git/list[1] front-page[2] threads[3] people[4] search[5] about
 

[PATCH v2 04/10] checkout-index: delay automatic setting of to_tempfile

From
Jeff King <peff@peff.net>
Date
Aug 31, 2023, 21:20 UTC
Message-ID
<20230831212051.GD949469@coredump.intra.peff.net>
In-Reply-To
<20230831211637.GA949188@coredump.intra.peff.net>

Using --stage=all requires writing to tempfiles, since we cannot put multiple stages into a single file. So --stage=all implies --temp.

But we do this by setting to_tempfile in the options callback for --stage, rather than after all options have been parsed. This leads to two bugs:

  1. If you run "checkout-index --stage=all --stage=2", this should not
     imply --temp, but it currently does. The callback cannot just unset
     to_tempfile when it sees the "2" value, because it no longer knows
     if its value was from the earlier --stage call, or if the user
     specified --temp explicitly.
  2. If you run "checkout-index --stage=all --no-temp", the --no-temp
     will overwrite the earlier implied --temp. But this mode of
     operation cannot work, and the command will fail with "<path>
     already exists" when trying to write the higher stages.

We can fix both by lazily setting to_tempfile. We'll make it a tristate, with -1 as "not yet given", and have --stage=all enable it only after all options are parsed. Likewise, after all options are parsed we can detect and reject the bogus "--no-temp" case.

Note that this does technically change the behavior for "--stage=all --no-temp" for paths which have only one stage present (which accidentally worked before, but is now forbidden). But this behavior was never intended, and you'd have to go out of your way to try to trigger it.

The new tests cover both cases, as well the general "--stage=all implies --temp", as most of the other tests explicitly say "--temp". Ironically, the test "checkout --temp within subdir" is the only one that _doesn't_ use "--temp", and so was implicitly covering this case. But it seems reasonable to have a more explicit test alongside the other related ones.

Suggested-by: Junio C Hamano <gitster@pobox.com>
Signed-off-by: Jeff King <peff@peff.net>
---
One other bug-let I noticed in this function: it only looks at the first
byte of the stage argument, so "--stage=13" is the same as "--stage=1".
We could tighten this, but I wonder if anybody does something weird like
"--stage=2ours" or something. That seems pretty unlikely, but I'd still
prefer to do it as a separate patch (if at all).
 builtin/checkout-index.c       |  8 ++++++--
 t/t2004-checkout-cache-temp.sh | 20 ++++++++++++++++++++
 2 files changed, 26 insertions(+), 2 deletions(-)
diff --git a/builtin/checkout-index.c b/builtin/checkout-index.c
index f62f13f2b5..6687a495ff 100644
--- a/builtin/checkout-index.c
+++ b/builtin/checkout-index.c
@@ -24,7 +24,7 @@
 static int nul_term_line;
 static int checkout_stage; /* default to checkout stage0 */
 static int ignore_skip_worktree; /* default to 0 */
-static int to_tempfile;
+static int to_tempfile = -1;
 static char topath[4][TEMPORARY_FILENAME_LENGTH + 1];
 
 static struct checkout state = CHECKOUT_INIT;
@@ -196,7 +196,6 @@ static int option_parse_stage(const struct option *opt,
 	BUG_ON_OPT_NEG(unset);
 
 	if (!strcmp(arg, "all")) {
-		to_tempfile = 1;
 		checkout_stage = CHECKOUT_ALL;
 	} else {
 		int ch = arg[0];
@@ -269,6 +268,11 @@ int cmd_checkout_index(int argc, const char **argv, const char *prefix)
 		state.base_dir = "";
 	state.base_dir_len = strlen(state.base_dir);
 
+	if (to_tempfile < 0)
+		to_tempfile = (checkout_stage == CHECKOUT_ALL);
+	if (!to_tempfile && checkout_stage == CHECKOUT_ALL)
+		die("--stage=all and --no-temp are incompatible");
+
 	/*
 	 * when --prefix is specified we do not want to update cache.
 	 */
diff --git a/t/t2004-checkout-cache-temp.sh b/t/t2004-checkout-cache-temp.sh
index b16d69ca4a..8eb2ef44eb 100755
--- a/t/t2004-checkout-cache-temp.sh
+++ b/t/t2004-checkout-cache-temp.sh
@@ -117,6 +117,26 @@ test_expect_success 'checkout all stages/one file to temporary files' '
 	test $(cat $s3) = tree3path1)
 '
 
