drm/i915: Tidy workaround batch buffer emission
Subsystems (2)
Tvrtko Ursulin
maintainerPATCH 1/2 drm/i915: Tidy workaround batch buffer emission
From: Tvrtko Ursulin <[email protected]>
Use the "*batch++ = " style as in the ring emission for better readability and also simplify the logic a bit by consolidating the offset and size calculations and overflow checking. The latter is a programming error so it is not required to check for it after each write to the object, but rather do it once the whole state has been written and fail the driver if something went wrong.
v2: Rebase.
--- drivers/gpu/drm/i915/intel_lrc.c | 309 ++++++++++++++------------------------- 1 file changed, 110 insertions(+), 199 deletions(-)
drivers/gpu/drm/i915/intel_lrc.c +110 -199
@@ -890,18 +890,6 @@ static int execlists_request_alloc(struct drm_i915_gem_request *request)return ret; } -#define wa_ctx_emit (batch, index, cmd) \ - do { \ - int __index = (index)++; \ - if (WARN_ON(__index >= (PAGE_SIZE / sizeof(uint32_t)))) { \ - return -ENOSPC; \ - } \ - batch[__index] = (cmd); \ - } while (0) - -#define wa_ctx_emit_reg (batch, index, reg) \ - wa_ctx_emit((batch), (index), i915_mmio_reg_offset(reg)) -/* * In this WA we need to set GEN8_L3SQCREG4[21:21] and reset it after * PIPE_CONTROL instruction. This is required for the flush to happen correctly
@@ -918,56 +906,31 @@ static int execlists_request_alloc(struct drm_i915_gem_request *request)* This WA is also required for Gen9 so extracting as a function avoids * code duplication. */ -static inline int gen8_emit_flush_coherentl3_wa (struct intel_engine_cs *engine, -uint32_t *batch, -uint32_t index) -{ -uint32_t l3sqc4_flush = (0x40400000 |GEN8_LQSC_FLUSH_COHERENT_LINES ); - -wa_ctx_emit (batch, index, (MI_STORE_REGISTER_MEM_GEN8 | -MI_SRM_LRM_GLOBAL_GTT )); -wa_ctx_emit_reg (batch, index,GEN8_L3SQCREG4 ); -wa_ctx_emit (batch, index,i915_ggtt_offset (engine->scratch) +256 ); -wa_ctx_emit (batch, index,0 ); - -wa_ctx_emit (batch, index,MI_LOAD_REGISTER_IMM (1 )); -wa_ctx_emit_reg (batch, index,GEN8_L3SQCREG4 ); -wa_ctx_emit (batch, index, l3sqc4_flush); - -wa_ctx_emit (batch, index,GFX_OP_PIPE_CONTROL (6 )); -wa_ctx_emit (batch, index, (PIPE_CONTROL_CS_STALL | -PIPE_CONTROL_DC_FLUSH_ENABLE )); -wa_ctx_emit (batch, index,0 ); -wa_ctx_emit (batch, index,0 ); -wa_ctx_emit (batch, index,0 ); -wa_ctx_emit (batch, index,0 ); - -wa_ctx_emit (batch, index, (MI_LOAD_REGISTER_MEM_GEN8 | -MI_SRM_LRM_GLOBAL_GTT )); -wa_ctx_emit_reg (batch, index,GEN8_L3SQCREG4 ); -wa_ctx_emit (batch, index,i915_ggtt_offset (engine->scratch) +256 ); -wa_ctx_emit (batch, index,0 ); - -return index; -} - -static inline uint32_t wa_ctx_start (struct i915_wa_ctx_bb *wa_ctx, -uint32_t offset, -uint32_t start_alignment) +static u32 * +gen8_emit_flush_coherentl3_wa (struct intel_engine_cs *engine,u32 *batch) { -return wa_ctx->offset =ALIGN (offset, start_alignment); -} - -static inline int wa_ctx_end (struct i915_wa_ctx_bb *wa_ctx, -uint32_t offset, -uint32_t size_alignment) -{ - wa_ctx->size = offset - wa_ctx->offset; - -WARN (wa_ctx->size % size_alignment, -"wa_ctx_bb failed sanity checks: size %d is not aligned to %d\n" , - wa_ctx->size, size_alignment); -return 0 ; + *batch++ =MI_STORE_REGISTER_MEM_GEN8 |MI_SRM_LRM_GLOBAL_GTT ; + *batch++ =i915_mmio_reg_offset (GEN8_L3SQCREG4 ); + *batch++ =i915_ggtt_offset (engine->scratch) +256 ; + *batch++ =0 ; + + *batch++ =MI_LOAD_REGISTER_IMM (1 ); + *batch++ =i915_mmio_reg_offset (GEN8_L3SQCREG4 ); + *batch++ =0x40400000 |GEN8_LQSC_FLUSH_COHERENT_LINES ; + + *batch++ =GFX_OP_PIPE_CONTROL (6 ); + *batch++ =PIPE_CONTROL_CS_STALL |PIPE_CONTROL_DC_FLUSH_ENABLE ; + *batch++ =0 ; + *batch++ =0 ; + *batch++ =0 ; + *batch++ =0 ; + + *batch++ =MI_LOAD_REGISTER_MEM_GEN8 |MI_SRM_LRM_GLOBAL_GTT ; + *batch++ =i915_mmio_reg_offset (GEN8_L3SQCREG4 ); + *batch++ =i915_ggtt_offset (engine->scratch) +256 ; + *batch++ =0 ; + +return batch; }/*
