[Beignet] [PATCH] GBE: fixup/refine a bug for image1D array's extra binding index handling.
Zhigang Gong
zhigang.gong at intel.com
Wed Sep 3 17:06:09 PDT 2014
Due to hardware limitation on Gen7/Gen75 when sampling a
surface with clamp address mode and nearest filter mode
on a integer image1Darray type surface, we have to bind
one buffer to to bti. The previous implementation hard
coded it to 128 + original index and when check whether
it is such type bti in driver layer, assume the bti reserved
is 3 which is wrong now.
This patch fixed those hard coded functions and use the
macros defined in the program.h.
Signed-off-by: Zhigang Gong <zhigang.gong at intel.com>
---
backend/src/backend/gen_insn_selection.cpp | 2 +-
backend/src/backend/program.h | 2 ++
backend/src/llvm/llvm_gen_backend.cpp | 13 +++++++++----
src/cl_command_queue.c | 5 ++++-
src/intel/intel_gpgpu.c | 28 +++++++++++++++++++---------
5 files changed, 35 insertions(+), 15 deletions(-)
diff --git a/backend/src/backend/gen_insn_selection.cpp b/backend/src/backend/gen_insn_selection.cpp
index abbff77..d631579 100644
--- a/backend/src/backend/gen_insn_selection.cpp
+++ b/backend/src/backend/gen_insn_selection.cpp
@@ -3862,7 +3862,7 @@ namespace gbe
msgLen = srcNum;
}
// We switch to a fixup bti for linear filter on a image1d array sampling.
- uint32_t bti = insn.getImageIndex() + (insn.getSamplerOffset() == 2 ? 128 : 0);
+ uint32_t bti = insn.getImageIndex() + (insn.getSamplerOffset() == 2 ? BTI_MAX_IMAGE_NUM : 0);
if (bti > 253) {
std::cerr << "Too large bti " << bti;
return false;
diff --git a/backend/src/backend/program.h b/backend/src/backend/program.h
index 254df92..c63ae6a 100644
--- a/backend/src/backend/program.h
+++ b/backend/src/backend/program.h
@@ -66,6 +66,8 @@ enum gbe_get_arg_info_value {
#define BTI_CONSTANT 0
#define BTI_PRIVATE 1
#define BTI_RESERVED_NUM 2
+#define BTI_MAX_IMAGE_NUM 128
+#define BTI_MAX_ID (BTI_MAX_IMAGE_NUM + BTI_RESERVED_NUM - 1)
/*! Constant buffer values (ie values to setup in the constant buffer) */
enum gbe_curbe_type {
diff --git a/backend/src/llvm/llvm_gen_backend.cpp b/backend/src/llvm/llvm_gen_backend.cpp
index 1604ede..738f7d3 100644
--- a/backend/src/llvm/llvm_gen_backend.cpp
+++ b/backend/src/llvm/llvm_gen_backend.cpp
@@ -500,6 +500,11 @@ namespace gbe
ir::ImmediateIndex processConstantImmIndex(Constant *CPV, int32_t index = 0u);
const ir::Immediate &processConstantImm(Constant *CPV, int32_t index = 0u);
+ uint32_t incBtiBase() {
+ GBE_ASSERT(btiBase <= BTI_MAX_ID);
+ return btiBase++;
+ }
+
bool runOnFunction(Function &F) {
// Do not codegen any 'available_externally' functions at all, they have
// definitions outside the translation unit.
@@ -1363,7 +1368,7 @@ namespace gbe
globalPointer.insert(std::make_pair(I, btiBase));
ctx.appendSurface(btiBase, reg);
ctx.input(argName, ir::FunctionArgument::GLOBAL_POINTER, reg, llvmInfo, ptrSize, align, btiBase);
- btiBase++;
+ incBtiBase();
break;
case ir::MEM_LOCAL:
ctx.input(argName, ir::FunctionArgument::LOCAL_POINTER, reg, llvmInfo, ptrSize, align, 0xfe);
@@ -1374,7 +1379,7 @@ namespace gbe
break;
case ir::IMAGE:
ctx.input(argName, ir::FunctionArgument::IMAGE, reg, llvmInfo, ptrSize, align, 0x0);
- ctx.getFunction().getImageSet()->append(reg, &ctx, btiBase++);
+ ctx.getFunction().getImageSet()->append(reg, &ctx, incBtiBase());
break;
default: GBE_ASSERT(addrSpace != ir::MEM_PRIVATE);
}
@@ -1720,12 +1725,12 @@ namespace gbe
if(v.getName().equals(StringRef("__gen_ocl_printf_buf"))) {
ctx.appendSurface(btiBase, ir::ocl::printfbptr);
ctx.getFunction().getPrintfSet()->setBufBTI(btiBase);
- globalPointer.insert(std::make_pair(&v, btiBase++));
+ globalPointer.insert(std::make_pair(&v, incBtiBase()));
regTranslator.newScalarProxy(ir::ocl::printfbptr, const_cast<GlobalVariable*>(&v));
} else if(v.getName().equals(StringRef("__gen_ocl_printf_index_buf"))) {
ctx.appendSurface(btiBase, ir::ocl::printfiptr);
ctx.getFunction().getPrintfSet()->setIndexBufBTI(btiBase);
- globalPointer.insert(std::make_pair(&v, btiBase++));
+ globalPointer.insert(std::make_pair(&v, incBtiBase()));
regTranslator.newScalarProxy(ir::ocl::printfiptr, const_cast<GlobalVariable*>(&v));
} else if(v.getName().str().substr(0, 4) == ".str") {
/* When there are multi printf statements in multi kernel fucntions within the same
diff --git a/src/cl_command_queue.c b/src/cl_command_queue.c
index 52e91ae..4cbb4eb 100644
--- a/src/cl_command_queue.c
+++ b/src/cl_command_queue.c
@@ -17,6 +17,7 @@
* Author: Benjamin Segovia <benjamin.segovia at intel.com>
*/
+#include "program.h" // for BTI_MAX_IMAGE_NUM
#include "cl_command_queue.h"
#include "cl_context.h"
#include "cl_program.h"
@@ -142,8 +143,10 @@ cl_command_queue_bind_image(cl_command_queue queue, cl_kernel k)
image->intel_fmt, image->image_type,
image->w, image->h, image->depth,
image->row_pitch, (cl_gpgpu_tiling)image->tiling);
+ // TODO, this workaround is for GEN7/GEN75 only, we may need to do it in the driver layer
+ // on demand.
if (image->image_type == CL_MEM_OBJECT_IMAGE1D_ARRAY)
- cl_gpgpu_bind_image(gpgpu, k->images[i].idx + 128, image->base.bo, image->offset,
+ cl_gpgpu_bind_image(gpgpu, k->images[i].idx + BTI_MAX_IMAGE_NUM, image->base.bo, image->offset,
image->intel_fmt, image->image_type,
image->w, image->h, image->depth,
image->row_pitch, (cl_gpgpu_tiling)image->tiling);
diff --git a/src/intel/intel_gpgpu.c b/src/intel/intel_gpgpu.c
index 867ab4c..c4b9156 100644
--- a/src/intel/intel_gpgpu.c
+++ b/src/intel/intel_gpgpu.c
@@ -35,6 +35,7 @@
#include "intel/intel_structs.h"
#include "intel/intel_batchbuffer.h"
#include "intel/intel_driver.h"
+#include "program.h" // for BTI_RESERVED_NUM
#include "cl_alloc.h"
#include "cl_utils.h"
@@ -819,6 +820,22 @@ intel_get_surface_type(cl_mem_object_type type)
return 0;
}
+/* Get fixed surface type. If it is a 1D array image with a large index,
+ we need to fixup it to 2D type due to a Gen7/Gen75's sampler issue
+ on a integer type surface with clamp address mode and nearest filter mode.
+*/
+static uint32_t get_surface_type(intel_gpgpu_t *gpgpu, int index, cl_mem_object_type type)
+{
+ uint32_t surface_type;
+ if (((IS_IVYBRIDGE(gpgpu->drv->device_id) || IS_HASWELL(gpgpu->drv->device_id))) &&
+ index >= 128 + BTI_RESERVED_NUM &&
+ type == CL_MEM_OBJECT_IMAGE1D_ARRAY)
+ surface_type = I965_SURFACE_2D;
+ else
+ surface_type = intel_get_surface_type(type);
+ return surface_type;
+}
+
static void
intel_gpgpu_bind_image_gen7(intel_gpgpu_t *gpgpu,
uint32_t index,
@@ -836,12 +853,8 @@ intel_gpgpu_bind_image_gen7(intel_gpgpu_t *gpgpu,
gen7_surface_state_t *ss = (gen7_surface_state_t *) heap->surface[index];
memset(ss, 0, sizeof(*ss));
-
ss->ss0.vertical_line_stride = 0; // always choose VALIGN_2
- if (index > 128 + 2 && type == CL_MEM_OBJECT_IMAGE1D_ARRAY)
- ss->ss0.surface_type = I965_SURFACE_2D;
- else
- ss->ss0.surface_type = intel_get_surface_type(type);
+ ss->ss0.surface_type = get_surface_type(gpgpu, index, type);
if (intel_is_surface_array(type)) {
ss->ss0.surface_array = 1;
ss->ss0.surface_array_spacing = 1;
@@ -886,10 +899,7 @@ intel_gpgpu_bind_image_gen75(intel_gpgpu_t *gpgpu,
gen7_surface_state_t *ss = (gen7_surface_state_t *) heap->surface[index];
memset(ss, 0, sizeof(*ss));
ss->ss0.vertical_line_stride = 0; // always choose VALIGN_2
- if (index > 128 + 2 && type == CL_MEM_OBJECT_IMAGE1D_ARRAY)
- ss->ss0.surface_type = I965_SURFACE_2D;
- else
- ss->ss0.surface_type = intel_get_surface_type(type);
+ ss->ss0.surface_type = get_surface_type(gpgpu, index, type);
if (intel_is_surface_array(type)) {
ss->ss0.surface_array = 1;
ss->ss0.surface_array_spacing = 1;
--
1.8.3.2
More information about the Beignet
mailing list