Backports 1.5.1 #36899
Backports 1.5.1 #36899
Conversation
|
Can we add #36819 as a backport? |
|
Sure, done. |
|
Backport for the arm cpu feature detection would be good too #36485 |
|
Upgrading libuv seems a bit too big for a patch release, in my opinion. |
|
Could you add #36705 to enable building on ARMv7l? |
KristofferC
added
release
|
Probably a good idea to backport the documentation footer for SEO/tiobe etc. #36983 |
|
Looks like we cannot use e438d0e, since it calls |
KristofferC
force-pushed
the
backports-release-1.5
branch
from
fbac504 to
317b2cf
Compare
|
|
@nanosoldier |
|
@nanosoldier |
|
Your benchmark job has completed - possible performance regressions were detected. A full report can be found here. cc @ararslan |
|
Your package evaluation job has completed - possible new issues were detected. A full report can be found here. cc @maleadt |
|
@nanosoldier |
|
Dropping #36856 since it changes exception types. |
KristofferC
force-pushed
the
backports-release-1.5
branch
from
317b2cf to
dea950e
Compare
|
Your package evaluation job has completed - possible new issues were detected. A full report can be found here. cc @maleadt |
(cherry picked from commit 87c609a)
When several Julia processes compile the same package concurrently (e.g. during a cluster run), they can conflict on the compile cache file. This change makes a Julia process create a compile cache in a temporary file and atomically rename it to the final cache file. Co-authored-by: Takafumi Arakaki <tkf@@users.noreply.github.com> (cherry picked from commit 3bbb582)
This was really just a regression when switching from grisu -> ryu algorithm for float printing. The fix is just putting [@vtjnash's code](5d2a0ec#diff-22819a3d47074bbe6530dfef9c94969cR75) for the grisu fix into the ryu branch. (cherry picked from commit ada82ac)
The code was using two different `std::pair` types to pass this info around as remnant of an earlier version and the auto conversion between the two types does not retain the meaning we want... Use a custom struct to hold the info to guard against this kind of errors in the future. (cherry picked from commit 2d3288e)
* Clean up old LLVM version check * Clean up old cores that'll probably not be used (and some that are removed from LLVM) * Add document about data sources * Update AArch64 feature list to match changes to the LLVM ones * ARMv8.4 - ARMv8.6 support * Add all known cores that I can find and all information I can find about them * Remove a few A64 only cores from AArch32 CPU list * Rename Apple cores (following LLVM) * SVE register size Co-authored-by: Ian <[email protected]> (cherry picked from commit e0e3825)
* Missing feature from Apple A13
* Enable Cortex-A78 and Cortex-X1 on LLVM 11
llvm/llvm-project@954db63
https://reviews.llvm.org/D83206
* More relaxed Zen detection: treat all family 23 as Zen* and treat all model >= 0x30 as Zen2.
GCC uses a similar fallback structure albeit based on feature.
This should still generate **correct** code since that is always controlled by
available features. It should be as good a scheduling model estimate as anything else.
Fix #36826
(cherry picked from commit cd3fb4d)
We might print ANSI color codes around ERROR now, making this test unreliable. (cherry picked from commit 537664f)
(cherry picked from commit a652cd8)
(cherry picked from commit 102c9ac)
…ad of 1.5.0-rc2.0 (cherry picked from commit 50c4e67)
(cherry picked from commit cd4330d)
Co-authored-by: Ian <[email protected]> Co-authored-by: Fredrik Ekre <[email protected]> (cherry picked from commit 5be3a54)
Not sure why this was first linked statically, as the commit that introduced this simply had the message "restoring stuff that seems to have been clobbered by the revert of the unintended merge to master". Nearly all other libraries that we use are linked dynamically. (cherry picked from commit 9267bbf)
…signed Julia (#37069) Without this entitlement, users that wish to attach a debugger to the codesign Julia executable they receive from `julialang.org` must strip the codesigning signature from the Julia executable (e.g. via `codesign --remove-signature Julia-1.5.app/Contents/Resources/bin/julia`). This has its disadvantages, of course, so much better to simply declare to the OS that it's alright for other processes to attach to this process. (cherry picked from commit bf8a8e8)
(cherry picked from commit 903542b)
KristofferC
force-pushed
the
backports-release-1.5
branch
from
04d7574 to
a197786
Compare
|
Should be good to go now. |
Backported PRs:
Unionin supertype #36996 - fix julia crashes during method disptch #36869, incorrect intersection withUnionin supertypeget-task-allowentitlement to allow debuggers to attach to codesigned Julia #37069 - Addget-task-allowentitlement to allow debuggers to attach to codesigned Juliamapreduce#37105 - workaround for problem with the unreliable approximation ofCore.Compiler.return_type#35800, inference issue inmapreduceNon-merged PRs with backport label: