#define CPU_JIT_AVX512_COMMON_CONVOLUTION_HPP
#include "c_types_map.hpp"
+#include "memory_tracking.hpp"
+#include "mkldnn_thread.hpp"
+#include "utils.hpp"
+
+#include "cpu_barrier.hpp"
#include "cpu_convolution_pd.hpp"
-#include "cpu_engine.hpp"
-#include "jit_avx512_common_conv_kernel.hpp"
-#include "jit_transpose_src_utils.hpp"
#include "cpu_reducer.hpp"
-#include "cpu_barrier.hpp"
+
+#include "jit_transpose_src_utils.hpp"
+#include "jit_avx512_common_conv_kernel.hpp"
namespace mkldnn {
namespace impl {
namespace cpu {
-template <bool with_relu, impl::data_type_t src_type,
+template <impl::data_type_t src_type,
impl::data_type_t wei_type = src_type,
impl::data_type_t dst_type = src_type>
-struct _jit_avx512_common_convolution_fwd_t : public cpu_primitive_t {
- struct pd_t : public _cpu_convolution_fwd_pd_t<with_relu> {
- pd_t(engine_t *engine, const typename pd_t::base_desc_t *adesc,
+struct jit_avx512_common_convolution_fwd_t : public cpu_primitive_t {
+ struct pd_t : public cpu_convolution_fwd_pd_t {
+ pd_t(engine_t *engine, const convolution_desc_t *adesc,
const primitive_attr_t *attr,
const typename pd_t::base_class *hint_fwd_pd)
- : _cpu_convolution_fwd_pd_t<with_relu>(engine, adesc, attr,
- hint_fwd_pd)
+ : cpu_convolution_fwd_pd_t(engine, adesc, attr, hint_fwd_pd)
, jcp_()
{
}
DECLARE_COMMON_PD_T(
JIT_IMPL_NAME_HELPER("jit:", avx512_common, ""),
- _jit_avx512_common_convolution_fwd_t);
+ jit_avx512_common_convolution_fwd_t);
virtual status_t init() override
{
using namespace prop_kind;
assert(this->engine()->kind() == engine_kind::cpu);
bool ok = true
- && utils::one_of(this->cdesc_().prop_kind, forward_training,
+ && utils::one_of(this->desc()->prop_kind, forward_training,
forward_inference)
- && this->cdesc_().alg_kind == alg_kind::convolution_direct
+ && utils::one_of(this->desc()->alg_kind,
+ alg_kind::convolution_auto,
+ alg_kind::convolution_direct)
&& !this->has_zero_dim_memory()
- && this->cdesc_().src_desc.data_type == src_type
- && this->cdesc_().weights_desc.data_type == wei_type
- && this->cdesc_().dst_desc.data_type == dst_type
+ && this->desc()->src_desc.data_type == src_type
+ && this->desc()->weights_desc.data_type == wei_type
+ && this->desc()->dst_desc.data_type == dst_type
&& IMPLICATION(this->with_bias(), dst_type
- == this->cdesc_().bias_desc.data_type)
- && !(with_relu && this->negative_slope()!= 0.
- && dst_type == data_type::s32
- && src_type == data_type::s16
- && wei_type == data_type::s16);
+ == this->desc()->bias_desc.data_type);
if (!ok)
return status::unimplemented;
- return jit_avx512_common_conv_fwd_kernel::init_conf(
- jcp_, this->cdesc_(), this->src_pd_, this->weights_pd_,
+ status_t status = jit_avx512_common_conv_fwd_kernel::init_conf(
+ jcp_, *this->desc(), this->src_pd_, this->weights_pd_,
this->dst_pd_,this->bias_pd_, *this->attr(),
- mkldnn_get_max_threads(), with_relu, this->negative_slope());
- }
+ mkldnn_get_max_threads());
+ if (status != status::success) return status;
- inline int ndims() { return this->cdesc_().src_desc.ndims; }
