Merge 'Add Miri support for turso_stress, with bash scripts to run' from Bob Peterson

It was mentioned in https://github.com/tursodatabase/turso/pull/3720
that adding Miri support for `turso_stress` would be useful. And, that a
bash script to start Miri with the right config would be a big help.
Notable changes:
- `antithesis_sdk`'s default features are disabled at the workspace
level, and only enabled as needed with the `antithesis` feature flag in
the various turso crates. Miri needs the noop version of
`antithesis_sdk` to run `turso_stress`, and feature unification
previously prevented this. I'm not able to ensure locally that all the
Antithesis stuff is still happy with these changes.
- Bash script to run `turso_stress` - this is barebones for now, see
below
- Bash script to run `simulator` - this passes any args to the `cargo
run` invocation inside, intercepting `--seed` if it's present, and
generating one from `/dev/random` if it's not. The seed is passed to
both Miri and the simulator to keep the overall execution reproducible.
(I checked this with a simple case)
- A `const fn`, `normal_or_miri` to supply different defaults in things
like CLI args for normal operation and Miri, since it's so slow. (An
idea I stole from tokio.) Right now the relevant values are 100x smaller
for Miri, although Miri is probably 1000 to 10,000x slower overall from
a rough estimation.
Caught UB from running `turso_stress` with Miri:
- An unsafe cast of a `*u8` to `*u32` inside the BTree implementation
resulted in the `*u32` making an unaligned read: `read()` ->
`read_unaligned()` fixes this
Future work - Making `turso_stress` reproducible under Miri:
- Right now `turso_stress` is plugged in to Antithesis, which is great!
But, `antithesis_sdk`'s noop mode (`default-features = false`) turns
`antithesis_sdk::random::get_random()` into `rand::random<u64>()`, which
isn't seedable/reproducible. It's more work than I wanted to take on in
this PR, but I'd like to instead conditionally replace `get_random` with
a seedable `ChaCha8Rng` like in the simulator, if Miri is being used.
Comment:
- On a machine without all necessary dependencies, running the bash
scripts fails in a way that cargo prompts you through installing the
nightly toolchain, Miri, etc. until it works
- Below is a snippet of the output from Miri on the Btree alignment
issue. Because turso_stress isn't yet deterministic/reproducible under
Miri, I can't always reproduce it. (It doesn't always happen like the
ones in my last MR)
```
error: Undefined Behavior: accessing memory based on pointer with alignment 1, but alignment 4 is required
    --> /home/rwp/git/turso/core/storage/btree.rs:2860:50
     |
2860 |                     let mut pgno: u32 = unsafe { right_pointer.cast::<u32>().read().swap_bytes() };
     |                                                  ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ Undefined Behavior occurred here
     |
     = help: this indicates a bug in the program: it performed an invalid operation, and caused Undefined Behavior
     = help: see https://doc.rust-lang.org/nightly/reference/behavior-considered-undefined.html for further information

```

Closes #3790

Cargo.toml

@@ -90,7 +90,7 @@ fallible-iterator = "0.3.0"
 criterion = "0.5"
 chrono = { version = "0.4.42", default-features = false }
 hex = "0.4"
-antithesis_sdk = "0.2"
+antithesis_sdk = { version = "0.2", default-features = false }
 cfg-if = "1.0.0"
 tracing-appender = "0.2.3"
 env_logger = { version = "0.11.6", default-features = false }

bindings/rust/src/lib.rs

@@ -153,14 +153,14 @@ impl Builder {
         match vfs_choice {
             "memory" => Ok(Arc::new(turso_core::MemoryIO::new())),
             "syscall" => {
-                #[cfg(target_family = "unix")]
+                #[cfg(all(target_family = "unix", not(miri)))]
                 {
                     Ok(Arc::new(
                         turso_core::UnixIO::new()
                             .map_err(|e| Error::SqlExecutionFailure(e.to_string()))?,
                     ))
                 }
-                #[cfg(not(target_family = "unix"))]
+                #[cfg(any(not(target_family = "unix"), miri))]
                 {
                     Ok(Arc::new(
                         turso_core::PlatformIO::new()
@@ -168,12 +168,12 @@ impl Builder {
                     ))
                 }
             }
-            #[cfg(target_os = "linux")]
+            #[cfg(all(target_os = "linux", not(miri)))]
             "io_uring" => Ok(Arc::new(
                 turso_core::UringIO::new()
                     .map_err(|e| Error::SqlExecutionFailure(e.to_string()))?,
             )),
-            #[cfg(not(target_os = "linux"))]
+            #[cfg(any(not(target_os = "linux"), miri))]
             "io_uring" => Err(Error::SqlExecutionFailure(
                 "io_uring is only available on Linux targets".to_string(),
             )),

core/Cargo.toml

@@ -15,7 +15,7 @@ path = "lib.rs"
 
