Fix laziness bug introduced in last release that breaks use of
--unlock-present and --hide-missing adjusted branches.
Since there is a writeFile of the same file immediately after readFile, it
may still have the file open for read (or may have happened to read it
already and closed it).
I was not able to reproduce the problem in brief testing, but this seems
obvious.
Sponsored-by: Luke Shumaker on Patreona
-- has that have not yet been propigated back to the
-- origbranch.
_ <- propigateAdjustedCommits' origbranch adj commitlck
-
- origheadfile <- inRepo $ readFile . Git.Ref.headFile
+
+ origheadfile <- inRepo $ readFileStrict . Git.Ref.headFile
-- Git normally won't do anything when asked to check
-- out the currently checked out branch, even when its
unless ok $ case newheadfile of
Nothing -> noop
Just v -> preventCommits $ \_commitlck -> inRepo $ \r -> do
- v' <- readFile (Git.Ref.headFile r)
+ v' <- readFileStrict (Git.Ref.headFile r)
when (v == v') $
writeFile (Git.Ref.headFile r) origheadfile
* git-annex.cabal: Prevent building with unix-compat 0.7 which
removed System.PosixCompat.User.
* Sped up sqlite inserts 2x when built with persistent 2.14.5.0
+ * Fix laziness bug introduced in last release that breaks use
+ of --unlock-present and --hide-missing adjusted branches.
-- Joey Hess <id@joeyh.name> Fri, 31 Mar 2023 12:48:54 -0400