Author: Nerixyz Date: 2026-05-07T12:15:16+02:00 New Revision: 605feeda1e50aa0064947e64d66d3351b9f9693e
URL: https://github.com/llvm/llvm-project/commit/605feeda1e50aa0064947e64d66d3351b9f9693e DIFF: https://github.com/llvm/llvm-project/commit/605feeda1e50aa0064947e64d66d3351b9f9693e.diff LOG: [lldb] Trim and show embedded zeros in `charN_t` arrays (#195514) When formatting `char[N]` (N > 0), the read buffer is limited to the array size and trailing zeros are trimmed. `charN_t[N]` was treated like a `charN_t *` and the formatter read until the first zero byte. If the array doesn't have any zero bytes in it, this will read too much. If the array has zero bytes in it, it will show too little. With this PR, `wchar_t[N]` and `charN_t[N]` are printed like `char[N]`. Added: Modified: lldb/include/lldb/DataFormatters/StringPrinter.h lldb/source/DataFormatters/StringPrinter.cpp lldb/source/Plugins/Language/CPlusPlus/CxxStringTypes.cpp lldb/source/Plugins/Language/ObjC/NSString.cpp lldb/source/ValueObject/ValueObject.cpp lldb/test/API/lang/cpp/char1632_t/TestChar1632T.py lldb/test/API/lang/cpp/char1632_t/main.cpp lldb/test/API/lang/cpp/char8_t/TestCxxChar8_t.py lldb/test/API/lang/cpp/char8_t/main.cpp lldb/test/API/lang/cpp/wchar_t/TestCxxWCharT.py lldb/test/API/lang/cpp/wchar_t/main.cpp lldb/unittests/DataFormatter/StringPrinterTests.cpp Removed: ################################################################################ diff --git a/lldb/include/lldb/DataFormatters/StringPrinter.h b/lldb/include/lldb/DataFormatters/StringPrinter.h index 4169f53e63f38..0a674b9fb303e 100644 --- a/lldb/include/lldb/DataFormatters/StringPrinter.h +++ b/lldb/include/lldb/DataFormatters/StringPrinter.h @@ -26,6 +26,15 @@ class StringPrinter { enum class EscapeStyle { CXX, Swift }; + enum class ZeroTermination { + /// Don't look for a terminator - print the whole buffer. + Ignore, + /// Stop printing at the first zero terminator. + ZeroTerminate, + /// Print embedded zeros, but ignore zeros at the end of the buffer. + TrimTrailingZeros, + }; + class DumpToStreamOptions { public: DumpToStreamOptions() = default; @@ -54,13 +63,9 @@ class StringPrinter { uint32_t GetSourceSize() const { return m_source_size; } - void SetNeedsZeroTermination(bool z) { m_needs_zero_termination = z; } - - bool GetNeedsZeroTermination() const { return m_needs_zero_termination; } - - void SetBinaryZeroIsTerminator(bool e) { m_zero_is_terminator = e; } + void SetZeroTermination(ZeroTermination z) { m_zero_termination = z; } - bool GetBinaryZeroIsTerminator() const { return m_zero_is_terminator; } + ZeroTermination GetZeroTermination() const { return m_zero_termination; } void SetEscapeNonPrintables(bool e) { m_escape_non_printables = e; } @@ -85,18 +90,15 @@ class StringPrinter { char m_quote = '"'; /// The length of the memory region that should be dumped in bytes. uint32_t m_source_size = 0; - bool m_needs_zero_termination = true; /// True iff non-printable characters should be escaped when dumping /// them to the stream. bool m_escape_non_printables = true; /// True iff the max-string-summary-length setting of the target should /// be ignored. bool m_ignore_max_length = false; - /// True iff a zero bytes ('\0') should terminate the memory region that - /// is being dumped. - bool m_zero_is_terminator = true; /// The language-specific style for escaping special characters. EscapeStyle m_escape_style = EscapeStyle::CXX; + ZeroTermination m_zero_termination = ZeroTermination::ZeroTerminate; }; class ReadStringAndDumpToStreamOptions : public DumpToStreamOptions { diff --git a/lldb/source/DataFormatters/StringPrinter.cpp b/lldb/source/DataFormatters/StringPrinter.cpp index 60cb0fc5d6876..3a6a55e9d4e02 100644 --- a/lldb/source/DataFormatters/StringPrinter.cpp +++ b/lldb/source/DataFormatters/StringPrinter.cpp @@ -279,9 +279,11 @@ static bool DumpEncodedBufferToStream( (const SourceDataType *)data.GetDataStart(); const SourceDataType *data_end_ptr = data_ptr + source_size; - const bool zero_is_terminator = dump_options.GetBinaryZeroIsTerminator(); + switch (dump_options.GetZeroTermination()) { + case StringPrinter::ZeroTermination::Ignore: + break; - if (zero_is_terminator) { + case StringPrinter::ZeroTermination::ZeroTerminate: { while (data_ptr < data_end_ptr) { if (!*data_ptr) { data_end_ptr = data_ptr; @@ -291,7 +293,19 @@ static bool DumpEncodedBufferToStream( } data_ptr = (const SourceDataType *)data.GetDataStart(); + } break; + + case StringPrinter::ZeroTermination::TrimTrailingZeros: { + while (data_end_ptr != data_ptr) { + if (*(data_end_ptr - 1)) + break; + data_end_ptr--; + } + } break; } + const bool zero_is_terminator = + dump_options.GetZeroTermination() == + StringPrinter::ZeroTermination::ZeroTerminate; lldb::WritableDataBufferSP utf8_data_buffer_sp; llvm::UTF8 *utf8_data_ptr = nullptr; @@ -384,7 +398,7 @@ lldb_private::formatters::StringPrinter::ReadBufferAndDumpToStreamOptions:: SetSuffixToken(options.GetSuffixToken()); SetQuote(options.GetQuote()); SetEscapeNonPrintables(options.GetEscapeNonPrintables()); - SetBinaryZeroIsTerminator(options.GetBinaryZeroIsTerminator()); + SetZeroTermination(options.GetZeroTermination()); SetEscapeStyle(options.GetEscapeStyle()); } @@ -419,7 +433,8 @@ static bool ReadEncodedBufferAndDumpToStream( if (origin_encoding != 8 && !ConvertFunction) return false; - bool needs_zero_terminator = options.GetNeedsZeroTermination(); + bool needs_zero_terminator = options.GetZeroTermination() == + StringPrinter::ZeroTermination::ZeroTerminate; bool is_truncated = false; const auto max_size = target_sp->GetMaximumSizeOfStringSummary(); @@ -427,8 +442,8 @@ static bool ReadEncodedBufferAndDumpToStream( uint32_t sourceSize; if (elem_type == StringElementType::ASCII && !options.GetSourceSize()) { // FIXME: The NSString formatter sets HasSourceSize(true) when the size is - // actually unknown, as well as SetBinaryZeroIsTerminator(false). IIUC the - // C++ formatter also sets SetBinaryZeroIsTerminator(false) when it doesn't + // actually unknown, as well as SetZeroTermination(Ignore). IIUC the + // C++ formatter also sets SetZeroTermination(Ignore) when it doesn't // mean to. I don't see how this makes sense: we should fix the formatters. // // Until then, the behavior that's expected for ASCII strings with unknown @@ -479,9 +494,10 @@ static bool ReadEncodedBufferAndDumpToStream( target_sp->GetArchitecture().GetAddressByteSize())); dump_options.SetSourceSize(sourceSize); dump_options.SetIsTruncated(is_truncated); - dump_options.SetNeedsZeroTermination(needs_zero_terminator); - if (needs_zero_terminator) - dump_options.SetBinaryZeroIsTerminator(true); + if (needs_zero_terminator) { + dump_options.SetZeroTermination( + StringPrinter::ZeroTermination::ZeroTerminate); + } GetPrintableElementType print_style = (elem_type == StringElementType::ASCII) ? GetPrintableElementType::ASCII diff --git a/lldb/source/Plugins/Language/CPlusPlus/CxxStringTypes.cpp b/lldb/source/Plugins/Language/CPlusPlus/CxxStringTypes.cpp index bf8c393445908..347c108a9a72a 100644 --- a/lldb/source/Plugins/Language/CPlusPlus/CxxStringTypes.cpp +++ b/lldb/source/Plugins/Language/CPlusPlus/CxxStringTypes.cpp @@ -60,6 +60,15 @@ static bool CharStringSummaryProvider(ValueObject &valobj, Stream &stream) { options.SetStream(&stream); options.SetPrefixToken(getElementTraits(ElemType).first); + CompilerType ty = valobj.GetCompilerType(); + uint64_t size = 0; + if (ty.IsArrayType(nullptr, &size) && size > 0) { + options.SetSourceSize(size); + options.SetHasSourceSize(true); + options.SetZeroTermination( + StringPrinter::ZeroTermination::TrimTrailingZeros); + } + if (!StringPrinter::ReadStringAndDumpToStream<ElemType>(options)) stream.Printf("Summary Unavailable"); @@ -89,7 +98,7 @@ static bool CharSummaryProvider(ValueObject &valobj, Stream &stream) { options.SetPrefixToken(ElemTraits.first); options.SetQuote('\''); options.SetSourceSize(1); - options.SetBinaryZeroIsTerminator(false); + options.SetZeroTermination(StringPrinter::ZeroTermination::Ignore); return StringPrinter::ReadBufferAndDumpToStream<ElemType>(options); } @@ -127,6 +136,15 @@ bool lldb_private::formatters::WCharStringSummaryProvider( options.SetStream(&stream); options.SetPrefixToken("L"); + CompilerType ty = valobj.GetCompilerType(); + uint64_t arr_size = 0; + if (ty.IsArrayType(nullptr, &arr_size) && arr_size > 0) { + options.SetSourceSize(arr_size); + options.SetHasSourceSize(true); + options.SetZeroTermination( + StringPrinter::ZeroTermination::TrimTrailingZeros); + } + switch (wchar_size) { case 1: return StringPrinter::ReadStringAndDumpToStream<StringElementType::UTF8>( @@ -180,7 +198,7 @@ bool lldb_private::formatters::WCharSummaryProvider( options.SetPrefixToken("L"); options.SetQuote('\''); options.SetSourceSize(1); - options.SetBinaryZeroIsTerminator(false); + options.SetZeroTermination(StringPrinter::ZeroTermination::Ignore); switch (wchar_size) { case 1: @@ -247,7 +265,7 @@ bool lldb_private::formatters::StringBufferSummaryProvider( options.SetPrefixToken(prefix_token); options.SetQuote('"'); options.SetSourceSize(size); - options.SetBinaryZeroIsTerminator(false); + options.SetZeroTermination(StringPrinter::ZeroTermination::Ignore); return StringPrinter::ReadBufferAndDumpToStream<element_type>(options); } diff --git a/lldb/source/Plugins/Language/ObjC/NSString.cpp b/lldb/source/Plugins/Language/ObjC/NSString.cpp index ce268ac7f3257..7a295119bd031 100644 --- a/lldb/source/Plugins/Language/ObjC/NSString.cpp +++ b/lldb/source/Plugins/Language/ObjC/NSString.cpp @@ -151,10 +151,9 @@ bool