AST file more lazy, so that we don't eagerly load that information for
all known identifiers each time a new AST file is loaded. The eager
reloading made some sense in the context of precompiled headers, since
very few identifiers were defined before PCH load time. With modules,
however, a huge amount of code can get parsed before we see an
@import, so laziness becomes important here.
The approach taken to make this information lazy is fairly simple:
when we load a new AST file, we mark all of the existing identifiers
as being out-of-date. Whenever we want to access information that may
come from an AST (e.g., whether the identifier has a macro definition,
or what top-level declarations have that name), we check the
out-of-date bit and, if it's set, ask the AST reader to update the
IdentifierInfo from the AST files. The update is a merge, and we now
take care to merge declarations before/after imports with declarations
from multiple imports.
The results of this optimization are fairly dramatic. On a small
application that brings in 14 non-trivial modules, this takes modules
from being > 3x slower than a "perfect" PCH file down to 30% slower
for a full rebuild. A partial rebuild (where the PCH file or modules
can be re-used) is down to 7% slower. Making the PCH file just a
little imperfect (e.g., adding two smallish modules used by a bunch of
.m files that aren't in the PCH file) tips the scales in favor of the
modules approach, with 24% faster partial rebuilds.
This is just a first step; the lazy scheme could possibly be improved
by adding versioning, so we don't search into modules we already
searched. Moreover, we'll need similar lazy schemes for all of the
other lookup data structures, such as DeclContexts.
llvm-svn: 143100
and DefaultFunctionArrayLvalueConversion. To prevent
significant regression for should-this-be-a-call fixits,
and to repair some such regression from the introduction of
bound member placeholders, make those placeholder checks
try to build calls appropriately. Harden the build-a-call
logic while we're at it.
llvm-svn: 141738
from unfriendly (== not at all modularized) directories. This is
temporary, and it only affects module construction until I'll figured
out how to deal with system headers.
llvm-svn: 140159
than conversions of C pointers to ObjC pointers. In order to ensure that
we've caught every case, add asserts to CastExpr that strictly determine
which cast kind is used for which kind of bit cast.
llvm-svn: 139352
from the given source. -emit-module behaves similarly to -emit-pch,
except that Sema is somewhat more strict about the contents of
-emit-module. In the future, there are likely to be more interesting
differences.
llvm-svn: 138595
type over into the AST context, then make that declaration a
predefined declaration in the AST format. This ensures that different
AST files will at least agree on the (global) declaration ID for 'id',
and eliminates one of the "special" types in the AST file format.
llvm-svn: 137429
ASTContext with accessors/mutators. The only functional change is that
the AST writer won't bother writing the id/Class/SEL redefinition type
if it hasn't been explicitly set; previously, it ended up being
written as a synonym for the built-in id/Class/SEL.
llvm-svn: 137349
- Added LazyVector::erase() to support this use case.
- Factored out the LazyDecl-of-Decls to RecordData translation in
the ASTWriter. There is still a pile of code duplication here to
eliminate.
llvm-svn: 136270
contents are lazily loaded on demand from an external source (e.g., an
ExternalASTSource or ExternalSemaSource). The "loaded" entities are
kept separate from the "local" entities, so that the two can grow
independently.
Switch Sema::TentativeDefinitions from a normal vector that is eagerly
populated by the ASTReader into one of these LazyVectors, making the
ASTReader a bit more like me (i.e., lazy).
llvm-svn: 136262
FullSourceLoc::getInstantiationLoc to ...::getExpansionLoc. This is part
of the API and documentation update from 'instantiation' as the term for
macros to 'expansion'.
llvm-svn: 135914
Special detail is added for uninitialized variable analysis as this has
serious performance problems than need to be tracked.
Computing some of this data is expensive, for example walking the CFG to
determine its size. To avoid doing that unless the stats data is going
to be used, we thread a bit into the Sema object to track whether
detailed stats should be collected or not. This bit is used to avoid
computations whereever the computations are likely to be more expensive
than checking the state of the flag. Thus, counters are in some cases
unconditionally updated, but the more expensive (and less frequent)
aggregation steps are skipped.
