QFileInfo.isRelative() deems any path starting with a slash to be
absolute; on MS-Win, such paths need a drive specifier (unless they're
UNC), so use IoUtils's more robust test for absolute paths.
Change-Id: I7d0872a87833cbf1cc1a6ef107941adc4c529624
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
For Q_OS_WIN, a path is only truly absolute if it includes a drive
letter; merely starting with a slash is not enough. (We can't support
UNC paths, so don't even try: qmake runs various commands in the
source directory using CMD.exe, which doesn't support UNC as PWD.)
This requires, when resolving a path relative to a root, transcribing
the root's drive to such not-quite-absolute paths.
Changed QMakeGlobals, $$absolute_path() and $$relative_path() to now
use IoUtils::resolvePath() rather than delegating to QDir's absolute
path method, since that doesn't correctly recognize the need for a
drive letter (and qmake did run into problems with some paths, from
splitPathList and a failing test, as a result).
Moved existing ioUtils tests for handling of relative / absolute paths
out into separate functions and expanded significantly. Fixed some
existing tests to use an absolute path where one is needed; added two
tests involving driveless (but rooted) paths; and fixed the test init
to set a value for QT_HOST_DATA/src property (the lack of which lead
to an assertion failure with this fix).
Task-number: QTBUG-50839
Change-Id: I2bfc13c1bfbe1ae09997274622ea55cb3de31b43
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Since Xcode 6.3, this must be set to NO because stripping on copy is no
longer fully supported due to the potential of input binaries being code
signed. In this case Xcode will simply ignore the strip step and issue
a warning since stripping would invalidate the code signature. This
change silences that annoying warning for release builds. Also, the
setting assignment is moved from being hardcoded in the generator, to
a QMAKE_MAC_XCODE_SETTINGS value.
Change-Id: If25511edddc12b7b0407e2992d80884b7d6437dc
Reviewed-by: Gabriel de Dietrich <gabriel.dedietrich@qt.io>
This problem does not affect the Xcode generator.
Task-number: QTBUG-65477
Change-Id: I6194edc5b679edad9ae1a25e35b71e5df9bd4c95
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
This was missed in 8bebded9.
Task-number: QTBUG-63637
Change-Id: I6be472430a9aa8f533def4fd6c14c8dbfe8b6f70
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
In Xcode, the default value for GCC_GENERATE_DEBUGGING_SYMBOLS is YES,
which causes Xcode to emit debug symbol bundles (.dSYM) on macOS and iOS
*if* DEBUG_INFORMATION_FORMAT is also set to dwarf-with-dsym. Since that
setting is already set to an appropriate value with debug vs release
builds, the default Xcode value for GCC_GENERATE_DEBUGGING_SYMBOLS is
already correct and in effect the only thing qmake was doing was always
setting GCC_GENERATE_DEBUGGING_SYMBOLS to a wrong value for release
builds - it should be YES in all cases, to allow the .dSYM bundles to
be generated in release mode, which is in fact the only case where
they're really needed in the first place.
Task-number: QTBUG-41246
Task-number: QTBUG-50896
Change-Id: I07639a3c4ff9f62d591cde3ad66748767d475e3b
Reviewed-by: Tor Arne Vestbø <tor.arne.vestbo@qt.io>
This is C++, not qmake code. Amends 5fa6438633.
Change-Id: Ie5b88c3a06dbe089948488ea3b4b297a08164113
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
SYSTEM is used for system() calls, while SHELL is used in the target
Makefiles.
Task-number: QTBUG-62985
Change-Id: Ia75d3939c59c98699359421166433e8b4a6ee35e
Reviewed-by: Frederik Gladhorn <frederik.gladhorn@qt.io>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
QMAKESPEC is added in makefile generator,
it is not in INCLUDEPATH.
Change-Id: I2451b3c7b30bc237157e68e5ce9de67f55e784b2
Reviewed-by: Leena Miettinen <riitta-leena.miettinen@qt.io>
When the source file is read-only then it would copy the file with that
attribute set when it is installed. However this will cause a problem
if it is installed a second time. Therefore the read-only attribute
needs to be manually reset before installing and again before touching
the file. Once the process is done then it is set back to be read-only
to preserve the state of the original.
Change-Id: I1c01f418ef3c9bd434acd2c2b8ee695544d7bb35
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
the replacement value may well constitute the whole output string - this
is in fact common, given this rather typical usage pattern:
BAR = $$replace(FOO, -flag, -otherflag)
this must be considered when constructing the return value.
compare 3c8134958c.
as of now, this is irrelevant, as QString::replace(QRegExp, QString) will
always memcpy the replacement into a detached copy of the target, but one
never knows.
Change-Id: Ia1f271f45023746040fc28ce6d88a6609e05e5c2
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
technically, we should not rely on the i/o classes not storing the
strings beyond the instantiated object's life time.
