assistant: Notice on startup when the index file is corrupt, and auto-repair.

Assistant/Threads/SanityChecker.hs

@@ -25,8 +25,10 @@ import Utility.Batch
 import Utility.NotificationBroadcaster
 import Config
 import Utility.HumanTime
+import Git.Repair
 
 import Data.Time.Clock.POSIX
+import qualified Data.Set as S
 
 {- This thread runs once at startup, and most other threads wait for it
  - to finish. (However, the webapp thread does not, to prevent the UI
@@ -36,6 +38,16 @@ sanityCheckerStartupThread startupdelay = namedThreadUnchecked "SanityCheckerSta
 	{- Stale git locks can prevent commits from happening, etc. -}
 	void $ repairStaleGitLocks =<< liftAnnex gitRepo
 
+	{- A corrupt index file can prevent the assistant from working at
+	 - all, so detect and repair. -}
+	unlessM (liftAnnex $ inRepo $ checkIndex S.empty) $ do
+		debug ["corrupt index found at startup; removing"]
+		liftAnnex $ inRepo nukeIndex
+		{- Normally the startup scan avoids re-staging files,
+		 - but with the index deleted, everything needs to be
+		 - restaged. -}
+		modifyDaemonStatus_ $ \s -> s { forceRestage = True }
+
 	{- If there's a startup delay, it's done here. -}
 	liftIO $ maybe noop (threadDelaySeconds . Seconds . fromIntegral . durationSeconds) startupdelay
 

Assistant/Threads/Watcher.hs

@@ -200,6 +200,9 @@ onAdd matcher file filestatus
 			add matcher file
 	| otherwise = noChange
 
+shouldRestage :: DaemonStatus -> Bool
+shouldRestage ds = scanComplete ds || forceRestage ds
+
 {- In direct mode, add events are received for both new files, and
  - modified existing files.
  -}
@@ -214,7 +217,7 @@ onAddDirect symlinkssupported matcher file fs = do
 				 - really modified, but it might have
 				 - just been deleted and been put back,
 				 - so it symlink is restaged to make sure. -}
-				( ifM (scanComplete <$> getDaemonStatus)
+				( ifM (shouldRestage <$> getDaemonStatus)
 					( do
 						link <- liftAnnex $ inRepo $ gitAnnexLink file key
 						addLink file link (Just key)
@@ -286,7 +289,7 @@ onAddSymlink' linktarget mk isdirect file filestatus = go mk
 	 - links too.)
 	 -}
 	ensurestaged (Just link) daemonstatus
-		| scanComplete daemonstatus = addLink file link mk
+		| shouldRestage daemonstatus = addLink file link mk
 		| otherwise = case filestatus of
 			Just s
 				| not (afterLastDaemonRun (statusChangeTime s) daemonstatus) -> noChange

Assistant/Types/DaemonStatus.hs

@@ -28,6 +28,8 @@ data DaemonStatus = DaemonStatus
 	{ startedThreads :: M.Map ThreadName (Async (), IO ())
 	-- False when the daemon is performing its startup scan
 	, scanComplete :: Bool
+	-- True when all files should be restaged.
+	, forceRestage :: Bool
 	-- Time when a previous process of the daemon was running ok
 	, lastRunning :: Maybe POSIXTime
 	-- True when the daily sanity checker is running
@@ -81,6 +83,7 @@ newDaemonStatus :: IO DaemonStatus
 newDaemonStatus = DaemonStatus
 	<$> pure M.empty
 	<*> pure False
+	<*> pure False
 	<*> pure Nothing
 	<*> pure False
 	<*> pure Nothing

Git/Repair.hs

@@ -13,6 +13,7 @@ module Git.Repair (
 	resetLocalBranches,
 	removeTrackingBranches,
 	checkIndex,
+	nukeIndex,
 	emptyGoodCommits,
 ) where
 
@@ -368,15 +369,18 @@ verifyTree missing treesha r
 			else cleanup
 
 {- Checks that the index file only refers to objects that are not missing,
- - and is not itself corrupt. -}
+ - and is not itself corrupt or missing. -}
 checkIndex :: MissingObjects -> Repo -> IO Bool
-checkIndex missing r = do
+checkIndex missing r = ifM (doesFileExist (localGitDir r </> "index"))
+	( do
 		(bad, _good, cleanup) <- partitionIndex missing r
 		if null bad
 			then cleanup
 			else do
 				void cleanup
 				return False
+	, return False
+	)
 
 partitionIndex :: MissingObjects -> Repo -> IO ([LsFiles.StagedDetails], [LsFiles.StagedDetails], IO Bool)
 partitionIndex missing r = do
@@ -396,7 +400,7 @@ rewriteIndex missing r
 	| otherwise = do
 		(bad, good, cleanup) <- partitionIndex missing r
 		unless (null bad) $ do
-			nukeFile (localGitDir r </> "index")
+			nukeIndex r
 			UpdateIndex.streamUpdateIndex r
 				=<< (catMaybes <$> mapM reinject good)
 		void cleanup
@@ -408,6 +412,9 @@ rewriteIndex missing r
 			UpdateIndex.stageFile sha blobtype file r
 	reinject _ = return Nothing
 
+nukeIndex :: Repo -> IO ()
+nukeIndex r = nukeFile (localGitDir r </> "index")
+
 newtype GoodCommits = GoodCommits (S.Set Sha)
 
 emptyGoodCommits :: GoodCommits
@@ -502,7 +509,7 @@ runRepairOf fsckresult forced referencerepo g = do
 				return (True, stillmissing, modifiedbranches)
 	
 	corruptedindex = do
-		nukeFile (localGitDir g </> "index")
+		nukeIndex g
 		putStrLn "Removed the corrupted index file. You should look at what files are present in your working tree and git add them back to the index when appropriate."
 		return (True, S.empty, [])
 

debian/changelog

@@ -27,6 +27,8 @@ git-annex (5.20131102) UNRELEASED; urgency=low
   * Work around Android linker problem that had prevented git-annex from
     running on Android 4.3 and 4.4.
   * repair: Handle case where index file is corrupt, but all objects are ok.
+  * assistant: Notice on startup when the index file is corrupt, and
+    auto-repair.
 
  -- Joey Hess <joeyh@debian.org>  Wed, 06 Nov 2013 16:14:14 -0400