Bug#651227: Compositing unuseable with Radeon r300
Package: kde-window-manager Version: 4:4.6.5-3 Severity: normal If I enable windows effects the glxgears framerate goes down to 20 fps, moving windows etc. is a pain. Doing "compiz --replace" works as expected: glxgears is at the expected 60 fps (synchronized with monitor refresh rate), wobbling windows are a real fun. Because compiz does not work well with KDE4 I have disabled windows effects. from /var/log/Xorg.0.log: (II) LoadModule: "radeon" (II) Loading /usr/lib/xorg/modules/drivers/radeon_drv.so (II) Module radeon: vendor="X.Org Foundation" compiled for 1.11.1.902, module version = 6.14.3 Module class: X.Org Video Driver ABI class: X.Org Video Driver, version 11.0 Chipset: "ATI Radeon X1200" (ChipID = 0x791e) from glxinfo: GLX version: 1.4 OpenGL vendor string: X.Org R300 Project OpenGL renderer string: Gallium 0.4 on ATI RS690 OpenGL version string: 2.1 Mesa 7.11 OpenGL shading language version string: 1.20 -- To UNSUBSCRIBE, email to debian-qt-kde-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/201112062216.57177.patchesthomas@web.de
Bug#651227: Compositing unuseable with Radeon r300
I retried with 4:4.8.4-3 but got similar bad results than a year ago: > glxgears Running synchronized to the vertical refresh. The framerate should be approximately the same as the monitor refresh rate. 62 frames in 5.0 seconds = 12.369 FPS 64 frames in 5.0 seconds = 12.778 FPS ... from glxinfo: OpenGL vendor string: X.Org R300 Project OpenGL renderer string: Gallium 0.4 on ATI RS690 OpenGL version string: 2.1 Mesa 8.0.4 OpenGL shading language version string: 1.20 -- To UNSUBSCRIBE, email to debian-qt-kde-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: http://lists.debian.org/201210172245.42174.patchesthomas@web.de
Bug#795006: sddm should provide x-display-manager
Package: sddm Version: 0.11.0-3 The package should have a Provides: x-display-manager field. Because of this if I look into the list of x-display-manager, sddm is missing. A Tag: field in the description would be nice, too. Suggestion (adapted from kdm): Tag: implemented-in::c++, interface::daemon, interface::x11, network::server, role::program, uitoolkit::qt, use::login, x11::application, x11::display-manager -- To UNSUBSCRIBE, email to debian-qt-kde-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org Archive: https://lists.debian.org/55c72b3e.8000...@web.de
Bug#846996: Unable to compile with g++-6
Package: qtbase5-dev Version: 5.7.1~20161021+dfsg-6 Severity: important Trying to compile lxqt-panel with g++-6 and qtbase5-dev_5.7.1 results in the following errors: In file included from /usr/include/x86_64-linux-gnu/qt5/QtCore/qlist.h:43:0, from /usr/include/x86_64-linux-gnu/qt5/QtCore/qobject.h:49, from /usr/include/x86_64-linux-gnu/qt5/QtCore/qplugin.h:43, from /usr/include/x86_64-linux-gnu/qt5/QtCore/QtPlugin:1, from /home/src/lxqt-panel-de/plugin-clock/../panel/ilxqtpanelplugin.h:32, from /home/src/lxqt-panel-de/plugin-clock/lxqtclock.h:34, from /home/src/lxqt-panel-de/plugin-clock/lxqtclock.cpp:32: /usr/include/x86_64-linux-gnu/qt5/QtCore/qalgorithms.h: In function ‘uint qCountTrailingZeroBits(quint16)’: /usr/include/x86_64-linux-gnu/qt5/QtCore/qalgorithms.h:630:32: error: ‘__builtin_ctzs’ was not declared in this scope return v ? __builtin_ctzs(v) : 16U; ^ /usr/include/x86_64-linux-gnu/qt5/QtCore/qalgorithms.h: In function ‘uint qCountLeadingZeroBits(quint16)’: /usr/include/x86_64-linux-gnu/qt5/QtCore/qalgorithms.h:693:32: error: ‘__builtin_clzs’ was not declared in this scope return v ? __builtin_clzs(v) : 16U; ^ This is because g++ removed __builtin_clzs and __builtin_ctzs. See for example https://www.mail-archive.com/gcc-bugs@gcc.gnu.org/msg511530.html
Bug#846996: Unable to compile with g++-6
Hi Dmitry, >> Trying to compile lxqt-panel with g++-6 and qtbase5-dev_5.7.1 results in the >> following errors: >> >> /usr/include/x86_64-linux-gnu/qt5/QtCore/qalgorithms.h: In function ‘uint >> qCountTrailingZeroBits(quint16)’: >> /usr/include/x86_64-linux-gnu/qt5/QtCore/qalgorithms.h:630:32: error: >> ‘__builtin_ctzs’ was not declared in this scope >> return v ? __builtin_ctzs(v) : 16U; >> ^ >> /usr/include/x86_64-linux-gnu/qt5/QtCore/qalgorithms.h: In function ‘uint >> qCountLeadingZeroBits(quint16)’: >> /usr/include/x86_64-linux-gnu/qt5/QtCore/qalgorithms.h:693:32: error: >> ‘__builtin_clzs’ was not declared in this scope >> return v ? __builtin_clzs(v) : 16U; >> ^ >> >> This is because g++ removed __builtin_clzs and __builtin_ctzs. > There is a potential fix for this at https://codereview.qt-project.org/175542, > but unfortunately it needs some more work. > > I have just asked the upstream author about the status, if he does not reply > in a few days, I will write my own patch and add it to the next Debian upload. according to Uroš Bizjak it should be sufficient to use return __tzcnt_u16(v); instead of return v ? __tzcnt_u16(v) : 16U; I haven't checked it...
