[Development] QInotifyFileSystemWatcherEngine might have incomplete data?
Hi,
I'm trying to make my first little steps in making a Qt patch to implement a:
signal created(const QString path)
because that would allow you to know which files get created in a
folder that is being watched. KDirWatch has that feature,
QFileSystemWatcher doesn't.
However, when trying to implement that feature i was faced with an
empty inotify_event,name.
According to the inotify docs [1]. Specifically:
-- QUOTE --
The name field is only present when an event is returned for a file
inside a watched directory; it identifies the file pathname relative
to the watched directory. This pathname is null-terminated, and may
include further null bytes to align subsequent reads to a suitable
address boundary.
-- END QUOTE --
I'm not quite sure of this is a bug, something intentional or
something i'm doing wrong. The steps to reproduce this are as follows:
1. Create a simple that watches a folder:
#include QCoreApplication
#include QFileSystemWatcher
int main(int argc, char *argv[])
{
QCoreApplication a(argc, argv);
QFileSystemWatcher* watcher = new QFileSystemWatcher(a);
watcher-addPath(/path/to/folder/);
return a.exec();
}
2. Set a breakpoint within the
QInotifyFileSystemWatcherEngine::readFromInotify() function after the
point where you have an event object.
3. Create a new file in the folder you're monitoring.
4. GDB (or whatever you used) should hit the bearkpoint. Now inspace
the event members (name specifically). It's empty while it shouldn't
be.
Just to be sure that inotify works, i tested the same by typing the command:
inotifywait -m -e create /path/to/folder/
Creating a new file then gives me a new line like:
/path/to/folder/ CREATE newFileName
Which leads me to think that inotify itself is working just fine.
I hope someone could take a look at this and see if i'm wrong or if
this is a bug somewhere?
CCing the people that changed the readFromInotify function.
Cheers,
Mark
[1] http://linux.die.net/man/7/inotify
___
Development mailing list
Development@qt-project.org
http://lists.qt-project.org/mailman/listinfo/development
Re: [Development] QInotifyFileSystemWatcherEngine might have incomplete data?
On Sunday 17 November 2013 19:07:48 Mark Gaiser wrote: 4. GDB (or whatever you used) should hit the bearkpoint. Now inspace the event members (name specifically). It's empty while it shouldn't be. Hi, I'd add a qDebug() to be sure. Maybe it's just GDB giving wronf information. Perhaps the data as been optimized. Perhaps gdb think the name has lengh 0 because it is declared as a char[]. -- Olivier Woboq - Qt services and support - http://woboq.com ___ Development mailing list Development@qt-project.org http://lists.qt-project.org/mailman/listinfo/development
Re: [Development] QInotifyFileSystemWatcherEngine might have incomplete data?
On Sun, Nov 17, 2013 at 8:03 PM, Olivier Goffart oliv...@woboq.com wrote: On Sunday 17 November 2013 19:07:48 Mark Gaiser wrote: 4. GDB (or whatever you used) should hit the bearkpoint. Now inspace the event members (name specifically). It's empty while it shouldn't be. Hi, I'd add a qDebug() to be sure. Maybe it's just GDB giving wronf information. Perhaps the data as been optimized. Perhaps gdb think the name has lengh 0 because it is declared as a char[]. Hi Oliver, I tried your suggestion and that indeed works though not as i would expect it. Yes, it does get printed on screen when i put event.name in a qDebug line. However, none of the emit signals in that function (readFromInotify) work anymore when i use event.name there. Anywhere! I just tried that out since you've had send the reply. You can find the diff of the stuff that's i've done thus far. Just adding this to your code breaks all the emits and i have no clue as to why that happens or how it's even possible. If you have any clue.. please do tell. ___ Development mailing list Development@qt-project.org http://lists.qt-project.org/mailman/listinfo/development
Re: [Development] QInotifyFileSystemWatcherEngine might have incomplete data?
On Sun, Nov 17, 2013 at 9:20 PM, Mark Gaiser mark...@gmail.com wrote: On Sun, Nov 17, 2013 at 8:03 PM, Olivier Goffart oliv...@woboq.com wrote: On Sunday 17 November 2013 19:07:48 Mark Gaiser wrote: 4. GDB (or whatever you used) should hit the bearkpoint. Now inspace the event members (name specifically). It's empty while it shouldn't be. Hi, I'd add a qDebug() to be sure. Maybe it's just GDB giving wronf information. Perhaps the data as been optimized. Perhaps gdb think the name has lengh 0 because it is declared as a char[]. Hi Oliver, I tried your suggestion and that indeed works though not as i would expect it. Yes, it does get printed on screen when i put event.name in a qDebug line. However, none of the emit signals in that function (readFromInotify) work anymore when i use event.name there. Anywhere! I just tried that out since you've had send the reply. You can find the diff of the stuff that's i've done thus far. Just adding this to your code breaks all the emits and i have no clue as to why that happens or how it's even possible. If you have any clue.. please do tell. Sorry, forgot the diff: http://pastebin.kde.org/pzqb2jph6 Lifetime of 30 days. ___ Development mailing list Development@qt-project.org http://lists.qt-project.org/mailman/listinfo/development
Re: [Development] StereoViewport for Qt3D
On Wed 06 Nov 2013 12:54:07 PM CET, Rutledge Shawn wrote: On 28 Oct 2013, at 10:35 PM, Svenn-Arne Dragly wrote: Secondly, I was missing a stereo option for the default QML Viewport element, and decided to have a look at the implementation in QGLView to see if I could port that to the QML Viewport element. Luckily this was quite easy (thanks to Qt3D being well designed) and I have uploaded the resulting StereoViewport element to a GitHub repository: https://github.com/dragly/stereoviewport This should compile with Qt5.1.1 and the latest master branch of Qt3D. Cool. Are you using it with an Oculus Rift, 3D TV or something else? Both and Oculus Rift and a 3D TV, actually. I have tested the stereo mode where the viewport is split in two thoroughly, with one half of the screen for the left eye and the other for the right. It works quite well and seems to be stable. Now, I'm wondering if this feature should be a part of the original Viewport element. Should I contribute this into the main Qt3D repository? If so, what steps should I take before doing this, apart from those listed in the Qt Contributing Guidelines? I'm for instance not sure if I might have introduced some bugs or damaged any features of the original Viewport element. How should I verify that I haven't done so? I don't know much about that but I'm sure you will get a reply from someone who does. But it sounds like a good feature to have. I hope so too, but if I don't hear anything I'll just try pushing it to gerrit and see how it goes. Svenn-Arne ___ Development mailing list Development@qt-project.org http://lists.qt-project.org/mailman/listinfo/development