+test_expect_success '--stage=all implies --temp' '
+	rm -f path* .merge_* actual &&
+	git checkout-index --stage=all -- path1 &&
+	test_path_is_missing path1
+'
+
+test_expect_success 'overriding --stage=all resets implied --temp' '
+	rm -f path* .merge_* actual &&
+	git checkout-index --stage=all --stage=2 -- path1 &&
+	echo tree2path1 >expect &&
+	test_cmp expect path1
+'
+
+test_expect_success '--stage=all --no-temp is rejected' '
+	rm -f path* .merge_* actual &&
+	test_must_fail git checkout-index --stage=all --no-temp -- path1 2>err &&
+	grep -v "already exists" err &&
+	grep "stage=all and --no-temp are incompatible" err
+'
+
 test_expect_success 'checkout some stages/one file to temporary files' '
 	rm -f path* .merge_* actual &&
 	git checkout-index --stage=all --temp -- path2 >actual &&
-- 
2.42.0.561.gaa987ecc69
Previous: Jeff KingNext: Junio C Hamano
Message 30 of 47 in “more unused parameters in parseopt callbacks”
  1. 0/8 more unused parameters in parseopt callbacksJeff King, Aug 31, 2023
  2. 1/8 merge: make xopts a strvecJeff King, Aug 31, 2023
  3. Jeff KingAug 31, 2023
  4. Phillip WoodAug 31, 2023
  5. Junio C HamanoAug 31, 2023
  6. Taylor BlauAug 31, 2023
  7. 2/8 merge: simplify parsing of "-n" optionJeff King, Aug 31, 2023
  8. Junio C HamanoAug 31, 2023
  9. 3/8 parse-options: prefer opt->value to globals in callbacksJeff King, Aug 31, 2023
  10. Junio C HamanoAug 31, 2023
  11. 4/8 parse-options: mark unused "opt" parameter in callbacksJeff King, Aug 31, 2023
  12. Junio C HamanoAug 31, 2023
  13. Jeff KingAug 31, 2023
  14. Jeff KingAug 31, 2023
  15. 5/8 merge: do not pass unused opt->value parameterJeff King, Aug 31, 2023
  16. Junio C HamanoAug 31, 2023
  17. 6/8 parse-options: add more BUG_ON() annotationsJeff King, Aug 31, 2023
  18. Junio C HamanoAug 31, 2023
  19. 7/8 interpret-trailers: mark unused "unset" parameters in option callbacksJeff King, Aug 31, 2023
  20. Junio C HamanoAug 31, 2023
  21. Jeff KingAug 31, 2023
  22. 8/8 parse-options: mark unused parameters in noop callbackJeff King, Aug 31, 2023
  23. Junio C HamanoAug 31, 2023
  24. 0/10 more unused parameters in parseopt callbacksJeff King, Aug 31, 2023
  25. 01/10 merge: make xopts a strvecJeff King, Aug 31, 2023
  26. 02/10 merge: simplify parsing of "-n" optionJeff King, Aug 31, 2023
  27. René ScharfeSep 2, 2023
  28. Jeff KingSep 5, 2023
  29. 03/10 format-patch: use OPT_STRING_LIST for to/cc optionsJeff King, Aug 31, 2023
  30. 04/10 checkout-index: delay automatic setting of to_tempfileJeff King, Aug 31, 2023
  31. Junio C HamanoAug 31, 2023
  32. René ScharfeSep 2, 2023
  33. 04/10 checkout-index: delay automatic setting of to_tempfileJeff King, Sep 5, 2023
  34. 05/10 parse-options: prefer opt->value to globals in callbacksJeff King, Aug 31, 2023
  35. René ScharfeSep 2, 2023
  36. Jeff KingSep 5, 2023
  37. 06/10 parse-options: mark unused "opt" parameter in callbacksJeff King, Aug 31, 2023
  38. René ScharfeSep 2, 2023
  39. Jeff KingSep 5, 2023
  40. René ScharfeSep 19, 2023
  41. 07/10 merge: do not pass unused opt->value parameterJeff King, Aug 31, 2023
  42. 08/10 parse-options: add more BUG_ON() annotationsJeff King, Aug 31, 2023
  43. 09/10 interpret-trailers: mark unused "unset" parameters in option callbacksJeff King, Aug 31, 2023
  44. 10/10 parse-options: mark unused parameters in noop callbackJeff King, Aug 31, 2023
  45. René ScharfeSep 2, 2023
  46. Jeff KingSep 5, 2023
  47. René ScharfeSep 7, 2023

Read the whole thread, see it on lore, or plain text.

$ cat FOOTERMessages come from the public archive at lore.kernel.org/git, fetched every hour. The front page is chosen and written each morning by an AI editor and can be wrong; the threads themselves are the record. About and API. For agents: an MCP server at https://gitlist.dev/mcp, and any thread, story or person page as Markdown by adding .md to its URL (or sending Accept: text/markdown). Details in /llms.txt.