summaryrefslogtreecommitdiffstats
path: root/src/gallium
Commit message (Collapse)AuthorAgeFilesLines
* nvc0: invalidate textures/samplers on GK104+Samuel Pitoiset2016-08-242-12/+22
| | | | | | | | | | | | | | Like Fermi, textures and samplers are aliased between 3D and compute, especially the TIC_FLUSH/TSC_FLUSH methods and we have to re-validate these resources when switching between the two pipelines. This fixes a GPU hang with Elemental (and most likely with other UE4 demos). Tested on GK107 and GM107. Signed-off-by: Samuel Pitoiset <[email protected]> Reviewed-by: Ilia Mirkin <[email protected]> CC: <[email protected]>
* gallium/ttn: Remove duplicated TGSI_OPCODE_DP2A initializationRhys Kidd2016-08-241-1/+0
| | | | | | | | | Duplicate line is currently on 1535. Identified by Clang, when run through Eric Anholt's Travis harness. Signed-off-by: Rhys Kidd <[email protected]> Reviewed-by: Eric Anholt <[email protected]>
* vc4: Fix GPU hangs with >16 varying values.Eric Anholt2016-08-242-19/+68
| | | | Fixes glsl-routing in piglit and hangs in glbenchmark 2.0.2.
* vl/rbsp: fix another three byte not detectedLeo Liu2016-08-241-1/+1
| | | | | | | | | | | | This happens when three byte "00 00 03" is partly loaded to vlc->buffer, thus at the bottom of buffer with valid bits is "00" or "00 00" and left like "00 03" or "03" in the data, so that it will not be detected by three byte emulation check. The reason for that is the escaped bit was set to 0 from the rbsp init. Signed-off-by: Leo Liu <[email protected]> Acked-by: Christian König <[email protected]>
* radeonsi: fix VM faults due NULL internal const buffers on CIKMarek Olšák2016-08-241-2/+11
| | | | | | | | They are harmless, but the interrupts do decrease performance. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=97039 Cc: 12.0 <[email protected]>
* gallium/winsys/kms: Look up the GEM handle after importing a prime FDTomasz Figa2016-08-241-0/+4
| | | | | | | | | | | | | | | | | | | | drmPrimeHandleToFD() will return the same GEM handle every time the same buffer is imported, even from a different prime FD. Since GEM handles are not reference counted, we need to make sure that each GEM handle is referenced only by one display target struct, by looking it up in kms_sw->bo_list first and bumping the refcount of the found dt on hit and falling back to creating a new dt only on miss. v2: Split into separate function. Use helper function for lookup. v3 [Emil Velikov]: Rename kms_sw_displaytarget_{lookup,find_and_ref} (Jordan) Signed-off-by: Tomasz Figa <[email protected]> CC: <[email protected]> Reviewed-by: Hans de Goede <[email protected]> (v2) Signed-off-by: Emil Velikov <[email protected]>
* gallium/winsys/kms: Move display target handle lookup to separate functionTomasz Figa2016-08-241-9/+24
| | | | | | | | | | | | | | | | | As a preparation to use the lookup in more than once place, move the code that looks up given KMS/GEM handle to a separate function. This change should not introduce any functional changes. v2: Split into separate patch. Move lookup code into separate function. v3 [Emil Velikov]: Rename kms_sw_displaytarget_{lookup,find_and_ref} (Jordan) Signed-off-by: Tomasz Figa <[email protected]> CC: <[email protected]> Reviewed-by: Hans de Goede <[email protected]> (v2) Signed-off-by: Emil Velikov <[email protected]>
* gallium/winsys/kms: Fully initialize kms_sw_dt at prime import time (v2)Tomasz Figa2016-08-241-7/+11
| | | | | | | | | | | | | | | Currently kms_sw_displaytarget_add_from_prime() allocates the struct and fills in only some of the fields, resulting in a half-baked struct that needs to be further completed by the caller. To make this a bit more consistent, pass width, height and stride to this function and fill in everything there, so that caller can take the returned struct as is. v2: Split from one big patch into four fixing one thing at a time. Signed-off-by: Tomasz Figa <[email protected]> CC: <[email protected]> Reviewed-by: Hans de Goede <[email protected]> Reviewed-by: Emil Velikov <[email protected]>
* gallium/winsys/kms: Fix double refcount when importing from prime FD (v2)Tomasz Figa2016-08-241-1/+0
