git.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Yiannis Marangos <yiannis.marangos@gmail.com>
To: git@vger.kernel.org
Cc: Yiannis Marangos <yiannis.marangos@gmail.com>
Subject: [PATCH v6] Verify index file before we opportunistically update it
Date: Thu, 10 Apr 2014 16:11:42 +0300	[thread overview]
Message-ID: <1397135502-9666-1-git-send-email-yiannis.marangos@gmail.com> (raw)
In-Reply-To: <1397081197-14803-1-git-send-email-yiannis.marangos@gmail.com>

Before we proceed to "opportunistic update" we must verify that the
current index file is the same as the one that we read before. There
is a possible race if we don't do this. In the example below git-status
does "opportunistic update" and git-rebase updates the index, but the
race can happen in general.

  1. process A calls git-rebase (or does anything that uses the index)

  2. process A applies 1st commit

  3. process B calls git-status (or does anything that updates the index)

  4. process B reads index

  5. process A applies 2nd commit

  6. process B takes the lock, then overwrites process A's changes.

  7. process A applies 3rd commit

As an end result the 3rd commit will have a revert of the 2nd commit.
When process B takes the lock, it needs to make sure that the index
hasn't changed since step 4.

Signed-off-by: Yiannis Marangos <yiannis.marangos@gmail.com>
---

Version 4 contains fixes based on Junio's comments.
Version 5 fixs a typo in commit message (git-show -> git-status)
Version 6 removes verify_hdr() and use pread() instead of mmap().

 cache.h      |  1 +
 read-cache.c | 47 ++++++++++++++++++++++++++++++++++++++++++++++-
 2 files changed, 47 insertions(+), 1 deletion(-)

diff --git a/cache.h b/cache.h
index 107ac61..0460f06 100644
--- a/cache.h
+++ b/cache.h
@@ -279,6 +279,7 @@ struct index_state {
 		 initialized : 1;
 	struct hashmap name_hash;
 	struct hashmap dir_hash;
+	unsigned char sha1[20];
 };
 
 extern struct index_state the_index;
diff --git a/read-cache.c b/read-cache.c
index ba13353..034865d 100644
--- a/read-cache.c
+++ b/read-cache.c
@@ -1477,6 +1477,7 @@ int read_index_from(struct index_state *istate, const char *path)
 	if (verify_hdr(hdr, mmap_size) < 0)
 		goto unmap;
 
+	hashcpy(istate->sha1, (unsigned char *)hdr + mmap_size - 20);
 	istate->version = ntohl(hdr->hdr_version);
 	istate->cache_nr = ntohl(hdr->hdr_entries);
 	istate->cache_alloc = alloc_nr(istate->cache_nr);
@@ -1760,6 +1761,50 @@ static int ce_write_entry(git_SHA_CTX *c, int fd, struct cache_entry *ce,
 	return result;
 }
 
+/*
+ * This function verifies if index_state has the correct sha1 of an index file.
+ * Don't die if we have any other failure, just return 0.
+ */
+static int verify_index_from(const struct index_state *istate, const char *path)
+{
+	int fd;
+	ssize_t n;
+	struct stat st;
+	unsigned char sha1[20];
+
+	if (!istate->initialized)
+		return 0;
+
+	fd = open(path, O_RDONLY);
+	if (fd < 0)
+		return 0;
+
+	if (fstat(fd, &st))
+		goto out;
+
+	if (st.st_size < sizeof(struct cache_header) + 20)
+		goto out;
+
+	n = pread(fd, sha1, 20, st.st_size - 20);
+	if (n != 20)
+		goto out;
+
+	if (hashcmp(istate->sha1, sha1))
+		goto out;
+
+	close(fd);
+	return 1;
+
+out:
+	close(fd);
+	return 0;
+}
+
+static int verify_index(const struct index_state *istate)
+{
+	return verify_index_from(istate, get_index_file());
+}
+
 static int has_racy_timestamp(struct index_state *istate)
 {
 	int entries = istate->cache_nr;
@@ -1779,7 +1824,7 @@ static int has_racy_timestamp(struct index_state *istate)
 void update_index_if_able(struct index_state *istate, struct lock_file *lockfile)
 {
 	if ((istate->cache_changed || has_racy_timestamp(istate)) &&
-	    !write_index(istate, lockfile->fd))
+	    verify_index(istate) && !write_index(istate, lockfile->fd))
 		commit_locked_index(lockfile);
 	else
 		rollback_lock_file(lockfile);
-- 
1.9.1

  parent reply	other threads:[~2014-04-10 13:12 UTC|newest]

Thread overview: 31+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-04-09 22:06 [PATCH] Verify index file before we opportunistically update it Yiannis Marangos
2014-04-09 22:06 ` Yiannis Marangos
2014-04-09 22:34 ` [PATCH v2] " Yiannis Marangos
2014-04-09 23:05   ` Junio C Hamano
2014-04-09 22:52 ` [PATCH v3] " Yiannis Marangos
2014-04-10  5:22 ` [PATCH v4] " Yiannis Marangos
2014-04-10  5:34 ` [PATCH v5] " Yiannis Marangos
2014-04-10 10:40   ` Duy Nguyen
2014-04-10 11:57     ` Yiannis Marangos
2014-04-10 16:57     ` Junio C Hamano
2014-04-10 13:11 ` Yiannis Marangos [this message]
2014-04-10 18:31 ` [PATCH v7 1/2] Add xpread() and xpwrite() Yiannis Marangos
2014-04-10 18:31   ` [PATCH v7 2/2] Verify index file before we opportunistically update it Yiannis Marangos
2014-04-10 19:28     ` Junio C Hamano
2014-04-11  2:57       ` Duy Nguyen
2014-04-11 19:24         ` Junio C Hamano
2014-04-11 20:43           ` Junio C Hamano
2014-04-11 23:30             ` Yiannis Marangos
2014-04-12  0:10             ` Duy Nguyen
2014-04-12  4:19               ` Junio C Hamano
2014-04-12  7:05                 ` Junio C Hamano
2014-04-12 10:13                 ` Duy Nguyen
2014-04-14 18:50                   ` Junio C Hamano
2014-04-11  7:47       ` Torsten Bögershausen
2014-04-11 15:58         ` Yiannis Marangos
2014-04-11 10:36       ` Duy Nguyen
2014-04-10 18:35   ` [PATCH v7 1/2] Add xpread() and xpwrite() Junio C Hamano
2014-04-10 18:44     ` Yiannis Marangos
2014-04-10 18:54 ` [PATCH v8 1/2] Add xpread() Yiannis Marangos
2014-04-10 19:12   ` Johannes Sixt
2014-04-10 19:20     ` Junio C Hamano

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=1397135502-9666-1-git-send-email-yiannis.marangos@gmail.com \
    --to=yiannis.marangos@gmail.com \
    --cc=git@vger.kernel.org \
    /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).