alx/mu4e
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

indexer: make faster

Browse Source 
Remove an (apparently) unneeded lock which help back much of the multi-core
handing.
This commit is contained in:
Dirk-Jan C. Binnema
2022-06-07 23:05:12 +03:00
parent 30c46fba61
commit 65e60a6ed5
1 changed files with 22 additions and 12 deletions
Download Patch File Download Diff File Expand all files Collapse all files

34
lib/index/mu-indexer.cc
View File

@ -99,8 +99,9 @@ struct Indexer::Private {
void item_worker(); void item_worker();
void scan_worker(); void scan_worker();
bool cleanup(); bool add_message(const std::string& path);
bool cleanup();
bool start(const Indexer::Config& conf); bool start(const Indexer::Config& conf);
bool stop(); bool stop();
@ -215,17 +216,25 @@ Indexer::Private::maybe_start_worker()
} }
} }
static bool bool
add_message(Store& store, const std::string& path) Indexer::Private::add_message(const std::string& path)
{ {
/*
* Having the lock here makes things a _lot_ slower.
*
* The reason for having the lock is some helgrind warnings;
* but it believed those are _false alarms_
* https://gitlab.gnome.org/GNOME/glib/-/issues/2662
*
* std::unique_lock lock{w_lock_};
*/
auto msg{Message::make_from_path(path)}; auto msg{Message::make_from_path(path)};
if (!msg) { if (!msg) {
g_warning("failed to create message from %s: %s", g_warning("failed to create message from %s: %s",
path.c_str(), msg.error().what()); path.c_str(), msg.error().what());
return false; return false;
} }
auto res = store_.add_message(msg.value(), true /*use-transaction*/);
auto res = store.add_message(msg.value(), true /*use-transaction*/);
if (!res) { if (!res) {
g_warning("failed to add message @ %s: %s", g_warning("failed to add message @ %s: %s",
path.c_str(), res.error().what()); path.c_str(), res.error().what());
@ -246,12 +255,11 @@ Indexer::Private::item_worker()
if (!todos_.pop(item, 250ms)) if (!todos_.pop(item, 250ms))
continue; continue;
try { try {
std::lock_guard lock{w_lock_};
switch (item.type) { switch (item.type) {
case WorkItem::Type::File: case WorkItem::Type::File: {
if (add_message(store_, item.full_path)) if (G_LIKELY(add_message(item.full_path)))
++progress_.updated; ++progress_.updated;
break; } break;
case WorkItem::Type::Dir: case WorkItem::Type::Dir:
store_.set_dirstamp(item.full_path, ::time(NULL)); store_.set_dirstamp(item.full_path, ::time(NULL));
break; break;
@ -352,9 +360,11 @@ Indexer::Private::start(const Indexer::Config& conf)
conf_ = conf; conf_ = conf;
if (conf_.max_threads == 0) { if (conf_.max_threads == 0) {
/* note, most time is spent in the (single) db thread /* benchmarking suggests that ~4 threads is the fastest (the
* but creating messages in parallel still helps a bit */ * real bottleneck is the database, so adding more threads just
max_workers_ = std::thread::hardware_concurrency()/2; * slows things down)
*/
max_workers_ = std::min(4U, std::thread::hardware_concurrency());
} else } else
max_workers_ = conf.max_threads; max_workers_ = conf.max_threads;