| | | | | | | | | | | | | Currently the code creates a display target struct with refcount field initialized to 1 and then the caller again increments it, leading to a leaked reference. Let's remove the unnecessary increment. v2: Split from one big patch into four fixing one thing at a time. Signed-off-by: Tomasz Figa <[email protected]> CC: <[email protected]> Reviewed-by: Hans de Goede <[email protected]> Reviewed-by: Emil Velikov <[email protected]>
* gallium/cso: avoid unnecessary null dereferenceEric Engestrom2016-08-241-1/+1
| | | | | | | | | | | The label `out:` calls `destroy()` which dereferences `ctx`. This is unnecessary as there is nothing to destroy. Immediately return instead. CovID: 1258255 Signed-off-by: Eric Engestrom <[email protected]> Reviewed-by: Nicolai Hähnle <[email protected]> Reviewed-by: Marek Olšák <[email protected]>
* st/xvmc: fix a couple 'unused-but-set-variable' warningsEric Engestrom2016-08-241-2/+3
| | | | | | Signed-off-by: Eric Engestrom <[email protected]> Reviewed-by: Christian König <[email protected]> Reviewed-by: Emil Velikov <[email protected]>
* st/va: Remove unused variable coded_size from vlVaEndPicture()Kai Wasserbäch2016-08-241-1/+0
| | | | | | | | | | | | Removes the following GCC warning: ../../../../../src/gallium/state_trackers/va/picture.c:542:17: warning: unused variable 'coded_size' [-Wunused-variable] unsigned int coded_size; ^~~~~~~~~~ Signed-off-by: Kai Wasserbäch <[email protected]> Reviewed-by: Christian König <[email protected]> Reviewed-by: Boyuan Zhang <[email protected]>
* st/va: Remove else case in vlVaEndPicture() made superfluous by c59628d11bKai Wasserbäch2016-08-241-3/+1
| | | | | | | | | | Commit c59628d11b134fc016388a170880f7646e100d6f made the else statement and duplication of the context->decoder->end_frame() call superfluous. Cc: Boyuan Zhang <[email protected]> Signed-off-by: Kai Wasserbäch <[email protected]> Reviewed-by: Christian König <[email protected]> Reviewed-by: Boyuan Zhang <[email protected]>
* st/va: add missing mutex_unlockEric Engestrom2016-08-241-1/+3
| | | | | | | | Fixes: c59628d11b134fc01638 ("st/va: enable dual instances encode by sync surface") Signed-off-by: Eric Engestrom <[email protected]> Reviewed-by: Boyuan Zhang <[email protected]> Reviewed-by: Christian König <[email protected]>
* st/dri: respect driver's request to avoid mixed color/depth bit configsIlia Mirkin2016-08-231-2/+6
| | | | | Signed-off-by: Ilia Mirkin <[email protected]> Reviewed-by: Marek Olšák <[email protected]>
* gallium: add a cap to expose whether driver supports mixed color/zs bitsIlia Mirkin2016-08-2317-0/+22
| | | | | | | | | | Some hardware can't render to color/depth buffers of mixed bitness. When that happens a fallback has to happen, but this allows the driver to express that this isn't an optimal scenario. The purpose of this is to remove such fbconfigs from the GLX/EGL config list. Signed-off-by: Ilia Mirkin <[email protected]> Reviewed-by: Marek Olšák <[email protected]>
* dri: add a way to request that modes have matching color/zs depthsIlia Mirkin2016-08-231-2/+2
| | | | | | | | | | Some GPUs, notably nv3x/nv4x can't render to mismatched color/zs framebuffer depths. Fallbacks can be done by the driver, with shadow surfaces, but no reason to encourage applications to select non-matching glx visuals. Signed-off-by: Ilia Mirkin <[email protected]> Reviewed-by: Marek Olšák <[email protected]>
* nv50/ir: make sure cfg iterator always hits all blocksIlia Mirkin2016-08-231-4/+4
| | | | | | | | | | | | In some very specially-crafted cases, we could attempt to visit a node that has already been visited, and then run out of bb's to visit, while there were still cross blocks on the list. Make sure that those get moved over in that case. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=96274 Signed-off-by: Ilia Mirkin <[email protected]> Reviewed-by: Samuel Pitoiset <[email protected]> Cc: [email protected]
* vc4: Tell state_tracker that we would prefer NIR.Eric Anholt2016-08-223-8/+31
| | | | | | | | | | Before this series, the code generation path was: GLSL IR -> TGSI -> NIR -> NIR clone -> QIR -> QPU Now it's (generally) GLSL IR -> NIR -> NIR clone -> QIR -> QPU
* vc4: Use proper type sizes for uniforms.Eric Anholt2016-08-221-4/+5
|
* vc4: Add VARYING_SLOT_PNTC support.Eric Anholt2016-08-221-4/+5
| | | | We end up with this when doing GLSL-to-NIR.
