As reported by tsan, SelectThreads could write to
is_context_switch_pending holding a mutex while SwitchToCurrent reads it
without holding any.
It is assumed that the author didn't want an atomic here, so the code is
reordered so that whenever is_context_switch_pending is read inside
SwitchToContext, the mutex is locked.
As reported by tsan, host_thread_ids could be read while
any of the RegisterHostThread variants were called.
To fix this, lock the register mutex when yuzu is running in multicore
mode and GetCurrentHostThreadID is called.
Migrates a remaining common file over to the Common namespace, making it
consistent with the rest of common files.
This also allows for high-traffic FS related code to alias the
filesystem function namespace as
namespace FS = Common::FS;
for more concise typing.
Allows the compiler to warn about cases where the constructor is used
but then immediately discarded, which is a potential cause of
locking/unlocking bugs.
This makes it more inline with its currently unavailable standardized
analogue std::derived_from.
While we're at it, we can also make the template match the requirements
of the standardized variant as well.
Previously the constructor for all of these would run at program
startup, consuming time before the application can enter main().
This is also particularly dangerous, given the logging system wouldn't
have been initialized properly yet, yet the program would use the logs
to signify an error.
To rectify this, we can replace the literals with constexpr functions
that perform the conversion at compile-time, completely eliminating the
runtime cost of initializing these arrays.
- In `SetCurrentThreadName`, when on Linux, truncate to 15 bytes, as (at
least on glibc) `pthread_set_name_np` will otherwise return `ERANGE` and
do nothing.
- Also, add logging in case `pthread_set_name_np` returns an error
anyway. This is Linux-specific, as the Apple and BSD versions of
`pthread_set_name_np return `void`.
- Change the name for CPU threads in multi-core mode from
"yuzu:CoreCPUThread_N" (19 bytes) to "yuzu:CPUCore_N" (14 bytes) so it
fits into the Linux limit. Some other thread names are also cut off,
but I didn't bother addressing them as you can guess them from the
truncated versions. For a CPU thread, truncation means you can't see
which core it is!
We can add a helper function to make creation of these files nicer.
While we're at it, we can eliminate an unnecessary std::array copy in
the constructor. This makes the overhead on some of these functions way
less intensive, given some arrays were quite large.
e.g. The timezone location names are 9633 bytes in size.
In a few places, the data to be set as the IV is already within an array.
We shouldn't require this data to be heap-allocated if it doesn't need
to be. This allows certain callers to reduce heap churn.
The general pattern is to mark mutexes as mutable when it comes to
matters of constness, given the mutex acts as a transient member of a
data structure.
* ipc: Allow all trivially copyable objects to be passed directly into WriteBuffer
With the support of C++20, we can use concepts to deduce if a type is an STL container or not.
* More agressive concept for stl containers
* Add -fconcepts
* Move to common namespace
* Add Common::IsBaseOf
This implements: Socket, Poll, Accept, Bind, Connect, GetPeerName,
GetSockName, Listen, Fcntl, SetSockOpt, Shutdown, Recv, RecvFrom,
Send, SendTo, Write, and Close
The implementation was done referencing: SwIPC, switchbrew, testing
with libnx and inspecting its code, general information about bsd
sockets online, and analysing official software.
Not everything from these service calls is implemented, but everything
that is not implemented will be logged in some way.
This abstraction allows executing blocking functions (like recvfrom on a
socket configured for blocking) without blocking the service thread.
It is intended to be used with SleepClientThread.
Makes the interface future-proofed for supporting other platforms in the event we ever support platforms with differing pointer sizes. This way, we have a type in place that is always guaranteed to be able to represent a pointer exactly.