@@ -985,42 +948,28 @@ static inline int wa_ctx_end(struct i915_wa_ctx_bb *wa_ctx,* MI_BATCH_BUFFER_END will be added to perctx batch and both of them together * makes a complete batch buffer. */ -static int gen8_init_indirectctx_bb (struct intel_engine_cs *engine, -struct i915_wa_ctx_bb *wa_ctx, -uint32_t *batch, -uint32_t *offset) +static u32 *gen8_init_indirectctx_bb (struct intel_engine_cs *engine,u32 *batch) { -uint32_t scratch_addr; -uint32_t index =wa_ctx_start (wa_ctx, *offset,CACHELINE_DWORDS ); -/* WaDisableCtxRestoreArbitration:bdw,chv */ -wa_ctx_emit (batch, index,MI_ARB_ON_OFF |MI_ARB_DISABLE ); + *batch++ =MI_ARB_ON_OFF |MI_ARB_DISABLE ;/* WaFlushCoherentL3CacheLinesAtContextSwitch:bdw */ -if (IS_BROADWELL (engine->i915)) { -int rc =gen8_emit_flush_coherentl3_wa (engine, batch, index); -if (rc <0 ) -return rc; - index = rc; - } +if (IS_BROADWELL (engine->i915)) + batch =gen8_emit_flush_coherentl3_wa (engine, batch); + *batch++ =GFX_OP_PIPE_CONTROL (6 ); + *batch++ =PIPE_CONTROL_FLUSH_L3 |PIPE_CONTROL_GLOBAL_GTT_IVB | +PIPE_CONTROL_CS_STALL |PIPE_CONTROL_QW_WRITE ;/* WaClearSlmSpaceAtContextSwitch:bdw,chv */ /* Actual scratch location is at 128 bytes offset */ - scratch_addr =i915_ggtt_offset (engine->scratch) +2 *CACHELINE_BYTES ; - -wa_ctx_emit (batch, index,GFX_OP_PIPE_CONTROL (6 )); -wa_ctx_emit (batch, index, (PIPE_CONTROL_FLUSH_L3 | -PIPE_CONTROL_GLOBAL_GTT_IVB | -PIPE_CONTROL_CS_STALL | -PIPE_CONTROL_QW_WRITE )); -wa_ctx_emit (batch, index, scratch_addr); -wa_ctx_emit (batch, index,0 ); -wa_ctx_emit (batch, index,0 ); -wa_ctx_emit (batch, index,0 ); + *batch++ =i915_ggtt_offset (engine->scratch) +2 *CACHELINE_BYTES ; + *batch++ =0 ; + *batch++ =0 ; + *batch++ =0 ;/* Pad to end of cacheline */ -while (index %CACHELINE_DWORDS ) -wa_ctx_emit (batch, index,MI_NOOP ); +while ((unsigned long )batch %CACHELINE_BYTES ) + *batch++ =MI_NOOP ;/* * MI_BATCH_BUFFER_END is not required in Indirect ctx BB because
@@ -1028,7 +977,7 @@ static int gen8_init_indirectctx_bb(struct intel_engine_cs *engine,* in the register CTX_RCS_INDIRECT_CTX */ -return wa_ctx_end (wa_ctx, *offset = index,CACHELINE_DWORDS ); +return batch; }/*
@@ -1040,58 +989,38 @@ static int gen8_init_indirectctx_bb(struct intel_engine_cs *engine,* This batch is terminated with MI_BATCH_BUFFER_END and so we need not add padding * to align it with cacheline as padding after MI_BATCH_BUFFER_END is redundant. */ -static int gen8_init_perctx_bb (struct intel_engine_cs *engine, -struct i915_wa_ctx_bb *wa_ctx, -uint32_t *batch, -uint32_t *offset) +static u32 *gen8_init_perctx_bb (struct intel_engine_cs *engine,u32 *batch) { -uint32_t index =wa_ctx_start (wa_ctx, *offset,CACHELINE_DWORDS ); -/* WaDisableCtxRestoreArbitration:bdw,chv */ -wa_ctx_emit (batch, index,MI_ARB_ON_OFF |MI_ARB_ENABLE ); - -wa_ctx_emit (batch, index,MI_BATCH_BUFFER_END ); + *batch++ =MI_ARB_ON_OFF |MI_ARB_ENABLE ; + *batch++ =MI_BATCH_BUFFER_END ; -return wa_ctx_end (wa_ctx, *offset = index,1 ); +return batch; } -static int gen9_init_indirectctx_bb (struct intel_engine_cs *engine, -struct i915_wa_ctx_bb *wa_ctx, -uint32_t *batch, -uint32_t *offset) +static u32 *gen9_init_indirectctx_bb (struct intel_engine_cs *engine,u32 *batch) { -int ret; -struct drm_i915_private *dev_priv = engine->i915; -uint32_t index =wa_ctx_start (wa_ctx, *offset,CACHELINE_DWORDS ); -/* WaFlushCoherentL3CacheLinesAtContextSwitch:skl,bxt,glk */ - ret =gen8_emit_flush_coherentl3_wa (engine, batch, index); -if (ret <0 ) -return ret; - index = ret; + batch =gen8_emit_flush_coherentl3_wa (engine, batch);/* WaDisableGatherAtSetShaderCommonSlice:skl,bxt,kbl,glk */ -wa_ctx_emit (batch, index,MI_LOAD_REGISTER_IMM (1 )); -wa_ctx_emit_reg (batch, index,COMMON_SLICE_CHICKEN2 ); -wa_ctx_emit (batch, index,_MASKED_BIT_DISABLE ( -GEN9_DISABLE_GATHER_AT_SET_SHADER_COMMON_SLICE )); -wa_ctx_emit (batch, index,MI_NOOP ); + *batch++ =MI_LOAD_REGISTER_IMM (1 ); + *batch++ =i915_mmio_reg_offset (COMMON_SLICE_CHICKEN2 ); + *batch++ =_MASKED_BIT_DISABLE ( +GEN9_DISABLE_GATHER_AT_SET_SHADER_COMMON_SLICE ); + *batch++ =MI_NOOP ;/* WaClearSlmSpaceAtContextSwitch:kbl */ /* Actual scratch location is at 128 bytes offset */ -if (IS_KBL_REVID (dev_priv,0 ,KBL_REVID_A0 )) { -u32 scratch_addr = -i915_ggtt_offset (engine->scratch) +2 *CACHELINE_BYTES ; - -wa_ctx_emit (batch, index,GFX_OP_PIPE_CONTROL (6 )); -wa_ctx_emit (batch, index, (PIPE_CONTROL_FLUSH_L3 | -PIPE_CONTROL_GLOBAL_GTT_IVB | -PIPE_CONTROL_CS_STALL | -PIPE_CONTROL_QW_WRITE )); -wa_ctx_emit (batch, index, scratch_addr); -wa_ctx_emit (batch, index,0 ); -wa_ctx_emit (batch, index,0 ); -wa_ctx_emit (batch, index,0 ); +if (IS_KBL_REVID (engine->i915,0 ,KBL_REVID_A0 )) { + *batch++ =GFX_OP_PIPE_CONTROL (6 ); + *batch++ =PIPE_CONTROL_FLUSH_L3 |PIPE_CONTROL_GLOBAL_GTT_IVB | +PIPE_CONTROL_CS_STALL |PIPE_CONTROL_QW_WRITE ; + *batch++ =i915_ggtt_offset (engine->scratch) + +2 *CACHELINE_BYTES ; + *batch++ =0 ; + *batch++ =0 ; + *batch++ =0 ; }/* WaMediaPoolStateCmdInWABB:bxt,glk */
@@ -1109,41 +1038,37 @@ static int gen9_init_indirectctx_bb(struct intel_engine_cs *engine,* possible configurations, to avoid duplication they are * not shown here again. */ -u32 eu_pool_config =0x00777000 ; -wa_ctx_emit (batch, index,GEN9_MEDIA_POOL_STATE ); -wa_ctx_emit (batch, index,GEN9_MEDIA_POOL_ENABLE ); -wa_ctx_emit (batch, index, eu_pool_config); -wa_ctx_emit (batch, index,0 ); -wa_ctx_emit (batch, index,0 ); -wa_ctx_emit (batch, index,0 ); + *batch++ =GEN9_MEDIA_POOL_STATE ; + *batch++ =GEN9_MEDIA_POOL_ENABLE ; + *batch++ =0x00777000 ; + *batch++ =0 ; + *batch++ =0 ; + *batch++ =0 ; }/* Pad to end of cacheline */ -while (index %CACHELINE_DWORDS ) -wa_ctx_emit (batch, index,MI_NOOP ); +while ((unsigned long )batch %CACHELINE_BYTES ) + *batch++ =MI_NOOP ; -return wa_ctx_end (wa_ctx, *offset = index,CACHELINE_DWORDS ); +return batch; } -static int gen9_init_perctx_bb (struct intel_engine_cs *engine, -struct i915_wa_ctx_bb *wa_ctx, -uint32_t *batch, -uint32_t *offset) +static u32 *gen9_init_perctx_bb (struct intel_engine_cs *engine,u32 *batch) { -uint32_t index =wa_ctx_start (wa_ctx, *offset,CACHELINE_DWORDS ); - -wa_ctx_emit (batch, index,MI_BATCH_BUFFER_END ); + *batch++ =MI_BATCH_BUFFER_END ; -return wa_ctx_end (wa_ctx, *offset = index,1 ); +return batch; } -static int lrc_setup_wa_ctx_obj (struct intel_engine_cs *engine,u32 size) +#define CTX_WA_BB_OBJ_SIZE (PAGE_SIZE) + +static int lrc_setup_wa_ctx (struct intel_engine_cs *engine) {struct drm_i915_gem_object *obj;struct i915_vma *vma;int err; - obj =i915_gem_object_create (engine->i915,PAGE_ALIGN (size)); + obj =i915_gem_object_create (engine->i915,CTX_WA_BB_OBJ_SIZE );if (IS_ERR (obj))return PTR_ERR (obj);
@@ -1165,80 +1090,66 @@ static int lrc_setup_wa_ctx_obj(struct intel_engine_cs *engine, u32 size)return err; } -static void lrc_destroy_wa_ctx_obj (struct intel_engine_cs *engine) +static void lrc_destroy_wa_ctx (struct intel_engine_cs *engine) {i915_vma_unpin_and_release (&engine->wa_ctx.vma); } +typedef u32 *(*wa_bb_func_t )(struct intel_engine_cs *engine,u32 *batch); +static int intel_init_workaround_bb (struct intel_engine_cs *engine) {struct i915_ctx_workarounds *wa_ctx = &engine->wa_ctx; -uint32_t *batch; -uint32_t offset; +struct i915_wa_ctx_bb *wa_bb[2 ] = { &wa_ctx->indirect_ctx, + &wa_ctx->per_ctx }; +wa_bb_func_t wa_bb_f[2 ];struct page *page; +u32 *batch, *batch_ptr; +unsigned int i;int ret; -WARN_ON (engine->id !=RCS ); +if (WARN_ON (engine->id !=RCS || !engine->scratch)) +return -EINVAL ; -/* update this when WA for higher Gen are added */ -if (INTEL_GEN (engine->i915) >9 ) { -DRM_ERROR ("WA batch buffer is not initialized for Gen%d\n" , -INTEL_GEN (engine->i915)); +switch (INTEL_GEN (engine->i915)) { +case 9 : + wa_bb_f[0 ] = gen9_init_indirectctx_bb; + wa_bb_f[1 ] = gen9_init_perctx_bb; +break ; +case 8 : + wa_bb_f[0 ] = gen8_init_indirectctx_bb; + wa_bb_f[1 ] = gen8_init_perctx_bb; +break ; +default : +MISSING_CASE (INTEL_GEN (engine->i915));return 0 ; } -/* some WA perform writes to scratch page, ensure it is valid */ -if (!engine->scratch) { -DRM_ERROR ("scratch page not allocated for %s\n" , engine->name); -return -EINVAL ; - } - - ret =lrc_setup_wa_ctx_obj (engine,PAGE_SIZE ); + ret =lrc_setup_wa_ctx (engine);if (ret) {DRM_DEBUG_DRIVER ("Failed to setup context WA page: %d\n" , ret);return ret; } page =i915_gem_object_get_dirty_page (wa_ctx->vma->obj,0 ); - batch =kmap_atomic (page); - offset =0 ; - -if (IS_GEN8 (engine->i915)) { - ret =gen8_init_indirectctx_bb (engine, - &wa_ctx->indirect_ctx, - batch, - &offset); -if (ret) - goto out; + batch = batch_ptr =kmap_atomic (page); - ret =gen8_init_perctx_bb (engine, - &wa_ctx->per_ctx, - batch, - &offset); -if (ret) - goto out; - }else if (IS_GEN9 (engine->i915)) { - ret =gen9_init_indirectctx_bb (engine, - &wa_ctx->indirect_ctx, - batch, - &offset); -if (ret) - goto out; - - ret =gen9_init_perctx_bb (engine, - &wa_ctx->per_ctx, - batch, - &offset); -if (ret) - goto out; +/* +* Emit the two workaround batch buffers, recording the offset from the +* start of the workaround batch buffer object for each and their +* respective sizes. +*/ +for (i =0 ; i <ARRAY_SIZE (wa_bb_f); i++) { + wa_bb[i]->offset =ALIGN (batch_ptr - batch,CACHELINE_DWORDS ); + batch_ptr = wa_bb_f[i](engine, batch_ptr); + wa_bb[i]->size = batch_ptr - &batch[wa_bb[i]->offset]; } -out: +BUG_ON (batch_ptr - batch >CTX_WA_BB_OBJ_SIZE /sizeof (u32)); +kunmap_atomic (batch); -if (ret) -lrc_destroy_wa_ctx_obj (engine); -return ret; +return 0 ; }static u32 port_seqno (struct execlist_port *port)
@@ -1685,7 +1596,7 @@ void intel_logical_ring_cleanup(struct intel_engine_cs *engine)intel_engine_cleanup_common (engine); -lrc_destroy_wa_ctx_obj (engine); +lrc_destroy_wa_ctx (engine); engine->i915 =NULL ; dev_priv->engine[engine->id] =NULL ;kfree (engine);
Signature · 6 lines
2.9.3 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Tvrtko Ursulin
maintainerPATCH 2/2 drm/i915: Consolidate gen8_emit_pipe_control
From: Tvrtko Ursulin <[email protected]>
We have a few open coded instances in the execlists code and an almost suitable helper in intel_ringbuf.c
We can consolidate to a single helper if we change the existing helper to emit directly to ring buffer memory and move the space reservation outside it.
--- drivers/gpu/drm/i915/intel_lrc.c | 77 +++++++++++---------------------- drivers/gpu/drm/i915/intel_ringbuffer.c | 45 +++++++------------ drivers/gpu/drm/i915/intel_ringbuffer.h | 12 +++++ 3 files changed, 53 insertions(+), 81 deletions(-)
drivers/gpu/drm/i915/intel_lrc.c +26 -51
@@ -918,12 +918,10 @@ gen8_emit_flush_coherentl3_wa(struct intel_engine_cs *engine, u32 *batch) *batch++ =i915_mmio_reg_offset (GEN8_L3SQCREG4 ); *batch++ =0x40400000 |GEN8_LQSC_FLUSH_COHERENT_LINES ; - *batch++ =GFX_OP_PIPE_CONTROL (6 ); - *batch++ =PIPE_CONTROL_CS_STALL |PIPE_CONTROL_DC_FLUSH_ENABLE ; - *batch++ =0 ; - *batch++ =0 ; - *batch++ =0 ; - *batch++ =0 ; + batch =gen8_emit_pipe_control (batch, +PIPE_CONTROL_CS_STALL | +PIPE_CONTROL_DC_FLUSH_ENABLE , +0 ); *batch++ =MI_LOAD_REGISTER_MEM_GEN8 |MI_SRM_LRM_GLOBAL_GTT ; *batch++ =i915_mmio_reg_offset (GEN8_L3SQCREG4 );
@@ -957,15 +955,15 @@ static u32 *gen8_init_indirectctx_bb(struct intel_engine_cs *engine, u32 *batch)if (IS_BROADWELL (engine->i915)) batch =gen8_emit_flush_coherentl3_wa (engine, batch); - *batch++ =GFX_OP_PIPE_CONTROL (6 ); - *batch++ =PIPE_CONTROL_FLUSH_L3 |PIPE_CONTROL_GLOBAL_GTT_IVB | -PIPE_CONTROL_CS_STALL |PIPE_CONTROL_QW_WRITE ;/* WaClearSlmSpaceAtContextSwitch:bdw,chv */ /* Actual scratch location is at 128 bytes offset */ - *batch++ =i915_ggtt_offset (engine->scratch) +2 *CACHELINE_BYTES ; - *batch++ =0 ; - *batch++ =0 ; - *batch++ =0 ; + batch =gen8_emit_pipe_control (batch, +PIPE_CONTROL_FLUSH_L3 | +PIPE_CONTROL_GLOBAL_GTT_IVB | +PIPE_CONTROL_CS_STALL | +PIPE_CONTROL_QW_WRITE , +i915_ggtt_offset (engine->scratch) + +2 *CACHELINE_BYTES );/* Pad to end of cacheline */ while ((unsigned long )batch %CACHELINE_BYTES )
