summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* radeonsi: move si_shader_dump call out of si_shader_binary_readMarek Olšák2016-01-073-20/+21
| | | | Reviewed-by: Nicolai Hähnle <[email protected]>
* radeonsi: separate shader dumping code to si_shader_dump and *_dump_statsMarek Olšák2016-01-071-12/+30
| | | | | | | Eventually, I'd like to dump stats for several combined binaries, which is why you don't see a binary parameter in si_shader_dump_stats Reviewed-by: Nicolai Hähnle <[email protected]>
* radeonsi: add si_shader_destroy_binaryMarek Olšák2016-01-072-5/+10
| | | | Reviewed-by: Nicolai Hähnle <[email protected]>
* radeonsi: don't pass si_shader to si_compile_llvmMarek Olšák2016-01-073-18/+28
| | | | Reviewed-by: Nicolai Hähnle <[email protected]>
* radeonsi: move si_shader_binary_upload out of si_compile_llvmMarek Olšák2016-01-072-4/+9
| | | | Reviewed-by: Nicolai Hähnle <[email protected]>
* radeonsi: always keep shader code, rodata, and relocs in memoryMarek Olšák2016-01-071-7/+3
| | | | | | | We won't compile shaders in draw calls, but we will concatenate shader binaries according to states in draw calls, so keep the binaries. Reviewed-by: Nicolai Hähnle <[email protected]>
* radeonsi: don't pass si_shader to si_shader_binary_readMarek Olšák2016-01-073-14/+19
| | | | Reviewed-by: Nicolai Hähnle <[email protected]>
* radeonsi: don't pass si_shader to si_shader_binary_read_configMarek Olšák2016-01-073-17/+19
| | | | Reviewed-by: Nicolai Hähnle <[email protected]>
* radeonsi: add struct si_shader_configMarek Olšák2016-01-075-64/+68
| | | | | | | There will be 1 config per variant, which will be a union of configs from {prolog, main, epilog}. For now, just add the structure. Reviewed-by: Nicolai Hähnle <[email protected]>
* radeonsi: move NULL exporting into a separate functionMarek Olšák2016-01-071-15/+22
| | | | Reviewed-by: Nicolai Hähnle <[email protected]>
* radeonsi: move MRT color exporting into a separate functionMarek Olšák2016-01-071-41/+57
| | | | | | This will be used by a fragment shader epilog. Reviewed-by: Nicolai Hähnle <[email protected]>
* radeonsi: use EXP_NULL for pixel shaders without outputsMarek Olšák2016-01-072-6/+3
| | | | | | This never happens currently. Reviewed-by: Nicolai Hähnle <[email protected]>
* radeonsi: only use LLVMBuildLoad once when updating color outputs at the endMarek Olšák2016-01-071-47/+20
| | | | | | | | | | | without LLVMBuildStore. So: - do LLVMBuildLoad - update the values as necessary - export Reviewed-by: Nicolai Hähnle <[email protected]>
* radeonsi: export "undef" values for undefined PS outputsMarek Olšák2016-01-071-9/+10
| | | | Reviewed-by: Nicolai Hähnle <[email protected]>
* radeonsi: move MRTZ export into a separate functionMarek Olšák2016-01-071-51/+62
| | | | Reviewed-by: Nicolai Hähnle <[email protected]>
* radeonsi: simplify setting the DONE bit for PS exportsMarek Olšák2016-01-072-73/+55
| | | | | | First find out what the last export is and simply set the DONE bit there. Reviewed-by: Nicolai Hähnle <[email protected]>
* radeonsi: set SPI color formats and CB_SHADER_MASK outside of compilationMarek Olšák2016-01-073-16/+28
| | | | Reviewed-by: Nicolai Hähnle <[email protected]>
* radeonsi: write all MRTs only if there is exactly one outputMarek Olšák2016-01-072-4/+5
| | | | | | | | This doesn't fix a known bug, but better safe than sorry. Also, simplify the expression in si_shader.c. Reviewed-by: Nicolai Hähnle <[email protected]>
* radeonsi: determine SPI_SHADER_Z_FORMAT outside of shader compilationMarek Olšák2016-01-073-9/+6
| | | | Reviewed-by: Nicolai Hähnle <[email protected]>
* radeonsi: determine DB_SHADER_CONTROL outside of shader compilationMarek Olšák2016-01-073-28/+40
| | | | | | | because the API pixel shader binary will not emulate alpha test one day, so the KILL_ENABLE bit must be determined elsewhere. Reviewed-by: Nicolai Hähnle <[email protected]>
* tgsi/scan: set which color components are read by a fragment shaderMarek Olšák2016-01-072-8/+23
| | | | | | This will be used by radeonsi. Reviewed-by: Nicolai Hähnle <[email protected]>
