git.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Junio C Hamano <gitster@pobox.com>
To: Salikh Zakirov <salikh@gmail.com>
Cc: git@vger.kernel.org
Subject: [PATCH] Fix "git add -u" data corruption.
Date: Wed, 15 Aug 2007 14:12:14 -0700	[thread overview]
Message-ID: <7v1we4pknl.fsf_-_@gitster.siamese.dyndns.org> (raw)
In-Reply-To: <7v643gplph.fsf@gitster.siamese.dyndns.org> (Junio C. Hamano's message of "Wed, 15 Aug 2007 13:49:30 -0700")

This applies to 'maint' to fix a rather serious data corruption
issue.  When "git add -u" affects a subdirectory in such a way
that the only changes to its contents are path removals, the
next tree object written out of that index was bogus, as the
remove codepath forgot to invalidate the cache-tree entry.

Kodos for noticing this breakage goes to Salikh Zakirov.

Signed-off-by: Junio C Hamano <gitster@pobox.com>
---
 builtin-add.c         |    1 +
 t/t2200-add-update.sh |   59 +++++++++++++++++++++++++++++++++++-------------
 2 files changed, 44 insertions(+), 16 deletions(-)

diff --git a/builtin-add.c b/builtin-add.c
index 1591171..a5fae7c 100644
--- a/builtin-add.c
+++ b/builtin-add.c
@@ -115,6 +115,7 @@ static void update_callback(struct diff_queue_struct *q,
 			break;
 		case DIFF_STATUS_DELETED:
 			remove_file_from_cache(path);
+			cache_tree_invalidate_path(active_cache_tree, path);
 			if (verbose)
 				printf("remove '%s'\n", path);
 			break;
diff --git a/t/t2200-add-update.sh b/t/t2200-add-update.sh
index 83005e7..4c7c6af 100755
--- a/t/t2200-add-update.sh
+++ b/t/t2200-add-update.sh
@@ -13,26 +13,53 @@ only the updates to dir/sub.'
 
 . ./test-lib.sh
 
-test_expect_success 'setup' '
-echo initial >top &&
-mkdir dir &&
-echo initial >dir/sub &&
-git-add dir/sub top &&
-git-commit -m initial &&
-echo changed >top &&
-echo changed >dir/sub &&
-echo other >dir/other
+test_expect_success setup '
+	echo initial >check &&
+	echo initial >top &&
+	mkdir dir1 dir2 &&
+	echo initial >dir1/sub1 &&
+	echo initial >dir1/sub2 &&
+	echo initial >dir2/sub3 &&
+	git add check dir1 dir2 top &&
+	test_tick
+	git-commit -m initial &&
+
+	echo changed >check &&
+	echo changed >top &&
+	echo changed >dir2/sub3 &&
+	rm -f dir1/sub1 &&
+	echo other >dir2/other
+'
+
+test_expect_success update '
+	git add -u dir1 dir2
 '
 
-test_expect_success 'update' 'git-add -u dir'
+test_expect_success 'update noticed a removal' '
+	test "$(git-ls-files dir1/sub1)" = ""
+'
 
-test_expect_success 'update touched correct path' \
-  'test "`git-diff-files --name-status dir/sub`" = ""'
+test_expect_success 'update touched correct path' '
+	test "$(git-diff-files --name-status dir2/sub3)" = ""
+'
 
-test_expect_success 'update did not touch other tracked files' \
-  'test "`git-diff-files --name-status top`" = "M	top"'
+test_expect_success 'update did not touch other tracked files' '
+	test "$(git-diff-files --name-status check)" = "M	check" &&
+	test "$(git-diff-files --name-status top)" = "M	top"
+'
 
-test_expect_success 'update did not touch untracked files' \
-  'test "`git-diff-files --name-status dir/other`" = ""'
+test_expect_success 'update did not touch untracked files' '
+	test "$(git-ls-files dir2/other)" = ""
+'
+
+test_expect_success 'cache tree has not been corrupted' '
+
+	git ls-files -s |
+	sed -e "s/ 0	/	/" >expect &&
+	git ls-tree -r $(git write-tree) |
+	sed -e "s/ blob / /" >current &&
+	diff -u expect current
+
+'
 
 test_done

  reply	other threads:[~2007-08-15 21:12 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2007-08-15 14:16 git-commit goes awry after git-add -u Salikh Zakirov
2007-08-15 14:21 ` Salikh Zakirov
2007-08-15 20:49 ` Junio C Hamano
2007-08-15 21:12   ` Junio C Hamano [this message]
2007-08-16  0:15     ` [PATCH] Fix "git add -u" data corruption Zakirov Salikh
2007-08-15 23:43   ` git-commit goes awry after git-add -u Salikh Zakirov

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=7v1we4pknl.fsf_-_@gitster.siamese.dyndns.org \
    --to=gitster@pobox.com \
    --cc=git@vger.kernel.org \
    --cc=salikh@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).