summaryrefslogtreecommitdiff
AgeCommit message (Collapse)Author
2013-11-10rs: rework etna_submit_rs_state for gc2000Christian Gmeiner
2013-11-06Merge pull request #19 from austriancoder/masterWladimir J. van der Laan
gc2000: fix RS_PIPE_OFFSET and RS_WINDOW_SIZE
2013-11-06rs: do computation of RS_WINDOW_SIZE in etna_compile_rs_stateChristian Gmeiner
Signed-off-by: Christian Gmeiner <christian.gmeiner@gmail.com>
2013-11-06make etna_ctx available in etna_compile_rs_state(..)Christian Gmeiner
Signed-off-by: Christian Gmeiner <christian.gmeiner@gmail.com>
2013-11-05rs: update RS_WINDOW_SIZE for two pixel pipesChristian Gmeiner
on a vivante gpu with only one pixel pipe the following gets used: 0x08010588, /* LOAD_STATE (1) Base: 0x01620 Size: 1 Fixp: 0 */ 0x010001c0, /* [01620] RS.WINDOW_SIZE := HEIGHT=256,WIDTH=448 */ with two pixel pipes the following gets used: 0x008001c0, /* [01620] RS.WINDOW_SIZE := HEIGHT=128,WIDTH=448 */ 0x080205c0, /* LOAD_STATE (1) Base: 0x01700 Size: 2 Fixp: 0 */ 0x00000000, /* [01700] RS.PIPE[0].OFFSET := X=0,Y=0 */ 0x00800000, /* [01704] RS.PIPE[1].OFFSET := X=0,Y=128 */ 0x00000000, /* PAD */
2013-11-05rs: set pipe_offsetChristian Gmeiner
2013-10-28etnaviv: fix build problem with GCABI=v2Wladimir J. van der Laan
GCABI_CONTEXT_HAS_PHYSICAL was still using the old ctx_info structure which doesn't exist anymore.
2013-10-28src: raise error if GCABI is not definedWladimir J. van der Laan
Defaulting to v2 (or any version, at that) is a bad idea.
2013-10-19Merge pull request #17 from austriancoder/masterWladimir J. van der Laan
small fixes for replay
2013-10-18replay: make use of using2D and using3DChristian Gmeiner
The imx6 kernel 1.1.0 makes use of these vars, so it is a good idea to set them. root@linaro-ubuntu-desktop:~/linux-imx6# grep using3D drivers/mxc/gpu-viv/ -R drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_command.c: if (Context->dirty || commandBufferObject->using3D) drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_command.c: if (Context->dirty || commandBufferObject->using3D) drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_command.c: if (commandBufferObject->using3D && Context->dirty3D) drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_command.c: if (commandBufferObject->using3D && Context->dirty3D) drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_kernel_buffer.h: gctBOOL using3D; root@linaro-ubuntu-desktop:~/linux-imx6# grep using2D drivers/mxc/gpu-viv/ -R drivers/mxc/gpu-viv/hal/kernel/gc_hal_kernel_command.c: if (commandBufferObject->using2D && Context->dirty2D) drivers/mxc/gpu-viv/hal/kernel/inc/gc_hal_kernel_buffer.h: gctBOOL using2D; root@linaro-ubuntu-desktop:~/linux-imx6# Signed-off-by: Christian Gmeiner <christian.gmeiner@gmail.com>
2013-10-18replay: set hardwareType if neededChristian Gmeiner
Signed-off-by: Christian Gmeiner <christian.gmeiner@gmail.com>
2013-10-17etnaviv: support importing external vidmem nodesWladimir J. van der Laan
Add support for importing external jgckVIDMEM nodes. This could be used to experiment with transferring textures between processes, at least until the new kernel driver.
2013-10-15driver: fix cubemapsWladimir J. van der Laan
Previous tiling changes broke cubemap support. This commit fixes the offsets and makes the cubemap test work again.
2013-10-15driver: clean up resource creation functionWladimir J. van der Laan
2013-10-14fbdemos: use new way of passing fbWladimir J. van der Laan
Pass buffer object here instead of raw physical address.
