The metadata reference to the pre-exsting VWT cannot be supported on PE/COFF
due to the direct reference to a value in an external module that is in
static data (the model requires indirecting through memory). Always
force the lazy initialization for the metadata on such platforms.
This requires a secondary change - to initialize the VWT as well. This
is ideally moved into the runtime where we can do this uniformly.
By default, the frontend tries to figure out if the built module is
likely to be distributed in some way, and uses that to decide whether
to include options that help with debugging (such as local search
paths). There's long been a -serialize-debugging-options that forces
those options to be included even when it looks like a framework is
being built, but the opposite has been absent until now.
Note that both of these options are still /frontend/ options, not
driver options, which means they could still change in the future.
(I'd really like to get to a point where debugging doesn't need to
sniff these options out from the module this way, but there are some
complications we'd need to work out. Swift 1 expediency coming back to
cause trouble again.)
rdar://problem/37954803
Because this feature mostly exists for the standard library, turn it on
specifically. This avoids having to serialize it in the flags necessary
for parseable interfaces.
...by printing them with a dummy, unsatisfiable condition. This
happens when a public type conforms to a public protocol with
non-public conditions; the conformance can't be used in inlinable
code, but neither is it okay for a client to declare their own
conformance (constrained or unconstrained).
rdar://problem/45657450
If runtime asserts are disabled in general, also disable runtime
exclusivity checks unless explicitly requested.
-Ounchecked is an unsupported benchmarking mode, but, if used, it
should disable all runtime assertions. Exclusivity checks are no more
essential than integer overflow or bounds checks.
In this code:
private protocol MyProto: Hashable {}
public struct MyStruct: MyProto {}
Being Hashable is part of MyStruct's public API, even though it's not
written explicitly. If we're not going to require people to write it
explicitly, we need to make sure it gets printed.
rdar://problem/44662501
A module compiled with `-enable-private-imports` allows other modules to
import private declarations if the importing source file uses an
``@_private(from: "SourceFile.swift") import statement.
rdar://29318654
`#assert` is a new static assertion statement that will let us write
tests for the new constant evaluation infrastructure that we are working
on. `#assert` works by lowering to a `Builtin.poundAssert` SIL
instruction. The constant evaluation infrastructure will look for these
SIL instructions, const-evaluate their conditions, and emit errors if
the conditions are non-constant or false.
This commit implements parsing, typechecking and SILGen for `#assert`.
This reverts commit 121f5b64be.
Sorry to revert this again. This commit makes some pretty big changes. After
messing with the merge-conflict created by this internally, I did not feel
comfortable landing this now. I talked with Saleem and he agreed with me that
this was the right thing to do.
This change could impact Swift programs that previously appeared
well-behaved, but weren't fully tested in debug mode. Now, when running
in release mode, they may trap with the message "error: overlapping
accesses...".
Recent optimizations have brought performance where I think it needs
to be for adoption. More optimizations are planned, and some
benchmarks should be further improved, but at this point we're ready
to begin receiving bug reports. That will help prioritize the
remaining work for Swift 5.
Of the 656 public microbenchmarks in the Swift repository, there are
still several regressions larger than 10%:
TEST OLD NEW DELTA RATIO
ClassArrayGetter2 139 1307 +840.3% **0.11x**
HashTest 631 1233 +95.4% **0.51x**
NopDeinit 21269 32389 +52.3% **0.66x**
Hanoi 1478 2166 +46.5% **0.68x**
Calculator 127 158 +24.4% **0.80x**
Dictionary3OfObjects 391 455 +16.4% **0.86x**
CSVParsingAltIndices2 526 604 +14.8% **0.87x**
Prims 549 626 +14.0% **0.88x**
CSVParsingAlt2 1252 1411 +12.7% **0.89x**
Dictionary4OfObjects 206 232 +12.6% **0.89x**
ArrayInClass 46 51 +10.9% **0.90x**
The common pattern in these benchmarks is to define an array of data
as a class property and to repeatedly access that array through the
class reference. Each of those class property accesses now incurs a
runtime call. Naturally, introducing a runtime call in a loop that
otherwise does almost no work incurs substantial overhead. This is
similar to the issue caused by automatic reference counting. In some
cases, more sophistacated optimization will be able to determine the
same object is repeatedly accessed. Furthermore, the overhead of the
runtime call itself can be improved. But regardless of how well we
optimize, there will always a class of microbenchmarks in which the
runtime check has a noticeable impact.
As a general guideline, avoid performing class property access within
the most performance critical loops, particularly on different objects
in each loop iteration. If that isn't possible, it may help if the
visibility of those class properties is private or internal.
Currently, the check for whether to serialize parseable interface
arguments doesn't handle the case where a supplementary output file map
is used, preferring only to check if the frontend is passed
`-emit*interface`. Instead, check if the frontend inputs and outputs
contains a parseable interface, and use that to determine if we need to
save args.
This also puts `-module-link-name` in the parseable interface arg list.
