Commit Graph

751 Commits

Author SHA1 Message Date
NAKAMURA Takumi 0981f73473 clang/test/Modules/ModuleDebugInfo.cpp: Add -triple %itanium to appease ms-targeted builds.
I think DebugInfo tests may avoid MS stuff for now.

llvm-svn: 247093
2015-09-08 22:47:32 +00:00
Adrian Prantl 748a6cd1ac Module Debugging: Emit debug type information into clang ObjC modules.
When -fmodule-format is set to "obj", emit debug info for all types
declared in a module or referenced by a declaration into the module's
object file container.

This patch adds support for Objective-C types and methods.

llvm-svn: 247068
2015-09-08 20:41:52 +00:00
Adrian Prantl 4aa2b3a311 Module Debugging: Emit debug type information into clang modules.
When -fmodule-format is set to "obj", emit debug info for all types
declared in a module or referenced by a declaration into the module's
object file container.

This patch adds support for C and C++ types.

llvm-svn: 247049
2015-09-08 19:20:27 +00:00
Sean Silva 0df0391fb6 [modules] Tighten up this test a bit.
llvm-svn: 246702
2015-09-02 20:16:09 +00:00
Richard Smith d88a7f1a92 Re-commit r246497 (and dependent changes r246524 and r246521), reverted in
r246546, with a workaround for an MSVC 2013 miscompile and an MSVC 2015
rejects-valid.

Original commit message:

[modules] Rework serialized DeclContext lookup table management. Instead of
walking the loaded ModuleFiles looking for lookup tables for the context, store
them all in one place, and merge them together if we find we have too many
(currently, more than 4). If we do merge, include the merged form in our
serialized lookup table, so that downstream readers never need to look at our
imports' tables.

This gives a huge performance improvement to builds with very large numbers of
modules (in some cases, more than a 2x speedup was observed).

llvm-svn: 246582
2015-09-01 20:35:42 +00:00
Aaron Ballman 6a4a210126 Reverting r246497 (which requires also reverting r246524 and r246521 to avoid merge conflicts). It broke the build on MSVC 2015. It also broke an MSVC 2013 bot with testing issues.
llvm\tools\clang\lib\serialization\MultiOnDiskHashTable.h(117):
error C2065: 'Files': undeclared identifier

http://bb.pgr.jp/builders/ninja-clang-i686-msc18-R/builds/2917

llvm-svn: 246546
2015-09-01 13:24:39 +00:00
Richard Smith 7fa450c6ae [modules] Preserve DeclID order when merging lookup tables to give a more
predictable diagnostic experience. The hash-of-DeclID order we were using
before gave different results on Win32 due to a different predefined
declaration of __builtin_va_list.

llvm-svn: 246521
2015-09-01 01:37:34 +00:00
Richard Smith 9ce2b45936 [modules] Rework serialized DeclContext lookup table management. Instead of
walking the loaded ModuleFiles looking for lookup tables for the context, store
them all in one place, and merge them together if we find we have too many
(currently, more than 4). If we do merge, include the merged form in our
serialized lookup table, so that downstream readers never need to look at our
imports' tables.

This gives a huge performance improvement to builds with very large numbers of
modules (in some cases, more than a 2x speedup was observed).

llvm-svn: 246497
2015-08-31 22:17:11 +00:00
Richard Smith 79bf920552 [modules] Stop updating all identifiers when writing a module. This is
unnecessary in C++ modules (where we don't need the identifiers for their
Decls) and expensive.

llvm-svn: 245821
2015-08-24 03:33:22 +00:00
Richard Smith d8a83718f0 [modules] Rearrange how redeclaration chains are loaded, to remove a walk over
all modules and reduce the number of declarations we load when loading a
redeclaration chain.

The new approach is:
 * when loading the first declaration of an entity within a module file, we
   first load all declarations of the entity that were imported into that
   module file, and then load all the other declarations of that entity from
   that module file and build a suitable decl chain from them
 * when loading any other declaration of an entity, we first load the first
   declaration from the same module file

As before, we complete redecl chains through name lookup where necessary.

To make this work, I also had to change the way that template specializations
are stored -- it no longer suffices to track only canonical specializations; we
now emit all "first local" declarations when emitting a list of specializations
for a template.