lldb_private::formatters::NSStringSummaryProvider( options.SetQuote('"'); options.SetSourceSize(explicit_length); options.SetHasSourceSize(has_explicit_length); - options.SetNeedsZeroTermination(false); + options.SetZeroTermination(StringPrinter::ZeroTermination::Ignore); options.SetIgnoreMaxLength(summary_options.GetCapping() == TypeSummaryCapping::eTypeSummaryUncapped); - options.SetBinaryZeroIsTerminator(false); return StringPrinter::ReadStringAndDumpToStream< StringPrinter::StringElementType::UTF16>(options); } else { @@ -163,10 +162,9 @@ bool lldb_private::formatters::NSStringSummaryProvider( options.SetStream(&stream); options.SetSourceSize(explicit_length); options.SetHasSourceSize(has_explicit_length); - options.SetNeedsZeroTermination(false); + options.SetZeroTermination(StringPrinter::ZeroTermination::Ignore); options.SetIgnoreMaxLength(summary_options.GetCapping() == TypeSummaryCapping::eTypeSummaryUncapped); - options.SetBinaryZeroIsTerminator(false); return StringPrinter::ReadStringAndDumpToStream< StringPrinter::StringElementType::ASCII>(options); } @@ -202,10 +200,12 @@ bool lldb_private::formatters::NSStringSummaryProvider( options.SetQuote('"'); options.SetSourceSize(explicit_length); options.SetHasSourceSize(has_explicit_length); - options.SetNeedsZeroTermination(!has_explicit_length); + if (has_explicit_length) + options.SetZeroTermination(StringPrinter::ZeroTermination::Ignore); + else + options.SetZeroTermination(StringPrinter::ZeroTermination::ZeroTerminate); options.SetIgnoreMaxLength(summary_options.GetCapping() == TypeSummaryCapping::eTypeSummaryUncapped); - options.SetBinaryZeroIsTerminator(!has_explicit_length); return StringPrinter::ReadStringAndDumpToStream< StringPrinter::StringElementType::UTF16>(options); } else if (is_path_store) { @@ -228,10 +228,12 @@ bool lldb_private::formatters::NSStringSummaryProvider( options.SetQuote('"'); options.SetSourceSize(explicit_length); options.SetHasSourceSize(has_explicit_length); - options.SetNeedsZeroTermination(!has_explicit_length); + if (has_explicit_length) + options.SetZeroTermination(StringPrinter::ZeroTermination::Ignore); + else + options.SetZeroTermination(StringPrinter::ZeroTermination::ZeroTerminate); options.SetIgnoreMaxLength(summary_options.GetCapping() == TypeSummaryCapping::eTypeSummaryUncapped); - options.SetBinaryZeroIsTerminator(!has_explicit_length); return StringPrinter::ReadStringAndDumpToStream< StringPrinter::StringElementType::UTF16>(options); } else if (is_inline) { @@ -250,10 +252,12 @@ bool lldb_private::formatters::NSStringSummaryProvider( options.SetStream(&stream); options.SetSourceSize(explicit_length); options.SetHasSourceSize(has_explicit_length); - options.SetNeedsZeroTermination(!has_explicit_length); + if (has_explicit_length) + options.SetZeroTermination(StringPrinter::ZeroTermination::Ignore); + else + options.SetZeroTermination(StringPrinter::ZeroTermination::ZeroTerminate); options.SetIgnoreMaxLength(summary_options.GetCapping() == TypeSummaryCapping::eTypeSummaryUncapped); - options.SetBinaryZeroIsTerminator(!has_explicit_length); if (has_explicit_length) return StringPrinter::ReadStringAndDumpToStream< StringPrinter::StringElementType::UTF8>(options); diff --git a/lldb/source/ValueObject/ValueObject.cpp b/lldb/source/ValueObject/ValueObject.cpp index d8ca5aad7d399..802790c09834c 100644 --- a/lldb/source/ValueObject/ValueObject.cpp +++ b/lldb/source/ValueObject/ValueObject.cpp @@ -1409,7 +1409,13 @@ bool ValueObject::DumpPrintableRepresentation( options.SetQuote('"'); options.SetSourceSize(buffer_sp->GetByteSize()); options.SetIsTruncated(read_string.second); - options.SetBinaryZeroIsTerminator(custom_format != eFormatVectorOfChar); + if (custom_format == eFormatVectorOfChar) { + options.SetZeroTermination( + formatters::StringPrinter::ZeroTermination::Ignore); + } else { + options.SetZeroTermination( + formatters::StringPrinter::ZeroTermination::ZeroTerminate); + } formatters::StringPrinter::ReadBufferAndDumpToStream< lldb_private::formatters::StringPrinter::StringElementType::ASCII>( options); diff --git a/lldb/test/API/lang/cpp/char1632_t/TestChar1632T.py b/lldb/test/API/lang/cpp/char1632_t/TestChar1632T.py index e2521d88258bc..cd4dc9a406f5c 100644 --- a/lldb/test/API/lang/cpp/char1632_t/TestChar1632T.py +++ b/lldb/test/API/lang/cpp/char1632_t/TestChar1632T.py @@ -80,6 +80,12 @@ def test(self): substrs=['u"ﺸﺵۻ"', 'U"ЕЙРГЖО"'], ) + # Check that embedded zeros show up in arrays + self.expect_var_path("aZero16", summary='u"I\\0have\\0zeros"') + self.expect_var_path("cZero16", summary='u"I"') + self.expect_var_path("aZero32", summary='U"I\\0have\\0zeros"') + self.expect_var_path("cZero32", summary='U"I"') + self.runCmd("next") # step to after the string is nullified # check that we don't crash on NULL diff --git a/lldb/test/API/lang/cpp/char1632_t/main.cpp b/lldb/test/API/lang/cpp/char1632_t/main.cpp index dc37df27fceb5..e38188cf85dee 100644 --- a/lldb/test/API/lang/cpp/char1632_t/main.cpp +++ b/lldb/test/API/lang/cpp/char1632_t/main.cpp @@ -10,7 +10,7 @@ void copy_char_seq (T (&arr)[N], const T* src) assert(src_len < N); std::char_traits<T>::copy(arr, src, src_len); - arr[src_len] = 0; + memset(&arr[src_len], 0, (N - src_len) * sizeof(T)); } int main (int argc, char const *argv[]) @@ -25,6 +25,12 @@ int main (int argc, char const *argv[]) char32_t *s32 = (char32_t *)U"ЕЙРГЖО"; copy_char_seq(as16, s16); copy_char_seq(as32, s32); + + char16_t aZero16[32] = u"I\0have\0zeros"; + const char16_t *cZero16 = u"I\0have\0zeros"; + char32_t aZero32[32] = U"I\0have\0zeros"; + const char32_t *cZero32 = U"I\0have\0zeros"; + s32 = nullptr; // breakpoint1 s32 = (char32_t *)U"෴"; s16 = (char16_t *)u"色ハ匂ヘト散リヌルヲ"; diff --git a/lldb/test/API/lang/cpp/char8_t/TestCxxChar8_t.py b/lldb/test/API/lang/cpp/char8_t/TestCxxChar8_t.py index 08f09b317b217..c727ae9cf6ab6 100644 --- a/lldb/test/API/lang/cpp/char8_t/TestCxxChar8_t.py +++ b/lldb/test/API/lang/cpp/char8_t/TestCxxChar8_t.py @@ -21,10 +21,18 @@ def test_without_process(self): self.expect("target variable a", substrs=["char8_t", "0x61 u8'a'"]) self.expect("target variable ab", substrs=["const char8_t *", 'u8"你好"']) self.expect("target variable abc", substrs=["char8_t[9]", 'u8"你好"']) + self.expect( + "target variable aZero", substrs=["char8_t[32]", 'u8"I\\0have\\0zeros"'] + ) + self.expect("target variable cZero", substrs=["const char8_t *", 'u8"I"']) self.expect_expr("a", result_type="char8_t", result_summary="0x61 u8'a'") self.expect_expr("ab", result_type="const char8_t *", result_summary='u8"你好"') self.expect_expr("abc", result_type="char8_t[9]", result_summary='u8"你好"') + self.expect_expr( + "aZero", result_type="char8_t[32]", result_summary='u8"I\\0have\\0zeros"' + ) + self.expect_expr("cZero", result_type="const