2013-10-14driver: use etna_bo_from_fbdev to get at framebufferWladimir J. van der Laan
etna_screen can now render to any buffer object
2013-10-14etnaviv: fix compile with v2Wladimir J. van der Laan
2013-10-14Merge pull request #16 from austriancoder/masterWladimir J. van der Laan
fix make gc2000 in src/replay
2013-10-14current gc2000 replay stuff is gc880 specificChristian Gmeiner
Signed-off-by: Christian Gmeiner <christian.gmeiner@gmail.com>
2013-10-14fix compile of cube_etna_gc2000.cChristian Gmeiner
Signed-off-by: Christian Gmeiner <christian.gmeiner@gmail.com>
2013-10-14fix compile of empty_screen_gc2000.cChristian Gmeiner
root@linaro-ubuntu-desktop:~/etna_viv/src/replay# make gc2000 gcc -D_POSIX_C_SOURCE=200809 -D_GNU_SOURCE -DLINUX -pthread -O2 -g3 -ggdb3 -Wall -Wmissing-prototypes -DGCABI_imx6 -D_POSIX_C_SOURCE=200809 -D_GNU_SOURCE -I../lib -I../include_imx6 -fPIC -I../minigallium/include -I../minigallium/auxiliary -I.. -fdiagnostics-show-option -Werror=missing-prototypes -Werror=missing-declarations -Werror=implicit-function-declaration -Werror=return-type -I../resources -I../driver -std=gnu99 -c -o cube_etna2_gc2000.o cube_etna2_gc2000.c gcc -D_POSIX_C_SOURCE=200809 -D_GNU_SOURCE -DLINUX -pthread -O2 -g3 -ggdb3 -Wall -Wmissing-prototypes -DGCABI_imx6 -D_POSIX_C_SOURCE=200809 -D_GNU_SOURCE -I../lib -I../include_imx6 -fPIC -I../minigallium/include -I../minigallium/auxiliary -I.. -fdiagnostics-show-option -Werror=missing-prototypes -Werror=missing-declarations -Werror=implicit-function-declaration -Werror=return-type -I../resources -I../driver -std=gnu99 -o cube_etna2_gc2000 cube_etna2_gc2000.o ../lib/write_bmp.o ../lib/esTransform.o ../lib/etna_bswap.o -O2 -g3 -ggdb3 -Wall -Wmissing-prototypes -DGCABI_imx6 -D_POSIX_C_SOURCE=200809 -D_GNU_SOURCE -I../lib -I../include_imx6 -fPIC -I../minigallium/include -I../minigallium/auxiliary -I.. -fdiagnostics-show-option -Werror=missing-prototypes -Werror=missing-declarations -Werror=implicit-function-declaration -Werror=return-type -I../resources -I../driver -lm -L../etnaviv -letnaviv gcc -D_POSIX_C_SOURCE=200809 -D_GNU_SOURCE -DLINUX -pthread -O2 -g3 -ggdb3 -Wall -Wmissing-prototypes -DGCABI_imx6 -D_POSIX_C_SOURCE=200809 -D_GNU_SOURCE -I../lib -I../include_imx6 -fPIC -I../minigallium/include -I../minigallium/auxiliary -I.. -fdiagnostics-show-option -Werror=missing-prototypes -Werror=missing-declarations -Werror=implicit-function-declaration -Werror=return-type -I../resources -I../driver -std=gnu99 -c -o empty_screen_gc2000.o empty_screen_gc2000.c In file included from ../etnaviv/etna.h:33:0, from empty_screen_gc2000.c:16: ../etnaviv/viv.h:43:6: error: nested redefinition of ‘enum viv_features_word’ ../etnaviv/viv.h:43:6: error: redeclaration of ‘enum viv_features_word’ viv_raw.h:35:6: note: originally defined here ../etnaviv/viv.h:45:5: error: redeclaration of enumerator ‘viv_chipFeatures’ viv_raw.h:37:5: note: previous definition of ‘viv_chipFeatures’ was here ../etnaviv/viv.h:46:5: error: redeclaration of enumerator ‘viv_chipMinorFeatures0’ viv_raw.h:38:5: note: previous definition of ‘viv_chipMinorFeatures0’ was here ../etnaviv/viv.h:47:5: error: redeclaration of enumerator ‘viv_chipMinorFeatures1’ viv_raw.h:39:5: note: previous definition of ‘viv_chipMinorFeatures1’ was here ../etnaviv/viv.h:48:5: error: redeclaration of enumerator ‘viv_chipMinorFeatures2’ viv_raw.h:40:5: note: previous definition of ‘viv_chipMinorFeatures2’ was here ../etnaviv/viv.h:49:5: error: redeclaration of enumerator ‘viv_chipMinorFeatures3’ viv_raw.h:41:5: note: previous definition of ‘viv_chipMinorFeatures3’ was here ../etnaviv/viv.h:50:5: error: redeclaration of enumerator ‘VIV_FEATURES_WORD_COUNT’ viv_raw.h:42:5: note: previous definition of ‘VIV_FEATURES_WORD_COUNT’ was here ../etnaviv/viv.h:55:6: error: nested redefinition of ‘enum viv_hw_type’ ../etnaviv/viv.h:55:6: error: redeclaration of ‘enum viv_hw_type’ viv_raw.h:46:6: note: originally defined here ../etnaviv/viv.h:57:5: error: redeclaration of enumerator ‘VIV_HW_3D’ viv_raw.h:48:5: note: previous definition of ‘VIV_HW_3D’ was here ../etnaviv/viv.h:58:5: error: redeclaration of enumerator ‘VIV_HW_2D’ viv_raw.h:49:5: note: previous definition of ‘VIV_HW_2D’ was here ../etnaviv/viv.h:59:5: error: redeclaration of enumerator ‘VIV_HW_VG’ viv_raw.h:50:5: note: previous definition of ‘VIV_HW_VG’ was here ../etnaviv/viv.h:61:5: error: redeclaration of enumerator ‘VIV_HW_2D3D’ viv_raw.h:52:5: note: previous definition of ‘VIV_HW_2D3D’ was here ../etnaviv/viv.h:163:8: error: redefinition of ‘struct viv_specs’ viv_raw.h:61:8: note: originally defined here ../etnaviv/viv.h:186:8: error: redefinition of ‘struct viv_conn’ viv_raw.h:78:8: note: originally defined here ../etnaviv/viv.h:215:5: error: conflicting types for ‘viv_open’ viv_raw.h:91:5: note: previous declaration of ‘viv_open’ was here ../etnaviv/viv.h:220:5: error: conflicting types for ‘viv_invoke’ viv_raw.h:96:5: note: previous declaration of ‘viv_invoke’ was here ../etnaviv/viv.h:225:5: error: conflicting types for ‘viv_close’ viv_raw.h:100:5: note: previous declaration of ‘viv_close’ was here ../etnaviv/viv.h:228:5: error: conflicting types for ‘viv_alloc_contiguous’ viv_raw.h:103:5: note: previous declaration of ‘viv_alloc_contiguous’ was here ../etnaviv/viv.h:233:5: error: conflicting types for ‘viv_alloc_linear_vidmem’ viv_raw.h:108:5: note: previous declaration of ‘viv_alloc_linear_vidmem’ was here In file included from ../etnaviv/etna.h:33:0, from empty_screen_gc2000.c:16: ../etnaviv/viv.h:238:5: error: conflicting types for ‘viv_lock_vidmem’ viv_raw.h:112:5: note: previous declaration of ‘viv_lock_vidmem’ was here ../etnaviv/viv.h:242:5: error: conflicting types for ‘viv_commit’ viv_raw.h:119:5: note: previous declaration of ‘viv_commit’ was here ../etnaviv/viv.h:251:5: error: conflicting types for ‘viv_unlock_vidmem’ viv_raw.h:124:5: note: previous declaration of ‘viv_unlock_vidmem’ was here ../etnaviv/viv.h:255:5: error: conflicting types for ‘viv_free_vidmem’ viv_raw.h:128:5: note: previous declaration of ‘viv_free_vidmem’ was here ../etnaviv/viv.h:263:5: error: conflicting types for ‘viv_map_user_memory’ viv_raw.h:132:5: note: previous declaration of ‘viv_map_user_memory’ was here ../etnaviv/viv.h:267:5: error: conflicting types for ‘viv_unmap_user_memory’ viv_raw.h:136:5: note: previous declaration of ‘viv_unmap_user_memory’ was here ../etnaviv/viv.h:271:5: error: conflicting types for ‘viv_event_commit’ viv_raw.h:140:5: note: previous declaration of ‘viv_event_commit’ was here ../etnaviv/viv.h:277:5: error: conflicting types for ‘viv_user_signal_create’ viv_raw.h:146:5: note: previous declaration of ‘viv_user_signal_create’ was here ../etnaviv/viv.h:281:5: error: conflicting types for ‘viv_user_signal_signal’ viv_raw.h:150:5: note: previous declaration of ‘viv_user_signal_signal’ was here ../etnaviv/viv.h:286:5: error: conflicting types for ‘viv_user_signal_wait’ viv_raw.h:154:5: note: previous declaration of ‘viv_user_signal_wait’ was here ../etnaviv/viv.h:290:5: error: conflicting types for ‘viv_user_signal_destroy’ viv_raw.h:158:5: note: previous declaration of ‘viv_user_signal_destroy’ was here ../etnaviv/viv.h:292:6: error: conflicting types for ‘viv_show_chip_info’ viv_raw.h:164:6: note: previous declaration of ‘viv_show_chip_info’ was here ../etnaviv/viv.h:297:5: error: conflicting types for ‘viv_reset’ viv_raw.h:168:5: note: previous declaration of ‘viv_reset’ was here empty_screen_gc2000.c: In function ‘main’: empty_screen_gc2000.c:51:5: warning: passing argument 6 of ‘viv_alloc_linear_vidmem’ from incompatible pointer type [enabled by default] ../etnaviv/viv.h:233:5: note: expected ‘viv_node_t *’ but argument is of type ‘union _gcuVIDMEM_NODE **’ empty_screen_gc2000.c:60:5: warning: passing argument 2 of ‘viv_lock_vidmem’ makes integer from pointer without a cast [enabled by default] ../etnaviv/viv.h:238:5: note: expected ‘viv_node_t’ but argument is of type ‘gcuVIDMEM_NODE_PTR’ empty_screen_gc2000.c:69:5: warning: passing argument 6 of ‘viv_alloc_linear_vidmem’ from incompatible pointer type [enabled by default] ../etnaviv/viv.h:233:5: note: expected ‘viv_node_t *’ but argument is of type ‘union _gcuVIDMEM_NODE **’ empty_screen_gc2000.c:78:5: warning: passing argument 2 of ‘viv_lock_vidmem’ makes integer from pointer without a cast [enabled by default] ../etnaviv/viv.h:238:5: note: expected ‘viv_node_t’ but argument is of type ‘gcuVIDMEM_NODE_PTR’ empty_screen_gc2000.c:87:5: warning: passing argument 6 of ‘viv_alloc_linear_vidmem’ from incompatible pointer type [enabled by default] ../etnaviv/viv.h:233:5: note: expected ‘viv_node_t *’ but argument is of type ‘union _gcuVIDMEM_NODE **’ empty_screen_gc2000.c:96:5: warning: passing argument 2 of ‘viv_lock_vidmem’ makes integer from pointer without a cast [enabled by default] ../etnaviv/viv.h:238:5: note: expected ‘viv_node_t’ but argument is of type ‘gcuVIDMEM_NODE_PTR’ empty_screen_gc2000.c:105:5: warning: passing argument 6 of ‘viv_alloc_linear_vidmem’ from incompatible pointer type [enabled by default] ../etnaviv/viv.h:233:5: note: expected ‘viv_node_t *’ but argument is of type ‘union _gcuVIDMEM_NODE **’ empty_screen_gc2000.c:114:5: warning: passing argument 2 of ‘viv_lock_vidmem’ makes integer from pointer without a cast [enabled by default] ../etnaviv/viv.h:238:5: note: expected ‘viv_node_t’ but argument is of type ‘gcuVIDMEM_NODE_PTR’ empty_screen_gc2000.c:123:5: warning: passing argument 6 of ‘viv_alloc_linear_vidmem’ from incompatible pointer type [enabled by default] ../etnaviv/viv.h:233:5: note: expected ‘viv_node_t *’ but argument is of type ‘union _gcuVIDMEM_NODE **’ empty_screen_gc2000.c:132:5: warning: passing argument 2 of ‘viv_lock_vidmem’ makes integer from pointer without a cast [enabled by default] ../etnaviv/viv.h:238:5: note: expected ‘viv_node_t’ but argument is of type ‘gcuVIDMEM_NODE_PTR’ empty_screen_gc2000.c:188:5: warning: passing argument 3 of ‘viv_commit’ makes integer from pointer without a cast [enabled by default] ../etnaviv/viv.h:242:5: note: expected ‘viv_context_t’ but argument is of type ‘gckCONTEXT’ empty_screen_gc2000.c:188:5: error: too few arguments to function ‘viv_commit’ ../etnaviv/viv.h:242:5: note: declared here empty_screen_gc2000.c:206:5: warning: passing argument 3 of ‘viv_commit’ makes integer from pointer without a cast [enabled by default] ../etnaviv/viv.h:242:5: note: expected ‘viv_context_t’ but argument is of type ‘gckCONTEXT’ empty_screen_gc2000.c:206:5: error: too few arguments to function ‘viv_commit’ ../etnaviv/viv.h:242:5: note: declared here empty_screen_gc2000.c:224:5: warning: passing argument 1 of ‘viv_event_queue_signal’ from incompatible pointer type [enabled by default] viv_raw.h:162:5: note: expected ‘struct viv_conn *’ but argument is of type ‘struct viv_conn *’ empty_screen_gc2000.c:239:5: warning: passing argument 6 of ‘viv_alloc_linear_vidmem’ from incompatible pointer type [enabled by default] ../etnaviv/viv.h:233:5: note: expected ‘viv_node_t *’ but argument is of type ‘union _gcuVIDMEM_NODE **’ empty_screen_gc2000.c:248:5: warning: passing argument 2 of ‘viv_lock_vidmem’ makes integer from pointer without a cast [enabled by default] ../etnaviv/viv.h:238:5: note: expected ‘viv_node_t’ but argument is of type ‘gcuVIDMEM_NODE_PTR’ empty_screen_gc2000.c:268:5: warning: passing argument 3 of ‘viv_commit’ makes integer from pointer without a cast [enabled by default] ../etnaviv/viv.h:242:5: note: expected ‘viv_context_t’ but argument is of type ‘gckCONTEXT’ empty_screen_gc2000.c:268:5: error: too few arguments to function ‘viv_commit’ ../etnaviv/viv.h:242:5: note: declared here empty_screen_gc2000.c:275:5: warning: passing argument 1 of ‘viv_event_queue_signal’ from incompatible pointer type [enabled by default] viv_raw.h:162:5: note: expected ‘struct viv_conn *’ but argument is of type ‘struct viv_conn *’ empty_screen_gc2000.c:289:5: warning: passing argument 2 of ‘viv_unlock_vidmem’ makes integer from pointer without a cast [enabled by default] ../etnaviv/viv.h:251:5: note: expected ‘viv_node_t’ but argument is of type ‘gcuVIDMEM_NODE_PTR’ empty_screen_gc2000.c:289:5: error: too few arguments to function ‘viv_unlock_vidmem’ ../etnaviv/viv.h:251:5: note: declared here make: *** [empty_screen_gc2000.o] Error 1 root@linaro-ubuntu-desktop:~/etna_viv/src/replay# Signed-off-by: Christian Gmeiner <christian.gmeiner@gmail.com>
2013-10-14driver: get rid of caching of gpu/cpu pointersWladimir J. van der Laan
No longer cache logical and GPU address for resource levels but store offsets and query only when needed. This is the first step toward submitting bo's with relocations (just search for `etna_bo_gpu_address` to find all places where GPU addresses are used).