+ auto scratchpad = scratchpad_registry().registrar();
+ jit_avx512_common_conv_fwd_kernel::init_scratchpad(scratchpad,
+ jcp_);
+
+ if (status == status::success
+ && this->desc()->alg_kind == alg_kind::convolution_auto)
+ CHECK(this->set_alg_kind(alg_kind::convolution_direct));
+ return status;
+ }
jit_conv_conf_t jcp_;
};
- _jit_avx512_common_convolution_fwd_t(const pd_t *pd,
+ jit_avx512_common_convolution_fwd_t(const pd_t *apd,
const input_vector &inputs, const output_vector &outputs)
- : cpu_primitive_t(&conf_, inputs, outputs), conf_(*pd)
- , padded_bias_(nullptr)
+ : cpu_primitive_t(apd, inputs, outputs)
{
- kernel_ = new jit_avx512_common_conv_fwd_kernel(conf_.jcp_,
- *conf_.attr());
-
- if (conf_.want_padded_bias()) {
- const auto &j = conf_.jcp_;
- assert(j.ngroups == 1);
- padded_bias_ = (dst_data_t *)malloc(sizeof(dst_data_t) * j.oc, 64);
- for (int oc = j.oc_without_padding; oc < j.oc; ++oc)
- padded_bias_[oc] = 0;
- }
+ kernel_ = new jit_avx512_common_conv_fwd_kernel(pd()->jcp_,
+ *pd()->attr());
}
- ~_jit_avx512_common_convolution_fwd_t() {
- delete kernel_;
- free(padded_bias_);
- };
+ ~jit_avx512_common_convolution_fwd_t() { delete kernel_; }
typedef typename prec_traits<src_type>::type src_data_t;
typedef typename prec_traits<wei_type>::type wei_data_t;
typedef typename prec_traits<dst_type>::type dst_data_t;
- virtual void execute(event_t *e)
+ virtual void execute(event_t *e) const
{
- if (conf_.ndims() == 3)
+ if (pd()->ndims() == 3)
execute_forward_1d();
- else if (conf_.ndims() == 4)
+ else if (pd()->ndims() == 4)
execute_forward_2d();
- else if (conf_.ndims() == 5)
+ else if (pd()->ndims() == 5)
execute_forward_3d();
else
assert(false);
+
+ if (pd()->wants_zero_pad_dst())
+ output_memory_primitive(0)->zero_pad();
+
e->set_state(event_t::ready);
}
private:
- void execute_forward_1d();
- void execute_forward_2d();
- void execute_forward_3d();
- pd_t conf_;
+ void prepare_padded_bias(const dst_data_t *&bias) const;
+ void execute_forward_1d() const;
+ void execute_forward_2d() const;
+ void execute_forward_3d() const;
+ const pd_t *pd() const { return (const pd_t *)primitive_t::pd(); }
+
jit_avx512_common_conv_fwd_kernel *kernel_;
- dst_data_t *padded_bias_;
};
-template <impl::data_type_t src_type, impl::data_type_t wei_type = src_type,
- impl::data_type_t dst_type = src_type>
-using jit_avx512_common_convolution_fwd_t =
- _jit_avx512_common_convolution_fwd_t<false, src_type, wei_type, dst_type>;
-
-template <impl::data_type_t src_type, impl::data_type_t wei_type = src_type,
- impl::data_type_t dst_type = src_type>
-using jit_avx512_common_convolution_relu_t =
- _jit_avx512_common_convolution_fwd_t<true, src_type, wei_type, dst_type>;
-
template <impl::data_type_t diff_dst_type,
impl::data_type_t wei_type = diff_dst_type,
impl::data_type_t diff_src_type = diff_dst_type>
bool ok = true
&& this->set_default_params() == status::success
&& utils::one_of(this->desc()->prop_kind, backward_data) // XXX (this->!)
