Sanjay Patel 138d4e067f transform fmin/fmax calls when possible (PR24314) ... If we can ignore NaNs, fmin/fmax libcalls can become compare and select (this is what we turn std::min / std::max into). This IR should then be optimized in the backend to whatever is best for any given target. Eg, x86 can use minss/maxss instructions. This should solve PR24314: https://llvm.org/bugs/show_bug.cgi?id=24314 Differential Revision: http://reviews.llvm.org/D11866 llvm-svn: 245187		2015-08-16 20:18:19 +00:00
..
InstCombine	Revert r240137 (Fixed/added namespace ending comments using clang-tidy. NFC)	2015-06-23 09:49:53 +00:00
IPO	LowerBitSets: Add debugging output.	2015-07-29 18:12:36 +00:00
Scalar	Revert r240137 (Fixed/added namespace ending comments using clang-tidy. NFC)	2015-06-23 09:49:53 +00:00
Utils	transform fmin/fmax calls when possible (PR24314)	2015-08-16 20:18:19 +00:00
Instrumentation.h	[sancov] Leave llvm.localescape in the entry block	2015-08-14 16:45:42 +00:00
IPO.h	Resubmit "Add new EliminateAvailableExternally module pass" (r239480)	2015-07-06 16:22:42 +00:00
ObjCARC.h	Revert r240137 (Fixed/added namespace ending comments using clang-tidy. NFC)	2015-06-23 09:49:53 +00:00
Scalar.h	[GlobalMerge] Allow targets to enable merging of extern variables, NFC.	2015-08-03 12:08:41 +00:00
Vectorize.h	Revert r240137 (Fixed/added namespace ending comments using clang-tidy. NFC)	2015-06-23 09:49:53 +00:00