summaryrefslogtreecommitdiffstats
path: root/src/gallium
Commit message (Collapse)AuthorAgeFilesLines
* gallium/radeon: consolidate PIPE_BIND_SHARED/SCANOUT handlingMarek Olšák2017-09-272-13/+4
| | | | Reviewed-by: Nicolai Hähnle <[email protected]>
* radeonsi: remove useless check in si_blit_decompress_color()Samuel Pitoiset2017-09-271-1/+3
| | | | | | | | That's unnecessary to double-check that dcc_offset is not 0 because all callers already check that. Signed-off-by: Samuel Pitoiset <[email protected]> Reviewed-by: Nicolai Hähnle <[email protected]>
* gallium/radeon: more use of vi_dcc_formats_are_incompatible()Samuel Pitoiset2017-09-271-2/+1
| | | | | | | Found by inspection. Signed-off-by: Samuel Pitoiset <[email protected]> Reviewed-by: Nicolai Hähnle <[email protected]>
* swr: Remove unneeeded comparisonGeorge Kyriazis2017-09-261-2/+1
| | | | | | No need to check if screen->pipe != pipe, so we can just assign it. Just do it. Reviewed-by: Bruce Cherniak <[email protected]>
* swr: Handle resource across context changesGeorge Kyriazis2017-09-264-10/+70
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Swr caches fb contents in tiles. Those tiles are stored on a per-context basis. When switching contexts that share resources we need to make sure that the tiles of the old context are being stored and the tiles of the new context are being invalidated (marked as invalid, hence contents need to be reloaded). The context does not get any dirty bits to identify this case. This has to be, then, coordinated by the resources that are being shared between the contexts. Add a "curr_pipe" hook in swr_resource that will allow us to identify a MakeCurrent of the above form during swr_update_derived(). At that time, we invalidate the tiles of the new context. The old context, will need to have already store its tiles by that time, which happens during glFlush(). glFlush() is being called at the beginning of MakeCurrent. So, the sequence of operations is: - At the beginning of glXMakeCurrent(), glFlush() will store the tiles of all bound surfaces of the old context. - After the store, a fence will guarantee that the all tile store make it to the surface - During swr_update_derived(), when we validate the new context, we check all resources to see what changed, and if so, we invalidate the current tiles. Fixes rendering problems with CEI/Ensight. Reviewed-by: Bruce Cherniak <[email protected]>
* gallium: Weaken assertion about u_mm's align2 field.Eric Anholt2017-09-261-1/+4
| | | | | | | | | vc5 MMU mappings are access-controlled at a 128kb boundary, so the 4kb here was too small for that purpose. Allowing any valid align2 value that u_mm's 32-bit addressing can represent will still catch most cases of people passing in a byte alignment. Reviewed-by: Nicolai Hähnle <[email protected]>
* broadcom/vc4: Fix infinite retry in vc4_bo_alloc()Boris Brezillon2017-09-261-6/+4
| | | | | | | | | | | | | | | cleared_and_retried is always reset to false when jumping to the retry label, thus leading to an infinite retry loop. Fix that by moving the cleared_and_retried variable definitions at the beginning of the function. While we're at it, move the create variable with the other local variables and explicitly reset its content in the retry path. Signed-off-by: Boris Brezillon <[email protected]> Reviewed-by: Eric Anholt <[email protected]> Reviewed-by: Eric Engestrom <[email protected]> Fixes: 78087676c98aa8884ba92 "vc4: Restructure the simulator mode."