2013-10-12documentation update for RSWladimir J. van der Laan
2013-10-12replay: some build fixesWladimir J. van der Laan
Minimal fixes to get cube_etna2 and cube_etna2_gc2000 to compile again.
2013-10-12Merge pull request #15 from tchebb/yuvfbWladimir J. van der Laan
fb: support YUY2 framebuffers using the FOURCC API
2013-10-12fb: support YUY2 framebuffers using the FOURCC APIThomas Hebb
This is a bit of a hack. FOURCC support should be done properly if more modes are ever added, but it's not worth it for just the one.
2013-10-12etnaviv: viv structure was not zeroed properlyWladimir J. van der Laan
Fixes hang bugs with new fence system.
2013-10-12driver/etnaviv: refactor etna_vidmem into etna_boWladimir J. van der Laan
First step towards making the libetnaviv API more like freedreno_drm. Bo's are not yet fully emulated, but this is a step closer.
2013-10-11old demos: don't call etna_vidmem_unlock directlyWladimir J. van der Laan
2013-10-11driver: use u_minify for mipmap scalingWladimir J. van der Laan
2013-10-11driver: get rid of want_fenceWladimir J. van der Laan
2013-10-11etnaviv: move fence handling to libetnavivWladimir J. van der Laan
This makes fence handles into globally increasing uint32_t's and thus simplifies the handling in the driver, and makes it possible to compare fence handles to see what happened before. The scheme emulates a saner kernel API (closer to adreno drm driver)
2013-10-10utils: include the correct headerWladimir J. van der Laan
Fixes build problem with 'dove' driver.
2013-10-08driver: check for rebind of already bound shader objectsWladimir J. van der Laan
Skip the dirty bit if re-binding already bound shader object.
2013-10-08driver: fix potential segfault in etna_fetch_uniformsWladimir J. van der Laan
2013-10-08driver: comments updates and small cleanupsWladimir J. van der Laan
2013-10-07driver: temp->output optimization can only be done if no swizzleWladimir J. van der Laan
Haven't noticed any cases yet where this causes problems, but it is a bug nevertheless if the MOV is optimized out when it does a swizzle.
2013-10-07rename native directory to srcWladimir J. van der Laan
This is more consistent with other projects.
2013-10-05documentation updateWladimir J. van der Laan
2013-10-05driver: don't ignore NPOT GPU capabilityWladimir J. van der Laan
This enables any kind of wrapping for NPOT textures on GC880/GC2000+ GPUs with the NON_POWER_OF_TWO feature bit set.
2013-10-04rnndb: guess meaning of NON_POWER_OF_TWO capWladimir J. van der Laan
To be tested on hardware that has this capability (gc880/gc2000).
2013-10-04driver: NPOT textures only support clamp to edge wrap modeWladimir J. van der Laan
Force wrapping mode to "clamp to edge" in case client requests otherwise. This fixes the texture issue in Dark Places.
2013-10-04fb_old: some RS experimentsWladimir J. van der Laan
And update documentation.
2013-10-04driver: skip draw when # elements and # VS inputs does not match upWladimir J. van der Laan
This can hang the GPU. This should never happen, but show a BUG and skip draw in case it does.
2013-10-04driver: no need to track RS state in gpu3dWladimir J. van der Laan
2013-10-03Merge pull request #13 from jluebbe/masterWladimir J. van der Laan
small fixes
2013-10-03parse_rng.py: fix self-testJan Luebbe
2013-10-03rnndb: fix misplaced slashJan Luebbe
2013-10-02etnaviv: add gcabi for imx6_v4_1_0Wladimir J. van der Laan
This is the version used in the Solidrun kernel tree for the carrier one.
2013-10-02driver: fix mipmap size compotation for npotWladimir J. van der Laan
GL uses a round-down convention.
generated by cgit (git 2.34.1) at 2025-05-09 21:17:21 +0000