fix: handle hard links in initramfs creation

- Track inodes to avoid duplicating data for hard-linked files
- Use fast compression level for faster initramfs creation

src/qemu_vm.rs

@@ -204,8 +204,11 @@ fn create_initramfs(rootfs: &PathBuf) -> Result<PathBuf> {
 fn create_cpio_archive(rootfs: &Path, pb: &ProgressBar) -> Result<Vec<u8>> {
     let mut archive = Vec::new();
     
+    // Track seen inodes to handle hard links properly
+    let mut seen_inodes = std::collections::HashMap::new();
+    
     // Collect all entries with their data
-    let entries = collect_entries(rootfs, rootfs, pb)?;
+    let entries = collect_entries(rootfs, rootfs, pb, &mut seen_inodes)?;
     
     // Collect entry names for checking existence later
     let entry_names: Vec<&str> = entries.iter().map(|(n, _, _, _, _)| n.as_str()).collect();
@@ -275,8 +278,15 @@ fn create_cpio_archive(rootfs: &Path, pb: &ProgressBar) -> Result<Vec<u8>> {
 }
 
 /// Collect all filesystem entries recursively
-fn collect_entries(base: &Path, current: &Path, pb: &ProgressBar) -> Result<Vec<(String, u32, u32, u32, Vec<u8>)>> {
+/// Uses a HashMap to track hard links by (device, inode) - only stores data for first occurrence
+fn collect_entries(
+    base: &Path,
+    current: &Path,
+    pb: &ProgressBar,
+    seen_inodes: &mut std::collections::HashMap<(u64, u64), String>,
+) -> Result<Vec<(String, u32, u32, u32, Vec<u8>)>> {
     let mut entries = Vec::new();
+    let mut total_data: u64 = 0;
     
     // Read directory entries
     let dir_entries: Vec<_> = match std::fs::read_dir(current) {
@@ -317,42 +327,69 @@ fn collect_entries(base: &Path, current: &Path, pb: &ProgressBar) -> Result<Vec<
             continue; // Skip other types (sockets, fifos, etc.)
         };
         
-        // Get file content or symlink target
+        // Build the entry name (relative path from base)
-        let data: Vec<u8> = if file_type.is_file() {
+        let relative = path.strip_prefix(base).unwrap();
-            match std::fs::read(&path) {
+        let entry_name = relative.to_string_lossy().into_owned();
+        
+        // Handle hard links: only store data for first occurrence
+        let (data, nlink) = if file_type.is_file() && metadata.nlink() > 1 {
+            // This file has multiple hard links - check if we've seen it before
+            let inode_key = (metadata.dev(), metadata.ino());
+            
+            if let Some(_first_path) = seen_inodes.get(&inode_key) {
+                // We've seen this inode before - create a hard link entry with no data
+                (Vec::new(), metadata.nlink() as u32)
+            } else {
+                // First occurrence - read the data and record this inode
+                seen_inodes.insert(inode_key, entry_name.clone());
+                let data = match std::fs::read(&path) {
                     Ok(data) => data,
                     Err(e) => {
                         veprintln!("Warning: cannot read file {}: {}", path.display(), e);
                         continue;
                     }
+                };
+                (data, metadata.nlink() as u32)
             }
+        } else if file_type.is_file() {
+            // Regular file with nlink=1
+            let data = match std::fs::read(&path) {
+                Ok(data) => data,
+                Err(e) => {
+                    veprintln!("Warning: cannot read file {}: {}", path.display(), e);
+                    continue;
+                }
+            };
+            (data, 1)
         } else if file_type.is_symlink() {
             match std::fs::read_link(&path) {
-                Ok(target) => target.to_string_lossy().into_owned().into_bytes(),
+                Ok(target) => (target.to_string_lossy().into_owned().into_bytes(), 1),
                 Err(e) => {
                     veprintln!("Warning: cannot read symlink {}: {}", path.display(), e);
                     continue;
                 }
             }
         } else {
-            Vec::new()
+            // Directory
+            (Vec::new(), 2)
         };
         
-        // Build the entry name (relative path from base)
+        total_data += data.len() as u64;
-        let relative = path.strip_prefix(base).unwrap();
-        let entry_name = relative.to_string_lossy().into_owned();
-        
-        // nlink: directories have 2 (. and ..), files/symlinks have 1
-        let nlink = if file_type.is_dir() { 2 } else { 1 };
-        
         entries.push((entry_name, mode, metadata.mtime() as u32, nlink, data));
         
         // Recurse into directories
         if file_type.is_dir() {
-            let mut sub_entries = collect_entries(base, &path, pb)?;
+            let mut sub_entries = collect_entries(base, &path, pb, seen_inodes)?;
+            for (_, _, _, _, d) in &sub_entries {
+                total_data += d.len() as u64;
+            }
             entries.append(&mut sub_entries);
         }
     }
     
+    // Only print summary for the root directory
+    if current == base {
+        veprintln!("Collected {} entries, {} bytes total data", entries.len(), total_data);
+    }
     Ok(entries)
 }