* vc4: Fix vc4_nir_lower_io for non-vec4 I/O.Eric Anholt2016-08-221-22/+12
| | | | | To support GLSL-to-NIR, we need to be able to support actual float/vec2/vec3 varyings.
* nir: Define system values for vc4's blending-lowering arguments.Eric Anholt2016-08-224-46/+54
| | | | | | | | | | | | | In the GLSL-to-NIR conversion of VC4, I had a bit of trouble with what I was calling the "state uniforms" that I was putting into the NIR fighting with its other lowering passes. Instead of using magic uniform base numbers in the backend, follow the lead of load_user_clip_plane and just define system values for them. v2: Fix unintended change to channel_num, drop unspecified const_index value on blend_const_color_r_float. Reviewed-by: Kenneth Graunke <[email protected]>
* gallium/hud: round max_value to print nicely rounded numbers next to graphsMarek Olšák2016-08-223-4/+80
| | | | | | This improves readability a lot. Reviewed-by: Brian Paul <[email protected]>
* gallium/hud: generalize code for drawing numbers next to graphsMarek Olšák2016-08-221-5/+8
| | | | Reviewed-by: Brian Paul <[email protected]>
* gallium/hud: draw numbers with 3 decimal places if those aren't 0Marek Olšák2016-08-221-5/+12
| | | | Reviewed-by: Brian Paul <[email protected]>
* gallium/hud: use sRGB for nicer AA linesMarek Olšák2016-08-221-0/+21
| | | | Reviewed-by: Brian Paul <[email protected]>
* gallium/hud: use AA lines for graphsMarek Olšák2016-08-221-1/+5
| | | | | | this looks a lot better (with the next patch) Reviewed-by: Brian Paul <[email protected]>
* gallium/hud: don't enable blending for all objectsMarek Olšák2016-08-221-2/+7
| | | | Reviewed-by: Brian Paul <[email protected]>
* llvmpipe: fix issues with depth clampRoland Scheidegger2016-08-203-49/+94
| | | | | | | | | | | | | | | | We only did depth clamp when the value was written from the fs. This is very wrong both for d3d10 and GL, and only passed the corresponding piglit test due to pure luck (it no longer does with the enhanced test). Also, interpolation clamped values to 1.0 always, which can legitimately happen if depth clip is disabled, so fix that as well (untested). There is one unresolved issue left, d3d10 always does depth clamping, whereas GL does not (but does [0,1] clamp instead for fs depth outputs) - this information isn't in any gallium state object, leave it as-is for now (though it looks like llvmpipe misses the [0,1] clamp as well). This (with the previous patch) fixes piglit depth-clamp-range test. Reviewed-by: Jose Fonseca <[email protected]>
* llvmpipe: fix depth clamping wrt reversed near/far valuesRoland Scheidegger2016-08-201-9/+3
| | | | | | | | | | | This wasn't handled before (the result was that no matter what value got clamped, it always ended up as the near value in this case) (if clamping actually happened). Fix this by using the util helper for that (the math is otherwise "mostly" the same, mostly because there could actually be differences due to float rounding, but I don't even know which one would be more correct). Reviewed-by: Jose Fonseca <[email protected]>
* a4xx: make sure to actually clamp depth as requestedIlia Mirkin2016-08-192-2/+29
| | | | | | | | | | | We were previously ... not clamping. I guess this meant that everything got clamped to 1/0, which was enough to pass the existing tests. Or perhaps the clamping would only happen to the rasterized depth value and not the frag shader's output depth value. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=97231 Signed-off-by: Ilia Mirkin <[email protected]> Cc: [email protected]
* a4xx: only disable depth clipping, not all clipping, when requestedIlia Mirkin2016-08-192-1/+4
| | | | | | | | | | The previous bit disables the whole clipper, including the regular viewport-related clipping that would go on. The two new bits disable near and far clipping (separately, as verified with the depth-clamp-range piglit). Signed-off-by: Ilia Mirkin <[email protected]> Cc: [email protected]
* vc4: Switch store_output to using nir_lower_io_to_scalar / component.Eric Anholt2016-08-192-44/+16
|
* vc4: Use the intrinsic's first_component for vattr VPM index.Eric Anholt2016-08-192-7/+3
| | | | Avoids another multiplication by 4 of the base in the NIR.