@@ -1013,14 +1011,13 @@ static u32 *gen9_init_indirectctx_bb(struct intel_engine_cs *engine, u32 *batch)/* WaClearSlmSpaceAtContextSwitch:kbl */ /* Actual scratch location is at 128 bytes offset */ if (IS_KBL_REVID (engine->i915,0 ,KBL_REVID_A0 )) { - *batch++ =GFX_OP_PIPE_CONTROL (6 ); - *batch++ =PIPE_CONTROL_FLUSH_L3 |PIPE_CONTROL_GLOBAL_GTT_IVB | -PIPE_CONTROL_CS_STALL |PIPE_CONTROL_QW_WRITE ; - *batch++ =i915_ggtt_offset (engine->scratch) + -2 *CACHELINE_BYTES ; - *batch++ =0 ; - *batch++ =0 ; - *batch++ =0 ; + batch =gen8_emit_pipe_control (batch, +PIPE_CONTROL_FLUSH_L3 | +PIPE_CONTROL_GLOBAL_GTT_IVB | +PIPE_CONTROL_CS_STALL | +PIPE_CONTROL_QW_WRITE , +i915_ggtt_offset (engine->scratch) + +2 *CACHELINE_BYTES ); }/* WaMediaPoolStateCmdInWABB:bxt,glk */
@@ -1450,39 +1447,17 @@ static int gen8_emit_flush_render(struct drm_i915_gem_request *request,if (IS_ERR (cs))return PTR_ERR (cs); -if (vf_flush_wa) { - *cs++ =GFX_OP_PIPE_CONTROL (6 ); - *cs++ =0 ; - *cs++ =0 ; - *cs++ =0 ; - *cs++ =0 ; - *cs++ =0 ; - } +if (vf_flush_wa) + cs =gen8_emit_pipe_control (cs,0 ,0 ); -if (dc_flush_wa) { - *cs++ =GFX_OP_PIPE_CONTROL (6 ); - *cs++ =PIPE_CONTROL_DC_FLUSH_ENABLE ; - *cs++ =0 ; - *cs++ =0 ; - *cs++ =0 ; - *cs++ =0 ; - } +if (dc_flush_wa) + cs =gen8_emit_pipe_control (cs,PIPE_CONTROL_DC_FLUSH_ENABLE , +0 ); - *cs++ =GFX_OP_PIPE_CONTROL (6 ); - *cs++ = flags; - *cs++ = scratch_addr; - *cs++ =0 ; - *cs++ =0 ; - *cs++ =0 ; + cs =gen8_emit_pipe_control (cs, flags, scratch_addr); -if (dc_flush_wa) { - *cs++ =GFX_OP_PIPE_CONTROL (6 ); - *cs++ =PIPE_CONTROL_CS_STALL ; - *cs++ =0 ; - *cs++ =0 ; - *cs++ =0 ; - *cs++ =0 ; - } +if (dc_flush_wa) + cs =gen8_emit_pipe_control (cs,PIPE_CONTROL_CS_STALL ,0 );intel_ring_advance (request, cs);
drivers/gpu/drm/i915/intel_ringbuffer.c +15 -30
65 lines not on this page · raw message
drivers/gpu/drm/i915/intel_ringbuffer.h +12 -0
@@ -632,4 +632,16 @@ void intel_engine_reset_breadcrumbs(struct intel_engine_cs *engine);void intel_engine_fini_breadcrumbs (struct intel_engine_cs *engine);unsigned int intel_breadcrumbs_busy (struct drm_i915_private *i915); +static inline u32 *gen8_emit_pipe_control (u32 *batch,u32 flags,u32 offset) +{ +static const u32 pc6[6 ] = {GFX_OP_PIPE_CONTROL (6 ),0 ,0 ,0 ,0 ,0 }; + +memcpy (batch, pc6,sizeof (pc6)); + + batch[1 ] = flags; + batch[2 ] = offset; + +return batch +6 ; +} +#endif /* _INTEL_RINGBUFFER_H_ */
Signature · 6 lines
2.9.3 _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Re: [PATCH 2/2] drm/i915: Consolidate gen8_emit_pipe_control
drivers/gpu/drm/i915/intel_ringbuffer.h void intel_engine_reset_breadcrumbs(struct intel_engine_cs *engine); · line 644
Quoted 10 lines · Tvrtko Ursulin
drivers/gpu/drm/i915/intel_ringbuffer.h +10
-0+static inline u32 *gen8_emit_pipe_control (u32 *batch,u32 flags,u32 offset) +{ +static const u32 pc6[6 ] = {GFX_OP_PIPE_CONTROL (6 ),0 ,0 ,0 ,0 ,0 }; + +memcpy (batch, pc6,sizeof (pc6)); + + batch[1 ] = flags; + batch[2 ] = offset; + +return batch +6 ;
godbolt would seem to say it is best to use
static inline u32 *gen8_emit_pipe_control(u32 *batch, u32 flags, u32 offset)
{
batch[0] = GFX_OP_PIPE_CONTROL(6);
batch[1] = flags;
batch[2] = offset;
batch[3] = 0;
batch[4] = 0;
batch[5] = 0;
return batch + 6;
}-Chris
Signature · 5 lines
Chris Wilson, Intel Open Source Technology Centre _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Re: [PATCH 2/2] drm/i915: Consolidate gen8_emit_pipe_control
Quoted 14 lines · Chris Wilson
drivers/gpu/drm/i915/intel_ringbuffer.h void intel_engine_reset_breadcrumbs(struct intel_engine_cs *engine); · line 644
Quoted 10 lines · Tvrtko Ursulin
drivers/gpu/drm/i915/intel_ringbuffer.h +10
-0+static inline u32 *gen8_emit_pipe_control (u32 *batch,u32 flags,u32 offset) +{ +static const u32 pc6[6 ] = {GFX_OP_PIPE_CONTROL (6 ),0 ,0 ,0 ,0 ,0 }; + +memcpy (batch, pc6,sizeof (pc6)); + + batch[1 ] = flags; + batch[2 ] = offset; + +return batch +6 ;godbolt would seem to say it is best to use
static inline u32 *gen8_emit_pipe_control(u32 *batch, u32 flags, u32 offset)
Quoted 10 lines
{
batch[0] = GFX_OP_PIPE_CONTROL(6);
batch[1] = flags;
batch[2] = offset;
batch[3] = 0;
batch[4] = 0;
batch[5] = 0;
return batch + 6;
}Yeah agreed, it was a bit silly. I falsely remember it had quite good effects on the optimisation gcc was able to do but couldn't repro that.
