search

ssvb / xf86-video-fbturbo

Xorg DDX driver for ARM devices (Allwinner, RPi and others)
Other
202 stars 80 forks source link

eglSwapBuffers #47

Open magicse opened 9 years ago

magicse commented 9 years ago

Driversr 4p0-00rel1 Device Orange Pi2 (Allwinner H3) Mali-400mp2 lsmod Module Size Used by mali_drm 6634 0 drm 180399 1 mali_drm mali 344514 0 ump 38853 1 mali

In test app triangle I get first frame and after seg fault. After debug I found that it crash in call eglSwapBuffers();

ssvb commented 9 years ago

Allwinner H3 is a rather new SoC and you are probably the first one who tried to test 3D acceleration on it in Linux. The r4p0 mali kernel driver also was never used with xf86-video-fbturbo ddx before.

I would suggest to try the xf86-video-mali driver to see if it does not crash for you and report the results.

I also have ordered an Orange Pi PC board with Allwinner H3 myself, but have no idea how long it will take to be delivered from China. Hopefully before Christmas.

magicse commented 9 years ago

Ok. You driver work normal and not crashed when starting X. Problems that I found in es2_info seg fault on call XCloseDisplay(x_dpy); And in triangle test seg fault after first frame on call eglSwapBuffers(), but I see first frame that mean is problem with buffers.

magicse commented 9 years ago

xf86-video-mali driver crashed on start X with next log in Xorg after satrt a got next errors in Xorg.0.log

13.931 Loading sub module "exa" 13.932 LoadModule: "exa" 13.932 Loading /usr/lib/xorg/modules/libexa.so 13.937 Module exa: vendor="X.Org Foundation" [ 13.937] compiled for 1.15.1, module version = 2.6.0 [ 13.937] ABI class: X.Org Video Driver, version 15.0

13.938 MALI(0): Mali EXA driver is loaded successfully

13.938 MALI(0): [MaliScreenInit:1355] Warning: Initializing EXA Driver! 13.938 EXA(0): Driver allocated offscreen pixmaps 13.938 EXA(0): Driver registered support for the following operations: 13.938 Solid 13.938 Copy 13.938 Composite (RENDER acceleration) 13.938 MALI(0): Backing store enabled 13.938 MALI(0): Silken mouse enabled 13.945 MALI(0): Changing mode to 720 576 720 1152 13.945 MALI(0): RandR 1.2 enabled, ignore the following RandR disabled message. 13.946 RandR disabled 13.978 SELinux: Disabled on system 13.984 AIGLX error: dlopen of /usr/lib/arm-linux-gnueabihf/dri/Mali DRI2_dri.so failed (/usr/lib/arm-linux-gnueabih

f/dri/Mali DRI2_dri.so: cannot open shared object file: No such file or directory) 13.984 AIGLX: reverting to software rendering 14.095 AIGLX: Loaded and initialized swrast 14.095 GLX: Initialized DRISWRAST GL provider for screen 0 14.097 MALI(0): [maliModifyPixmapHeader:184] Error: GET_UMP_SECURE_ID_BUF1 returned 0xffffffff offset: 0 virt addre

ss: 0xb63c6000 fb_virt: 0xb63c6000

14.097 MALI(0): [maliModifyPixmapHeader:196] Error: UMP failed to retrieve secure id 14.097 MALI(0): Setting screen physical size to 190 x 152 14.097 MALI(0): fbdev_crtc_config_resize: width = 720 height = 576

14.097 14.097 Backtrace: 14.097 14.097 Segmentation fault at address 0x20 14.097

Fatal server error: 14.097 Caught signal 11 (Segmentation fault). Server aborting 14.098 14.098

ssvb commented 9 years ago

Yes, The XCloseDisplay is an old mali gles blob problem, which they never fixed. It indeed may cause segfaults on exit from applications.

Anyway, if the crashes are on the client side, then there is probably nothing to fix in xf86-video-fbturbo which runs in the X server process.

ssvb commented 9 years ago

xf86-video-mali driver crashed on start X with next log in Xorg after satrt a got next errors in Xorg.0.log

Thanks for this information. It looks like your kernel is probably not implementing the GET_UMP_SECURE_ID_BUF1 ioctl, this upsets xf86-video-mali and makes it crash in a spectacular way. The xf86-video-fbturbo ddx is handling this more gracefully.

magicse commented 9 years ago

If need I could make more logs and debugs. Now I try resolve eglSwapBuffers issue. And disp module build with kernel.... may be it need build like external disp.ko?

magicse commented 9 years ago

Back traces from running test app colour triangle with FBTURBO and mali r4p0 drivers. (gdb) bt

0 __GI___libc_malloc (bytes=) at malloc.c:2907

1 0xb6f59cc4 in __egl_mali_post_to_window_surface () from /usr/lib/libEGL.so

2 0xb6f59294 in _egl_swap_buffers () from /usr/lib/libEGL.so

3 0xb6f5836c in eglSwapBuffers () from /usr/lib/libEGL.so

4 0x00009050 in Redraw (width=200, height=200) at test.c:135

5 0x00009790 in main (argc=1, argv=0xbefff6f4) at test.c:328

//==================================================

//================================================== Single stepping until exit from function _egl_surface_job_incref, which has no line number information. 0xb6f59dbe in __egl_mali_post_to_window_surface () from /usr/lib/libEGL.so (gdb) s Single stepping until exit from function __egl_mali_post_to_window_surface, which has no line number information.

Program received signal SIGSEGV, Segmentation fault. [Switching to Thread 0xb69d1460 (LWP 2377)] 0xb6f5a224 in _egl_surface_job_decref () from /usr/lib/libEGL.so (gdb) bt

0 0xb6f5a224 in _egl_surface_job_decref () from /usr/lib/libEGL.so

1 0xb6f596be in __egl_worker_callback () from /usr/lib/libEGL.so

2 0xb6f62238 in _mali_base_worker_thread () from /usr/lib/libEGL.so

3 0xb6d10fbc in start_thread (arg=0xb69d1460) at pthread_create.c:314

4 0xb6dc720c in ?? ()

at ../ports/sysdeps/unix/sysv/linux/arm/nptl/../clone.S:92

from /lib/arm-linux-gnueabihf/libc.so.6 Backtrace stopped: previous frame identical to this frame (corrupt stack?) (gdb)

magicse commented 8 years ago

With patched blob drivers worked well http://orangepi.pp.ua/index.php/topic,17.msg569.html#msg569