Kenneth Graunke [Thu, 19 Apr 2018 20:08:37 +0000 (13:08 -0700)]
iris: vma fixes - don't free binder address
Kenneth Graunke [Thu, 19 Apr 2018 20:03:21 +0000 (13:03 -0700)]
iris: bo reuse
Kenneth Graunke [Thu, 19 Apr 2018 19:52:51 +0000 (12:52 -0700)]
iris: crazy pipe control code
imported from ~kwg/mesa pcx-2, gen < 8 code dropped
Kenneth Graunke [Thu, 19 Apr 2018 19:07:44 +0000 (12:07 -0700)]
iris: fixes
Kenneth Graunke [Thu, 19 Apr 2018 19:01:09 +0000 (12:01 -0700)]
iris: fixes from i965
Kenneth Graunke [Tue, 10 Apr 2018 22:59:57 +0000 (15:59 -0700)]
iris: port bug fix from i965
Kenneth Graunke [Tue, 10 Apr 2018 07:39:27 +0000 (00:39 -0700)]
iris: fix index
Kenneth Graunke [Tue, 10 Apr 2018 07:20:32 +0000 (00:20 -0700)]
iris: increase allocator alignment
Kenneth Graunke [Sat, 7 Apr 2018 14:07:19 +0000 (07:07 -0700)]
iris: better BT asserts
Probably nothing is working because texture upload isn't implemented
Kenneth Graunke [Sat, 7 Apr 2018 13:59:22 +0000 (06:59 -0700)]
iris: decoder fixes
Kenneth Graunke [Sat, 7 Apr 2018 13:35:51 +0000 (06:35 -0700)]
iris: set sampler views
Kenneth Graunke [Sat, 7 Apr 2018 13:16:51 +0000 (06:16 -0700)]
iris: isv freeing fixes
Kenneth Graunke [Sat, 7 Apr 2018 08:18:12 +0000 (01:18 -0700)]
iris: TES stash
TODO: key setup
Kenneth Graunke [Sat, 7 Apr 2018 08:01:24 +0000 (01:01 -0700)]
iris: SBA once at context creation, not per batch
hooray!
Kenneth Graunke [Sat, 7 Apr 2018 07:58:07 +0000 (00:58 -0700)]
iris: fix a scissor bug
Kenneth Graunke [Sat, 7 Apr 2018 07:49:12 +0000 (00:49 -0700)]
iris: assemble SAMPLER_STATE table at bind time
It's useless to allocate SAMPLER_STATEs in GPU memory on creation like
we do for SURFACE_STATES, because they need to be organized into a
contiguous block of memory. But we can do that at bind time, rather
than draw time.
Kenneth Graunke [Sat, 7 Apr 2018 07:11:14 +0000 (00:11 -0700)]
iris: same treatment for sampler views
Kenneth Graunke [Sat, 7 Apr 2018 06:57:45 +0000 (23:57 -0700)]
iris: allocate SURFACE_STATEs up front and stop streaming them
Kenneth Graunke [Sat, 7 Apr 2018 06:16:11 +0000 (23:16 -0700)]
iris: delete more trash
Kenneth Graunke [Sat, 7 Apr 2018 06:08:52 +0000 (23:08 -0700)]
iris: canonicalize addresses.
Back to working! Woo!
Kenneth Graunke [Sat, 7 Apr 2018 05:39:01 +0000 (22:39 -0700)]
iris: validation dumping improvements
backported from i965. don't bother with (pinned) because everything is.
