Commit Graph

20 Commits

Author SHA1 Message Date
Michael Gottesman
f854547c55 [ownership] Enable ownership verification by default.
I also removed the -verify-sil-ownership flag in favor of a disable flag
-disable-sil-ownership-verifier. I used this on only two tests that still need
work to get them to pass with ownership, but whose problems are well understood,
small corner cases. I am going to fix them in follow on commits. I detail them
below:

1. SILOptimizer/definite_init_inout_super_init.swift. This is a test case where
DI is supposed to error. The only problem is that we crash before we error since
the code emitting by SILGen to trigger this error does not pass ownership
invariants. I have spoken with JoeG about this and he suggested that I fix this
earlier in the compiler. Since we do not run the ownership verifier without
asserts enabled, this should not affect compiler users. Given that it has
triggered DI errors previously I think it is safe to disable ownership here.

2. PrintAsObjC/extensions.swift. In this case, the signature generated by type
lowering for one of the thunks here uses an unsafe +0 return value instead of
doing an autorelease return. The ownership checker rightly flags this leak. This
is going to require either an AST level change or a change to TypeLowering. I
think it is safe to turn this off since it is such a corner case that it was
found by a test that has nothing to do with it.

rdar://43398898
2019-03-25 00:11:52 -07:00
Arnold Schwaighofer
e113ef8c93 Add a mode to test implicit dynamic with private imports 2019-03-20 14:34:01 -07:00
Michael Gottesman
0dfaa19f9f [ownership] Rename enable-sil-ownership => verify-sil-ownership.
I have been meaning to do this change for a minute, but kept on putting it off.
This describes what is actually happening and is a better name for the option.
2019-03-18 01:31:44 -07:00
Michael Gottesman
3ebd8df493 [gardening] Remove unnecessary -enable-sil-ownership from tests that now just get it from their pattern.
This just eliminates -enable-sil-ownership from all target-swift-frontend and
target-swift-emit-silgen RUN lines. Both of those now include
enable-sil-ownership in their expansion.
2019-03-12 20:39:18 -07:00
Vedant Kumar
a9da803033 [SILProfiler] Set up profiling for lazily-emitted functions (#22403)
A SIL function that's initially only emitted as a declaration may later
be prepared for definition. When this happens, set up a profiler for the
definition.

This makes code coverage visible for private methods (the frontend
follows a declare-then-define pattern for these).

rdar://47759243
2019-02-05 19:57:18 -08:00
Vedant Kumar
63c02ce073 [SILProfiler] Skip source regions with invalid locations
It's not possible to emit code coverage mappings for such regions.

rdar://47254122
2019-01-14 13:55:09 -08:00
Rintaro Ishizaki
22652f9e88 [Parse] Eliminate backtracking in collection expression parsing
Parsing collection literal expression used to take exponential time
depending on the nesting level of the first element.

Stop using 'parseList()' because using it complicates libSyntax parsing.

rdar://problem/45221238 / https://bugs.swift.org/browse/SR-9220
rdar://problem/38913395 / https://bugs.swift.org/browse/SR-7283
2018-12-25 11:05:23 +09:00
Michael Gottesman
fd4828e40a Eliminate -assume-parsing-unqualified-ownership-sil from tests.
I am doing this separately from the actual change to eliminate the option to
make it easier to review.
2018-12-19 12:54:13 -08:00
Michael Gottesman
40a09c9c21 Fixup tests for -assume-parsing-unqualified-ownership-sil => [ossa] transition. 2018-12-18 00:49:32 -08:00
Pavel Yaskevich
a1df238d47 Associate @autoclosure only with parameter types
Mangled implicit closures shouldn't have `@autoclosure` flag
associated with them, because their type is just a regular
function type.
2018-11-10 11:59:28 -08:00
Erik Eckstein
39bb14b094 change mangling prefix from $S to $s
This is the final ABI mangling prefix

rdar://problem/38471478
2018-09-19 13:55:11 -07:00
Jordan Rose
01a0de27ec [test] Update for remote-run-ing tests on a different macOS (#18966)
Most of this is just "remember to specify the inputs and outputs on
the command line, so remote-run can see them". A bit is "prefix
environment variables with '%env-'". And the last few are "yeah,
this was never going to work in a remote environment".

In the few cases where I couldn't think of anything reasonable, I just
marked the test as "UNSUPPORTED: remote_run", a new "feature".
2018-08-27 14:50:40 -07:00
Vedant Kumar
02aa4af6b2 Require macOS/x86_64 in a TSan IRGen test 2018-08-17 10:59:25 -07:00
Arnold Schwaighofer
df8cb665cf XFAIL test/Profiler/instrprof_tsan.swift
Fails on simulator with  unsupported option '-sanitize=thread' for target 'i386-apple-ios7.0'
rdar://43422035
2018-08-17 06:34:10 -07:00
Vedant Kumar
c76494b02e Use atomic profile counter updates when TSan is enabled
This suppresses TSan diagnostics about racy profile counter updates.

rdar://40477803
2018-08-16 16:23:25 -07:00
Vedant Kumar
6d467fd31f [SILProfiler] Don't crash when coverage info for lazy getters is incomplete (#18744)
ASTWalker visits a lazy_initializer_expr once within its associated
var_decl (by way of the parent nominal type). However, SILGen visits the
lazy_initializer_expr while inside of the var_decl's getter. The result
is that there is no coverage mapping information for the contents of the
lazy init within the getter's SILProfiler.

Fixing this will require reworking how profile counters are assigned to
be more in line with what SILGen needs.

As a stop-gap, this patch prevents SILGen from asserting that coverage
mappings are complete with a defensive check which prevents a crash seen
in SR-8429.

rdar://42792053
2018-08-15 17:48:21 -07:00
Slava Pestov
301d9f0cf3 AST: Give destructors a second parameter list
Constructors and methods had two parameter lists, one for self and one
for the formal parameters. Destructors only had one parameter list,
which introduced an annoying corner case.
2018-07-19 12:28:13 -07:00
Vedant Kumar
d3ea09fafc [Coverage] Fix mappings of regions following do-catch stmts
Coverage reporting for source regions following do-catch statements was
incorrect, because it did not properly reflect early exits.

rdar://34244637
2018-05-18 12:11:14 -07:00
Vedant Kumar
d71d3fb0f7 [Coverage] Test lowering of ill-formed SIL profiling intrinsics (#16607)
* [Coverage] Parse SIL coverage maps for top-level code decls

This adds SIL printer/parser support for SILCoverageMaps representing
top-level code decls.

* [Coverage] Test lowering of ill-formed SIL profiling intrinsics

This adds a test case to exercise a path in IRGen which discards
ill-formed profiling intrinsics.

rdar://40133800 & r://39146527
2018-05-15 11:16:31 -07:00
Vedant Kumar
c3f4d99e57 Group the the code coverage, profiling, and pgo tests together (#16070)
* Group tests for profiling instrumentation together, NFC

This will make it easier to test changes to the code coverage logic.

There are a handful of tests which relate to profiling which I have not
moved. These include tests for the driver and for the SIL optimizer. It
makes more sense to keep those tests where they are.

* Rename a test file, NFC

This file tests code coverage of primary files, so I've changed the name
of the file to reflect that.

* Simplify the check lines in a test, NFC

This file tests code coverage of closures. It had several check lines
which obscured the meaning of the test, and its check lines were in a
strange order.

Remove the extra checks and disable -emit-sorted-sil.
2018-04-20 15:15:52 -07:00