With this patch, we're able to see that for 'gcc.c':
*** Analysis Based Warnings Stats:
232 functions analyzed (0 w/o CFGs).
7151 CFG blocks built.
30 average CFG blocks per function.
1167 max CFG blocks per function.
163 functions analyzed for uninitialiazed variables
640 variables analyzed.
3 average variables per function.
94 max variables per function.
96409 block visits.
591 average block visits per function.
61546 max block visits per function.
And for the reduced testcase in PR10183:
*** Analysis Based Warnings Stats:
98 functions analyzed (0 w/o CFGs).
8526 CFG blocks built.
87 average CFG blocks per function.
7277 max CFG blocks per function.
68 functions analyzed for uninitialiazed variables
1359 variables analyzed.
19 average variables per function.
1196 max variables per function.
2540494 block visits.
37360 average block visits per function.
2536495 max block visits per function.
That last number is the somewhat scary one that indicates the problem in
PR10183.
llvm-svn: 134494
vector<int>
to
std::vector<int>
Patch by Kaelyn Uhrain, with minor tweaks + PCH support from me. Fixes
PR5776/<rdar://problem/8652971>.
Thanks Kaelyn!
llvm-svn: 134007
Language-design credit goes to a lot of people, but I particularly want
to single out Blaine Garst and Patrick Beard for their contributions.
Compiler implementation credit goes to Argyrios, Doug, Fariborz, and myself,
in no particular order.
llvm-svn: 133103
We can use this to produce nice diagnostics (and try to fixit-and-recover) in
various cases where we might see "MyFunction" instead of "MyFunction()". The
changes in SemaExpr are an example of how to use isExprCallable.
llvm-svn: 130878
This is more efficient as it's all done at once at the end of the TU.
This could still get expensive, so a flag is provided to disable it. As
an added bonus, the diagnostics will now print out a cycle.
The PCH test is XFAILed because we currently can't deal with a note
emitted in the header and I, being tired, see no other way to verify the
serialization of delegating constructors. We should probably address
this problem /somehow/ but no good solution comes to mind.
llvm-svn: 130836
new templates that need to be instantiated and vice-versa. Iterate
until we've instantiated all required templates and defined all
required vtables. Fixed PR9325 / <rdar://problem/9055177>.
llvm-svn: 130023
gcc's unused warnings which don't get emitted if the function is referenced even in an unevaluated context
(e.g. in templates, sizeof, etc.). Also, saying that a function is 'unused' because it won't get codegen'ed
is somewhat misleading.
- Don't emit 'unused' warnings for functions that are referenced in any part of the user's code.
- A warning that an internal function/variable won't get emitted is useful though, so introduce
-Wunneeded-internal-declaration which will warn if a function/variable with internal linkage is not
"needed" ('used' from the codegen perspective), e.g:
static void foo() { }
template <int>
void bar() {
foo();
}
test.cpp:1:13: warning: function 'foo' is not needed and will not be emitted
static void foo() { }
^
Addresses rdar://8733476.
llvm-svn: 129794
This patch authored by Eric Niebler.
Many methods on the Sema class (e.g. ConvertPropertyForRValue) take Expr
pointers as in/out parameters (Expr *&). This is especially true for the
routines that apply implicit conversions to nodes in-place. This design is
workable only as long as those conversions cannot fail. If they are allowed
to fail, they need a way to report their failures. The typical way of doing
this in clang is to use an ExprResult, which has an extra bit to signal a
valid/invalid state. Returning ExprResult is de riguour elsewhere in the Sema
interface. We suggest changing the Expr *& parameters in the Sema interface
to ExprResult &. This increases interface consistency and maintainability.
