quite new thing for processors topology). This time bug is present in DSDT one.

Sounds familiar? If you had x86 laptop about 10 years ago then it could. DSDT stands for Differentiated System Description Table. It is a major ACPI table used to describe what peripherals the machine has. And serial ports are described wrong there so kernel ignores them.

One of solutions is bundling fixed DSDT to kernel/initrd but that would require adding support for it into Debian and probably not get merged as no one needs that nowadays (unless they have XGene1).

So far I decided to stay on 4.14 for my development cartridges. It works and allows me to continue my Nova work. Do not plan to move to other platform as at Linaro we have probably over hundred XGene1 systems (M400 and Mustangs) which will stay there for development (hard to replace 4.3U case with 45 cartridges by something else).

Shenzhen trip

Few months ago, at the end of previous Linaro Connect gathering, there was announcement that next one will take place in Hong Kong. This gave me idea of repeating Shenzhen trip but in a bit longer version.

So I mailed people at Linaro and there were some responses. We quickly agreed on going there before Connect. Alex, Arnd, Green and me were landing around noon, Riku a few hours later so we decided that we will meet in Shenzhen.

We crossed border in Lok Ma Chau, my visa had the highest price again and then we took a taxi to the Maker Hotel (still called “Quchuang Hotel” in Google Maps and on Booking.com) next to all those shops we wanted to visit. Then went for quick walk through Seg Electronics Market. Lot of mining gear. 2000W power supplies, strange PCI Express expanders etc. Dinner, meeting with Riku and day ended.

I have woken up at 02:22 and was not able to fall asleep. Around 6:00 it turned out that rest of team is awake as well so we decided to go around and search for some breakfast. Deserted streets looked a bit weird.

Back at hotel we were discussing random things. Then someone from Singapore joined and we were talking about changes in how Shenzhen stores/factories operate. He told us that there is less and less of stores as business moves to the Internet. Then some Chinese family came with about seven years old boy. He said something, his mother translated and it turned out that he wanted to touch my beard. As it was not the first time my beard got such attention I allowed him. That surprise on his face was worth it. And then we realized that we have not seen bearded Chinese man on a street.

As stores were opening at 10:00 we still had a lot of time so went for random walk. Including Shenzhen Center Park which is really nice place:

Then stores started to open. Fake phones, real phones, tablets, components, devices, misc things… Walking there was fun itself. Bought some items from my list.

They also had a lot of old things. Intel Overdrive system for example or 386/486 era processors and FPUs.

From weird things: 3.5″ floppy disks and Intel Xeon Platinum 8175 made for Amazon cloud only.

Lot and lot of stuff everywhere. Need power supply? There were several stores with industrial ones, regulated ones etc. Used computers/laptops? Piles after piles. New components? Lot to choose from. Etc, etc, etc…

After several hours we finally decided to go back to Hong Kong and rest. The whole trip was fun. I really enjoyed it. Even without getting half of items from my ‘buy while in Shenzhen’ list ;D

And ordered Shenzhen fridge magnet on Aliexpress… They were not available to buy at any place we were.

SnowpenStack in Dublin

During last week I was in Dublin, Ireland on OpenStack PTG. It was also the worst weather since 1982. There was snow and strong wind so conference quickly got renamed to SnowpenStack.

The main reasons for me to be there were:

  • meet all those developers who took some time and looked at my changes
  • discuss some other changes/plans
  • share aarch64 knowledge with OpenStack projects

Conference took place in the Croke Park stadium. We used meeting rooms on 4th, 5th, 6th floors. One day by mistake I took wrong stairs and ended on top of the stadium in just T-shirt… Quickly ran to an elevator to get back to proper floor ;D

The schedule was split into two parts: Monday and Tuesday were for mixed teams sessions while Wednesday — Friday were for discussions in teams. I spoke with Kolla, Nova and Infra teams mostly. There were some discussions with Ironic, Kuryr and some other ones too. Also met several Polish developers so there was a time to speak in native language ;D

On Tuesday I went to the city centre to buy some souvenirs for family (and 99th fridge magnet for myself). Launched Ingress, did one mosaic to see more of a city and after 11 kilometres I was back at the hotel just in time for a small party in GAA museum. And then pub trip with Polish guys. When I finally reached the hotel (about 01:30) there were still discussions in the lobby and I took a part in one of them.

