libunwind: Use APIs exposed by RTLD to unwind the trusted stack

Assembly stubs for _rtld_unw_{get,set}context are no longer needed. Due to the significantly simplified implementation, the LIBUNWIND_CHERI_C18N_SUPPORT option has been removed and c18n support is now included by default for supported architectures (currently Morello only).
CTSRD-CHERI · Oct 24, 2024 · 8112874 · 8112874
1 parent 0621ed3
commit 8112874
Show file tree

Hide file tree

Showing 11 changed files with 94 additions and 203 deletions.
diff --git a/libunwind/CMakeLists.txt b/libunwind/CMakeLists.txt
@@ -50,7 +50,6 @@ option(LIBUNWIND_IS_BAREMETAL "Build libunwind for baremetal targets." OFF)
 option(LIBUNWIND_USE_FRAME_HEADER_CACHE "Cache frame headers for unwinding. Requires locking dl_iterate_phdr." OFF)
 option(LIBUNWIND_REMEMBER_HEAP_ALLOC "Use heap instead of the stack for .cfi_remember_state." OFF)
 option(LIBUNWIND_INSTALL_HEADERS "Install the libunwind headers." OFF)
-option(LIBUNWIND_CHERI_C18N_SUPPORT "Use a libunwind implementation that supports a CHERI c18n RTLD." OFF)
 
 set(LIBUNWIND_LIBDIR_SUFFIX "${LLVM_LIBDIR_SUFFIX}" CACHE STRING
     "Define suffix of library directory name (32/64)")
@@ -294,14 +293,6 @@ if (NOT LIBUNWIND_ENABLE_THREADS)
   add_compile_flags(-D_LIBUNWIND_HAS_NO_THREADS)
 endif()
 
-# Sandboxing and c18n support
-if (LIBUNWIND_CHERI_C18N_SUPPORT)
-  if (NOT CMAKE_SYSTEM_PROCESSOR MATCHES "ARM64" OR NOT CMAKE_SYSTEM_NAME MATCHES "FreeBSD")
-    message(FATAL_ERROR "LIBUNWIND_CHERI_C18N_SUPPORT is not supported for ${CMAKE_SYSTEM_NAME}-${CMAKE_SYSTEM_PROCESSOR}")
-  endif()
-  add_compile_flags(-D_LIBUNWIND_CHERI_C18N_SUPPORT)
-endif()
-
 # ARM WMMX register support
 if (LIBUNWIND_ENABLE_ARM_WMMX)
   # __ARM_WMMX is a compiler pre-define (as per the ACLE 2.0). Clang does not
@@ -337,6 +328,10 @@ if (C_SUPPORTS_COMMENT_LIB_PRAGMA)
   endif()
 endif()
 
+if (LIBUNWIND_HAS_CHERI_LIB_C18N)
+  add_compile_definitions(_LIBUNWIND_HAS_CHERI_LIB_C18N)
+endif()
+
 #===============================================================================
 # Setup Source Code
 #===============================================================================

diff --git a/libunwind/cmake/config-ix.cmake b/libunwind/cmake/config-ix.cmake
@@ -111,3 +111,5 @@ else()
   check_library_exists(dl dladdr "" LIBUNWIND_HAS_DL_LIB)
   check_library_exists(pthread pthread_once "" LIBUNWIND_HAS_PTHREAD_LIB)
 endif()
+
+check_symbol_exists(dl_c18n_is_trampoline link.h LIBUNWIND_HAS_CHERI_LIB_C18N)
diff --git a/libunwind/include/__libunwind_config.h b/libunwind/include/__libunwind_config.h
@@ -235,9 +235,4 @@
 # define _LIBUNWIND_HIGHEST_DWARF_REGISTER 287
 #endif // _LIBUNWIND_IS_NATIVE_ONLY
 
-#if defined(_LIBUNWIND_CHERI_C18N_SUPPORT) &&                                  \
-    !defined(_LIBUNWIND_TARGET_AARCH64)
-# error "LIBUNWIND_CHERI_C18N_SUPPORT is only supported on Morello"
-#endif
-
 #endif // ____LIBUNWIND_CONFIG_H__
diff --git a/libunwind/src/AddressSpace.hpp b/libunwind/src/AddressSpace.hpp
@@ -321,12 +321,6 @@ class _LIBUNWIND_HIDDEN LocalAddressSpace {
     return get<v128>(addr);
   }
   capability_t     getCapability(pint_t addr) { return get<capability_t>(addr); }
-#if defined(__CHERI_PURE_CAPABILITY__) && defined(_LIBUNWIND_CHERI_C18N_SUPPORT)
-  static pint_t getUnwindSealer();
-  static bool isValidSealer(pint_t sealer) {
-    return __builtin_cheri_tag_get(sealer);
-  }
-#endif // __CHERI_PURE_CAPABILITY__ && _LIBUNWIND_CHERI_C18N_SUPPORT
   __attribute__((always_inline))
   uintptr_t       getP(pint_t addr);
   uint64_t        getRegister(pint_t addr);
@@ -415,24 +409,6 @@ inline uint64_t LocalAddressSpace::getRegister(pint_t addr) {
 #endif
 }
 
