Skip to content

Instantly share code, notes, and snippets.

@guneysus
Created November 3, 2022 19:26
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save guneysus/0f4e09fbf463ec6685abc07b2ffb7c0d to your computer and use it in GitHub Desktop.
Save guneysus/0f4e09fbf463ec6685abc07b2ffb7c0d to your computer and use it in GitHub Desktop.
PE/COFF Specification Addendum

PE/COFF Specification Addendum

Deterministic PE/COFF File

PE/COFF file is deterministic if it is produced by a tool that guarantees that the entire content of the file is based solely on documented inputs given to the tool (such as source files, resource files, compiler options, etc.) rather than ambient environment variables (such as the current time, the operating system, the bitness of the process running the tool, etc.). The content of the file can be reproduced exactly (bit for bit) given the same inputs.

The value of field TimeDateStamp in COFF File Header of a deterministic PE/COFF file does not indicate the date and time when the file was produced and should not be interpreted that way. Instead the value of the field is derived from a hash of the file content. The algorithm used to calculate this value is an implementation detail of the tool that produced the file.

Debug Directory

Image files contain an optional debug directory that indicates what form of debug information is present and where it is. This directory consists of an array of debug directory entries whose location and size are indicated in the image optional header.

PE/COFF Specification defines the structure of Debug Directory in section 5.1.1. Each entry has a Type and a pointer to data. The format of the data is specific to the Type of the debug entry and not prescribed by the PE/COFF specification. The following paragraphs specify format of data of debug directory entries produced and consumed by CLI (Common Language Infrastructure) compilers, debuggers and other tools.

CodeView Debug Directory Entry (type 2)

Version Major=0, Minor=0 of the data format:

Offset Size Field Description
0 4 Signature 0x52 0x53 0x44 0x53 (ASCII string: "RSDS")
4 16 Guid GUID (Globally Unique Identifier) of the associated PDB.
20 4 Age Iteration of the PDB. The first iteration is 1. The iteration is incremented each time the PDB content is augmented.
24 Path UTF-8 NUL-terminated path to the associated .pdb file

Guid and Age are used to match PE/COFF image with the associated PDB.

The associated .pdb file may not exist at the path indicated by Path field. If it doesn't the Path, Guid and Age can be used to find the corresponding PDB file locally or on a symbol server. The exact search algorithm used by tools to locate the PDB depends on the tool and its configuration.

If the containing PE/COFF file is deterministic the Guid field above and DateTimeStamp field of the directory entry are calculated deterministically based solely on the content of the associated .pdb file. Otherwise the value of Guid is random and the value of DateTimeStamp indicates the time and date that the debug data was created.

Version Major=any, Minor=0x504d of the data format has the same structure as above. The Age shall be 1. The format of the .pdb file that this PE/COFF file was built with is Portable PDB. The Major version specified in the entry indicates the version of the Portable PDB format. Together 16B of the Guid concatenated with 4B of the TimeDateStamp field of the entry form a PDB ID that should be used to match the PE/COFF image with the associated PDB (instead of Guid and Age). Matching PDB ID is stored in the #Pdb stream of the .pdb file.

A matching PDB may be found whose format is different than the format of the PDB the PE/COFF file was built with. This may happen when the original PDB file is converted to the other format without updating the PE/COFF file. This scenario is fully supported. A tool looking for the associated PDB shall determine the actual format of the found PDB based on the signature at the start of the PDB file. The tool may use the version in CodeView entry as a hint to prefer the original format over the converted one if both are available.

Deterministic Debug Directory Entry (type 16)

The entry doesn't have any data associated with it. All fields of the entry, but Type shall be zero.

Presence of this entry indicates that the containing PE/COFF file is deterministic.

Embedded Portable PDB Debug Directory Entry (type 17)

Declares that debugging information is embedded in the PE file at location specified by PointerToRawData.

Version Major=any, Minor=0x0100 of the data format:

Offset Size Field Description
0 4 Signature 0x4D 0x50 0x44 0x42
4 4 UncompressedSize The size of decompressed Portable PDB image
8 SizeOfData - 8 PortablePdbImage Portable PDB image compressed using Deflate algorithm

If both CodeView and Embedded Portable PDB entries are present then they shall represent the same data.

Note: The reader may prefer to read from the file if it exists. This may be more efficient as it avoids in-memory decompression.

Note: Including both entries enables a tool that does not recognize Embedded Portable PDB entry to locate debug information as long as it is also available in a file specified in CodeView entry. Such file can be created by extracting the embedded Portable PDB image to a separate file.

UncompressedSize shall be greater than 0. Other values are reserved for future use, in case the format of the data changes.

Note: Some tools and APIs only work with the data blob and don't have access to the Debug Directory Entry itself to determine the version of the data format. To enable these tools to adopt new versions of the data blob the UncompressedSize highest bit shall be used to indicate version change and the new version shall be included in the data blob.

The Major version specified in the entry indicates the version of the Portable PDB format. The Minor version indicates the version of the Embedded Portable PDB data format.

The value of Stamp field in the entry shall be 0.

PDB Checksum Debug Directory Entry (type 19)

Stores crypto hash of the content of the symbol file the PE/COFF file was built with.

The hash can be used to validate that a given PDB file was built with the PE/COFF file and not altered in any way.

More than one entry can be present, in case multiple PDBs were produced during the build of the PE/COFF file (e.g. private and public symbols).

Version Major=0x0001, Minor=0x0000 of the entry data format is following:

Offset Size Field Description
0 AlgNameLength AlgorithmName Zero-terminated UTF8 string. The name of the crypho hash algorithm.
AlgNameLength ChecksumSize Checksum Blob. Checksum of the PDB content.

AlgorithmName is the name of the crypto hash algorithm used to calculate the checksum. The name is case-sensitive.

Supported algorithm names include at least:

AlgorithmName ChecksumSize Description
SHA256 32 The 256-bit secure hash algorithm. Standard: FIPS 180-2, FIPS 198.
SHA384 48 The 384-bit secure hash algorithm. Standard: FIPS 180-2, FIPS 198.
SHA512 64 The 512-bit secure hash algorithm. Standard: FIPS 180-2, FIPS 198.

Portable PDB Checksum

If the symbol format is Portable PDB the checksum is calculated by hashing the entire content of the PDB file with the PDB ID set to 0 (20 zeroed bytes).

When validating that Portable PDB matches the debug directory record check that the checksums match and that the PDB ID match the data in the corresponding Portable CodeView record.

Excluding the PDB ID allows the compiler to calculate a single hash of the PDB file content and use it to generate both deterministic PDB ID and PDB checksum.

Windows PDB Checksum

If the symbol format is Windows PDB the checksum is calculated by hashing the entire content of the PDB file with the PDB signature comprising of 16B GUID and 4B timestamp zeroed.

When validating that Windows PDB matches the debug directory record check that the checksums match and that the PDB signature (both GUID and timestamp values) match the data in the corresponding CodeView record.

Note that when the debugger (or other tool) searches for the PDB only GUID and Age fields are used to match the PDB, but the timestamp of the CodeView debug directory entry does not need to match the timestamp stored in the PDB. Therefore, to verify byte-for-byte identity of the PDB, the timestamp field should also be checked.

R2R PerfMap Debug Directory Entry (type 21)

Declares that the image has an associated PerfMap file containing a table mapping symbols to offsets for ready to run compilations.

Version Major=0x0001, Minor=0x0000 of the entry data format is following:

Offset Size Field Description
0 4 Magic 0x52 0x32 0x52 0x4D (ASCII string: "R2RM")
4 16 Signature Byte sequence uniquely identifying the associated PerfMap
20 4 Version Version number of the PerfMap. Currently only version 1 is supported.
24 Path UTF-8 NUL-terminated path to the associated .r2rmap file.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment