Lines Matching full:radeonsi
90 - [Regression][Bisected][20.2][radeonsi] American Truck Simulator continually allocates memory unti…
114 - radeonsi: radeonsi crashes in Chrome on chromeos
243 - [radeonsi] [glthread] Crash with glthread enabled
252 - [RadeonSI] Glitches on VEGA8 + RX 560X after MR 4863
253 - RadeonSI OpenGL broken for GFX8 after unify code for overriding offset
268 - radeonsi: "Torchlight II" trace showing regression on mesa-20.0.6 [bisected]
831 - radeonsi: Enable tgsi to nir disk cache
866 - radeonsi: Explicitly map Z16_UNORM_S8_UINT to None for GFX10.
867 - amd/common,radeonsi: Move gfx10_format_table to common.
868 - radeonsi: Define gfx10_format in the common header.
897 - radeonsi: Inhibit clock-gating for perf counters.
924 - radeonsi: Work around Wasteland 2 bug.
982 - radeonsi: Fix omitted flush when moving suballocated texture
2670 - radeonsi: Only call nir_lower_var_copies at the end of the opt loop
2881 - radeonsi: Use TRUNC_COORD on samplers
3195 - radeonsi: unify and align down the max SSBO/TBO/UBO buffer binding size
3196 - radeonsi: clean up and deduplicate code around internal compute dispatches
3197 - radeonsi: bind shader images after DCC is disabled for image stores
3198 - radeonsi: add SI_IMAGE_ACCESS_DCC_OFF to ignore DCC for shader images
3199 - radeonsi: implement and use compute-based DCC decompression on gfx9-10
3200 - radeonsi: add a workaround to fix KHR-GL45.texture_view.view_classes on gfx9
3201 - radeonsi: fix si_compute_clear_render_target with render condition enabled
3202 - radeonsi: revert an accidental change in si_clear_buffer
3204 - Revert "ac: reassociate FP expressions for inexact instructions for radeonsi"
3206 - radeonsi: don't wait for idle at the end of gfx IBs
3208 - radeonsi/gfx9: always use IMG_DATA_FORMAT_S8_32 for 8-bit stencil
3209 - radeonsi: allow tc_compatible_htile to be mutable
3210 - radeonsi: enable TC-compatible HTILE on demand for best Z/S performance
3212 - radeonsi: fix compilation of monolithic PS
3217 - ac/surface,radeonsi: move the set/get_bo_metadata code to ac_surface.c
3218 - ac/surface,radeonsi: move the set/get_umd_metadata code into ac_surface.c
3226 - radeonsi: also enable tgsi_to_nir caching for compute shaders
3228 - radeonsi: don't expose 16xAA on chips with 1 RB due to an occlusion query issue
3230 - radeonsi: use correct clear value size for EQAA in expand_fmask
3231 - radeonsi: optimize access pattern for compute blits with linear textures
3232 - radeonsi: tweak clear/copy_buffer limits when to use compute
3233 - radeonsi: simplify setting resource usage for si_init_temp_resource_from_box
3234 - radeonsi: rename SI_RESOURCE_FLAG_TRANSFER to FORCE_LINEAR
3235 - radeonsi: use vi_dcc_enabled instead of using tex->surface.dcc_offset directly
3236 - radeonsi: use display_dcc_offset for setting displayable_dcc_cb_mask
3238 - radeonsi: disable the L2 cache for most CPU mappings of textures
3239 - radeonsi: disable the L2 cache for CPU read mappings of buffers
3240 - radeonsi: compute perf tests - don't test 1 wave/SA limit, test no limit first
3241 - radeonsi: test uncached clear/copy buffer performance with compute shaders
3245 - radeonsi: don't use INDIRECT_BUFFER within IBs
3246 - radeonsi: decrease the max GS invocation count to 32
3247 - Revert "radeonsi: don't wait for idle at the end of gfx IBs"
3249 - radeonsi: move resetting tracked registers into a new function
3250 - radeonsi: split si_all_descriptors_begin_new_cs and rename functions
3251 - radeonsi: don't enable TC-compatible HTILE for stencil if stencil doesn't use it
3252 - radeonsi/gfx8: enable TC-compatible HTILE from the beginning as before
3253 - radeonsi: don't hardcode most perf counter block counts
3256 - radeonsi/gfx10: implement most performance counters
3277 - ac/nir: select v_cvt_pkrtz for all conversions from f32 to f16 for radeonsi
3283 - radeonsi: remove unused leftover code for INDIRECT_BUFFER inside IBs
3284 - radeonsi: remove const_buffers_declared hacks
3285 - radeonsi: pass at most 3 images and/or shader buffers via user SGPRs for compute
3286 - radeonsi: add a hack to disable TRUNC_COORD for shadow samplers
3290 - radeonsi: enable ARB_sparse_buffer
3291 - ac,radeonsi: replace == GFX10 with >= GFX10 where it's needed
3292 - ac,radeonsi: start adding support for gfx10.3
3294 - radeonsi: honor a user-specified pitch on gfx10.3
3295 - radeonsi: enable larger SDMA clears and copies on gfx10.3
3296 - radeonsi: implement R9G9B9E5 render target and image store support on gfx10.3
3297 - radeonsi: move L2_CACHE_CONTROL registers into si_emit_framebuffer_state
3298 - radeonsi: set BIG_PAGE fields on gfx10.3
3299 - radeonsi: don't set any XNACK options on gfx10.3
3301 - radeonsi: add support for Sienna Cichlid
3302 - radeonsi: require LLVM 11 for gfx10.3
