search for: isatomic

On 11/19/2014 4:05 AM, Chandler Carruth wrote: > > On Fri, Nov 14, 2014 at 1:09 PM, Sahasrabuddhe, Sameer > <sameer.sahasrabuddhe at amd.com <mailto:sameer.sahasrabuddhe at amd.com>> > wrote: > > 1. Update the synchronization scope field in atomic instructions > from a > single bit to a wider field, say 32-bit unsigned integer. > > > I

...eline, we’re losing at least ~1% of total time to avoidable DenseMap::grow() operations, which feels a little bit… unnecessary. > > Slightly OT, but it looks like the LoadValue (value type of the AvailableLoads structure) is relatively memory inefficient. At minimum, we could get rid of the IsAtomic space by using a tagged pointer. That would at least bring us down to a 128 bits (a nice power of two). That might help speed up some of the copying. Just to note, it looks like I was testing on a somewhat older LLVM version that didn’t have LoadValue at all, so my guess is this means it’s eve...

There’s a few passes in LLVM that make heavy use of a big DenseMap, one that potentially gets filled with up to 1 entry for each instruction in the function. EarlyCSE is the best example, but Reassociate and MachineCSE have this to some degree as well (there might be others?). To put it simply: at least in my profile, EarlyCSE spends ~1/5 of its time growing DenseMaps. This is kind of… bad.

...e cmpxchg address does not alias the location, it does not access it. diff --git a/lib/AsmParser/LLParser.cpp b/lib/AsmParser/LLParser.cpp index a4bbcfc..fddada9 100644 --- a/lib/AsmParser/LLParser.cpp +++ b/lib/AsmParser/LLParser.cpp @@ -1517,6 +1517,15 @@ bool LLParser::ParseScopeAndOrdering(bool isAtomic, SynchronizationScope &Scope, Scope = CrossThread; if (EatIfPresent(lltok::kw_singlethread)) Scope = SingleThread; + + return ParseOrdering(Ordering); +} + +/// ParseOrdering +/// ::= AtomicOrdering +/// +/// This sets Ordering to the parsed value. +bool LLParser::ParseOrdering(At...