[XRay] Create an Index of sleds per function

[dberris]

Summary

This change adds a new section to the xray-instrumented binary that
stores an index into ranges of the instrumentation map, where sleds
associated with the same function can be accessed as an array. At
runtime, we can get access to this index by function ID offset allowing
for selective patching and unpatching by function ID.

Each entry in this new section (xray_fn_idx) will include two pointers
indicating the start and one past the end of the sleds associated with
the same function. These entries will be 16 bytes long on x86 and
aarch64. On arm, we align to 16 bytes anyway so the runtime has to take
that into consideration.

__{start,stop}_xray_fn_idx will be the symbols that the runtime will
look for when we implement the selective patching/unpatching by function
id APIs. Because XRay synthesizes the function id's in a monotonically
increasing manner at runtime now, implementations (and users) can use
this table to look up the sleds associated with a specific function.
This is useful in implementations that want to do things like:

• Implement coverage mode for functions by patching everything pre-main, then as functions are encountered, the installed handler can unpatch the function that's been encountered after recording that it's been called.
• Do "learning mode", so that the implementation can figure out some statistical information about function calls by function id for a time being, and then determine which functions are worth uninstrumenting at runtime.
• Do "selective instrumentation" where an implementation can specifically instrument only certain function id's at runtime (either based on some external data, or through some other heuristics) instead of patching all the instrumented functions at runtime.

[dblaikie]

Looks reasonable to me. (others can speak up if they want to hold off on committing this until they've had a chance to look)

[dberris]

Thanks @dblaikie -- I'll land and deal with follow-ups as required in the meantime.

Cheers

include/llvm/CodeGen/AsmPrinter.h

struct XRayFunctionEntry {
     SledKind Kind;
     bool AlwaysInstrument;
     const class Function *Fn;
 
     void emit(int, MCStreamer *, const MCSymbol *) const;
   };
 
   // All the sleds to be emitted.
-  std::vector<XRayFunctionEntry> Sleds;
+  SmallVector<XRayFunctionEntry, 4> Sleds;
 
   // Helper function to record a given XRay sled.
   void recordSled(MCSymbol *Sled, const MachineInstr &MI, SledKind Kind);
 
   /// Emit a table with all XRay instrumentation points.
   void emitXRayTable();
 
   //===------------------------------------------------------------------===//

lib/CodeGen/AsmPrinter/AsmPrinter.cpp

 }
 
 void AsmPrinter::emitXRayTable() {
   if (Sleds.empty())
     return;
 
   auto PrevSection = OutStreamer->getCurrentSectionOnly();
   auto Fn = MF->getFunction();
-  MCSection *Section = nullptr;
+  MCSection *InstMap = nullptr;
+  MCSection *FnSledIndex = nullptr;
   if (MF->getSubtarget().getTargetTriple().isOSBinFormatELF()) {
     if (Fn->hasComdat()) {
-      Section = OutContext.getELFSection("xray_instr_map", ELF::SHT_PROGBITS,
+      InstMap = OutContext.getELFSection("xray_instr_map", ELF::SHT_PROGBITS,
+                                         ELF::SHF_ALLOC | ELF::SHF_GROUP, 0,
+                                         Fn->getComdat()->getName());
+      FnSledIndex = OutContext.getELFSection("xray_fn_idx", ELF::SHT_PROGBITS,
                                              ELF::SHF_ALLOC | ELF::SHF_GROUP, 0,
                                              Fn->getComdat()->getName());
     } else {
-      Section = OutContext.getELFSection("xray_instr_map", ELF::SHT_PROGBITS,
+      InstMap = OutContext.getELFSection("xray_instr_map", ELF::SHT_PROGBITS,
+                                         ELF::SHF_ALLOC);
+      FnSledIndex = OutContext.getELFSection("xray_fn_idx", ELF::SHT_PROGBITS,
                                              ELF::SHF_ALLOC);
     }
   } else if (MF->getSubtarget().getTargetTriple().isOSBinFormatMachO()) {
-    Section = OutContext.getMachOSection("__DATA", "xray_instr_map", 0,
+    InstMap = OutContext.getMachOSection("__DATA", "xray_instr_map", 0,
+                                         SectionKind::getReadOnlyWithRel());
+    FnSledIndex = OutContext.getMachOSection("__DATA", "xray_fn_idx", 0,
                                              SectionKind::getReadOnlyWithRel());
   } else {
     llvm_unreachable("Unsupported target");
   }
 
