e90372a0 cmCommonTargetGenerator: Factor out Fortran module directory computation
70c21301 cmCommonTargetGenerator: Store working directory for relative paths
7371d8f3 cmCommonTargetGenerator: Return string from GetFortranModuleDirectory
613bc08a cmDependsFortran: Use string to store module directory
The Makefile generators run tools with the current working directory set
to the subdirectory of the build tree for the each target. The Ninja
generator runs tools with the current working directory set to the top
of the build tree. Tell cmCommonTargetGenerator where the working
directory will be so it can compute proper relative paths.
Factor a GetLinkedTargetDirectories method out of
cmMakefileTargetGenerator::WriteTargetDependRules to compute the list of
directories associated with targets to which the current target links.
Teach the Makefile and Ninja generators to substitute for an <INCLUDES>
placeholder instead of putting -I in <FLAGS>. Update our values for
CMAKE_<LANG>_COMPILE_OBJECT,
CMAKE_<LANG>_CREATE_ASSEMBLY_SOURCE, and
CMAKE_<LANG>_CREATE_PREPROCESSED_SOURCE
to place <INCLUDES> just before <FLAGS>.
Create a <LANG>_COMPILER_LAUNCHER target property (initialized by a
CMAKE_<LANG>_COMPILER_LAUNCHER variable) to specify a compiler launcher
tool. This will supersede the CMAKE_<LANG>_COMPILER_ARG1 approach to
using such tools. The old approach set CMAKE_<LANG>_COMPILER to the
launcher tool while the new approach leaves this variable set to the
actual compiler.
Implement this property for Makefile and Ninja generators. It cannot be
implemented for VS or Xcode generators as the IDE build tools offer no
such hooks.
The test for this variable was removed in commit v2.8.8~330^2~7 (complex:
Remove ancient unused ComplexRelativePaths test, 2011-12-23).
Commit v3.1.0-rc1~425^2~2 (backtrace: Convert to local paths in
IssueMessage, 2014-03-12) appears to have accidentally made some backtraces
print relative paths with the variable because conversions which used to be
done at configure time, before the variable had an effect are now potentially
done at generate time.
The documentation of the variable says not to use it, and the docs are wrong in
that the variable actually applies in per-directory scope.
The read of the variable makes it harder to split conversion methods from
cmLocalGenerator where they don't belong. Remove it now.
The 'ctest --launch' command must be placed before the IWYU launcher on
the compiler command line. Extend the RunCMake.IncludeWhatYouUse test
to cover this case. The Ninja generator already does it correctly.
Create a <LANG>_INCLUDE_WHAT_YOU_USE target property (initialized by a
CMAKE_<LANG>_INCLUDE_WHAT_YOU_USE variable) to specify an IWYU command
line to be run along with the compiler.
When given the command line
tool a\ b c
mingw32-make incorrectly passes "a b" and "c" to the tool. When given
the command line
tool a\ b "c"
mingw32-make correctly passes "a\", "b", and "c" to the tool.
Since commit v3.1.0-rc1~861^2 (MSVC: Add properties to configure
compiler PDB files, 2014-02-24) we pass the compiler pdb option to
MS-style compiler tools as "/Fd<dir>\" but mingw32-make may consume
the backslash as escaping a following space as described above.
Workaround this problem by changing the backslash to a forward
slash as had been used prior to the above commit.
9660a3cc Makefile: Fix multiple custom command outputs with one missing
5c08e255 KWSys SystemTools: Teach Touch with !create to succeed on missing file
The use of "cmake -E touch_nocreate" added in commit v3.2.1~4^2
(Makefile: Fix multiple custom command outputs regression, 2015-03-06)
caused builds to fail when one of the outputs is intentionally not
created. This was fixed by our parent commit by making touch_nocreate
succeed when the file is missing. Add a test case covering it.
For the Watcom WMake generator, check for the SYMBOLIC source file
property separately on each output. The mark is needed on outputs that
are not really created to tell 'wmake' not to complain that it is
missing. The mark is also needed on outputs that are created or 'wmake'
will not consider them out of date when they exist.
Inspired-by: Ben Boeckel <ben.boeckel@kitware.com>
Add .DELETE_ON_ERROR to the "build.make" files that contain the actual
build rules that generate files. This tells GNU make to delete the
output of a rule if the recipe modifies the output but returns failure.
This is particularly useful for custom commands that use shell
redirection to produce a file.
Do not add .DELETE_ON_ERROR for Borland or Watcom make tools because
they may not tolerate it and would not honor it anyway. Other make
tools that do not understand .DELETE_ON_ERROR will not be hurt.
Suggested-by: Andrey Vihrov <andrey.vihrov@gmail.com>
232a6883 Help: Add release notes for target-language-genex.
