ARM 32-bit progress #5460

manu-st · 2016-03-30T00:24:31Z

I'm opening an issue to track the progress of ARM 32-bit with respects to the regression tests.

Currently I'm getting the following results on Ubuntu 14.04 running on a NVidia Jetson TK1. This is unfortunately not the results you get from a clean checkout, the only difference is that it contains a fix for issue #5422 (PR dotnet/coreclr#3879), but overall it should provide a good base to monitor our progress:

=======================
     Test Results
=======================
# Tests Discovered : 6018
# Passed           : 4313
# Failed           : 1361
# Skipped          : 344
=======================

I've attached the XML results:
coreclrtests.zip

The text was updated successfully, but these errors were encountered:

leemgs · 2016-03-30T01:29:43Z

This is unfortunately not the results you get from a clean checkout,

Even though you could get the 1,361 Failed result, this result is meaningful to me. Actually, we have been meeting a lot of "NYI" functions from CoreCLR.:( It means that we have to do a lot of things for stablization of CoreCLR.

Currently I'm getting the following results on Ubuntu 14.04 running on a NVidia Jetson TK1

BTW, Do you also focus on the Ubuntu/ARM32bit without the Ubuntu/ARM64bit?
http://elinux.org/Jetson_TK1

CPU: NVIDIA "4-Plus-1" 2.32GHz ARM quad-core Cortex-A15 CPU with Cortex-A15 battery-saving shadow-core
GPU: NVIDIA Kepler "GK20a" GPU with 192 SM3.2 CUDA cores (upto 326 GFLOPS)
Memory: ~ 8GiB DDR3

\CC: @myungjoo , @lemmaa
They also interested in the regression test of CoreCLR.

manu-st · 2016-03-30T02:13:10Z

@leemgs I'm not sure to understand what you mean by Ubuntu/ARM32bit without the Ubuntu/ARM32bit part?

leemgs · 2016-03-30T02:36:37Z

@manu-silicon Sorry, Typo, I have modified.

manu-st · 2016-03-30T02:39:57Z

@leemgs My board is the 32-bit variant, so it is just Ubuntu/ARM32bit.

https://en.wikipedia.org/wiki/Tegra#Tegra_K1

manu-st · 2016-03-31T00:24:39Z

With the fix for PR dotnet/coreclr#3981 the result is:

=======================
     Test Results
=======================
# Tests Discovered : 6019
# Passed           : 4793
# Failed           : 882
# Skipped          : 344
=======================

coreclrtests.zip

prajwal-aithal · 2016-03-31T05:16:32Z

@manu-silicon I am trying to do the same (run coreclr tests on ARM environment) by following this guide https://github.com/dotnet/coreclr/blob/master/Documentation/building/unix-test-instructions.md. According to this, building tests for ARM results in an error (as building the native code using cmake in Windows does not support arm architecture yet). I can make some changes to the buildtest script to build only the C# based tests for ARM architecture.

I wanted to ask how you are building the test binaries for ARM environment (using buildtest.cmd). Have you made some local changes (to build only the C# code and ignore the CMake build) or is there some other way to do it?

manu-st · 2016-03-31T05:33:02Z

@prajwal-aithal I've followed the instructions and build the tests on Windows and copy the tests folder on Linux ARM. I built on ARM the native code for CoreFX (just the native code as the rest won't build). I built CoreFX on Linux x64 and copied the assemblies over. I built mscorlib on Linux x64 and copied it over.

Once I had all the paths setup properly, I launched the command to run the tests.

prajwal-aithal · 2016-03-31T05:43:28Z

@manu-silicon Oh, ok. I will try this then. Thank you :)

manu-st · 2016-03-31T23:42:44Z

Today's result with FEATURE_STUBS_AS_IL disabled:

=======================
     Test Results
=======================
# Tests Discovered : 6019
# Passed           : 4836
# Failed           : 839
# Skipped          : 344
=======================

coreclrtests.zip

jyoungyun · 2016-04-12T10:16:46Z

Target: Raspberry Pi2
CoreClr commit id : ff26d68 (around april 9)

=======================
     Test Results
=======================
# Tests Discovered : 7422
# Passed           : 6074
# Failed           : 1017
# Skipped          : 331
=======================