On one testcase with several thousand imported module files, this reduces the
total runtime by 72%.

llvm-svn: 245779
2015-08-22 01:47:18 +00:00
Richard Smith 401cd25b23 [modules] Tests for r245390.
llvm-svn: 245391
2015-08-18 23:42:50 +00:00
Richard Smith 6dc8ae10aa [modules] When parsing the base specifiers of a parse-merged class, the current
context is the class itself but lookups should be performed starting with the
lookup parent of the class (class and base members don't shadow types from the
surrounding context because they have not been declared yet).

llvm-svn: 245236
2015-08-17 20:24:17 +00:00
Sean Silva 8b7c0398b6 [modules] PR20507: Avoid silent textual inclusion.
Summary:
If a module was unavailable (either a missing requirement on the module
being imported, or a missing file anywhere in the top-level module (and
not dominated by an unsatisfied `requires`)), we would silently treat
inclusions as textual. This would cause all manner of crazy and
confusing errors (and would also silently "work" sometimes, making the
problem difficult to track down).

I'm really not a fan of the `M->isAvailable(getLangOpts(), getTargetInfo(),
Requirement, MissingHeader)` function; it seems to do too many things at
once, but for now I've done things in a sort of awkward way.

The changes to test/Modules/Inputs/declare-use/module.map
were necessitated because the thing that was meant to be tested there
(introduced in r197805) was predicated on silently falling back to textual
inclusion, which we no longer do.

The changes to test/Modules/Inputs/macro-reexport/module.modulemap
are just an overlooked missing header that seems to have been missing since
this code was committed (r213922), which is now caught.

Reviewers: rsmith, benlangmuir, djasper

Subscribers: cfe-commits

Differential Revision: http://reviews.llvm.org/D10423

llvm-svn: 245228
2015-08-17 16:39:30 +00:00
Richard Smith e75ee0f0c8 [modules] When explicitly building a module file, don't include timestamps in
the produced pcm file for stable file creation across distributed build
systems.

llvm-svn: 245199
2015-08-17 07:13:32 +00:00
Richard Smith 3938f0c728 [modules] Stop dropping 'module.timestamp' files into the current directory
when building with implicit modules disabled.

llvm-svn: 245136
2015-08-15 00:34:15 +00:00
Reid Kleckner 397a17bfeb Don't run explicit-modules-missing-files.cpp on Windows
It is flaky due to inability to remove files with open handles.  We
could paper over it with rm -f, but then the file would still be
present.

This is more evidence to me that we should roll our own 'rm'
implementation in LLVM.

llvm-svn: 245083
2015-08-14 19:03:02 +00:00
Richard Smith fb1e7f7d1a [modules] Add an experimental -cc1 feature to embed the contents of an input
file in the .pcm files. This allows a smaller set of files to be sent to a
remote build worker when building with explicit modules (for instance, module
map files need not be sent along with the corresponding precompiled modules).

This doesn't actually make the embedded files visible to header search, so
it's not useful as a packaging format for public header files.

llvm-svn: 245028
2015-08-14 05:02:58 +00:00
Richard Smith f7b41371d9 [modules] When writing a module file built with -fmodule-map-file-home-is-cwd,
via a module map found by -fmodule-map-file=, the home directory of the module
is the current working directory, even if that's a different directory on
reload.

llvm-svn: 244988
2015-08-13 23:47:44 +00:00
Richard Smith be9b6c75c3 [modules] For explicit module file dependencies, only list direct dependency module files.
llvm-svn: 244931
2015-08-13 18:30:25 +00:00
Richard Smith 1dbb95151b Fix wrong expected test output in r244923.
llvm-svn: 244927
2015-08-13 18:11:20 +00:00
Richard Smith 216a3bd70d [modules] Change the way we deal with .d output for explicitly-specified module
files: include the .pcm file itself in the .d output, rather than including its
own input files. Other forms of module file continue to be transparent for .d
output.

Arguably, the input files for the .pcm file are still inputs to the
compilation, but that's unnecessary for make-like build systems (where the
mtime of the .pcm file is sufficient) and harmful for smarter build systems
that know about module files and want to track only the local dependencies.

llvm-svn: 244923
2015-08-13 17:57:10 +00:00
Ben Langmuir 7ff29148ac [Modules] Add Darwin-specific compatibility module map parsing hacks
This preserves backwards compatibility for two hacks in the Darwin
system module map files:

1. The use of 'requires excluded' to make headers non-modular, which
should really be mapped to 'textual' now that we have this feature.

2. Silently removes a bogus cplusplus requirement from IOKit.avc.

Once we start diagnosing missing requirements and headers on
auto-imports these would have broken compatibility with existing Darwin
SDKs.

llvm-svn: 244912
2015-08-13 17:13:33 +00:00
Richard Smith dd55b95fb2 [modules] Fix thread safety analysis to cope with merging of FieldDecls across modules.
llvm-svn: 244714
2015-08-12 02:17:52 +00:00
Richard Smith 9c9173dcc8 [modules] When instantiating the contents of an imported CXXRecordDecl, we can
emit lexical contents for a declaration for another module. Track which module
those contents came from, and ensure that we only grab the lexical contents
from a single such instantiation.

llvm-svn: 244682
2015-08-11 22:00:24 +00:00
Richard Smith d61c1f8280 [modules] Don't suggest deleting the module cache on an error if there is no module cache for this build.
llvm-svn: 244547
2015-08-11 00:32:42 +00:00
Richard Smith a24ff5545b [modules] Properly diagnose errors in module files for which we have no
corresponding include location (those specified on the command line).

llvm-svn: 244538
2015-08-11 00:05:21 +00:00
NAKAMURA Takumi 4b81164810 clang/test/Modules: Satisfy win32 hosts due to dospath issue.
llvm-svn: 244427
2015-08-09 22:55:50 +00:00
Richard Smith 0f99d6a441 [modules] PR22534: Load files specified by -fmodule-file= eagerly. In particular, this avoids the need to re-parse module map files when using such a module.
llvm-svn: 244416
2015-08-09 08:48:41 +00:00
Richard Smith 38c1e6d355 Unrevert r244412 (reverted in r244414), and delete the bogus line left behind
in the unit test that was checking a file the test no longer creates.

llvm-svn: 244415
2015-08-09 06:03:55 +00:00
Justin Bogner b102d1a432 Revert "[modules] Don't leak -M flags for dependency file generation into the module"
This was failing tests on a bunch of bots:

http://lab.llvm.org:8011/builders/clang-hexagon-elf/builds/29919/steps/check-all
http://lab.llvm.org:8011/builders/clang-x86_64-debian-fast/builds/29627/steps/check-all
http://lab.llvm.org:8080/green/job/clang-stage1-configure-RA_check/9959/
http://lab.llvm.org:8080/green/job/clang-stage2-configure-Rlto_check/5591/

This reverts r244412

llvm-svn: 244414
2015-08-09 05:40:38 +00:00
Richard Smith 2a6edb30d9 [modules] When building a dependency file, include module maps parsed in the
current compilation, not just those from imported modules.

llvm-svn: 244413
2015-08-09 04:46:57 +00:00
Richard Smith 2d7fbbc7be [modules] Don't leak -M flags for dependency file generation into the module
build process when we implicitly build a module. Previously, we'd create the
specified .d file once for each implicitly-built module and then finally
overwrite it with the correct contents after the requested build completes.
(This fails if you use stdout as a dependency file, which is what the provided
testcase does, and is how I discovered this brokenness.)

llvm-svn: 244412
2015-08-09 02:44:20 +00:00
Richard Smith d9a0e64f66 Add file missed from r244409.
llvm-svn: 244410
2015-08-09 01:50:14 +00:00
Richard Smith 2bb3c34ed4 [modules] When loading a template specialization, re-canonicalize its template
arguments because the reloaded form might have become non-canonical across the
serialization/deserialization step (this particularly happens when the
canonical form of the type involves an expression).

llvm-svn: 244409
2015-08-09 01:05:31 +00:00
Ben Langmuir 587aeafe2c Make 'arm' cover both 32 and 64 bit architecutres
... and add aarch32 to specifically refer to the 32-bit ones.

Previously, 'arm' meant only 32-bit architectures and there was no way
for a module to build with both 32 and 64 bit ARM architectures.

Now a module that is intended to work on both architectures can specify
    requires arm
whereas a module only for 32-bit platforms can say
    requires aarch32
and just like before, 64-bit only can say
    requires aarch64

llvm-svn: 244306
2015-08-07 01:59:56 +00:00
Richard Smith 8cbd895947 [modules] Make IndirectFieldDecl mergeable to avoid lookup ambiguity when the same anonymous union is defined across multiple modules.
llvm-svn: 243940
2015-08-04 02:05:09 +00:00
Argyrios Kyrtzidis c56419ed00 [modules] Fix issue where building a module from a relative path when -working-directory option is set, results in error.
The error was "module '<name>' was built in directory '<path>' but now resides in directory '<path>'
rdar://21330027

llvm-svn: 243718
2015-07-31 00:58:32 +00:00
Pete Cooper bfcf676737 Add fmodules-cache-path to 2 tests which were missing it.
These tests were creating a modules.idx file in my clang checkout, not
the build directory or temp.