char8_t *", result_summary='u8"I"') @skipIf(compiler="clang", compiler_version=["<", "7.0"]) def test_with_process(self): @@ -38,3 +46,7 @@ def test_with_process(self): self.expect_expr("a", result_type="char8_t", result_summary="0x61 u8'a'") self.expect_expr("ab", result_type="const char8_t *", result_summary='u8"你好"') self.expect_expr("abc", result_type="char8_t[9]", result_summary='u8"你好"') + self.expect_expr( + "aZero", result_type="char8_t[32]", result_summary='u8"I\\0have\\0zeros"' + ) + self.expect_expr("cZero", result_type="const char8_t *", result_summary='u8"I"') diff --git a/lldb/test/API/lang/cpp/char8_t/main.cpp b/lldb/test/API/lang/cpp/char8_t/main.cpp index d109751cc2314..36a2bbef95e2b 100644 --- a/lldb/test/API/lang/cpp/char8_t/main.cpp +++ b/lldb/test/API/lang/cpp/char8_t/main.cpp @@ -2,6 +2,9 @@ char8_t a = u8'a'; const char8_t* ab = u8"你好"; char8_t abc[9] = u8"你好"; +char8_t aZero[32] = u8"I\0have\0zeros"; +const char8_t *cZero = u8"I\0have\0zeros"; + int main (int argc, char const *argv[]) { return 0; // break here } diff --git a/lldb/test/API/lang/cpp/wchar_t/TestCxxWCharT.py b/lldb/test/API/lang/cpp/wchar_t/TestCxxWCharT.py index 35cb512142ca0..f17d5a707c9e3 100644 --- a/lldb/test/API/lang/cpp/wchar_t/TestCxxWCharT.py +++ b/lldb/test/API/lang/cpp/wchar_t/TestCxxWCharT.py @@ -48,3 +48,6 @@ def test(self): self.expect("frame variable wchar_zero", substrs=["L'\\0'"]) self.expect("expression wchar_zero", substrs=["L'\\0'"]) + + self.expect_var_path("aZero", summary='L"I\\0have\\0zeros"') + self.expect_var_path("cZero", summary='L"I"') diff --git a/lldb/test/API/lang/cpp/wchar_t/main.cpp b/lldb/test/API/lang/cpp/wchar_t/main.cpp index 778642e0133f0..f6031ae9f7b9d 100644 --- a/lldb/test/API/lang/cpp/wchar_t/main.cpp +++ b/lldb/test/API/lang/cpp/wchar_t/main.cpp @@ -19,8 +19,13 @@ int main (int argc, char const *argv[]) const wchar_t *mazeltov = L"מזל טוב"; wchar_t *ws_NULL = nullptr; wchar_t *ws_empty = L""; - wchar_t array[200], * array_source = L"Hey, I'm a super wchar_t string, éõñž"; + wchar_t array[200], + *array_source = L"Hey, I'm a super wchar_t string, éõñž"; wchar_t wchar_zero = (wchar_t)0; - memcpy(array, array_source, 39 * sizeof(wchar_t)); + memcpy(array, array_source, 39 * sizeof(wchar_t)); + + wchar_t aZero[32] = L"I\0have\0zeros"; + const wchar_t *cZero = L"I\0have\0zeros"; + return 0; // break here } diff --git a/lldb/unittests/DataFormatter/StringPrinterTests.cpp b/lldb/unittests/DataFormatter/StringPrinterTests.cpp index ac4116139beb2..64789bccd3e87 100644 --- a/lldb/unittests/DataFormatter/StringPrinterTests.cpp +++ b/lldb/unittests/DataFormatter/StringPrinterTests.cpp @@ -32,7 +32,7 @@ format(StringRef input, StringPrinter::EscapeStyle escape_style) { StringPrinter::ReadBufferAndDumpToStreamOptions opts; opts.SetStream(&out); opts.SetSourceSize(input.size()); - opts.SetNeedsZeroTermination(true); + opts.SetZeroTermination(StringPrinter::ZeroTermination::ZeroTerminate); opts.SetEscapeNonPrintables(true); opts.SetIgnoreMaxLength(false); opts.SetEscapeStyle(escape_style); _______________________________________________ lldb-commits mailing list [email protected] https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits
