feat(runner): add detailed TOML config validation with miette errors

- Use miette's NamedSource and SourceSpan for precise error locations
    - Implement field-level validation (mode, payload, iters, concurrency)
    - Show exact line and field causing validation errors in CLI output

runner/src/config.rs

@@ -1,7 +1,11 @@
-use crate::{args::Args, error};
+use crate::{
+    args::Args,
+    error::{self, ConfigError},
+};
 use common::{self, KeyExchangeMode};
+use miette::{NamedSource, SourceSpan};
 use serde::Deserialize;
-use std::{fs::read_to_string, net::SocketAddr, path::PathBuf};
+use std::{fs::read_to_string, net::SocketAddr, path::Path};
 
 #[derive(Debug, Clone, Deserialize)]
 pub struct BenchmarkConfig {
@@ -22,9 +26,28 @@ pub struct Config {
 ///
 /// # Errors
 /// Returns an error if the file cannot be read or parsed.
-pub fn load_from_file(path: &PathBuf) -> error::Result<Config> {
-    let content = read_to_string(path).map_err(common::Error::Io)?;
-    let config = toml::from_str::<Config>(&content).map_err(common::Error::Toml)?;
+pub fn load_from_file(path: &Path) -> error::Result<Config> {
+    let content = read_to_string(path).map_err(|source| ConfigError::ReadError {
+        source,
+        path: path.to_owned(),
+    })?;
+
+    let src = NamedSource::new(path.display().to_string(), content.clone());
+
+    let config = toml::from_str::<Config>(&content).map_err(|source| {
+        let span = source
+            .span()
+            .map(|s| SourceSpan::new(s.start.into(), s.end - s.start));
+
+        ConfigError::TomlParseError {
+            src: src.clone(),
+            span,
+            source,
+        }
+    })?;
+
+    validate_config(&config, &content, path)?;
+
     Ok(config)
 }
 
@@ -57,3 +80,163 @@ impl Config {
             .unwrap_or(KeyExchangeMode::X25519)
     }
 }
+
+/// Validate the configuration after parsing.
+fn validate_config(config: &Config, content: &str, path: &Path) -> error::Result<()> {
+    if config.benchmarks.is_empty() {
+        return Err(ConfigError::EmptyBenchmarks {
+            src: NamedSource::new(path.display().to_string(), content.to_string()),
+        }
+        .into());
+    }
+
+    for (idx, benchmark) in config.benchmarks.iter().enumerate() {
+        validate_benchmark(benchmark, idx, content, path)?;
+    }
+
+    Ok(())
+}
+
+/// Validate a single benchmark configuration.
+fn validate_benchmark(
+    benchmark: &BenchmarkConfig,
+    idx: usize,
+    content: &str,
+    path: &Path,
+) -> error::Result<()> {
+    let src = NamedSource::new(path.display().to_string(), content.to_string());
+
+    // Validate mode
+    if benchmark.mode.parse::<KeyExchangeMode>().is_err() {
+        return Err(ConfigError::ValidationError {
+            src,
+            span: find_field_span(content, idx, "mode"),
+            field: "mode".into(),
+            idx,
+            message: format!(
+                "Invalid key exchange mode '{}'. Valid values: 'x25519', 'x25519mlkem768'",
+                benchmark.mode
+            ),
+        }
+        .into());
+    }
+
+    if benchmark.payload == 0 {
+        return Err(ConfigError::ValidationError {
+            src,
+            span: find_field_span(content, idx, "payload"),
+            field: "payload".into(),
+            idx,
+            message: "Must be greater than 0".into(),
+        }
+        .into());
+    }
+
+    if benchmark.iters == 0 {
+        return Err(ConfigError::ValidationError {
+            src,
+            span: find_field_span(content, idx, "iters"),
+            field: "iters".into(),
+            idx,
+            message: "Must be greater than 0".into(),
+        }
+        .into());
+    }
+
+    if benchmark.concurrency == 0 {
+        return Err(ConfigError::ValidationError {
+            src,
+            span: find_field_span(content, idx, "concurrency"),
+            field: "concurrency".into(),
+            idx,
+            message: "Must be greater than 0".into(),
+        }
+        .into());
+    }
+
+    Ok(())
+}
+
+/// Find the span of a field in the TOML content for better error reporting.
+fn find_field_span(content: &str, benchmark_idx: usize, field_name: &str) -> Option<SourceSpan> {
+    let benchmark_section = find_benchmakr_section(content, benchmark_idx)?;
+    find_field_in_section(content, benchmark_section, field_name)
+}
+
+fn find_benchmakr_section(content: &str, target_idx: usize) -> Option<(usize, usize)> {
+    let mut current_idx = 0;
+    let mut section_start = None;
+    let mut byte_offset = 0;
+
+    for line in content.lines() {
+        let line_len = line.len() + 1; // +1 for newline
+
+        if line.trim_start().starts_with("[[benchmarks]]") {
+            if current_idx == target_idx {
+                section_start = Some(byte_offset);
+            } else if section_start.is_some() {
+                return Some((section_start?, byte_offset));
+            }
+            current_idx += 1;
+        }
+        byte_offset += line_len;
+    }
+    section_start.map(|start| (start, content.len()))
+}
+
+fn find_field_in_section(
+    content: &str,
+    (start, end): (usize, usize),
+    field_name: &str,
+) -> Option<SourceSpan> {
+    let section = &content[start..end];
+    let mut byte_offset = start;
+
+    for line in section.lines() {
+        let line_len = line.len() + 1; // +1 for newline
+        if line.trim_start().starts_with('[') && byte_offset != start {
+            break;
+        }
+
+        if let Some(span) = extract_field_value_span(line, field_name, byte_offset) {
+            return Some(span);
+        }
+
+        byte_offset += line_len;
+    }
+
+    None
+}
+
+fn extract_field_value_span(
+    line: &str,
+    field_name: &str,
+    line_offset: usize,
+) -> Option<SourceSpan> {
+    let eq_idx = line.find('=')?;
+    let field = line[..eq_idx].trim();
+
+    if field != field_name {
+        return None;
+    }
+
+    let after_eq = &line[eq_idx + 1..];
+    let value_start_offset = after_eq.len() - after_eq.trim_start().len();
+    let value_str = after_eq.trim_start();
+
+    let value_start = line_offset + eq_idx + 1 + value_start_offset;
+    let value_len = calculate_value_len(value_str);
+
+    Some(SourceSpan::new(value_start.into(), value_len))
+}
+
+fn calculate_value_len(value_str: &str) -> usize {
+    if let Some(stripped) = value_str.strip_prefix('"') {
+        return stripped.find('"').map_or(value_str.len(), |i| i + 2);
+    }
+
+    value_str.find('#').map_or_else(
+        || value_str.trim_end().len(),
+        |i| value_str[..i].trim_end().len(),
+    )
+}

