Skip to content
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

Simplify scalar indexing #19958

Merged
merged 1 commit into from
Jan 28, 2017
Merged

Conversation

mbauman
Copy link
Member

@mbauman mbauman commented Jan 10, 2017

This removes the generated functions for scalar indexing and modifies Array indexing to get it ready for the deprecation of generalized linear indexing (#14770). Let's see if CI likes this. Last time I tried this simplification Win64 didn't like it. And then we can send it through nanosoldier.

This is currently targeting my indexing conversion branch, so nanosoldier should be compared against that, not master.

@mbauman mbauman changed the base branch from mb/too_many_indices to master January 10, 2017 00:33
@mbauman
Copy link
Member Author

mbauman commented Jan 10, 2017

Hm, retargeted to master to get CI to run.

@mbauman mbauman closed this Jan 10, 2017
@mbauman mbauman reopened this Jan 10, 2017
@mbauman
Copy link
Member Author

mbauman commented Jan 10, 2017

@nanosoldier runbenchmarks(ALL, vs = "9cd522c948b4c1e4d4d75eeaa402dbfbe7d1ebe7")

@KristofferC
Copy link
Member

@ in front of SHA for commit benchmarks

@mbauman
Copy link
Member Author

mbauman commented Jan 10, 2017

Oh my goodness. I should not be trying to do this on my phone. Thanks Kristoffer.

@nanosoldier runbenchmarks(ALL, vs = "@9cd522c948b4c1e4d4d75eeaa402dbfbe7d1ebe7")

@nanosoldier
Copy link
Collaborator

Your benchmark job has completed - possible performance regressions were detected. A full report can be found here. cc @jrevels

@mbauman
Copy link
Member Author

mbauman commented Jan 10, 2017

Ah, of course, Nanosoldier uses the merge commit from the PR. Those improvements are certainly unrelated. The regressions are likely real.

@mbauman mbauman force-pushed the mb/simpler-scalar-indexing branch from 1154b92 to bc565e1 Compare January 13, 2017 16:56
@mbauman
Copy link
Member Author

mbauman commented Jan 13, 2017

@nanosoldier runbenchmarks(ALL, vs = ":master")

@nanosoldier
Copy link
Collaborator

Your benchmark job has completed - possible performance regressions were detected. A full report can be found here. cc @jrevels

@StefanKarpinski
Copy link
Member

Windows failure is a timeout. Might be worth restarting it.

@tkelman
Copy link
Contributor

tkelman commented Jan 16, 2017

What's the status on this one, will this be refactored?

@mbauman
Copy link
Member Author

mbauman commented Jan 16, 2017

The first commit is still a nice simplification. The second commit is messy and unnecessary if we do the partial linear indexing deprecation within checkbounds… and it's also likely the cause of the performance regressions.

@mbauman mbauman force-pushed the mb/simpler-scalar-indexing branch 2 times, most recently from e9477be to 853e920 Compare January 16, 2017 17:30
@mbauman
Copy link
Member Author

mbauman commented Jan 16, 2017

@nanosoldier runbenchmarks(ALL, vs = ":master")

@mbauman mbauman force-pushed the mb/simpler-scalar-indexing branch from 853e920 to bed4377 Compare January 16, 2017 17:32
@mbauman mbauman changed the title Simplify scalar indexing and prepare for #14770 Simplify scalar indexing Jan 16, 2017
@nanosoldier
Copy link
Collaborator

Your benchmark job has completed - possible performance regressions were detected. A full report can be found here. cc @jrevels

Copy link
Contributor

@tkelman tkelman left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

lgtm!

@StefanKarpinski
Copy link
Member

Bump – Is this ready to go?

@mbauman
Copy link
Member Author

mbauman commented Jan 27, 2017

Yup, but it's just an internal refactoring that touches some pretty foundational methods, so I had been thinking it could wait for the branch. It's totally ready to go, though… so anyone is welcome to push that button! I'll re-start CI since the last result is getting a little stale.

@mbauman mbauman closed this Jan 27, 2017
@mbauman mbauman reopened this Jan 27, 2017
@KristofferC
Copy link
Member

+37 -98 is a reason to merge in my opinion...

@tkelman tkelman merged commit 1445962 into JuliaLang:master Jan 28, 2017
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

5 participants