Funnel all places where we create a generic signature builder to compute
the generic signature through a single entry point in the GSB
(`computeGenericSignature()`), and make `finalize` and `getGenericSignature`
private so no new uses crop up.
Tighten up the signature of `computeGenericSignature()` so it only works on
GSB rvalues, and ensure that all clients consider the GSB dead after that
point by clearing out the internal representation of the GSB.
introduce a common superclass, SILNode.
This is in preparation for allowing instructions to have multiple
results. It is also a somewhat more elegant representation for
instructions that have zero results. Instructions that are known
to have exactly one result inherit from a class, SingleValueInstruction,
that subclasses both ValueBase and SILInstruction. Some care must be
taken when working with SILNode pointers and testing for equality;
please see the comment on SILNode for more information.
A number of SIL passes needed to be updated in order to handle this
new distinction between SIL values and SIL instructions.
Note that the SIL parser is now stricter about not trying to assign
a result value from an instruction (like 'return' or 'strong_retain')
that does not produce any.
Once we compute a generic signature from a generic signature builder,
all queries involving that generic signature will go through a separate
(canonicalized) builder, and the original builder can no longer be used.
The canonicalization process then creates a new, effectively identical
generic signature builder. How silly.
Once we’ve computed the signature of a generic signature builder, “register”
it with the ASTContext, allowing us to move the existing generic signature
builder into place as the canonical generic signature builder. The builder
requires minimal patching but is otherwise fully usable.
Thanks to Slava Pestov for the idea!
Funnel all places where we create a generic signature builder to compute
the generic signature through a single entry point in the GSB
(`computeGenericSignature()`), and make `finalize` and `getGenericSignature`
private so no new uses crop up.
Tighten up the signature of `computeGenericSignature()` so it only works on
GSB rvalues, and ensure that all clients consider the GSB dead after that
point by clearing out the internal representation of the GSB.
This patch implements collection and dumping of statistics about SILModules, SILFunctions and memory consumption during the execution of SIL optimization pipelines.
The following statistics can be collected:
* For SILFunctions: the number of SIL basic blocks, the number of SIL instructions, the number of SIL instructions of a specific kind, duration of a pass
* For SILModules: the number of SIL basic blocks, the number of SIL instructions, the number of SIL instructions of a specific kind, the number of SILFunctions, the amount of memory used by the compiler, duration of a pass
By default, any collection of statistics is disabled to avoid affecting compile times.
One can enable the collection of statistics and dumping of these statistics for the whole SILModule and/or for SILFunctions.
To reduce the amount of produced data, one can set thresholds in such a way that changes in the statistics are only reported if the delta between the old and the new values are at least X%. The deltas are computed as using the following formula:
Delta = (NewValue - OldValue) / OldValue
Thresholds provide a simple way to perform a simple filtering of the collected statistics during the compilation. But if there is a need for a more complex analysis of collected data (e.g. aggregation by a pipeline stage or by the type of a transformation), it is often better to dump as much data as possible into a file using e.g. -sil-stats-dump-all -sil-stats-modules -sil-stats-functions and then e.g. use the helper scripts to store the collected data into a database and then perform complex queries on it. Many kinds of analysis can be then formulated pretty easily as SQL queries.
The main loop of mandatory inlining is spending a lot of time managing complex
iterator invalidation issues. This is the first in a series of commits that move
the main inlining loop to only delete the callee and to do all cleanups after we
have finished inlining.
This specific optimization (the quick retain/release peephole), I am not going
to do in MandatoryInlining, we already have guaranteed arc opts afterwards that
will be able to hit such a peephole so no perf should be lost.
*NOTE* The reason why I had to touch some of the code motion tests is that the
routine I am using to ensure that strong_retain/release_value is emitted as
appropriate is also used by codemotion. Code motion tests had cargo culted some
code from previous tests that retained Builtin.Int32. I changed the routines
though so that when a retain/release is inserted, if it is trivial, nothing is
inserted. No routine was relying on the actual usage of the inserted
retain/releases, so everything will be safe. This addition to the relevant code
caused me to need to change the tests in code motion to use actual non-trivial
values. The same code paths are being tested in terms of blocking code
motion/etc.
rdar://31521023
The reason to do this is:
1. The check in SILInliner if we can inline can be done without triggering
side-effects.