How about though replacing the last three assignments with memset(&batch[3], 0, 3 * sizeof(u32))? That is indeed helpful on 64-bit.
Regards,
Tvrtko
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Re: [PATCH 2/2] drm/i915: Consolidate gen8_emit_pipe_control
Quoted 27 lines · Tvrtko Ursulin
Quoted 14 lines · Chris Wilson
drivers/gpu/drm/i915/intel_ringbuffer.h void intel_engine_reset_breadcrumbs(struct intel_engine_cs *engine); · line 644
Quoted 10 lines · Tvrtko Ursulin
drivers/gpu/drm/i915/intel_ringbuffer.h +10
-0+static inline u32 *gen8_emit_pipe_control (u32 *batch,u32 flags,u32 offset) +{ +static const u32 pc6[6 ] = {GFX_OP_PIPE_CONTROL (6 ),0 ,0 ,0 ,0 ,0 }; + +memcpy (batch, pc6,sizeof (pc6)); + + batch[1 ] = flags; + batch[2 ] = offset; + +return batch +6 ;godbolt would seem to say it is best to use
static inline u32 *gen8_emit_pipe_control(u32 *batch, u32 flags, u32 offset)Quoted 10 lines
{ batch[0] = GFX_OP_PIPE_CONTROL(6); batch[1] = flags; batch[2] = offset; batch[3] = 0; batch[4] = 0; batch[5] = 0; return batch + 6; }
Quoted 7 lines
Yeah agreed, it was a bit silly. I falsely remember it had quite good effects on the optimisation gcc was able to do but couldn't repro that.
How about though replacing the last three assignments with
memset(&batch[3], 0, 3 * sizeof(u32))? That is indeed helpful on 64-bit.
Hah. Yes. Probably something to do with C preventing combining adjoining writes to memory? With memset it uses a *(uint64_t *)&batch[3] = 0, and we are not going to write that ugly code ourselves ;)
Once we accept that gcc will inline the memset, it becomes equally as good just to use memset(batch, 0, 6*sizeof(u32)). Just need to double check that the kernel cflags don't prevent that magic. -Chris
--
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Re: [PATCH 2/2] drm/i915: Consolidate gen8_emit_pipe_control
In reply to Tvrtko UrsulinQuoted 46 lines · Tvrtko Ursulin
drivers/gpu/drm/i915/intel_ringbuffer.c +6
-18@@ -334,35 +334,16 @@ gen7_render_ring_flush(struct drm_i915_gem_request *req, u32 mode) }static int -gen8_emit_pipe_control (struct drm_i915_gem_request *req, -u32 flags,u32 scratch_addr) +gen8_render_ring_flush (struct drm_i915_gem_request *req,u32 mode) { +u32 flags;u32 *cs; - cs =intel_ring_begin (req,6 ); + cs =intel_ring_begin (req, mode &EMIT_INVALIDATE ?12 :6 );if (IS_ERR (cs))return PTR_ERR (cs); -static int -gen8_render_ring_flush (struct drm_i915_gem_request *req,u32 mode) -{ -u32 scratch_addr = -i915_ggtt_offset (req->engine->scratch) +2 *CACHELINE_BYTES ; -u32 flags =0 ; -int ret; - - flags |=PIPE_CONTROL_CS_STALL ; + flags =PIPE_CONTROL_CS_STALL ;if (mode &EMIT_FLUSH ) { flags |=PIPE_CONTROL_RENDER_TARGET_CACHE_FLUSH ;@@ -381,15 +362,19 @@ gen8_render_ring_flush(struct drm_i915_gem_request *req, u32 mode) flags |=PIPE_CONTROL_GLOBAL_GTT_IVB ;/* WaCsStallBeforeStateCacheInvalidate:bdw,chv */ - ret =gen8_emit_pipe_control (req, -PIPE_CONTROL_CS_STALL | -PIPE_CONTROL_STALL_AT_SCOREBOARD , -0 ); -if (ret) -return ret; + cs = gen8_emit_pipe_control(cs, + PIPE_CONTROL_CS_STALL |
drivers/gpu/drm/i915/intel_ringbuffer.c gen8_render_ring_flush(struct drm_i915_gem_request *req, u32 mode) · line 375
Quoted 10 lines
drivers/gpu/drm/i915/intel_ringbuffer.c +7
-1+PIPE_CONTROL_STALL_AT_SCOREBOARD , + 0); } -return gen8_emit_pipe_control (req, flags, scratch_addr); + cs =gen8_emit_pipe_control (cs, flags, +i915_ggtt_offset (req->engine->scratch) + +2 *CACHELINE_BYTES ); + +intel_ring_advance (req, cs);
That looks scary in the diff, but correct.
drivers/gpu/drm/i915/intel_ringbuffer.c gen8_render_ring_flush(struct drm_i915_gem_request *req, u32 mode)
Quoted 11 lines
drivers/gpu/drm/i915/intel_ringbuffer.c +11
-0+static inline u32 *gen8_emit_pipe_control (u32 *batch,u32 flags,u32 offset) +{ +static const u32 pc6[6 ] = {GFX_OP_PIPE_CONTROL (6 ),0 ,0 ,0 ,0 ,0 }; + +memcpy (batch, pc6,sizeof (pc6)); + + batch[1 ] = flags; + batch[2 ] = offset; + +return batch +6 ; +}
Other than quibbling over code generation here,
Coverage of pipecontrol side-effects is lax in igt, piglit is the best bet to see if it has any unintentional impact.