-#if defined(__CHERI_PURE_CAPABILITY__) && defined(_LIBUNWIND_CHERI_C18N_SUPPORT)
-extern "C" {
-/// Call into the RTLD to get a sealer capability. This sealer will be used to
-/// seal information in the unwinding context.
-uintptr_t _rtld_unw_getsealer();
-uintptr_t __rtld_unw_getsealer();
-_LIBUNWIND_HIDDEN uintptr_t __rtld_unw_getsealer() {
-  return (uintptr_t)0;
-}
-_LIBUNWIND_WEAK_ALIAS(__rtld_unw_getsealer, _rtld_unw_getsealer)
-}
-
-/// C++ wrapper for calling into RTLD.
-inline LocalAddressSpace::pint_t LocalAddressSpace::getUnwindSealer() {
-  return _rtld_unw_getsealer();
-}
-#endif // __CHERI_PURE_CAPABILITY__ && _LIBUNWIND_CHERI_C18N_SUPPORT
-
 /// Read a ULEB128 into a 64-bit word.
 inline uint64_t LocalAddressSpace::getULEB128(pint_t &addr, pint_t end) {
   const uint8_t *p = (uint8_t *)addr;

diff --git a/libunwind/src/CompartmentInfo.hpp b/libunwind/src/CompartmentInfo.hpp
@@ -13,21 +13,52 @@
 #ifndef __COMPARTMENT_INFO_HPP__
 #define __COMPARTMENT_INFO_HPP__
 
+#include <link.h>
+
 namespace libunwind {
-class _LIBUNWIND_HIDDEN CompartmentInfo {
-public:
-#if defined(__CHERI_PURE_CAPABILITY__) && defined(_LIBUNWIND_CHERI_C18N_SUPPORT)
-  static CompartmentInfo sThisCompartmentInfo;
-  // Per-architecture trusted stack frame layout.
-#if defined(_LIBUNWIND_TARGET_AARCH64)
-  static const uint32_t kNewSPOffset = 12 * sizeof(void *);
-  static const uint32_t kNextOffset = 14 * sizeof(void *);
-  static const uint32_t kCalleeSavedOffset = 2 * sizeof(void *);
-  static const uint32_t kCalleeSavedCount = 10;
-  static const uint32_t kReturnAddressOffset = 15 * sizeof(void *) + 8;
-  static const uint32_t kPCOffset = sizeof(void *);
-#endif // _LIBUNWIND_TARGET_AARCH64
-#endif // __CHERI_PURE_CAPABILITY__ && _LIBUNWIND_CHERI_C18N_SUPPORT
+
+// A wrapper for RTLD APIs related to library-based compartmentalisation (c18n).
+template <typename A, typename R>
+struct CompartmentInfo {
+  typedef typename A::pint_t pint_t;
+
+  static int unwindIfAtBoundary(R &registers) {
+#ifdef _LIBUNWIND_HAS_CHERI_LIB_C18N
+#ifdef _LIBUNWIND_TARGET_AARCH64
+    struct dl_c18n_compart_state state;
+    pint_t pc = registers.getIP();
+    pint_t tf = registers.getTrustedStack();
+
+    if (!dl_c18n_is_trampoline(pc, (void *)tf))
+      return UNW_STEP_SUCCESS;
+
+    CHERI_DBG("COMPARTMENT BOUNDARY %#p\n", (void *)pc);
+
+    tf = (pint_t)dl_c18n_pop_trusted_stack(&state, (void *)tf);
+
+    registers.setTrustedStack(tf);
+    CHERI_DBG("C18N: SET TRUSTED STACK %#p\n", (void *)tf);
+
+    registers.setFP((pint_t)state.fp);
+    CHERI_DBG("C18N: SET FP %#p\n", state.fp);
+
+    registers.setSP((pint_t)state.sp);
+    CHERI_DBG("C18N: SET SP: %#p\n", state.sp);
+
+    registers.setIP((pint_t)state.pc);
+    CHERI_DBG("C18N: SET IP: %#p\n", state.pc);
+
+    for (size_t i = 0; i < sizeof(state.regs) / sizeof(*state.regs); ++i) {
+      registers.setCapabilityRegister(UNW_ARM64_C19 + i, (pint_t)state.regs[i]);
+      CHERI_DBG("C18N: SET REGISTER: %lu (%s): %#p\n",
+                UNW_ARM64_C19 + i,
+                registers.getRegisterName(UNW_ARM64_C19 + i),
+                state.regs[i]);
+    }
+#endif
+#endif // _LIBUNWIND_HAS_CHERI_LIB_C18N
+    return UNW_STEP_SUCCESS;
+  }
 };
 } // namespace libunwind
 #endif // __COMPARTMENT_INFO_HPP__
diff --git a/libunwind/src/DwarfInstructions.hpp b/libunwind/src/DwarfInstructions.hpp
@@ -20,7 +20,6 @@
 #include "Registers.hpp"
 #include "DwarfParser.hpp"
 #include "config.h"
-#include "CompartmentInfo.hpp"
 
 
 namespace libunwind {
@@ -55,14 +54,6 @@ class DwarfInstructions {
   typedef typename CFI_Parser<A>::FDE_Info          FDE_Info;
   typedef typename CFI_Parser<A>::CIE_Info          CIE_Info;
 
-#if defined(__CHERI_PURE_CAPABILITY__) && defined(_LIBUNWIND_CHERI_C18N_SUPPORT)
-  static pint_t restoreRegistersFromSandbox(pint_t csp, A &addressSpace,
-                                            R &newRegisters,
-                                            CompartmentInfo &CI, pint_t sealer);
-  static bool isCompartmentTransitionTrampoline(pint_t ecsp, A &addressSpace,
-                                                CompartmentInfo &CI,
-                                                pint_t returnAddress);
-#endif
   static pint_t evaluateExpression(pint_t expression, A &addressSpace,
                                    const R &registers,
                                    pint_t initialStackValue);
@@ -255,75 +246,6 @@ bool DwarfInstructions<A, R>::getRA_SIGN_STATE(A &addressSpace, R registers,
 }
 #endif
 
-#if defined(__CHERI_PURE_CAPABILITY__) && defined(_LIBUNWIND_CHERI_C18N_SUPPORT)
-#if defined(_LIBUNWIND_TARGET_AARCH64)
-template <typename A, typename R>
-typename A::pint_t DwarfInstructions<A, R>::restoreRegistersFromSandbox(
-    pint_t csp, A &addressSpace, R &newRegisters, CompartmentInfo &CI,
-    pint_t sealer) {
-  // Get the unsealed executive CSP
-  assert(__builtin_cheri_tag_get((void *)csp) &&
-         "Executive stack should be tagged!");
-  // Derive the new executive CSP
-  pint_t nextCSP = addressSpace.getP(csp + CI.kNextOffset);
-  // Seal ECSP
-  nextCSP = __builtin_cheri_seal(nextCSP, sealer);
-  assert(__builtin_cheri_tag_get((void *)nextCSP) &&
-         "Next executive stack should be tagged!");
-  CHERI_DBG("SANDBOX: SETTING EXECUTIVE CSP %#p\n", (void *)nextCSP);
-  newRegisters.setTrustedStack(nextCSP);
-  // Restore the next RCSP
-  pint_t nextRCSP = addressSpace.getP(csp + CI.kNewSPOffset);
-  newRegisters.setSP(nextRCSP);
-  CHERI_DBG("SANDBOX: SETTING RESTRICTED CSP: %#p\n",
-            (void *)newRegisters.getSP());
-  // Restore callee-saved registers
-  // Restore: c19-c28
-  for (size_t i = 0, offset = CI.kCalleeSavedOffset; i < CI.kCalleeSavedCount;
-       ++i, offset += sizeof(void *)) {
-    pint_t regValue = addressSpace.getP(csp + offset);
-    newRegisters.setCapabilityRegister(UNW_ARM64_C19 + i, regValue);
-    CHERI_DBG("SETTING CALLEE SAVED CAPABILITY REGISTER: %lu (%s): %#p "
-              "(offset=%zu)\n",
-              UNW_ARM64_C19 + i,
-              newRegisters.getRegisterName(UNW_ARM64_C19 + i), (void *)regValue,
-              offset);
-  }
-  // Restore the frame pointer
-  pint_t newFP = addressSpace.getP(csp);
-  CHERI_DBG("SANDBOX: SETTING CFP %#p\n", (void *)newFP);
-  newRegisters.setFP(newFP);
-  // Get the new return address.
-  return addressSpace.getP(csp + CI.kPCOffset);
-}
-
-template <typename A, typename R>
-bool DwarfInstructions<A, R>::isCompartmentTransitionTrampoline(
-    pint_t ecsp, A &addressSpace, CompartmentInfo &CI, pint_t returnAddress) {
-  ptraddr_t expectedReturnAddress =
-      addressSpace.template get<ptraddr_t>(ecsp + CI.kReturnAddressOffset);
-  CHERI_DBG(
-      "isCompartmentTransitionTrampoline(): expectedReturnAddress: 0x%lx\n",
-      expectedReturnAddress);
-  return expectedReturnAddress == returnAddress;
-}
-#else // _LIBUNWIND_TARGET_AARCH64
-template <typename A, typename R>
-typename A::pint_t DwarfInstructions<A, R>::restoreRegistersFromSandbox(
-    pint_t csp, A &addressSpace, R &newRegisters, CompartmentInfo &CI,
-    pint_t sealer) {
-  assert(0 && "not implemented on this architecture");
-  return (pint_t)0;
-}
-template <typename A, typename R>
-bool DwarfInstructions<A, R>::isCompartmentTransitionTrampoline(
-    pint_t ecsp, A &addressSpace, CompartmentInfo &CI, pint_t returnAddress) {
-  assert(0 && "not implemented on this architecture");
-  return false;
-}
-#endif // _LIBUNWIND_TARGET_AARCH64
-#endif // __CHERI_PURE_CAPABILITY__ && _LIBUNWIND_CHERI_C18N_SUPPORT
-
 template <typename A, typename R>
 int DwarfInstructions<A, R>::stepWithDwarf(A &addressSpace, pc_t pc,
                                            pint_t fdeStart, R &registers,
@@ -483,28 +405,6 @@ int DwarfInstructions<A, R>::stepWithDwarf(A &addressSpace, pc_t pc,
       }
 #endif
 