* broadcom/vc4: Keep pipe_sampler_view->texture matching the original texture.Eric Anholt2017-09-266-39/+42
| | | | | | | | | | | I was overwriting view->texture with the shadow resource when we need to do shadow copies (retiling or baselevel rebase), but that tripped up some critical new sanity checking in state_tracker (making sure that stObj->pt hasn't changed from view->texture through TexImage-related paths). To avoid that, move the shadow resource to the vc4_sampler_view struct. Fixes: f0ecd36ef8e1 ("st/mesa: add an entirely separate codepath for setting up buffer views")
* svga: silence unused var warning in optimized build with MAYBE_UNUSEDBrian Paul2017-09-261-1/+1
| | | | Trivial
* r600: fork and import gallium/radeonMarek Olšák2017-09-2665-977/+15239
| | | | | | | | | | | This marks the end of code sharing between r600 and radeonsi. It's getting difficult to work on radeonsi without breaking r600. A lot of functions had to be renamed to prevent linker conflicts. There are also minor cleanups. Acked-by: Dave Airlie <[email protected]> Reviewed-by: Nicolai Hähnle <[email protected]>
* swr/rast: Handle instanceID offset / Instance Stride enableTim Rowley2017-09-251-7/+39
| | | | | | | | | | Supported in JitGatherVertices(); FetchJit::JitLoadVertices() may require similar changes, will need address this if it is determined that this path is still in use. Handle Force Sequential Access in FetchJit::Create. Reviewed-by: Bruce Cherniak <[email protected]>
* swr/rast: Remove code supporting legacy llvm (<3.9)Tim Rowley2017-09-253-105/+15
| | | | Reviewed-by: Bruce Cherniak <[email protected]>
* swr/rast: Fix allocation of DS output data for USE_SIMD16_FRONTENDTim Rowley2017-09-251-10/+6
| | | | Reviewed-by: Bruce Cherniak <[email protected]>
* swr/rast: Slightly more efficient blend jitTim Rowley2017-09-251-20/+10
| | | | Reviewed-by: Bruce Cherniak <[email protected]>
* swr/rast: Properly sized null GS bufferTim Rowley2017-09-251-1/+1
| | | | Reviewed-by: Bruce Cherniak <[email protected]>
* swr/rast: Move SWR_GS_CONTEXT from thread local storage to stackTim Rowley2017-09-251-12/+11
| | | | | | | Move structure, as the size is significantly reduced due to dynamic allocation of the GS buffers. Reviewed-by: Bruce Cherniak <[email protected]>
* swr/rast: Fetch compile state changesTim Rowley2017-09-252-1/+12
| | | | | | | Add ForceSequentialAccessEnable and InstanceIDOffsetEnable bools to FETCH_COMPILE_STATE. Reviewed-by: Bruce Cherniak <[email protected]>
* swr/rast: New GS state/context APITim Rowley2017-09-253-212/+253
| | | | | | | One piglit regression, which was a false pass: [email protected]@execution@geometry@dynamic_input_array_index Reviewed-by: Bruce Cherniak <[email protected]>
* swr/rast: Add support for R10G10B10_FLOAT_A2_UNORM pixel formatTim Rowley2017-09-253-17/+28
| | | | Reviewed-by: Bruce Cherniak <[email protected]>
* st/va/postproc: implement the DRM prime grabberLeo Liu2017-09-251-2/+16
| | | | Acked-by: Christian König <[email protected]>
* vl/compositor: convert RGB buffer to YUV with color conversionLeo Liu2017-09-252-0/+81
| | | | Acked-by: Christian König <[email protected]>
* vl/csc: add a RGB to YUV CSC matrixLeo Liu2017-09-252-1/+11
| | | | Acked-by: Christian König <[email protected]>
* vl/compositor: create RGB to YUV fragment shaderLeo Liu2017-09-252-0/+51
| | | | Acked-by: Christian König <[email protected]>
* st/va/postproc: use progressive target buffer for scalingLeo Liu2017-09-251-0/+18
| | | | | | | | | Scaling between interlaced buffers, esp. for scale-up, because blit will scale up top filed and bottom field separately. it'll result in the weaving for these buffer with lack of accuracy. So use shader deint for the case. Acked-by: Christian König <[email protected]>
* st/va: make internal func vlVaHandleSurfaceAllocate() call simplerLeo Liu2017-09-253-7/+4
| | | | Acked-by: Christian König <[email protected]>
* st/va/postproc: add a full NV12 deint support from buffer I to PLeo Liu2017-09-251-1/+21
| | | | | | | | | | | Before it's impossible to transcode an interlaced video, becasue if in order for encoder to work, we have to force buffer to progessive, but the deint with buffer from I to P is missing. Now along With the new YUV deint full function, it works with weave and bob deint. Also this will benefit transcoding video with scaling parameters. Acked-by: Christian König <[email protected]>
* vl/compositor: add Bob top and bottom to YUV deint functionLeo Liu2017-09-251-6/+28