I'm curious why my test discovered count is different with others.
CoreClr_UnitTest_Results_160412_01.zip

myungjoo · 2016-04-12T10:40:09Z

@jyoungyun Next time, please attach the result XML file as well.

manu-st · 2016-04-12T23:37:26Z

@jyoungyun The difference in the number of tests is related on how I build first the test on Windows and then copied them over to the Linux ARM device. As @myungjoo is saying, having the XML file would enable us to compare and see what the differences are.

jyoungyun · 2016-04-19T08:35:26Z

@manu-silicon I attached xml file. Thank you.

jyoungyun · 2016-04-27T09:46:36Z

Target: Raspberry Pi2
Coreclr commit id : 31fada1 + dotnet/coreclr#4460 + dotnet/coreclr#4503

==========================
      Test Results
==========================
# Tests Discovered : 7421
# Passed           : 6547
# Failed           : 543
# Skipped          : 331
==========================

CoreClr_UnitTest_Results_160427.zip

masonwheeler · 2016-04-27T10:14:22Z

Wow, so those two fixes cleared up about half of the failing tests? Nice!

jyoungyun · 2016-04-28T04:38:31Z

@masonwheeler Yes! The dotnet/coreclr#4460 patch reduced half of failing tests. :)

jyoungyun · 2016-04-28T04:44:28Z

Target: Raspberry Pi2
Coreclr commit id: 18268be + dotnet/coreclr#4503 + dotnet/coreclr#4581

==========================                                                      
      Test Results                                                              
==========================                                                      
# Tests Discovered : 6018                                                       
# Passed           : 5295                                                       
# Failed           : 395                                                        
# Skipped          : 328                                                        
==========================

And I found the reason why my tests discovered count is different with others.
My test directory structure was wrong. Now it's the same with others. :)

CoreClr_UnitTest_Results_160428.zip

leemgs · 2016-04-28T05:05:39Z

@jyoungyun , It's amazing. : ) BTW, Is dotnet/coreclr@18268be related to the number of the reduced failure? It seems that it is not related to the failure numbers of the unit test for Linux/ARM.

manu-st · 2016-04-28T05:05:48Z

@jyoungyun If possible, it would be nice to know how many tests each bug fix addresses.

manu-st · 2016-04-28T05:06:55Z

@jyoungyun Out of curiosity, how long does it take you to run all the tests. For me on the Jetson TK1, it takes roughly half a day.

myungjoo · 2016-04-28T05:15:52Z

@manu-silicon We will be pushing out patches that accelerate the execution of corerun at ARM soon. @leemgs is preparing PR for that (release build bugfix)

leemgs · 2016-04-28T05:18:06Z

For reference, @manu-silicon has shared the hardware specification that he used for getting the result of unit test. Here is the detail contents of his development board.

http://elinux.org/Jetson_TK1 (using Quad-core Cortex-A15 CPU and 2GiB RAM).

jyoungyun · 2016-04-28T09:06:31Z

@leemgs Oh, The number(18268be) means the base commit. Someone want to reproduce the result, the base commit info. will be helpful. Reducing failure test cases are more to do with dotnet/coreclr#4460 patch, I think.
@manu-silicon It takes a lot of time to run unix test on Raspberry pi2. In my case, about 12 hours is the usaual time required. But when I use a release mode binary, it takes only 4 hours! I hope that @leemgs patch will be ready soon.

hqueue · 2016-05-03T13:06:45Z

In the lastet commit(92d7091), I got much better result. I was investigating several failures and many of them are fixed in the latest master branch.
Can anybody run the whole test again?

manu-st · 2016-05-04T02:29:32Z

@hqueue I can do it on Friday.

manu-st · 2016-05-09T00:57:27Z

Tried to run the test suite on Friday, but it got stuck on 2 tests that I had to manually kill:

FAILED   - JIT/Regression/CLR-x86-JIT/V2.0-Beta2/b425314/b425314/b425314.sh
               BEGIN EXECUTION
               /home/ubuntu/local/Microsoft/tests/Tests/coreoverlay/corerun b425314.exe
               ./b425314.sh: line 62: 32428 Killed                  $_DebuggerFullPath "$CORE_ROOT/corerun" b425314.exe $CLRTestExecutionArguments $Host_Args
               Expected: 100
               Actual: 137
               END EXECUTION - FAILED

