man5/gitformat-commit-graph.5

'\" t
.\"     Title: gitformat-commit-graph
.\"    Author: [FIXME: author] [see http://www.docbook.org/tdg5/en/html/author]
.\" Generator: DocBook XSL Stylesheets vsnapshot <http://docbook.sf.net/>
.\"      Date: 2024-04-29
.\"    Manual: Git Manual
.\"    Source: Git 2.45.0
.\"  Language: English
.\"
.TH "GITFORMAT\-COMMIT\-GRAPH" "5" "2024\-04\-29" "Git 2\&.45\&.0" "Git Manual"
.\" -----------------------------------------------------------------
.\" * Define some portability stuff
.\" -----------------------------------------------------------------
.\" ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
.\" http://bugs.debian.org/507673
.\" http://lists.gnu.org/archive/html/groff/2009-02/msg00013.html
.\" ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
.ie \n(.g .ds Aq \(aq
.el       .ds Aq '
.\" -----------------------------------------------------------------
.\" * set default formatting
.\" -----------------------------------------------------------------
.\" disable hyphenation
.nh
.\" disable justification (adjust text to left margin only)
.ad l
.\" -----------------------------------------------------------------
.\" * MAIN CONTENT STARTS HERE *
.\" -----------------------------------------------------------------
.SH "NAME"
gitformat-commit-graph \- Git commit\-graph format
.SH "SYNOPSIS"
.sp
.nf
$GIT_DIR/objects/info/commit\-graph
$GIT_DIR/objects/info/commit\-graphs/*
.fi
.sp
.SH "DESCRIPTION"
.sp
The Git commit\-graph stores a list of commit OIDs and some associated metadata, including:
.sp
.RS 4
.ie n \{\
\h'-04'\(bu\h'+03'\c
.\}
.el \{\
.sp -1
.IP \(bu 2.3
.\}
The generation number of the commit\&.
.RE
.sp
.RS 4
.ie n \{\
\h'-04'\(bu\h'+03'\c
.\}
.el \{\
.sp -1
.IP \(bu 2.3
.\}
The root tree OID\&.
.RE
.sp
.RS 4
.ie n \{\
\h'-04'\(bu\h'+03'\c
.\}
.el \{\
.sp -1
.IP \(bu 2.3
.\}
The commit date\&.
.RE
.sp
.RS 4
.ie n \{\
\h'-04'\(bu\h'+03'\c
.\}
.el \{\
.sp -1
.IP \(bu 2.3
.\}
The parents of the commit, stored using positional references within the graph file\&.
.RE
.sp
.RS 4
.ie n \{\
\h'-04'\(bu\h'+03'\c
.\}
.el \{\
.sp -1
.IP \(bu 2.3
.\}
The Bloom filter of the commit carrying the paths that were changed between the commit and its first parent, if requested\&.
.RE
.sp
These positional references are stored as unsigned 32\-bit integers corresponding to the array position within the list of commit OIDs\&. Due to some special constants we use to track parents, we can store at most (1 << 30) + (1 << 29) + (1 << 28) \- 1 (around 1\&.8 billion) commits\&.
.SH "COMMIT\-GRAPH FILES HAVE THE FOLLOWING FORMAT:"
.sp
In order to allow extensions that add extra data to the graph, we organize the body into "chunks" and provide a binary lookup table at the beginning of the body\&. The header includes certain values, such as number of chunks and hash type\&.
.sp
All multi\-byte numbers are in network byte order\&.
.SS "HEADER:"
.sp
.if n \{\
.RS 4
.\}
.nf
4\-byte signature:
    The signature is: {\*(AqC\*(Aq, \*(AqG\*(Aq, \*(AqP\*(Aq, \*(AqH\*(Aq}
.fi
.if n \{\
.RE
.\}
.sp
.if n \{\
.RS 4
.\}
.nf
1\-byte version number:
    Currently, the only valid version is 1\&.