Team discussions started on Wednesday. Visited Nova one summarizing ‘Queens’ release. Turned out that it went better than previous ones. The main problem was lack of reviews — not everyone likes to pester developers on IRC to get some attention to patches. I was asked few times for opinion as I was one of few fresh contributors.

Kolla sessions were a bit chaotic in my opinion. Recently chosen PTL was not present and the person supposed to replace him got stuck at home due to weather. One of the discussions I remember was about Ceph: should we keep using our images or rather move to ‘ceph-ansible’ instead. Final idea was to keep as it looked like there were more cons than pros with moving to ‘ceph-ansible’ images.

Discussed Arm64 support with Infra team. We (Linaro) provided them resources on one of our developer clouds to get aarch64 present in OpenStack CI gates. Turned out that machines work and some initial tests were done. I also got informed that diskimage-builder patches to add GPT/UEFI support will be reviewed soon.

And then there were some weather related issues. On Wednesday every attendee got an email with information that Irish government issued the Red Alert which strongly suggest to stay inside if you do not really have to go outside. And as attendance was not mandatory then people should first check are they comfortable with going to Croke Park (especially those who not stayed in the hotel nearby). Next day organization team announced that the venue we used will close after lunch to make sure that everyone is safe. And the whole conference moved to the hotel…

Imagine developers discussing EVERYWHERE in a hotel. Lobby was occupied by few teams, Infra found a table in library corner, Nova took Neutron and occupied breakfast room. Bar area was quite popular and soon some beers were visible here and there. Few teams went to meeting rooms etc. WiFi bandwidth was gone… Some time later hotel staff created a separate wireless network for our use. And then similar situation was on Friday.

On Wednesday other thing happened too: people started receiving information that their flights are cancelled. There were some connections on Thursday and then nothing was flying on Friday. Kudos to hotel staff to be aware of it — they stopped taking external reservations to make sure that PTG attendees have a place to stay for longer (as some people got rebooked to even Thursday).

And even on Saturday it was hard to get to the airport. No taxi going to the hotel due to snow on a street. But if you walked 500 meters then cab could be hailed on a street. Many people went for buses (line 700 was the only working one). The crowd on the airport was huge. Some of those people looked like they lived there (which was probably true). Several flights were delayed (even by 4-5 hours), other got cancelled but most of them were flying.

Despite the weather sitting in a hotel in Dublin was safe, walking around too as there were about 15-20 centimetres of snow on a street. There were several snowmen around, people had fun playing with snow. But at same time local news were informing that 30 000 homes lacked electricity and some people got stuck in their cars. There was no public transport, no trains, no buses. Much smaller amount of people on streets.

Was it worth attending? Yes. Will I attend next ones? Probably not as this is very developer related where I spend most of my OpenStack time around building it’s components or doing some testing.

OpenStack ‘Queens’ release done

OpenStack community released ‘queens’ version this week. IMHO it is quite important moment for AArch64 community as well because it works out of the box for us.

Gone are things like setting hw_firmware_type=uefi for each image you upload to Glance — Nova assumes UEFI to be the default firmware on AArch64 (unless you set the variable to different value for some reason). This simplifies things as users does not have to worry about and we should have less support questions on new setups of Linaro Developer Cloud (which will be based on ‘Queens’ instead of ‘Newton’).

There is a working graphical console if your guest image uses properly configured kernel (4.14 from Debian/stretch-backports works fine, 4.4 from Ubuntu/xenial (used by CirrOS) does not have graphics enabled). Handy feature which we were asked already by some users.

Sad thing is state of live migration on AArch64. It simply does not work through the whole stack (Nova, libvirt, QEMU) because we have no idea what exactly cpu we are running on and how it is compatible with other cpu cores. In theory live migration between same type of processors (like XGene1 -> XGene1) should be possible but we do not have even that level of information available. More information can be found in bug 1430987 reported against libvirt.

Less sad part? We set cpu_model to ‘host-passthrough’ by default now (in Nova) so nevermind which deployment method is used it should work out of the box.

When it comes to building (Kolla) and deploying (Kolla Ansible) most of the changes were done during Pike cycle. During Queens’ one most of the changes were small tweaks here and there. I think that our biggest change was convincing everyone in Kolla(-ansible) to migrate from MariaDB 10.0.x (usually from external repositories) to 10.1.x taken from distribution (Debian) or from RDO.