2. This enables us to know if inlining will succeed before attempting to inline.
This enables for arguments to be adjusted with new SILInstructions and the like
before inlining occurs. I use this in a forthcoming patch that updates mandatory
inlining for ownership.
rdar://31521023
The etymology of these terms isn't about race, but "black" = "blocked"
and "white" = "allowed" isn't really a good look these days. In most
cases we weren't using these terms particularly precisely anyway, so
the rephrasing is actually an improvement.
"Accessibility" has a different meaning for app developers, so we've
already deliberately excised it from our diagnostics in favor of terms
like "access control" and "access level". Do the same in the compiler
now that we aren't constantly pulling things into the release branch.
Rename AccessibilityAttr to AccessControlAttr and
SetterAccessibilityAttr to SetterAccessAttr, then track down the last
few uses of "accessibility" that don't have to do with
NSAccessibility. (I left the SourceKit XPC API alone because that's
supposed to be more stable.)
"Accessibility" has a different meaning for app developers, so we've
already deliberately excised it from our diagnostics in favor of terms
like "access control" and "access level". Do the same in the compiler
now that we aren't constantly pulling things into the release branch.
This commit changes the 'Accessibility' enum to be named 'AccessLevel'.
This commit contains:
-) adding the new instructions + infrastructure, like parsing, printing, etc.
-) support in IRGen to generate global object-variables (i.e. "heap" objects) which are statically initialized in the data section.
-) IRGen for global_value which lazily initializes the object header and returns a reference to the object.
For details see the documentation of the new instructions in SIL.rst.
Remove the cast consumption kind from all unconditional casts. It
doesn't make sense for unconditional casts, complicates SIL ownership,
and wasn't fully supported for all variants. Copies should be
explicit.
Count how many generic specializations were prevented due to the possibility of creating an infinite generic specialization loop and due to the complexity of their generic type parameters.
Allow for cases, where the old substitution type `T1` is partially contained in the new substitution type `T2`. Partially contained means that if you drop the common structural "prefix" of `T1` and `T2` and get `T1'` and `T2'` then `T1'` is strictly contained in `T2'`. E.g. `Outer<Start>` is partially contained in `Outer<Step<Start>>` if you drop the common prefix `Outer`, then `Start` is contained in `Step<Start>`
The existing simple mechanism for avoiding infinite generic specialization loops is based on checking the structural depth and width of types passed as generic type parameters. If the depth or the width of a type is above a certain threshold, the type is considered too complex for generic specialization and no specialization is produced. While this approach prevents the possibility of producing an infinite number of generic specializations for ever-growing generic type parameters, it catches the issue too late in some cases, leading to excessive CPU and memory usage.
Therefore, the new method tries to solve the problem at its root. An infinite generic specialization loop can be triggered by specializing a given generic call-site if and only if:
- Doing so would result in a loop inside the specialization graph represented by the `GenericSpecializationInformations`, i.e. it would produce direct or indirect recursion involving a generic call
- The substitutions used by the current generic call-site are structurally more complex than the substitutions used by the same call-site in the previous iteration inside specialization graph. More complex in this context means that the new generic type parameter structurally contains the generic type parameter from a previous iteration inside the specialization graph and has greater structural depth, e.g. `Array<Int>` is more complex than `Int`.
The generic specializer now records all the required information about specializations it produces and uses it later to detect and prevent any generic specializations which would result in an infinite specialization loop. It detects them as early as possible and thus reduces compile times, memory consumption and potentially also reduces the code-size by not generating useless specializations.
Since the return type of the callee is not changed and the old apply instruction is not removed by devirtualizeClassMethod, there is no need to insert an "unreachable" instruction, because by construction rules of SIL it should be already there right after the old apply.
This bug was caught by the SIL verifier. Any invocation of a NoReturn function should be followed by an `unreachable` instruction.
Fixes rdar://problem/33591235
In dead-array elimination we assume that the array allocation is post-dominated by all its final releases.
The only exception are branches to dead-end ("unreachable") blocks. So we just ignored all paths which didn't end up in a final release.
Now we explicitly pass the set of dead-end blocks and just ignore those blocks.
This is safer and it's also needed in the upcoming re-write of StackPromotion.