All the other tests in this directory use -fmodules-cache-path=%t so
updated these to match.

llvm-svn: 243657
2015-07-30 16:44:07 +00:00
Sean Silva 5a45222745 Enable accidentally-disabled tests.
I think I was isolating the failure in the last RUN: line during
development, and forgot to re-enable the others.

llvm-svn: 243624
2015-07-30 07:23:17 +00:00
Sean Silva b963dedbe8 Avoid failure to canonicalize '..'.
Also fix completely broken and untested code which was hiding the
primary bug. The !LLVM_ON_UNIX branch of the ifdef was actually a no-op.

I ran into this in the wild. It was causing failures in our SDK build.

Ideally we'd have a perfect llvm::sys::fs::canonical, but at least this
is a step in the right direction, and fixes an obviously broken case.
In some sense the test case I've added here is an integration test. We
should have these routines thoroughly unit tested in llvm::sys::fs.

llvm-svn: 243597
2015-07-30 00:26:34 +00:00
Richard Smith 826711d456 [modules] When performing redeclaration lookup for a using declaration, prefer
UsingShadowDecls over other declarations of the same entity in the lookup
results. This ensures that we build correct redeclaration chains for the
UsingShadowDecls (otherwise we could see assertions and other misbehavior in
modules builds, when merging combines multiple redeclaration chains for the
same entity from the same module into one chain).

llvm-svn: 243592
2015-07-29 23:38:25 +00:00
Richard Smith a7c535b3c1 [modules] Change module manager visitation order to be a bit more stable when
more modules are added: visit modules depth-first rather than breadth-first.
The visitation is still (approximately) oldest-to-newest, and still guarantees
that a module is visited before anything it imports, so modules that are
imported by others sometimes need to jump to a later position in the visitation
order when more modules are loaded, but independent module trees don't
interfere with each other any more.

llvm-svn: 242863
2015-07-22 01:28:05 +00:00
Richard Smith a534a31c5e [modules] In C++, stop serializing and deserializing a list of declarations in
the identifier table. This is redundant, since the TU-scope lookups are also
serialized as part of the TU DeclContext, and wasteful in a number of ways. We
still emit the decls for PCH / preamble builds, since for those we want
identical results, not merely semantically equivalent ones.

llvm-svn: 242855
2015-07-21 23:54:07 +00:00
Richard Smith d520a250b6 [modules] Produce an error if -cc1 wants to implicitly build a module and no
module cache has been provided, rather than creating one in the current
directory.

llvm-svn: 242819
2015-07-21 18:07:47 +00:00
Richard Smith 842d61f50c Attempt to fix greendragon buildbot failures; apparently wc sometimes inserts a leading space into its output. Just check for a matching number anywhere in the output.
llvm-svn: 242661
2015-07-20 08:40:51 +00:00
Richard Smith 4440d6eae0 [modules] Add (already passing) test for template default argument merging when the template is declared in a namespace.
llvm-svn: 242653
2015-07-19 23:44:27 +00:00
Richard Smith 9c25418424 [modules] Don't save uninteresting identifiers, and don't consider identifiers
to be interesting just because they are the name of a builtin. Reduces the size
of an empty module by over 80% (~100KB).

llvm-svn: 242650
2015-07-19 21:41:12 +00:00
Adrian Prantl fb2398d0c4 Make the clang module container format selectable from the command line.
- introduces a new cc1 option -fmodule-format=[raw,obj]
  with 'raw' being the default
- supports arbitrary module container formats that libclang is agnostic to
- adds the format to the module hash to avoid collisions
- splits the old PCHContainerOperations into PCHContainerWriter and
  a PCHContainerReader.

Thanks to Richard Smith for reviewing this patch!

llvm-svn: 242499
2015-07-17 01:19:54 +00:00
David Majnemer 38a50c0a16 [Sema] Emit a better diagnostic when variable redeclarations disagree
We referred to all declaration in definitions in our diagnostic messages
which is can be inaccurate.  Instead, classify the declaration and emit
an appropriate diagnostic for the new declaration and an appropriate
note pointing to the old one.

This fixes PR24116.

llvm-svn: 242190
2015-07-14 20:08:49 +00:00
Richard Smith 8a53a0ddbd Add file missed from r242105.
llvm-svn: 242109
2015-07-14 02:40:33 +00:00