summaryrefslogtreecommitdiffstats
Commit message (Collapse)AuthorAgeFilesLines
* i965: Use predicate enable bit for conditional rendering w/o stallingNeil Roberts2015-05-129-12/+240
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Previously whenever a primitive is drawn the driver would call _mesa_check_conditional_render which blocks waiting for the result of the query to determine whether to render. On Gen7+ there is a bit in the 3DPRIMITIVE command which can be used to disable the primitive based on the value of a state bit. This state bit can be set based on whether two registers have different values using the MI_PREDICATE command. We can load these two registers with the pixel count values stored in the query begin and end to implement conditional rendering without stalling. Unfortunately these two source registers were not in the whitelist of available registers in the kernel driver until v3.19. This patch uses the command parser version from intel_screen to detect whether to attempt to set the predicate data registers. The predicate enable bit is currently only used for drawing 3D primitives. For blits, clears, bitmaps, copypixels and drawpixels it still causes a stall. For most of these it would probably just work to call the new brw_check_conditional_render function instead of _mesa_check_conditional_render because they already work in terms of rendering primitives. However it's a bit trickier for blits because it can use the BLT ring or the blorp codepath. I think these operations are less useful for conditional rendering than rendering primitives so it might be best to leave it for a later patch. v2: Use the command parser version to detect whether we can write to the predicate data registers instead of trying to execute a register load command. v3: Simple rebase v4: Changes suggested by Kenneth Graunke: Split the load_64bit_register function out to a separate patch so it can be a shared public function. Avoid calling _mesa_check_conditional_render if we've already determined that there's no query object. Some styling fixes. Reviewed-by: Kenneth Graunke <[email protected]>
* i956: Add a function to load a 64-bit register from a bufferNeil Roberts2015-05-122-14/+46
| | | | | | | | | | | | Adds brw_load_register_mem64 which is similar to brw_load_register_mem except that it queues two GEN7_MI_LOAD_REGISTER_MEM commands in order to load both halves of a 64-bit register. The function is implemented by splitting the 32-bit version into an internal helper function which takes a size. This will later be used to set the 64-bit predicate source registers. Reviewed-by: Kenneth Graunke <[email protected]>
* i965: Store the command parser version number in intel_screenNeil Roberts2015-05-122-1/+14
| | | | | | | | | In order to detect whether the predicate source registers can be used in a later patch we will need to know the version number for the command parser. This patch just adds a member to intel_screen and does an ioctl to get the version. Reviewed-by: Kenneth Graunke <[email protected]>
* docs/GL3: (trivial) mark some tf extensions as done for softpipe/llvmpipeRoland Scheidegger2015-05-121-3/+3
| | | | Those extensions were enabled for ages already.
* docs: add news item and link release notes for mesa 10.5.5Emil Velikov2015-05-112-0/+7
| | | | Signed-off-by: Emil Velikov <[email protected]>
* docs: Add sha256 sums for the 10.5.5 releaseEmil Velikov2015-05-111-1/+2
| | | | | Signed-off-by: Emil Velikov <[email protected]> (cherry picked from commit 8ee1a1c08b168d7583b806a2f8a2dc2ae28be62a)
* Add release notes for the 10.5.5 releaseEmil Velikov2015-05-111-0/+94
| | | | | Signed-off-by: Emil Velikov <[email protected]> (cherry picked from commit d88fb4050561a62fa824bec59ffedf2a826c2083)
* st/mesa: make sure to create a "clean" bool when doing i2bIlia Mirkin2015-05-111-1/+1
