From 6aa4adae0e05cc08f25060637bc0497bdab1ab02 Mon Sep 17 00:00:00 2001 From: Christian Couder Date: Tue, 8 Sep 2026 18:41:25 +0200 Subject: [PATCH 1/5] promisor-remote: factor out lazy_fetch_objects() In "promisor-remote.c:fetch_objects()", there is a check to disable lazy fetching when the `GIT_NO_LAZY_FETCH` environment variable is set. The fetch_objects() function is called once per promisor remote though. So the check might be performed more times than necessary. Also promisor_remote_get_direct() mixes up the logic deciding which promisor remotes to try with the logic checking that the objects that could not be fetched are promisor objects. Let's refactor the lazy fetching logic out of these two functions into a new lazy_fetch_objects() function. This is a pure refactoring with no intended behavior change. Two things shift in ways that are observably equivalent though: - the `GIT_NO_LAZY_FETCH` check is now performed once up front, instead of once per promisor remote, and - promisor_remote_init() is no longer called when lazy fetching is disabled, which is fine as nothing downstream of it, like is_promisor_object(), needs it in that case. While at it, let's also convert try_promisor_remotes() to return 'bool' instead of 'int', as it just returns whether all the objects could be fetched, and document its return value. Signed-off-by: Christian Couder Signed-off-by: Junio C Hamano --- promisor-remote.c | 74 +++++++++++++++++++++++++++-------------------- 1 file changed, 43 insertions(+), 31 deletions(-) diff --git a/promisor-remote.c b/promisor-remote.c index 43505d1e1a..df17fec3bb 100644 --- a/promisor-remote.c +++ b/promisor-remote.c @@ -31,15 +31,6 @@ static int fetch_objects(struct repository *repo, FILE *child_in; int quiet; - if (git_env_bool(NO_LAZY_FETCH_ENVIRONMENT, 0)) { - static int warning_shown; - if (!warning_shown) { - warning_shown = 1; - warning(_("lazy fetching disabled; some objects may not be available")); - } - return -1; - } - child.git_cmd = 1; child.in = -1; if (repo != the_repository) @@ -270,10 +261,15 @@ static int remove_fetched_oids(struct repository *repo, return remaining_nr; } -static int try_promisor_remotes(struct repository *repo, - struct object_id **remaining_oids, - int *remaining_nr, int *to_free, - bool accepted_only) +/* + * Return 'true' if all the objects could be fetched from the + * (non-)accepted remotes, 'false' otherwise. + */ +static bool try_promisor_remotes(struct repository *repo, + struct object_id **remaining_oids, + int *remaining_nr, + int *to_free, + bool accepted_only) { struct promisor_remote *r = repo->promisor_remote_config->promisors; @@ -290,9 +286,35 @@ static int try_promisor_remotes(struct repository *repo, continue; } } - return 1; /* all fetched */ + return true; /* all fetched */ } - return 0; + return false; +} + +/* + * Return 'true' if all the objects could be fetched, 'false' otherwise. + */ +static bool lazy_fetch_objects(struct repository *repo, + struct object_id **remaining_oids, + int *remaining_nr, + int *to_free) +{ + if (git_env_bool(NO_LAZY_FETCH_ENVIRONMENT, 0)) { + static int warning_shown; + if (!warning_shown) { + warning_shown = 1; + warning(_("lazy fetching disabled; some objects may not be available")); + } + return false; + } + + promisor_remote_init(repo); + + /* Try accepted remotes first (those the server told us to use) */ + return try_promisor_remotes(repo, remaining_oids, remaining_nr, + to_free, true) || + try_promisor_remotes(repo, remaining_oids, remaining_nr, + to_free, false); } void promisor_remote_get_direct(struct repository *repo, @@ -302,28 +324,18 @@ void promisor_remote_get_direct(struct repository *repo, struct object_id *remaining_oids = (struct object_id *)oids; int remaining_nr = oid_nr; int to_free = 0; - int i; if (oid_nr == 0) return; - promisor_remote_init(repo); - - /* Try accepted remotes first (those the server told us to use) */ - if (try_promisor_remotes(repo, &remaining_oids, &remaining_nr, - &to_free, true)) - goto all_fetched; - if (try_promisor_remotes(repo, &remaining_oids, &remaining_nr, - &to_free, false)) - goto all_fetched; - - for (i = 0; i < remaining_nr; i++) { - if (is_promisor_object(repo, &remaining_oids[i])) - die(_("could not fetch %s from promisor remote"), - oid_to_hex(&remaining_oids[i])); + if (!lazy_fetch_objects(repo, &remaining_oids, &remaining_nr, &to_free)) { + for (int i = 0; i < remaining_nr; i++) { + if (is_promisor_object(repo, &remaining_oids[i])) + die(_("could not fetch %s from promisor remote"), + oid_to_hex(&remaining_oids[i])); + } } -all_fetched: if (to_free) free(remaining_oids); } From e63bb828500aa12e6d066f8d18e86643ddcde5bd Mon Sep 17 00:00:00 2001 From: Christian Couder Date: Tue, 8 Sep 2026 18:41:26 +0200 Subject: [PATCH 2/5] setup: extract path_allowlist_apply() In a following commit we are going to check whether a repository is part of an allowlist specified in a config variable. To prepare for that let's extract existing code from safe_directory_cb() into a new path_allowlist_apply() helper that will help with such checks. While at it let's make the helper's code simpler and more generic, by passing it a `bool (*allow_path)(const char *path, void *cbdata)` function that decides if a path is acceptable by the caller. To further simplify how to reuse that new helper, and avoid duplicating the config-value handling in a future commit, let's also introduce a path_allowlist_config_apply() helper. For clarity, let's change the `int is_safe` to `bool safe` in `struct safe_directory_data`. Signed-off-by: Christian Couder Signed-off-by: Junio C Hamano --- setup.c | 138 ++++++++++++++++++++++++++++++++++++-------------------- setup.h | 50 ++++++++++++++++++++ 2 files changed, 138 insertions(+), 50 deletions(-) diff --git a/setup.c b/setup.c index dfe05d9a03..366a7dc5c0 100644 --- a/setup.c +++ b/setup.c @@ -1338,67 +1338,105 @@ static int canonicalize_ceiling_entry(struct string_list_item *item, } } +void path_allowlist_apply(const char *allowed, const char *target_path, + bool *matches, + bool (*allow_path)(const char *path, void *cbdata), + void *allow_path_cbdata) +{ + char *normalized = NULL; + + if (!allowed || !*allowed) { + *matches = false; + return; + } + + if (!strcmp(allowed, "*")) { + *matches = true; + return; + } + + if (!allow_path(allowed, allow_path_cbdata)) + return; + + /* + * A .gitconfig in $HOME may be shared across different + * machines and the config variable entries may or may not + * exist as paths on all of these machines. In other words, + * it is not a warning worthy event when there is no such path + * on this machine---the entry may be useful elsewhere. + */ + normalized = real_pathdup(allowed, 0); + if (!normalized) + return; + + if (ends_with(normalized, "/*")) { + size_t len = strlen(normalized); + if (!fspathncmp(normalized, target_path, len - 1)) + *matches = true; + } else if (!fspathcmp(target_path, normalized)) { + *matches = true; + } + + free(normalized); +} + +void path_allowlist_config_apply(const char *key, const char *value, + const char *target_path, bool *matches, + bool (*allow_path)(const char *path, void *cbdata), + void *allow_path_cbdata) +{ + char *allowed = NULL; + + if (!value || !*value || !strcmp(value, "*")) { + path_allowlist_apply(value, target_path, matches, + allow_path, allow_path_cbdata); + return; + } + + if (git_config_pathname(&allowed, key, value) || !allowed) + return; + + path_allowlist_apply(allowed, target_path, matches, + allow_path, allow_path_cbdata); + + free(allowed); +} + +/* + * Setting the config variable to a non-absolute path makes + * little sense---it won't be relative to the configuration + * file the item is defined in. Except for ".", which means + * "if we are at the top level of a repository, then it is + * OK", which is slightly tighter than "*" that allows + * discovery. + */ +static bool allow_safe_dir(const char *path, void *cbdata_) +{ + struct path_allowlist_cb_data *cbdata = cbdata_; + + if (is_absolute_path(path) || !strcmp(path, ".")) + return true; + + warning(_("%s '%s' not absolute"), cbdata->key, path); + return false; +} + struct safe_directory_data { char *path; - int is_safe; + bool safe; }; static int safe_directory_cb(const char *key, const char *value, const struct config_context *ctx UNUSED, void *d) { struct safe_directory_data *data = d; + struct path_allowlist_cb_data cbdata = { .key = key }; if (strcmp(key, "safe.directory")) return 0; - if (!value || !*value) { - data->is_safe = 0; - } else if (!strcmp(value, "*")) { - data->is_safe = 1; - } else { - char *allowed = NULL; - - if (!git_config_pathname(&allowed, key, value) && allowed) { - char *normalized = NULL; - - /* - * Setting safe.directory to a non-absolute path - * makes little sense---it won't be relative to - * the configuration file the item is defined in. - * Except for ".", which means "if we are at the top - * level of a repository, then it is OK", which is - * slightly tighter than "*" that allows discovery. - */ - if (!is_absolute_path(allowed) && strcmp(allowed, ".")) { - warning(_("safe.directory '%s' not absolute"), - allowed); - goto next; - } - - /* - * A .gitconfig in $HOME may be shared across - * different machines and safe.directory entries - * may or may not exist as paths on all of these - * machines. In other words, it is not a warning - * worthy event when there is no such path on this - * machine---the entry may be useful elsewhere. - */ - normalized = real_pathdup(allowed, 0); - if (!normalized) - goto next; - - if (ends_with(normalized, "/*")) { - size_t len = strlen(normalized); - if (!fspathncmp(normalized, data->path, len - 1)) - data->is_safe = 1; - } else if (!fspathcmp(data->path, normalized)) { - data->is_safe = 1; - } - next: - free(normalized); - free(allowed); - } - } + path_allowlist_config_apply(key, value, data->path, &data->safe, + allow_safe_dir, &cbdata); return 0; } @@ -1440,7 +1478,7 @@ static int ensure_valid_ownership(const char *gitfile, git_protected_config(safe_directory_cb, &data); free(data.path); - return data.is_safe; + return data.safe; } void die_upon_dubious_ownership(const char *gitfile, const char *worktree, diff --git a/setup.h b/setup.h index 763fd384e8..6b84fbe507 100644 --- a/setup.h +++ b/setup.h @@ -304,4 +304,54 @@ struct startup_info { extern struct startup_info *startup_info; extern const char *tmp_original_cwd; +/* Path allowlist */ + +struct path_allowlist_cb_data { + const char *key; +}; + +/* + * Check the allowlist entry in `allowed` against `target_path`, + * updating `*matches` accordingly. + * + * `allowed` is a single entry of an allowlist of paths, typically one + * value of a multi-valued config variable, already expanded by + * git_config_pathname(). `target_path` is the (normalized) path being + * tested. `*matches` is updated in place: + * + * - an empty `allowed` resets it to 'false' (so a later, more + * specific config scope can clear entries from a broader one), + * - "*" sets it to 'true' (allow everything), + * - "" sets it to 'true' if equals `target_path`, + * - "" + "/" + "*" sets it to 'true' if is a leading + * directory of `target_path`, + * - anything else leaves `*matches` unchanged. + * + * `allow_path` is called with `allowed` and `allow_path_cbdata`, and + * should return 'true' if the entry is acceptable to the caller. It + * lets each caller decide which paths it is willing to consider, and + * whether to warn about the ones it rejects. Returning 'false' leaves + * `*matches` unchanged. + * + * Callers are expected to invoke this once per allowlist entry, + * typically from a protected-config callback, so that untrusted + * repository config cannot influence the decision. + */ +void path_allowlist_apply(const char *allowed, const char *target_path, + bool *matches, + bool (*allow_path)(const char *path, void *cbdata), + void *allow_path_cbdata); + +/* + * Apply one value of a multi-valued config variable holding an + * allowlist of paths, expanding it with git_config_pathname() before + * checking it against `target_path`. Empty and "*" values are passed + * through without expansion, as interpolating them is not + * meaningful. See path_allowlist_apply(). + */ +void path_allowlist_config_apply(const char *key, const char *value, + const char *target_path, bool *matches, + bool (*allow_path)(const char *path, void *cbdata), + void *allow_path_cbdata); + #endif /* SETUP_H */ From 6a22bd1cf68164d25652f6fc1f334e1370691f18 Mon Sep 17 00:00:00 2001 From: Christian Couder Date: Tue, 8 Sep 2026 18:41:27 +0200 Subject: [PATCH 3/5] upload-pack: read uploadpack.lazyFetchTrusted Previous commits created and prepared the path_allowlist_apply() and path_allowlist_config_apply() functions, but used them only for the "safe.directory" configuration variable. Let's reuse these functions for a new "uploadpack.lazyFetchTrusted" configuration variable. It allows us to: - read an allowlist from that config variable, - check if the current repo is in that list, and - return the result from a new upload_pack_lazy_fetch_trusted() function. As path_allowlist_config_apply() lets each caller decide which paths it is willing to accept using a callback, let's pass it a new allow_trusted_path() callback. Unlike the "safe.directory" callback, it accepts only absolute paths, and not ".", as `upload-pack` always serves a repository given by an absolute path, so there is no "current repository" for "." to refer to. Note that a served repository is identified by its git directory, and not by its worktree. This is because `upload-pack` uses enter_repo() instead of the usual repository discovery, so it never learns about a worktree and `r->worktree` is always NULL there. In practice this means that a non-bare repository served as "/srv/repo" has to be allowlisted as "/srv/repo/.git". The new upload_pack_lazy_fetch_trusted() function will be used in a following commit. Note that the new config variable should be read only from protected configuration files. Signed-off-by: Christian Couder Signed-off-by: Junio C Hamano --- upload-pack.c | 59 +++++++++++++++++++++++++++++++++++++++++++++++++++ upload-pack.h | 3 +++ 2 files changed, 62 insertions(+) diff --git a/upload-pack.c b/upload-pack.c index 22573ad365..a300870fa9 100644 --- a/upload-pack.c +++ b/upload-pack.c @@ -34,6 +34,8 @@ #include "json-writer.h" #include "strmap.h" #include "promisor-remote.h" +#include "setup.h" +#include "abspath.h" /* Remember to update object flag allocation in object.h */ #define THEY_HAVE (1u << 11) @@ -1343,6 +1345,63 @@ static int upload_pack_config(const char *var, const char *value, return parse_hide_refs_config(var, value, "uploadpack", &data->hidden_refs); } +/* + * Only absolute paths make sense here. Unlike 'safe.directory', "." + * is not accepted, as the served repository is always identified by + * an absolute path. + */ +static bool allow_trusted_path(const char *path, void *cbdata_) +{ + struct path_allowlist_cb_data *cbdata = cbdata_; + + if (is_absolute_path(path)) + return true; + + warning(_("%s '%s' not absolute"), cbdata->key, path); + return false; +} + +struct lazy_fetch_trusted { + char *repo_path; + bool trusted; +}; + +static int upload_pack_protected_lazy_fetch_config(const char *var, const char *value, + const struct config_context *ctx UNUSED, + void *cb_data) +{ + struct lazy_fetch_trusted *data = cb_data; + struct path_allowlist_cb_data cbdata = { .key = var }; + + if (strcmp("uploadpack.lazyfetchtrusted", var)) + return 0; + + path_allowlist_config_apply(var, value, data->repo_path, &data->trusted, + allow_trusted_path, &cbdata); + + return 0; +} + +bool upload_pack_lazy_fetch_trusted(struct repository *r) +{ + struct lazy_fetch_trusted data = { 0 }; + + /* + * A served repository is identified by its git directory, as + * `upload-pack` uses enter_repo() instead of the usual repository + * discovery, so its worktree, if any, is never known here. + */ + data.repo_path = real_pathdup(r->gitdir, 0); + if (!data.repo_path) + return false; + + git_protected_config(upload_pack_protected_lazy_fetch_config, &data); + + free(data.repo_path); + + return !!data.trusted; +} + static int upload_pack_protected_config(const char *var, const char *value, const struct config_context *ctx UNUSED, void *cb_data) diff --git a/upload-pack.h b/upload-pack.h index d6ee25ea98..b2212992c3 100644 --- a/upload-pack.h +++ b/upload-pack.h @@ -12,4 +12,7 @@ struct strbuf; int upload_pack_advertise(struct repository *r, struct strbuf *value); +/* Is this repo trusted for lazy fetching? */ +bool upload_pack_lazy_fetch_trusted(struct repository *r); + #endif /* UPLOAD_PACK_H */ From fb65aca90a3a9fa179ad28e57b3e34afa18088d3 Mon Sep 17 00:00:00 2001 From: Christian Couder Date: Tue, 8 Sep 2026 18:41:28 +0200 Subject: [PATCH 4/5] promisor-remote: prevent infinite recursion when lazy fetching If a repository R is configured to lazy fetch from a promisor remote P which is also configured to in turn lazy fetch from R, there is an infinite recursion: R asks P for a missing object, P asks R for it, and so on. The simplest case of this is a repository configured as its own promisor remote. This is not reachable when serving a repository by default, as `upload-pack` sets `GIT_NO_LAZY_FETCH` to 1, which makes the nested `upload-pack` refuse to lazily fetch. A following commit will let server operators allow lazy fetching for repositories they trust though, and as `GIT_NO_LAZY_FETCH` is then set to 0 and passed down to child processes, nothing stops the recursion anymore. It does not recurse forever in practice, but only because each level adds one more variable to the environment of the child process, so after a while `exec()` fails with: fatal: cannot exec 'git-upload-pack ...': Argument list too long fatal: unable to fork To avoid this pathological case altogether, let's use a new `GIT_INTERNAL_LAZY_FETCH_DEPTH` to count the recursion depth, and let's check that it doesn't exceed a MAX_LAZY_FETCH_DEPTH limit (set to 5 for now). Note that some nesting is legitimate: when `git fetch` runs `index-pack`, it can lazily fetch REF_DELTA bases that are missing locally, so the limit should not be 1. Signed-off-by: Christian Couder Signed-off-by: Junio C Hamano --- environment.h | 8 ++++++++ promisor-remote.c | 26 ++++++++++++++++++++++---- t/t0410-partial-clone.sh | 33 +++++++++++++++++++++++++++++++++ 3 files changed, 63 insertions(+), 4 deletions(-) diff --git a/environment.h b/environment.h index e7ec5b0437..f2833be9fe 100644 --- a/environment.h +++ b/environment.h @@ -52,6 +52,14 @@ */ #define GIT_ADVICE_ENVIRONMENT "GIT_ADVICE" +/* + * Environment variable used to detect that a lazy fetch is already in + * progress in a parent process, to prevent infinite recursion when a + * promisor remote resolves back to the repository being served. + * This is an internal variable that should not be set by the user. + */ +#define LAZY_FETCH_DEPTH_ENVIRONMENT "GIT_INTERNAL_LAZY_FETCH_DEPTH" + /* * Environment variable used in handshaking the wire protocol. * Contains a colon ':' separated list of keys with optional values diff --git a/promisor-remote.c b/promisor-remote.c index df17fec3bb..e9c5b413f1 100644 --- a/promisor-remote.c +++ b/promisor-remote.c @@ -24,7 +24,7 @@ struct promisor_remote_config { static int fetch_objects(struct repository *repo, const char *remote_name, const struct object_id *oids, - int oid_nr) + int oid_nr, unsigned long depth) { struct child_process child = CHILD_PROCESS_INIT; int i; @@ -41,6 +41,7 @@ static int fetch_objects(struct repository *repo, "--filter=blob:none", "--stdin", NULL); if (!repo_config_get_bool(repo, "promisor.quiet", &quiet) && quiet) strvec_push(&child.args, "--quiet"); + strvec_pushf(&child.env, "%s=%lu", LAZY_FETCH_DEPTH_ENVIRONMENT, depth + 1); if (start_command(&child)) die(_("promisor-remote: unable to fork off fetch subprocess")); child_in = xfdopen(child.in, "w"); @@ -269,6 +270,7 @@ static bool try_promisor_remotes(struct repository *repo, struct object_id **remaining_oids, int *remaining_nr, int *to_free, + unsigned long depth, bool accepted_only) { struct promisor_remote *r = repo->promisor_remote_config->promisors; @@ -276,7 +278,8 @@ static bool try_promisor_remotes(struct repository *repo, for (; r; r = r->next) { if (accepted_only != r->accepted) continue; - if (fetch_objects(repo, r->name, *remaining_oids, *remaining_nr) < 0) { + if (fetch_objects(repo, r->name, + *remaining_oids, *remaining_nr, depth) < 0) { if (*remaining_nr == 1) continue; *remaining_nr = remove_fetched_oids(repo, remaining_oids, @@ -291,6 +294,8 @@ static bool try_promisor_remotes(struct repository *repo, return false; } +#define MAX_LAZY_FETCH_DEPTH 5 + /* * Return 'true' if all the objects could be fetched, 'false' otherwise. */ @@ -299,6 +304,8 @@ static bool lazy_fetch_objects(struct repository *repo, int *remaining_nr, int *to_free) { + unsigned long depth = git_env_ulong(LAZY_FETCH_DEPTH_ENVIRONMENT, 0); + if (git_env_bool(NO_LAZY_FETCH_ENVIRONMENT, 0)) { static int warning_shown; if (!warning_shown) { @@ -308,13 +315,24 @@ static bool lazy_fetch_objects(struct repository *repo, return false; } + if (depth >= MAX_LAZY_FETCH_DEPTH) { + static int warning_shown; + if (!warning_shown) { + warning_shown = 1; + warning(_("too many nested lazy fetches (%lu); " + "is a promisor remote pointing at the repository itself?"), + depth); + } + return false; + } + promisor_remote_init(repo); /* Try accepted remotes first (those the server told us to use) */ return try_promisor_remotes(repo, remaining_oids, remaining_nr, - to_free, true) || + to_free, depth, true) || try_promisor_remotes(repo, remaining_oids, remaining_nr, - to_free, false); + to_free, depth, false); } void promisor_remote_get_direct(struct repository *repo, diff --git a/t/t0410-partial-clone.sh b/t/t0410-partial-clone.sh index 788e9a1631..a54685e3c7 100755 --- a/t/t0410-partial-clone.sh +++ b/t/t0410-partial-clone.sh @@ -709,6 +709,39 @@ test_expect_success 'lazy-fetch when accessing object not in the_repository' ' test_grep ! "[?]$FILE_HASH" out ' +test_expect_success 'lazy-fetch does not recurse infinitely between two promisor remotes' ' + rm -rf full partial1.git partial2.git && + + # Create a repo with a blob + test_create_repo full && + test_config -C full uploadpack.allowfilter 1 && + test_config -C full uploadpack.allowanysha1inwant 1 && + test_commit -C full create-a-file file.txt && + FILE_HASH=$(git -C full rev-parse HEAD:file.txt) && + + # Create partial clone repos without blobs + git clone --filter=blob:none --bare "file://$(pwd)/full" partial1.git && + git clone --filter=blob:none --bare "file://$(pwd)/full" partial2.git && + test_config -C partial1.git uploadpack.allowfilter 1 && + test_config -C partial1.git uploadpack.allowanysha1inwant 1 && + test_config -C partial2.git uploadpack.allowfilter 1 && + test_config -C partial2.git uploadpack.allowanysha1inwant 1 && + + # Configure the partial repos as remotes of each other + git -C partial2.git remote set-url origin "file://$(pwd)/partial1.git" && + git -C partial1.git remote set-url origin "file://$(pwd)/partial2.git" && + + # Make sure lazy fetching fails + test_must_fail env GIT_TRACE="$(pwd)/trace" GIT_NO_LAZY_FETCH=0 \ + git -C