   // Before we switch over, we force a reference to a label inside the
-  // xray_instr_map section. Since this function is always called just
-  // before the function's end, we assume that this is happening after
-  // the last return instruction.
-
+  // xray_instr_map and xray_fn_idx sections. Since this function is always
+  // called just before the function's end, we assume that this is happening
+  // after the last return instruction. We also use the synthetic label in the
+  // xray_inster_map as a delimeter for the range of sleds for this function in
+  // the index.
   auto WordSizeBytes = MAI->getCodePointerSize();
-  MCSymbol *Tmp = OutContext.createTempSymbol("xray_synthetic_", true);
+  MCSymbol *SledsStart = OutContext.createTempSymbol("xray_synthetic_", true);
+  MCSymbol *IdxRef = OutContext.createTempSymbol("xray_fn_idx_synth_", true);
   OutStreamer->EmitCodeAlignment(16);
-  OutStreamer->EmitSymbolValue(Tmp, WordSizeBytes, false);
-  OutStreamer->SwitchSection(Section);
-  OutStreamer->EmitLabel(Tmp);
+  OutStreamer->EmitSymbolValue(SledsStart, WordSizeBytes, false);
+  OutStreamer->EmitSymbolValue(IdxRef, WordSizeBytes, false);
+
+  // Now we switch to the instrumentation map section. Because this is done
+  // per-function, we are able to create an index entry that will represent the
+  // range of sleds associated with a function.
+  OutStreamer->SwitchSection(InstMap);
+  OutStreamer->EmitLabel(SledsStart);
   for (const auto &Sled : Sleds)
     Sled.emit(WordSizeBytes, OutStreamer.get(), CurrentFnSym);
+  MCSymbol *SledsEnd = OutContext.createTempSymbol("xray_synthetic_end", true);
+  OutStreamer->EmitLabel(SledsEnd);
 
+  // We then emit a single entry in the index per function. We use the symbols
+  // that bound the instrumentation map as the range for a specific function.
+  // Each entry here will be 16-byte aligned, as we're writing down two
+  // pointers.
+  OutStreamer->SwitchSection(FnSledIndex);
+  OutStreamer->EmitCodeAlignment(16);
+  OutStreamer->EmitLabel(IdxRef);
+  OutStreamer->EmitSymbolValue(SledsStart, WordSizeBytes);
+  OutStreamer->EmitSymbolValue(SledsEnd, WordSizeBytes);
   OutStreamer->SwitchSection(PrevSection);
   Sleds.clear();
 }
 
 void AsmPrinter::recordSled(MCSymbol *Sled, const MachineInstr &MI,
   SledKind Kind) {
   auto Fn = MI.getParent()->getParent()->getFunction();
   auto Attr = Fn->getFnAttribute("function-instrument");

test/CodeGen/AArch64/xray-attribute-instrumentation.ll

 ; CHECK-NEXT:  nop
 ; CHECK-NEXT:  nop
 ; CHECK-NEXT:  nop
 ; CHECK-LABEL: Ltmp1:
 ; CHECK-NEXT:  ret
 }
 ; CHECK:       .p2align 4
 ; CHECK-NEXT:  .xword .Lxray_synthetic_0
+; CHECK-NEXT:  .xword .Lxray_fn_idx_synth_0
 ; CHECK-NEXT:  .section xray_instr_map,{{.*}}
 ; CHECK-LABEL: Lxray_synthetic_0:
 ; CHECK:       .xword .Lxray_sled_0
 ; CHECK:       .xword .Lxray_sled_1

