When assigning between physical lvalues, emit a 'copy_addr' instead of a load + assign sequence. This provides a higher-level semantic instruction to SIL passes, and also produces better code in many cases.
It unfortunately exposes another bug in DI which affects the test/IRGen/objc.swift test, which I've XFAILed until it can be fixed.
Swift SVN r10564
Route global var refs (except for those referencing top-level code vars) through the lazy-initializing global accessor functions for those globals.
Swift SVN r10519
For every global pattern binding, emit a lazy initializer token and function that initializes the global variables in that binding. For each of those vars, create an accessor that Builtin.once's the lazy initializer before producing the address. Hide this all behind a switch till the surrounding serialization and IRGen infrastructure catches up.
Swift SVN r10511
of having to lower to an RValue.
This is valuable because we can often emit an expression to a
desired abstraction level more efficiently than just emitting
it to minimal abstraction and then generalizing.
Swift SVN r10455
- Enhance SILBuilder::emitStrongRelease to be smarter.
- Start using emitStrongRelease in type lowering, SILGen,
CapturePromotion (replacing its implementation of the
same logic), and MandatoryInlining (one more place)
- Rename the primitive createStrongRetain/ReleaseInst
instructions to lose their suffix.
- Now that createStrongRetain/ReleaseInst are not special
cases from the naming perspective, remove some special cases
from DeserializeSIL and ParseSIL.
Swift SVN r10449
They are the same as createStrongRetainInst and createStrongReleaseInst, but
peephole away FunctionRefInst. It turns out that there is only a couple
places in SILGen where this behavior is necessary, and this tramples on the
general pattern used in SILBuilder.
Swift SVN r10448
Previously, we were only looking at the innermost level, ignoring
outer generic types. Fixes the SILGen part of
<rdar://problem/15345742>, but IRGen issues remain.
Swift SVN r10418
(the same way alloc_box returns two) instead of returning a tuple.
This eliminates a ton of tuple_extract instructions, which just
bloat the generated SIL. This resolves rdar://15378135.
Swift SVN r10416
<rdar://problem/14835992> happens because, when we emitted a chain of checked casts for 'is' patterns in a 'switch', we would emit abstraction changes for each individual cast that required them. This is both redundant and makes cleaning up messier, because the number of branches with different stack heights increases. Instead, let's emit one abstraction change up front for all of the casts. No functionality change yet, but this is a step on the way to fixing <rdar://problem/14835992>.
Swift SVN r9921
I tried hard find all references to 'func' in documentation, comments and
diagnostics, but I am sure that I missed a few. If you find something, please
let me know.
rdar://15346654
Swift SVN r9886
If a local function tried to reference a local typealias, we were trying to emit a SIL variable reference and freaking out, causing <rdar://problem/15360605>. For now, don't emit any capture arguments for local typealiases. Any context a local typealias requires should be handled by capturing the external context's generic parameters.
Swift SVN r9865
The empty tuple used for return value when none is supplied does not really correspond to user code. (We don’t have epilog, so mark it as pert of cleanup.)
The destroy_addr instructions are doing the cleanup.
Swift SVN r9765
We already know how to call external functions in SILGen, why reimplement all that in IRGen? Instead of representing the thunk operation as a SIL instruction, let's just emit the thunk using existing SILGen machinery. Fixes <rdar://problem/14097136>.
Swift SVN r9576
And tweak the verifier to allow 'convert_cc' over a static FunctionRefInst to remain thin. The SILGen part of <rdar://problem/14097136>.
Swift SVN r9556
Rather than wrapping a DeclRefExpr in a SpecializeExpr for a reference
to a generic free function, just use the substitutions stored within
the DeclRefEXpr. Such DeclRefExprs will always have a concrete
type. One tiny nail in the SpecializeExpr coffin.
Swift SVN r9539
location rather than a regular location to avoid the linetable jumping
back to the beginning of the function.
Add a large number of testcases that check the sanity of the return
locations in various scenarios involving implicit/explicit returns,
cleanups, and multiple return locations per function.
rdar://problem/14845534
Swift SVN r9511
Now that we have a solid Optional-based story for dynamic casts, it's no longer needed, and can be expressed as '(x as T)!'. Future refinement of the 'as' syntax will deal with the unfortunate extra parens.
Swift SVN r9181
This is a simple peephole optimization for cases where we're
performing a conditional checked cast and then immediately forcing it
to succeed with postfix '!'. It also happens with the new forced
downcast of 'id', e.g.,
var w : NSView = foo.views()[0]!
Swift SVN r9180
If we're using emitExprInto to emit a LoadExpr, we can handle that by emitting a copy_addr from the underlying lvalue to the initialization instead of building and storing the rvalue, which plays better with the direction we're taking for value semantics optimizations.
This is currently guarded behind a flag -enable-silgen-lvalue-peepholes because it introduces a miscompile in the stdlib that crashes some tests; requires further investigation before making live.
Swift SVN r9027
The other end of r8989. Canonicalize toward destroy_addr when destroying lvalues already in memory, instead of trying to emit a more specific sequence of operations.
Swift SVN r9022
Emit the 'self' argument of constructors first so that its box's lifetime follows proper stack discipline relative to the other arguments.
Swift SVN r8984
We were leaking dealloc_stacks when the enum constructor had to create temporary allocations in order to implode address-only tuple payloads.
Swift SVN r8971
preceding copy_addr instruction when totally trivial. Adopt this in SILGen, eliminating
a couple dozen destroy_addr instructions from the stdlib and producing more canonical SIL.
Swift SVN r8968
As with the monadic '?', we treat any left-bound '!' as a postfix
operator. Currently, it extracts the value of its optional
subexpression, failing at run-time if the optional is empty.
Swift SVN r8948
Parse '_' as a DiscardAssignmentExpr. Type-check it as an lvalue, and check that it only appears in the LHS of AssignExprs. During matching pattern resolution, convert it into an AnyPattern. In SILGen, when we see '_' in the LHS of an assignment, ignore the corresponding RHS rvalue.
Swift SVN r8848
Though we plan to revamp the casting syntax, our general plan is for this form of cast, which does a conditional cast and returns an Optional<T> result, to be the one that survives. Parse the status-quo syntax 'x as? T' and type-check it. While we're here, refresh some fixits for redundant casts that referred to the now defunct 'as T' coercion syntax to completely remove whatever cast was in the source code.
Swift SVN r8805
