1Mesa 13.0.3 Release Notes / January 5, 2017 2=========================================== 3 4Mesa 13.0.3 is a bug fix release which fixes bugs found since the 13.0.2 5release. 6 7Mesa 13.0.3 implements the OpenGL 4.4 API, but the version reported by 8glGetString(GL_VERSION) or glGetIntegerv(GL_MAJOR_VERSION) / 9glGetIntegerv(GL_MINOR_VERSION) depends on the particular driver being 10used. Some drivers don't support all the features required in OpenGL 114.4. OpenGL 4.4 is **only** available if requested at context creation 12because compatibility contexts are not supported. 13 14SHA256 checksums 15---------------- 16 17:: 18 19 55b07d056f9b855ba9d7c8b2ddc7d3b220a61c6ab1bdc73cbfc2f607721094c2 mesa-13.0.3.tar.gz 20 d9aa8be5c176d00d0cd503cb2f64a5a403ea471ec819c022581414860d7ba40e mesa-13.0.3.tar.xz 21 22New features 23------------ 24 25None 26 27Bug fixes 28--------- 29 30- `Bug 77662 <https://bugs.freedesktop.org/show_bug.cgi?id=77662>`__ - 31 Fail to render to different faces of depth-stencil cube map 32- `Bug 92234 <https://bugs.freedesktop.org/show_bug.cgi?id=92234>`__ - 33 [BDW] GPU hang in Shogun2 34- `Bug 98329 <https://bugs.freedesktop.org/show_bug.cgi?id=98329>`__ - 35 [dEQP, EGL, SKL, BDW, BSW] 36 dEQP-EGL.functional.image.render_multiple_contexts.gles2_renderbuffer_depth16_depth_buffer 37- `Bug 99038 <https://bugs.freedesktop.org/show_bug.cgi?id=99038>`__ - 38 [dEQP, EGL, SKL, BDW, BSW] 39 dEQP-EGL.functional.negative_api.create_pixmap_surface crashes 40 41Changes 42------- 43 44Chad Versace (2): 45 46- i965/mt: Disable aux surfaces after making miptree shareable 47- egl: Fix crashes in eglCreate*Surface() 48 49Dave Airlie (4): 50 51- anv: set maxFragmentDualSrcAttachments to 1 52- radv: set maxFragmentDualSrcAttachments to 1 53- radv: fix another regression since shadow fixes. 54- radv: add missing license file to radv_meta_bufimage. 55 56Emil Velikov (5): 57 58- docs: add sha256 checksums for 13.0.2 59- anv: don't double-close the same fd 60- anv: don't leak memory if anv_init_wsi() fails 61- radv: don't leak the fd if radv_physical_device_init() succeeds 62- Update version to 13.0.3 63 64Eric Anholt (1): 65 66- vc4: In a loop break/continue, jump if everyone has taken the path. 67 68Gwan-gyeong Mun (3): 69 70- anv: Add missing error-checking to anv_block_pool_init (v2) 71- anv: Update the teardown in reverse order of the anv_CreateDevice 72- vulkan/wsi: Fix resource leak in success path of wsi_queue_init() 73 74Haixia Shi (1): 75 76- compiler/glsl: fix precision problem of tanh 77 78Ilia Mirkin (1): 79 80- mesa: only verify that enabled arrays have backing buffers 81 82Jason Ekstrand (8): 83 84- anv/cmd_buffer: Re-emit MEDIA_CURBE_LOAD when CS push constants are 85 dirty 86- anv/image: Rename hiz_surface to aux_surface 87- anv/cmd_buffer: Remove the 1-D case from the HiZ QPitch calculation 88- genxml/gen9: Change the default of 89 MI_SEMAPHORE_WAIT::RegisterPoleMode 90- anv/device: Return the right error for failed maps 91- anv/device: Implicitly unmap memory objects in FreeMemory 92- anv/descriptor_set: Write the state offset in the surface state free 93 list. 94- spirv: Use a simpler and more correct implementaiton of tanh() 95 96Kenneth Graunke (1): 97 98- i965: Allocate at least some URB space even when max_vertices = 0. 99 100Marek Olšák (17): 101 102- radeonsi: always set all blend registers 103- radeonsi: set CB_BLEND1_CONTROL.ENABLE for dual source blending 104- radeonsi: disable RB+ blend optimizations for dual source blending 105- radeonsi: consolidate max-work-group-size computation 106- radeonsi: apply a multi-wave workgroup SPI bug workaround to affected 107 CIK chips 108- radeonsi: apply a TC L1 write corruption workaround for SI 109- radeonsi: apply a tessellation bug workaround for SI 110- radeonsi: add a tess+GS hang workaround for VI dGPUs 111- radeonsi: apply the double EVENT_WRITE_EOP workaround to VI as well 112- cso: don't release sampler states that are bound 113- radeonsi: always restore sampler states when unbinding sampler views 114- radeonsi: fix incorrect FMASK checking in bind_sampler_states 115- radeonsi: allow specifying simm16 of emit_waitcnt at call sites 116- radeonsi: wait for outstanding memory instructions in TCS barriers 117- tgsi: fix the src type of TGSI_OPCODE_MEMBAR 118- radeonsi: wait for outstanding LDS instructions in memory barriers if 119 needed 120- radeonsi: disable the constant engine (CE) on Carrizo and Stoney 121 122Matt Turner (3): 123 124- i965/fs: Rename opt_copy_propagate -> opt_copy_propagation. 125- i965/fs: Add unit tests for copy propagation pass. 126- i965/fs: Reject copy propagation into SEL if not min/max. 127 128Nanley Chery (1): 129 130- mesa/fbobject: Update CubeMapFace when reusing textures 131 132Nicolai Hähnle (4): 133 134- radeonsi: fix isolines tess factor writes to control ring 135- radeonsi: update all GSVS ring descriptors for new buffer allocations 136- radeonsi: do not kill GS with memory writes 137- radeonsi: fix an off-by-one error in the bounds check for 138 max_vertices 139 140Rhys Kidd (1): 141 142- glsl: Add pthread libs to cache_test 143 144Timothy Arceri (2): 145 146- mesa: fix active subroutine uniforms properly 147- Revert "nir: Turn imov/fmov of undef into undef." 148