test/CodeGen/AArch64/xray-tail-call-sled.ll

 ; CHECK-NEXT:  nop
 ; CHECK-NEXT:  nop
 ; CHECK-NEXT:  nop
 ; CHECK-LABEL: .Ltmp1:
 ; CHECK-NEXT:  ret
 }
 ; CHECK:       .p2align 4
 ; CHECK-NEXT:  .xword .Lxray_synthetic_0
+; CHECK-NEXT:  .xword .Lxray_fn_idx_synth_0
 ; CHECK-NEXT:  .section xray_instr_map,{{.*}}
 ; CHECK-LABEL: Lxray_synthetic_0:
 ; CHECK:       .xword .Lxray_sled_0
 ; CHECK:       .xword .Lxray_sled_1
+; CHECK-LABEL: Lxray_synthetic_end0:
+; CHECK:       .section xray_fn_idx,{{.*}}
+; CHECK-LABEL: Lxray_fn_idx_synth_0:
+; CHECK:       .xword .Lxray_synthetic_0
+; CHECK-NEXT:  .xword .Lxray_synthetic_end0
 
 define i32 @caller() nounwind noinline uwtable "function-instrument"="xray-always" {
 ; CHECK:       .p2align 2
 ; CHECK-LABEL: .Lxray_sled_2:
 ; CHECK-NEXT:  b  #32
 ; CHECK-NEXT:  nop
 ; CHECK-NEXT:  nop
 ; CHECK-NEXT:  nop

 ; CHECK-NEXT:  nop
 ; CHECK-LABEL: .Ltmp3:
   %retval = tail call i32 @callee()
 ; CHECK:       b  callee
   ret i32 %retval
 }
 ; CHECK:       .p2align 4
 ; CHECK-NEXT:  .xword .Lxray_synthetic_1
+; CHECK-NEXT:  .xword .Lxray_fn_idx_synth_1
 ; CHECK-NEXT:  .section xray_instr_map,{{.*}}
 ; CHECK-LABEL: Lxray_synthetic_1:
 ; CHECK:       .xword .Lxray_sled_2
 ; CHECK:       .xword .Lxray_sled_3
+; CHECK-LABEL: Lxray_synthetic_end1:
+; CHECK:       .section xray_fn_idx,{{.*}}
+; CHECK-LABEL: Lxray_fn_idx_synth_1:
+; CHECK:       .xword .Lxray_synthetic_1
+; CHECK-NEXT:  .xword .Lxray_synthetic_end1

test/CodeGen/ARM/xray-armv6-attribute-instrumentation.ll

 ; CHECK-NEXT:  mov  r0, r0
 ; CHECK-NEXT:  mov  r0, r0
 ; CHECK-NEXT:  mov  r0, r0
 ; CHECK-LABEL: Ltmp1:
 ; CHECK-NEXT:  bx lr
 }
 ; CHECK:       .p2align 4
 ; CHECK-NEXT:  .long {{.*}}Lxray_synthetic_0
+; CHECK-NEXT:  .long {{.*}}Lxray_fn_idx_synth_0
 ; CHECK-NEXT:  .section {{.*}}xray_instr_map{{.*}}
 ; CHECK-LABEL: Lxray_synthetic_0:
 ; CHECK:       .long {{.*}}Lxray_sled_0
 ; CHECK:       .long {{.*}}Lxray_sled_1
+; CHECK-LABEL: Lxray_synthetic_end0:
+; CHECK:       .section {{.*}}xray_fn_idx{{.*}}
+; CHECK-LABEL: Lxray_fn_idx_synth_0:
+; CHECK:       .long {{.*}}Lxray_synthetic_0
+; CHECK-NEXT:  .long {{.*}}Lxray_synthetic_end0