* vc4: Convert to using nir_lower_io_scalar for FS inputs.Eric Anholt2016-08-192-44/+62
| | | | | The scalarizing of FS inputs can be done in a non-driver-dependent manner, so extract it out of the driver.
* vc4: Switch to using the intrinsic accessors.Eric Anholt2016-08-193-23/+29
| | | | | The const_index[] values have always felt magic, and this documents them a bit better.
* ttn: Use nir_load_front_face instead of the TGSI-style input.Eric Anholt2016-08-193-63/+28
| | | | | | | This reduces the diff between GLSL-to-NIR and TGSI-to-NIR, and gives NIR more optimization to work on. Reviewed-by: Kenneth Graunke <[email protected]>
* ttn: Make FRAG_RESULT_DEPTH be a float variable to match gtn and ptn.Eric Anholt2016-08-194-11/+17
| | | | | | | This lets TTN-using drivers handle FRAG_RESULT_DEPTH the same between all their source paths. Reviewed-by: Rob Clark <[email protected]>
* vc4: Dump the TGSI before trying to convert it to NIR.Eric Anholt2016-08-191-4/+3
| | | | In the case of debugging a crash in TTN, this is nice to have.
* radeon/vce: set flag based on dual instance enablementBoyuan Zhang2016-08-191-2/+4
| | | | | | | Set the flag on when dual instance encoding is supported, otherwise set it to off. Signed-off-by: Boyuan Zhang <[email protected]>
* st/va: enable dual instances encode by sync surfaceBoyuan Zhang2016-08-193-3/+46
| | | | | | | | | | | This patch improves the performance of Vaapi Encode by enabling dual instances encoding. flush function is not called after each end_frame call. radeon/vce will do flush whenever 2 frames are submitted for encoding. Implement sync surface function to flush only if the frame hasn't been flushed yet. Signed-off-by: Boyuan Zhang <[email protected]> Reviewed-by: Christian König <[email protected]>
* radeonsi: initialize and finalize the LLVM function pass managerMarek Olšák2016-08-181-0/+2
| | | | Reviewed-by: Tom Stellard <[email protected]>
* swr: [rasterizer core] only use Viewport/Scissors during SwrDraw* operationsTim Rowley2016-08-1712-415/+400
| | | | | | | | | | | Add explicit rects for: - SwrClearRenderTarget - SwrDiscardRect - SwrInvalidateTiles - SwrStoreTiles Signed-off-by: Tim Rowley <[email protected]>
* swr: [rasterizer common] reorder SWR_FORMAT_INFOTim Rowley2016-08-172-825/+1433
| | | | Signed-off-by: Tim Rowley <[email protected]>
* swr: [rasterizer core] make dirtytile list point directly to macrotilequeuesTim Rowley2016-08-173-14/+15
| | | | | | Speeds up high geometry HPC workloads. Signed-off-by: Tim Rowley <[email protected]>
* swr: [rasterizer core] portability - remove use of INT64Tim Rowley2016-08-171-2/+2
| | | | Signed-off-by: Tim Rowley <[email protected]>
* swr: [rasterizer core] viewport transform disabled fixTim Rowley2016-08-171-4/+11
| | | | | | | When viewport transform is disabled (ie. screen space coords are passed in directly), the W component should be interpreted as RHW. Signed-off-by: Tim Rowley <[email protected]>
* swr: [rasterizer core] clamp scissor rects to current tile rectTim Rowley2016-08-171-0/+18
| | | | Signed-off-by: Tim Rowley <[email protected]>
* swr: [rasterizer core] align stats structuresTim Rowley2016-08-171-2/+2
| | | | Signed-off-by: Tim Rowley <[email protected]>
generated by cgit v1.2.3 (git 2.39.1) at 2024-12-15 02:23:56 +0000