-Chris
Signature · 5 lines
Chris Wilson, Intel Open Source Technology Centre _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Patchwork
CI✗ Fi.CI.BAT: failure for series starting with [1/2] drm/i915: Tidy workaround batch buffer emission
In reply to Tvrtko Ursulin== Series Details ==
== Series Details ==
Series: series starting with [1/2] drm/i915: Tidy workaround batch buffer emission
URL : https://patchwork.freedesktop.org/series/19715/
State : failure
== Summary ==
Series 19715v1 Series without cover letter
https://patchwork.freedesktop.org/api/1.0/series/19715/revisions/1/mbox/
Test kms_pipe_crc_basic:
Subgroup read-crc-pipe-b:
pass -> DMESG-FAIL (fi-snb-2520m)
Subgroup read-crc-pipe-b-frame-sequence:
pass -> DMESG-FAIL (fi-snb-2520m)
Subgroup suspend-read-crc-pipe-a:
pass -> DMESG-WARN (fi-snb-2520m)fi-bdw-5557u total:252 pass:238 dwarn:3 dfail:0 fail:0 skip:11 fi-bsw-n3050 total:252 pass:210 dwarn:3 dfail:0 fail:0 skip:39 fi-bxt-j4205 total:252 pass:230 dwarn:3 dfail:0 fail:0 skip:19 fi-bxt-t5700 total:83 pass:70 dwarn:0 dfail:0 fail:0 skip:12 fi-byt-j1900 total:252 pass:222 dwarn:3 dfail:0 fail:0 skip:27 fi-byt-n2820 total:252 pass:218 dwarn:3 dfail:0 fail:0 skip:31 fi-hsw-4770 total:252 pass:233 dwarn:3 dfail:0 fail:0 skip:16 fi-hsw-4770r total:252 pass:233 dwarn:3 dfail:0 fail:0 skip:16 fi-ilk-650 total:252 pass:199 dwarn:3 dfail:0 fail:0 skip:50 fi-ivb-3520m total:252 pass:231 dwarn:3 dfail:0 fail:0 skip:18 fi-ivb-3770 total:252 pass:231 dwarn:3 dfail:0 fail:0 skip:18 fi-kbl-7500u total:252 pass:231 dwarn:3 dfail:0 fail:0 skip:18 fi-skl-6260u total:252 pass:239 dwarn:3 dfail:0 fail:0 skip:10 fi-skl-6700hq total:252 pass:232 dwarn:3 dfail:0 fail:0 skip:17 fi-skl-6700k total:252 pass:230 dwarn:4 dfail:0 fail:0 skip:18 fi-skl-6770hq total:252 pass:239 dwarn:3 dfail:0 fail:0 skip:10 fi-snb-2520m total:252 pass:218 dwarn:4 dfail:2 fail:0 skip:28 fi-snb-2600 total:252 pass:220 dwarn:3 dfail:0 fail:0 skip:29
890e171e84eb11944701de9d53c1162dd5c38142 drm-tip: 2017y-02m-15d-15h-34m-13s UTC integration manifest 17d9d3f9 drm/i915: Consolidate gen8_emit_pipe_control
b3317e0 drm/i915: Tidy workaround batch buffer emission
== Logs ==
For more details see: https://intel-gfx-ci.01.org/CI/Patchwork_3830/
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Quoted 11 lines · Tvrtko Ursulin
From: Tvrtko Ursulin <[email protected]>
Use the "*batch++ = " style as in the ring emission for better readability and also simplify the logic a bit by consolidating the offset and size calculations and overflow checking. The latter is a programming error so it is not required to check for it after each write to the object, but rather do it once the whole state has been written and fail the driver if something went wrong.
v2: Rebase.
I did not spot any mistakes in the mechanical translations.
drivers/gpu/drm/i915/intel_lrc.c static int lrc_setup_wa_ctx_obj(struct intel_engine_cs *engine, u32 size) · line 1144
Quoted 8 lines
drivers/gpu/drm/i915/intel_lrc.c +8
-0+/* +* Emit the two workaround batch buffers, recording the offset from the +* start of the workaround batch buffer object for each and their +* respective sizes. +*/ +for (i =0 ; i <ARRAY_SIZE (wa_bb_f); i++) { + wa_bb[i]->offset =ALIGN (batch_ptr - batch,CACHELINE_DWORDS ); + batch_ptr = wa_bb_f[i](engine, batch_ptr);
I'm just a bit more familar with using _fn for function pointers.
drivers/gpu/drm/i915/intel_lrc.c static int lrc_setup_wa_ctx_obj(struct intel_engine_cs *engine, u32 size) · line 1145
Quoted 1 line
drivers/gpu/drm/i915/intel_lrc.c +1
-0+ wa_bb[i]->size = batch_ptr - &batch[wa_bb[i]->offset];
Surprised me that we store the offset/size in dwords not bytes. And then convert to bytes to store in the registers.
drivers/gpu/drm/i915/intel_lrc.c static int lrc_setup_wa_ctx_obj(struct intel_engine_cs *engine, u32 size) · line 1148
Quoted 4 lines
drivers/gpu/drm/i915/intel_lrc.c +1
-1} -out: +BUG_ON (batch_ptr - batch >CTX_WA_BB_OBJ_SIZE /sizeof (u32));
Would it make sense going forward just use void *batch, batch_ptr here and compute bytes?
-Chris
Signature · 5 lines
Chris Wilson, Intel Open Source Technology Centre _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Quoted 15 lines · Chris Wilson
Quoted 11 lines · Tvrtko Ursulin
From: Tvrtko Ursulin <[email protected]>
Use the "*batch++ = " style as in the ring emission for better readability and also simplify the logic a bit by consolidating the offset and size calculations and overflow checking. The latter is a programming error so it is not required to check for it after each write to the object, but rather do it once the whole state has been written and fail the driver if something went wrong.
v2: Rebase.
I did not spot any mistakes in the mechanical translations.