FAILED   - JIT/Regression/CLR-x86-JIT/V2.0-Beta2/b426654/b426654/b426654.sh
               BEGIN EXECUTION
               /home/ubuntu/local/Microsoft/tests/Tests/coreoverlay/corerun b426654.exe
               ./b426654.sh: line 62: 10285 Killed                  $_DebuggerFullPath "$CORE_ROOT/corerun" b426654.exe $CLRTestExecutionArguments $Host_Args
               Expected: 100
               Actual: 137
               END EXECUTION - FAILED

manu-st · 2016-05-09T02:35:00Z

The run finally completed. Results on Jetson TK1 commit 3ddd9c4

=======================
     Test Results
=======================
# Tests Discovered : 6019
# Passed           : 5478
# Failed           : 234
# Skipped          : 307
=======================

coreclrtests_2016_05_06.zip

jyoungyun · 2016-05-10T07:14:14Z

The latest unit test results!

Target: Raspberry Pi2
Coreclr commit id: 6b92cff

==========================
      Test Results
==========================
# Tests Discovered : 6530
# Passed           : 5981
# Failed           : 215
# Skipped          : 334
==========================

CoreClr_UnitTest_Results_160510.zip

@manu-silicon When I run unix tests, some tc takes such a long time. And the blocked tc is changed. Have you seen the same case ?

myungjoo · 2016-05-10T07:42:22Z

Note that we have about 120 fails due to regression being hopefully fixed easily with dotnet/coreclr#4888. We may see under 100 fails soon.

manu-st · 2016-05-10T08:18:41Z

@jyoungyun I'm not sure what you mean by "tc is changed". But as I said it just before posting my results, I had 2 tests that was hanging preventing the rest of the test cases to run. I had to kill them manually to get to the completion.

jaestevan · 2017-03-30T18:18:36Z

Then is Ubuntu in the Pi the only way to run the project? I was guessing it can be ran in Windows 10 IoT too.

Petermarcu · 2017-03-30T18:20:18Z

@jaestevan take a look here: https://github.com/dotnet/core/blob/master/samples/RaspberryPiInstructions.md

Petermarcu · 2017-03-30T18:23:25Z

I'm curious what you hit on Win10 IoT? You shouldn't need to install any core runtime. The instruction put the runtime with your application.

darxis · 2017-03-30T18:26:53Z

In case it would help anyone, I'm posting my instruction notes on how to install the dotnet tool on RPi3 to run .net core apps.

Clone coreclr repo
Compile coreclr
1. Install everything from https://github.com/dotnet/coreclr/blob/master/Documentation/building/windows-instructions.md
2. build.cmd arm release skiptests skipbuildpackages
Clone corefx repo
Compile corefx
1. build-native.cmd -release -buildArch=arm
2. build-managed.cmd -release -buildArch=arm -skiptests
Clone core-setup repo
Compile core-setup
1. build.cmd -Configuration Release -TargetArch arm
Copy output files to Windows 10 IoT
Get-ChildItem "U:\netcore\corefx" -Directory | ForEach-Object { Copy-Item $("U:\netcore\corefx{0}{0}.dll" -f $_.name) "U:\netcore\corefx" }
Get-ChildItem "U:\netcore\corefx" -Directory | ForEach-Object {Write-Host $_.name } Remove-Item -recurse

With this I got working an .NET Core console app and an ASP.NET Core web app that is running 24/7 for about 2 months.

EDIT: My notes were taken when there were no official instructions on how to install on RPi3. Consider first trying these ones https://github.com/dotnet/core/blob/master/samples/RaspberryPiInstructions.md

Gonkers · 2017-03-30T18:31:52Z

Sorry @jaestevan I should have been more clear. I meant to point out the screen shots of how/were you can update the framework version. To see if that would help you get it running. That was my biggest hang up for a while, figuring out building for the correct version of the framework.

jaestevan · 2017-03-30T18:41:26Z

@Petermarcu by following this instructions I got a CLR error: Failed to initialize CoreCLR, HRESULT: 0X80131534

@Gonkers oh, I see what you mean. Thanks.

