...where "inline" (in its meaning of "this function can be defined in multiple
translation units") thus doesn't make much sense. (As discussed in
compilerplugins/clang/redundantinline.cxx, exempt such "static inline" functions
in include files for now.)
All the rewriting has been done automatically by the plugin, except for one
instance in sw/source/ui/frmdlg/column.cxx that used to involve an #if), plus
some subsequent solenv/clang-format/reformat-formatted-files.
Change-Id: Ib8b996b651aeafc03bbdc8890faa05ed50517224
Reviewed-on: https://gerrit.libreoffice.org/61573
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
...warning about (for now only) functions and variables with external linkage
that likely don't need it.
The problems with moving entities into unnamed namespacs and breaking ADL
(as alluded to in comments in compilerplugins/clang/external.cxx) are
illustrated by the fact that while
struct S1 { int f() { return 0; } };
int f(S1 s) { return s.f(); }
namespace N {
struct S2: S1 { int f() { return 1; } };
int f(S2 s) { return s.f(); }
}
int main() { return f(N::S2()); }
returns 1, both moving just the struct S2 into an nunnamed namespace,
struct S1 { int f() { return 0; } };
int f(S1 s) { return s.f(); }
namespace N {
namespace { struct S2: S1 { int f() { return 1; } }; }
int f(S2 s) { return s.f(); }
}
int main() { return f(N::S2()); }
as well as moving just the function f overload into an unnamed namespace,
struct S1 { int f() { return 0; } };
int f(S1 s) { return s.f(); }
namespace N {
struct S2: S1 { int f() { return 1; } };
namespace { int f(S2 s) { return s.f(); } }
}
int main() { return f(N::S2()); }
would each change the program to return 0 instead.
Change-Id: I4d09f7ac5e8f9bcd6e6bde4712608444b642265c
Reviewed-on: https://gerrit.libreoffice.org/60539
Tested-by: Jenkins
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
It seems that on iOS, where we don't have any Java, Python, BASIC, or
other scripting, the only thing that would use the C++/UNO bridge
functionality that invokes codeSnippet() was cppu::throwException().
codeSnippet() is part of what corresponds to the code that uses
run-time-generated machine code on other platforms. We can't generate
code at run-time on iOS, that has been known forever. Instead we have
used some manually written assembler to handle it instead. We used to
have a Perl script to generate a set of code snippets for different
cases, different numbers of parameters of the called function and
whatnot, but that went away at some stage some year ago. (It is
unclear whether that broke the C++/UNO bridge on iOS, or whether the
stuff continued to work even after that.)
Anyway, this handwritten assembly, or the manual construction of
internal data structures for exceptions, or something else, seemed to
have bit-rotten. Exceptions thrown with cppu::throwException() were
not catchable properly any longer.
Instead of digging in and trying to understand what is wrong, I chose
another solution. It turns out that the number of types of exception
objects thrown by cppu::throwException() is fairly small. During
startup of the LibreOffice code, and loading of an .odt document, only
one kind of exception is thrown this way... (The lovely
css::ucb:InteractiveAugmentedIOException.)
So we can simply have code that checks what the type of object being
thrown is, and explicitgly throws such an object then with a normal
C++ throw statement. Seems to work.
Sadly the cppu::getCaughtException() API still needs some inline
assembly in the C++/UNO brige. That seems to work though, knock on
wood.
This commit also adds a small "unit test" for iOS, copied from
cppuhelperm to ImplSVMain(). Ideally we should not copy code around of
course, but have a separate unit test app for iOS that would somehow
include relevant unit tests from source files all over the place.
Later.
Change-Id: Ib6d9d5b6fb8cc684ec15c97a312ca2f720e87069
Reviewed-on: https://gerrit.libreoffice.org/60506
Tested-by: Jenkins
Reviewed-by: Tor Lillqvist <tml@collabora.com>
Found with bin/find-unneeded-includes
Only removal proposals are dealt with here.
Change-Id: Ib420e9216b8313f5ed7634ec375e39ceb741fd45
Reviewed-on: https://gerrit.libreoffice.org/59297
Tested-by: Jenkins
Reviewed-by: Miklos Vajna <vmiklos@collabora.co.uk>
where used directly, since rtl_allocateMemory now just calls into std::malloc
Change-Id: I59f85bdb7efdf6baa30e8fcd2370c0f8e9c999ad
Reviewed-on: https://gerrit.libreoffice.org/59685
Tested-by: Jenkins
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
rtl/string.hxx and rtl/ustring.hxx both unnecessarily #include <sal/log.hxx> (and don't make use of it themselves), but many other files happen to depend on it.
This is a continuation of commit 6ff2d84ade to be able to remove those unneeded includes.