Change-Id: I0990769b3cf86860184869036c096c531160e9be
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
property values are de-facto guaranteed to be backed by full QStrings,
so there is nothing to be gained from using the raw data optimization,
while doing so risks raw data leaks.
Change-Id: I3d43da9aaadd4d5811c4b1a9d7ac734049da423c
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
it's not clear why detaching would be necessary; there is no danger of a
raw data leak here.
concatenating a QStringRef with a non-empty QLatin1String (the only
expected use of this overload) will yield a detached QString anyway, so
this makes little difference in practice.
amends f137957e08.
Change-Id: I521c0e89a8b0c1ae62b1450e81b0ae91a931bcfa
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
no m_tmp is involved any more in this code path; it uses QStringRef.
amends 11d957d043.
Change-Id: Ib272d61edfb150a549c5e6a9a60d53502702e802
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
the m_tmp array is a member, so the index toggle for accessing it also
needs to be one - otherwise, odd iteration counts will defeat the
mechanism.
Change-Id: If7a800ed5a4b4168625daf1ebbd5d2d164569d8e
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
... in $$basename(), $$dirname(), and contains(). the latter case is
marginal, as it only applies to mutuals which are regexes, which i don't
remember ever seeing used.
QRegExp saves a copy of the matched string, so it's necessary to
alternate between two temporaries to avoid detaching. we already
did that in most places.
Change-Id: I97b8294585c17c76d1756f83971f42cb88353af0
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
in most cases, the main advantage is not using toQString(m_tmp), which
reduces the possibility of raw data leaks. in cases where we used
toQString() without temporary, this is a slight optimization.
Change-Id: Ib343acffd383aa2c4fefab75fb52762fb534dfc6
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
A macro name ending in R might expand to a string; if this precedes a
string constant, we're juxtaposing the strings. My first parser for
raw strings would mistake it for a raw string instead, ignoring the
part of the identifier before R. Re-worked the exploration of what
came before the string to catch these cases, too.
The backwards parsing would also allow any messy jumble of [RLUu8]* as
prefix for the string; but in fact R must (if present) be last in the
prefix and *it* can have at most one prefix, [LUu] or u8. Anything
else is an identifier that happens to precede the string. Reworked
the parsing to allow only one prefix and not treat R specially unless
it's immediately (modulo BSNL) before the string's open-quotes.
Add link to the cppreference page about string literals, on which the
grammar now parsed is based.
Added a test for the issue this addresses.
Verified that this fails on 5.6, dev and 5.9 without the fix.
Expanded the existing test to cover R-with-prefix cases.
Task-number: QTBUG-55633
Change-Id: I541486c2ec909cfb42050907c84bee83ead4a2f4
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
The $${TARGET}_manifest.bak file was not removed on "nmake clean".
Task-number: QTBUG-59827
Change-Id: Ia5b636f4917f3e7a2df8d753824b72e63d278005
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
QMAKE_LINK_OBJECT_MAX is actually a property of the host, not the
target.
this works around binutil's inability to use thin LTO objects in
conjunction with an MRI script
(https://sourceware.org/bugzilla/show_bug.cgi?id=21702).
Task-number: QTBUG-61335
Change-Id: I90a1334b9c905c433b35546e8f3f3b5089d2c65b
Reviewed-by: Jake Petroules <jake.petroules@qt.io>
The strings remember in which file they were created/assigned.
However, this used a non-counting reference to a ProFile, which could
become dangling. If a subsequent ProFile re-used the exact same address,
a string's source would be mis-identified, which would be fatal in
conjunction with discard_from().
Since we actually need only a unique id for comparison, let's use an
integer for that.
Task-number: QTBUG-62434
Started-by: Simon Hausmann <simon.hausmann@qt.io>
Change-Id: I395153afaf7c835d0119690ee7f4b915e6f90d4a
Reviewed-by: Simon Hausmann <simon.hausmann@qt.io>
Task-number: QTBUG-59301
Change-Id: I2562b862465a52ecc56f551bcdb98fa7279ebfcf
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Reviewed-by: Gabriel de Dietrich <gabriel.dedietrich@qt.io>
If these rules are not added to the solution, Visual Studio will
complain, that the project has to be deployed before it can be run.
Change-Id: I6d3fbc949c85b11a92f78e13e2f6a1b92a5cfdc7
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
The old docs contained obsolete information. Also, the new docs
match the new docs for the other Apple OS deployment targets.
Change-Id: Id773fa2086f291d8a2552fe1b339ec1e13c19d74
Reviewed-by: Jake Petroules <jake.petroules@qt.io>
The following variables were not documented:
- QMAKE_IOS_DEPLOYMENT_TARGET
- QMAKE_TVOS_DEPLOYMENT_TARGET
- QMAKE_WATCHOS_DEPLOYMENT_TARGET
Change-Id: I5cfb6c0024d92e943aed882fd01bc2a4f2c7c042
Reviewed-by: Jake Petroules <jake.petroules@qt.io>
If several Makefiles are used in the same directory (for example, for
multiple projects in the same directory or different build configurations),
they all reference the same object_script, which is obviously wrong.