runner/src/error.rs

@@ -1,6 +1,8 @@
+#![allow(unused_assignments)]
 //! Error types for the benchmark runner.
 
-use miette::Diagnostic;
+use miette::{Diagnostic, NamedSource, SourceSpan};
+use std::path::PathBuf;
 use thiserror::Error;
 
 /// Result type using the runner's custom error type.
@@ -13,6 +15,10 @@ pub enum Error {
     #[diagnostic(code(runner::common_error))]
     Common(#[from] common::Error),
 
+    #[error(transparent)]
+    #[diagnostic(transparent)]
+    Config(#[from] Box<ConfigError>),
+
     /// Network connection failure.
     #[error("Network error: {0}")]
     #[diagnostic(code(runner::network_error))]
@@ -26,3 +32,56 @@ impl Error {
         Self::Network(error.into())
     }
 }
+
+#[derive(Debug, Error, Diagnostic)]
+pub enum ConfigError {
+    #[error("Failed to read config file: {path}")]
+    #[diagnostic(
+        code(config::read_error),
+        help("Make sure the file exists and you have read permissions")
+    )]
+    ReadError {
+        #[source]
+        source: std::io::Error,
+        path: PathBuf,
+    },
+
+    #[error("Invalid TOML syntax in config file")]
+    #[diagnostic(code(config::toml_parse_error))]
+    TomlParseError {
+        #[source_code]
+        src: NamedSource<String>,
+        #[label("here")]
+        span: Option<SourceSpan>,
+        #[source]
+        source: toml::de::Error,
+    },
+
+    #[error("Invalid value for field '{field}' in benchmarks[{idx}]")]
+    #[diagnostic(code(config::validation_error))]
+    ValidationError {
+        #[source_code]
+        src: NamedSource<String>,
+        #[label("{message}")]
+        span: Option<SourceSpan>,
+        field: String,
+        idx: usize,
+        message: String,
+    },
+
+    #[error("Configuration must contain at least one [[benchmarks]] entry")]
+    #[diagnostic(
+        code(config::empty_benchmarks),
+        help("Add at least one [[benchmarks]] section to your config file")
+    )]
+    EmptyBenchmarks {
+        #[source_code]
+        src: NamedSource<String>,
+    },
+}
+
+impl From<ConfigError> for Error {
+    fn from(err: ConfigError) -> Self {
+        Self::Config(Box::new(err))
+    }
+}