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.

Sign up for GitHub

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

Jump to bottom

[release-0.5] fix type stability of unrelated function after precompile #18869

Merged
merged 5 commits into from
Nov 4, 2016
Merged

[release-0.5] fix type stability of unrelated function after precompile #18869

merged 5 commits into from
Nov 4, 2016

Conversation

vtjnash
Copy link
Member

@vtjnash vtjnash commented Oct 11, 2016

(this bug also exists on master, but in a different form so it needs a different fix there)

@tkelman tkelman mentioned this pull request Oct 11, 2016
Merged
@vtjnash
Copy link
Member Author

vtjnash commented Oct 12, 2016

I assume @tkelman will merge this when he wants it (since it is directly targeting the release-0.5 branch)

@tkelman
Copy link
Contributor

tkelman commented Oct 12, 2016

I'll want to run pkgeval, make sure it doesn't break anything. And figure out what's wrong on osx travis...

@kshyatt kshyatt added compiler:precompilation Precompilation of modules bugfix This change fixes an existing bug labels Oct 12, 2016
@vtjnash
reduce the stack-space used by unoptimized jl_deserialize_value_
15459bf 
with -O0 builds, the number of local variables quick blows out the available stack space
pushing these into separate functions helps cut down on the stack requirement in such cases

this is perhaps just a stopgap measure to rewriting the serializer to be
faster, non-recursive, and non-allocating

(cherry picked from commit b4d49e5, PR #18811)
@vtjnash
fix an incremental deserializer bug dealing with methods instances th…
ed7002e 
…at don't dominate the subtype signature of their method definition

(cherry picked from commit f6db1cc, PR #18979,
and commit 3b33217)
@vtjnash
type-intersection may not be computable before jl_recache_types has f…
4d4ca5b 
…inished

for example, computing subtypes might construct one of the types that we are trying to later unique
and mess up on that logic

also be more permissive of type-intersection failures in jl_recache_method_instance

(cherry picked from commit 39afcc3, PR #19029)
This was referenced Oct 20, 2016
Merged
Merged
@KristofferC KristofferC mentioned this pull request Oct 22, 2016
Closed
@tkelman
Copy link
Contributor

tkelman commented Nov 4, 2016

I'll run pkgeval in a separate PR and ping you if it turns out this causes any issues.

@tkelman tkelman merged commit 8da61fe into release-0.5 Nov 4, 2016
@tkelman tkelman deleted the jn/release-0.5-infer-precompile branch November 4, 2016 07:58
@tkelman tkelman mentioned this pull request Nov 4, 2016
Closed
@iamed2 iamed2 mentioned this pull request Nov 8, 2016
Closed
@iamed2
Copy link
Contributor

iamed2 commented Nov 8, 2016

This broke Atom.jl (and thus Juno):
#19272
http://discuss.junolab.org/t/error-loading-atom-package/1027/2

@yuyichao yuyichao mentioned this pull request Feb 1, 2017
Closed
@joshbode joshbode mentioned this pull request Feb 1, 2017
Closed
@joshbode joshbode mentioned this pull request Mar 26, 2017
Closed
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@JeffBezanson JeffBezanson JeffBezanson approved these changes

Assignees
No one assigned
Labels
bugfix This change fixes an existing bug compiler:precompilation Precompilation of modules
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
5 participants
@vtjnash @tkelman @iamed2 @JeffBezanson @kshyatt