test/CodeGen/ARM/xray-armv7-attribute-instrumentation.ll

 ; CHECK-NEXT:  nop
 ; CHECK-NEXT:  nop
 ; CHECK-NEXT:  nop
 ; CHECK-LABEL: Ltmp1:
 ; CHECK-NEXT:  bx lr
 }
 ; CHECK:       .p2align 4
 ; CHECK-NEXT:  .long {{.*}}Lxray_synthetic_0
+; CHECK-NEXT:  .long {{.*}}Lxray_fn_idx_synth_0
 ; CHECK-NEXT:  .section {{.*}}xray_instr_map{{.*}}
 ; CHECK-LABEL: Lxray_synthetic_0:
 ; CHECK:       .long {{.*}}Lxray_sled_0
 ; CHECK:       .long {{.*}}Lxray_sled_1
+; CHECK-LABEL: Lxray_synthetic_end0:
+; CHECK:       .section {{.*}}xray_fn_idx{{.*}}
+; CHECK-LABEL: Lxray_fn_idx_synth_0:
+; CHECK:       .long {{.*}}xray_synthetic_0
+; CHECK-NEXT:  .long {{.*}}xray_synthetic_end0
+

test/CodeGen/X86/xray-attribute-instrumentation.ll

; CHECK-LABEL: Ltmp0:
   ret i32 0
 ; CHECK:       .p2align 1, 0x90
 ; CHECK-LABEL: Lxray_sled_1:
 ; CHECK-NEXT:  retq
 ; CHECK-NEXT:  nopw %cs:512(%rax,%rax)
 }
 ; CHECK:       .p2align 4, 0x90
 ; CHECK-NEXT:  .quad {{.*}}xray_synthetic_0
+; CHECK-NEXT:  .quad {{.*}}xray_fn_idx_synth_0
 ; CHECK-NEXT:  .section {{.*}}xray_instr_map
 ; CHECK-LABEL: Lxray_synthetic_0:
 ; CHECK:       .quad {{.*}}xray_sled_0
 ; CHECK:       .quad {{.*}}xray_sled_1
+; CHECK-LABEL: Lxray_synthetic_end0:
+; CHECK:       .section {{.*}}xray_fn_idx
+; CHECK-LABEL: Lxray_fn_idx_synth_0:
+; CHECK:       .quad {{.*}}xray_synthetic_0
+; CHECK-NEXT:  .quad {{.*}}xray_synthetic_end0
+
 
 ; We test multiple returns in a single function to make sure we're getting all
 ; of them with XRay instrumentation.
 define i32 @bar(i32 %i) nounwind noinline uwtable "function-instrument"="xray-always" {
 ; CHECK:       .p2align 1, 0x90
 ; CHECK-LABEL: Lxray_sled_2:
 ; CHECK-NEXT:  .ascii "\353\t"
 ; CHECK-NEXT:  nopw 512(%rax,%rax)
NotEqual:
   ret i32 1
 ; CHECK:       .p2align 1, 0x90
 ; CHECK-LABEL: Lxray_sled_4:
 ; CHECK-NEXT:  retq
 ; CHECK-NEXT:  nopw %cs:512(%rax,%rax)
 }
 ; CHECK:       .p2align 4, 0x90
 ; CHECK-NEXT:  .quad {{.*}}xray_synthetic_1
+; CHECK-NEXT:  .quad {{.*}}xray_fn_idx_synth_1
 ; CHECK-NEXT:  .section {{.*}}xray_instr_map
 ; CHECK-LABEL: Lxray_synthetic_1:
 ; CHECK:       .quad {{.*}}xray_sled_2
 ; CHECK:       .quad {{.*}}xray_sled_3
 ; CHECK:       .quad {{.*}}xray_sled_4
+; CHECK-LABEL: Lxray_synthetic_end1:
+; CHECK:       .section {{.*}}xray_fn_idx
+; CHECK-LABEL: Lxray_fn_idx_synth_1:
+; CHECK:       .quad {{.*}}xray_synthetic_1
+; CHECK-NEXT:  .quad {{.*}}xray_synthetic_end1

