Mon Jan 12 04:10:24 PST 2009 Simon Marlow <[email protected]>
* Keep the remembered sets local to each thread during parallel GC
This turns out to be quite vital for parallel programs:
- The way we discover which threads to traverse is by finding
dirty threads via the remembered sets (aka mutable lists).
- A dirty thread will be on the remembered set of the capability
that was running it, and we really want to traverse that thread's
stack using the GC thread for the capability, because it is in
that CPU's cache. If we get this wrong, we get penalised badly by
the memory system.
Previously we had per-capability mutable lists but they were
aggregated before GC and traversed by just one of the GC threads.
This resulted in very poor performance particularly for parallel
programs with deep stacks.
Now we keep per-capability remembered sets throughout GC, which also
removes a lock (recordMutableGen_sync).
M ./includes/Storage.h -32 +10
M ./rts/Capability.c +3
M ./rts/Capability.h -3 +5
M ./rts/Stats.c -1
M ./rts/Updates.h -1 +1
M ./rts/sm/Compact.c +9
M ./rts/sm/GC.c -38 +50
M ./rts/sm/GCThread.h +8
M ./rts/sm/GCUtils.h +19
M ./rts/sm/Scav.c -17 +32
M ./rts/sm/Scav.h -2 +4
M ./rts/sm/Storage.c -2 +1
View patch online:
http://darcs.haskell.org/ghc/_darcs/patches/20090112121024-12142-5e1f5762b5ba553388493e7057132e4e2a5155a8.gz
_______________________________________________
Cvs-ghc mailing list
[email protected]
http://www.haskell.org/mailman/listinfo/cvs-ghc