3313 - radeonsi: compact MRTs to save PS export memory space
3318 - radeonsi: don't wait for idle at the end of gfx IBs
3319 - radeonsi: replace ctx->screen with sscreen in si_flush_gfx_cs
3324 - radeonsi: don't add the border color buffer into the init_config state
3325 - radeonsi: rename init_config states to cs_preamble states
3326 - radeonsi: don't add the tess ring buffers into the cs_preamble state
3327 - radeonsi: make wait_mem_scratch unmappable
3328 - radeonsi: disallow adding BOs into si_pm4_state except 1 shader BO per state
3329 - radeonsi: make si_pm4_cmd_begin/end static and simplify all usages
3330 - radeonsi: clear per-context buffers at the end of si_create_context
3331 - radeonsi: remove tabs
3332 - radeonsi: don't flush in fence_server_sync
3334 - radeonsi: fix NGG culling for Wave64
3335 - radeonsi: always use Wave32 for GS fast launch, because Wave64 hangs
3336 - radeonsi: always use Wave64 for HS/GS/VS shader stages (except GS fast launch)
3337 - radeonsi: don't try to enable NGG culling for GS
3338 - radeonsi: add a debug option to enable NGG culling for tessellation
3359 - radeonsi: prevent a gfx10_ngg_calculate_subgroup_info failure for TES+NGG GS
3360 - radeonsi: add missing initialization of registers
3361 - radeonsi/gfx10: set the correct value for OFFCHIP_BUFFERING
3362 - radeonsi: sort registers in si_emit_initial_compute_regs according to GPU gen
3363 - radeonsi: sort registers in si_init_cs_preamble_state according to GPU gen
3367 - radeonsi: make cs_preamble_state optional
3368 - radeonsi: reorder code in update_gs_ring_buffers and init_tess_factor_ring
3369 - radeonsi: implement CP register shadowing
3370 - radeonsi: add reg shadowing codepaths to GS and tess ring setup
3371 - radeonsi: add debug code for register shadowing
3372 - radeonsi: don't restore states at the beginning of IBs if they're shadowed
3373 - radeonsi: set up IBs for preemption
3374 - radeonsi: enable preemption if the kernel enabled it
3378 - radeonsi: disable SDMA on gfx9
3379 - radeonsi: reorder NIR optimizations
3380 - radeonsi: call nir_split_array_vars/shrink_vec_array_vars/opt_find_array_copies
3387 - Revert "radeonsi: honor a user-specified pitch on gfx10.3"
3388 - radeonsi: use correct wave size in gfx10_ngg_calculate_subgroup_info
3389 - radeonsi: use the same units for esgs_ring_size and ngg_emit_size
3390 - radeonsi: increase minimum NGG vertex count requirement per workgroup on gfx 10.3
3391 - radeonsi: fix applying the NGG minimum vertex count requirement
3392 - radeonsi: don't count unusable vertices to the NGG LDS size
3393 - radeonsi: add a common function for getting the size of gs_ngg_scratch
3394 - radeonsi: remove the NGG hack decreasing LDS usage to deal with overflows
3395 - radeonsi: various fixes for gfx10.3
3396 - radeonsi: disable NGG culling on gfx10.3 because of hangs
3398 - radeonsi: fix tess levels coming as scalar arrays from SPIR-V
3401 - Revert "ac: generate FMA for inexact instructions for radeonsi"
3472 - Revert "ac,radeonsi: fix compilations issues with LLVM 11"
3502 - ci: Use ignore_scheduled_pipelines anchor in .radeonsi-rules
3693 - radeonsi: fix export count
3696 - radeonsi: add workaround for issue 2647
3697 - radeonsi: don't print gs_copy_shader stats for shaderdb
3706 - radeonsi: add AMD_DEBUG=tmz option
3708 - radeonsi: allocate framebuffer texture as secure when using tmz
3710 - radeonsi: force using staging texture when uploading to secure texture
3711 - radeonsi/sdma: implement tmz support
3713 - radeonsi: add support for PIPE_RESOURCE_FLAG_ENCRYPTED
3715 - radeonsi: determine secure flag must be set for gfx IB
3716 - radeonsi: do not use cmask with encrypted texture
3718 - radeonsi: fix inversed arguments in si_test_gds_memory_management
3720 - radeonsi/sdma: remove useless compare
3721 - radeonsi/drirc: enable zerovram option for 7 Days to Die
3723 - radeonsi: add return value to gfx10_ngg_calculate_subgroup_info
3724 - radeonsi/ngg: try GS multi-cycling mode if default mode failed
3729 - radeonsi: bump SI_NUM_SHADER_BUFFERS to 32
3743 - radeonsi: adjust epitch for PIPE_FORMAT_R8G8_R8B8_UNORM
3744 - radeonsi: extend workaround for KHR-GL45.texture_view.view_classes on gfx9
3747 - radeonsi: enable PIPE_CAP_NO_CLIP_ON_COPY_TEX
3749 - radeonsi,driconf: add clamp_div_by_zero option
3750 - radeonsi: use radeonsi_clamp_div_by_zero for SPECviewperf13, Road Redemption
3763 - radeonsi: remove emacs style config file
3765 - radeonsi: fix syncobj wait timeout
3766 - radeonsi: fix user fence space when MCBP is enabled
3767 - radeonsi: fix max syncobj wait timeout
3768 - radeonsi: fix user fence GPU address
4554 - radeonsi: use PIPE_FORMAT_P010 for 10-bit VP9 decoding
4597 - radeonsi: add missing fallthrough comment
4612 - r600/radeonsi: silence zero-length-bounds gcc warnings
4613 - radeonsi: fix SI_NUM_ATOMS