This commit adds missing headers to every file found by:
grep -FwL sal/log.hxx $(git grep -Elw 'SAL_INFO|SAL_INFO_IF|SAL_WARN|SAL_WARN_IF|SAL_DETAIL_LOG_STREAM|SAL_WHERE|SAL_STREAM|SAL_DEBUG')
to directories from connectivity to cui
Change-Id: I9903c10d0a04bbeb93d0f776d1d252b152459499
Reviewed-on: https://gerrit.libreoffice.org/57408
Tested-by: Jenkins
Reviewed-by: Miklos Vajna <vmiklos@collabora.co.uk>
(*) if we are already throwing a Wrapped*Exception, get the
exception using cppu::getCaughtexception.
(*) when catching and then immediately throwing UNO exceptions,
use cppu::getCaughtException to prevent exception slicing
(*) if we are going to catch an exception and then
immediately throw a RuntimeException, rather throw a
WrappedTargetRuntimeException and preserve the original exception information.
Change-Id: Ia7a501a50ae0e6f4d05186333c8517fdcb17d558
Reviewed-on: https://gerrit.libreoffice.org/54692
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Noel Grandin <noel.grandin@collabora.co.uk>
auto-rewrite with <https://gerrit.libreoffice.org/#/c/47798/> "Enable
loplugin:cstylecast for some more cases" plus
solenv/clang-format/reformat-formatted-files
Change-Id: I3966d302241a52014aaad41f72924a2095ea6621
This reverts commit ab07f81d0b.
> Jan 12 10:21:35 <sberg> tml_, is ab07f81d0b
> still relevant? (one could imagine that if one wants such debug output during
> development in a non-debug build, one could build with --enable-sal-log?)
> Jan 12 10:22:01 <tml_> I have no idea what's relevant on iOS any longer
> Jan 12 10:22:26 <tml_> ... but my gut feeling would be no, can be removed
> Jan 12 10:23:14 <tml_> the rationale in the commit message sounds a bit thin
> Jan 12 10:24:02 <tml_> surely one should test an ap well enough before
> submitting, to verify that it isn't possible to get it to try to invoke a
> constructor that isn't included
Change-Id: I20ce8aa40a00205c1cae9d6ed4a87a6093745dc4
Reviewed-on: https://gerrit.libreoffice.org/47791
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Stephan Bergmann <sbergman@redhat.com>
That *is* what we want, I guess, so better I do it proactively before
I am told to do it. Sorry for the back and forth, but my main point in
touching this code was anyway the filtering out of the component (the
KDE4Backend thing) that causes a crash for me, at least, and which it
presumably is fairly pointless to preload anyway.
Also, filter out libraries that for some reason have an empty name.
Change-Id: I2cffd51e67c23977e0779e7f0d6faaf59d8819d9
As the 1aedb6c434 commit message said,
the point is that we want this output even with a LO built for
production where SAL_INFO() and SAL_WARN() are no-ops.
Change-Id: I5f788c3914286ca8df2c7e290150adae0be42820
Use just SAL_INFO(), with tag "cppuhelper.preload". No unconditional
writes to std::cerr. Also, log the loading attempt beforehand, so that
it is printed before any crash during the loading of a component.
Change-Id: Ifde8be8cb5d18dd88df21c45c906deb575d36960
...than 25313923b0 "Fix CppunitTest_services for
constructor-based implementations..." (which this commit reverts again). My
claim that "the 'factory' would be the object itself" is nonsense, it would
rather be an ImplementationWrapper (but one that was freshly created for each
ServiceManager::createContentEnumeration).
Change-Id: I85c683cff6f9ba78d0f8567a53f8fcbc56fe55cf
Previosly (since commit 9ac98e6e34)
it was expected to gradually remove SAL_U/W usage in Windows code
by replacing with reinterpret_cast or changing to some bettertypes.
But as it's useful to make use of fact that LibreOffice and Windows
use compatible representation of strings, this commit puts these
functions to a better-suited o3tl, and recommends that the functions
be consistently used throughout Windows-specific code to reflect the
compatibility and keep the casts safe.
Change-Id: I2f7c65606d0e2d0c01a00f08812bb4ab7659c5f6
Reviewed-on: https://gerrit.libreoffice.org/43150
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
On Windows, UTF-8 is never current locale encoding; so using 8-bit
strings will always fail for paths containing characters outside
of current codepage.
Also fix leaks caused by failing to release its result: previously
it could return either result of getenv (that shouldn't get freed),
or an allocated string, but never got freed; now the result is
always allocated and properly freed.
Change-Id: I8b255dea20040eec0572de2b34280749fe8f071c
Reviewed-on: https://gerrit.libreoffice.org/42743
Tested-by: Jenkins <ci@libreoffice.org>
Reviewed-by: Mike Kaganski <mike.kaganski@collabora.com>