* tgsi/scan: fix tgsi_shader_info::reads_zMarek Olšák2016-01-071-2/+3
| | | | | | This has no users in Mesa. Reviewed-by: Nicolai Hähnle <[email protected]>
* tgsi/scan: set if a fragment shader writes sample maskMarek Olšák2016-01-072-0/+3
| | | | | | This will be used by radeonsi. Reviewed-by: Nicolai Hähnle <[email protected]>
* glsl: Disallow vectorization of vector_insert/extract.Kenneth Graunke2016-01-061-0/+2
| | | | | | | | | | | | | | | | | | vector_insert takes a vector, a scalar location, and a scalar value, and produces a new vector with that component updated. As such, it can't be vectorized properly. vector_extract takes a vector and a scalar location, and returns that scalar component of the vector. Vectorization doesn't really make any sense. Treating both as horizontal operations makes sure the vectorizer won't try to touch these. Found by inspection. Signed-off-by: Kenneth Graunke <[email protected]> Reviewed-by: Matt Turner <[email protected]>
* softpipe: tell draw about the vertex layout we wantRoland Scheidegger2016-01-074-58/+105
| | | | | | | | | | | | | | | This makes it more similar to llvmpipe. It also allows us to let draw emit code handle things like getting zeros for non-existing vs outputs automatically. There probably isn't really any overhead either way, there isn't really any "simply copy everything" code in the emit path it would copy each attrib individually just the same. Likewise, we still do another mapping step in softpipe as the layout may still not match exactly (same as in llvmpipe, should probably nuke the pointless mapping in both drivers). This fixes the piglit arb_fragment_layer_viewport no_gs/no_write tests. Reviewed-by: Brian Paul <[email protected]> Reviewed-by: Edward O'Callaghan <[email protected]>
* llvmpipe: use ints not unsigned for slotsRoland Scheidegger2016-01-076-67/+73
| | | | | | | | | | | | | They can't actually be 0 (as position is there) but should avoid confusion. This was supposed to have been done by af7ba989fb5a39925a0a1261ed281fe7f48a16cf but I accidentally pushed an older version of the patch in the end... Also prettify slightly. And make some notes about the confusing and useless fs input "map". Reviewed-by: Jose Fonseca <[email protected]> Reviewed-by: Brian Paul <[email protected]> Reviewed-by: Edward O'Callaghan <[email protected]>
* draw: nuke the interp parameter from vertex_infoRoland Scheidegger2016-01-0711-92/+72
| | | | | | | | | | | | | draw emit couldn't care less what the interpolation mode is... This somehow looked like it would matter, all drivers more or less dutifully filled that in correctly. But this is only used for emit, if draw needs to know about interpolation mode (for clipping for instance) it will get that information from the vs anyway. softpipe actually used to depend on that interpolation parameter, as it abused that structure quite a bit but no longer. Reviewed-by: Brian Paul <[email protected]> Reviewed-by: Edward O'Callaghan <[email protected]>
* softpipe: don't abuse the draw vertex_info struct for something differentRoland Scheidegger2016-01-075-45/+53
| | | | | | | | | | softpipe would calculate two "vertex layouts". The second one was however just used for internal purposes, draw would know nothing about it even though it looked exactly the same as the other one we tell draw about. So, store that information separately as this was just confusing. Reviewed-by: Brian Paul <[email protected]> Reviewed-by: Edward O'Callaghan <[email protected]>
* softpipe: fix mapping of "special" vs outputsRoland Scheidegger2016-01-071-12/+17
| | | | | | | | | | | | | Unlike llvmpipe, softpipe always tells draw to emit the vertices as-is. The two vertex layouts it calculates are a bit confusing, one which is just used to tell draw to emit vertices as-is, and the other which has draw written all over it but draw is completely unaware of and is used only to look up the correct interpolation info later in setup. Thus, the slots used are different to what llvmpipe does (I'm going to clean up the confusing two layout stuff). Reviewed-by: Brian Paul <[email protected]> Reviewed-by: Edward O'Callaghan <[email protected]>
