diff options
author | Reid Kleckner <rnk@google.com> | 2018-06-25 18:16:27 +0000 |
---|---|---|
committer | Reid Kleckner <rnk@google.com> | 2018-06-25 18:16:27 +0000 |
commit | 494d9457845a5733428f00bd0cd525d6bce6e556 (patch) | |
tree | 20fea7c29b7a3e0281bd81b4a5724b285060b491 /llvm/lib/Target/X86/X86TargetMachine.cpp | |
parent | fe36919fcdd641cb343f7983719b1742e891ffe8 (diff) |
Re-land r335297 "[X86] Implement more of x86-64 large and medium PIC code models"
The large code model allows code and data segments to exceed 2GB, which
means that some symbol references may require a displacement that cannot
be encoded as a displacement from RIP. The large PIC model even relaxes
the assumption that the GOT itself is within 2GB of all code. Therefore,
we need a special code sequence to materialize it:
.LtmpN:
leaq .LtmpN(%rip), %rbx
movabsq $_GLOBAL_OFFSET_TABLE_-.LtmpN, %rax # Scratch
addq %rax, %rbx # GOT base reg
From that, non-local references go through the GOT base register instead
of being PC-relative loads. Local references typically use GOTOFF
symbols, like this:
movq extern_gv@GOT(%rbx), %rax
movq local_gv@GOTOFF(%rbx), %rax
All calls end up being indirect:
movabsq $local_fn@GOTOFF, %rax
addq %rbx, %rax
callq *%rax
The medium code model retains the assumption that the code segment is
less than 2GB, so calls are once again direct, and the RIP-relative
loads can be used to access the GOT. Materializing the GOT is easy:
leaq _GLOBAL_OFFSET_TABLE_(%rip), %rbx # GOT base reg
DSO local data accesses will use it:
movq local_gv@GOTOFF(%rbx), %rax
Non-local data accesses will use RIP-relative addressing, which means we
may not always need to materialize the GOT base:
movq extern_gv@GOTPCREL(%rip), %rax
Direct calls are basically the same as they are in the small code model:
They use direct, PC-relative addressing, and the PLT is used for calls
to non-local functions.
This patch adds reasonably comprehensive testing of LEA, but there are
lots of interesting folding opportunities that are unimplemented.
I restricted the MCJIT/eh-lg-pic.ll test to Linux, since the large PIC
code model is not implemented for MachO yet.
Differential Revision: https://reviews.llvm.org/D47211
Diffstat (limited to 'llvm/lib/Target/X86/X86TargetMachine.cpp')
-rw-r--r-- | llvm/lib/Target/X86/X86TargetMachine.cpp | 8 |
1 files changed, 7 insertions, 1 deletions
diff --git a/llvm/lib/Target/X86/X86TargetMachine.cpp b/llvm/lib/Target/X86/X86TargetMachine.cpp index a713bddf7bf..8402434cd47 100644 --- a/llvm/lib/Target/X86/X86TargetMachine.cpp +++ b/llvm/lib/Target/X86/X86TargetMachine.cpp @@ -156,9 +156,15 @@ static std::string computeDataLayout(const Triple &TT) { } static Reloc::Model getEffectiveRelocModel(const Triple &TT, + bool JIT, Optional<Reloc::Model> RM) { bool is64Bit = TT.getArch() == Triple::x86_64; if (!RM.hasValue()) { + // JIT codegen should use static relocations by default, since it's + // typically executed in process and not relocatable. + if (JIT) + return Reloc::Static; + // Darwin defaults to PIC in 64 bit mode and dynamic-no-pic in 32 bit mode. // Win64 requires rip-rel addressing, thus we force it to PIC. Otherwise we // use static relocation model by default. @@ -210,7 +216,7 @@ X86TargetMachine::X86TargetMachine(const Target &T, const Triple &TT, CodeGenOpt::Level OL, bool JIT) : LLVMTargetMachine( T, computeDataLayout(TT), TT, CPU, FS, Options, - getEffectiveRelocModel(TT, RM), + getEffectiveRelocModel(TT, JIT, RM), getEffectiveCodeModel(CM, JIT, TT.getArch() == Triple::x86_64), OL), TLOF(createTLOF(getTargetTriple())) { // Windows stack unwinder gets confused when execution flow "falls through" |