Age | Commit message (Collapse) | Author | |
---|---|---|---|
2013-10-08 | driver: comments updates and small cleanups | Wladimir J. van der Laan | |
2013-10-05 | documentation update | Wladimir J. van der Laan | |
2013-08-29 | driver: comment update, small refactor of sync_context | Wladimir J. van der Laan | |
2013-08-23 | remove lots of trailing spaces | Wladimir J. van der Laan | |
whitespace only changes | |||
2013-08-23 | comments/documentation update | Wladimir J. van der Laan | |
2013-08-03 | tools: add output to file option to gdb gpu-trace | Wladimir J. van der Laan | |
2013-08-02 | trivial comment + demo updates | Wladimir J. van der Laan | |
2013-08-01 | tools: improve gpu-trace | Wladimir J. van der Laan | |
Use our old friend memcpy() to make temporary copy of command buffer memory in client process. No longer needs cooporation from target, which means the hacky _viv_read_u32 can be removed again. An added advantage is that this is *much* faster. | |||
2013-07-31 | tools: add gpu-trace gdb command | Wladimir J. van der Laan | |
This command traces and dumps all submitted command buffers, along with the physcial address of each command. This should come in handy for searching back the physical address that the rendering is stuck on according to the kernel. | |||
2013-07-29 | fb_rawshader: fix compile warnings | Wladimir J. van der Laan | |
2013-07-28 | documentation and comment updates | Wladimir J. van der Laan | |
2013-07-28 | readme update | Michał Ściubidło | |
2013-07-25 | rnndb: add DMA debug register description | Wladimir J. van der Laan | |
Info from imx6_v4_0_0 driver | |||
2013-07-25 | driver: fix the build on !imx6_v4 | Wladimir J. van der Laan | |
gc_abi.h used `uint32_t` which was not yet defined there, resulting in a compile error. Fix by simply using `unsigned int` as that's what the kernel does too. | |||
2013-07-24 | README update | Wladimir J. van der Laan | |
2013-07-24 | add mth in to authors in README | Wladimir J. van der Laan | |
2013-07-18 | Add @mike29 to authors list | Wladimir J. van der Laan | |
2013-07-10 | documentation updates | Wladimir J. van der Laan | |
2013-07-05 | documentation update | Wladimir J. van der Laan | |
2013-07-04 | driver: fix stencil alpha in mesa | Wladimir J. van der Laan | |
Set stencil operations to KEEP if write mask is 0. When the write mask is 0 and the operations are not set to KEEP, the GPU writes the depth buffer for the entire primitive and not just the area where the stencil test condition holds. | |||
2013-04-28 | Some build fixes | Wladimir J. van der Laan | |
- don't build gc600 specific tests with a plain "make", as they error out during compile for other platforms - stdlib.h is needed for size_t in c++ | |||
2013-04-18 | small doc update | Wladimir J. van der Laan | |
2013-03-30 | etna: chip specs/features handling | Wladimir J. van der Laan | |
- further plumbing to allow switching stuff in the etna driver based on chip specifications - add kernel-independent specs structure to viv - figure out some bits related to shader state and specs | |||
2013-03-16 | beginnings of shader code generation | Wladimir J. van der Laan | |
still needs quite some work, but the first steps toward a usable shader generator are there (can compile the vertex shader of cube_companion). Fragment shader is still a todo. | |||
2013-03-10 | add sjhill to authors in README.md | Wladimir J. van der Laan | |
2013-03-07 | etna: prepare for tgsi->vivante shader conversion | Wladimir J. van der Laan | |
- integrate part of gallium, outside of mesa for now, this will help porting | |||
