-
Notifications
You must be signed in to change notification settings - Fork 5.1k
Added PowerShell.Benchmarks to SuperPMI collections #93493
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Conversation
Tagging subscribers to this area: @JulieLeeMSFT, @jakobbotsch Issue DetailsThis adds the PowerShell.Benchmarks, from the recently merged dotnet/performance#3000, to the SuperPMI collections.
|
@kunalspathak this is ready. |
Seems to fail at the moment actually:
|
Current failure is the targetframework is set to net7, made a PR dotnet/performance#3420 to update it to net8. |
@BruceForstall This is ready. See successful pipeline: https://dev.azure.com/dnceng/internal/_build/results?buildId=2293539&view=logs&j=6e5fe0ba-f6d5-5f75-5db7-b07a5a8eee39&t=b89c5eaa-be13-560b-8e20-13db295b7d16
NativeAOT failure is not related. |
Something seems wrong. There should be more than 8 new, unique, functions. It doesn't look like the benchmark actually ran. In the per-run log it looks like all the collected functions are either System.* or BenchmarkDotNet. E.g., I see this failure:
|
Ok, looking into it. |
It's missing the |
So, the @BruceForstall @kunalspathak You know of anything special that I may need to do here? It does fail locally when I try to run through SuperPMI:
|
I am wondering if the benchmark runs properly after updating dotnet/performance#3420? If yes, you might want to download the payload of other real-world benchmark to see its folder structure and command line parameters for superpmi. |
@kunalspathak I think there is a bug when running benchmarks when specifying a custom |
so this is outside of superpmi and just normally running the benchmarks? |
Running benchmarks normally works if I do this: Supplying a
|
@dotnet/jit-contrib @BruceForstall This is ready. It was able to get a collection: https://dev.azure.com/dnceng/internal/_build/results?buildId=2315554&view=results |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
It looks like there's a problem with the Linux collections, e.g.:
---> System.DllNotFoundException: Unable to load shared library 'advapi32.dll' or one of its dependencies.
also, the Linux collections only add about 140 MC versus about 4300 for Windows.
@BruceForstall , this is ready again. I checked the OSX and Linux pipelines and they are successfully running the benchmarks and collecting. Now, the Windows benchmarks do run two more than Linux and OSX but this is due to the configuration of the tests themselves and not due to issues of missing assemblies. (Was able to replicate that locally) https://dev.azure.com/dnceng/internal/_build/results?buildId=2317989&view=results |
Not related to this PR, but it looks like the ILLink benchmark is broken / not supported on almost ever platform:
and broken on linux/win-x64:
|
Thanks @BruceForstall . Should we file a report to dotnet/performance since it may be a benchmark issue itself? |
Not sure. We should open a bug against the SPMI collection though |
This adds the PowerShell.Benchmarks, from the recently merged dotnet/performance#3000, to the SuperPMI collections.