Kenneth Graunke [Sat, 7 Apr 2018 03:33:17 +0000 (20:33 -0700)]
iris: update vb BO handling now that we have softpin
Kenneth Graunke [Sat, 7 Apr 2018 00:01:45 +0000 (17:01 -0700)]
iris: decoder fixes
Kenneth Graunke [Fri, 6 Apr 2018 23:31:01 +0000 (16:31 -0700)]
iris: binder fixes
Kenneth Graunke [Fri, 6 Apr 2018 23:21:21 +0000 (16:21 -0700)]
iris: hook up batch decoder
Kenneth Graunke [Fri, 6 Apr 2018 18:44:59 +0000 (11:44 -0700)]
iris: binders
Kenneth Graunke [Fri, 6 Apr 2018 18:41:14 +0000 (11:41 -0700)]
iris: include p_defines.h in iris_bufmgr.h
for PIPE_TRANSFER_WRITE and friends
Kenneth Graunke [Fri, 6 Apr 2018 07:19:57 +0000 (00:19 -0700)]
iris: set EXEC_OBJECT_WRITE
Kenneth Graunke [Fri, 6 Apr 2018 07:05:24 +0000 (00:05 -0700)]
iris: rewrite to use memzones and not relocs
Kenneth Graunke [Fri, 6 Apr 2018 04:48:33 +0000 (21:48 -0700)]
iris: more uploaders
Kenneth Graunke [Fri, 6 Apr 2018 00:04:54 +0000 (17:04 -0700)]
iris: Also set SUPPORTS_48B? Not sure if necessary.
Kenneth Graunke [Fri, 6 Apr 2018 00:03:07 +0000 (17:03 -0700)]
iris: dump gtt offset in dump_validation_list
Kenneth Graunke [Thu, 5 Apr 2018 21:52:53 +0000 (14:52 -0700)]
iris: fix icache memzone
Kenneth Graunke [Tue, 3 Apr 2018 08:40:23 +0000 (01:40 -0700)]
iris: Soft-pin the universe
Breaks everything, woo!
Kenneth Graunke [Mon, 26 Mar 2018 21:11:55 +0000 (14:11 -0700)]
iris: some thinking about binding tables
Kenneth Graunke [Mon, 26 Mar 2018 23:14:29 +0000 (16:14 -0700)]
iris: bufmgr updates.
Drop BO_ALLOC_BUSY (best not to hand people a loaded gun...)
Drop vestiges of alignment
Kenneth Graunke [Sat, 10 Feb 2018 07:38:08 +0000 (23:38 -0800)]
iris: stop adding 9 to our varyings
Kenneth Graunke [Sat, 10 Feb 2018 06:46:28 +0000 (22:46 -0800)]
iris: set strides on transfers
Kenneth Graunke [Sat, 10 Feb 2018 06:18:27 +0000 (22:18 -0800)]
iris: enable a few more formats
Kenneth Graunke [Sat, 10 Feb 2018 05:58:08 +0000 (21:58 -0800)]
iris: decode batches if they fail to submit
Kenneth Graunke [Sat, 10 Feb 2018 05:58:00 +0000 (21:58 -0800)]
iris: NOOP pad batches correctly
Kenneth Graunke [Sat, 10 Feb 2018 00:45:04 +0000 (16:45 -0800)]
iris: warn if execbuf fails
Kenneth Graunke [Fri, 9 Feb 2018 22:21:54 +0000 (14:21 -0800)]
iris: uniform bits...badly
Kenneth Graunke [Wed, 31 Jan 2018 01:44:26 +0000 (17:44 -0800)]
iris: sample mask...not 0.
We now have a first triangle!
