In #10527 performance was measured and ResetEvent did take a lot of time
in these measurements. The reset is only required once though, so use a
more fine grained trigger condition.
1)
Added missing checks for CreateEvent which also required the
following related changes:
- changed freerdp_context_new API to BOOL
- changed freerdp_peer_context_new API to BOOL
- changed pRdpClientNew callback to BOOL
- changed pContextNew callback to BOOL
- changed psPeerAccepted callback to BOOL
- changed psPeerContextNew callback to BOOL
2)
Fixed lots of missing alloc and error checks in the
changed code's neighbourhood.
3)
Check freerdp_client_codecs_prepare result to avoid segfaults
caused by using non-initialized codecs.
4)
Fixed deadlocks in x11 caused by missing xf_unlock_x11() calls
in some error handlers
5)
Some fixes in thread pool:
- DEFAULT_POOL assignment did not match TP_POOL definition
- don't free the pool pointer if it points to the static DEFAULT_POOL
- added error handling and cleanup in InitializeThreadpool
Use InitializeCriticalSectionAndSpinCount instead of IntializeCriticalSection.
Using spin counts for critical sections of short duration enables the calling
thread to avoid the wait operation in most situations which can dramatically
improve the overall performance on multiprocessor systems.
On Linux this change has no effect because the new winpr critical section
implementation does not use the SpinCount field under Linux because the NPTL
synchronization primitives are implemented using the extremely performant
futex system calls which have this magic already built in.
However, on Mac OS X this change improved the overall performance of the
multithreaded RemoteFX decoder by 25 percent.
I've used a SpinCount of 4000 which avoided 99 percent of the wait calls.
This value is also used by Microsoft's heap manager for its per-heap
critical sections.
Note: This change requires pull request #1397 to be merged.
- Improved/completed(almost) winpr's critical section implementation
- Replaced WaitForSingleObject locking with critical sections
Note:
WaitForSingleObject should _never_ be used for granular low-contention
locks as it _always_ enters the kernel.
Just replacing WaitForSingleObject locking in Bufferpool with
EnterCriticalSection boosts the multithreaded rfx decoder
performance by almost 400% on win32.