.fi
.if n \{\
.RE
.\}
.sp
.if n \{\
.RS 4
.\}
.nf
1\-byte Hash Version
    We infer the hash length (H) from this value:
      1 => SHA\-1
      2 => SHA\-256
    If the hash type does not match the repository\*(Aqs hash algorithm, the
    commit\-graph file should be ignored with a warning presented to the
    user\&.
.fi
.if n \{\
.RE
.\}
.sp
.if n \{\
.RS 4
.\}
.nf
1\-byte number (C) of "chunks"
.fi
.if n \{\
.RE
.\}
.sp
.if n \{\
.RS 4
.\}
.nf
1\-byte number (B) of base commit\-graphs
    We infer the length (H*B) of the Base Graphs chunk
    from this value\&.
.fi
.if n \{\
.RE
.\}
.SS "CHUNK LOOKUP:"
.sp
.if n \{\
.RS 4
.\}
.nf
(C + 1) * 12 bytes listing the table of contents for the chunks:
    First 4 bytes describe the chunk id\&. Value 0 is a terminating label\&.
    Other 8 bytes provide the byte\-offset in current file for chunk to
    start\&. (Chunks are ordered contiguously in the file, so you can infer
    the length using the next chunk position if necessary\&.) Each chunk
    ID appears at most once\&.
.fi
.if n \{\
.RE
.\}
.sp
.if n \{\
.RS 4
.\}
.nf
The CHUNK LOOKUP matches the table of contents from
the chunk\-based file format, see linkgit:gitformat\-chunk[5]
.fi
.if n \{\
.RE
.\}
.sp
.if n \{\
.RS 4
.\}
.nf
The remaining data in the body is described one chunk at a time, and
these chunks may be given in any order\&. Chunks are required unless
otherwise specified\&.
.fi
.if n \{\
.RE
.\}
.SS "CHUNK DATA:"
.sp
.it 1 an-trap
.nr an-no-space-flag 1
.nr an-break-flag 1
.br
.ps +1
\fBOID Fanout (ID: {O, I, D, F}) (256 * 4 bytes)\fR
.RS 4
.sp
.if n \{\
.RS 4
.\}
.nf
The ith entry, F[i], stores the number of OIDs with first
byte at most i\&. Thus F[255] stores the total
number of commits (N)\&.
.fi
.if n \{\
.RE
.\}
.RE
.sp
.it 1 an-trap
.nr an-no-space-flag 1
.nr an-break-flag 1
.br
.ps +1
\fBOID Lookup (ID: {O, I, D, L}) (N * H bytes)\fR
.RS 4
.sp
.if n \{\
.RS 4
.\}
.nf
The OIDs for all commits in the graph, sorted in ascending order\&.
.fi
.if n \{\
.RE
.\}
.RE
.sp
.it 1 an-trap
.nr an-no-space-flag 1
.nr an-break-flag 1
.br
.ps +1
\fBCommit Data (ID: {C, D, A, T }) (N * (H + 16) bytes)\fR
.RS 4
.sp
.RS 4
.ie n \{\
\h'-04'\(bu\h'+03'\c
.\}
.el \{\
.sp -1
.IP \(bu 2.3
.\}
The first H bytes are for the OID of the root tree\&.
.RE
.sp
.RS 4
.ie n \{\
\h'-04'\(bu\h'+03'\c
.\}
.el \{\
.sp -1
.IP \(bu 2.3
.\}
The next 8 bytes are for the positions of the first two parents of the ith commit\&. Stores value 0x70000000 if no parent in that position\&. If there are more than two parents, the second value has its most\-significant bit on and the other bits store an array position into the Extra Edge List chunk\&.
.RE
.sp
.RS 4
.ie n \{\
\h'-04'\(bu\h'+03'\c
.\}
.el \{\
.sp -1
.IP \(bu 2.3
.\}
The next 8 bytes store the topological level (generation number v1) of the commit and the commit time in seconds since EPOCH\&. The generation number uses the higher 30 bits of the first 4 bytes, while the commit time uses the 32 bits of the second 4 bytes, along with the lowest 2 bits of the lowest byte, storing the 33rd and 34th bit of the commit time\&.
