Chris Lattner 21585221b6 Handle passing constant integers to functions much more efficiently. Instead ... of generating this code: mov %EAX, 4 mov DWORD PTR [%ESP], %EAX mov %AX, 123 movsx %EAX, %AX mov DWORD PTR [%ESP + 4], %EAX call Y we now generate: mov DWORD PTR [%ESP], 4 mov DWORD PTR [%ESP + 4], 123 call Y Which hurts the eyes less. :) Considering that register pressure around call sites is already high (with all of the callee clobber registers n stuff), this may help a lot. git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@12028 91177308-0d34-0410-b5e6-96231b3b80d8		2004-03-01 02:42:43 +00:00
..
Analysis	* Remove function to find "main" in a Module, there's a method for that	2004-02-29 23:09:10 +00:00
Archive
AsmParser	It turns out that the two dimensional vectors were causing big slowdowns	2004-02-09 21:03:38 +00:00
Bytecode	Adjustments to support the new ConstantAggregateZero class	2004-02-15 05:55:15 +00:00
CodeGen	Add an assert	2004-02-29 22:01:51 +00:00
Debugger
ExecutionEngine	remove obsolete comment	2004-02-26 07:59:22 +00:00
Linker	Include Config/config.h for SHLIBEXT.	2004-02-23 22:42:51 +00:00
Support	Make sure that at least one virtual method is defined in a .cpp file to avoid	2004-02-26 07:24:18 +00:00
Target	Handle passing constant integers to functions much more efficiently. Instead	2004-03-01 02:42:43 +00:00
Transforms	Disable tail duplication in a case that breaks on Olden/tsp	2004-03-01 01:12:13 +00:00
VMCore	Fix -debug-pass=Executions, which relied on Function, Module, and BasicBlock	2004-02-29 22:37:04 +00:00
Makefile	CBackend is no longer here	2004-02-13 23:29:37 +00:00