Re-enable needs_finalize intrinsic #43
Conversation
|
Please squash. |
|
Squashed |
jacob-hughes
force-pushed
the
fix_needs_finalize
branch
from
e75980f to
8b4374d
Compare
|
bors r+ |
|
Build failed: |
|
Pushed a rustfmt fix, LMK if you're happy for me to squash. |
|
Please squash. |
jacob-hughes
force-pushed
the
fix_needs_finalize
branch
from
1f788f5 to
6e69c3c
Compare
|
Squashed |
|
bors r+ |
|
Build failed: |
|
Lets give this another kick now that we are dockerized. bors try |
|
Am I ok to rebase this with master? |
tryBuild failed: |
|
Took me a while to spot:
|
|
@jacob-hughes We've still got a build failure (well, a test failure), so we need to fix that before considering squashing. |
|
Ok, this test is very irritating because I get the correct output on my Arch machine. Let me see if I can wildcard the broken bit, it's not the relevant part of the test anyway. |
|
bors try |
tryBuild failed: |
|
bors try |
tryBuild failed: |
|
bors try |
tryBuild failed: |
|
I had a look at using a make test after lunch but it looks like it's going require a bit more effort at the moment. In the meantime I've ignored this test and I'll open an issue so that we can move on. When I squash, I'll make note of this in the commit message. |
|
Please squash. |
jacob-hughes
force-pushed
the
fix_needs_finalize
branch
from
c1d6c23 to
5e09347
Compare
|
Squashed |
|
bors r+ |
|
Build failed: |
|
bors try |
|
If bors succeeds, then please squash. If it doesn't... don't :) |
tryBuild succeeded: |
The GcSmartPointer trait was a hangover which let the compiler identify `Gc<T>` structs when they were defined as part of a non-rustc library. Now that the `Gc` implementation is part of the standard library, this is unnecessary and adds complexity. This requires the prevent_early_finalization mir-opt tests to be temporarily ignored due to an issue with non-deterministic builds when bringing in the alloc crate across different machines. [1] [1]: softdevteam#45
This fixes the broken implementation of `std::mem::needs_finalize()` which because of changes in upstream to the needs_drop implementation in rustc. This introduces a new module where needs_finalize is tested in isolation instead of being bolted onto the ever-changing needs_drop internals.
jacob-hughes
force-pushed
the
fix_needs_finalize
branch
from
8e92b6a to
2978eef
Compare
|
Squashed |
|
bors r+ |
|
Build succeeded: |
This fixes the broken implementation of
std::mem::needs_finalize()which was disabled because of changes in upstream to the needs_drop implementation in rustc.This introduces a new module where needs_finalize is tested in isolation instead of being bolted onto the ever-changing needs_drop internals.