What will Rocky bring? Better hotplug for PCI Express machines (AArch64/virt, x86/q35 models) is one thing. I hope that live migration stuff situation will improve as well.

Hotplug in VM. Easy to say…

You run VM instance. Nevermind is it part of OpenStack setup or just local one started using Boxes, virt-manager, virsh or other that kind of fronted to libvirt daemon. And then you want to add some virtual hardware to it. And another card and one more controller…

Easy to imagine scenario, right? What can go wrong, you say? “No more available PCI slots.” message can happen. On second/third card/controller… But how? Why?

Like I wrote in one of my previous posts most of VM instances are 90s pc hardware virtual boxes. With simple PCI bus which accepts several cards to be added/removed at any moment.

But not on AArch64 architecture. Nor on x86-64 with Q35 machine type. What is a difference? Both are PCI Express machines. And by default they have far too small amount of pcie slots (called pcie-root-port in qemu/libvirt language). More about PCI Express support can be found in PCI topology and hotplug page of libvirt documentation.

So I wrote a patch to Nova to make sure that enough slots will be available. And then started testing. Tried few different approaches, discussed with upstream libvirt developers about ways of solving the problem and finally we selected the one and only proper way of doing it. Then discussed failures with UEFI developers. And went for help to Qemu authors. And explained what I want to achieve and why to everyone in each of those four projects. At some point I had seen pcie-root-port things everywhere…

Turned out that the method of fixing it is kind of simple: we have to create whole pcie structure with root port and slots. This tells libvirt to not try any automatic adding of slots (which may be tricky if not configured properly as you may end with too small amount of slots for basic addons).

Then I went with idea of using insane values. VM with one hundred PCIe slots? Sure. So I made one, booted it and then something weird happen: landed in UEFI shell instead of getting system booted. Why? How? Where is my storage? Network? Etc?

Turns out that Qemu has limits. And libvirt has limits… All ports/slots went into one bus and memory for MMCONFIG and/or I/O space was gone. There are two interesting threads about it on qemu-devel mailing list.

So I added magic number into my patch: 28 — this amount of pcie-root-port entries in my aarch64 VM instance was giving me bootable system. Have to check it on x86-64/q35 setup still but it should be more or less the same. I expect this patch to land in ‘Rocky’ (the next OpenStack release) and probably will have to find a way to get it into ‘Queens’ as well because this is what we are planning to use for next edition of Linaro Developer Cloud.

Conclusion? Hotplug may be complicated. But issues with it can be solved.

Developers planet is online

People write blogs. People read blogs. But sometimes it is hard to find blogs of all those interesting people. That’s where so called “planets” are solution.

Years ago there was “Planet Linaro” website filled with blog posts from Linaro developers. Then it vanished. Later it got replaced by poor substitute.

But I do not want to have to track each Linaro developer to find their blog and add it into Feedly. So instead I decided to create new planet website. And that’s how Developers Planet got born.

So far it lists a bunch of blogs of Linaro developers. I used venus to run it. Few years old code but runs. Will adapt HTML/CSS template to be a bit more modern.

And why .cf domain? It is free — that’s why.

Graphical console in OpenStack/aarch64

OpenStack users are used to have graphical console available. They take it for granted even. But we lacked it…

When we started working on OpenStack on 64-bit ARM there were many things missing. Most of them got sorted out already. One thing was still in a queue: graphical console. So two weeks ago I started looking at the issue.

Whenever someone tried to use it Nova reported one simple message: “No free USB ports.” You can ask what it has to console? I thought similar and started digging…

As usual reason was simple: yet another aarch64<>x86 difference in libvirt. Turned out that arm64 is one of those few architectures which do not have USB host controller in default setup. When Nova is told to provide graphical console it adds Spice (or VNC) graphics, video card and USB tablet. But in our case VM instance does not have any USB ports so VM start fails with “No free USB ports” message.

Solution was simple: let’s add USB host controller into VM instance. But where? Should libvirt do that or should Nova? I discussed it with libvirt developers and got mixed answers. Opened a bug for it and went to Nova hacking.