| | | | Acked-by: Christian König <[email protected]>
* vl/compositor: remove vl_compositor_yuv_deint() functionLeo Liu2017-09-252-40/+0
| | | | | | No longer used. Acked-by: Christian König <[email protected]>
* st/va: use new vl_compositor_yuv_deint_full() to deintLeo Liu2017-09-251-3/+11
| | | | | | | We also set src rectangle explicitly just in case of the mismatch of size between interlaced buffer and progressive buffer Acked-by: Christian König <[email protected]>
* st/omx: use new vl_compositor_yuv_deint_full() to deintLeo Liu2017-09-251-2/+9
| | | | | | v2: add dst rect to make sure no scale Acked-by: Christian König <[email protected]>
* vl/compositor: add a new function for YUV deintLeo Liu2017-09-252-0/+42
| | | | | | | It will replace previous deint function with abilities of scaling and field deinterlacing Acked-by: Christian König <[email protected]>
* vl/compositor: extend YUV deint function to do field deintLeo Liu2017-09-252-12/+26
| | | | | | It will add Bob deint ability to interlaced video for HW encoder Acked-by: Christian König <[email protected]>
* vl/compositor: separate YUV part from shader video buffer functionLeo Liu2017-09-251-13/+18
| | | | | | So that it can be re-used Acked-by: Christian König <[email protected]>
* st/va/postproc: use video original size for postprocessingLeo Liu2017-09-251-6/+8
| | | | | | | Otherwise the aligned size will make video scaled Cc: [email protected] Reviewed-by: Christian König <[email protected]>
* scons: use python3-compatible generatorEric Engestrom2017-09-251-4/+2
| | | | | | | These changes were generated using python's `2to3` tool. Suggested-by: Ilia Mirkin <[email protected]> Signed-off-by: Eric Engestrom <[email protected]>
* scons: use python3-compatible print()Eric Engestrom2017-09-254-6/+6
| | | | | | | | | These changes were generated using python's `2to3` tool. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=102852 Reported-by: Alex Granni <[email protected]> Signed-off-by: Eric Engestrom <[email protected]> Reviewed-by: Jose Fonseca <[email protected]>
* etnaviv: Add missing includes after 6ace0b8Wladimir J. van der Laan2017-09-221-0/+2
| | | | | | | | | | Add missing includes after 6ace0b8 (etnaviv: don't enable RT full-overwrite when logicop is enabled), otherwise the etnaviv driver won't build because of missing macros. Signed-off-by: Wladimir J. van der Laan <[email protected]> Reviewed-by: Christian Gmeiner <[email protected]> Tested-by: Andres Gomez <[email protected]>
* etnaviv: fix 16bpp clearsLucas Stach2017-09-221-1/+1
| | | | | | | | | | | | | | | util_pack_color may leave undefined values in the upper half of the packed integer. As our hardware needs the upper 16 bits to mirror the lower 16bits, this breaks clears of those formats if the undefined values aren't masked off. I've only observed the issue with R5G6B5_UNORM surfaces, other 16bpp formats seem to work fine. Fixes: d6aa2ba2b2 (etnaviv: replace translate_clear_color with util_pack_color) Cc: [email protected] Signed-off-by: Lucas Stach <[email protected]> Reviewed-by: Wladimir J. van der Laan <[email protected]> Reviewed-by: Christian Gmeiner <[email protected]>
* swr/rast: remove llvm fence/atomics from generated filesTim Rowley2017-09-221-0/+8
| | | | | | | | | | | We currently don't use these instructions, and since their API changed in llvm-5.0 having them in the autogen files broke the mesa release tarballs which ship with generated autogen files. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=102847 CC: [email protected] Tested-by: Laurent Carlier <[email protected]> Reviewed-by: Bruce Cherniak <[email protected]>
* etnaviv: don't enable RT full-overwrite when logicop is enabledLucas Stach2017-09-222-6/+14
| | | | | | | | | Logicop is a form of blending with the framebuffer, so we must allow framebuffer reads when logicop is enabled. Fixes: piglit gl-1.0-logicop on GC3000, which has logicop support Signed-off-by: Lucas Stach <[email protected]>
* gallium/util: Remove unused keymapThomas Helland2017-09-213-388/+0
| | | | | | | | | | This is not used anywhere in the codebase. It's a hashtable implementation that is based around cso_hash, and is therefore (and as mentioned in a comment in the source) quite similar to u_hash_table. CC: Brian Paul<[email protected]> Reviewed-by: Marek Olšák <[email protected]>
* gallium: Add PIPE_SHADER_CAP_INT64_ATOMICSJan Vesely2017-09-2116-0/+17
| | | | | | | Denotes availability of 64bit int atomic instructions Signed-off-by: Jan Vesely <[email protected]> Reviewed-by: Marek Olšák <[email protected]>