Change-Id: I9b499ceb6b6bd6058f54b452fa44bfb2313eec26
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
When generating a project, the directories can be specified as arguments
to the qmake call. As a result files can either be incorrectly added to
the project with a leading slash, or can end up duplicated. By passing
the absolute path with the file, it ensures that the file is added
correctly and no duplicates occur as a result.
Task-number: QTBUG-48342
Change-Id: If774de8d7f5cceca80042a25a3aa4e5b045249da
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
This was already done on unix, but not for MinGW.
If the archive already exists, it is appended rather than replaced.
This can cause invalid references when whole-archive linking is used
and some object file that was already linked was deleted.
Change-Id: Ie265371f197d996d57002b248043736544ee641e
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
adding shared install paths via QMAKE_LFLAGS in the spec has the tiny
side effect that they are searched _first_, which is generally a really
bad idea - they should be _last_.
for that purpose, introduce QMAKE_RPATHLINKDIR_POST, and migrate all
specs to use it.
QMAKE_RPATHDIR_POST is added for consistency, but not actually used.
Task-number: QTBUG-59457
Change-Id: Iac6cda5e9111ef8cca454a69861fe8408bb40589
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
adding shared install paths to QMAKE_{INCDIR,LIBDIR} in the spec has the
tiny side effect that they are searched _first_, which is generally a
really bad idea - they should be _last_.
for that purpose, make QMAKE_{INCDIR,LIBDIR}_POST live up to their names
(i.e., search them actually last) and migrate all affected specs to use
them.
Task-number: QTBUG-40825
Change-Id: Ie0de81c3cc49e193186d2fedd7d6c77590c8ef79
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Every /Zc:XXX option qmake doesn't know about yields a "WARNING: Could
not parse Compiler option '-Zc:XXX'; added to AdditionalOptions."
Put all /Zc:XXX options we don't handle into AdditionalOptions without
printing a warning. There's no point in making all options known to
qmake and updating them for every MSVC release.
Change-Id: I319e027791a7b0a29d139ee3074ab1aed8ce8a63
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
As the directory installation command also works with files as a source
we can unify the external commands, resulting in simpler command lines.
Change-Id: I65013626eedbdb3ce1c77ed230d46edd1603b986
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
The same as for other visual studio versions use "x86" as arch instead
of win32. arch is used to determine library paths and these use x86 and
not win32.
As compilerArch is not used in MSVC 2017 it can be removed.
Task-number: QTBUG-60530
Change-Id: I47157eb1d7ae9d913461210d34858ffb37c81586
Reviewed-by: Maurice Kalinowski <maurice.kalinowski@qt.io>
Similar to the two parent commits, this patchs preserves the time stamps
of files we install as a result of recursive directory copying.
Change-Id: Id5931a467196d5cd67acfa0deffc2488af8a3669
Task-number: QTBUG-59004
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
The result of "make install" should be the same regardless of whether it
has been run multiple times and the destination exists already. This is
done by making the file installation calls always take canonical source
and target paths and not look at the target directory.
Task-number: QTBUG-60370
Change-Id: I83a584c0dbc4fd10c79976d4169bf6bc051884a1
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
The original commit only added support for GCC and Clang, but not ICC.
Amends 73331eeb
Change-Id: Id7638cf1b538edb1008fb3aa10754c1f517a994f
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
Similar to the parent commit, this patch adds a unified code path in
qmake itself for installing program files while preserving their
original last modification timestamp.
Change-Id: I7b7dcfa6228c2bfd48ea6036549398bb6f90032f
Task-number: QTBUG-59004
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
On non-windows platforms, we use the "-p" parameter of install(1) to
preserve the last modification timestamps of files. On Windows the use
of copy does not preserve them. As a cross-platform solution, this patch
introduces a simple built-in install command in qmake to copy files.
Task-number: QTBUG-59004
Change-Id: I3064d29a2b8c7b009a1efbf8f00b84c079ea5417
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
projects using a dynamic build are not supposed to access this variable
anyway.
Task-number: QTBUG-51598
Change-Id: I81b55ea9ba460b80919f40ed7fe3d52129636b9e
Reviewed-by: Michal Klocek <michal.klocek@qt.io>
Reviewed-by: Jonathan Liu <net147@gmail.com>
When having Qt sources in a symbolic link "shadowed" did not work
because _PRO_FILE_PWD_ used canonical path, while source_root did not.
Due to this mix it was possible that shadowedPath did not find any
"common denominator" and always returned and empty string. The first
place where things broke was while running config.tests.
Task-number: QTBUG-59319
Change-Id: If73ecbc58290ee9113f887a73c312ebfb5e20a33
Reviewed-by: Joerg Bornemann <joerg.bornemann@qt.io>
Reviewed-by: Oswald Buddenhagen <oswald.buddenhagen@qt.io>