partial1.git cat-file -e "$FILE_HASH" 2>err && + test_grep "too many nested lazy fetches" err && + + # Make sure the recursion was bounded, i.e. that only + # MAX_LAZY_FETCH_DEPTH "git fetch" subprocesses were spawned + grep "run_command: GIT_INTERNAL_LAZY_FETCH_DEPTH" trace >fetches && + test_line_count = 5 fetches +' + test_expect_success 'push should not fetch new commit objects' ' rm -rf server client && test_create_repo server && From e7f98afab48f971ebecd6956271cddf0ecd4c1e2 Mon Sep 17 00:00:00 2001 From: Christian Couder Date: Tue, 8 Sep 2026 18:41:29 +0200 Subject: [PATCH 5/5] builtin/upload-pack: set GIT_NO_LAZY_FETCH to 0 on trusted repo A previous commit added a new "uploadpack.lazyFetchTrusted" protected config variable that can contain an allowlist of repos, as well as functions to check if the current repo is in that list. But when the current repo is in that list, we currently do nothing. Let's instead set `GIT_NO_LAZY_FETCH` to `0`, which allows `upload-pack` and its `pack-objects` child process to lazily fetch the objects they need to serve a client, for example when the filter used by the client and the one used by the server don't match. This allows server operators to properly control lazy fetching. It is their responsibility, not the client's, to decide if the served repo is trusted, as the main security issue is that lazily fetching runs `git fetch`, which may execute arbitrary commands specified in the configuration and hooks of the served repo. As `GIT_NO_LAZY_FETCH` is passed down to child processes through the environment, this works for `pack-objects`, which performs the lazy fetch when serving a client, without any further plumbing. Now that "uploadpack.lazyFetchTrusted" is actually doing something, let's document it and reference it from GIT_NO_LAZY_FETCH's docs. Signed-off-by: Christian Couder Signed-off-by: Junio C Hamano --- Documentation/config/uploadpack.adoc | 49 +++++++++ Documentation/git-upload-pack.adoc | 5 + Documentation/git.adoc | 4 +- builtin/upload-pack.c | 11 ++ t/t5710-promisor-remote-capability.sh | 142 ++++++++++++++++++++++++++ 5 files changed, 210 insertions(+), 1 deletion(-) diff --git a/Documentation/config/uploadpack.adoc b/Documentation/config/uploadpack.adoc index 0e1dda944a..e143de93aa 100644 --- a/Documentation/config/uploadpack.adoc +++ b/Documentation/config/uploadpack.adoc @@ -86,3 +86,52 @@ uploadpack.allowRefInWant:: is intended for the benefit of load-balanced servers which may not have the same view of what OIDs their refs point to due to replication delay. + +uploadpack.lazyFetchTrusted:: + A multi-valued configuration variable, each of which contains the + absolute local path of a repository that `upload-pack` is allowed to + lazily fetch missing objects for. ++ +A repository is identified by its git directory, i.e. the `.git` +directory of a repository that has a worktree, or the repository itself +if it is bare. So a non-bare repository served as `/srv/repo` has to be +allowlisted as `/srv/repo/.git`. Giving a path with `/*` appended to it +will trust all repositories under the named directory. To trust all +served repositories, set `uploadpack.lazyFetchTrusted` to the string +`*`. ++ +The value of this setting is interpolated, i.e. `~/` expands to a +path relative to the home directory and `%(prefix)/` expands to a +path relative to Git's (runtime) prefix. ++ +By default, `upload-pack` refuses to lazily fetch (see the description +of the `GIT_NO_LAZY_FETCH` environment variable in +linkgit:git-upload-pack[1]), because doing so would run `git fetch`, +which may execute arbitrary commands specified in the configuration +and hooks of the served repository. Listing a repository here tells +`upload-pack` that it is trusted, so lazy fetching from the promisor +remotes configured in it is allowed. This is equivalent to setting +`GIT_NO_LAZY_FETCH` to `0` for the matching repositories. An +explicitly set `GIT_NO_LAZY_FETCH` takes precedence over this setting. ++ +Note that this allows lazy fetching from any promisor remote +configured in the served repository, not only from the promisor +remotes that the client accepted using the "promisor-remote" protocol +v2 capability (see linkgit:gitprotocol-v2[5]). The served repository +is trusted as a whole, including its configuration, so the promisor +remotes it configures are trusted too. It is the server operator's +responsibility to make sure that the promisor remotes of a trusted +repository are also trustworthy. In particular, a trusted repository +should not be configured as its own promisor remote, as `upload-pack` +would then try to lazily fetch missing objects from the repository +itself, which is pointless. ++ +As this is a multi-valued setting, you can add more than one +repository via `git config (--global|--system) --add`. To reset the +list of trusted repositories (e.g. to override any such repositories +specified in the system config), add an `uploadpack.lazyFetchTrusted` +entry with an empty value. ++ +Note that this configuration variable is only respected when it is +specified in protected configuration (see <>). This prevents +untrusted repositories from tampering with this value. diff --git a/Documentation/git-upload-pack.adoc b/Documentation/git-upload-pack.adoc index 9167a321d0..90c2ba1194 100644 --- a/Documentation/git-upload-pack.adoc +++ b/Documentation/git-upload-pack.adoc @@ -71,6 +71,11 @@ This is implemented by having `upload-pack` internally set the (because you are fetching from a partial clone, and you are sure you trust it), you can explicitly set `GIT_NO_LAZY_FETCH` to `0`. ++ +Instead of setting `GIT_NO_LAZY_FETCH` to `0` in the environment, a +server operator can allow lazy fetching on a per-repository basis by +listing trusted repositories in the `uploadpack.lazyFetchTrusted` +configuration variable. See linkgit:git-config[1]. SECURITY -------- diff --git a/Documentation/git.adoc b/Documentation/git.adoc index 8a5cdd3b3d..2e763d1f93 100644 --- a/Documentation/git.adoc +++ b/Documentation/git.adoc @@ -949,7 +949,9 @@ for full details. `GIT_NO_LAZY_FETCH`:: Setting this Boolean environment variable to true tells Git not to lazily fetch missing objects from the promisor remote - on demand. + on demand. On the server side, the `uploadpack.lazyFetchTrusted` + configuration variable can control this per-repository. See + linkgit:git-upload-pack[1]. `GIT_REFLOG_ACTION`:: When a ref is updated, reflog entries are created to keep diff --git a/builtin/upload-pack.c b/builtin/upload-pack.c index 32831fb879..8b531ca724 100644 --- a/builtin/upload-pack.c +++ b/builtin/upload-pack.c @@ -42,10 +42,13 @@ int cmd_upload_pack(int argc, OPT_END() }; unsigned enter_repo_flags = ENTER_REPO_ANY_OWNER_OK; + bool no_lazy_fetch_set; packet_trace_identity("upload-pack"); disable_replace_refs(); save_commit_buffer = 0; + + no_lazy_fetch_set = !!getenv(NO_LAZY_FETCH_ENVIRONMENT); xsetenv(NO_LAZY_FETCH_ENVIRONMENT, "1", 0); argc = parse_options(argc, argv, prefix, options, upload_pack_usage, 0); @@ -62,6 +65,14 @@ int cmd_upload_pack(int argc, if (!enter_repo(the_repository, dir, enter_repo_flags)) die("'%s' does not appear to be a git repository", dir); + /* + * Relax the GIT_NO_LAZY_FETCH=1 default if the served repo is in + * the "uploadpack.lazyFetchTrusted" protected allowlist and + * GIT_NO_LAZY_FETCH was not already set explicitly. + */ + if (!no_lazy_fetch_set && upload_pack_lazy_fetch_trusted(the_repository)) + xsetenv(NO_LAZY_FETCH_ENVIRONMENT, "0", 1); + switch (determine_protocol_version_server()) { case protocol_v2: if (advertise_refs) diff --git a/t/t5710-promisor-remote-capability.sh b/t/t5710-promisor-remote-capability.sh index 549acff23f..62f4b56006 100755 --- a/t/t5710-promisor-remote-capability.sh +++ b/t/t5710-promisor-remote-capability.sh @@ -173,6 +173,148 @@ test_expect_success "clone with promisor.acceptfromserver set to 'None'" ' initialize_server 1 "$oid" ' +test_expect_success "clone with uploadpack.lazyFetchTrusted" ' + # No promisors are advertised + git -C server config promisor.advertise false && + test_when_finished "rm -rf client" && + + # The served repo is trusted for lazy fetching + test_config_global uploadpack.lazyFetchTrusted "$(pwd)/server" && + + # Clone without GIT_NO_LAZY_FETCH=0 + git clone --no-local --filter="blob:limit=5k" server client && + + # Check that the largest object is not missing on the server + # This means the server lazy fetched it + check_missing_objects server 0 "" && + + # Reinitialize server so that the largest object is missing again + initialize_server 1 "$oid" +' + +test_expect_success "clone without uploadpack.lazyFetchTrusted fails" ' + # No promisors are advertised + git -C server config promisor.advertise false && + test_when_finished "rm -rf client" && + + # Note: no uploadpack.lazyFetchTrusted config is set here, so + # the served repo is NOT trusted for lazy fetching. + + # Clone without GIT_NO_LAZY_FETCH=0 fails + test_must_fail git clone --no-local --filter="blob:limit=5k" server client 2>err && + test_grep "lazy fetching disabled" err && + + # Check that the largest object is still missing on the server + check_missing_objects server 1 "$oid" +' + +test_expect_success "uploadpack.lazyFetchTrusted is ignored in repo config" ' + # No promisors are advertised + git -C server config promisor.advertise false && + test_when_finished "rm -rf client" && + + # The served repo is trusted for lazy fetching, but this is + # done in the repo config, not in protected config, so this is + # ignored. + test_config -C server uploadpack.lazyFetchTrusted "$(pwd)/server" && + + # Clone without GIT_NO_LAZY_FETCH=0 fails + test_must_fail git clone --no-local --filter="blob:limit=5k" server client 2>err && + test_grep "lazy fetching disabled" err && + + # Check that the largest object is still missing on the server + check_missing_objects server 1 "$oid" +' + +test_expect_success "explicit GIT_NO_LAZY_FETCH overrides uploadpack.lazyFetchTrusted" ' + # No promisors are advertised + git -C server config promisor.advertise false && + test_when_finished "rm -rf client" && + + # The served repo is trusted for lazy fetching + test_config_global uploadpack.lazyFetchTrusted "$(pwd)/server" && + + # But GIT_NO_LAZY_FETCH=1 disables lazy fetching, so clone fails + test_must_fail env GIT_NO_LAZY_FETCH=1 git clone --no-local \ + --filter="blob:limit=5k" server client 2>err && + test_grep "lazy fetching disabled" err && + + # Check that the largest object is still missing on the server + check_missing_objects server 1 "$oid" +' + +test_expect_success "trusted repo as its own promisor remote does not recurse" ' + # No promisors are advertised + git -C server config promisor.advertise false && + test_when_finished "rm -rf client" && + + # Add itself as its own remote + git -C server remote add self "$TRASH_DIRECTORY_URL/server" && + git -C server config remote.self.promisor true && + test_when_finished "git -C server remote remove self" && + + # Make "self" the only promisor remote of the server, so that it + # cannot get the missing object from "lop". Note that + # "remote.lop.partialCloneFilter" also makes "lop" a promisor + # remote, so it has to be unset too. + git -C server config --unset remote.lop.promisor && + test_when_finished "git -C server config remote.lop.promisor true" && + lop_filter="$(git -C server config remote.lop.partialCloneFilter)" && + git -C server config --unset remote.lop.partialCloneFilter && + test_when_finished "git -C server config remote.lop.partialCloneFilter \"$lop_filter\"" && + + # Allow lazy fetching from itself + test_config_global uploadpack.lazyFetchTrusted "$(pwd)/server" && + + # Check that lazy fetching fails + test_must_fail git clone --no-local --filter="blob:limit=5k" server client 2>err && + test_grep "too many nested lazy fetches" err && + + # Check that the largest object is still missing on the server + check_missing_objects server 1 "$oid" +' + +test_expect_success "uploadpack.lazyFetchTrusted needs the git dir of a non-bare repo" ' + test_when_finished "rm -rf nonbare client client2" && + + # Create a non-bare repo, without any worktree content, so that + # its largest object can be filtered out below + git init nonbare && + git -C nonbare remote add origin "$TRASH_DIRECTORY_URL/template" && + git -C nonbare fetch origin && + git -C nonbare update-ref HEAD FETCH_HEAD && + + git -C nonbare remote add lop "$TRASH_DIRECTORY_URL/lop" && + git -C nonbare config remote.lop.promisor true && + git -C nonbare config uploadpack.allowFilter true && + git -C nonbare config uploadpack.allowAnySHA1InWant true && + git -C nonbare config promisor.advertise false && + + # Repack everything, then repack without the largest object and + # create a promisor pack, like initialize_server() does + git -C nonbare -c repack.writebitmaps=false repack -a -d && + rm -f nonbare/.git/objects/pack/*.promisor && + git -C nonbare -c repack.writebitmaps=false repack -a -d \ + --filter=blob:limit=5k --filter-to="$(pwd)/nonbare-pack" && + promisor_file=$(ls nonbare/.git/objects/pack/*.pack | sed "s/\.pack/.promisor/") && + >"$promisor_file" && + check_missing_objects nonbare 1 "$oid" && + + # The worktree path does not identify the repo, so it is not + # trusted and the clone fails + test_config_global uploadpack.lazyFetchTrusted "$(pwd)/nonbare" && + test_must_fail git clone --no-local --filter="blob:limit=1k" \ + nonbare client 2>err && + test_grep "lazy fetching disabled" err && + check_missing_objects nonbare 1 "$oid" && + + # The git dir identifies the repo, so it is trusted and the + # clone succeeds + test_config_global uploadpack.lazyFetchTrusted "$(pwd)/nonbare/.git" && + git clone --no-local --filter="blob:limit=1k" nonbare client2 && + check_missing_objects nonbare 0 "" +' + test_expect_success "init + fetch with promisor.advertise set to 'true'" ' git -C server config promisor.advertise true && test_when_finished "rm -rf client" &&