 [features]
 default = ["fs", "uuid", "time", "json", "series", "encryption"]
-antithesis = ["dep:antithesis_sdk"]
+antithesis = ["dep:antithesis_sdk", "antithesis_sdk?/full"]
 tracing_release = ["tracing/release_max_level_info"]
 conn_raw_api = []
 fs = ["turso_ext/vfs"]

core/storage/btree.rs

@@ -2857,7 +2857,8 @@ impl BTreeCursor {
 
                     // load sibling pages
                     // start loading right page first
-                    let mut pgno: u32 = unsafe { right_pointer.cast::<u32>().read().swap_bytes() };
+                    let mut pgno: u32 =
+                        unsafe { right_pointer.cast::<u32>().read_unaligned().swap_bytes() };
                     let current_sibling = sibling_pointer;
                     let mut group = CompletionGroup::new(|_| {});
                     for i in (0..=current_sibling).rev() {

simulator/run-miri.sh

@@ -0,0 +1,31 @@
+#!/bin/bash
+
+ARGS=("$@")
+
+# Intercept the seed if it's passed
+while [[ $# -gt 0 ]]; do
+  case $1 in
+    -s=*|--seed=*)
+      seed="${1#*=}"
+      shift
+      ;;
+    -s|--seed)
+      seed="$2"
+      shift 2
+      ;;
+    *)
+      shift
+      ;;
+  esac
+done
+# Otherwise make one up
+if [ -z "$seed" ]; then
+  # Dump 8 bytes of /dev/random as decimal u64
+  seed=$(od -An -N8 -tu8 /dev/random | tr -d ' ')
+  ARGS+=("--seed" "${seed}")
+  echo "Generated seed for Miri and simulator: ${seed}"
+else
+  echo "Intercepted simulator seed to pass to Miri: ${seed}"
+fi
+
+MIRIFLAGS="-Zmiri-disable-isolation -Zmiri-disable-stacked-borrows -Zmiri-seed=${seed}" cargo +nightly miri run --bin limbo_sim -- "${ARGS[@]}"

simulator/runner/cli.rs

@@ -30,7 +30,7 @@ pub struct SimulatorCLI {
         short = 'n',
         long,
         help = "change the maximum size of the randomly generated sequence of interactions",
-        default_value_t = 5000,
+        default_value_t = normal_or_miri(5000, 50),
         value_parser = clap::value_parser!(u32).range(1..)
     )]
     pub maximum_tests: u32,
@@ -38,7 +38,7 @@ pub struct SimulatorCLI {
         short = 'k',
         long,
         help = "change the minimum size of the randomly generated sequence of interactions",
-        default_value_t = 1000,
+        default_value_t = normal_or_miri(1000, 10),
         value_parser = clap::value_parser!(u32).range(1..)
     )]
     pub minimum_tests: u32,
@@ -149,7 +149,8 @@ pub struct SimulatorCLI {
     pub keep_files: bool,
     #[clap(
         long,
-        help = "Disable the SQLite integrity check at the end of a simulation"
+        help = "Disable the SQLite integrity check at the end of a simulation",
+        default_value_t = normal_or_miri(false, true)
     )]
     pub disable_integrity_check: bool,
     #[clap(
@@ -279,3 +280,7 @@ impl ValueParserFactory for ProfileType {
         ProfileTypeParser
     }
 }
+
+const fn normal_or_miri<T: Copy>(normal_val: T, miri_val: T) -> T {
+    if cfg!(miri) { miri_val } else { normal_val }
+}

stress/Cargo.toml

@@ -16,7 +16,7 @@ path = "main.rs"
 
 [features]
 default = ["experimental_indexes"]
-antithesis = ["turso/antithesis"]
+antithesis = ["turso/antithesis", "antithesis_sdk/full"]
 experimental_indexes = []
 
 [dependencies]

stress/main.rs

@@ -642,9 +642,11 @@ async fn main() -> Result<(), Box<dyn std::error::Error + Send + Sync>> {
     println!("Done. SQL statements written to {}", opts.log_file);
     println!("Database file: {db_file}");
 
-    println!("Running SQLite Integrity check");
-
-    integrity_check(std::path::Path::new(&db_file))?;
+    #[cfg(not(miri))]
+    {
+        println!("Running SQLite Integrity check");
+        integrity_check(std::path::Path::new(&db_file))?;
+    }
 
     Ok(())
 }

stress/opts.rs

@@ -21,7 +21,7 @@ pub struct Opts {
         short = 'i',
         long,
         help = "the number of iterations",
-        default_value_t = 100000
+        default_value_t = normal_or_miri(100_000, 1000)
     )]
     pub nr_iterations: usize,
 
@@ -75,3 +75,11 @@ pub struct Opts {
     )]
     pub busy_timeout: u64,
 }
+
+const fn normal_or_miri<T: Copy>(normal_val: T, miri_val: T) -> T {
+    if cfg!(miri) {
+        miri_val
+    } else {
+        normal_val
+    }
+}

stress/run-miri.sh

@@ -0,0 +1,4 @@
+#!/bin/bash
+
+
+MIRIFLAGS="-Zmiri-disable-isolation -Zmiri-disable-stacked-borrows" cargo +nightly miri run -p turso_stress -- "$@"