-#if defined(__CHERI_PURE_CAPABILITY__) && defined(_LIBUNWIND_CHERI_C18N_SUPPORT)
-      // If the sealer is not valid (only the case when we're running without
-      // c18n), check if the return address has the executive mode bit set.
-      // If so, we should be calling into the c18n RTLD as this is a
-      // compartment boundary. We need to restore registers from the executive
-      // stack and ask rtld for it.
-      uintptr_t sealer = addressSpace.getUnwindSealer();
-      if (addressSpace.isValidSealer(sealer)) {
-        pint_t csp = registers.getTrustedStack();
-        if (__builtin_cheri_sealed_get(csp))
-          csp = __builtin_cheri_unseal(csp, sealer);
-        CompartmentInfo &CI = CompartmentInfo::sThisCompartmentInfo;
-        if (csp != 0 && isCompartmentTransitionTrampoline(csp, addressSpace, CI,
-                                                          returnAddress)) {
-          CHERI_DBG("%#p: detected a trampoline, unwinding from sandbox\n",
-                    (void *)returnAddress);
-          returnAddress = restoreRegistersFromSandbox(
-              csp, addressSpace, newRegisters, CI, sealer);
-        }
-      }
-#endif
-
       // Return address is address after call site instruction, so setting IP to
       // that does simualates a return.
       newRegisters.setIP(returnAddress);

diff --git a/libunwind/src/Registers.hpp b/libunwind/src/Registers.hpp
@@ -1874,10 +1874,10 @@ class _LIBUNWIND_HIDDEN Registers_arm64 {
   void       setIP(uintptr_t value) { _registers.__pc = value; }
   uintptr_t  getFP() const          { return _registers.__fp; }
   void       setFP(uintptr_t value) { _registers.__fp = value; }
-#if defined(__CHERI_PURE_CAPABILITY__) && defined(_LIBUNWIND_CHERI_C18N_SUPPORT)
+#ifdef __CHERI_PURE_CAPABILITY__
   uintptr_t  getTrustedStack() const { return _registers.__ecsp; }
   void       setTrustedStack(uintptr_t value) { _registers.__ecsp = value; }
-#endif // __CHERI_PURE_CAPABILITY__ && _LIBUNWIND_CHERI_C18N_SUPPORT
+#endif
 
 private:
   struct GPRs {

diff --git a/libunwind/src/UnwindCursor.hpp b/libunwind/src/UnwindCursor.hpp
@@ -86,6 +86,7 @@ extern "C" _Unwind_Reason_Code __libunwind_seh_personality(
 #include "Registers.hpp"
 #include "RWMutex.hpp"
 #include "Unwind-EHABI.h"
+#include "CompartmentInfo.hpp"
 
 namespace libunwind {
 
@@ -1294,6 +1295,10 @@ class UnwindCursor : public AbstractUnwindCursor{
   }
 #endif // defined(_LIBUNWIND_SUPPORT_TBTAB_UNWIND)
 
+  int stepThroughIfAtC18NBoundary() {
+    return CompartmentInfo<A, R>::unwindIfAtBoundary(_registers);
+  }
+
   A               &_addressSpace;
   R                _registers;
   unw_proc_info_t  _info;
@@ -2867,6 +2872,11 @@ int UnwindCursor<A, R>::step() {
 #endif
   }
 
+  // If we are at a compartment boundary, step through it by asking RTLD to
+  // restore registers from the trusted stack.
+  if (result == UNW_STEP_SUCCESS)
+    result = stepThroughIfAtC18NBoundary();
+
   // update info based on new PC
   if (result == UNW_STEP_SUCCESS) {
     this->setInfoBasedOnIPRegister(true);

diff --git a/libunwind/src/UnwindRegistersRestore.S b/libunwind/src/UnwindRegistersRestore.S
@@ -703,25 +703,6 @@ Lnovec:
 
 #elif defined(__aarch64__)
 
-//
-// extern "C" void __rtld_unw_setcontext(void *c0, void *c1,
-//                                       void *rcsp, void **sealed_ecsp);
-//
-#if defined(__CHERI_PURE_CAPABILITY__)
-DEFINE_LIBUNWIND_FUNCTION(__rtld_unw_setcontext)
-  mov    c16, c2
-  ldp    c2, c3,  [c3, #(-0x210 + 0x20)]
-  mov    csp, c16
-#ifdef __ARM_MORELLO_PURECAP_BENCHMARK_ABI
-  and    x30, x30, #~1
-  ret    x30
-#else
-  ret
-#endif
-END_LIBUNWIND_FUNCTION(__rtld_unw_setcontext)
-WEAK_ALIAS(__rtld_unw_setcontext, _rtld_unw_setcontext)
-#endif
-
 //
 // extern "C" void __libunwind_Registers_arm64_jumpto(Registers_arm64 *);
 //
@@ -731,8 +712,18 @@ WEAK_ALIAS(__rtld_unw_setcontext, _rtld_unw_setcontext)
   .p2align 2
 DEFINE_LIBUNWIND_FUNCTION(__libunwind_Registers_arm64_jumpto)
 #ifdef __CHERI_PURE_CAPABILITY__
+#ifdef _LIBUNWIND_HAS_CHERI_LIB_C18N
+  // Preserve the argument in a callee-saved register instead of pushing it onto
+  // the stack because stack unwinding will switch the stack.
+  mov    c19, c0
+  ldr    c0,      [c19, #0x1f0]  // Pass the target untrusted stack pointer
+  ldr    c1,      [c19, #0x210]  // Pass the target trusted stack pointer
+  bl     dl_c18n_unwind_trusted_stack
+  mov    c0, c19
+#endif
+
   // skip restore of c0,c1 for now
-  // also skip restoring c2 and c3 because they will get clobbered later on
+  ldp    c2, c3,  [c0, #0x020]
   ldp    c4, c5,  [c0, #0x040]
   ldp    c6, c7,  [c0, #0x060]
   ldp    c8, c9,  [c0, #0x080]
@@ -772,17 +763,14 @@ DEFINE_LIBUNWIND_FUNCTION(__libunwind_Registers_arm64_jumpto)
   // context struct, because it is allocated on the stack, and an exception
   // could clobber the de-allocated portion of the stack after csp has been
   // restored.
-  ldr    c2,      [c0, #0x1f0]
-  add    c3, c0,  #0x210
-  ldp    c0, c1,  [c0, #0x000]
-  // XXX: variant PCS is not yet supported by rtld, work around it
-  // using a function pointer.
-  adrp   c16,     :got:_rtld_unw_setcontext
-  ldr    c16,     [c16, :got_lo12:_rtld_unw_setcontext]
+  ldr    c16,     [c0, #0x1f0]
+  ldp    c0, c1,  [c0, #0x000]  // restore c0,c1
+  mov    csp,c16                // restore csp
 #ifdef __ARM_MORELLO_PURECAP_BENCHMARK_ABI
-  br     x16
+  and    x30, x30, #~1
+  ret    x30                    // jump to pc
 #else
-  br     c16
+  ret                           // jump to pcc
 #endif
 #else
   // skip restore of x0,x1 for now