Messages

46

Window / Memory Leaks in sf::RenderWindow

« on: August 05, 2008, 11:21:56 am »

I noticed that a while ago, but I'm not sure if it really is a SFML related problem.
Anyway, here is the valgrind result (Ubuntu 8.04.1 32-bit, but I remember that the same happened with 8.10 alpha versions):

Code: [Select]

int main(int argc, char **argv)
{
sf::VideoMode Mode(640, 480);

sf::RenderWindow* RenderWindow = new sf::RenderWindow(Mode,"Test",sf::Style::Close|sf::Style::Resize);

delete RenderWindow;

return 0;
}


This is the valgrind output (run with  --leak-check=full):

Code: [Select]

==9284== Memcheck, a memory error detector.
==9284== Copyright (C) 2002-2007, and GNU GPL'd, by Julian Seward et al.
==9284== Using LibVEX rev 1804, a library for dynamic binary translation.
==9284== Copyright (C) 2004-2007, and GNU GPL'd, by OpenWorks LLP.
==9284== Using valgrind-3.3.0-Debian, a dynamic binary instrumentation framework.
==9284== Copyright (C) 2000-2007, and GNU GPL'd, by Julian Seward et al.
==9284== For more details, rerun with: -v
==9284==
==9284== Syscall param writev(vector[...]) points to uninitialised byte(s)
==9284==    at 0x40007F2: (within /lib/ld-2.7.so)
==9284==    by 0x4718EC2: (within /usr/lib/libxcb.so.1.0.0)
==9284==    by 0x47194AA: (within /usr/lib/libxcb.so.1.0.0)
==9284==    by 0x47195DE: (within /usr/lib/libxcb.so.1.0.0)
==9284==    by 0x471AB36: xcb_wait_for_reply (in /usr/lib/libxcb.so.1.0.0)
==9284==    by 0x44B61EA: _XReply (in /usr/lib/libX11.so.6.2.0)
==9284==    by 0x4499661: XInternAtom (in /usr/lib/libX11.so.6.2.0)
==9284==    by 0x406482A: sf::priv::WindowImplX11::Initialize() (WindowImplX11.cpp:691)
==9284==    by 0x4066306: sf::priv::WindowImplX11::WindowImplX11(sf::VideoMode, std::string const&, unsigned long, sf::WindowSettings&) (WindowImplX11.cpp:304)
==9284==    by 0x406107C: sf::priv::WindowImpl::New(sf::VideoMode, std::string const&, unsigned long, sf::WindowSettings&) (WindowImpl.cpp:113)
==9284==    by 0x405E0A8: sf::Window::Create(sf::VideoMode, std::string const&, unsigned long, sf::WindowSettings const&) (Window.cpp:121)
==9284==    by 0x40EF5E8: sf::RenderWindow::RenderWindow(sf::VideoMode, std::string const&, unsigned long, sf::WindowSettings const&) (RenderWindow.cpp:51)
==9284==  Address 0x533c034 is 4,380 bytes inside a block of size 8,556 alloc'd
==9284==    at 0x4021BDE: calloc (vg_replace_malloc.c:397)
==9284==    by 0x4719130: xcb_connect_to_fd (in /usr/lib/libxcb.so.1.0.0)
==9284==    by 0x471BA50: xcb_connect (in /usr/lib/libxcb.so.1.0.0)
==9284==    by 0x44B5716: _XConnectXCB (in /usr/lib/libX11.so.6.2.0)
==9284==    by 0x449E028: XOpenDisplay (in /usr/lib/libX11.so.6.2.0)
==9284==    by 0x4063399: sf::priv::WindowImplX11::OpenDisplay(bool) (WindowImplX11.cpp:1139)
==9284==    by 0x4066A79: sf::priv::WindowImplX11::WindowImplX11() (WindowImplX11.cpp:71)
==9284==    by 0x40610DA: sf::priv::WindowImpl::New() (WindowImpl.cpp:80)
==9284==    by 0x405D82C: sf::Window::ForceContextInit() (Window.cpp:391)
==9284==    by 0x405DF96: sf::Window::Create(sf::VideoMode, std::string const&, unsigned long, sf::WindowSettings const&) (Window.cpp:104)
==9284==    by 0x40EF5E8: sf::RenderWindow::RenderWindow(sf::VideoMode, std::string const&, unsigned long, sf::WindowSettings const&) (RenderWindow.cpp:51)
==9284==    by 0x804C5C5: main (main.cpp:53)
==9284==
==9284== Invalid read of size 4
==9284==    at 0x4015220: (within /lib/ld-2.7.so)
==9284==    by 0x4005C69: (within /lib/ld-2.7.so)
==9284==    by 0x4007A97: (within /lib/ld-2.7.so)
==9284==    by 0x4011543: (within /lib/ld-2.7.so)
==9284==    by 0x400D5D5: (within /lib/ld-2.7.so)
==9284==    by 0x4010F5D: (within /lib/ld-2.7.so)
==9284==    by 0x4729C18: (within /lib/tls/i686/cmov/libdl-2.7.so)
==9284==    by 0x400D5D5: (within /lib/ld-2.7.so)
==9284==    by 0x472A2BB: (within /lib/tls/i686/cmov/libdl-2.7.so)
==9284==    by 0x4729B50: dlopen (in /lib/tls/i686/cmov/libdl-2.7.so)
==9284==    by 0x448C786: (within /usr/lib/libX11.so.6.2.0)
==9284==    by 0x448CA6E: _XNoticeCreateBitmap (in /usr/lib/libX11.so.6.2.0)
==9284==  Address 0x562fea0 is 24 bytes inside a block of size 25 alloc'd
==9284==    at 0x4022AB8: malloc (vg_replace_malloc.c:207)
==9284==    by 0x4008031: (within /lib/ld-2.7.so)
==9284==    by 0x4011543: (within /lib/ld-2.7.so)
==9284==    by 0x400D5D5: (within /lib/ld-2.7.so)
==9284==    by 0x4010F5D: (within /lib/ld-2.7.so)
==9284==    by 0x4729C18: (within /lib/tls/i686/cmov/libdl-2.7.so)
==9284==    by 0x400D5D5: (within /lib/ld-2.7.so)
==9284==    by 0x472A2BB: (within /lib/tls/i686/cmov/libdl-2.7.so)
==9284==    by 0x4729B50: dlopen (in /lib/tls/i686/cmov/libdl-2.7.so)
==9284==    by 0x448C786: (within /usr/lib/libX11.so.6.2.0)
==9284==    by 0x448CA6E: _XNoticeCreateBitmap (in /usr/lib/libX11.so.6.2.0)
==9284==    by 0x448CE6C: XCreatePixmap (in /usr/lib/libX11.so.6.2.0)
==9284==
==9284== ERROR SUMMARY: 2 errors from 2 contexts (suppressed: 71 from 1)
==9284== malloc/free: in use at exit: 359,080 bytes in 803 blocks.
==9284== malloc/free: 5,906 allocs, 5,103 frees, 3,470,686 bytes allocated.
==9284== For counts of detected errors, rerun with: -v
==9284== searching for pointers to 803 not-freed blocks.
==9284== checked 815,648 bytes.
==9284==
==9284==
==9284== 262,140 bytes in 1 blocks are possibly lost in loss record 100 of 100
==9284==    at 0x4022AB8: malloc (vg_replace_malloc.c:207)
==9284==    by 0x45AAD56: (within /usr/lib/libGL.so.169.12)
==9284==
==9284== LEAK SUMMARY:
==9284==    definitely lost: 0 bytes in 0 blocks.
==9284==      possibly lost: 262,140 bytes in 1 blocks.
==9284==    still reachable: 96,940 bytes in 802 blocks.
==9284==         suppressed: 0 bytes in 0 blocks.
==9284== Reachable blocks (those to which a pointer was found) are not shown.
==9284== To see them, rerun with: --leak-check=full --show-reachable=yes

47

Graphics / Draw(Text) performance issue

« on: July 30, 2008, 08:46:53 am »

lollll 320X240... I will draw an fps string in a 10X100 window and we will see wich fps i will get XD.

Any results? :wink:

48

Graphics / Draw(Text) performance issue

« on: July 29, 2008, 06:20:07 pm »

Nope, it's a 8800GTX, rendering nothing but a message box (only a few lines of OpenGL code) + fps string in a 320x240 window (it drops to an average 15000 fps if I change the resolution to 640x480).
Don't worry, it gets slower the more I draw on the screen. :wink:

49

Graphics / Draw(Text) performance issue

« on: July 29, 2008, 05:45:33 pm »

It definitely got faster. I got 5000 fps before, now I get 24000 fps.
I almost thought that there was something wrong with my message box rendering code.

50

Feature requests / wstring in Window title, unicode strings, String baseline

« on: July 29, 2008, 03:02:54 pm »

http://muffin.cias.osakafu-u.ac.jp/~matumoto/cgi-bin/xt.cgi?prog/squirrel
It seems to be down at random times. If there is nothing on that site, try again later...

That + Squirrel source code provides pretty much all information I need for the system I intend to create.

51

Feature requests / wstring in Window title, unicode strings, String baseline

« on: July 29, 2008, 12:14:14 pm »

If you decide to switch, the source of the Squirrel standard library also turned out to be helpful (+ some japanese site ).

52

Feature requests / wstring in Window title, unicode strings, String baseline

« on: July 29, 2008, 10:49:30 am »

I love Squirrel. Such fun  

I agree. Squirrel is quite nice.
I started to embed it just a few days ago. :wink:
It lacks documentation though.

53

Feature requests / wstring in Window title, unicode strings, String baseline

« on: July 28, 2008, 08:34:14 pm »

If it's only for the binding purpose you could load the file into memory and then use sf::Image::LoadFromMemory.

54

Graphics / OpenGL performance issue

« on: July 27, 2008, 11:26:28 am »

You could use a profiler to find out what the bottleneck is.
A short question: Are you running Vista?
I get only about half of the performance I get on XP or Linux (on the same system). I always thought that it was a driver related problem, but maybe it is something different...

55

General discussions / Idling a program

« on: July 21, 2008, 10:19:04 pm »

You might have a look at sf::RenderWindow::SetFramerateLimit().

56

Audio / Audio "Clicks" on Linux

« on: July 13, 2008, 03:41:05 pm »

You're right. The problem is gone when I select the debug build.
Can you tell what causes the problem? Has it something to do with optimization?

57

Audio / Audio "Clicks" on Linux

« on: July 13, 2008, 02:14:33 pm »

Here is what ldd says:

Code: [Select]

linux-gate.so.1 =>  (0xb7f3d000)
libsfml-audio.so.1.3 => /usr/lib/libsfml-audio.so.1.3 (0xb7efe000)
libsfml-system.so.1.3 => /usr/lib/libsfml-system.so.1.3 (0xb7ef9000)
libstdc++.so.6 => /usr/lib/libstdc++.so.6 (0xb7e0a000)
libm.so.6 => /lib/tls/i686/cmov/libm.so.6 (0xb7de4000)
libgcc_s.so.1 => /lib/libgcc_s.so.1 (0xb7dd6000)
libc.so.6 => /lib/tls/i686/cmov/libc.so.6 (0xb7c78000)
libpthread.so.0 => /lib/tls/i686/cmov/libpthread.so.0 (0xb7c5f000)
libsndfile.so.1 => /usr/lib/libsndfile.so.1 (0xb7c03000)
libopenal.so.1 => /usr/lib/libopenal.so.1 (0xb78bf000)
/lib/ld-linux.so.2 (0xb7f3e000)
libFLAC.so.8 => /usr/lib/libFLAC.so.8 (0xb786c000)
libdl.so.2 => /lib/tls/i686/cmov/libdl.so.2 (0xb7868000)
libogg.so.0 => /usr/lib/libogg.so.0 (0xb7862000)

I assume that those are the right ones.
I'm using the latest revision (rev 768).

58

Audio / Audio "Clicks" on Linux

« on: July 13, 2008, 01:33:28 pm »

It still clicks for me...

59

General / Something strange... (Mouse click)

« on: July 09, 2008, 09:47:44 pm »

I tried it under linux (wine and linux native build) and everything seemed to work.

60

Window / [Solved] Stretch to fit screen

« on: July 04, 2008, 03:55:28 pm »

It should be sufficient to create a window with the wanted resolution:

Code: [Select]

sf::RenderWindow App(sf::VideoMode(1024, 768), sTitle, sf::Style::Fullscreen);