Pipe created by boost::process::child execProg can not be closed #6
Comments
|
This is the issue that I knew of: boostorg/process#62. The initial fix for it was here: klemens-morgenstern/boost-process@318439a and I believe made it into Boost 1.71. If I remember correctly, there was still a small corner case with this fix that should be closed with this change: klemens-morgenstern/boost-process@574d9e0, but I don't know if that one made it into a release. There is a similar issue opened here: klemens-morgenstern/boost-process#207, but it sounds like a pipe_guard was supposed to fix it. Not sure if it's the same pipe_guard as the above fix. |
|
I did a bit of digging and it looks like the pipe_guard fix and some follow-up fixes were merged by April 2019 and are included in Boost 1.71. |
|
In that case, this is no longer an issue with openbmc after the update to Poky Dunfell, which is using boost-1.72. |
dkodihal
pushed a commit
to NVIDIA/phosphor-user-manager
that referenced
this issue
May 7, 2024
```
Changes Added : Added event loop support in test code
problem : After adding support for sending events from phosphor-user-manager it is obsderved that it is calling sendEvent which internally calls async_send_handler and allocates memeory for context
since the event loop is not present in test code, callback is never called and
the CI was throwing memory leak error
Direct leak of 1 byte(s) in 1 object(s) allocated from:
#0 0x7ffa787b91e7 in operator new(unsigned long) ../../../../src/libsanitizer/asan/asan_new_delete.cpp:99
openbmc#1 0x7ffa77ad6383 in operator() /usr/local/include/sdbusplus/asio/detail/async_send_handler.hpp:40
openbmc#2 0x7ffa77ad6383 in async_send<sdbusplus::asio::connection::async_method_call_timed<phosphor::logging::sendEvent(phosphor::logging::MESSAGE_TYPE, sdbusplus::xyz::openbmc_project::Logging::server::Entry::Level, const std::vector<std::__cxx11::basic_string<char> >&, const string&)::<lambda(boost::system::error_code)>, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, std::map<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, std::less<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > >, std::allocator<std::pair<const std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > > > > >
(phosphor::logging::sendEvent(phosphor::logging::MESSAGE_TYPE, sdbusplus::xyz::openbmc_project::Logging::server::Entry::Level, const std::vector<std::__cxx11::basic_string<char> >&, const string&)::<lambda(boost::system::error_code)>&&, const string&, const string&, const string&, const string&, uint64_t, const std::__cxx11::basic_string<char>&, const std::__cxx11::basic_string<char>&, const std::map<std::__cxx11::basic_string<char>, std::__cxx11::basic_string<char> >&)::<lambda(boost::system::error_code, sdbusplus::message_t&)> > /usr/local/include/sdbusplus/asio/connection.hpp:98
openbmc#3 0x7ffa77ad6383 in async_method_call_timed<phosphor::logging::sendEvent(phosphor::logging::MESSAGE_TYPE, sdbusplus::xyz::openbmc_project::Logging::server::Entry::Level, const std::vector<std::__cxx11::basic_string<char> >&, const string&)::<lambda(boost::system::error_code)>, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, std::map<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, std::less<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > >, std::allocator<std::pair<const std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > > > > > /usr/local/include/sdbusplus/asio/connection.hpp:192
openbmc#4 0x7ffa77ad6383 in async_method_call<phosphor::logging::sendEvent(phosphor::logging::MESSAGE_TYPE, sdbusplus::xyz::openbmc_project::Logging::server::Entry::Level, const std::vector<std::__cxx11::basic_string<char> >&, const string&)::<lambda(boost::system::error_code)>, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, std::map<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, std::less<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > >, std::allocator<std::pair<const std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > > > > > /usr/local/include/sdbusplus/asio/connection.hpp:221
openbmc#5 0x7ffa77ad6383 in phosphor::logging::sendEvent(phosphor::logging::MESSAGE_TYPE, sdbusplus::xyz::openbmc_project::Logging::server::Entry::Level, std::vector<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >, std::allocator<std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > > > const&, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) ../lib/redfish_event_log.cpp:112
openbmc#6 0x55dfba9084a3 in phosphor::certs::Manager::replaceCertificate(phosphor::certs::Certificate*, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) ../certs_manager.cpp:493
openbmc#7 0x55dfba8ab09d in phosphor::certs::Certificate::replace(std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >) ../certificate.cpp:315
openbmc#8 0x55dfba7981e0 in TestBody ../test/certs_manager_test.cpp:677
openbmc#9 0x7ffa786e3f2e in void testing::internal::HandleSehExceptionsInMethodIfSupported<testing::Test, void>(testing::Test*, void (testing::Test::*)(), char const*) /googletest-662fe38e44900c007eccb65a5d2ea19df7bd520e/googletest/src/gtest.cc:2607
openbmc#10 0x7ffa786e3f2e in void testing::internal::HandleExceptionsInMethodIfSupported<testing::Test, void>(testing::Test*, void (testing::Test::*)(), char const*) /googletest-662fe38e44900c007eccb65a5d2ea19df7bd520e/googletest/src/gtest.cc:2643
```
Solution : The memory leak error was thrown because the
memory allocated by "async_send_handler" in sdbusplus was not getting de-allocated
because the callback is never getting called called since there was no event loop
present in test code.
Added event loop support in test code
Fixes jira https://jirasw.nvidia.com/browse/DGXOPENBMC-8881
|
given the suggested fix above from 2020 can we close this? |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
In long time testing with phosphor-user-manager.service, we found many pipes created by this process and these pipes can not be closed. We find pipe is created by function "executeCmd" in which boost::process::child execProg is used. There is some problem with this boost::process::child execProg library function. Is this problem caused by low boost::process library version?
We should fix this problem, otherwise there will be so many file descriptors created by phosphor-user-manager.service in long time running. @vmauery @williamspatrick
The text was updated successfully, but these errors were encountered: