gc: refactor the gc server thread out into a class without changing it
This removes a *whole load* of variables from scope and enforces thread
boundaries with the type system.
There is not much change of significance in here, so the things to watch
out for while reviewing it are primarily that the destructor ordering
may have changed inadvertently, I think.
Change-Id: I3cd87e6d5a08dfcf368637407251db22a8906316
This commit is contained in:
parent
22252825c4
commit
77ff799cc8
|
|
@ -22,8 +22,8 @@
|
|||
namespace nix {
|
||||
|
||||
|
||||
static std::string gcSocketPath = "/gc-socket/socket";
|
||||
static std::string gcRootsDir = "gcroots";
|
||||
constexpr static const std::string_view gcSocketPath = "/gc-socket/socket";
|
||||
constexpr static const std::string_view gcRootsDir = "gcroots";
|
||||
|
||||
|
||||
static void makeSymlink(const Path & link, const Path & target)
|
||||
|
|
@ -359,16 +359,34 @@ void LocalStore::findRuntimeRoots(Roots & roots, bool censor)
|
|||
}
|
||||
|
||||
|
||||
struct GCLimitReached { };
|
||||
struct GCLimitReached : std::exception { };
|
||||
|
||||
|
||||
void LocalStore::collectGarbage(const GCOptions & options, GCResults & results)
|
||||
{
|
||||
bool shouldDelete = options.action == GCOptions::gcDeleteDead || options.action == GCOptions::gcDeleteSpecific;
|
||||
bool gcKeepOutputs = settings.gcKeepOutputs;
|
||||
bool gcKeepDerivations = settings.gcKeepDerivations;
|
||||
/**
|
||||
* Delegate class to expose just the operations required to perform GC on a store.
|
||||
*/
|
||||
class GCStoreDelegate {
|
||||
LocalStore const & store;
|
||||
|
||||
StorePathSet roots, dead, alive;
|
||||
public:
|
||||
GCStoreDelegate(LocalStore const & store) : store(store) {}
|
||||
|
||||
std::optional<StorePath> maybeParseStorePath(std::string_view path) const {
|
||||
return store.maybeParseStorePath(path);
|
||||
}
|
||||
};
|
||||
|
||||
|
||||
/**
|
||||
* Class holding a server to receive new GC roots.
|
||||
*/
|
||||
class GCOperation {
|
||||
const GCStoreDelegate store;
|
||||
|
||||
std::thread serverThread;
|
||||
Pipe shutdownPipe;
|
||||
|
||||
AutoCloseFD fdServer;
|
||||
|
||||
struct Shared
|
||||
{
|
||||
|
|
@ -381,9 +399,165 @@ void LocalStore::collectGarbage(const GCOptions & options, GCResults & results)
|
|||
std::optional<std::string> pending;
|
||||
};
|
||||
|
||||
Sync<Shared> _shared;
|
||||
void runServerThread();
|
||||
|
||||
std::condition_variable wakeup;
|
||||
Sync<Shared> _shared;
|
||||
|
||||
public:
|
||||
GCOperation(LocalStore const & store, Path stateDir) : store(store)
|
||||
{
|
||||
/* Start the server for receiving new roots. */
|
||||
shutdownPipe.create();
|
||||
|
||||
auto socketPath = stateDir + gcSocketPath;
|
||||
createDirs(dirOf(socketPath));
|
||||
fdServer = createUnixDomainSocket(socketPath, 0666);
|
||||
|
||||
if (fcntl(fdServer.get(), F_SETFL, fcntl(fdServer.get(), F_GETFL) | O_NONBLOCK) == -1) {
|
||||
throw SysError("making socket '%1%' non-blocking", socketPath);
|
||||
}
|
||||
|
||||
serverThread = std::thread([this]() { runServerThread(); });
|
||||
}
|
||||
|
||||
void addTempRoot(std::string rootHashPart)
|
||||
{
|
||||
_shared.lock()->tempRoots.insert(rootHashPart);
|
||||
}
|
||||
|
||||
void releasePending()
|
||||
{
|
||||
auto shared(_shared.lock());
|
||||
shared->pending.reset();
|
||||
wakeup.notify_all();
|
||||
}
|
||||
|
||||
/**
|
||||
* Marks a path as pending deletion if it is not in tempRoots.
