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

[release-0.5] backports for 0.5.0 rc2 #17960

Merged
merged 80 commits into from
Aug 12, 2016
Merged

Conversation

tkelman
Copy link
Contributor

@tkelman tkelman commented Aug 11, 2016

A few things aren't here yet. The curl build still needs a bit more work on master and might have to wait until rc3, and the SSH tests are disabled here with a commit that we'll hopefully be able to revert at some point, since they're causing issues on the buildbots.

yuyichao and others added 30 commits August 10, 2016 21:11
The byte gets sign extended when passing to the vararg `jl_printf` and then
printed as an unsigned int which might come with unwanted `0xff` prefix...

(cherry picked from commit bffa84e)
ref #17803
introduced by #17522

(cherry picked from commit a3f288c)
ref #17815
Fixed missing UnicodeCharacters doc/conf.py, see #17813
(cherry picked from commit f566f0d)
ref #17814

remove patch copy-paste + artifacts from doc/conf.py

(cherry picked from commit 6ba6d8d)
Lots of our documentation for some relatively simple iterator methods
had no examples. I added a bunch, moved docstrings out of HelpDB, and
tried to illustrate some common use cases.

(cherry picked from commit ad5d244)
ref #17790
Lots of examples for things that are probably confusing for new
Julia users. Made `reducedim` and `mapreducedim` arguments more
similar to the ones for `reduce` and `mapreduce`.
(cherry picked from commit 3e00217)
The manual entry for doctests was way out of date! Updated
with an example of what to do.
(cherry picked from commit 6d3e337)
* Move math docs out of HelpDB, more examples, fix typos

Found a series of typos in `cov` and friends. Added more notes
about `NaN` and Julia. Made the function signatures reflect
what's actually in the code. More examples for quite a few
functions.

* Move quadgk docs out, update formatting

* Moved special functions out of HelpDB, insert some links

* Updated docs for some array ops as well

* Updated in response to feedback

Removed calls to `rand` in doctests. Made examples
better. Cleaned up function signatures.

(cherry picked from commit 2d24eda)
It was already working for matrices

(cherry picked from commit 9114ae6)
ref #17816
* Apparently `ifunc` is not supported by gcc 5 on ARM
* Try even harder to workaround LLVM bug when dealing with returntwice function

(cherry picked from commit f90eab3)
ref #17800
(cherry picked from commit f2863ef)
ref #17836
The `edit` function on windows fails if the editor command is `atom`. It succeeds if it is instead `atom.cmd`, but then Julia fails to recognize the `atom` editor. This fixes that.
(cherry picked from commit e79bee7)
ref #17857
It is now in a separate repo and not in `contrib/`

(cherry picked from commit c367b42)
ref #17849
Yet MORE examples from the REPL. Fixed and added links,
changd some wording.

(cherry picked from commit e4a377c)
ref #17797
Removed calls to `rand` in the doctests.

(cherry picked from commit 6df3ec0)
ref #17797
(cherry picked from commit 071cd66)
ref #17797
(cherry picked from commit fdef48d)
ref #17797
(cherry picked from commit 6ffe1ab)
ref #17797
(cherry picked from commit 03e9262)
ref #17864
(cherry picked from commit c7fc048)
ref #17871
Keno and others added 9 commits August 10, 2016 22:28
…17909)

and Symmetric, it is redundant to also have an uplo argument and
occasionally, it also gave the wrong result. This can therefore be
considered a bugfix.
(cherry picked from commit 35df9b8)
* Add cross-release compatibility note to CONTRIBUTING.md

* tag -> specify

(cherry picked from commit dcffb4a)
* improve heuristic for when to refresh line

* fix trailing whitespace...

* use indent when not on first line

* remove unused argument and change local var to not share name with function

(cherry picked from commit f84b94c)
See #16766. `ndigits(big(0),1)` doesn't raise an error, but we were
sometimes (randomly, rarely) testing that it does. This change makes
the test reliable. The issue of how this case should actually
behave is still open.

(cherry picked from commit dd09d19)
ref #17952
fix output of getaddrinfo in networking and streams documentation for v0.5.
(cherry picked from commit 068aa9d)
@tkelman tkelman added this to the 0.5.0 milestone Aug 11, 2016
@tkelman
Copy link
Contributor Author

tkelman commented Aug 11, 2016

@nanosoldier runbenchmarks(ALL, vs = "@0030eec2f332f353e6890ca289ac2aca55532dde")

(PR merge commit 92c85fd vs 0.5.0-rc0)

@nanosoldier
Copy link
Collaborator

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

@tkelman
Copy link
Contributor Author

tkelman commented Aug 11, 2016

PkgEval comparison (vs rc0) finished: https://gist.github.com/c5e9489fdacaeed4a31abfe69ebbbe1b

This is a few more failures than I would like to see (9, though 8 packages went from failing to passing), would be good to get a few more eyes looking through the failure logs.

edit: BlockArrays, DistributedArrays, ForwardDiff, and PowerSeries failures are caused by this typo: 2912c9e#commitcomment-18608997

Git.jl was terminated, timeout apparently?

@KristofferC
Copy link
Member

For BlockArrays it is because a typo:

(:acot, :atan, :acoth, :atahn, "cotangent"))

@swt30
Copy link
Contributor

swt30 commented Aug 11, 2016

Same for ForwardDiff, I think

ERROR: LoadError: LoadError: UndefVarError: atahn not defined
in macro expansion; at /home/vagrant/.julia/v0.5/ForwardDiff/test/DualTest.jl:388 [inlined]

@tkelman
Copy link
Contributor Author

tkelman commented Aug 11, 2016

BiomolecularStructures and NullableArrays pass for me locally, so those may have been flukes. That leaves
CovarianceMatrices (cc @gragusa) and MixedModels (cc @dmbates) that look like real failures we don't know the cause of, worth bisecting.

@tkelman
Copy link
Contributor Author

tkelman commented Aug 11, 2016

@andreasnoack CovarianceMatrices was broken by the backport of #17909, can you take a look?

@andreasnoack
Copy link
Member

Sure. I'll do that.

@tkelman
Copy link
Contributor Author

tkelman commented Aug 11, 2016

MixedModels bisected to the same commit.

@andreasnoack
Copy link
Member

@tkelman Thanks for catching this before it got backported. This check is comparing a Char and a Symbol so it ends up exactly the opposite of what it should be. The tests actually captured this but unfortunately I convinced myself that the tests were wrong. #17985 reverts the old commit and adds a new commit without the bug.

Keno and others added 4 commits August 12, 2016 01:26
…mitian (#17909)"

This reverts commit d992f3d.

(cherry picked from commit 62b55ce)
ref #17985
and Symmetric, it is redundant to also have an uplo argument and
occasionally, it also gave the wrong result. This can therefore be
considered a bugfix.

(cherry picked from commit 6085103)
ref #17985
* Fix atahn to atanh

* More tests

(cherry picked from commit 5838a12)
@tkelman tkelman merged commit e3d9acc into release-0.5 Aug 12, 2016
@tkelman tkelman deleted the tk/backports-0.5.0-rc2 branch August 12, 2016 10:11
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.