Check for stack space more intelligently.
libgcc sets the stack limit field in TCB to 256 bytes above the actual
allocated stack limit. This means if the function's stack frame needs
less than 256 bytes, we can just compare the stack pointer with the
stack limit. This should result in lesser calls to __morestack.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@145766 91177308-0d34-0410-b5e6-96231b3b80d8
diff --git a/lib/Target/X86/X86FrameLowering.cpp b/lib/Target/X86/X86FrameLowering.cpp
index 182af84..27b762a 100644
--- a/lib/Target/X86/X86FrameLowering.cpp
+++ b/lib/Target/X86/X86FrameLowering.cpp
@@ -1306,6 +1306,10 @@
}
}
+// The stack limit in the TCB is set to this many bytes above the actual stack
+// limit.
+static const uint64_t kSplitStackAvailable = 256;
+
void
X86FrameLowering::adjustForSegmentedStacks(MachineFunction &MF) const {
MachineBasicBlock &prologueMBB = MF.front();
@@ -1360,16 +1364,24 @@
TlsReg = X86::FS;
TlsOffset = 0x70;
- BuildMI(checkMBB, DL, TII.get(X86::LEA64r), ScratchReg).addReg(X86::RSP)
- .addImm(0).addReg(0).addImm(-StackSize).addReg(0);
+ if (StackSize < kSplitStackAvailable)
+ ScratchReg = X86::RSP;
+ else
+ BuildMI(checkMBB, DL, TII.get(X86::LEA64r), ScratchReg).addReg(X86::RSP)
+ .addImm(0).addReg(0).addImm(-StackSize).addReg(0);
+
BuildMI(checkMBB, DL, TII.get(X86::CMP64rm)).addReg(ScratchReg)
.addReg(0).addImm(0).addReg(0).addImm(TlsOffset).addReg(TlsReg);
} else {
TlsReg = X86::GS;
TlsOffset = 0x30;
- BuildMI(checkMBB, DL, TII.get(X86::LEA32r), ScratchReg).addReg(X86::ESP)
- .addImm(0).addReg(0).addImm(-StackSize).addReg(0);
+ if (StackSize < kSplitStackAvailable)
+ ScratchReg = X86::ESP;
+ else
+ BuildMI(checkMBB, DL, TII.get(X86::LEA32r), ScratchReg).addReg(X86::ESP)
+ .addImm(0).addReg(0).addImm(-StackSize).addReg(0);
+
BuildMI(checkMBB, DL, TII.get(X86::CMP32rm)).addReg(ScratchReg)
.addReg(0).addImm(0).addReg(0).addImm(TlsOffset).addReg(TlsReg);
}