| | | | | | | | | | | | i2b has to work for all integers, not just 1. INEG would not necessarily result with all bits set, which is something that other operations can rely on by e.g. using AND (or INEG for b2i). Signed-off-by: Ilia Mirkin <[email protected]> Reviewed-by: Jason Ekstrand <[email protected]> Reviewed-by: Marek Olšák <[email protected]> Reviewed-by: Roland Scheidegger <[email protected]> Cc: [email protected]
* clover: Fix a bug with multi-threaded events v2Tom Stellard2015-05-112-3/+17
| | | | | | | | | | | | | | | | | | | | It was possible for some events never to get triggered if one thread was creating events and another threads was waiting for them. This patch consolidates soft_event::wait() and hard_event::wait() into event::wait() so that hard_event objects will now wait for all their dependencies to be submitted before flushing the command queue. v2: - Rename variables - Use mutable varibales so we can keep event::wait() const - Open code signalled() call so mutex can be atted to signalled without deadlocking. CC: 10.5 <[email protected]> Reviewed-by: Francisco Jerez <[email protected]>
* clover: Add a mutex to guard queue::queued_eventsTom Stellard2015-05-112-0/+4
| | | | | | | | | | | | This fixes a potential crash where on a sequence like this: Thread 0: Check if queue is not empty. Thread 1: Remove item from queue, making it empty. Thread 0: Do something assuming queue is not empty. CC: 10.5 <[email protected]> Reviewed-by: Francisco Jerez <[email protected]>
* i965/fs: Add missing initializer in fs_visitor().Matt Turner2015-05-111-1/+1
|
* egl: Remove skeleton implementation of EGL_MESA_screen_surfaceAdam Jackson2015-05-1113-1242/+0
| | | | | | | | No backend wires this up to anything, and the extension spec has been marked obsolete for 4+ years. Reviewed-by: Marek Olšák <[email protected]> Signed-off-by: Adam Jackson <[email protected]>
* egl/swrast: Enable config extension for swrastAxel Davy2015-05-113-14/+15
| | | | | | | | Enables to use dri config for swrast, like vblank_mode. Reviewed-by: Dave Airlie <[email protected]> Signed-off-by: Axel Davy <[email protected]>
* egl/wayland: Implement swrast supportAxel Davy2015-05-113-17/+696
| | | | | | | Reviewed-by: Dave Airlie <[email protected]> Reviewed-by: Daniel Stone <[email protected]>. Signed-off-by: Axel Davy <[email protected]>
* egl/wayland: Simplify dri2_wl_create_surfaceAxel Davy2015-05-111-25/+13
| | | | | | | | | This function is always used with EGL_WINDOW_BIT. Pixmaps are forbidden for Wayland, and PBuffers are unimplemented. Reviewed-by: Daniel Stone <[email protected]>. Signed-off-by: Axel Davy <[email protected]>
* egl/x11: move dri2_x11_swrast_create_image_khr to egl_dri2_fallback.hAxel Davy2015-05-112-10/+10
| | | | | | | Reviewed-by: Dave Airlie <[email protected]> Reviewed-by: Daniel Stone <[email protected]> Signed-off-by: Axel Davy <[email protected]>
* egl/wayland: Implement DRI_PRIME supportAxel Davy2015-05-112-15/+92
| | | | | | | | | | | | | | | | | | When the server gpu and requested gpu are different: . They likely don't support the same tiling modes . They likely do not have fast access to the same locations Thus we do: . render to a tiled buffer we do not share with the server . Copy the content at every swap to a buffer with no tiling that we share with the server. This is similar to the glx dri3 DRI_PRIME implementation. Reviewed-by: Dave Airlie <[email protected]> Reviewed-by: Daniel Stone <[email protected]> Signed-off-by: Axel Davy <[email protected]>
* egl/wayland: Add support for render-nodesAxel Davy2015-05-112-14/+56
| | | | | | | | | | | | | | It is possible the server advertises a render-node. In that case no authentication is needed, and Gem names are forbidden. Reviewed-by: Dave Airlie <[email protected]> Reviewed-by: Daniel Stone <[email protected]> Signed-off-by: Axel Davy <[email protected]> v2: do not check for __DRI_IMAGE_DRIVER, but instead do not advertise __DRI_DRI2_LOADER when on a render-node.
