git.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: "Shawn O. Pearce" <spearce@spearce.org>
To: Robin Rosenberg <robin.rosenberg@dewire.com>
Cc: git@vger.kernel.org
Subject: [JGIT PATCH 10/13] Add compare-and-swap semantics to RefUpdate
Date: Mon, 22 Dec 2008 16:27:20 -0800	[thread overview]
Message-ID: <1229992043-1053-11-git-send-email-spearce@spearce.org> (raw)
In-Reply-To: <1229992043-1053-10-git-send-email-spearce@spearce.org>

This permits the caller to set the value it expects to find in the
Ref upon obtaining the update lock.  If the ref value doesn't match
then the update is aborted with Result.LOCK_FAILURE, so the caller
can recover gracefully without making unexpected changes.

Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
---
 .../src/org/spearce/jgit/lib/RefUpdate.java        |   30 ++++++++++++++++++++
 1 files changed, 30 insertions(+), 0 deletions(-)

diff --git a/org.spearce.jgit/src/org/spearce/jgit/lib/RefUpdate.java b/org.spearce.jgit/src/org/spearce/jgit/lib/RefUpdate.java
index 235c2fd..0c9ce91 100644
--- a/org.spearce.jgit/src/org/spearce/jgit/lib/RefUpdate.java
+++ b/org.spearce.jgit/src/org/spearce/jgit/lib/RefUpdate.java
@@ -149,6 +149,9 @@
 	/** Old value of the ref, obtained after we lock it. */
 	private ObjectId oldValue;
 
+	/** If non-null, the value {@link #oldValue} must have to continue. */
+	private ObjectId expValue;
+
 	/** Result of the update operation. */
 	private Result result = Result.NOT_ATTEMPTED;
 
@@ -190,6 +193,27 @@ public void setNewObjectId(final AnyObjectId id) {
 	}
 
 	/**
+	 * @return the expected value of the ref after the lock is taken, but before
+	 *         update occurs. Null to avoid the compare and swap test. Use
+	 *         {@link ObjectId#zeroId()} to indicate expectation of a
+	 *         non-existant ref.
+	 */
+	public ObjectId getExpectedOldObjectId() {
+		return expValue;
+	}
+
+	/**
+	 * @param id
+	 *            the expected value of the ref after the lock is taken, but
+	 *            before update occurs. Null to avoid the compare and swap test.
+	 *            Use {@link ObjectId#zeroId()} to indicate expectation of a
+	 *            non-existant ref.
+	 */
+	public void setExpectedOldObjectId(final AnyObjectId id) {
+		expValue = id != null ? id.toObjectId() : null;
+	}
+
+	/**
 	 * Check if this update wants to forcefully change the ref.
 	 * 
 	 * @return true if this update should ignore merge tests.
@@ -370,6 +394,12 @@ private Result updateImpl(final RevWalk walk, final Store store)
 			return Result.LOCK_FAILURE;
 		try {
 			oldValue = db.idOf(getName());
+			if (expValue != null) {
+				final ObjectId o;
+				o = oldValue != null ? oldValue : ObjectId.zeroId();
+				if (!expValue.equals(o))
+					return Result.LOCK_FAILURE;
+			}
 			if (oldValue == null)
 				return store.store(lock, Result.NEW);
 
-- 
1.6.1.rc4.301.g5497a

  reply	other threads:[~2008-12-23  0:30 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2008-12-23  0:27 [JGIT PATCH 00/13] Add receive-pack server side support Shawn O. Pearce
2008-12-23  0:27 ` [JGIT PATCH 01/13] Fix invalid "double checked locking" in InflaterCache Shawn O. Pearce
2008-12-23  0:27   ` [JGIT PATCH 02/13] Cleanup stupid release of the cached Inflater in IndexPack Shawn O. Pearce
2008-12-23  0:27     ` [JGIT PATCH 03/13] Cache an Inflater inside a WindowCursor and reuse it as much as possible Shawn O. Pearce
2008-12-23  0:27       ` [JGIT PATCH 04/13] Make RefDatabase thread-safe Shawn O. Pearce
2008-12-23  0:27         ` [JGIT PATCH 05/13] Make PackFile thread-safe Shawn O. Pearce
2008-12-23  0:27           ` [JGIT PATCH 06/13] Make Repository thread-safe Shawn O. Pearce
2008-12-23  0:27             ` [JGIT PATCH 07/13] Don't open a PackFile multiple times on scanForPacks Shawn O. Pearce
2008-12-23  0:27               ` [JGIT PATCH 08/13] Expose RepositoryConfig.getBoolean so applications can use it Shawn O. Pearce
2008-12-23  0:27                 ` [JGIT PATCH 09/13] Add AnyObjectId.copyTo(StringBuilder) Shawn O. Pearce
2008-12-23  0:27                   ` Shawn O. Pearce [this message]
2008-12-23  0:27                     ` [JGIT PATCH 11/13] Allow null new ObjectId during RefUpdate.delete Shawn O. Pearce
2008-12-23  0:27                       ` [JGIT PATCH 12/13] Implement the git-receive-pack process in Java Shawn O. Pearce
2008-12-23  0:27                         ` [JGIT PATCH 13/13] Add basic git daemon support to publish receive-pack Shawn O. Pearce
2009-01-03 23:48                           ` Robin Rosenberg
2009-01-05  2:46                             ` [PATCH] Permit a wider range of repository names in jgit daemon requests Shawn O. Pearce
2009-01-05 23:07                               ` Robin Rosenberg

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=1229992043-1053-11-git-send-email-spearce@spearce.org \
    --to=spearce@spearce.org \
    --cc=git@vger.kernel.org \
    --cc=robin.rosenberg@dewire.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).