|
Java example source code file (debugInfoRec.cpp)
The debugInfoRec.cpp Java example source code/* * Copyright (c) 1998, 2013, Oracle and/or its affiliates. All rights reserved. * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER. * * This code is free software; you can redistribute it and/or modify it * under the terms of the GNU General Public License version 2 only, as * published by the Free Software Foundation. * * This code is distributed in the hope that it will be useful, but WITHOUT * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or * FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License * version 2 for more details (a copy is included in the LICENSE file that * accompanied this code). * * You should have received a copy of the GNU General Public License version * 2 along with this work; if not, write to the Free Software Foundation, * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA. * * Please contact Oracle, 500 Oracle Parkway, Redwood Shores, CA 94065 USA * or visit www.oracle.com if you need additional information or have any * questions. * */ #include "precompiled.hpp" #include "code/debugInfoRec.hpp" #include "code/scopeDesc.hpp" #include "prims/jvmtiExport.hpp" // Private definition. // There is one DIR_Chunk for each scope and values array. // A chunk can potentially be used more than once. // We keep track of these chunks in order to detect // repetition and enable sharing. class DIR_Chunk { friend class DebugInformationRecorder; int _offset; // location in the stream of this scope int _length; // number of bytes in the stream int _hash; // hash of stream bytes (for quicker reuse) void* operator new(size_t ignore, DebugInformationRecorder* dir) throw() { assert(ignore == sizeof(DIR_Chunk), ""); if (dir->_next_chunk >= dir->_next_chunk_limit) { const int CHUNK = 100; dir->_next_chunk = NEW_RESOURCE_ARRAY(DIR_Chunk, CHUNK); dir->_next_chunk_limit = dir->_next_chunk + CHUNK; } return dir->_next_chunk++; } DIR_Chunk(int offset, int length, DebugInformationRecorder* dir) { _offset = offset; _length = length; unsigned int hash = 0; address p = dir->stream()->buffer() + _offset; for (int i = 0; i < length; i++) { if (i == 6) break; hash *= 127; hash += p[i]; } _hash = hash; } DIR_Chunk* find_match(GrowableArray<DIR_Chunk*>* arr, int start_index, DebugInformationRecorder* dir) { int end_index = arr->length(); int hash = this->_hash, length = this->_length; address buf = dir->stream()->buffer(); for (int i = end_index; --i >= start_index; ) { DIR_Chunk* that = arr->at(i); if (hash == that->_hash && length == that->_length && 0 == memcmp(buf + this->_offset, buf + that->_offset, length)) { return that; } } return NULL; } }; static inline bool compute_recording_non_safepoints() { if (JvmtiExport::should_post_compiled_method_load() && FLAG_IS_DEFAULT(DebugNonSafepoints)) { // The default value of this flag is taken to be true, // if JVMTI is looking at nmethod codes. // We anticipate that JVMTI may wish to participate in profiling. return true; } // If the flag is set manually, use it, whether true or false. // Otherwise, if JVMTI is not in the picture, use the default setting. // (This is true in debug, just for the exercise, false in product mode.) return DebugNonSafepoints; } DebugInformationRecorder::DebugInformationRecorder(OopRecorder* oop_recorder) : _recording_non_safepoints(compute_recording_non_safepoints()) { _pcs_size = 100; _pcs = NEW_RESOURCE_ARRAY(PcDesc, _pcs_size); _pcs_length = 0; _prev_safepoint_pc = PcDesc::lower_offset_limit; _stream = new DebugInfoWriteStream(this, 10 * K); // make sure that there is no stream_decode_offset that is zero _stream->write_byte((jbyte)0xFF); // make sure that we can distinguish the value "serialized_null" from offsets assert(_stream->position() > serialized_null, "sanity"); _oop_recorder = oop_recorder; _all_chunks = new GrowableArray<DIR_Chunk*>(300); _shared_chunks = new GrowableArray<DIR_Chunk*>(30); _next_chunk = _next_chunk_limit = NULL; add_new_pc_offset(PcDesc::lower_offset_limit); // sentinel record debug_only(_recording_state = rs_null); } void DebugInformationRecorder::add_oopmap(int pc_offset, OopMap* map) { // !!!!! Preserve old style handling of oopmaps for now _oopmaps->add_gc_map(pc_offset, map); } void DebugInformationRecorder::add_safepoint(int pc_offset, OopMap* map) { assert(!_oop_recorder->is_complete(), "not frozen yet"); // Store the new safepoint // Add the oop map add_oopmap(pc_offset, map); add_new_pc_offset(pc_offset); assert(_recording_state == rs_null, "nesting of recording calls"); debug_only(_recording_state = rs_safepoint); } void DebugInformationRecorder::add_non_safepoint(int pc_offset) { assert(!_oop_recorder->is_complete(), "not frozen yet"); assert(_recording_non_safepoints, "must be recording non-safepoints"); add_new_pc_offset(pc_offset); assert(_recording_state == rs_null, "nesting of recording calls"); debug_only(_recording_state = rs_non_safepoint); } void DebugInformationRecorder::add_new_pc_offset(int pc_offset) { assert(_pcs_length == 0 || last_pc()->pc_offset() < pc_offset, "must specify a new, larger pc offset"); // add the pcdesc if (_pcs_length == _pcs_size) { // Expand int new_pcs_size = _pcs_size * 2; PcDesc* new_pcs = NEW_RESOURCE_ARRAY(PcDesc, new_pcs_size); for (int index = 0; index < _pcs_length; index++) { new_pcs[index] = _pcs[index]; } _pcs_size = new_pcs_size; _pcs = new_pcs; } assert(_pcs_size > _pcs_length, "There must be room for after expanding"); _pcs[_pcs_length++] = PcDesc(pc_offset, DebugInformationRecorder::serialized_null, DebugInformationRecorder::serialized_null); } int DebugInformationRecorder::serialize_monitor_values(GrowableArray<MonitorValue*>* monitors) { if (monitors == NULL || monitors->is_empty()) return DebugInformationRecorder::serialized_null; assert(_recording_state == rs_safepoint, "must be recording a safepoint"); int result = stream()->position(); stream()->write_int(monitors->length()); for (int index = 0; index < monitors->length(); index++) { monitors->at(index)->write_on(stream()); } assert(result != serialized_null, "sanity"); // (See comment below on DebugInformationRecorder::describe_scope.) int shared_result = find_sharable_decode_offset(result); if (shared_result != serialized_null) { stream()->set_position(result); result = shared_result; } return result; } int DebugInformationRecorder::serialize_scope_values(GrowableArray<ScopeValue*>* values) { if (values == NULL || values->is_empty()) return DebugInformationRecorder::serialized_null; assert(_recording_state == rs_safepoint, "must be recording a safepoint"); int result = stream()->position(); assert(result != serialized_null, "sanity"); stream()->write_int(values->length()); for (int index = 0; index < values->length(); index++) { values->at(index)->write_on(stream()); } // (See comment below on DebugInformationRecorder::describe_scope.) int shared_result = find_sharable_decode_offset(result); if (shared_result != serialized_null) { stream()->set_position(result); result = shared_result; } return result; } #ifndef PRODUCT // These variables are put into one block to reduce relocations // and make it simpler to print from the debugger. static struct dir_stats_struct { int chunks_queried; int chunks_shared; int chunks_reshared; int chunks_elided; void print() { tty->print_cr("Debug Data Chunks: %d, shared %d+%d, non-SP's elided %d", chunks_queried, chunks_shared, chunks_reshared, chunks_elided); } } dir_stats; #endif //PRODUCT int DebugInformationRecorder::find_sharable_decode_offset(int stream_offset) { // Only pull this trick if non-safepoint recording // is enabled, for now. if (!recording_non_safepoints()) return serialized_null; NOT_PRODUCT(++dir_stats.chunks_queried); int stream_length = stream()->position() - stream_offset; assert(stream_offset != serialized_null, "should not be null"); assert(stream_length != 0, "should not be empty"); DIR_Chunk* ns = new(this) DIR_Chunk(stream_offset, stream_length, this); // Look in previously shared scopes first: DIR_Chunk* ms = ns->find_match(_shared_chunks, 0, this); if (ms != NULL) { NOT_PRODUCT(++dir_stats.chunks_reshared); assert(ns+1 == _next_chunk, ""); _next_chunk = ns; return ms->_offset; } // Look in recently encountered scopes next: const int MAX_RECENT = 50; int start_index = _all_chunks->length() - MAX_RECENT; if (start_index < 0) start_index = 0; ms = ns->find_match(_all_chunks, start_index, this); if (ms != NULL) { NOT_PRODUCT(++dir_stats.chunks_shared); // Searching in _all_chunks is limited to a window, // but searching in _shared_chunks is unlimited. _shared_chunks->append(ms); assert(ns+1 == _next_chunk, ""); _next_chunk = ns; return ms->_offset; } // No match. Add this guy to the list, in hopes of future shares. _all_chunks->append(ns); return serialized_null; } // must call add_safepoint before: it sets PcDesc and this routine uses // the last PcDesc set void DebugInformationRecorder::describe_scope(int pc_offset, ciMethod* method, int bci, bool reexecute, bool is_method_handle_invoke, bool return_oop, DebugToken* locals, DebugToken* expressions, DebugToken* monitors) { assert(_recording_state != rs_null, "nesting of recording calls"); PcDesc* last_pd = last_pc(); assert(last_pd->pc_offset() == pc_offset, "must be last pc"); int sender_stream_offset = last_pd->scope_decode_offset(); // update the stream offset of current pc desc int stream_offset = stream()->position(); last_pd->set_scope_decode_offset(stream_offset); // Record flags into