2013-03-04 | rename THREAD_WALKER to CL | Wladimir J. van der Laan | |
also small documentation updates | |||
2013-03-02 | command stream dumper: i.mx6 kernel driver compat | Wladimir J. van der Laan | |
2013-02-26 | harmonize PIPE_* format usage | Wladimir J. van der Laan | |
PIPE_FORMAT convention is LSB to MSB, VIVS convention is MSB to LSB, so pixel formats appear reversed when converted | |||
2013-02-24 | minor reorganizations and corrections | Wladimir J. van der Laan | |
2013-02-23 | etna_pipe: add displacement mapping demo | Wladimir J. van der Laan | |
2013-02-17 | etna: fb stencil test | Wladimir J. van der Laan | |
2013-02-10 | etna: demonstrate cubemap | Wladimir J. van der Laan | |
2013-02-10 | etna: alpha blending | Wladimir J. van der Laan | |
In a bizarre twist of fate, alpha blending now works. Interpretation of bit 16 of PE_COLOR_FORMAT was the wrong way around. | |||
2013-02-08 | gallium pipe: putting it all together | Wladimir J. van der Laan | |
everything except fb/buffer swapping and shader state is now driven using gallium pipe commands TODO: move all the stuff to separate files | |||
2013-01-29 | smarter handling of GCABI | Wladimir J. van der Laan | |
gc_abi.h defines the following flags describing the kernel interface to etna: - GCABI_CONTEXT_HAS_PHYSICAL: struct _gcoCONTEXT has `physical` and `bytes` fields - GCABI_HAS_MINOR_FEATURES_2: struct _gcsHAL_QUERY_CHIP_IDENTITY has `chipMinorFeatures2` field - GCABI_HAS_MINOR_FEATURES_3: struct _gcsHAL_QUERY_CHIP_IDENTITY has `chipMinorFeatures3` field - GCABI_USER_SIGNAL_HAS_TYPE: struct _gcsHAL_USER_SIGNAL has `signalType` field - GCABI_HAS_CONTEXT: struct _gcsHAL_COMMIT has `contextBuffer` field - GCABI_HAS_STATE_DELTAS: struct _gcsHAL_COMMIT has `delta` field | |||
2013-01-28 | etna: name a few states more appropriately | Wladimir J. van der Laan | |
2013-01-27 | cubox: fb rendering beginnings | Wladimir J. van der Laan | |
try to support old as well as new Vivante kernel drivers | |||
2013-01-26 | collect feature bits to compare the various GPUs | Wladimir J. van der Laan | |
2013-01-26 | dove/cubox compatiblity preparation | Wladimir J. van der Laan | |
2013-01-24 | add ETC1 support to fb/mipcube demo, figure out supertiling | Wladimir J. van der Laan | |
Also figure out some other resolve state. Add a tool (genmipmaps_etc1.py) to generate a dds file with ETC1 texture compression and auto-generate mipmaps (GIMP dds plugin cannot do ETC1). | |||
2013-01-22 | etna: add mip_cube demo | Wladimir J. van der Laan | |
try out mipmapping... | |||
2013-01-21 | fb/cube_companion: implement MSAA | Wladimir J. van der Laan | |
XXX need to figure out what this extra PS input is | |||
2013-01-21 | etna: memory management, map user memory | Wladimir J. van der Laan | |
2013-01-20 | figure out extended texture state | Wladimir J. van der Laan | |
2013-01-20 | documentation update, clarify texture states | Wladimir J. van der Laan | |
2013-01-17 | etna driver: general low level command buffer building | Wladimir J. van der Laan | |
Also implements synchronization commands such as flush, finish, semaphores and pipe switching. - Create N command buffers, with a signal for each buffer - Before starting to write to a buffer, wait for buffer's sync signal - After a buffer is full, queue the buffer's sync signal and switch to next buffer This is a first step towards an animated, rotated framebuffer demo. | |||
2013-01-17 | Add Ingenic JZ4770 to list of devices with Vivante GPU | Wladimir J. van der Laan | |
2013-01-17 | add dove gc600 driver | Wladimir J. van der Laan | |
2013-01-12 | mipmapping state | Wladimir J. van der Laan | |