i965: Prefer using streaming reads from WC mmaps

For buffer objects, where we primarily expect to be writing to them and
so already have a WC mmap (for !llc access) reusing the existing mmap
and keeping the buffer out of the CPU cache seems preferable.

Cc: Kenneth Graunke <kenneth@whitecape.org>
Cc: Matt Turner <mattst88@gmail.com>
Reviewed-by: Matt Turner <mattst88@gmail.com>
This commit is contained in:
Chris Wilson 2017-07-22 10:28:14 +01:00
parent 27ba094a4a
commit b4f639d02a

View file

@ -32,12 +32,15 @@
#include "main/imports.h"
#include "main/mtypes.h"
#include "main/macros.h"
#include "main/streaming-load-memcpy.h"
#include "main/bufferobj.h"
#include "x86/common_x86_asm.h"
#include "brw_context.h"
#include "intel_blit.h"
#include "intel_buffer_objects.h"
#include "intel_batchbuffer.h"
#include "intel_tiled_memcpy.h"
static void
mark_buffer_gpu_usage(struct intel_buffer_object *intel_obj,
@ -337,14 +340,23 @@ brw_get_buffer_subdata(struct gl_context *ctx,
intel_batchbuffer_flush(brw);
}
void *map = brw_bo_map(brw, intel_obj->buffer, MAP_READ);
unsigned int map_flags = MAP_READ;
mem_copy_fn memcpy_fn = memcpy;
if (!intel_obj->buffer->cache_coherent && cpu_has_sse4_1) {
/* Rather than acquire a new WB mmaping of the buffer object and pull
* it into the CPU cache, keep using the WC mmap that we have for writes,
* and use the magic movntd instructions instead.
*/
map_flags |= MAP_COHERENT;
memcpy_fn = (mem_copy_fn) _mesa_streaming_load_memcpy;
}
void *map = brw_bo_map(brw, intel_obj->buffer, map_flags);
if (unlikely(!map)) {
_mesa_error_no_memory(__func__);
return;
}
memcpy(data, map + offset, size);
memcpy_fn(data, map + offset, size);
brw_bo_unmap(intel_obj->buffer);
mark_buffer_inactive(intel_obj);