WIP: Backports 1.2.0 rc2 #32270
WIP: Backports 1.2.0 rc2 #32270
Conversation
- The default mingw32 gfortran on ubuntu prints the version as `GNU Fortran (GCC) 7.3-win32 20180312`, rather than `7.3.0` most other gfortrans do, so be robust to that in normalize_triplet - When shelling out to python use the host path rather than the windows-equivalent path, since python is a host tool. (cherry picked from commit 36d6d7a)
(cherry picked from commit d8798be)
(cherry picked from commit 9628242)
(cherry picked from commit 2ebe0de)
* readdlm(bytearray) shouldn't modify bytearray * Update stdlib/DelimitedFiles/src/DelimitedFiles.jl Co-Authored-By: Jeff Bezanson <[email protected]> (cherry picked from commit 7038210)
KristofferC
added
DO NOT MERGE
(cherry picked from commit d02ad0a)
KristofferC
added
the
needs pkgeval
|
@nanosoldier |
|
Your benchmark job has completed - possible performance regressions were detected. A full report can be found here. cc @ararslan |
* fixes for getindex tab-completion * test fix (cherry picked from commit 1eca37e)
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
This comment has been minimized.
|
PkgEval failures (each link is to the corresponding log):
Looks like a number of these are failures when loading undeclared dependencies. |
|
A lot of packages (the ones I marked) are missing (test) dependencies in their |
|
The EponymTuples seems real and the MWE is var_and_value_form(ex::Symbol) = (e = esc(ex); :($(e) = $(e)))
macro eponymtuple(args::Vararg{Symbol})
forms = map(var_and_value_form, args)
Expr(:tuple, forms...)
end
@macroexpand @eponymtuple(a, b)which on 1.1 gives :((a = a, b = b))while on 1.2 gives :((a = (a = a), b = (b = b)))Perhaps you can look at this @JeffBezanson |
|
@nanosoldier |
|
#32464 should fix EponymTuples |
|
Your benchmark job has completed - possible performance regressions were detected. A full report can be found here. cc @ararslan |
|
Things that needs looking into:
StochasticPrograms |
|
DistributedArrays just needs a new tag. |
DataFrames failing is pretty bad. Again, I think we should just test dependencies anyway, even if their parent fail. |
|
MIDI: the |
|
I believe the NaturallyUnitful problem is MasonProtter/NaturallyUnitful.jl#5. |
|
Matrixnetwork passses locally multiple times on this branch. |
|
StochasticModels error can be reprod with: using StochasticModels
farmer_model = @stochastic_model begin
@stage 1 begin
end
@stage 2 begin
@uncertain ξ::YieldScenario = begin end
end
endon this branch it erros with ERROR: UndefVarError: YieldScenario not definedWhen I run @scenario YieldScenario = begin
end
StochasticPrograms.StochasticModel(StageParameters(), StageParameters()) do sp
@stage 1 sp = begin
end
@stage 2 sp = @uncertain(ξ::YieldScenario = begin
end)
endand this executes correctly. So not sure what is happening. @JeffBezanson, perhaps you could take a look. |
(cherry picked from commit 07c2ecc)
|
Including those commits here seems reasonable to me, so I've added the backport 1.2 label to the PRs. I assume they should be backported to 1.0 as well? |
points to the last non-opt argument of argv, but in the case where there are only options optind does not go beyond argc, except on musl libc, where it becomes argc + 1. (cherry picked from commit 9de07ce)
KristofferC
removed
DO NOT MERGE
|
The macOS buildbot seems to be consistently failing, but it works on Travis, so it's probably fine. Similarly, AppVeyor is failing but the Windows buildbots are working. The GC analysis build fails because the infrastructure for that is not present on the |
….0-rc2 Backports 1.2.0 rc2
Backported PRs:
import Distributed; Distributed.@everywhere x = 2does not work #28966.v0.3.5-2, to fix FreeBSD #32192 - Bump OpenBLAS BB tarballs tov0.3.5-2, to fix FreeBSDgetipaddrs#32260 - Remove internal sort fromgetipaddrsAbstractMatrixas the eltype ofSymmetricfor array of arrays #32041 - RFC: useAbstractMatrixas the eltype ofSymmetricfor array of arrayslet#32326 - fix Defining a struct with an outer constructor in local scope gives bad error #32325, struct plus outer ctor insideletmalloc_trimsometimes #32428 - fix Memory leak with function redefinition #30653, callmalloc_trimsometimesNon-merged PRs with backport label: