Merge branch 'kn/fetch-push-bulk-ref-update' into jch

"git push" and "git fetch" are taught to update refs in batches to
gain performance.

* kn/fetch-push-bulk-ref-update:
  receive-pack: use batched reference updates
  send-pack: fix memory leak around duplicate refs
  fetch: use batched reference updates
  refs: add function to translate errors to strings

Author: gitster

builtin/fetch.c

@@ -640,17 +640,13 @@ static struct ref *get_ref_map(struct remote *remote,
 	return ref_map;
 }
 
-#define STORE_REF_ERROR_OTHER 1
-#define STORE_REF_ERROR_DF_CONFLICT 2
-
 static int s_update_ref(const char *action,
 			struct ref *ref,
 			struct ref_transaction *transaction,
 			int check_old)
 {
 	char *msg;
 	char *rla = getenv("GIT_REFLOG_ACTION");
-	struct ref_transaction *our_transaction = NULL;
 	struct strbuf err = STRBUF_INIT;
 	int ret;
 
@@ -660,43 +656,10 @@ static int s_update_ref(const char *action,
 		rla = default_rla.buf;
 	msg = xstrfmt("%s: %s", rla, action);
 
-	/*
-	 * If no transaction was passed to us, we manage the transaction
-	 * ourselves. Otherwise, we trust the caller to handle the transaction
-	 * lifecycle.
-	 */
-	if (!transaction) {
-		transaction = our_transaction = ref_store_transaction_begin(get_main_ref_store(the_repository),
-									    0, &err);
-		if (!transaction) {
-			ret = STORE_REF_ERROR_OTHER;
-			goto out;
-		}
-	}
-
 	ret = ref_transaction_update(transaction, ref->name, &ref->new_oid,
 				     check_old ? &ref->old_oid : NULL,
 				     NULL, NULL, 0, msg, &err);
-	if (ret) {
-		ret = STORE_REF_ERROR_OTHER;
-		goto out;
-	}
-
-	if (our_transaction) {
-		switch (ref_transaction_commit(our_transaction, &err)) {
-		case 0:
-			break;
-		case REF_TRANSACTION_ERROR_NAME_CONFLICT:
-			ret = STORE_REF_ERROR_DF_CONFLICT;
-			goto out;
-		default:
-			ret = STORE_REF_ERROR_OTHER;
-			goto out;
-		}
-	}
 
-out:
-	ref_transaction_free(our_transaction);
 	if (ret)
 		error("%s", err.buf);
 	strbuf_release(&err);
@@ -1139,7 +1102,6 @@ N_("it took %.2f seconds to check forced updates; you can use\n"
    "to avoid this check\n");
 
 static int store_updated_refs(struct display_state *display_state,
-			      const char *remote_name,
 			      int connectivity_checked,
 			      struct ref_transaction *transaction, struct ref *ref_map,
 			      struct fetch_head *fetch_head,
@@ -1277,11 +1239,6 @@ static int store_updated_refs(struct display_state *display_state,
 		}
 	}
 
-	if (rc & STORE_REF_ERROR_DF_CONFLICT)
-		error(_("some local refs could not be updated; try running\n"
-		      " 'git remote prune %s' to remove any old, conflicting "
-		      "branches"), remote_name);
-
 	if (advice_enabled(ADVICE_FETCH_SHOW_FORCED_UPDATES)) {
 		if (!config->show_forced_updates) {
 			warning(_(warn_show_forced_updates));
@@ -1365,9 +1322,8 @@ static int fetch_and_consume_refs(struct display_state *display_state,
 	}
 
 	trace2_region_enter("fetch", "consume_refs", the_repository);
-	ret = store_updated_refs(display_state, transport->remote->name,
-				 connectivity_checked, transaction, ref_map,
-				 fetch_head, config);
+	ret = store_updated_refs(display_state, connectivity_checked,
+				 transaction, ref_map, fetch_head, config);
 	trace2_region_leave("fetch", "consume_refs", the_repository);
 
 out:
@@ -1687,6 +1643,37 @@ static int set_head(const struct ref *remote_refs, struct remote *remote)
 	return result;
 }
 
+struct ref_rejection_data {
+	int *retcode;
+	int conflict_msg_shown;
+	const char *remote_name;
+};
+
+static void ref_transaction_rejection_handler(const char *refname,
+					      const struct object_id *old_oid UNUSED,
+					      const struct object_id *new_oid UNUSED,
+					      const char *old_target UNUSED,
+					      const char *new_target UNUSED,
+					      enum ref_transaction_error err,
+					      void *cb_data)
+{
+	struct ref_rejection_data *data = (struct ref_rejection_data *)cb_data;
+
+	if (err == REF_TRANSACTION_ERROR_NAME_CONFLICT && !data->conflict_msg_shown) {
+		error(_("some local refs could not be updated; try running\n"
+			" 'git remote prune %s' to remove any old, conflicting "
+			"branches"), data->remote_name);
+		data->conflict_msg_shown = 1;
+	} else {
+		char *reason = ref_transaction_error_msg(err);
+
+		error(_("fetching ref %s failed: %s"), refname, reason);
+		free(reason);
+	}
+
+	*data->retcode = 1;
+}
+
 static int do_fetch(struct transport *transport,
 		    struct refspec *rs,
 		    const struct fetch_config *config)
@@ -1807,6 +1794,24 @@ static int do_fetch(struct transport *transport,
 			retcode = 1;
 	}
 