Turned out that Nova code for building guest configuration is not that complicated. I created a patch to add USB host controller and waited for code reviews. There were many suggestions, opinions and questions. So I rewrote code. And then again. And again. Finally 15th version got “looks good” opinion from all reviewers and got merged.

And how result looks? Boring as it should:

OpenStack graphical console on aarch64

Devconf.cz? FOSDEM? PTG? Linaro Connect?

What connects those names? All of them are conference or team meeting names. Spread over Europe + Asia. And I will be on all of them this year.

First will be in Brno, Czechia. Terrible city to travel to but amount of people I can meet there is enormous. Some guys from my Red Hat team will be there, my boss’ boss (and boss of my boss’ boss) and several people from CentOS, Fedora, RHEL (and some other names) communities. Meetings, sessions… Nice to be there. Will be in A-Sport hotel as it is closest to Red Hat office where I have some meetings to attend.

Then FOSDEM. The only “week-long conference squeezed into two days” I know. Good luck with meeting me probably. As usual my list of sessions covers all buildings and have several conflicts. Will meet many people, miss even more, do some beers. This year I stay in Floris Arlequin Grand-Place hotel.

Next one? OpenStack PTG in Dublin, Ireland. Finally will meet all those developers reviewing my patches, helping me with understanding source code of several OpenStack projects. And probably answering several questions about state of AArch64 support. Conference hotel.

Linaro Connect. Hong Kong again. Not registered yet, not looked at flights. Have to do that sooner than later and checking which airline sucks less on intercontinental connections sucks too. With few members of SDI team I will talk about our journey through/with OpenStack on our beloved architecture. Conference hotel.

What after those? Probably OpenSource Day in Poland, maybe some other ones. Will see.

I am now core reviewer in Kolla

Months of work, tens of patches, hundreds of changed lines. My whole work in Kolla project got rewarded this week. I am now one of core reviewers 🙂

What does it mean? I think that Gema summarised it best:

For those of you who don’t know, this means Kolla has recognised our contributions to the project as first class and are giving Marcin and ARM64 a vote of confidence, they realise we are there to stay.

I found it helpful in my daily work as now I can suggest my coworkers to send their patches directly instead of proxying it through me ;D

Can Socionext SynQuacer be first 96boards desktop machine?

During Linaro Connect SFO17 I had an occasion to take a look at first 96boards Enterprise Edition MicroATX format board: Socionext SynQuacer. Can it be called first 96boards desktop machine?

Just to remind — 96boards EE specification defined two form factors:

  • custom 160x120mm
  • MicroATX

There were attempts to build boards in that custom format (Husky, Cello) but they both failed terribly. Turns out that companies which are able to produce 96boards CE boards are not able to make more complicated ones.

Connect ago I wrote about Systart Oxalis LS1020A board as being first 96boards EE one but it used that custom format.

So going back to SynQuacer board…

I would say that it looks like typical MicroATX mainboard:

  • four memory slots (DDR4, up to 64GB, ECC or not ECC)
  • CPU under heatsink (24 Cortex-A53 cores, 1GHz clock)
  • PCI-Express slots (x1, x1, x16 with just 4 lanes)
  • two SATA ports
  • Gigabit Ethernet port
  • two USB 3.0 ports at the back
  • connector for another USB 3.0 ports
  • 96boards low-speed connector (think sensors, serial console, tpm etc)
  • 24pin ATX power connector (no extra +12V ones)
  • power and reset buttons
  • fan connector
  • JTAG port

Socionext SynQuacer

The official announcement did not provide information about price. Only info present was that it will available in December 2017. During discussions with Socionext representatives I was told that full developer box will cost around 1000 USD and involve mainboard, memory, storage (rather not SSD), case and graphics card. Price for just mainboard was not provided as it looked like such option is not planned.

From software point of view there was UEFI presented. With graphical boot. Upstreaming kernel support is in progress (Linaro provides 4.14-rc tree with required changes).

Will it satisfy a need for AArch64 desktop? Time will show. From what I got from developers using it already performance is quite ok as long as it is multithreaded (so kernel build goes nice with -j24 until linking phase kicks in).

Other option for AArch64 desktop would be Macchiatobin. Latest revisions are needed as PCI support got fixed (I was told that first revisions were unable to fully use PCI Express port). Bernhard Rosenkränzer was demoing such setup and it was running nicely.