Bug#846996: The mini-if is really not needed
I just tested with the following test program. The result is as expected:
#include
#include
#include
int main(void)
{
printf("0 -> %d\n\n", 16U);
puts("__tzcnt_u16:");
printf("0x%x -> %d\n", 0, __tzcnt_u16(0));
for (uint16_t i = 1; i > 0; i <<= 1)
{
printf("0x%x -> %d\n", i, __tzcnt_u16(i));
}
puts("\n__lzcnt16:");
printf("0x%x -> %d\n", 0, __lzcnt16(0));
for (uint16_t i = 1; i > 0; i <<= 1)
{
printf("0x%x -> %d\n", i, __lzcnt16(i));
}
return 0;
}
g++ -march=native -o intrins intrinsictest.cpp
./intrins
Attached is the patch for
/usr/include/x86_64-linux-gnu/qt5/QtCore/qalgorithms.h
--- qalgorithms.h.orig 2016-12-18 15:57:30.279325472 +0100
+++ qalgorithms.h 2016-12-18 23:09:33.129324260 +0100
@@ -47,6 +47,10 @@ QT_WARNING_PUSH
QT_WARNING_DISABLE_GCC("-Wdeprecated-declarations")
QT_WARNING_DISABLE_CLANG("-Wdeprecated-declarations")
+#if defined(__BMI__) && defined(__LZCNT__) && !QT_HAS_BUILTIN(__builtin_ctzs)
+# include
+#endif
+
/*
Warning: The contents of QAlgorithmsPrivate is not a part of the public Qt API
and may be changed from version to version or even be completely removed.
@@ -626,8 +630,10 @@ Q_DECL_RELAXED_CONSTEXPR inline uint qCo
Q_DECL_RELAXED_CONSTEXPR inline uint qCountTrailingZeroBits(quint16 v) Q_DECL_NOTHROW
{
#if defined(Q_CC_GNU)
-# if QT_HAS_BUILTIN(__builtin_ctzs) || (defined(__LZCNT__) && defined(__BMI__))
+# if QT_HAS_BUILTIN(__builtin_ctzs)
return v ? __builtin_ctzs(v) : 16U;
+# elif (defined(__LZCNT__) && defined(__BMI__))
+return __tzcnt_u16(v);
# else
return v ? __builtin_ctz(v) : 16U;
# endif
@@ -689,8 +695,10 @@ Q_DECL_RELAXED_CONSTEXPR inline uint qCo
Q_DECL_RELAXED_CONSTEXPR inline uint qCountLeadingZeroBits(quint16 v) Q_DECL_NOTHROW
{
#if defined(Q_CC_GNU)
-# if QT_HAS_BUILTIN(__builtin_clzs) || (defined(__LZCNT__) && defined(__BMI__))
+# if QT_HAS_BUILTIN(__builtin_clzs)
return v ? __builtin_clzs(v) : 16U;
+# elif (defined(__LZCNT__) && defined(__BMI__))
+return __lzcnt16(v);
# else
return v ? __builtin_clz(v)-16U : 16U;
# endif
Bug#907280: Package should recommend fonts-hack instead of fonts-hack-ttf
Package: plasma-integration Version: 5.13.4-1 The description of fonts-hack-ttf says: This package used to contain the TTF variants of the Hack font, which are now part of the consolidated fonts-hack package. This package is a dummy transitional package. It can be safely removed. Please update the Recommends field.
Bug#907281: Package should recommend fonts-hack instead of fonts-hack-ttf
Package: plasma-desktop Version: 4:5.13.4-1 The description of fonts-hack-ttf says: This package used to contain the TTF variants of the Hack font, which are now part of the consolidated fonts-hack package. This package is a dummy transitional package. It can be safely removed. Please update the Recommends field.
