1GIT-CVSSERVER(1) Git Manual GIT-CVSSERVER(1)
2
6 git-cvsserver - A CVS server emulator for Git
7
9 SSH:
10 export CVS_SERVER="git cvsserver"
12 cvs -d :ext:user@server/path/repo.git co <HEAD_name>
13 pserver (/etc/inetd.conf):
15 cvspserver stream tcp nowait nobody /usr/bin/git-cvsserver git-cvsserver pserver
17 Usage:
19 git-cvsserver [<options>] [pserver|server] [<directory> ...]
21
23 All these options obviously only make sense if enforced by the server
24 side. They have been implemented to resemble the git-daemon(1) options
25 as closely as possible.
26 --base-path <path>
28 Prepend path to requested CVSROOT
29 --strict-paths
31 Don’t allow recursing into subdirectories
32 --export-all
34 Don’t check for gitcvs.enabled in config. You also have to specify
35 a list of allowed directories (see below) if you want to use this
36 option.
37 -V, --version
39 Print version information and exit
40 -h, -H, --help
42 Print usage information and exit
43 <directory>
45 You can specify a list of allowed directories. If no directories
46 are given, all are allowed. This is an additional restriction,
47 gitcvs access still needs to be enabled by the gitcvs.enabled
48 config option unless --export-all was given, too.
49
51 This application is a CVS emulation layer for Git.
52 It is highly functional. However, not all methods are implemented, and
54 for those methods that are implemented, not all switches are
55 implemented.
56 Testing has been done using both the CLI CVS client, and the Eclipse
58 CVS plugin. Most functionality works fine with both of these clients.
59
61 CVS clients cannot tag, branch or perform Git merges.
62 git-cvsserver maps Git branches to CVS modules. This is very different
64 from what most CVS users would expect since in CVS modules usually
65 represent one or more directories.
66
68 1. If you are going to offer CVS access via pserver, add a line in
69 /etc/inetd.conf like
70 cvspserver stream tcp nowait nobody git-cvsserver pserver
72 Note: Some inetd servers let you specify the name of the executable
74 independently of the value of argv[0] (i.e. the name the program
75 assumes it was executed with). In this case the correct line in
76 /etc/inetd.conf looks like
77 cvspserver stream tcp nowait nobody /usr/bin/git-cvsserver git-cvsserver pserver
79 Only anonymous access is provided by pserve by default. To commit
81 you will have to create pserver accounts, simply add a
82 gitcvs.authdb setting in the config file of the repositories you
83 want the cvsserver to allow writes to, for example:
84 [gitcvs]
86 authdb = /etc/cvsserver/passwd
87 The format of these files is username followed by the encrypted
89 password, for example:
90 myuser:$1Oyx5r9mdGZ2
92 myuser:$1$BA)@$vbnMJMDym7tA32AamXrm./
93 You can use the htpasswd facility that comes with Apache to make
95 these files, but Apache’s MD5 crypt method differs from the one
96 used by most C library’s crypt() function, so don’t use the -m
97 option.
98 Alternatively you can produce the password with perl’s crypt()
100 operator:
101 perl -e 'my ($user, $pass) = @ARGV; printf "%s:%s\n", $user, crypt($user, $pass)' $USER password
103 Then provide your password via the pserver method, for example:
105 cvs -d:pserver:someuser:somepassword <at> server/path/repo.git co <HEAD_name>
107 No special setup is needed for SSH access, other than having Git
109 tools in the PATH. If you have clients that do not accept the
110 CVS_SERVER environment variable, you can rename git-cvsserver to
111 cvs.
112 Note: Newer CVS versions (>= 1.12.11) also support specifying
114 CVS_SERVER directly in CVSROOT like
115 cvs -d ":ext;CVS_SERVER=git cvsserver:user@server/path/repo.git" co <HEAD_name>
117 This has the advantage that it will be saved in your CVS/Root files
119 and you don’t need to worry about always setting the correct
120 environment variable. SSH users restricted to git-shell don’t need
121 to override the default with CVS_SERVER (and shouldn’t) as
122 git-shell understands cvs to mean git-cvsserver and pretends that
123 the other end runs the real cvs better.
124 2. For each repo that you want accessible from CVS you need to edit
126 config in the repo and add the following section.
127 [gitcvs]
129 enabled=1
130 # optional for debugging
131 logFile=/path/to/logfile
132 Note: you need to ensure each user that is going to invoke
134 git-cvsserver has write access to the log file and to the database
135 (see Database Backend. If you want to offer write access over SSH,
136 the users of course also need write access to the Git repository
137 itself.
138 You also need to ensure that each repository is "bare" (without a
140 Git index file) for cvs commit to work. See gitcvs-migration(7).
141 All configuration variables can also be overridden for a specific
143 method of access. Valid method names are "ext" (for SSH access) and
144 "pserver". The following example configuration would disable
145 pserver access while still allowing access over SSH.
146 [gitcvs]
148 enabled=0
149 [gitcvs "ext"]
151 enabled=1
152 3. If you didn’t specify the CVSROOT/CVS_SERVER directly in the
154 checkout command, automatically saving it in your CVS/Root files,
155 then you need to set them explicitly in your environment. CVSROOT
156 should be set as per normal, but the directory should point at the
157 appropriate Git repo. As above, for SSH clients not restricted to
158 git-shell, CVS_SERVER should be set to git-cvsserver.
159 export CVSROOT=:ext:user@server:/var/git/project.git
161 export CVS_SERVER="git cvsserver"
162 4. For SSH clients that will make commits, make sure their server-side
164 .ssh/environment files (or .bashrc, etc., according to their
165 specific shell) export appropriate values for GIT_AUTHOR_NAME,
166 GIT_AUTHOR_EMAIL, GIT_COMMITTER_NAME, and GIT_COMMITTER_EMAIL. For
167 SSH clients whose login shell is bash, .bashrc may be a reasonable
168 alternative.
169 5. Clients should now be able to check out the project. Use the CVS
171 module name to indicate what Git head you want to check out. This
172 also sets the name of your newly checked-out directory, unless you
173 tell it otherwise with -d <dir_name>. For example, this checks out
174 master branch to the project-master directory:
175 cvs co -d project-master master
177
179 git-cvsserver uses one database per Git head (i.e. CVS module) to store
180 information about the repository to maintain consistent CVS revision
181 numbers. The database needs to be updated (i.e. written to) after every
182 commit.
183 If the commit is done directly by using git (as opposed to using
185 git-cvsserver) the update will need to happen on the next repository
186 access by git-cvsserver, independent of access method and requested
187 operation.
188 That means that even if you offer only read access (e.g. by using the
190 pserver method), git-cvsserver should have write access to the database
191 to work reliably (otherwise you need to make sure that the database is
192 up to date any time git-cvsserver is executed).
193 By default it uses SQLite databases in the Git directory, named
195 gitcvs.<module_name>.sqlite. Note that the SQLite backend creates
196 temporary files in the same directory as the database file on write so
197 it might not be enough to grant the users using git-cvsserver write
198 access to the database file without granting them write access to the
199 directory, too.
200 The database cannot be reliably regenerated in a consistent form after
202 the branch it is tracking has changed. Example: For merged branches,
203 git-cvsserver only tracks one branch of development, and after a git
204 merge an incrementally updated database may track a different branch
205 than a database regenerated from scratch, causing inconsistent CVS
206 revision numbers. git-cvsserver has no way of knowing which branch it
207 would have picked if it had been run incrementally pre-merge. So if you
208 have to fully or partially (from old backup) regenerate the database,
209 you should be suspicious of pre-existing CVS sandboxes.
210 You can configure the database backend with the following configuration
212 variables:
213 Configuring database backend
215 git-cvsserver uses the Perl DBI module. Please also read its
216 documentation if changing these variables, especially about
217 DBI->connect().
218 gitcvs.dbName
220 Database name. The exact meaning depends on the selected database
221 driver, for SQLite this is a filename. Supports variable
222 substitution (see below). May not contain semicolons (;). Default:
223 %Ggitcvs.%m.sqlite
224 gitcvs.dbDriver
226 Used DBI driver. You can specify any available driver for this
227 here, but it might not work. cvsserver is tested with DBD::SQLite,
228 reported to work with DBD::Pg, and reported not to work with
229 DBD::mysql. Please regard this as an experimental feature. May not
230 contain colons (:). Default: SQLite
231 gitcvs.dbuser
233 Database user. Only useful if setting dbDriver, since SQLite has no
234 concept of database users. Supports variable substitution (see
235 below).
236 gitcvs.dbPass
238 Database password. Only useful if setting dbDriver, since SQLite
239 has no concept of database passwords.
240 gitcvs.dbTableNamePrefix
242 Database table name prefix. Supports variable substitution (see
243 below). Any non-alphabetic characters will be replaced with
244 underscores.
245 All variables can also be set per access method, see above.
247 Variable substitution
249 In dbDriver and dbUser you can use the following variables:
250 %G
252 Git directory name
253 %g
255 Git directory name, where all characters except for
256 alphanumeric ones, ., and - are replaced with _ (this should
257 make it easier to use the directory name in a filename if
258 wanted)
259 %m
261 CVS module/Git head name
262 %a
264 access method (one of "ext" or "pserver")
265 %u
267 Name of the user running git-cvsserver. If no name can be
268 determined, the numeric uid is used.
269
271 These variables obviate the need for command-line options in some
272 circumstances, allowing easier restricted usage through git-shell.
273 GIT_CVSSERVER_BASE_PATH takes the place of the argument to --base-path.
275 GIT_CVSSERVER_ROOT specifies a single-directory whitelist. The
277 repository must still be configured to allow access through
278 git-cvsserver, as described above.
279 When these environment variables are set, the corresponding
281 command-line arguments may not be used.
282
284 To get a checkout with the Eclipse CVS client:
285 1. Select "Create a new project → From CVS checkout"
287 2. Create a new location. See the notes below for details on how to
289 choose the right protocol.
290 3. Browse the modules available. It will give you a list of the heads
292 in the repository. You will not be able to browse the tree from
293 there. Only the heads.
294 4. Pick HEAD when it asks what branch/tag to check out. Untick the
296 "launch commit wizard" to avoid committing the .project file.
297 Protocol notes: If you are using anonymous access via pserver, just
299 select that. Those using SSH access should choose the ext protocol, and
300 configure ext access on the Preferences→Team→CVS→ExtConnection pane.
301 Set CVS_SERVER to "git cvsserver". Note that password support is not
302 good when using ext, you will definitely want to have SSH keys setup.
303 Alternatively, you can just use the non-standard extssh protocol that
305 Eclipse offer. In that case CVS_SERVER is ignored, and you will have to
306 replace the cvs utility on the server with git-cvsserver or manipulate
307 your .bashrc so that calling cvs effectively calls git-cvsserver.
308
310 · CVS 1.12.9 on Debian
311 · CVS 1.11.17 on MacOSX (from Fink package)
313 · Eclipse 3.0, 3.1.2 on MacOSX (see Eclipse CVS Client Notes)
315 · TortoiseCVS
317
319 All the operations required for normal use are supported, including
320 checkout, diff, status, update, log, add, remove, commit.
321 Most CVS command arguments that read CVS tags or revision numbers
323 (typically -r) work, and also support any git refspec (tag, branch,
324 commit ID, etc). However, CVS revision numbers for non-default branches
325 are not well emulated, and cvs log does not show tags or branches at
326 all. (Non-main-branch CVS revision numbers superficially resemble CVS
327 revision numbers, but they actually encode a git commit ID directly,
328 rather than represent the number of revisions since the branch point.)
329 Note that there are two ways to checkout a particular branch. As
331 described elsewhere on this page, the "module" parameter of cvs
332 checkout is interpreted as a branch name, and it becomes the main
333 branch. It remains the main branch for a given sandbox even if you
334 temporarily make another branch sticky with cvs update -r.
335 Alternatively, the -r argument can indicate some other branch to
336 actually checkout, even though the module is still the "main" branch.
337 Tradeoffs (as currently implemented): Each new "module" creates a new
338 database on disk with a history for the given module, and after the
339 database is created, operations against that main branch are fast. Or
340 alternatively, -r doesn’t take any extra disk space, but may be
341 significantly slower for many operations, like cvs update.
342 If you want to refer to a git refspec that has characters that are not
344 allowed by CVS, you have two options. First, it may just work to supply
345 the git refspec directly to the appropriate CVS -r argument; some CVS
346 clients don’t seem to do much sanity checking of the argument. Second,
347 if that fails, you can use a special character escape mechanism that
348 only uses characters that are valid in CVS tags. A sequence of 4 or 5
349 characters of the form (underscore ("_"), dash ("-"), one or two
350 characters, and dash ("-")) can encode various characters based on the
351 one or two letters: "s" for slash ("/"), "p" for period ("."), "u" for
352 underscore ("_"), or two hexadecimal digits for any byte value at all
353 (typically an ASCII number, or perhaps a part of a UTF-8 encoded
354 character).
355 Legacy monitoring operations are not supported (edit, watch and
357 related). Exports and tagging (tags and branches) are not supported at
358 this stage.
359 CRLF Line Ending Conversions
361 By default the server leaves the -k mode blank for all files, which
362 causes the CVS client to treat them as a text files, subject to
363 end-of-line conversion on some platforms.
364 You can make the server use the end-of-line conversion attributes to
366 set the -k modes for files by setting the gitcvs.usecrlfattr config
367 variable. See gitattributes(5) for more information about end-of-line
368 conversion.
369 Alternatively, if gitcvs.usecrlfattr config is not enabled or the
371 attributes do not allow automatic detection for a filename, then the
372 server uses the gitcvs.allBinary config for the default setting. If
373 gitcvs.allBinary is set, then file not otherwise specified will default
374 to -kb mode. Otherwise the -k mode is left blank. But if
375 gitcvs.allBinary is set to "guess", then the correct -k mode will be
376 guessed based on the contents of the file.
377 For best consistency with cvs, it is probably best to override the
379 defaults by setting gitcvs.usecrlfattr to true, and gitcvs.allBinary to
380 "guess".
381
383 git-cvsserver depends on DBD::SQLite.
384
386 Part of the git(1) suite
387Git 2.30.2 2021-03-08 GIT-CVSSERVER(1)