* clover: Wait for requested operation if blocking flag is setJan Vesely2017-09-201-2/+28
| | | | | | | | | | v2: wait in map_buffer and map_image as well v3: use event::wait instead of wait (skips fence wait for hard_event) v4: use wait_signalled() Signed-off-by: Jan Vesely <[email protected]> Reviewed-by: Francisco Jerez <[email protected]> Reviewed-by: Aaron Watry <[email protected]>
* clover: Run the associated action before an event is signalled.Francisco Jerez2017-09-202-11/+12
| | | | | | | | | | | | | | | | And define a method for other threads to wait until the action function associated with an event has been executed to completion. For hard events, this will mean waiting until the corresponding command has been submitted to the pipe driver, without necessarily flushing the pipe_context and waiting for the actual command to be processed by the GPU (which is what hard_event::wait() already does). This weaker kind of event wait will allow implementing blocking memory transfers efficiently. Acked-by: Aaron Watry <[email protected]> Reviewed-by: Jan Vesely <[email protected]>
* clover: Wrap event::wait_count in a method taking care of the required locking.Francisco Jerez2017-09-202-8/+14
| | | | | Acked-by: Aaron Watry <[email protected]> Reviewed-by: Jan Vesely <[email protected]>
* llvmpipe, gallivm: implement lod queries (LODQ opcode)Roland Scheidegger2017-09-205-58/+144
| | | | | | | | | | | | | | | | | | | | | | This uses all the existing code to calculate lod values for mip linear filtering. Though we'll have to disable the simplifications (if we know some parts of the lod calculation won't actually matter for filtering purposes due to mip clamps etc.). For better or worse, we'll also disable lod calculation hacks (mostly should make a difference for cube maps) always - the issue with per-pixel lod being difficult is mostly because we then have different mipmaps needed for the actual texel fetch, which isn't a problem with lodq. We still use approximation for the log2 - for that reason I believe the float part of the lod is only accurate to about 4-5 bits (and one bit less with 1d textures actually) which is hopefully good enough (though d3d10 technically requires 6 bits - could use quadratic interpolation instead of linear to get 8 bits or so). Since lodq requires unclamped lod, we also have to move some sampler key calculations to texture sampling code - even if we know we're going to access mipmap 0 we still have to calculate lod and apply lod_bias for lodq. Passes piglit ARB_texture_query_lod tests (after having fixed the test). Reviewed-by: Jose Fonseca <[email protected]>
* radeonsi: set MIP_POINT_PRECLAMP to 0Nicolai Hähnle2017-09-201-1/+1
| | | | | | | | | | | | | | | | | This fixes a bug with nearest ("point") mip selection when the fractional part of max_lod is in (0.5,1). In this case, the spec mandates that we still select the mip level ceil(max_lod) in the clamping case. However, MIP_POINT_PRECLAMP will clamp before the mip selection, which is wrong. Supposedly this setting was originally copied from the closed Vulkan driver, but as far as I can tell, closed Vulkan was actually changed back recently :) Fixes dEQP-GLES3.functional.texture.mipmap.2d.max_lod.{nearest,linear}_nearest Fixes: f7420ef5b464 ("radeonsi: enable some sampler fields to match the closed driver") Reviewed-by: Marek Olšák <[email protected]> Reviewed-by: Dave Airlie <[email protected]>
* radeonsi: fix array textures layer coordinateNicolai Hähnle2017-09-201-1/+10
| | | | | | | | | Like for cube map (array) gather, we need to round to nearest on <= VI. Fixes tests in dEQP-GLES3.functional.shaders.texture_functions.texture.* Cc: [email protected] Reviewed-by: Marek Olšák <[email protected]>
* etnaviv: move sw query defines to etnaviv_query_sw.hChristian Gmeiner2017-09-202-2/+4
| | | | | | | Also add new define ETNA_SW_QUERY_BASE. Signed-off-by: Christian Gmeiner <[email protected]> Reviewed-by: Wladimir J. van der Laan <[email protected]>
* etnaviv: move sw get_driver_query_info(..)Christian Gmeiner2017-09-203-12/+28
| | | | | | | | | This change makes etna_get_driver_query_info(..) more generic and puts the knowledge of supported queries directly besides the implementation. Signed-off-by: Christian Gmeiner <[email protected]> Reviewed-by: Wladimir J. van der Laan <[email protected]>
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-22 15:00:10 +0000