@darxis I will try to build it myself when I have time, thanks.

@ALL: Found this link: http://developers.de/blogs/damir_dobric/archive/2015/09/22/how-to-run-net-core-application-on-pi2-win10-iot-core.aspx Still have not tried it myself but I will. Anyway I guess it should be a cleaner way when ARM support is official.

Petermarcu · 2017-03-30T18:51:23Z

@jaestevan , I'll give it another try, hadn't done it on Win10 IoT in a few weeks. You shouldn't have to build the full stack but its always an option.

jaestevan · 2017-03-30T19:59:04Z

@Petermarcu Thanks Peter, let me know how it works. I can send you the details I followed if that helps.

Petermarcu · 2017-03-31T17:15:11Z

I tried it again yesterday and everything worked really well.

<Project Sdk="Microsoft.NET.Sdk">
  <PropertyGroup>
    <OutputType>Exe</OutputType>
    <TargetFramework>netcoreapp2.0</TargetFramework>
    <RuntimeIdentifiers>win10-arm</RuntimeIdentifiers>
  </PropertyGroup>
</Project>

dotnet restore
dotnet publish -r win10-arm

Then I copy bin\Debug\netcoreapp2.0\win10-arm\publish\* to the Pi and execute .\Helloworld.exe

I have one guess as to what is happening. Can you run dotnet --version to make sure you are on a 2.0.0 SDK and make sure you have the latest one? What I do for the pre-release SDK's is I extract the zip to some location, open up a command windows, make sure the path to where I exctracted the zip is the first thing on my path, and then run dotnet commands. Right now, 1.0.0 is winning over 2.0.0 because of an issue in the SDK that is going to be fixed in the next couple weeks.

nrandell · 2017-04-10T10:28:38Z

I've just been experimenting with this and have it working with SDK version 2.0.0-preview1-005743 with runtime version 2.0.0-preview1-001915-00.

I've been able to edit and build using that latest Visual Studio 2017! All looking really good.

I've got a repository at https://github.com/sceneskope/PiDotnet which has some projects where this works. I'm actually using docker as containers for my images as the projects I'm working on have to run on multiple PIs.

TAGC · 2017-04-10T10:43:22Z

I've also created a .NET Core app that runs on an RPI 3 using SDK 2.0.0-preview1-005685 and runtime framework 2.0.0-beta-001834-00.

Runs fine natively (although I get exec format errors if trying to run it within a Docker container based on microsoft/dotnet-nightly:2.0.0-beta-runtime).

nrandell · 2017-04-10T10:57:45Z

The Microsoft images won't be for the ARM architecture, which is why I've created my own set of docker images! It would be great to have proper Microsoft images though.

TAGC · 2017-04-10T11:07:07Z

@nrandell Oh nice, I missed that. I'll have to try it out at some point.

MihaMarkic · 2017-04-11T07:41:59Z

@nrandell Is that a fact written down somewhere? I mean no ARM docker images from MS.

nrandell · 2017-04-11T08:49:48Z

@MihaMarkic have a look at dotnet/dotnet-docker#223. I guess this is as official as it gets.

MihaMarkic · 2017-04-12T10:43:17Z

@nrandell So, they'll provide one for .net core 2.0 eventually but skip 1.x.

jaestevan · 2017-04-14T17:35:47Z

Upgraded and rebuilt everything with 2.0.0-preview1-001978-00 and now it works in the Pi 💃

AndreiRinea · 2017-04-17T10:58:39Z

Excuse me if this has been discussed before, but will this run on Raspberry Pi Zero too?

benpye · 2017-04-17T11:01:34Z

Unfortunately not, the Pi Zero is only ARMv6, Mono will work there though.

