clover: try userptr for CL_MEM_USE_HOST_PTR

According to spec, CL_MEM_USE_HOST_PTR should directly use host memory,
if possible. This is just what userptr is for, so use it.

In case the memory cannot be mapped, a fallback similar to
CL_MEM_COPY_HOST_PTR is used.

v2: constify, drop unneeded cast

Reviewed-by: Francisco Jerez <currojerez@riseup.net>
This commit is contained in:
Grigori Goronzy
2015-05-19 09:28:30 +02:00
parent 5c495e8638
commit f972b223c4
2 changed files with 15 additions and 4 deletions
@@ -30,7 +30,7 @@ memory_obj::memory_obj(clover::context &ctx, cl_mem_flags flags,
size_t size, void *host_ptr) :
context(ctx), _flags(flags),
_size(size), _host_ptr(host_ptr) {
if (flags & (CL_MEM_COPY_HOST_PTR | CL_MEM_USE_HOST_PTR))
if (flags & CL_MEM_COPY_HOST_PTR)
data.append((char *)host_ptr, size);
}
@@ -118,6 +118,8 @@ root_resource::root_resource(clover::device &dev, memory_obj &obj,
command_queue &q, const std::string &data) :
resource(dev, obj) {
pipe_resource info {};
const bool user_ptr_support = dev.pipe->get_param(dev.pipe,
PIPE_CAP_RESOURCE_FROM_USER_MEMORY);
if (image *img = dynamic_cast<image *>(&obj)) {
info.format = translate_format(img->format());
@@ -137,7 +139,15 @@ root_resource::root_resource(clover::device &dev, memory_obj &obj,
PIPE_BIND_TRANSFER_READ |
PIPE_BIND_TRANSFER_WRITE);
if (obj.flags() & CL_MEM_ALLOC_HOST_PTR) {
if (obj.flags() & CL_MEM_USE_HOST_PTR && user_ptr_support) {
// Page alignment is normally required for this, just try, hope for the
// best and fall back if it fails.
pipe = dev.pipe->resource_from_user_memory(dev.pipe, &info, obj.host_ptr());
if (pipe)
return;
}
if (obj.flags() & (CL_MEM_ALLOC_HOST_PTR | CL_MEM_USE_HOST_PTR)) {
info.usage = PIPE_USAGE_STAGING;
}
@@ -145,12 +155,13 @@ root_resource::root_resource(clover::device &dev, memory_obj &obj,
if (!pipe)
throw error(CL_OUT_OF_RESOURCES);
if (!data.empty()) {
if (obj.flags() & (CL_MEM_USE_HOST_PTR | CL_MEM_COPY_HOST_PTR)) {
const void *data_ptr = !data.empty() ? data.data() : obj.host_ptr();
box rect { {{ 0, 0, 0 }}, {{ info.width0, info.height0, info.depth0 }} };
unsigned cpp = util_format_get_blocksize(info.format);
q.pipe->transfer_inline_write(q.pipe, pipe, 0, PIPE_TRANSFER_WRITE,
rect, data.data(), cpp * info.width0,
rect, data_ptr, cpp * info.width0,
cpp * info.width0 * info.height0);
}
}