* llvmpipe: scratch some special handling of vp_index/layerRoland Scheidegger2016-01-074-38/+7
| | | | | | | | | It was actually slightly buggy (missing initialization / setup not dependent on new vs albeit I didn't see issues), but the case of non-existing attributes is now handled by draw emit code so don't need that anymore. Reviewed-by: Brian Paul <[email protected]> Reviewed-by: Edward O'Callaghan <[email protected]>
* draw: rework handling of non-existing outputs in emit codeRoland Scheidegger2016-01-073-23/+46
| | | | | | | | | | | | | Previously the code would just redirect requests for attributes which don't exist to use output 0. Rework this to output all zeros instead which seems more useful - in particular some extensions like ARB_fragment_layer_viewport require 0 in the fs even if it wasn't output by previous stages. That way, drivers don't have to special case this depending if the vs/gs outputs some attribute or not. Reviewed-by: Brian Paul <[email protected]> Reviewed-by: Jose Fonseca <[email protected]> Reviewed-by: Edward O'Callaghan <[email protected]>
* mesa: Add KBL PCI IDs and platform information.Sarah Sharp2016-01-062-0/+82
| | | | | | | | | | | | | | | | | | | | | | | | | Add PCI IDs for the Intel Kabylake platforms. The IDs are taken directly from the Linux kernel patches, which are under review: http://lists.freedesktop.org/archives/intel-gfx/2015-October/078967.html http://cgit.freedesktop.org/~vivijim/drm-intel/log/?h=kbl-upstream-v2 The Kabylake PCI IDs taken from the kernel are rearranged to be in order of GT type, then PCI ID. Please note that if this patch is backported, the following fixes will need to be added before this patch: commit 28ed1e08e8ba98e "i965/skl: Remove early platform support" commit c1e38ad37042b0e "i965/skl: Use larger URB size where available." Thanks to Ben for fixing a bug around setting urb.size, and being patient with my questions about what the various fields mean. Signed-off-by: Sarah Sharp <[email protected]> Suggested-by: Ben Widawsky <[email protected]> Tested-by: Rodrigo Vivi <[email protected]> (KBL-GT2) Cc: "11.1" <[email protected]>
* svga: Rename SVGA_HINT_FLAG_DRAW_EMITTEDSinclair Yeh2016-01-066-21/+21
| | | | | | | | Rename SVGA_HINT_FLAG_DRAW_EMITTED to SVGA_HINT_FLAG_CAN_PRE_FLUSH because preemptive flush can be unblocked by more commands than draw. Reviewed-by: Brian Paul <[email protected]>
* svga: allow preemptive flushing on DMA, update, and readback commandsSinclair Yeh2016-01-062-0/+9
| | | | | | | | | | | | | | | The existing code effectively turns off preemptive flushing for all but the regions used for draws. This turns out to be overly restrictive as some memory regions, e.g. GMR, may never get a draw when used as a DMA upload staging area, causing problems for apps that upload a large amount of textures, e.g. Unigine Heaven. This patch fixes the Unigine Heaven memory allocation error and has been verified to not cause a regression in the previous extended retina display issue. Reviewed-by: Thomas Hellstrom <[email protected]> Reviewed-by: Brian Paul <[email protected]>
* svga: skip vertex attribute instruction with zero usage_maskCharmaine Lee2016-01-061-0/+5
| | | | | | | | | | | | | | | | In emit_input_declarations(), we are skipping declarations for those registers that are not being used. But in emit_vertex_attrib_instructions(), we are still emitting instructions to tweak the vertex attributes even if they are not being used. This causes an assert in the backend because an input register is not declared in the shader. This patch fixes the problem by skipping the instruction if the vertex attribute is not being used. Changes in this patch is originated from the code snippet from Jose as suggested in bug 1530161. Tested with piglit, Heaven, Turbine, glretrace. Reviewed-by: Jose Fonseca <[email protected]> Reviewed-by: Brian Paul <[email protected]>
* st/mesa: minor clean-ups in st_atom.cBrian Paul2016-01-061-14/+10
| | | | Remove useless comment. Reformat code.
