convert Key to ShortByteString

This adds the overhead of a copy when serializing and deserializing keys.
I have not benchmarked much, but runtimes seem barely changed at all by that.

When a lot of keys are in memory, it improves memory use.

And, it prevents keys sometimes getting PINNED in memory and failing to GC,
which is a problem ByteString has sometimes. In particular, git-annex sync
from a borg special remote had that problem and this improved its memory
use by a large amount.

Sponsored-by: Shae Erisson on Patreon
This commit is contained in:
Joey Hess 2021-10-05 20:20:08 -04:00
parent 012b71e471
commit 19e78816f0
No known key found for this signature in database
GPG key ID: DB12DB0FF05F8F38
15 changed files with 65 additions and 36 deletions

View file

@ -20,6 +20,7 @@ import Utility.Metered
import qualified Utility.SimpleProtocol as Proto
import qualified Data.ByteString as S
import qualified Data.ByteString.Short as S (toShort, fromShort)
import qualified Data.Map.Strict as M
import Data.Char
import Control.Concurrent
@ -285,7 +286,7 @@ toProtoKey k = ProtoKey $ alterKey k $ \d -> d
-- The extension can be easily removed, because the protocol
-- documentation does not allow '.' to be used in the keyName,
-- so the first one is the extension.
{ keyName = S.takeWhile (/= dot) (keyName d)
{ keyName = S.toShort (S.takeWhile (/= dot) (S.fromShort (keyName d)))
, keyVariety = setHasExt (HasExt False) (keyVariety d)
}
where

View file

@ -24,6 +24,7 @@ import Utility.Metered
import qualified Utility.RawFilePath as R
import qualified Data.ByteString as S
import qualified Data.ByteString.Short as S (toShort, fromShort)
import qualified Data.ByteString.Char8 as S8
import qualified Data.ByteString.Lazy as L
import Control.DeepSeq
@ -106,7 +107,7 @@ keyValue hash source meterupdate = do
filesize <- liftIO $ getFileSize file
s <- hashFile hash file meterupdate
return $ mkKey $ \k -> k
{ keyName = encodeBS s
{ keyName = S.toShort (encodeBS s)
, keyVariety = hashKeyVariety hash (HasExt False)
, keySize = Just filesize
}
@ -160,7 +161,7 @@ needsUpgrade :: Key -> Bool
needsUpgrade key = or
[ "\\" `S8.isPrefixOf` keyHash key
, S.any (not . validInExtension) (snd $ splitKeyNameExtension key)
, not (hasExt (fromKey keyVariety key)) && keyHash key /= fromKey keyName key
, not (hasExt (fromKey keyVariety key)) && keyHash key /= S.fromShort (fromKey keyName key)
]
trivialMigrate :: Key -> Backend -> AssociatedFile -> Annex (Maybe Key)
@ -171,14 +172,14 @@ trivialMigrate' :: Key -> Backend -> AssociatedFile -> Maybe Int -> Maybe Key
trivialMigrate' oldkey newbackend afile maxextlen
{- Fast migration from hashE to hash backend. -}
| migratable && hasExt oldvariety = Just $ alterKey oldkey $ \d -> d
{ keyName = keyHash oldkey
{ keyName = S.toShort (keyHash oldkey)
, keyVariety = newvariety
}
{- Fast migration from hash to hashE backend. -}
| migratable && hasExt newvariety = case afile of
AssociatedFile Nothing -> Nothing
AssociatedFile (Just file) -> Just $ alterKey oldkey $ \d -> d
{ keyName = keyHash oldkey
{ keyName = S.toShort $ keyHash oldkey
<> selectExtension maxextlen file
, keyVariety = newvariety
}
@ -186,9 +187,9 @@ trivialMigrate' oldkey newbackend afile maxextlen
- non-extension preserving key, with an extension
- in its keyName. -}
| newvariety == oldvariety && not (hasExt oldvariety) &&
keyHash oldkey /= fromKey keyName oldkey =
keyHash oldkey /= S.fromShort (fromKey keyName oldkey) =
Just $ alterKey oldkey $ \d -> d
{ keyName = keyHash oldkey
{ keyName = S.toShort (keyHash oldkey)
}
| otherwise = Nothing
where

View file

@ -16,6 +16,7 @@ import Types.Key
import Types.KeySource
import qualified Data.ByteString as S
import qualified Data.ByteString.Short as S (ShortByteString, toShort)
import qualified Data.ByteString.Lazy as L
import qualified System.FilePath.ByteString as P
import Data.Char
@ -25,13 +26,13 @@ import Data.Word
- If it's not too long, the full string is used as the keyName.
- Otherwise, it's truncated, and its md5 is prepended to ensure a unique
- key. -}
genKeyName :: String -> S.ByteString
genKeyName :: String -> S.ShortByteString
genKeyName s
-- Avoid making keys longer than the length of a SHA256 checksum.
| bytelen > sha256len = encodeBS $
| bytelen > sha256len = S.toShort $ encodeBS $
truncateFilePath (sha256len - md5len - 1) s' ++ "-" ++
show (md5 bl)
| otherwise = encodeBS s'
| otherwise = S.toShort $ encodeBS s'
where
s' = preSanitizeKeyName s
bl = encodeBL s
@ -47,7 +48,7 @@ addE source sethasext k = do
maxlen <- annexMaxExtensionLength <$> Annex.getGitConfig
let ext = selectExtension maxlen (keyFilename source)
return $ alterKey k $ \d -> d
{ keyName = keyName d <> ext
{ keyName = keyName d <> S.toShort ext
, keyVariety = sethasext (keyVariety d)
}

View file

@ -17,6 +17,7 @@ import Utility.Metered
import qualified Data.ByteString.Char8 as S8
import qualified Utility.RawFilePath as R
import qualified Data.ByteString.Short as S (toShort, fromShort)
backends :: [Backend]
backends = [backend]
@ -53,12 +54,13 @@ keyValue source _ = do
{- Old WORM keys could contain spaces and carriage returns,
- and can be upgraded to remove them. -}
needsUpgrade :: Key -> Bool
needsUpgrade key = any (`S8.elem` fromKey keyName key) [' ', '\r']
needsUpgrade key =
any (`S8.elem` S.fromShort (fromKey keyName key)) [' ', '\r']
removeProblemChars :: Key -> Backend -> AssociatedFile -> Annex (Maybe Key)
removeProblemChars oldkey newbackend _
| migratable = return $ Just $ alterKey oldkey $ \d -> d
{ keyName = encodeBS $ reSanitizeKeyName $ decodeBS $ keyName d }
{ keyName = S.toShort $ encodeBS $ reSanitizeKeyName $ decodeBS $ S.fromShort $ keyName d }
| otherwise = return Nothing
where
migratable = oldvariety == newvariety