Because `io_uring` may have many other I/O submission events queued
(that are relevant to the operation) when we experience an error,
marking our `Completion` objects as aborted is not sufficient, the
kernel will still execute queued I/O, which can mutate WAL or DB state
after we’ve declared failure and keep references (iovec arrays, buffers)
alive and stall reuse. We need to stop those in-flight SQEs at the
kernel and then drain the ring to a known-empty state before reusing any
resources.
The following methods were added to the `IO` trait:
`cancel`: which takes a slice of `Completion` objects and has a default
implementation that simply marks them as `aborted`.
`drain`: which has a default noop implementation, but the `io_uring`
backend implements this method to drain the ring.
CC @sivukhin
Reviewed-by: Pere Diaz Bou <pere-altea@homail.com>
Closes#2787
If upgrade from read to write transaction fails, don't roll back the
transaction. Instead restore the transaction into its original state,
which allows deferred transactions that have not read anything to
restart automatically.
Fixes#2984
This PR restructure JS packages and also adds support for OPFS for
tursodatabase in browser.
The new structure looks like this:
1. `@tursodatabase/database-common` - contains abstract JS code for
bindings which depends only on `NativeDB` interface and not on the
explicit native bindings
2. `@tursodatabase/database` - contains native bindings for the database
and re-use `core` package
3. `@tursodatabase/database-browser` - contains bindings for browser
(WASM + OPFS)
As OPFS sync API (which is the most performant one in the web) works
only in the web worker - this PR also make few operations async in order
to run them as `napi-rs` AsyncTask. The following operations became
async in `promise.ts` for node and browser: `pragma` / `exec` / `close`.
Also, as few code pathes during initialization are non-async - they
complicates integration of sync constructor in the browser with OPFS.
So, right now - turso support only `connect` method for browser in non-
memory mode.
Closes#2927
The MakeRecord instruction now accepts an optional affinity_str
parameter that applies column-specific type conversions before creating
records. When provided, the affinity string is applied
character-by-character to each register using the existing
apply_affinity_char() function, matching SQLite's behavior.
Fixes#2040Fixes#2041
On the syscall IO backend, on TPC-H query 12, the _dominating_ part
of the stack trace is trying to construct affinities from a character,
failing, allocating an error&string, and then immediately falling back to
Blob affinity and dropping the error&string.
Since I'm on vacation I won't spend cycles on figuring out why we are passing
an incorrect affinity in `flags.get_affinity()` and instead make this lazy PR
just to improve performance and stop doing silly things :]
This fairly long commit implements persistence for materialized view.
It is hard to split because of all the interdependencies between components,
so it is a one big thing. This commit message will at least try to go into
details about the basic architecture.
Materialized Views as tables
============================
Materialized views are now a normal table - whereas before they were a virtual
table. By making a materialized view a table, we can reuse all the
infrastructure for dealing with tables (cursors, etc).
One of the advantages of doing this is that we can create indexes on view
columns. Later, we should also be able to write those views to separate files
with ATTACH write.
Materialized Views as Zsets
===========================
The contents of the table are a ZSet: rowid, values, weight. Readers will
notice that because of this, the usage of the ZSet data structure dwindles
throughout the codebase. The main difference between our materialized ZSet and
the standard DBSP ZSet, is that obviously ours is backed by a BTree, not a Hash
(since SQLite tables are BTrees)
Aggregator State
================
In DBSP, the aggregator nodes also have state. To store that state, there is a
second table. The table holds all aggregators in the view, and there is one
table per view. That is __turso_internal_dbsp_state_{view_name}. The format of
that table is similar to a ZSet: rowid, serialized_values, weight. We serialize
the values because there will be many aggregators in the table. We can't rely
on a particular format for the values.
The Materialized View Cursor
============================
Reading from a Materialized View essentially means reading from the persisted
ZSet, and enhancing that with data that exists within the transaction.
Transaction data is ephemeral, so we do not materialize this anywhere: we have
a carefully crafted implementation of seek that takes care of merging weights
and stitching the two sets together.
`commit_txn` in MVCC was hacking its way through I/O until now. After
adding this and the test for concurrent writers we now see `busy` errors
returning as expected because there is no `commit` queueing happening
yet until next PR I open.
Closes#2895
@penberg this PR try to clean up `turso_parser`'s`fmt` code.
- `get_table_name` and `get_column_name` should return None when
table/column does not exist.
```rust
/// Context to be used in ToSqlString
pub trait ToSqlContext {
/// Given an id, get the table name
/// First Option indicates whether the table exists
///
/// Currently not considering aliases
fn get_table_name(&self, _id: TableInternalId) -> Option<&str> {
None
}
/// Given a table id and a column index, get the column name
/// First Option indicates whether the column exists
/// Second Option indicates whether the column has a name
fn get_column_name(&self, _table_id: TableInternalId, _col_idx: usize) -> Option<Option<&str>> {
None
}
// help function to handle missing table/column names
fn get_table_and_column_names(
&self,
table_id: TableInternalId,
col_idx: usize,
) -> (String, String) {
let table_name = self
.get_table_name(table_id)
.map(|s| s.to_owned())
.unwrap_or_else(|| format!("t{}", table_id.0));
let column_name = self
.get_column_name(table_id, col_idx)
.map(|opt| {
opt.map(|s| s.to_owned())
.unwrap_or_else(|| format!("c{col_idx}"))
})
.unwrap_or_else(|| format!("c{col_idx}"));
(table_name, column_name)
}
}
```
- remove `FmtTokenStream` because it is same as `WriteTokenStream `
- remove useless functions and simplify `ToTokens`
```rust
/// Generate token(s) from AST node
/// Also implements Display to make sure devs won't forget Display
pub trait ToTokens: Display {
/// Send token(s) to the specified stream with context
fn to_tokens<S: TokenStream + ?Sized, C: ToSqlContext>(
&self,
s: &mut S,
context: &C,
) -> Result<(), S::Error>;
// Return displayer representation with context
fn displayer<'a, 'b, C: ToSqlContext>(&'b self, ctx: &'a C) -> SqlDisplayer<'a, 'b, C, Self>
where
Self: Sized,
{
SqlDisplayer::new(ctx, self)
}
}
```
Closes#2748
Because we can abort a read_page completion, this means a page can be in
the cache but be unloaded and unlocked. However, if we do not evict that
page from the page cache, we will return an unloaded page later which
will trigger assertions later on. This is worsened by the fact that page
cache is not per `Statement`, so you can abort a completion in one
Statement, and trigger some error in the next one if we don't evict the
page in these circumstances.
Also, to propagate IO errors we need to return the Error from
IOCompletions on step.
Closes#2785
Using `usize` to compute file offsets caps us at ~16GB on 32-bit
systems. For example, with 4 KiB pages we can only address up to 1048576
pages; attempting the next page overflows a 32-bit usize and can wrap
the write offset, corrupting data. Switching our I/O APIs and offset
math to u64 avoids this overflow on 32-bit targets
Closes#2791
- check free list trunk and pages
- use shared hash map to check for duplicate references for pages
- properly check overflow pages
Reviewed-by: Pere Diaz Bou <pere-altea@homail.com>
Closes#2816
