maxjhandsome
/
llvm-project
forked from OSchip/llvm-project
1
0
Fork 0
Code Issues Pull Requests Packages Releases Wiki Activity
llvm-project/compiler-rt/lib/builtins/arm/subsf3vfp.S

35 lines
1015 B
ArmAsm
Raw Normal View History

start adding implementation of arm *vfp routines with test cases llvm-svn: 81558
2009-09-12 04:02:01 +08:00
//===-- subsf3vfp.S - Implement subsf3vfp ---------------------------------===//
//
// The LLVM Compiler Infrastructure
//
license change llvm-svn: 119397
2010-11-17 06:13:33 +08:00
// This file is dual licensed under the MIT and the University of Illinois Open
// Source Licenses. See LICENSE.TXT for details.
start adding implementation of arm *vfp routines with test cases llvm-svn: 81558
2009-09-12 04:02:01 +08:00
//
//===----------------------------------------------------------------------===//
Add assembly.h for use in .S files. llvm-svn: 85263
2009-10-28 01:49:50 +08:00
#include "../assembly.h"
start adding implementation of arm *vfp routines with test cases llvm-svn: 81558
2009-09-12 04:02:01 +08:00
//
// extern float __subsf3vfp(float a, float b);
//
builtins: support building ARM builtins for a HF target The `-target` impacts the CC for the builtins. HF targets (with either floating point ABI) always use AAPCS VFP for the builtins unless they are AEABI builtins, in which case they use AAPCS. Non-HF targets (with either floating point ABI) always use AAPCS for the builtins and AAPCS for the AEABI builtins. This introduces the thunks necessary to switch CC for the floating point operations. This is not currently enabled, and should be dependent on the target being used to build compiler-rt. However, as a stop-gap, a define can be added for ASFLAGS to get the thunks. llvm-svn: 291677
2017-01-12 00:19:25 +08:00
// Returns the difference between two single precision floating point numbers
start adding implementation of arm *vfp routines with test cases llvm-svn: 81558
2009-09-12 04:02:01 +08:00
// using the Darwin calling convention where single arguments are passsed
// like 32-bit ints.
//
Change ARM vfp assembly functions to use unified syntax. llvm-svn: 138332
2011-08-24 00:40:18 +08:00
.syntax unified
[CompilerRT] use .p2align, .balign instead of .align The .align statements in ARM assembly routines is actually meant to be a power of 2 alignment (e.g. .align 2 == 4 byte alignment, not 2). Switch to using .p2align. .p2align is guaranteed to be a power-of-two alignment always and much more explicit. The .align in the case of x86_64 is byte alignment, use .balign instead of .align. llvm-svn: 208578
2014-05-12 23:23:37 +08:00
.p2align 2
Switch to using DEFINE_COMPILERRT_[PRIVATE_]FUNCTION to define function symbols inside .S files. llvm-svn: 85264
2009-10-28 01:50:21 +08:00
DEFINE_COMPILERRT_FUNCTION(__subsf3vfp)
builtins: support building ARM builtins for a HF target The `-target` impacts the CC for the builtins. HF targets (with either floating point ABI) always use AAPCS VFP for the builtins unless they are AEABI builtins, in which case they use AAPCS. Non-HF targets (with either floating point ABI) always use AAPCS for the builtins and AAPCS for the AEABI builtins. This introduces the thunks necessary to switch CC for the floating point operations. This is not currently enabled, and should be dependent on the target being used to build compiler-rt. However, as a stop-gap, a define can be added for ASFLAGS to get the thunks. llvm-svn: 291677
2017-01-12 00:19:25 +08:00
#if defined(COMPILER_RT_ARMHF_TARGET)
vsub.f32 s0, s0, s1
[Builtin][ARM] Fix subsf3vfp for non-HF target Reviewers: rengolin, compnerd Reviewed By: compnerd Subscribers: hans, aemerson Differential Revision: https://reviews.llvm.org/D29869 llvm-svn: 294886
2017-02-12 13:10:25 +08:00
#else
Change ARM vfp assembly functions to use unified syntax. llvm-svn: 138332
2011-08-24 00:40:18 +08:00
vmov s14, r0 // move first param from r0 into float register
vmov s15, r1 // move second param from r1 into float register
vsub.f32 s14, s14, s15
vmov r0, s14 // move result back to r0
builtins: support building ARM builtins for a HF target The `-target` impacts the CC for the builtins. HF targets (with either floating point ABI) always use AAPCS VFP for the builtins unless they are AEABI builtins, in which case they use AAPCS. Non-HF targets (with either floating point ABI) always use AAPCS for the builtins and AAPCS for the AEABI builtins. This introduces the thunks necessary to switch CC for the floating point operations. This is not currently enabled, and should be dependent on the target being used to build compiler-rt. However, as a stop-gap, a define can be added for ASFLAGS to get the thunks. llvm-svn: 291677
2017-01-12 00:19:25 +08:00
#endif
start adding implementation of arm *vfp routines with test cases llvm-svn: 81558
2009-09-12 04:02:01 +08:00
bx lr
Mechanically add end-of-function markers. llvm-svn: 200005
2014-01-24 22:33:42 +08:00
END_COMPILERRT_FUNCTION(__subsf3vfp)
builtins: tag with noexecstack These routines do not require executable stacks. However, by default ELFish linkers may assume an executable stack on GNUish environments (and some non-GNU ones too!). The GNU extension to add a note to indicate a non-executable stack is honoured by these environments to mark the stack as non-executable (the compiler normally emits this directive on appropriate targets whenever possible). This allows normal builds from getting executable stacks due to linking to the compiler rt builtins. llvm-svn: 273500
2016-06-23 06:09:42 +08:00
NO_EXEC_STACK_DIRECTIVE