Not using the return value of these functions are undeniably the source
of a bug. This way we allow compilers to loudly make any future misuses
evident.
src/core/network/network.cpp:112:28: error: use of undeclared identifier 'SHUT_RD'
constexpr int SD_RECEIVE = SHUT_RD;
^
src/core/network/network.cpp:113:25: error: use of undeclared identifier 'SHUT_WR'
constexpr int SD_SEND = SHUT_WR;
^
src/core/network/network.cpp:114:25: error: use of undeclared identifier 'SHUT_RDWR'
constexpr int SD_BOTH = SHUT_RDWR;
^
src/core/network/network.cpp:120:37: error: unknown type name 'in_addr'; did you mean 'in_addr_t'?
constexpr IPv4Address TranslateIPv4(in_addr addr) {
^~~~~~~
in_addr_t
/usr/include/netdb.h:66:20: note: 'in_addr_t' declared here
typedef __uint32_t in_addr_t;
^
src/core/network/network.cpp:121:27: error: member reference base type 'in_addr_t' (aka 'unsigned int') is not a structure or union
const u32 bytes = addr.s_addr;
~~~~^~~~~~~
src/core/network/network.cpp:121:15: error: variables defined in a constexpr function must be initialized
const u32 bytes = addr.s_addr;
^
src/core/network/network.cpp:126:10: error: incomplete result type 'sockaddr' in function definition
sockaddr TranslateFromSockAddrIn(SockAddrIn input) {
^
/usr/include/netdb.h:142:9: note: forward declaration of 'sockaddr'
struct sockaddr *ai_addr; /* binary address */
^
src/core/network/network.cpp:127:5: error: unknown type name 'sockaddr_in'; did you mean 'sockaddr'?
sockaddr_in result;
^~~~~~~~~~~
sockaddr
/usr/include/netdb.h:142:9: note: 'sockaddr' declared here
struct sockaddr *ai_addr; /* binary address */
^
src/core/network/network.cpp:127:17: error: variable has incomplete type 'sockaddr'
sockaddr_in result;
^
/usr/include/netdb.h:142:9: note: forward declaration of 'sockaddr'
struct sockaddr *ai_addr; /* binary address */
^
src/core/network/network.cpp:131:29: error: use of undeclared identifier 'AF_INET'
result.sin_family = AF_INET;
^
src/core/network/network.cpp:135:29: error: use of undeclared identifier 'AF_INET'
result.sin_family = AF_INET;
^
src/core/network/network.cpp:139:23: error: use of undeclared identifier 'htons'
result.sin_port = htons(input.portno);
^
src/core/network/network.cpp:143:14: error: variable has incomplete type 'sockaddr'
sockaddr addr;
^
/usr/include/netdb.h:142:9: note: forward declaration of 'sockaddr'
struct sockaddr *ai_addr; /* binary address */
^
src/core/network/network.cpp:156:1: error: unknown type name 'linger'
linger MakeLinger(bool enable, u32 linger_value) {
^
src/core/network/network.cpp:157:5: error: unknown type name 'linger'
linger value;
^
src/core/network/network.cpp:185:16: error: use of undeclared identifier 'AF_INET'
return AF_INET;
^
src/core/network/network.cpp:195:16: error: use of undeclared identifier 'SOCK_STREAM'
return SOCK_STREAM;
^
src/core/network/network.cpp:197:16: error: use of undeclared identifier 'SOCK_DGRAM'
return SOCK_DGRAM;
^
src/core/network/network.cpp:207:16: error: use of undeclared identifier 'IPPROTO_TCP'
return IPPROTO_TCP;
^
fatal error: too many errors emitted, stopping now [-ferror-limit=]
This commit adds a network abstraction designed to implement bsd:s but
at the same time work as a generic abstraction to implement any
networking code we have to use from core.
This is implemented on top of BSD sockets on Unix systems and winsock on
Windows. The code is designed around winsocks having compatibility
definitions to support both BSD and Windows sockets.
Previously, the method wasn't modifying any class state and therefore not having any effects when called.
Since this has been the case for a very long time now, I'm not sure if we couldn't just remove this method altogether.