9e168941 File(GENERATE): Process genex evaluation files for each language.
b734fa44 Genex: Allow COMPILE_LANGUAGE when processing include directories.
0b945ea9 Genex: Allow COMPILE_LANGUAGE when processing compile definitions.
5c559f11 Genex: Enable use of COMPILE_LANGUAGE for compile options.
e387ce7d Genex: Add a COMPILE_LANGUAGE generator expression.
4a0128f4 VS6: Compute CMAKE_*_FLAGS and COMPILE_DEFINITIONS* only when needed
In commit v3.2.0-rc1~272^2~2 (Makefile: Fix rebuild with multiple custom
command outputs, 2014-12-05) we changed the generated makefile pattern
for multiple outputs from
out1: depends...
commands...
out2: out1
to
out1 out2: depends...
commands...
This was based on the incorrect assumption that make tools would treat
this as a combined output rule and run the command(s) exactly once for
them. It turns out that instead this new pattern is equivalent to
out1: depends...
commands...
out2: depends...
commands...
so the commands may be run more than once.
Some documents suggest using a "dedicated witness" stamp file:
stamp: depends...
rm -f stamp
touch stamp.tmp
commands...
mv stamp.tmp stamp
out1 out2: stamp
However, if the commands fail the error message will refer to the stamp
instead of any of the real outputs, which may be confusing to readers.
Also, this approach seems to have the same behavior of the original
approach that motiviated the above commit: multiple invocations are
needed to bring consumers of the outputs up to date.
Instead we can return to the original approach but add an explicit
touch to each extra output rule:
out1: depends...
commands...
out2: out1
touch -c out2
This causes make tools to recognize that all outputs have changed and
therefore to execute any commands that consume them.
Replace use of separate "cmake -E cmake_progress_report" and "cmake -E
cmake_echo_color" commands to report the progress and message portions
of build output lines with --progress-* options to the latter to print
everything with a single command. The line buffering of the stdout FILE
stream should cause the whole line to be printed with one atomic write.
This will avoid inter-mixing of line-wise messages from different
processes during a parallel build.
Fix the generated makefiles for custom commands with multiple outputs to
list all the outputs on the left hand side of the build rule. This is
much simpler and more reliable than the old multiple-output-pair
infrastructure.
Since commit v3.1.0-rc1~821^2 (Windows: Use response files to specify
link libraries for GNU tools, 2014-03-04) we use a response file to pass
possibly long linker flag lists to the GNU linker on Windows. On MinGW,
this may cause gfortran to use a response file to pass some flags to its
own internal invocation. This is okay except when we are parsing
implicit link flags from the compiler ABI detection build. If gfortran
uses a response file in that case then we may miss extracting some of
the implicit link flags, such as -lgfortran. Fortunately, in the
compiler ABI detection case we do not actually link to anything so the
response file is empty. Work around this problem by simply not using a
response file when the list of flags it is used to pass is empty (or
just whitespace).
Reported-by: Bill Somerville <bill@classdesign.com>
The checks are now split into languages that are able to generate
assembly listings, languages that are able to generate preprocessed
listings, and languages that are able to export the compile commands.
Changes in commit b9aa5041 (cmLocalGenerator: Simplify GetIncludeFlags
output formatting, 2014-03-04) caused Windows Resource Compiler include
directories to be computed as relative paths in the Ninja generator.
This breaks the cmcldeps handling of include paths. The reason for the
regression is that several cmLocalGenerator::GetIncludeFlags callers
treated the fourth "bool forResponseFile" argument as if it controlled
whether include directories were a full path. It actually did control
that by accident until the above commit.
Add an explicit "bool forceFullPaths" argument to GetIncludeFlags
and thread the value through ConvertToIncludeReference as needed.
Update GetIncludeFlags call sites that really wanted to control the
forResponseFile setting to be aware of the new argument. Extend the
VSResource test to cover this case.
Teach the Makefile generators to escape '#' characters on the right hand
side of variable assignments in flags.make. This is needed for flags
like '-Wno-error=#warnings'. Otherwise the make tool treats them as
comments and leaves them out of the _FLAGS variable value.
Add a case to the CompileOptions test covering '#' in a COMPILE_OPTIONS
value, at least on compilers where it is known to be supported.
86be733f cmGeneratorExpression: Add workaround for Borland compiler
3495ab0a tests: update unused variable test expected output
2a1b2d84 backtrace: Convert to local paths in IssueMessage
a0829205 genex: remove the need for backtraces
efc20569 cmake: remove dummy backtraces for IssueMessage
d46c650d cmMakefile: return a backtrace
Some compilers do not offer an option to specify the path to the object
file, but rather only to the directory in which to place the object
file. See issue 14876 for some examples. Add a new OBJECT_FILE_DIR
placeholder to specify the directory containing the object file for the
current compilation. This may differ from the main target OBJECT_DIR
when the object corresponds to a source in a subdirectory.