+ && utils::one_of(this->desc()->alg_kind,
+ alg_kind::convolution_auto,
+ alg_kind::convolution_direct)
&& !this->has_zero_dim_memory()
- && this->desc()->alg_kind == alg_kind::convolution_direct
&& this->desc()->diff_dst_desc.data_type == diff_dst_type
&& this->desc()->weights_desc.data_type == wei_type
&& this->desc()->diff_src_desc.data_type == diff_src_type;
if (!ok) return status::unimplemented;
- return jit_avx512_common_conv_bwd_data_kernel_f32::init_conf(
- jcp_,*this->desc(), *this->diff_src_pd_.desc(),
- *this->weights_pd_.desc(), *this->diff_dst_pd_.desc());
- }
+ status_t status =
+ jit_avx512_common_conv_bwd_data_kernel_f32::init_conf(jcp_,
+ *this->desc(), *this->diff_src_pd_.desc(),
+ *this->weights_pd_.desc(), *this->diff_dst_pd_.desc());
+ if (status != status::success) return status;
- inline int ndims() { return this->desc()->diff_src_desc.ndims; }
+ auto scratchpad = scratchpad_registry().registrar();
+ jit_avx512_common_conv_bwd_data_kernel_f32::init_scratchpad(
+ scratchpad, jcp_);
+
+ return status::success;
+ }
inline memory_format_t src_format()
{
CHECK(this->diff_dst_pd_.set_format(src_format()));
if (this->weights_pd_.desc()->format == any)
CHECK(this->weights_pd_.set_format(wei_format()));
+ if (this->desc()->alg_kind == alg_kind::convolution_auto)
+ CHECK(this->set_alg_kind(alg_kind::convolution_direct));
return status::success;
}
};
- jit_avx512_common_convolution_bwd_data_t(const pd_t *pd,
+ jit_avx512_common_convolution_bwd_data_t(const pd_t *apd,
const input_vector &inputs, const output_vector &outputs)
- : cpu_primitive_t(&conf_, inputs, outputs), conf_(*pd)
- {
- kernel_ = new jit_avx512_common_conv_bwd_data_kernel_f32(conf_.jcp_);
- }
+ : cpu_primitive_t(apd, inputs, outputs)
+ { kernel_ = new jit_avx512_common_conv_bwd_data_kernel_f32(pd()->jcp_); }
~jit_avx512_common_convolution_bwd_data_t() { delete kernel_; };
typedef typename prec_traits<diff_dst_type>::type diff_dst_data_t;
typedef typename prec_traits<wei_type>::type wei_data_t;
typedef typename prec_traits<diff_src_type>::type diff_src_data_t;
- virtual void execute(event_t *e) {
- switch (conf_.desc()->prop_kind) {
+ virtual void execute(event_t *e) const {
+ switch (pd()->desc()->prop_kind) {
case prop_kind::backward_data:
- if (conf_.ndims() == 3)
+ if (pd()->ndims() == 3)
execute_backward_data_1d();
- else if (conf_.ndims() == 4)
+ else if (pd()->ndims() == 4)
execute_backward_data_2d();
- else if (conf_.ndims() == 5)
+ else if (pd()->ndims() == 5)
execute_backward_data_3d();
else
assert(false);
}
private:
- void execute_backward_data_1d();
- void execute_backward_data_2d();
- void execute_backward_data_3d();
- pd_t conf_;
+ void execute_backward_data_1d() const;
+ void execute_backward_data_2d() const;
+ void execute_backward_data_3d() const;
+ const pd_t *pd() const { return (const pd_t *)primitive_t::pd(); }
+
jit_avx512_common_conv_bwd_data_kernel_f32 *kernel_;
};
assert(this->engine()->kind() == engine_kind::cpu);
bool ok = true
&& this->desc()->prop_kind == prop_kind::backward_weights
- && this->desc()->alg_kind == alg_kind::convolution_direct
+ && utils::one_of(this->desc()->alg_kind,
+ alg_kind::convolution_auto,
+ alg_kind::convolution_direct)
&& !this->has_zero_dim_memory()
&& this->desc()->src_desc.data_type == src_type
&& this->desc()->diff_dst_desc.data_type == diff_dst_type
== diff_weights_type;
if (!ok) return status::unimplemented;
- return jit_avx512_common_conv_bwd_weights_kernel_f32::init_conf(
- jcp_, *this->desc(), this->src_pd_, this->diff_weights_pd_,
- this->diff_bias_pd_, this->diff_dst_pd_);
- }
+ status_t status =
+ jit_avx512_common_conv_bwd_weights_kernel_f32::init_conf(jcp_,
+ *this->desc(), this->src_pd_, this->diff_weights_pd_,
+ this->diff_bias_pd_, this->diff_dst_pd_);
+ if (status != status::success) return status;
+
+ init_balancers();
- inline int ndims() { return this->desc()->src_desc.ndims; }
+ auto scratchpad = scratchpad_registry().registrar();
+ jit_avx512_common_conv_bwd_weights_kernel_f32::init_scratchpad(