pcDesc. last_pd->set_should_reexecute(reexecute); last_pd->set_is_method_handle_invoke(is_method_handle_invoke); last_pd->set_return_oop(return_oop); // serialize sender stream offest stream()->write_int(sender_stream_offset); // serialize scope Metadata* method_enc = (method == NULL)? NULL: method->constant_encoding(); stream()->write_int(oop_recorder()->find_index(method_enc)); stream()->write_bci(bci); assert(method == NULL || (method->is_native() && bci == 0) || (!method->is_native() && 0 <= bci && bci < method->code_size()) || (method->is_compiled_lambda_form() && bci == -99) || // this might happen in C1 bci == -1, "illegal bci"); // serialize the locals/expressions/monitors stream()->write_int((intptr_t) locals); stream()->write_int((intptr_t) expressions); stream()->write_int((intptr_t) monitors); // Here's a tricky bit. We just wrote some bytes. // Wouldn't it be nice to find that we had already // written those same bytes somewhere else? // If we get lucky this way, reset the stream // and reuse the old bytes. By the way, this // trick not only shares parent scopes, but also // compresses equivalent non-safepoint PcDescs. int shared_stream_offset = find_sharable_decode_offset(stream_offset); if (shared_stream_offset != serialized_null) { stream()->set_position(stream_offset); last_pd->set_scope_decode_offset(shared_stream_offset); } } void DebugInformationRecorder::dump_object_pool(GrowableArray<ScopeValue*>* objects) { guarantee( _pcs_length > 0, "safepoint must exist before describing scopes"); PcDesc* last_pd = &_pcs[_pcs_length-1]; if (objects != NULL) { for (int i = objects->length() - 1; i >= 0; i--) { ((ObjectValue*) objects->at(i))->set_visited(false); } } int offset = serialize_scope_values(objects); last_pd->set_obj_decode_offset(offset); } void DebugInformationRecorder::end_scopes(int pc_offset, bool is_safepoint) { assert(_recording_state == (is_safepoint? rs_safepoint: rs_non_safepoint), "nesting of recording calls"); debug_only(_recording_state = rs_null); // Try to compress away an equivalent non-safepoint predecessor. // (This only works because we have previously recognized redundant // scope trees and made them use a common scope_decode_offset.) if (_pcs_length >= 2 && recording_non_safepoints()) { PcDesc* last = last_pc(); PcDesc* prev = prev_pc(); // If prev is (a) not a safepoint and (b) has the same // stream pointer, then it can be coalesced into the last. // This is valid because non-safepoints are only sought // with pc_desc_near, which (when it misses prev) will // search forward until it finds last. // In addition, it does not matter if the last PcDesc // is for a safepoint or not. if (_prev_safepoint_pc < prev->pc_offset() && prev->is_same_info(last)) { assert(prev == last-1, "sane"); prev->set_pc_offset(pc_offset); _pcs_length -= 1; NOT_PRODUCT(++dir_stats.chunks_elided); } } // We have just recorded this safepoint. // Remember it in case the previous paragraph needs to know. if (is_safepoint) { _prev_safepoint_pc = pc_offset; } } #ifdef ASSERT bool DebugInformationRecorder::recorders_frozen() { return _oop_recorder->is_complete() || _oop_recorder->is_complete(); } void DebugInformationRecorder::mark_recorders_frozen() { _oop_recorder->freeze(); } #endif // PRODUCT DebugToken* DebugInformationRecorder::create_scope_values(GrowableArray<ScopeValue*>* values) { assert(!recorders_frozen(), "not frozen yet"); return (DebugToken*) (intptr_t) serialize_scope_values(values); } DebugToken* DebugInformationRecorder::create_monitor_values(GrowableArray<MonitorValue*>* monitors) { assert(!recorders_frozen(), "not frozen yet"); return (DebugToken*) (intptr_t) serialize_monitor_values(monitors); } int DebugInformationRecorder::data_size() { debug_only(mark_recorders_frozen()); // mark it "frozen" for asserts return _stream->position(); } int DebugInformationRecorder::pcs_size() { debug_only(mark_recorders_frozen()); // mark it "frozen" for asserts if (last_pc()->pc_offset() != PcDesc::upper_offset_limit) add_new_pc_offset(PcDesc::upper_offset_limit); return _pcs_length * sizeof(PcDesc); } void DebugInformationRecorder::copy_to(nmethod* nm) { nm->copy_scopes_data(stream()->buffer(), stream()->position()); nm->copy_scopes_pcs(_pcs, _pcs_length); } void DebugInformationRecorder::verify(const nmethod* code) { Unimplemented(); } #ifndef PRODUCT void DebugInformationRecorder::print_statistics() { dir_stats.print(); } #endif //PRODUCT Other Java examples (source code examples)Here is a short list of links related to this Java debugInfoRec.cpp source code file: |
... this post is sponsored by my books ... | |
#1 New Release! |
FP Best Seller |
Copyright 1998-2024 Alvin Alexander, alvinalexander.com
All Rights Reserved.
A percentage of advertising revenue from
pages under the /java/jwarehouse
URI on this website is
paid back to open source projects.