|
||||
*
|
||||
* Returns whether it was marked for deletion.
|
||||
*/
|
||||
bool markPendingIfPresent(std::string const & hashPart)
|
||||
{
|
||||
auto shared(_shared.lock());
|
||||
if (shared->tempRoots.count(hashPart)) {
|
||||
return false;
|
||||
}
|
||||
shared->pending = hashPart;
|
||||
return true;
|
||||
}
|
||||
|
||||
~GCOperation();
|
||||
};
|
||||
|
||||
void GCOperation::runServerThread()
|
||||
{
|
||||
Sync<std::map<int, std::thread>> connections;
|
||||
|
||||
Finally cleanup([&]() {
|
||||
debug("GC roots server shutting down");
|
||||
fdServer.close();
|
||||
while (true) {
|
||||
auto item = remove_begin(*connections.lock());
|
||||
if (!item) break;
|
||||
auto & [fd, thread] = *item;
|
||||
shutdown(fd, SHUT_RDWR);
|
||||
thread.join();
|
||||
}
|
||||
});
|
||||
|
||||
while (true) {
|
||||
std::vector<struct pollfd> fds;
|
||||
fds.push_back({.fd = shutdownPipe.readSide.get(), .events = POLLIN});
|
||||
fds.push_back({.fd = fdServer.get(), .events = POLLIN});
|
||||
auto count = poll(fds.data(), fds.size(), -1);
|
||||
assert(count != -1);
|
||||
|
||||
if (fds[0].revents)
|
||||
/* Parent is asking us to quit. */
|
||||
break;
|
||||
|
||||
if (fds[1].revents) {
|
||||
/* Accept a new connection. */
|
||||
assert(fds[1].revents & POLLIN);
|
||||
AutoCloseFD fdClient{accept(fdServer.get(), nullptr, nullptr)};
|
||||
if (!fdClient) continue;
|
||||
|
||||
debug("GC roots server accepted new client");
|
||||
|
||||
/* Process the connection in a separate thread. */
|
||||
auto fdClient_ = fdClient.get();
|
||||
std::thread clientThread([&, fdClient = std::move(fdClient)]() {
|
||||
Finally cleanup([&]() {
|
||||
auto conn(connections.lock());
|
||||
auto i = conn->find(fdClient.get());
|
||||
if (i != conn->end()) {
|
||||
i->second.detach();
|
||||
conn->erase(i);
|
||||
}
|
||||
});
|
||||
|
||||
/* On macOS, accepted sockets inherit the
|
||||
non-blocking flag from the server socket, so
|
||||
explicitly make it blocking. */
|
||||
if (fcntl(fdClient.get(), F_SETFL, fcntl(fdClient.get(), F_GETFL) & ~O_NONBLOCK) == -1)
|
||||
abort();
|
||||
|
||||
while (true) {
|
||||
try {
|
||||
auto path = readLine(fdClient.get());
|
||||
auto storePath = store.maybeParseStorePath(path);
|
||||
if (storePath) {
|
||||
debug("got new GC root '%s'", path);
|
||||
auto hashPart = std::string(storePath->hashPart());
|
||||
auto shared(_shared.lock());
|
||||
shared->tempRoots.insert(hashPart);
|
||||
/* If this path is currently being
|
||||
deleted, then we have to wait until
|
||||
deletion is finished to ensure that
|
||||
the client doesn't start
|
||||
re-creating it before we're
|
||||
done. FIXME: ideally we would use a
|
||||
FD for this so we don't block the
|
||||
poll loop. */
|
||||
while (shared->pending == hashPart) {
|
||||
debug("synchronising with deletion of path '%s'", path);
|
||||
shared.wait(wakeup);
|
||||
}
|
||||
} else
|
||||