* glx/dri3: Add additional check for gpu offloading caseAxel Davy2015-05-111-0/+5
| | | | | | | | | | | | | | | Checks blitImage is implemented. Initially having the __DRIimageExtension extension at version 9 at least meant blitImage was supported. However some implementation do advertise version >= 9 without implementing it. CC: 10.5 <[email protected]> Reviewed-by: Dave Airlie <[email protected]> Reviewed-by: Daniel Stone <[email protected]> Signed-off-by: Axel Davy <[email protected]>
* doc/egl: Remove depreciated EGL_SOFTWAREAxel Davy2015-05-111-8/+0
| | | | | | | | | | EGL_SOFTWARE is not supported anywhere in the code, whereas LIBGL_ALWAYS_SOFTWARE is. Reviewed-by: Dave Airlie <[email protected]> Reviewed-by: Daniel Stone <[email protected]> Signed-off-by: Axel Davy <[email protected]>
* egl/wayland: properly destroy wayland objectsAxel Davy2015-05-112-1/+6
| | | | | | | | | | | the wl_registry and the wl_queue allocated weren't destroyed. CC: 10.5 <[email protected]> Reviewed-by: Dave Airlie <[email protected]> Reviewed-by: Daniel Stone <[email protected]> Signed-off-by: Axel Davy <[email protected]>
* i965/fs: Disable opt_sampler_eot for textureGatherNeil Roberts2015-05-111-0/+10
| | | | | | | | | | | | | The opt_sampler_eot optimisation seems to break when the last instruction is SHADER_OPCODE_TG4. A bunch of Piglit tests end up doing this so it causes a lot of regressions. I can't find any documentation or known workarounds to indicate that this is expected behaviour, but considering that this is probably a pretty unlikely situation in a real use case we might as well disable it in order to avoid the regressions. In total this fixes 451 tests. Reviewed-by: Ben Widawsky <[email protected]> Reviewed-by: Chris Forbes <[email protected]>
* mesa: use _mesa_has_compute_shaders instead of extension checkTapani Pälli2015-05-111-1/+3
| | | | | | | | | | | This was really the original purpose, for enabling the path for ES3.1 tests without the extension being set. Set also fallthrough comment for Coverity (caught by Matt). v2: .. and test the right way, not wrong one (Ilia Mirkin) Signed-off-by: Tapani Pälli <[email protected]> Reviewed-by: Ilia Mirkin <[email protected]>
* main: glGetIntegeri_v fails for GL_VERTEX_BINDING_STRIDEMarta Lofstedt2015-05-111-0/+1
| | | | | | | | | | The return type for GL_VERTEX_BINDING_STRIDE is missing, this cause glGetIntegeri_v to fail. Signed-off-by: Marta Lofstedt <[email protected]> Reviewed-by: Tapani Pälli <[email protected]> Reviewed-by: Emil Velikov <[email protected]> Cc: "10.4 10.5" <[email protected]>
* r600: use pipe->hw prim convert from radeonsiDave Airlie2015-05-111-15/+16
| | | | | | | | This avoids future addition to PIPE_PRIM_ from causing regressions on r600g. Reviewed-by: Marek Olšák <[email protected]> Signed-off-by: Dave Airlie <[email protected]>
* freedreno/ir3/nir: fix build break after f752effaRob Clark2015-05-101-4/+3
| | | | | | | Our lower if/else pass was missed when converting NIR to use linked lists rather than hashsets to track use/def sets. Signed-off-by: Rob Clark <[email protected]>
* nv50/ir: only enable mul saturate on G200+Ilia Mirkin2015-05-091-1/+4
| | | | | | | | | | | | | Commit 44673512a84 enabled support for saturating fmul. However experimentally this does not seem to work on the older chips. Restrict the feature to G200 (NVA0) and later. Reported-by: Pierre Moreau <[email protected]> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=90350 Signed-off-by: Ilia Mirkin <[email protected]> Tested-by: Pierre Moreau <[email protected]> Reviewed-by: Tobias Klausmann <[email protected]> Cc: [email protected]
* nvc0: reset the instanced elements state when doing blit using 3d engineIlia Mirkin2015-05-091-0/+6
| | | | | | | | | | | | | | Since we update num_vtxelts here, we could otherwise end up with stale instancing information in the upper bits which wouldn't otherwise get reset. (Also we run the risk of the previous draw having set the first element as instanced.) This appears as one of the causes for the test pointed out in fdo#90363 to fail on nvc0. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=90363 Signed-off-by: Ilia Mirkin <[email protected]> Cc: [email protected]
* nvc0: keep track of PGRAPH state in nvc0_screenIlia Mirkin2015-05-094-24/+34
| | | | | | | | | | See identical commit for nv50. Destroying the current context and then creating a new one or switching to another existing context would cause the "current" state to not be properly initialized, so we save it off in the screen. Signed-off-by: Ilia Mirkin <[email protected]> Cc: [email protected]
* nv50: keep track of PGRAPH state in nv50_screenIlia Mirkin2015-05-094-30/+36
| | | | | | | | | | | | | | Normally this is kept in nv50_context, and on switching the active context, the state is copied from the previous context. However when the last context is destroyed, this is lost, and a new context might later be created. When the currently-active context is destroyed, save its state in the screen, and restore it when setting the current context. Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=90363 Reported-by: Matteo Bruni <[email protected]> Signed-off-by: Ilia Mirkin <[email protected]> Tested-by: Matteo Bruni <[email protected]> Cc: [email protected]
* nir: Fix aggressive typos in nir_from_ssa.c.Kenneth Graunke2015-05-081-5/+5
| | | | | | s/agressive/aggressive/g Trivial.