Kenneth Graunke [Wed, 31 Jan 2018 01:40:44 +0000 (17:40 -0800)]
iris: write DISABLES are not write ENABLES...whoops
Kenneth Graunke [Wed, 31 Jan 2018 01:36:24 +0000 (17:36 -0800)]
iris: fix extents
Kenneth Graunke [Wed, 31 Jan 2018 01:18:17 +0000 (17:18 -0800)]
iris: catastrophic state pointer mistake
Kenneth Graunke [Mon, 17 Dec 2018 23:17:54 +0000 (15:17 -0800)]
iris: more SF CL VPs
Kenneth Graunke [Tue, 30 Jan 2018 20:30:34 +0000 (12:30 -0800)]
iris: fix dmabuf retval comparisons
0 means success
Kenneth Graunke [Tue, 30 Jan 2018 20:30:30 +0000 (12:30 -0800)]
iris: more sketchy SBE
Kenneth Graunke [Tue, 30 Jan 2018 10:44:25 +0000 (02:44 -0800)]
iris: compctrl
oh, also run things
Kenneth Graunke [Tue, 30 Jan 2018 10:16:34 +0000 (02:16 -0800)]
iris: actually pin the instruction cache buffers
Kenneth Graunke [Tue, 30 Jan 2018 09:50:44 +0000 (01:50 -0800)]
iris: smaller blend state
Kenneth Graunke [Tue, 30 Jan 2018 09:44:07 +0000 (01:44 -0800)]
iris: don't do samplers for disabled stages
Kenneth Graunke [Tue, 30 Jan 2018 09:40:14 +0000 (01:40 -0800)]
iris: render targets!
Kenneth Graunke [Tue, 30 Jan 2018 09:09:53 +0000 (01:09 -0800)]
iris: fix silly unused batch with addr macro
Kenneth Graunke [Mon, 29 Jan 2018 23:06:08 +0000 (15:06 -0800)]
iris: warning fixes
Kenneth Graunke [Mon, 29 Jan 2018 23:06:04 +0000 (15:06 -0800)]
iris: basic SBE code
Kenneth Graunke [Fri, 26 Jan 2018 05:58:31 +0000 (21:58 -0800)]
iris: alpha testing in PSB
Kenneth Graunke [Fri, 26 Jan 2018 05:53:41 +0000 (21:53 -0800)]
iris: blend state
Kenneth Graunke [Fri, 26 Jan 2018 05:39:44 +0000 (21:39 -0800)]
iris: dummy constants
Kenneth Graunke [Fri, 26 Jan 2018 05:23:45 +0000 (21:23 -0800)]
iris: URB configs.
Kenneth Graunke [Fri, 26 Jan 2018 04:12:37 +0000 (20:12 -0800)]
iris: actually set KSP offsets
Kenneth Graunke [Fri, 26 Jan 2018 04:04:37 +0000 (20:04 -0800)]
iris: actually softpin at an address
Kenneth Graunke [Fri, 26 Jan 2018 03:43:45 +0000 (19:43 -0800)]
iris: actually destroy the cache
Kenneth Graunke [Fri, 26 Jan 2018 03:39:10 +0000 (19:39 -0800)]
iris: rewrite program cache to use u_upload_mgr
Kenneth Graunke [Thu, 25 Jan 2018 18:57:13 +0000 (10:57 -0800)]
iris: no NEW_SBA
Kenneth Graunke [Thu, 25 Jan 2018 10:13:28 +0000 (02:13 -0800)]
iris: shuffle comments
Kenneth Graunke [Thu, 25 Jan 2018 10:09:59 +0000 (02:09 -0800)]
iris: bits of WM key
Kenneth Graunke [Thu, 25 Jan 2018 10:03:18 +0000 (02:03 -0800)]
iris: move key pop to state module
shader key population needs to read state
Kenneth Graunke [Thu, 25 Jan 2018 09:53:52 +0000 (01:53 -0800)]
iris: fix SBA
Kenneth Graunke [Thu, 25 Jan 2018 09:36:49 +0000 (01:36 -0800)]
iris: use vtbl to avoid multiple symbols, fix state base address
Kenneth Graunke [Thu, 25 Jan 2018 09:19:58 +0000 (01:19 -0800)]
iris: softpin some things
Kenneth Graunke [Wed, 24 Jan 2018 04:01:28 +0000 (20:01 -0800)]
iris: drop const from prog data parameters
we ralloc steal things, which makes it a little bogus
Kenneth Graunke [Tue, 23 Jan 2018 09:23:54 +0000 (01:23 -0800)]
iris: more comes from bits filled in
tomorrow, fix the build system to avoid symbol clashes somehow...
we're getting gen9 functions because they happen to be listed before 10
in the link list.
