search for: get_rt_format

https://bugs.freedesktop.org/show_bug.cgi?id=49148 Bug #: 49148 Summary: nv20_state_fb.c:50: get_rt_format: Assertion `0' failed. Classification: Unclassified Product: xorg Version: unspecified Platform: x86 (IA32) OS/Version: Linux (All) Status: NEW Severity: normal Priority: medium Component: D...

https://bugs.freedesktop.org/show_bug.cgi?id=73459 Priority: medium Bug ID: 73459 Assignee: nouveau at lists.freedesktop.org Summary: nv10 get_rt_format: Assertion `0' failed. Severity: normal Classification: Unclassified OS: Linux (All) Reporter: dnietoc at gmail.com Hardware: x86 (IA32) Status: NEW Version: git Component: Drivers/DRI/nouveau Product:...

...MAT_COLOR_X8R8G8B8 }, + { PICT_a8r8g8b8, NV10TCL_RT_FORMAT_COLOR_A8R8G8B8 }, {}, }; @@ -432,7 +435,7 @@ setup_render_target(NVPtr pNv, PicturePtr pict, PixmapPtr pixmap) struct nouveau_bo *bo = nouveau_pixmap_bo(pixmap); BEGIN_RING(chan, celsius, NV10TCL_RT_FORMAT, 2); - OUT_RING (chan, get_rt_format(pict)); + OUT_RING (chan, get_rt_format(pict) | NV10TCL_RT_FORMAT_TYPE_LINEAR); OUT_RING (chan, (exaGetPixmapPitch(pixmap) << 16 | exaGetPixmapPitch(pixmap))); -- 1.7.0.5

This fixes issues where get_rt_format would see a 0 format because the nouveau_surface had not been properly initialized. Fixes crash on supertuxkart startup (which still fails due to out-of-vram issues). Signed-off-by: Ilia Mirkin <imirkin at alum.mit.edu> --- src/mesa/drivers/dri/nouveau/nouveau_texture.c | 9 +++++++++ 1 fil...

...s told that it was a problem with the nouveau driver, not Wine: > > This is an assertion in nouveau driver, so you have to try to update it or use > > proprietary one. > To add to that, the backtrace is missing function names and line numbers, but > this is probably the assert in get_rt_format() in nv10_state_fb.c. It gets > triggered because during wined3d startup we try to attach a GL_RGB5_A1 texture > (among various others) to an FBO. The driver should probably fail the FBO > completeness test for formats it doesn't know how to handle instead of > asserting. I am runn...

...observed on nv25, where in _mesa_prepare_mipmap_level call to Driver.FreeTextureImageBuffer unreferences nouveau_teximage's nouveau_surface, which then doesnt get reallocated in Driver.AllocTextureImageBuffer, because vieux uses swrast implementation, ultimately leading to assertion failure in get_rt_format in nv20_buffer_emit. Apart from adding nouveau_alloc_texture_image_buffer I renamed nouveau_teximage_free. What I am unsure of is the code concerning swrast_texture_image and the test for early return in nouveau_alloc_texture_image_buffer. Viktor src/mesa/drivers/dri/nouveau/nouveau_texture.c...

...at *format = + pict->repeat != RepeatNone ? nv10_tex_format_pot : + pNv->Architecture == NV_ARCH_20 ? nv20_tex_format_rect : + nv10_tex_format_rect; + + for (; format->hw; format++) { + if (format->exa == pict->format) + return format->hw; + } + + return 0; +} +static int +get_rt_format(PicturePtr pict) +{ + struct pict_format *format = nv10_rt_format; + + for (; format->hw; format++) { + if (format->exa == pict->format) + return format->hw; + } + + return 0; +} + +/* Blending functions. */ #define SF(x) NV10TCL_BLEND_FUNC_SRC_##x #define DF(x) NV10TCL_BLEND_FUNC_...