1 //===-- dfsan_interface.h -------------------------------------------------===// 2 // 3 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions. 4 // See https://llvm.org/LICENSE.txt for license information. 5 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception 6 // 7 //===----------------------------------------------------------------------===// 8 // 9 // This file is a part of DataFlowSanitizer. 10 // 11 // Public interface header. 12 //===----------------------------------------------------------------------===// 13 #ifndef DFSAN_INTERFACE_H 14 #define DFSAN_INTERFACE_H 15 16 #include <stddef.h> 17 #include <stdint.h> 18 #include <sanitizer/common_interface_defs.h> 19 20 #ifdef __cplusplus 21 extern "C" { 22 #endif 23 24 typedef uint8_t dfsan_label; 25 typedef uint32_t dfsan_origin; 26 27 /// Signature of the callback argument to dfsan_set_write_callback(). 28 typedef void (*dfsan_write_callback_t)(int fd, const void *buf, size_t count); 29 30 /// Computes the union of \c l1 and \c l2, resulting in a union label. 31 dfsan_label dfsan_union(dfsan_label l1, dfsan_label l2); 32 33 /// Sets the label for each address in [addr,addr+size) to \c label. 34 void dfsan_set_label(dfsan_label label, void *addr, size_t size); 35 36 /// Sets the label for each address in [addr,addr+size) to the union of the 37 /// current label for that address and \c label. 38 void dfsan_add_label(dfsan_label label, void *addr, size_t size); 39 40 /// Retrieves the label associated with the given data. 41 /// 42 /// The type of 'data' is arbitrary. The function accepts a value of any type, 43 /// which can be truncated or extended (implicitly or explicitly) as necessary. 44 /// The truncation/extension operations will preserve the label of the original 45 /// value. 46 dfsan_label dfsan_get_label(long data); 47 48 /// Retrieves the immediate origin associated with the given data. The returned 49 /// origin may point to another origin. 50 /// 51 /// The type of 'data' is arbitrary. 52 dfsan_origin dfsan_get_origin(long data); 53 54 /// Retrieves the label associated with the data at the given address. 55 dfsan_label dfsan_read_label(const void *addr, size_t size); 56 57 /// Returns whether the given label label contains the label elem. 58 int dfsan_has_label(dfsan_label label, dfsan_label elem); 59 60 /// Flushes the DFSan shadow, i.e. forgets about all labels currently associated 61 /// with the application memory. Use this call to start over the taint tracking 62 /// within the same process. 63 /// 64 /// Note: If another thread is working with tainted data during the flush, that 65 /// taint could still be written to shadow after the flush. 66 void dfsan_flush(void); 67 68 /// Sets a callback to be invoked on calls to write(). The callback is invoked 69 /// before the write is done. The write is not guaranteed to succeed when the 70 /// callback executes. Pass in NULL to remove any callback. 71 void dfsan_set_write_callback(dfsan_write_callback_t labeled_write_callback); 72 73 /// Interceptor hooks. 74 /// Whenever a dfsan's custom function is called the corresponding 75 /// hook is called it non-zero. The hooks should be defined by the user. 76 /// The primary use case is taint-guided fuzzing, where the fuzzer 77 /// needs to see the parameters of the function and the labels. 78 /// FIXME: implement more hooks. 79 void dfsan_weak_hook_memcmp(void *caller_pc, const void *s1, const void *s2, 80 size_t n, dfsan_label s1_label, 81 dfsan_label s2_label, dfsan_label n_label); 82 void dfsan_weak_hook_strncmp(void *caller_pc, const char *s1, const char *s2, 83 size_t n, dfsan_label s1_label, 84 dfsan_label s2_label, dfsan_label n_label); 85 86 /// Prints the origin trace of the label at the address addr to stderr. It also 87 /// prints description at the beginning of the trace. If origin tracking is not 88 /// on, or the address is not labeled, it prints nothing. 89 void dfsan_print_origin_trace(const void *addr, const char *description); 90 91 /// Prints the origin trace of the label at the address \p addr to a 92 /// pre-allocated output buffer. If origin tracking is not on, or the address is 93 /// not labeled, it prints nothing. 94 /// 95 /// Typical usage: 96 /// \code 97 /// char kDescription[] = "..."; 98 /// char buf[1024]; 99 /// dfsan_sprint_origin_trace(&tainted_var, kDescription, buf, sizeof(buf)); 100 /// \endcode 101 /// 102 /// Typical usage that handles truncation: 103 /// \code 104 /// char buf[1024]; 105 /// int len = dfsan_sprint_origin_trace(&var, nullptr, buf, sizeof(buf)); 106 /// 107 /// if (len < sizeof(buf)) { 108 /// ProcessOriginTrace(buf); 109 /// } else { 110 /// char *tmpbuf = new char[len + 1]; 111 /// dfsan_sprint_origin_trace(&var, nullptr, tmpbuf, len + 1); 112 /// ProcessOriginTrace(tmpbuf); 113 /// delete[] tmpbuf; 114 /// } 115 /// \endcode 116 /// 117 /// \param addr The tainted memory address whose origin we are printing. 118 /// \param description A description printed at the beginning of the trace. 119 /// \param [out] out_buf The output buffer to write the results to. 120 /// \param out_buf_size The size of \p out_buf. 121 /// 122 /// \returns The number of symbols that should have been written to \p out_buf 123 /// (not including trailing null byte '\0'). Thus, the string is truncated iff 124 /// return value is not less than \p out_buf_size. 125 size_t dfsan_sprint_origin_trace(const void *addr, const char *description, 126 char *out_buf, size_t out_buf_size); 127 128 /// Prints the stack trace leading to this call to a pre-allocated output 129 /// buffer. 130 /// 131 /// For usage examples, see dfsan_sprint_origin_trace. 132 /// 133 /// \param [out] out_buf The output buffer to write the results to. 134 /// \param out_buf_size The size of \p out_buf. 135 /// 136 /// \returns The number of symbols that should have been written to \p out_buf 137 /// (not including trailing null byte '\0'). Thus, the string is truncated iff 138 /// return value is not less than \p out_buf_size. 139 size_t dfsan_sprint_stack_trace(char *out_buf, size_t out_buf_size); 140 141 /// Retrieves the very first origin associated with the data at the given 142 /// address. 143 dfsan_origin dfsan_get_init_origin(const void *addr); 144 145 /// Returns the value of -dfsan-track-origins. 146 /// * 0: do not track origins. 147 /// * 1: track origins at memory store operations. 148 /// * 2: track origins at memory load and store operations. 149 int dfsan_get_track_origins(void); 150 #ifdef __cplusplus 151 } // extern "C" 152 153 template <typename T> 154 void dfsan_set_label(dfsan_label label, T &data) { // NOLINT 155 dfsan_set_label(label, (void *)&data, sizeof(T)); 156 } 157 158 #endif 159 160 #endif // DFSAN_INTERFACE_H 161