forked from OSchip/llvm-project
parent
f7ce11cf71
commit
92df9c3323
|
@ -1932,3 +1932,23 @@ Replacing an icmp+select with a shift should always be considered profitable in
|
|||
instcombine.
|
||||
|
||||
//===---------------------------------------------------------------------===//
|
||||
|
||||
Re-implement atomic builtins __sync_add_and_fetch() and __sync_sub_and_fetch
|
||||
properly.
|
||||
|
||||
When the return value is not used (i.e. only care about the value in the
|
||||
memory), x86 does not have to use add to implement these. Instead, it can use
|
||||
add, sub, inc, dec instructions with the "lock" prefix.
|
||||
|
||||
This is currently implemented using a bit of instruction selection trick. The
|
||||
issue is the target independent pattern produces one output and a chain and we
|
||||
want to map it into one that just output a chain. The current trick is to select
|
||||
it into a MERGE_VALUES with the first definition being an implicit_def. The
|
||||
proper solution is to add new ISD opcodes for the no-output variant. DAG
|
||||
combiner can then transform the node before it gets to target node selection.
|
||||
|
||||
Problem #2 is we are adding a whole bunch of x86 atomic instructions when in
|
||||
fact these instructions are identical to the non-lock versions. We need a way to
|
||||
add target specific information to target nodes and have this information
|
||||
carried over to machine instructions. Asm printer (or JIT) can use this
|
||||
information to add the "lock" prefix.
|
||||
|
|
Loading…
Reference in New Issue