* nir/search: Save/restore the variables_seen bitmask when matchingJason Ekstrand2015-05-081-0/+12
| | | | | | | | | | | | | Shader-db results on Broadwell: total instructions in shared programs: 7152330 -> 7137006 (-0.21%) instructions in affected programs: 1330548 -> 1315224 (-1.15%) helped: 5797 HURT: 76 GAINED: 0 LOST: 8 Reviewed-by: Connor Abbott <[email protected]>
* nir/search: Assert that variable id's are in rangeJason Ekstrand2015-05-081-0/+1
| | | | Reviewed-by: Connor Abbott <[email protected]>
* nir/search: handle explicitly sized sources in match_valueJason Ekstrand2015-05-081-8/+8
| | | | | | | | | | Previously, this case was being handled in match_expression prior to calling match_value. However, there is really no good reason for this given that match_value has all of the information it needs. Also, they weren't being handled properly in the commutative case and putting it in match_value gives us that for free. Reviewed-by: Connor Abbott <[email protected]>
* nir/nir: Use a linked list instead of a hash set for use/def setsJason Ekstrand2015-05-0814-306/+281
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This commit switches us from the current setup of using hash sets for use/def sets to using linked lists. Doing so should save us quite a bit of memory because we aren't carrying around 3 hash sets per register and 2 per SSA value. It should also save us CPU time because adding/removing things from use/def sets is 4 pointer manipulations instead of a hash lookup. Running shader-db 50 times with USE_NIR=0, NIR, and NIR + use/def lists: GLSL IR Only: 586.4 +/- 1.653833 NIR with hash sets: 675.4 +/- 2.502108 NIR + use/def lists: 641.2 +/- 1.557043 I also ran a memory usage experiment with Ken's patch to delete GLSL IR and keep NIR. This patch cuts an aditional 42.9 MiB of ralloc'd memory over and above what we gained by deleting the GLSL IR on the same dota trace. On the code complexity side of things, some things are now much easier and others are a bit harder. One of the operations we perform constantly in optimization passes is to replace one source with another. Due to the fact that an instruction can use the same SSA value multiple times, we had to iterate through the sources of the instruction and determine if the use we were replacing was the only one before removing it from the set of uses. With this patch, uses are per-source not per-instruction so we can just remove it safely. On the other hand, trying to iterate over all of the instructions that use a given value is more difficult. Fortunately, the two places we do that are the ffma peephole where it doesn't matter and GCM where we already gracefully handle duplicates visits to an instruction. Another aspect here is that using linked lists in this way can be tricky to get right. With sets, things were quite forgiving and the worst that happened if you didn't properly remove a use was that it would get caught in the validator. With linked lists, it can lead to linked list corruption which can be harder to track. However, we do just as much validation of the linked lists as we did of the sets so the validator should still catch these problems. While working on this series, the vast majority of the bugs I had to fix were caught by assertions. I don't think the lists are going to be that much worse than the sets. Reviewed-by: Connor Abbott <[email protected]>
* util/list: Add a list validation functionJason Ekstrand2015-05-081-0/+9
| | | | | Acked-by: Connor Abbott <[email protected]> Reviewed-by: Rob Clark <[email protected]>
* util/list: Add list_empty and list_length functionsJason Ekstrand2015-05-081-0/+15