Kenneth Graunke [Tue, 23 Jan 2018 09:10:05 +0000 (01:10 -0800)]
iris: index buffer BO
Kenneth Graunke [Tue, 23 Jan 2018 09:07:09 +0000 (01:07 -0800)]
iris: WM.
I could have added a dirty bit for this, but it doesn't seem worth it
Kenneth Graunke [Tue, 23 Jan 2018 08:44:44 +0000 (00:44 -0800)]
iris: initial gpu state
Kenneth Graunke [Tue, 23 Jan 2018 07:39:38 +0000 (23:39 -0800)]
iris: reorganize commands to match brw
Kenneth Graunke [Tue, 23 Jan 2018 07:25:18 +0000 (23:25 -0800)]
iris: don't forget about TE
Kenneth Graunke [Tue, 23 Jan 2018 07:03:54 +0000 (23:03 -0800)]
iris: convert IRIS_DIRTY_* to #defines
enums are SIGNED. so IRIS_DIRTY_VS << 4 gets sign extended, making it
not equal to IRIS_DIRTY_FS. Surprising!
Kenneth Graunke [Tue, 23 Jan 2018 06:40:51 +0000 (22:40 -0800)]
iris: emit shader packets
Kenneth Graunke [Tue, 23 Jan 2018 06:31:27 +0000 (22:31 -0800)]
iris: actually save derived state
Kenneth Graunke [Mon, 22 Jan 2018 19:52:58 +0000 (11:52 -0800)]
iris: promote iris_program_cache_item to iris_compiled_shader
Kenneth Graunke [Mon, 22 Jan 2018 07:55:04 +0000 (23:55 -0800)]
iris: some shader bits
Kenneth Graunke [Mon, 22 Jan 2018 05:23:48 +0000 (21:23 -0800)]
iris: scissor slots
Kenneth Graunke [Mon, 22 Jan 2018 05:20:44 +0000 (21:20 -0800)]
iris: bind_state -> compute state
Kenneth Graunke [Mon, 22 Jan 2018 05:19:26 +0000 (21:19 -0800)]
iris: 3DPRIMITIVE fields
Kenneth Graunke [Mon, 22 Jan 2018 05:09:30 +0000 (21:09 -0800)]
iris: fix VF instancing length so we don't get garbage in batch
Kenneth Graunke [Mon, 22 Jan 2018 05:02:05 +0000 (21:02 -0800)]
iris: vertex packet fixes
Kenneth Graunke [Mon, 22 Jan 2018 04:10:57 +0000 (20:10 -0800)]
iris: fix VBs
Kenneth Graunke [Mon, 22 Jan 2018 04:10:43 +0000 (20:10 -0800)]
iris: fix assert
Kenneth Graunke [Mon, 22 Jan 2018 02:12:18 +0000 (18:12 -0800)]
iris: fix indentation
Kenneth Graunke [Mon, 22 Jan 2018 02:08:15 +0000 (18:08 -0800)]
iris: hack to stop crashing on samplers for now
Kenneth Graunke [Mon, 22 Jan 2018 02:04:05 +0000 (18:04 -0800)]
iris: initialize dirty bits to ~0ull
Kenneth Graunke [Mon, 22 Jan 2018 02:03:58 +0000 (18:03 -0800)]
iris: actually advance forward when emitting commands
Kenneth Graunke [Mon, 22 Jan 2018 01:44:08 +0000 (17:44 -0800)]
iris: actually flush the commands
Kenneth Graunke [Mon, 22 Jan 2018 01:43:49 +0000 (17:43 -0800)]
iris: actually APPEND commands, not stomp over the top and never incr
Kenneth Graunke [Mon, 22 Jan 2018 01:34:41 +0000 (17:34 -0800)]
iris: VB fixes
Kenneth Graunke [Mon, 22 Jan 2018 00:49:02 +0000 (16:49 -0800)]
iris: DEBUG=bat
Deleted in the interest of making the branch compile at each step