test/CodeGen/X86/xray-tail-call-sled.ll

 ; RUN: llc -filetype=asm -o - -mtriple=x86_64-unknown-linux-gnu < %s | FileCheck %s
 ; RUN: llc -filetype=asm -o - -mtriple=x86_64-darwin-unknown    < %s | FileCheck %s
 
 define i32 @callee() nounwind noinline uwtable "function-instrument"="xray-always" {
 ; CHECK:       .p2align 1, 0x90
 ; CHECK-LABEL: Lxray_sled_0:
 ; CHECK-NEXT:  .ascii "\353\t"
 ; CHECK-NEXT:  nopw 512(%rax,%rax)
 ; CHECK-LABEL: Ltmp0:
   ret i32 0
 ; CHECK:       .p2align 1, 0x90
 ; CHECK-LABEL: Lxray_sled_1:
 ; CHECK-NEXT:  retq
 ; CHECK-NEXT:  nopw %cs:512(%rax,%rax)
 }
 ; CHECK:       .p2align 4, 0x90
-; CHECK-NEXT:  .quad {{.*}}xray_synthetic_0
+; CHECK-NEXT:  .quad {{.*}}xray_synthetic_0{{.*}}
+; CHECK-NEXT:  .quad {{.*}}xray_fn_idx_synth_0{{.*}}
 ; CHECK-NEXT:  .section {{.*}}xray_instr_map
 ; CHECK-LABEL: Lxray_synthetic_0:
 ; CHECK:       .quad {{.*}}xray_sled_0
 ; CHECK:       .quad {{.*}}xray_sled_1
+; CHECK-LABEL: Lxray_synthetic_end0:
+; CHECK-NEXT:  .section {{.*}}xray_fn_idx
+; CHECK-LABEL: Lxray_fn_idx_synth_0:
+; CHECK:       .quad {{.*}}xray_synthetic_0
+; CHECK-NEXT:  .quad {{.*}}xray_synthetic_end0
 
 define i32 @caller() nounwind noinline uwtable "function-instrument"="xray-always" {
 ; CHECK:       .p2align 1, 0x90
 ; CHECK-LABEL: Lxray_sled_2:
 ; CHECK-NEXT:  .ascii "\353\t"
 ; CHECK-NEXT:  nopw 512(%rax,%rax)
 ; CHECK-LABEL: Ltmp1:
 ; CHECK:       .p2align 1, 0x90
 ; CHECK-LABEL: Lxray_sled_3:
 ; CHECK-NEXT:  .ascii "\353\t"
 ; CHECK-NEXT:  nopw 512(%rax,%rax)
 ; CHECK-LABEL: Ltmp2:
   %retval = tail call i32 @callee()
 ; CHECK:       jmp {{.*}}callee {{.*}}# TAILCALL
   ret i32 %retval
 }
 ; CHECK:       .p2align 4, 0x90
-; CHECK-NEXT:  .quad {{.*}}xray_synthetic_1
+; CHECK-NEXT:  .quad {{.*}}xray_synthetic_1{{.*}}
+; CHECK-NEXT:  .quad {{.*}}xray_fn_idx_synth_1{{.*}}
 ; CHECK-LABEL: Lxray_synthetic_1:
 ; CHECK:       .quad {{.*}}xray_sled_2
 ; CHECK:       .quad {{.*}}xray_sled_3
+; CHECK-LABEL: Lxray_synthetic_end1:
+; CHECK:       .section {{.*}}xray_fn_idx
+; CHECK-LABEL: Lxray_fn_idx_synth_1:
+; CHECK:       .quad {{.*}}xray_synthetic_1
+; CHECK:       .quad {{.*}}xray_synthetic_end1