+ scratchpad, jcp_);
+
+ auto reducer_bia_scratchpad = memory_tracking::registrar_t(
+ scratchpad, memory_tracking::names::prefix_reducer_bia);
+ reducer_bia_conf_.init_scratchpad(reducer_bia_scratchpad);
+
+ if (status == status::success &&
+ this->desc()->alg_kind == alg_kind::convolution_auto)
+ CHECK(this->set_alg_kind(alg_kind::convolution_direct));
+ return status;
+ }
inline memory_format_t src_format()
{
OIdhw16o16i);
}
-
jit_conv_conf_t jcp_;
+ typename cpu_reducer_t<diff_weights_type>::conf_t reducer_bia_conf_;
- protected:
- virtual status_t set_default_params() override {
- using namespace memory_format;
+ protected:
+ virtual status_t set_default_params() override {
+ using namespace memory_format;
- if (this->src_pd_.desc()->format == any)
- CHECK(this->src_pd_.set_format(src_format()));
- if (this->diff_weights_pd_.desc()->format == any)
- CHECK(this->diff_weights_pd_.set_format(wei_format()));
- if (this->diff_dst_pd_.desc()->format == any)
- CHECK(this->diff_dst_pd_.set_format(src_format()));
+ if (this->src_pd_.desc()->format == any)
+ CHECK(this->src_pd_.set_format(src_format()));
+ if (this->diff_weights_pd_.desc()->format == any)
+ CHECK(this->diff_weights_pd_.set_format(wei_format()));
+ if (this->diff_dst_pd_.desc()->format == any)
+ CHECK(this->diff_dst_pd_.set_format(src_format()));
- return status::success;
- }
+ return status::success;
+ }
+ private:
+ void init_balancers() {
+ const size_t max_buffer_size = jcp_.nthr * 3 * 5 * 5 * 16 * 16;
+ if (with_bias()) {
+ reducer_bia_conf_.init(reduce_balancer_t(jcp_.nthr,
+ jcp_.oc_block, jcp_.ngroups * jcp_.nb_oc, jcp_.mb,
+ max_buffer_size));
+ }
+ }
};
- jit_avx512_common_convolution_bwd_weights_t(const pd_t *pd,
+ jit_avx512_common_convolution_bwd_weights_t(const pd_t *apd,
const input_vector &inputs, const output_vector &outputs);
~jit_avx512_common_convolution_bwd_weights_t() {
-
delete kernel_;
if (trans_kernel_)
delete trans_kernel_;
if (acc_ker_)
delete acc_ker_;
delete reducer_bias_;
- free(padded_bias_);
-
- free(tr_src_);
- free(ws_reduction_);
-
- free(tr_src_bctx_);
- free(tr_diff_dst_bctx_);
-
- free(tr_diff_dst_);
}
typedef typename prec_traits<src_type>::type src_data_t;
typedef typename prec_traits<diff_dst_type>::type diff_dst_data_t;
typedef typename prec_traits<diff_weights_type>::type diff_weights_data_t;
- virtual void execute(event_t *e) {
+ virtual void execute(event_t *e) const {
execute_backward_weights();
e->set_state(event_t::ready);
}
private:
- void execute_backward_weights();
- void balance();
-
+ void execute_backward_weights() const;
+ void prepare_scratchpad_data() const;
struct thread_info_t;
- void compute_diff_weights(const thread_info_t *);
- void compute_diff_weights_3d(const thread_info_t *);
- void reduce_diff_weights(const thread_info_t *);
- void reduce_diff_weights_3d(const thread_info_t *);
- void compute_diff_bias(const thread_info_t *);
- void compute_diff_bias_3d(const thread_info_t *);
+ void compute_diff_weights(const thread_info_t *) const;
+ void compute_diff_weights_3d(const thread_info_t *) const;
+ void reduce_diff_weights(const thread_info_t *) const;
+ void reduce_diff_weights_3d(const thread_info_t *) const;
+ void compute_diff_bias(const thread_info_t *) const;
+ void compute_diff_bias_3d(const thread_info_t *) const;
- pd_t conf_;
+ const pd_t *pd() const { return (const pd_t *)primitive_t::pd(); }
+
+ int nthr_, nthr_mb_, nthr_g_, nthr_oc_b_, nthr_ic_b_;
jit_avx512_common_conv_bwd_weights_kernel_f32 *kernel_;
jit_trans_src_t *trans_kernel_;
jit_trans_dst_t *trans_dst_kernel_;
cpu_accumulator_1d_t<diff_weights_type> *acc_ker_;
cpu_reducer_t<diff_weights_type> *reducer_bias_;
- diff_weights_data_t *padded_bias_;
-
- src_data_t *tr_src_;
- diff_dst_data_t *tr_diff_dst_;
- diff_weights_data_t *ws_reduction_;
-
- int nthr_, nthr_mb_, nthr_g_, nthr_oc_b_, nthr_ic_b_;
- simple_barrier::ctx_t *tr_src_bctx_, *tr_diff_dst_bctx_, reduction_bctx_;
};
}
projects / platform / upstream / dldt.git / blobdiff