+	/*
+	 * If not atomic, we can still use batched updates, which would be much
+	 * more performant. We don't initiate the transaction before pruning,
+	 * since pruning must be an independent step, to avoid F/D conflicts.
+	 *
+	 * TODO: if reference transactions gain logical conflict resolution, we
+	 * can delete and create refs (with F/D conflicts) in the same transaction
+	 * and this can be moved about the 'prune_refs()' block.
+	 */
+	if (!transaction) {
+		transaction = ref_store_transaction_begin(get_main_ref_store(the_repository),
+							  REF_TRANSACTION_ALLOW_FAILURE, &err);
+		if (!transaction) {
+			retcode = -1;
+			goto cleanup;
+		}
+	}
+
 	if (fetch_and_consume_refs(&display_state, transport, transaction, ref_map,
 				   &fetch_head, config)) {
 		retcode = 1;
@@ -1838,16 +1843,31 @@ static int do_fetch(struct transport *transport,
 		free_refs(tags_ref_map);
 	}
 
-	if (transaction) {
-		if (retcode)
-			goto cleanup;
+	if (retcode)
+		goto cleanup;
 
-		retcode = ref_transaction_commit(transaction, &err);
+	retcode = ref_transaction_commit(transaction, &err);
+	if (retcode) {
+		/*
+		 * Explicitly handle transaction cleanup to avoid
+		 * aborting an already closed transaction.
+		 */
+		ref_transaction_free(transaction);
+		transaction = NULL;
+		goto cleanup;
+	}
+
+	if (!atomic_fetch) {
+		struct ref_rejection_data data = {
+			.retcode = &retcode,
+			.conflict_msg_shown = 0,
+			.remote_name = transport->remote->name,
+		};
+
+		ref_transaction_for_each_rejected_update(transaction,
+							 ref_transaction_rejection_handler,
+							 &data);
 		if (retcode) {
-			/*
-			 * Explicitly handle transaction cleanup to avoid
-			 * aborting an already closed transaction.
-			 */
 			ref_transaction_free(transaction);
 			transaction = NULL;
 			goto cleanup;

builtin/receive-pack.c

@@ -1845,35 +1845,67 @@ static void BUG_if_skipped_connectivity_check(struct command *commands,
 	BUG_if_bug("connectivity check skipped???");
 }
 
+static void ref_transaction_rejection_handler(const char *refname,
+					      const struct object_id *old_oid UNUSED,
+					      const struct object_id *new_oid UNUSED,
+					      const char *old_target UNUSED,
+					      const char *new_target UNUSED,
+					      enum ref_transaction_error err,
+					      void *cb_data)
+{
+	struct strmap *failed_refs = cb_data;
+
+	strmap_put(failed_refs, refname, ref_transaction_error_msg(err));
+}
+
 static void execute_commands_non_atomic(struct command *commands,
 					struct shallow_info *si)
 {
 	struct command *cmd;
 	struct strbuf err = STRBUF_INIT;
+	const char *reported_error = NULL;
+	struct strmap failed_refs = STRMAP_INIT;
+
+	transaction = ref_store_transaction_begin(get_main_ref_store(the_repository),
+						  REF_TRANSACTION_ALLOW_FAILURE, &err);
+	if (!transaction) {
+		rp_error("%s", err.buf);
+		strbuf_reset(&err);
+		reported_error = "transaction failed to start";
+		goto failure;
+	}
 
 	for (cmd = commands; cmd; cmd = cmd->next) {
 		if (!should_process_cmd(cmd) || cmd->run_proc_receive)
 			continue;
 
-		transaction = ref_store_transaction_begin(get_main_ref_store(the_repository),
-							  0, &err);
-		if (!transaction) {
-			rp_error("%s", err.buf);
-			strbuf_reset(&err);
-			cmd->error_string = "transaction failed to start";
-			continue;
-		}
-
 		cmd->error_string = update(cmd, si);
+	}
 
-		if (!cmd->error_string
-		    && ref_transaction_commit(transaction, &err)) {
-			rp_error("%s", err.buf);
-			strbuf_reset(&err);
-			cmd->error_string = "failed to update ref";
-		}
-		ref_transaction_free(transaction);
+	if (ref_transaction_commit(transaction, &err)) {
+		rp_error("%s", err.buf);
+		reported_error = "failed to update refs";
+		goto failure;
+	}
+
+	ref_transaction_for_each_rejected_update(transaction,
+						 ref_transaction_rejection_handler,
+						 &failed_refs);
+
+	if (strmap_empty(&failed_refs))
+		goto cleanup;
+
+failure:
+	for (cmd = commands; cmd; cmd = cmd->next) {
+		if (reported_error)
+			cmd->error_string = reported_error;
+		else if (strmap_contains(&failed_refs, cmd->ref_name))
+			cmd->error_string = xstrdup(strmap_get(&failed_refs, cmd->ref_name));
 	}
+
+cleanup:
+	ref_transaction_free(transaction);
+	strmap_clear(&failed_refs, 1);
 	strbuf_release(&err);
 }
 

builtin/update-ref.c

@@ -575,37 +575,15 @@ static void print_rejected_refs(const char *refname,
 				void *cb_data UNUSED)
 {
 	struct strbuf sb = STRBUF_INIT;
-	const char *reason = "";
-
-	switch (err) {
-	case REF_TRANSACTION_ERROR_NAME_CONFLICT:
-		reason = "refname conflict";
-		break;
-	case REF_TRANSACTION_ERROR_CREATE_EXISTS:
-		reason = "reference already exists";
-		break;
-	case REF_TRANSACTION_ERROR_NONEXISTENT_REF:
-		reason = "reference does not exist";
-		break;
-	case REF_TRANSACTION_ERROR_INCORRECT_OLD_VALUE:
-		reason = "incorrect old value provided";
-		break;
-	case REF_TRANSACTION_ERROR_INVALID_NEW_VALUE:
-		reason = "invalid new value provided";
-		break;
-	case REF_TRANSACTION_ERROR_EXPECTED_SYMREF:
-		reason = "expected symref but found regular ref";
-		break;
-	default:
-		reason = "unkown failure";
-	}
+	char *reason = ref_transaction_error_msg(err);
 
 	strbuf_addf(&sb, "rejected %s %s %s %s\n", refname,
 		    new_oid ? oid_to_hex(new_oid) : new_target,
 		    old_oid ? oid_to_hex(old_oid) : old_target,
 		    reason);
 
 	fwrite(sb.buf, sb.len, 1, stdout);
+	free(reason);
 	strbuf_release(&sb);
 }
 

refs.c

@@ -3314,3 +3314,33 @@ int ref_update_expects_existing_old_ref(struct ref_update *update)
 	return (update->flags & REF_HAVE_OLD) &&
 		(!is_null_oid(&update->old_oid) || update->old_target);
 }
+
+char *ref_transaction_error_msg(enum ref_transaction_error err)
+{
+	const char *reason = "";
+
+	switch (err) {
+	case REF_TRANSACTION_ERROR_NAME_CONFLICT:
+		reason = "refname conflict";
+		break;
+	case REF_TRANSACTION_ERROR_CREATE_EXISTS:
+		reason = "reference already exists";
+		break;
+	case REF_TRANSACTION_ERROR_NONEXISTENT_REF:
+		reason = "reference does not exist";
+		break;
+	case REF_TRANSACTION_ERROR_INCORRECT_OLD_VALUE:
+		reason = "incorrect old value provided";
+		break;
+	case REF_TRANSACTION_ERROR_INVALID_NEW_VALUE:
+		reason = "invalid new value provided";
+		break;
+	case REF_TRANSACTION_ERROR_EXPECTED_SYMREF:
+		reason = "expected symref but found regular ref";
+		break;
+	default:
+		reason = "unkown failure";
+	}
+
+	return xstrdup(reason);
+}

refs.h

@@ -907,6 +907,11 @@ void ref_transaction_for_each_rejected_update(struct ref_transaction *transactio
 					      ref_transaction_for_each_rejected_update_fn cb,
 					      void *cb_data);
 
+/*
+ * Translate errors to human readable error messages.
+ */
+char *ref_transaction_error_msg(enum ref_transaction_error err);
+
 /*
  * Free `*transaction` and all associated data.
  */

send-pack.c

@@ -257,6 +257,13 @@ static int receive_status(struct repository *r,
 				refname);
 			continue;
 		}
+
+		/*
+		 * Clients sending duplicate refs can cause the same value
+		 * to be overridden, causing a memory leak.
+		 */
+		free(hint->remote_status);
+
 		if (!strcmp(head, "ng")) {
 			hint->status = REF_STATUS_REMOTE_REJECT;
 			if (p)

t/t1416-ref-transaction-hooks.sh

@@ -120,8 +120,6 @@ test_expect_success 'interleaving hook calls succeed' '
 
 	cat >expect <<-EOF &&
 		hooks/update refs/tags/PRE $ZERO_OID $PRE_OID
-		hooks/reference-transaction prepared
-		hooks/reference-transaction committed
 		hooks/update refs/tags/POST $ZERO_OID $POST_OID
 		hooks/reference-transaction prepared
 		hooks/reference-transaction committed