This interface change is important for work supporting MS-style C++
properties. For reasons explained here
<http://lists.cs.uiuc.edu/pipermail/cfe-dev/2011-February/013180.html>,
seemingly trivial operations like rvalue/lvalue conversions that formerly
could not fail now can. (The reason is that given the semantics of the
feature, getter/setter method lookup cannot happen until the point of use, at
which point it may be found that the method does not exist, or it may have the
wrong type, or overload resolution may fail, or it may be inaccessible.)
llvm-svn: 129143
diagnostics that occur in unreachable code (e.g., -Warray-bound).
We only pay the cost of doing the reachability analysis when we issue one of these diagnostics.
llvm-svn: 126290
* Flag indicating 'we're parsing this auto typed variable's initializer' moved from VarDecl to Sema
* Temporary template parameter list for auto deduction is now allocated on the stack.
* Deduced 'auto' types are now uniqued.
llvm-svn: 126139
without defining them. This should be an error, but I'm paranoid about
"uses" that end up not actually requiring a definition. I'll revisit later.
Also, teach IR generation to not set internal linkage on variable
declarations, just for safety's sake. Doing so produces an invalid module
if the variable is not ultimately defined.
Also, fix several places in the test suite where we were using internal
functions without definitions.
llvm-svn: 126016
making them be template instantiated in a more normal way and
make them handle attributes like other decls.
This fixes the used/unused label handling stuff, making it use
the same infrastructure as other decls.
llvm-svn: 125771
LabelDecl and LabelStmt. There is a 1-1 correspondence between the
two, but this simplifies a bunch of code by itself. This is because
labels are the only place where we previously had references to random
other statements, causing grief for AST serialization and other stuff.
This does cause one regression (attr(unused) doesn't silence unused
label warnings) which I'll address next.
This does fix some minor bugs:
1. "The only valid attribute " diagnostic was capitalized.
2. Various diagnostics printed as ''labelname'' instead of 'labelname'
3. This reduces duplication of label checking between functions and blocks.
Review appreciated, particularly for the cindex and template bits.
llvm-svn: 125733
access-control diagnostics which arise from the portion of the declarator
following the scope specifier, just in case access is granted by
friending the individual method. This can also happen with in-line
member function declarations of class templates due to templated-scope
friend declarations.
We were really playing fast-and-loose before with this sort of thing,
and it turned out to work because *most* friend functions are in file
scope. Making us delay regardless of context exposed several bugs with
how we were manipulating delay. I ended up needing a concept of a
context that's independent of the declarations in which it appears,
and then I actually had to make some things save contexts correctly,
but delay should be much cleaner now.
I also encapsulated all the delayed-diagnostics machinery in a single
subobject of Sema; this is a pattern we might want to consider rolling
out to other components of Sema.
llvm-svn: 125485
access control errors into SFINAE errors, so that the trait provides
enough support to implement the C++0x std::is_convertible type trait.
To get there, the SFINAETrap now knows how to set up a SFINAE context
independent of any template instantiations or template argument
deduction steps, and (separately) can set a Sema flag to translate
access control errors into SFINAE errors. The latter can also be
useful if we decide that access control errors during template argument
deduction should cause substitution failure (rather than a hard error)
as has been proposed for C++0x.
llvm-svn: 124446
whose patterns are template arguments. We can now instantiate, e.g.,
typedef tuple<pair<OuterTypes, InnerTypes>...> type;
where OuterTypes and InnerTypes are template type parameter packs.
There is a horrible inefficiency in
TemplateArgumentLoc::getPackExpansionPattern(), where we need to
create copies of TypeLoc data because our interfaces traffic in
TypeSourceInfo pointers where they should traffic in TypeLocs
instead. I've isolated in efficiency in this one routine; once we
refactor our interfaces to traffic in TypeLocs, we can eliminate it.
llvm-svn: 122278
a translation unit to the ActOnEndOfTranslationUnit function instead of doing
it at the start of DefineUsedVTables. The latter is now called *recursively*
during template instantiation, which causes an absolutely insane number of
walks of every record decl in the translation unit.
After this patch, an extremely template instantiation heavy test case's compile
time drops by 10x, and we see between 15% and 20% improvement in average
compile times across a project. This is just recovering a regression, it
doesn't make anything faster than it was several weeks ago.
llvm-svn: 121644
a useful template instantiation stack. Fixes PR8640.
This also causes a slight change to where the "instantianted from" note shows up
in truly esoteric cases (see the change to test/SemaCXX/destructor.cpp), but
that isn't directly the fault of this patch.
llvm-svn: 120135
-Move the stuff of Diagnostic related to creating/querying diagnostic IDs into a new DiagnosticIDs class.
-DiagnosticIDs can be shared among multiple Diagnostics for multiple translation units.
-The rest of the state in Diagnostic object is considered related and tied to one translation unit.
-Have Diagnostic point to the SourceManager that is related with. Diagnostic can now accept just a
SourceLocation instead of a FullSourceLoc.
-Reflect the changes to various interfaces.
llvm-svn: 119730
that are suppressed during template argument deduction. This change
queues diagnostics computed during template argument deduction. Then,
if the resulting function template specialization or partial
specialization is chosen by overload resolution or partial ordering
(respectively), we will emit the queued diagnostics at that point.
This addresses most of PR6784. However, the check for unnamed/local
template arguments (which existed before this change) is still only
skin-deep, and needs to be extended to look deeper into types. It must
be improved to finish PR6784.
llvm-svn: 116373
typeid expressions:
- make sure we have a proper source location for the closing ')'
- cache the declaration of std::type_info once we've found it
llvm-svn: 113441
One who seeks the Tao unlearns something new every day.
Less and less remains until you arrive at non-action.
When you arrive at non-action,
nothing will be left undone.
llvm-svn: 112244
This works courtesy of the new SmallVector<..., 0> specialization that
doesn't require a complete type. Note that you'll need to pull at least
SmallVector.h from LLVM to compile successfully.
llvm-svn: 112114
token. The first token might be something that ends up triggering code
completion, which in turn requires a valid Scope. Test case forthcoming.
llvm-svn: 112066
This option is not part of the Unused diagnostic group until the warnings on llvm codebase are fixed
and we are ready to turn it on. Suggestion by Daniel.
llvm-svn: 111298
Unused warnings for functions:
-static functions
-functions in anonymous namespace
-class methods in anonymous namespace
-class method specializations in anonymous namespace
-function specializations in anonymous namespace
Unused warnings for variables:
-static variables
-variables in anonymous namespace
-static data members in anonymous namespace
-static data members specializations in anonymous namespace
Reveals lots of opportunities for dead code removal in llvm codebase that will
interest my esteemed colleagues.
llvm-svn: 111086
when the CXTranslationUnit_CacheCompletionResults option is given to
clang_parseTranslationUnit(). Essentially, we compute code-completion
results for macro definitions after we have parsed the file, then
store an ASTContext-agnostic version of those results (completion
string, cursor kind, priority, and active contexts) in the
ASTUnit. When performing code completion in that ASTUnit, we splice
the macro definition results into the results provided by the actual
code-completion (which has had macros turned off) before libclang gets
those results. We use completion context information to only splice in
those results that make sense for that context.
With a completion involving all of the macros from Cocoa.h and a few other
system libraries (totally ~8500 macro definitions) living in a
precompiled header, we get about a 9% performance improvement from
code completion, since we no longer have to deserialize all of the
macro definitions from the precompiled header.
Note that macro definitions are merely the canary; the cache is
designed to also support other top-level declarations, which should be
a bigger performance win. That optimization will be next.
Note also that there is no mechanism for determining when to throw
away the cache and recompute its contents.
llvm-svn: 111051
-static variables
-variables in anonymous namespace (fixes rdar://7794535)
-static data members in anonymous namespace
-static data members specializations in anonymous namespace
llvm-svn: 111027
-static function declarations
-functions in anonymous namespace
-class methods in anonymous namespace
-class method specializations in anonymous namespace
-function specializations in anonymous namespace
llvm-svn: 111026
used when parsing (or re-parsing) a file. Also, when loading a
precompiled header into ASTUnit, create a Sema object that holds onto
semantic-analysis information.
llvm-svn: 111003
John McCall