printError("received garbage instead of a root from client");
|
||||
writeFull(fdClient.get(), "1", false);
|
||||
} catch (Error & e) {
|
||||
debug("reading GC root from client: %s", e.msg());
|
||||
break;
|
||||
}
|
||||
}
|
||||
});
|
||||
|
||||
connections.lock()->insert({fdClient_, std::move(clientThread)});
|
||||
}
|
||||
}
|
||||
}
|
||||
|
||||
GCOperation::~GCOperation()
|
||||
{
|
||||
writeFull(shutdownPipe.writeSide.get(), "x", false);
|
||||
{
|
||||
auto shared(_shared.lock());
|
||||
wakeup.notify_all();
|
||||
}
|
||||
if (serverThread.joinable()) serverThread.join();
|
||||
}
|
||||
|
||||
|
||||
void LocalStore::collectGarbage(const GCOptions & options, GCResults & results)
|
||||
{
|
||||
bool shouldDelete = options.action == GCOptions::gcDeleteDead || options.action == GCOptions::gcDeleteSpecific;
|
||||
bool gcKeepOutputs = settings.gcKeepOutputs;
|
||||
bool gcKeepDerivations = settings.gcKeepDerivations;
|
||||
|
||||
StorePathSet roots, dead, alive;
|
||||
|
||||
/* Using `--ignore-liveness' with `--delete' can have unintended
|
||||
consequences if `keep-outputs' or `keep-derivations' are true
|
||||
|
|
@ -395,7 +569,7 @@ void LocalStore::collectGarbage(const GCOptions & options, GCResults & results)
|
|||
}
|
||||
|
||||
if (shouldDelete)
|
||||
deletePath(reservedPath);
|
||||
deletePath(reservedSpacePath);
|
||||
|
||||
/* Acquire the global GC root. Note: we don't use fdGCLock
|
||||
here because then in auto-gc mode, another thread could
|
||||
|
|
@ -408,110 +582,7 @@ void LocalStore::collectGarbage(const GCOptions & options, GCResults & results)
|
|||
if (auto p = getEnv("_NIX_TEST_GC_SYNC_1"))
|
||||
readFile(*p);
|
||||
|
||||
/* Start the server for receiving new roots. */
|
||||
auto socketPath = stateDir.get() + gcSocketPath;
|
||||
createDirs(dirOf(socketPath));
|
||||
auto fdServer = createUnixDomainSocket(socketPath, 0666);
|
||||
|
||||
if (fcntl(fdServer.get(), F_SETFL, fcntl(fdServer.get(), F_GETFL) | O_NONBLOCK) == -1)
|
||||
throw SysError("making socket '%1%' non-blocking", socketPath);
|
||||
|
||||
Pipe shutdownPipe;
|
||||
shutdownPipe.create();
|
||||
|
||||
std::thread serverThread([&]() {
|
||||
Sync<std::map<int, std::thread>> connections;
|
||||
|
||||
Finally cleanup([&]() {
|
||||
debug("GC roots server shutting down");
|
||||
fdServer.close();
|
||||
while (true) {
|
||||
auto item = remove_begin(*connections.lock());
|
||||
if (!item) break;
|
||||
auto & [fd, thread] = *item;
|
||||
shutdown(fd, SHUT_RDWR);
|
||||
thread.join();
|
||||
}
|
||||
});
|
||||
|
||||
while (true) {
|
||||
std::vector<struct pollfd> fds;
|
||||
fds.push_back({.fd = shutdownPipe.readSide.get(), .events = POLLIN});
|
||||
fds.push_back({.fd = fdServer.get(), .events = POLLIN});
|
||||
auto count = poll(fds.data(), fds.size(), -1);
|
||||
assert(count != -1);
|
||||
|
||||
if (fds[0].revents)
|
||||
/* Parent is asking us to quit. */
|
||||
break;
|
||||
|
||||