9407174b target_sources: New command to add sources to target.
81ad69e0 Make the SOURCES target property writable.
6e636f2e cmTarget: Make the SOURCES origin tracable.
3676fb49 cmTarget: Allow transitive evaluation of SOURCES property.
e6971df6 cmTarget: Make the source files depend on the config.
df753df9 cmGeneratorTarget: Don't add computed sources to the target.
869328aa cmComputeTargetDepends: Use valid config to compute target depends.
5de63265 Genex: Only evaluate TARGET_OBJECTS to determine target sources.
aa0a3562 cmGeneratorTarget: Compute target objects on demand
042c1c83 cmTarget: Compute languages from object libraries on demand.
fdcefe3c cmGeneratorTarget: Compute consumed object libraries on demand.
c355d108 cmComputeTargetDepends: Track object library depends.
e5da9e51 cmTarget: Allow any generator expression in SOURCES property.
5702e106 cmTarget: Include TARGET_OBJECTS genex in target SOURCES property.
857d30b5 cmGlobalGenerator: Add interface to call ForceLinkerLanguages
28e1d2f8 cmStringCommand: Add GENEX_STRIP subcommand.
bf98cc25 Genex: Evaluate TARGET_OBJECTS as a normal expression.
8cd113ad cmTarget: Store strings instead of cmSourceFile* to represent SOURCES.
4959f341 cmSourceFileLocation: Collapse full path for directory comparisons.
fcc92878 cmSourceFileLocation: Remove unused Update method.
59e8740a cmTarget: Remove AddSourceFile method
26d494ba cmTarget: Use string API to add sources to cmTarget objects.
d38423ec cmTarget: Add a method to obtain list of filenames for sources.
...
Disallow the use of config-specific source files with
the Visual Studio and Xcode generators. They don't have
any way to represent the condition currently.
Use the same common-config API in cmQtAutoGenerators. While
it accepts config-specific files, it doesn't have to support
multiple configurations yet.
Loop over the configs in cmTargetTraceDependencies
and cmGlobalGenerator::WriteSummary and consume all source
files.
Loop over the configs in cmComputeTargetDepends and compute the
object library dependencies for each config.
Remove use of UseObjectLibraries from Makefile and Ninja generators. It
is not needed now because those generators use GetExternalObjects
which already contains the objects from object libraries.
The VS10 generator calls both the UseObjectLibraries and the GetExternalObjects
methods. Ensure that duplicates are not created by skipping objects
from object libraries in handling of GetExternalObjects.
Similarly, fix VS6, VS7 and Xcode object handling by skipping
external objects from OBJECT_LIBRARY usage as appropriate.
The error message in the BadSourceExpression1 test is now reported
by the generator expression evaluator, so it has different text.
Drop the CMAKE_NO_QUOTED_OBJECTS internal variable from the Makefile
generators. The underlying problem is with the Watcom linker, not with
WMake. The Watcom linker wants object files to be single-quoted. Add
<LINK-RULE>_USE_WATCOM_QUOTE platform information variables to tell the
generators to use Watcom-style single quotes for object files on link
lines.
On Windows, Watcom uses the GetCommandLine API to get the original
command-line string and do custom parsing that expects single quotes.
On POSIX systems, Watcom approximates the original command line by
joining all argv[] entries separated by a single space. Therefore we
need to double-quote the single-quoted arguments so that the shell does
not consume them and they are available for the parser to see.
Until now the cmCustomCommandGenerator was used only to compute the
command lines of a custom command. Generalize it to get the comment,
working directory, dependencies, and outputs of custom commands. Update
use in all generators to support this.
Use the clang RemoveCStrCalls tool to automatically migrate the
code. This was only run on linux, so does not have any positive or
negative effect on other platforms.
Casts from std::string -> cmStdString were high on the list of things
taking up time. Avoid such implicit casts across function calls by just
using std::string everywhere.
The comment that the symbol name is too long is no longer relevant since
modern debuggers alias the templates anyways and the size is a
non-issue since the underlying methods are generated since it's
inherited.
Work around the command-line-length limit by using an @linklibs.rsp
response file to pass the flags for link libraries. This allows
very long lists of libraries to be used in addition to the existing
support for passing object files via response file.
Suggested-by: Peter Keuschnigg <peter.keuschnigg@pmu.ac.at>
Stephen Kelly
Brad King
Bill Hoffman
Tim Gallagher
Nils Gladitz
Ben Boeckel
Jiri Malak