Git-Mirrors/monero
1
0
Fork 0
mirror of https://github.com/monero-project/monero.git synced 2025-08-09 23:32:24 -04:00
Code Issues Projects Releases Packages Wiki Activity

2014 network limit 1.2 +utils +toc -doc -drmonero

Browse source 
new update of the pr with network limits

more debug options:
discarding downloaded blocks all or after given height.
trying to trigger the locking errors.

debug levels polished/tuned to sane values.
debug/logging improved.

warning: this pr should be correct code, but it could make
an existing (in master version) locking error appear more often.

it's a race on the list (map) of peers, e.g. between closing/deleting
them versus working on them in net-limit sleep in sending chunk.

the bug is not in this code/this pr, but in the master version.

the locking problem of master will be fixed in other pr.

problem is ub, and in practice is seems to usually cause program abort
(tested on debian stable with updated gcc). see --help for option
to add sleep to trigger the error faster.
This commit is contained in:
rfree2monero 2015-02-20 22:28:03 +01:00
parent 0f06dca831
commit ae2a50659f
23 changed files with 251 additions and 178 deletions
Download patch file Download diff file Expand all files Collapse all files

21
contrib/epee/include/net/abstract_tcp_server2.inl
View file

@ -198,8 +198,9 @@ PRAGMA_WARNING_DISABLE_VS(4355)
bool connection<t_protocol_handler>::add_ref()
{
TRY_ENTRY();
//_info("[sock " << socket_.native_handle() << "] add_ref");
//_dbg3("[sock " << socket_.native_handle() << "] add_ref, m_peer_number=" << mI->m_peer_number);
CRITICAL_REGION_LOCAL(m_self_refs_lock);
//_dbg3("[sock " << socket_.native_handle() << "] add_ref 2, m_peer_number=" << mI->m_peer_number);
// Use safe_shared_from_this, because of this is public method and it can be called on the object being deleted
auto self = safe_shared_from_this();
@ -364,8 +365,7 @@ PRAGMA_WARNING_DISABLE_VS(4355)
{ // LOCK: chunking
epee::critical_region_t<decltype(m_chunking_lock)> send_guard(m_chunking_lock); // *** critical ***
_mark_c("net/out/size", "do_send() will SPLIT into small chunks, from packet="<<cb<<" B for ptr="<<ptr);
_mark("do_send() will SPLIT into small chunks, from packet="<<cb<<" B for ptr="<<ptr);
_dbg3_c("net/out/size", "do_send() will SPLIT into small chunks, from packet="<<cb<<" B for ptr="<<ptr);
t_safe all = cb; // all bytes to send
t_safe pos = 0; // current sending position
// 01234567890
@ -393,14 +393,13 @@ PRAGMA_WARNING_DISABLE_VS(4355)
ASRT(chunk_start >= ptr); // not wrapped around address?
//std::memcpy( (void*)buf, chunk_start, len);
_dbg1_c("net/out/size", "part of " << lenall << ": pos="<<pos << " len="<<len);
_dbg3_c("net/out/size", "part of " << lenall << ": pos="<<pos << " len="<<len);
bool ok = do_send_chunk(chunk_start, len); // <====== ***
all_ok = all_ok && ok;
if (!all_ok) {
_mark_c("net/out/size", "do_send() DONE ***FAILED*** from packet="<<cb<<" B for ptr="<<ptr);
_mark ( "do_send() DONE ***FAILED*** from packet="<<cb<<" B for ptr="<<ptr);
_dbg1_c("net/out/size", "do_send() DONE ***FAILED*** from packet="<<cb<<" B for ptr="<<ptr);
_dbg1("do_send() SEND was aborted in middle of big package - this is mostly harmless "
<< " (e.g. peer closed connection) but if it causes trouble tell us at #monero-dev. " << cb);
return false; // partial failure in sending
@ -410,8 +409,8 @@ PRAGMA_WARNING_DISABLE_VS(4355)
// (in catch block, or uniq pointer) delete buf;
} // each chunk
_mark_c("net/out/size", "do_send() DONE SPLIT from packet="<<cb<<" B for ptr="<<ptr);
_mark ( "do_send() DONE SPLIT from packet="<<cb<<" B for ptr="<<ptr);
_dbg3_c("net/out/size", "do_send() DONE SPLIT from packet="<<cb<<" B for ptr="<<ptr);
_dbg3 ( "do_send() DONE SPLIT from packet="<<cb<<" B for ptr="<<ptr);
_info_c("net/sleepRPC", "do_send() m_connection_type = " << m_connection_type);
@ -469,7 +468,7 @@ PRAGMA_WARNING_DISABLE_VS(4355)
if (retry > retry_limit) {
send_guard.unlock();
_erro("send que size is more than ABSTRACT_SERVER_SEND_QUE_MAX_COUNT(" << ABSTRACT_SERVER_SEND_QUE_MAX_COUNT << "), shutting down connection");
// _mark_c("net/sleep", "send que size is more than ABSTRACT_SERVER_SEND_QUE_MAX_COUNT(" << ABSTRACT_SERVER_SEND_QUE_MAX_COUNT << "), shutting down connection");
// _dbg1_c("net/sleep", "send que size is more than ABSTRACT_SERVER_SEND_QUE_MAX_COUNT(" << ABSTRACT_SERVER_SEND_QUE_MAX_COUNT << "), shutting down connection");
close();
return false;
}
@ -496,7 +495,7 @@ PRAGMA_WARNING_DISABLE_VS(4355)
}
auto size_now = m_send_que.front().size();
_mark_c("net/out/size", "do_send() NOW SENSD: packet="<<size_now<<" B");
_dbg1_c("net/out/size", "do_send() NOW SENSD: packet="<<size_now<<" B");
do_send_handler_write( ptr , size_now ); // (((H)))
ASRT( size_now == m_send_que.front().size() );
@ -580,7 +579,7 @@ PRAGMA_WARNING_DISABLE_VS(4355)
{
//have more data to send
auto size_now = m_send_que.front().size();
_mark_c("net/out/size", "handle_write() NOW SENDS: packet="<<size_now<<" B" <<", from queue size="<<m_send_que.size());
_dbg1_c("net/out/size", "handle_write() NOW SENDS: packet="<<size_now<<" B" <<", from queue size="<<m_send_que.size());
do_send_handler_write_from_queue(e, m_send_que.front().size() , m_send_que.size()); // (((H)))
ASRT( size_now == m_send_que.front().size() );
boost::asio::async_write(socket_, boost::asio::buffer(m_send_que.front().data(), size_now) ,