| | | | | | | v2: Don't use C99 when iterating over the list Acked-by: Connor Abbott <[email protected]> Reviewed-by: Rob Clark <[email protected]>
* util/list: Add C99-based iterator macrosJason Ekstrand2015-05-081-0/+34
| | | | | | | v2: Use LIST_ENTRY instead of container_of in iterators Acked-by: Connor Abbott <[email protected]> Reviewed-by: Rob Clark <[email protected]>
* util: Move gallium's linked list to utilJason Ekstrand2015-05-0844-55/+50
| | | | | | | | | The linked list in gallium is pretty much the kernel list and we would like to have a C-based linked list for all of mesa. Let's not duplicate and just steal the gallium one. Acked-by: Connor Abbott <[email protected]> Reviewed-by: Rob Clark <[email protected]>
* gallium/double_list: s/INLINE/inline and remove the p_compiler includeJason Ekstrand2015-05-081-7/+6
| | | | | Acked-by: Connor Abbott <[email protected]> Reviewed-by: Rob Clark <[email protected]>
* nir: Use nir_instr_rewrite_src in copy propagationJason Ekstrand2015-05-081-61/+10
| | | | | | | We were rolling our own rewrite_src variant in copy-propagation. Let's stop doing that and use the ones in core NIR. Reviewed-by: Connor Abbott <[email protected]>
* nir: Add a function for rewriting the condition of an if statementJason Ekstrand2015-05-082-0/+23
| | | | Reviewed-by: Connor Abbott <[email protected]>
* nir: Add and use initializer #defines for nir_src and nir_destJason Ekstrand2015-05-086-18/+15
| | | | Reviewed-by: Connor Abbott <[email protected]>
* nir: Modernize the out-of-SSA passJason Ekstrand2015-05-081-103/+36
| | | | | | | | | The out-of-SSA pass was one of the first passes written when getting SSA up-and-going (for obvious reasons). As such, it came before a lot of the nifty SSA-based helpers were introduced. This commit modernizes it so that we're no longer doing nearly as much manual banging on use/def sets. Reviewed-by: Connor Abbott <[email protected]>
* nir/validate: Validate SSA def parent instructionsJason Ekstrand2015-05-081-0/+2
| | | | Reviewed-by: Connor Abbott <[email protected]>
* nv50/ir: only propagate saturate up if some actual folding took placeIlia Mirkin2015-05-081-1/+2
| | | | | | | | | | The former logic would copy the saturate up to any mul with an immediate if there was a subsequent mul with a saturate. However we only want to do that if we collapsed 2 muls by multiplying their immediates (or were able to put the immediate in as a post-multiplier). Signed-off-by: Ilia Mirkin <[email protected]> Cc: [email protected]
* nir: Delete all traces of nir_op_flogIan Romanick2015-05-083-10/+0
| | | | | | | | Nothing produces it, and nothing can consume it. Signed-off-by: Ian Romanick <[email protected]> Reviewed-by: Matt Turner <[email protected]> Acked-by: Jason Ekstrand <[email protected]>
* nir: Don't produce nir_op_flog from GLSL IRIan Romanick2015-05-081-1/+1
| | | | | | | | | All paths that produce GLSL IR for NIR lower ir_unop_log. All paths that consume NIR will explode if they geta nir_op_flog. Signed-off-by: Ian Romanick <[email protected]> Reviewed-by: Matt Turner <[email protected]> Acked-by: Jason Ekstrand <[email protected]>
* nir: Delete all traces of nir_op_fexpIan Romanick2015-05-083-9/+0
| | | | | | | | Nothing produces it, and nothing can consume it. Signed-off-by: Ian Romanick <[email protected]> Reviewed-by: Matt Turner <[email protected]> Acked-by: Jason Ekstrand <[email protected]>
* nir: Don't produce nir_op_fexp from GLSL IRIan Romanick2015-05-081-1/+1
| | | | | | | | | All paths that produce GLSL IR for NIR lower ir_unop_exp. All paths that consume NIR will explode if they geta nir_op_fexp. Signed-off-by: Ian Romanick <[email protected]> Reviewed-by: Matt Turner <[email protected]> Acked-by: Jason Ekstrand <[email protected]>