* st/mesa: replace bitmap size checks with assertionBrian Paul2016-01-061-2/+2
| | | | The _mesa_Bitmap() caller already checks for zero-sized bitmaps.
* st/mesa: check texture target in allocate_full_mipmap()Brian Paul2016-01-061-0/+14
| | | | | | | Some kinds of textures never have mipmaps. 3D textures seldom have mipmaps. Reviewed-by: José Fonseca <[email protected]>
* st/mesa: move mipmap allocation check logic into a functionBrian Paul2016-01-061-12/+42
| | | | | | Better readability and easier to extend. Reviewed-by: José Fonseca <[email protected]>
* main: s/GLuint/GLbitfield for state bitmasksBrian Paul2016-01-062-3/+3
| | | | Reviewed-by: José Fonseca <[email protected]>
* vbo: s/GLuint/GLbitfield/ for state bitmasksBrian Paul2016-01-064-4/+4
| | | | Reviewed-by: José Fonseca <[email protected]>
* st/mesa: use GLbitfield in st_state_flags, add commentsBrian Paul2016-01-061-2/+2
| | | | | | Use GLbitfield instead of GLuint to be consistent with other variables. Reviewed-by: José Fonseca <[email protected]>
* s/GLuint/GLbitfield/ for st_invalidate_state() parameterBrian Paul2016-01-062-2/+2
| | | | | | To match dd_function_table::UpdateState(). Reviewed-by: José Fonseca <[email protected]>
* st/mesa: be more careful about state validation in st_Bitmap()Brian Paul2016-01-061-1/+8
| | | | | | | | | | | If the only dirty state is mesa's _NEW_PROGRAM_CONSTANTS flag, we can skip state validation before drawing a bitmap since that state doesn't effect bitmap rendering. This further increases the performance of the ipers demo on llvmpipe to about what it was before commit 36c93a6fae27561. Reviewed-by: José Fonseca <[email protected]>
* st/mesa: move bitmap cache flushing out of state validationBrian Paul2016-01-066-4/+17
| | | | | | Just do it where needed (before drawing, clearing, etc). Reviewed-by: José Fonseca <[email protected]>
* st/mesa: check state->mesa in early return check in st_validate_state()Brian Paul2016-01-061-1/+1
| | | | | | | | | | | | | | | We were checking the dirty->st flags but not the dirty->mesa flags. When we took the early return, we didn't clear the dirty->mesa flags so the next time we called st_validate_state() we'd often flush the glBitmap cache. And since st_validate_state() is called from st_Bitmap(), it meant we flushed the bitmap cache for every glBitmap() call. This change seems to recover most of the performance loss observed with the ipers demo on llvmpipe since commit commit 36c93a6fae27561. Cc: [email protected] Reviewed-by: José Fonseca <[email protected]>
* st/mesa: protect debug printf() with a conditional instead of commentBrian Paul2016-01-061-5/+5
|
* st/mesa: fix comment indentation in st_flush_bitmap_cache()Brian Paul2016-01-061-2/+2
|
* glsl: fix varying slot allocation for blocks and structs with explicit locationsTimothy Arceri2016-01-071-4/+5
| | | | | | | | | Previously each member was being counted as using a single slot, count_attribute_slots() fixes the count for array and struct members. Also don't assign a negitive to the unsigned expl_location variable. Reviewed-by: Ilia Mirkin <[email protected]>
* glsl: don't try adding built-ins to explicit locations bitmaskTimothy Arceri2016-01-071-1/+3
| | | | | Reviewed-by: Anuj Phogat <[email protected]> Reviewed-by: Edward O'Callaghan <[email protected]>
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-26 10:09:13 +0000