Commits

Author Commit Message Labels Comments Date
seun...@gmail.com
implemented a persistent string map. doesn't integrate with the other STMs yet.
seun...@gmail.com
a simple persistent map
seun...@gmail.com
private -> protected for some fields
seun...@gmail.com
implemented correctly using Condition objects
seun...@gmail.com
first attempt at non-lazy concurrent transactional queue
seun...@gmail.com
comment
seun...@gmail.com
renamed some functions, etc
seun...@gmail.com
NA
seun...@gmail.com
exponential backoff for 2pc;
seun...@gmail.com
[no commit message]
seun...@gmail.com
move to single-folder organization. again.
seun...@gmail.com
move to single-folder organization.
seun...@gmail.com
don't allow write lock acquisition in a read-only transaction.
seun...@gmail.com
better deadlock detection
seun...@gmail.com
little bugfix.
seun...@gmail.com
deadlock detection via the original method
seun...@gmail.com
TVar, etc -> Ref
seun...@gmail.com
added map support
seun...@gmail.com
s2 cleanup
seun...@gmail.com
transactional buffer
seun...@gmail.com
s2pl -> s2 and implementation of transactional set
seun...@gmail.com
s2pl compiles. not tested yet tho.
seun...@gmail.com
s2pl (doesn't yet compile though)
seun...@gmail.com
push changes.
seun...@gmail.com
simpler alternative to threadlocals, as suggested by Peter
seun...@gmail.com
TransactionExecutors should be passed to transactionmanagers in a constructor since they are never modified.
seun...@gmail.com
failed to eliminate threadlocals; reverted.
seun...@gmail.com
attempt to eliminate threadlocals
seun...@gmail.com
before the attempt to eliminate threadlocals
seun...@gmail.com
Central support for read-only transactions: (it's not safe to combine optimized read-only transactions with transactions involving writes. The resulting schedule won't be serializable.)
  1. Prev
  2. Next