Quoted 10 lines
drivers/gpu/drm/i915/intel_lrc.c static int lrc_setup_wa_ctx_obj(struct intel_engine_cs *engine, u32 size) · line 1144
Quoted 8 lines
drivers/gpu/drm/i915/intel_lrc.c +8
-0+/* +* Emit the two workaround batch buffers, recording the offset from the +* start of the workaround batch buffer object for each and their +* respective sizes. +*/ +for (i =0 ; i <ARRAY_SIZE (wa_bb_f); i++) { + wa_bb[i]->offset =ALIGN (batch_ptr - batch,CACHELINE_DWORDS ); + batch_ptr = wa_bb_f[i](engine, batch_ptr);I'm just a bit more familar with using
_fnfor function pointers.
Will change it, pattern must have escaped me.
Quoted 4 lines
drivers/gpu/drm/i915/intel_lrc.c static int lrc_setup_wa_ctx_obj(struct intel_engine_cs *engine, u32 size) · line 1145
Quoted 1 line
drivers/gpu/drm/i915/intel_lrc.c +1
-0+ wa_bb[i]->size = batch_ptr - &batch[wa_bb[i]->offset];Surprised me that we store the offset/size in dwords not bytes. And then convert to bytes to store in the registers.
I'll check it out.
Quoted 7 lines
drivers/gpu/drm/i915/intel_lrc.c static int lrc_setup_wa_ctx_obj(struct intel_engine_cs *engine, u32 size) · line 1148
Quoted 4 lines
drivers/gpu/drm/i915/intel_lrc.c +1
-1} -out: +BUG_ON (batch_ptr - batch >CTX_WA_BB_OBJ_SIZE /sizeof (u32));Would it make sense going forward just use void *batch,
batch_ptrhere and compute bytes?
Together with this.
Quoted 1 line
Reviewed-by Chris Wilson <[email protected]>
Thanks. So you think it is worth it? I am just annoyed with the BUG_ON. I don't like to add them, but GEM_BUG_ON seemed to weak. On balance it felt that it is OK. Overall I think it is easier to follow the flow from the code with this organisation and it is more compact in all respects.
Regards,
Tvrtko
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Quoted 5 lines · Tvrtko Ursulin
Quoted 1 line · Chris Wilson
Reviewed-by Chris Wilson <[email protected]>Thanks. So you think it is worth it?
Yes. As you say, it brings it into line with the rest of the command emission sequences - and using a consistent pattern is worth the investment.
Quoted 2 lines
I am just annoyed with the
BUG_ON. I don't like to add them, butGEM_BUG_ONseemed to weak.
That came across in your choice. At the point we hit the BUG_ON, the system is probably/should be dead already.
Quoted 4 lines
On
balance it felt that it is OK. Overall I think it is easier to follow the flow from the code with this organisation and it is more compact in all respects.
gen8_emit_pipe_control() is all the justification I need. That makes the benefits very clear.
-Chris
Signature · 5 lines
Chris Wilson, Intel Open Source Technology Centre _______________________________________________ Intel-gfx mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/intel-gfx
Patchwork
CI✓ Fi.CI.BAT: success for series starting with [1/2] drm/i915: Tidy workaround batch buffer emission
In reply to Tvrtko Ursulin== Series Details ==
== Series Details ==
Series: series starting with [1/2] drm/i915: Tidy workaround batch buffer emission
URL : https://patchwork.freedesktop.org/series/19715/
State : success
== Summary ==
Series 19715v1 Series without cover letter
https://patchwork.freedesktop.org/api/1.0/series/19715/revisions/1/mbox/
fi-bdw-5557u total:252 pass:241 dwarn:0 dfail:0 fail:0 skip:11 fi-bsw-n3050 total:252 pass:213 dwarn:0 dfail:0 fail:0 skip:39 fi-bxt-j4205 total:252 pass:233 dwarn:0 dfail:0 fail:0 skip:19 fi-bxt-t5700 total:83 pass:70 dwarn:0 dfail:0 fail:0 skip:12 fi-byt-j1900 total:252 pass:225 dwarn:0 dfail:0 fail:0 skip:27 fi-byt-n2820 total:252 pass:221 dwarn:0 dfail:0 fail:0 skip:31 fi-hsw-4770 total:252 pass:236 dwarn:0 dfail:0 fail:0 skip:16 fi-hsw-4770r total:252 pass:236 dwarn:0 dfail:0 fail:0 skip:16 fi-ilk-650 total:252 pass:202 dwarn:0 dfail:0 fail:0 skip:50 fi-ivb-3520m total:252 pass:234 dwarn:0 dfail:0 fail:0 skip:18 fi-ivb-3770 total:252 pass:234 dwarn:0 dfail:0 fail:0 skip:18 fi-kbl-7500u total:252 pass:234 dwarn:0 dfail:0 fail:0 skip:18 fi-skl-6260u total:252 pass:242 dwarn:0 dfail:0 fail:0 skip:10 fi-skl-6700hq total:252 pass:235 dwarn:0 dfail:0 fail:0 skip:17 fi-skl-6700k total:252 pass:230 dwarn:4 dfail:0 fail:0 skip:18 fi-skl-6770hq total:252 pass:242 dwarn:0 dfail:0 fail:0 skip:10 fi-snb-2520m total:252 pass:224 dwarn:0 dfail:0 fail:0 skip:28 fi-snb-2600 total:252 pass:223 dwarn:0 dfail:0 fail:0 skip:29
89a932d98b6e1733011019c9872583a9c7c8fda3 drm-tip: 2017y-02m-16d-10h-06m-42s UTC integration manifest 971886b drm/i915: Consolidate gen8_emit_pipe_control
14969ae drm/i915: Tidy workaround batch buffer emission
== Logs ==
For more details see: https://intel-gfx-ci.01.org/CI/Patchwork_3843/
_______________________________________________
Intel-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/intel-gfx