if (fds[1].revents) {
|
||||
/* Accept a new connection. */
|
||||
assert(fds[1].revents & POLLIN);
|
||||
AutoCloseFD fdClient{accept(fdServer.get(), nullptr, nullptr)};
|
||||
if (!fdClient) continue;
|
||||
|
||||
debug("GC roots server accepted new client");
|
||||
|
||||
/* Process the connection in a separate thread. */
|
||||
auto fdClient_ = fdClient.get();
|
||||
std::thread clientThread([&, fdClient = std::move(fdClient)]() {
|
||||
Finally cleanup([&]() {
|
||||
auto conn(connections.lock());
|
||||
auto i = conn->find(fdClient.get());
|
||||
if (i != conn->end()) {
|
||||
i->second.detach();
|
||||
conn->erase(i);
|
||||
}
|
||||
});
|
||||
|
||||
/* On macOS, accepted sockets inherit the
|
||||
non-blocking flag from the server socket, so
|
||||
explicitly make it blocking. */
|
||||
if (fcntl(fdClient.get(), F_SETFL, fcntl(fdClient.get(), F_GETFL) & ~O_NONBLOCK) == -1)
|
||||
abort();
|
||||
|
||||
while (true) {
|
||||
try {
|
||||
auto path = readLine(fdClient.get());
|
||||
auto storePath = maybeParseStorePath(path);
|
||||
if (storePath) {
|
||||
debug("got new GC root '%s'", path);
|
||||
auto hashPart = std::string(storePath->hashPart());
|
||||
auto shared(_shared.lock());
|
||||
shared->tempRoots.insert(hashPart);
|
||||
/* If this path is currently being
|
||||
deleted, then we have to wait until
|
||||
deletion is finished to ensure that
|
||||
the client doesn't start
|
||||
re-creating it before we're
|
||||
done. FIXME: ideally we would use a
|
||||
FD for this so we don't block the
|
||||
poll loop. */
|
||||
while (shared->pending == hashPart) {
|
||||
debug("synchronising with deletion of path '%s'", path);
|
||||
shared.wait(wakeup);
|
||||
}
|
||||
} else
|
||||
printError("received garbage instead of a root from client");
|
||||
writeFull(fdClient.get(), "1", false);
|
||||
} catch (Error & e) {
|
||||
debug("reading GC root from client: %s", e.msg());
|
||||
break;
|
||||
}
|
||||
}
|
||||
});
|
||||
|
||||
connections.lock()->insert({fdClient_, std::move(clientThread)});
|
||||
}
|
||||
}
|
||||
});
|
||||
|
||||
Finally stopServer([&]() {
|
||||
writeFull(shutdownPipe.writeSide.get(), "x", false);
|
||||
wakeup.notify_all();
|
||||
if (serverThread.joinable()) serverThread.join();
|
||||
});
|
||||
GCOperation gcServer {*this, stateDir.get()};
|
||||
|
||||
/* Find the roots. Since we've grabbed the GC lock, the set of
|
||||
permanent roots cannot increase now. */
|
||||
|
|
@ -527,7 +598,7 @@ void LocalStore::collectGarbage(const GCOptions & options, GCResults & results)
|
|||
Roots tempRoots;
|
||||
findTempRoots(tempRoots, true);
|
||||
for (auto & root : tempRoots) {
|
||||
_shared.lock()->tempRoots.insert(std::string(root.first.hashPart()));
|
||||
gcServer.addTempRoot(std::string(root.first.hashPart()));
|
||||
roots.insert(root.first);
|
||||
}
|
||||
|
||||
|
|
@ -580,9 +651,7 @@ void LocalStore::collectGarbage(const GCOptions & options, GCResults & results)
|
|||
/* Wake up any GC client waiting for deletion of the paths in
|
||||
'visited' to finish. */
|
||||
Finally releasePending([&]() {
|
||||
auto shared(_shared.lock());
|
||||
shared->pending.reset();
|
||||
wakeup.notify_all();
|
||||
gcServer.releasePending();
|
||||
});
|
||||
|
||||
auto enqueue = [&](const StorePath & path) {
|
||||
|
|
@ -629,14 +698,9 @@ void LocalStore::collectGarbage(const GCOptions & options, GCResults & results)
|
|||
&& !options.pathsToDelete.count(*path))
|
||||
return;
|
||||
|
||||
{
|
||||
auto hashPart = std::string(path->hashPart());
|
||||
auto shared(_shared.lock());
|
||||
if (shared->tempRoots.count(hashPart)) {
|
||||
debug("cannot delete '%s' because it's a temporary root", printStorePath(*path));
|
||||
return markAlive();
|
||||
}
|
||||
shared->pending = hashPart;
|
||||
if (!gcServer.markPendingIfPresent(std::string(path->hashPart()))) {
|
||||
debug("cannot delete '%s' because it's a temporary root", printStorePath(*path));
|
||||
return markAlive();
|
||||
}
|
||||
|
||||
if (isValidPath(*path)) {
|
||||
|
|
|
|||
|
|
@ -181,7 +181,7 @@ LocalStore::LocalStore(const Params & params)
|
|||
, LocalFSStore(params)
|
||||
, dbDir(stateDir + "/db")
|
||||
, linksDir(realStoreDir + "/.links")
|
||||
, reservedPath(dbDir + "/reserved")
|
||||
, reservedSpacePath(dbDir + "/reserved")
|
||||
, schemaPath(dbDir + "/schema")
|
||||
, tempRootsDir(stateDir + "/temproots")
|
||||
, fnTempRoots(fmt("%s/%d", tempRootsDir, getpid()))
|
||||
|
|
@ -259,10 +259,10 @@ LocalStore::LocalStore(const Params & params)
|
|||
before doing a garbage collection. */
|
||||
try {
|
||||
struct stat st;
|
||||
if (stat(reservedPath.c_str(), &st) == -1 ||
|
||||
if (stat(reservedSpacePath.c_str(), &st) == -1 ||
|
||||
st.st_size != settings.reservedSize)
|
||||
{
|
||||
AutoCloseFD fd{open(reservedPath.c_str(), O_WRONLY | O_CREAT | O_CLOEXEC, 0600)};
|
||||
AutoCloseFD fd{open(reservedSpacePath.c_str(), O_WRONLY | O_CREAT | O_CLOEXEC, 0600)};
|
||||
int res = -1;
|
||||
#if HAVE_POSIX_FALLOCATE
|
||||
res = posix_fallocate(fd.get(), 0, settings.reservedSize);
|
||||
|
|
|
|||
|
|
@ -119,7 +119,8 @@ public:
|
|||
|
||||
const Path dbDir;
|
||||
const Path linksDir;
|
||||
const Path reservedPath;
|
||||
/** Path kept around to reserve some filesystem space to be able to begin a garbage collection */
|
||||
const Path reservedSpacePath;
|
||||
const Path schemaPath;
|
||||
const Path tempRootsDir;
|
||||
const Path fnTempRoots;
|
||||
|
|
|
|||
|
|
@ -33,7 +33,7 @@ sleep 2
|
|||
pid2=$!
|
||||
|
||||
# Start a build. This should not be blocked by the GC in progress.
|
||||
outPath=$(nix-build --max-silent-time 60 -o "$TEST_ROOT/result" -E "
|
||||
outPath=$(nix-build --max-silent-time 60 --debug -o "$TEST_ROOT/result" -E "
|
||||
with import ./config.nix;
|
||||
mkDerivation {
|
||||
name = \"non-blocking\";
|
||||
|
|
|
|||
Loading…
Reference in a new issue