Hey,
I wanted to look at the locking strategy of the DUChain lock for a
while, but never got around to do it. Now, at 3:30 am on a friday, I
finally did (what else would you do at this time) and the results are at
least ... interesting:
Replacing QThread::usleep(500) with sched_yield() makes the number of
effective used CPUs go up from ~1.7 to ~3.3 and the wall clock time down
from ~170 to ~110 seconds on duchainify kdevplatform [1]. See:
https://paste.kde.org/pncqjowxp
For kdev-python, the effect is somewhat less severe but still noticeable:
https://paste.kde.org/pcirhvys9
Could anyone try the attached (5-line) patch and look if you see this
speedup as well? I feel like I might be doing something wrong.
On the other hand, I think the "sleep 500us" strategy is quite terrible,
if only because 500us is relatively a long time and nothing else will
happen during that wait for _any_ thread iirc.
Good night,
Sven
________________
[1] Conditions: everything built in release mode, CLEAR_DUCHAIN_DIR=1
set before invocation, -r -u flags passed to duchainify, 4 threads, CPU
is a 4-core (8 threads with hyperthreading) Intel i7 3720 QM
-------------- next part --------------
diff --git a/language/duchain/duchainlock.cpp b/language/duchain/duchainlock.cpp
index aa842ce..12f2ef6 100644
--- a/language/duchain/duchainlock.cpp
+++ b/language/duchain/duchainlock.cpp
@@ -26,6 +26,9 @@
#include <QThreadStorage>
#include <QElapsedTimer>
+#ifdef Q_OS_LINUX
+#include <sched.h>
+#endif
///@todo Always prefer exactly that lock that is requested by the thread that has the foreground mutex,
/// to reduce the amount of UI blocking.
@@ -94,7 +97,7 @@ bool DUChainLock::lockForRead(unsigned int timeout)
while (d->m_writer.loadAcquire()) {
if (!timeout || t.elapsed() < timeout) {
- QThread::usleep(uSleepTime);
+ sched_yield();
} else {
//Fail!
d->changeOwnReaderRecursion(-1);
@@ -149,7 +152,7 @@ bool DUChainLock::lockForWrite(uint timeout)
}
if (!timeout || t.elapsed() < timeout) {
- QThread::usleep(uSleepTime);
+ sched_yield();
} else {
//Fail!
return false;
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: OpenPGP digital signature
URL: <http://mail.kde.org/pipermail/kdevelop-devel/attachments/20160716/5fa887db/attachment.sig>