.RE
.RE
.sp
.it 1 an-trap
.nr an-no-space-flag 1
.nr an-break-flag 1
.br
.ps +1
\fBGeneration Data (ID: {G, D, A, 2 }) (N * 4 bytes) [Optional]\fR
.RS 4
.sp
.RS 4
.ie n \{\
\h'-04'\(bu\h'+03'\c
.\}
.el \{\
.sp -1
.IP \(bu 2.3
.\}
This list of 4\-byte values store corrected commit date offsets for the commits, arranged in the same order as commit data chunk\&.
.RE
.sp
.RS 4
.ie n \{\
\h'-04'\(bu\h'+03'\c
.\}
.el \{\
.sp -1
.IP \(bu 2.3
.\}
If the corrected commit date offset cannot be stored within 31 bits, the value has its most\-significant bit on and the other bits store the position of corrected commit date into the Generation Data Overflow chunk\&.
.RE
.sp
.RS 4
.ie n \{\
\h'-04'\(bu\h'+03'\c
.\}
.el \{\
.sp -1
.IP \(bu 2.3
.\}
Generation Data chunk is present only when commit\-graph file is written by compatible versions of Git and in case of split commit\-graph chains, the topmost layer also has Generation Data chunk\&.
.RE
.RE
.sp
.it 1 an-trap
.nr an-no-space-flag 1
.nr an-break-flag 1
.br
.ps +1
\fBGeneration Data Overflow (ID: {G, D, O, 2 }) [Optional]\fR
.RS 4
.sp
.RS 4
.ie n \{\
\h'-04'\(bu\h'+03'\c
.\}
.el \{\
.sp -1
.IP \(bu 2.3
.\}
This list of 8\-byte values stores the corrected commit date offsets for commits with corrected commit date offsets that cannot be stored within 31 bits\&.
.RE
.sp
.RS 4
.ie n \{\
\h'-04'\(bu\h'+03'\c
.\}
.el \{\
.sp -1
.IP \(bu 2.3
.\}
Generation Data Overflow chunk is present only when Generation Data chunk is present and atleast one corrected commit date offset cannot be stored within 31 bits\&.
.RE
.RE
.sp
.it 1 an-trap
.nr an-no-space-flag 1
.nr an-break-flag 1
.br
.ps +1
\fBExtra Edge List (ID: {E, D, G, E}) [Optional]\fR
.RS 4
.sp
.if n \{\
.RS 4
.\}
.nf
This list of 4\-byte values store the second through nth parents for
all octopus merges\&. The second parent value in the commit data stores
an array position within this list along with the most\-significant bit
on\&. Starting at that array position, iterate through this list of commit
positions for the parents until reaching a value with the most\-significant
bit on\&. The other bits correspond to the position of the last parent\&.
.fi
.if n \{\
.RE
.\}
.RE
.sp
.it 1 an-trap
.nr an-no-space-flag 1
.nr an-break-flag 1
.br
.ps +1
\fBBloom Filter Index (ID: {B, I, D, X}) (N * 4 bytes) [Optional]\fR
.RS 4
.sp
.RS 4
.ie n \{\
\h'-04'\(bu\h'+03'\c
.\}
.el \{\
.sp -1
.IP \(bu 2.3
.\}
The ith entry, BIDX[i], stores the number of bytes in all Bloom filters from commit 0 to commit i (inclusive) in lexicographic order\&. The Bloom filter for the i\-th commit spans from BIDX[i\-1] to BIDX[i] (plus header length), where BIDX[\-1] is 0\&.
.RE
.sp
.RS 4
.ie n \{\
\h'-04'\(bu\h'+03'\c
.\}
.el \{\
.sp -1
.IP \(bu 2.3
.\}
The BIDX chunk is ignored if the BDAT chunk is not present\&.
.RE
.RE
.sp
.it 1 an-trap
.nr an-no-space-flag 1
.nr an-break-flag 1
.br
.ps +1
\fBBloom Filter Data (ID: {B, D, A, T}) [Optional]\fR
.RS 4
.sp
.RS 4
.ie n \{\
\h'-04'\(bu\h'+03'\c
.\}
.el \{\
.sp -1
.IP \(bu 2.3
.\}
It starts with header consisting of three unsigned 32\-bit integers:
.sp
.RS 4
.ie n \{\
\h'-04'\(bu\h'+03'\c
.\}
.el \{\
.sp -1
.IP \(bu 2.3
.\}
Version of the hash algorithm being used\&. We currently only support value 1 which corresponds to the 32\-bit version of the murmur3 hash implemented exactly as described in
\m[blue]\fBhttps://en\&.wikipedia\&.org/wiki/MurmurHash#Algorithm\fR\m[]
and the double hashing technique using seed values 0x293ae76f and 0x7e646e2 as described in
\m[blue]\fBhttps://doi\&.org/10\&.1007/978\-3\-540\-30494\-4_26\fR\m[]
"Bloom Filters in Probabilistic Verification"
.RE
.sp
.RS 4
.ie n \{\
\h'-04'\(bu\h'+03'\c
.\}
.el \{\
.sp -1
.IP \(bu 2.3
.\}
The number of times a path is hashed and hence the number of bit positions that cumulatively determine whether a file is present in the commit\&.
.RE
.sp
.RS 4
.ie n \{\
\h'-04'\(bu\h'+03'\c
.\}
.el \{\
.sp -1
.IP \(bu 2.3
.\}
The minimum number of bits
\fIb\fR
per entry in the Bloom filter\&. If the filter contains
\fIn\fR
entries, then the filter size is the minimum number of 64\-bit words that contain n*b bits\&.
.RE
.RE
.sp
.RS 4
.ie n \{\
\h'-04'\(bu\h'+03'\c
.\}
.el \{\
.sp -1
.IP \(bu 2.3
.\}
The rest of the chunk is the concatenation of all the computed Bloom filters for the commits in lexicographic order\&.
.RE
.sp
.RS 4
.ie n \{\
\h'-04'\(bu\h'+03'\c
.\}
.el \{\
.sp -1
.IP \(bu 2.3
.\}
Note: Commits with no changes or more than 512 changes have Bloom filters of length one, with either all bits set to zero or one respectively\&.
.RE
.sp
.RS 4
.ie n \{\
\h'-04'\(bu\h'+03'\c
.\}
.el \{\
.sp -1
.IP \(bu 2.3
.\}
The BDAT chunk is present if and only if BIDX is present\&.
.RE
.RE
.sp
.it 1 an-trap
.nr an-no-space-flag 1
.nr an-break-flag 1
.br
.ps +1
\fBBase Graphs List (ID: {B, A, S, E}) [Optional]\fR
.RS 4
.sp
.if n \{\
.RS 4
.\}
.nf
This list of H\-byte hashes describe a set of B commit\-graph files that
form a commit\-graph chain\&. The graph position for the ith commit in this
file\*(Aqs OID Lookup chunk is equal to i plus the number of commits in all
base graphs\&.  If B is non\-zero, this chunk must exist\&.
.fi
.if n \{\
.RE
.\}
.RE
.SS "TRAILER:"
.sp
.if n \{\
.RS 4
.\}
.nf
H\-byte HASH\-checksum of all of the above\&.
.fi
.if n \{\
.RE
.\}
.SH "HISTORICAL NOTES:"
.sp
The Generation Data (GDA2) and Generation Data Overflow (GDO2) chunks have the number \fI2\fR in their chunk IDs because a previous version of Git wrote possibly erroneous data in these chunks with the IDs "GDAT" and "GDOV"\&. By changing the IDs, newer versions of Git will silently ignore those older chunks and write the new information without trusting the incorrect data\&.
.SH "GIT"
.sp
Part of the \fBgit\fR(1) suite