Hello,
I tested the r200 driver from the latest CVS trunk. My config
is a dual athlon box, radeon 8500. I'm running a SMP
redhat kernel (2.4.18-10smp). I have R200_NO_USLEEP set to 1.
I overwrited the radeon.o kernel module with the one
from the CVS dri sources. I compiled both XFree86 and the drm
module with gcc-2.96-112 from redhat 7.3.
I noticed that the X server becomes unresponsive when
I start glxgears while running flightgear for example.
I still can connect remotely. glxgears takes 99% of one
CPU. The backtrace of glxgears is :
(gdb) bt
#0 0x4027cc64 in ioctl () from /lib/libc.so.6
#1 0x0804f140 in ?? ()
#2 0x4043dd5e in r200WaitForFrameCompletion (rmesa=0x804f140)
at r200_ioctl.c:556
#3 0x4043deb5 in r200CopyBuffer (dPriv=0x804cfe0) at
r200_ioctl.c:608
#4 0x40069c8a in glXSwapBuffers (dpy=0x804bd88, drawable=31457282)
at glxcmds.c:578
#5 0x0804a67b in XOpenDisplay ()
#6 0x0804a893 in XOpenDisplay ()
#7 0x401be1c4 in __libc_start_main () from /lib/libc.so.6
glxgears never returns from drmCommandWriteRead() while attached
in gdb.
The backtrace of flightgear is :
(gdb) bt
#0 0x404c2c64 in ioctl () from /lib/libc.so.6
#1 0xbffff358 in ?? ()
#2 0x47778d50 in r200GetLock (rmesa=0x86fa2f8, flags=0) at
r200_lock.c:86
#3 0x477777b7 in r200FlushCmdBuf (rmesa=0x86fa2f8,
caller=0x4779c98c "r200Flush") at r200_ioctl.c:375
#4 0x47778bf3 in r200Flush (ctx=0x870d988) at r200_ioctl.c:936
#5 0x47766036 in r200SetDrawBuffer (ctx=0x870d988, mode=1026)
at r200_state.c:1550
#6 0x476450d9 in _mesa_DrawBuffer (mode=1029) at buffers.c:264
#7 0x4764300c in _mesa_PopAttrib () at attrib.c:767
#8 0x08338c80 in std::string*
std::__uninitialized_copy_aux<__gnu_cxx::__normal_iterator<std::string
const*, std::vector<std::string, std::allocator<std::string> > >,
std::string*>(__gnu_cxx::__normal_iterator<std::string const*,
std::vector<std::string, std::allocator<std::string> > >,
__gnu_cxx::__normal_iterator<std::string const*,
std::vector<std::string, std::allocator<std::string> > >, std::string*,
__false_type) ()
When flightgear is running alone, everything works fine. Any idea ?
Can I help to provide more useful debugging trace ?
--
fabrice
-------------------------------------------------------
This SF.NET email is sponsored by: AMD - Your access to the experts
on Hammer Technology! Open Source & Linux Developers, register now
for the AMD Developer Symposium. Code: EX8664
http://www.developwithamd.com/developerlab
_______________________________________________
Dri-devel mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/dri-devel
