Re: [Qemu-devel] TCP Segementation Offloading

2016-05-05 Thread Ingo Krabbe
> On Sun, May 01, 2016 at 02:31:57PM +0200, Ingo Krabbe wrote: >> Good Mayday Qemu Developers, >> >> today I tried to find a reference to a networking problem, that seems to be >> of quite general nature: TCP Segmentation Offloading (TSO) in virtual >> envi

[Qemu-devel] [Bug 1202289] Re: Windows 2008/7 Guest to Guest Very slow 10-20Mbit/s

2016-05-01 Thread Ingo Krabbe
I'm not sure yet, but I think the TSO problem affects much more setups, not only windows guests. You might try to set GSO to off in windows guests too, if there is such an option (GSO might be a linux kernel feature). I currently think that we have similar problems with centos7 hosts and guests

[Qemu-devel] TCP Segementation Offloading

2016-05-01 Thread Ingo Krabbe
xt week anyway and I woud be happy to share my observation with you. (Where can I register for the wiki, or whom should I sent my reports about this topic?). Regards, Ingo Krabbe

[Qemu-devel] [Bug 1412098] Re: qemu crashes when ctrl-alt-u is pressed

2015-02-20 Thread Ingo Krabbe
A build from the current master attached in gdb reveals Program received signal SIGSEGV, Segmentation fault. sdl_switch (dcl=0x7f4db26e4b20, new_surface=new_surface@entry=0x0) at ui/sdl.c:128 128 PixelFormat pf = qemu_pixelformat_from_pixman(new_surface-format); (gdb) bt #0 sdl_switch

[Qemu-devel] [Bug 1412098] Re: qemu crashes when ctrl-alt-u is pressed

2015-02-20 Thread Ingo Krabbe
Actually in any version this can never work, as you call sdl_switch(dcl,NULL); in ui/sdl.c:552. So the dereferncing statement new_surface-format must SEGFAULT. The obvious patch is very simple, of course, as just the statement below line 128 asks if(new_surface). So pf should be