…

On Mon, 17 Apr 2017, 11:58 Andrei Rînea, ***@***.***> wrote: Excuse me if this has been discussed before, but will this run on Raspberry Pi Zero too? — You are receiving this because you commented. Reply to this email directly, view it on GitHub <https://github.com/dotnet/coreclr/issues/3977#issuecomment-294452108>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AAbA_0ZBbplbgWzERjKkrw-MV7buPiDxks5rw0XvgaJpZM4H7Qi3> .

wanton7 · 2017-04-17T11:36:06Z

I think pi 1 and zero have ARMv6k that has some enhancements compared to ARMv6.

masonwheeler · 2017-04-17T11:40:03Z

@wanton7 Hoo boy, even more versionitis for us to keep straight?

wanton7 · 2017-04-17T11:56:08Z

@masonwheeler not sure what the differences are but example next golang (1.9 i think) is going to still support ARMv6k but they are dropping support for ARMv6. So must be something significant.

BruceForstall · 2018-04-25T00:55:13Z

@Petermarcu I propose to close this issue; we're far past the "ARM 32-bit progress" title -- we have many thousands of tests running on ARM Linux in our CI. Agreed?

myungjoo · 2018-04-25T01:14:01Z

I agree. ;)

Petermarcu closed this as completed Apr 25, 2018

msftgits transferred this issue from dotnet/coreclr Jan 30, 2020

msftgits added this to the Future milestone Jan 30, 2020

lemmaa mentioned this issue Jan 30, 2020

[ARM64/Linux] Cross building CoreCLR ARM64 on Linux #6012

Closed

8 tasks

BruceForstall mentioned this issue Jan 31, 2020

ARM32 LEGACY_BACKEND status #8604

Closed

ghost locked as resolved and limited conversation to collaborators Jan 1, 2021

ARM 32-bit progress #5460

ARM 32-bit progress #5460

Comments

manu-st commented Mar 30, 2016

leemgs commented Mar 30, 2016

manu-st commented Mar 30, 2016

leemgs commented Mar 30, 2016

manu-st commented Mar 30, 2016

manu-st commented Mar 31, 2016

prajwal-aithal commented Mar 31, 2016

manu-st commented Mar 31, 2016

prajwal-aithal commented Mar 31, 2016

manu-st commented Mar 31, 2016

jyoungyun commented Apr 12, 2016 • edited

myungjoo commented Apr 12, 2016

manu-st commented Apr 12, 2016

jyoungyun commented Apr 19, 2016

jyoungyun commented Apr 27, 2016

masonwheeler commented Apr 27, 2016

jyoungyun commented Apr 28, 2016 • edited

jyoungyun commented Apr 28, 2016

leemgs commented Apr 28, 2016

manu-st commented Apr 28, 2016

manu-st commented Apr 28, 2016

myungjoo commented Apr 28, 2016

leemgs commented Apr 28, 2016

jyoungyun commented Apr 28, 2016

hqueue commented May 3, 2016 • edited

manu-st commented May 4, 2016

manu-st commented May 9, 2016

manu-st commented May 9, 2016

jyoungyun commented May 10, 2016 • edited

myungjoo commented May 10, 2016

manu-st commented May 10, 2016

jaestevan commented Mar 30, 2017

Petermarcu commented Mar 30, 2017

Petermarcu commented Mar 30, 2017

darxis commented Mar 30, 2017 • edited

Gonkers commented Mar 30, 2017

jaestevan commented Mar 30, 2017 • edited

Petermarcu commented Mar 30, 2017

jaestevan commented Mar 30, 2017

Petermarcu commented Mar 31, 2017 • edited

nrandell commented Apr 10, 2017

TAGC commented Apr 10, 2017

nrandell commented Apr 10, 2017

TAGC commented Apr 10, 2017

MihaMarkic commented Apr 11, 2017

nrandell commented Apr 11, 2017

MihaMarkic commented Apr 12, 2017

jaestevan commented Apr 14, 2017

AndreiRinea commented Apr 17, 2017

benpye commented Apr 17, 2017 via email

wanton7 commented Apr 17, 2017

masonwheeler commented Apr 17, 2017

wanton7 commented Apr 17, 2017

BruceForstall commented Apr 25, 2018

myungjoo commented Apr 25, 2018

jyoungyun commented Apr 12, 2016 •

edited

jyoungyun commented Apr 28, 2016 •

edited

hqueue commented May 3, 2016 •

edited

jyoungyun commented May 10, 2016 •

edited

darxis commented Mar 30, 2017 •

edited

jaestevan commented Mar 30, 2017 •

edited

Petermarcu commented Mar 31, 2017 •

edited