Fix volume init to always reconcile the host directory's own attributes

resolve_volume_mount() was supposed to initialize a fresh host directory
from the image's own content at the mounted container path, but only ever
did anything when that image-side directory was non-empty. Two bugs
followed: an image declaring an *empty* directory with specific ownership/
permissions (e.g. a data directory owned by a non-root uid/gid) got a host
directory with plain create_directories() defaults instead, and even the
non-empty case never reconciled the directory's own attributes -- only
each copied entry's.

Fixed by splitting into two independent steps in the embedded cp script:
copy contents only when non-empty (as before), then always reconcile the
host directory's own mode/ownership/timestamps/xattrs via
`cp -a --attributes-only -T`. -T/--no-target-directory turned out to be
required -- caught by direct testing: without it, cp nests the image
directory *into* the already-existing host directory instead of
reconciling its attributes, which silently produced a spurious nested
copy and left the host directory's own attributes untouched.

Verified end-to-end against images/gitea.tar's real declared
/etc/gitea and /var/lib/gitea volumes under a real rootless mount: the
resulting host directories' mode/ownership now match the image's own
declared values, with no leftover mounts/layers afterward.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Gv3s5jckJKzh6JkMoi2Akz

CLAUDE.md

@@ -471,20 +471,38 @@ Source layout (all under `src/`):
   `resolve_volume_mount()`, called once per `-v` occurrence from `run_container()`
   when running with `-r`. A spec with no `/` is looked up in `config.volumes` by
   name (error if unknown); one with `/` is treated as a host directory path and
-  `create_directories()`'d if missing. If the resulting host directory is empty and
-  the image already has non-empty content at the given container path, that content
-  is copied in first. Both the existence check and the copy run as a single
-  `sh -c '[ -d ... ] && cp -a ...'` invocation wrapped through
-  `wrap_for_root_namespace()` (`bwrap.h`) — **not** a plain `std::filesystem` check
-  — because a rootless `containers-storage mount`'s content isn't visible to this
-  process at all without `nsenter`, the same constraint `run_bwrap()` itself works
-  around (see the root-vs-rootless paragraph below). `cp -a
-  --preserve=mode,ownership,timestamps,links[,xattr]` does the copy; whether
-  `,xattr` is included is decided by a direct `setxattr()`/`removexattr()` probe on
-  the host directory (no new library dependency — Linux POSIX ACLs are themselves
-  stored as xattrs, so this one probe stands in for both, logging a single
-  `spdlog::warn` if unsupported). A nonzero `cp` exit is only ever a warning, never
-  fatal — often just an ownership-preservation shortfall when not running as root.
+  `create_directories()`'d if missing. If the resulting host directory is empty,
+  `initialize_volume_directory()` reconciles it against the image's own directory
+  at the given container path: if that image directory is non-empty, its contents
+  are copied in first; then, **whether or not there was content to copy**, the
+  host directory's own mode/ownership/timestamps (and xattrs/ACLs where
+  supported) are always set to match the image directory's own — **real bug
+  fixed by the user, not assumed**: an earlier version only ever copied when the
+  image directory was non-empty, so an image declaring an *empty* directory with
+  specific ownership/permissions (e.g. a data directory owned by a non-root
+  uid/gid) got a host directory with default `create_directories()` permissions
+  instead, and even the non-empty case never reconciled the directory's *own*
+  attributes (only each copied entry's). The existence check, content copy, and
+  attribute reconciliation all run as a single `sh -c` invocation wrapped through
+  `wrap_for_root_namespace()` (`bwrap.h`) — **not** a plain `std::filesystem`
+  check — because a rootless `containers-storage mount`'s content isn't visible
+  to this process at all without `nsenter`, the same constraint `run_bwrap()`
+  itself works around (see the root-vs-rootless paragraph below). `cp -a
+  --preserve=mode,ownership,timestamps,links[,xattr] --attributes-only -T` does
+  the attribute-reconciliation step; whether `,xattr` is included is decided by
+  a direct `setxattr()`/`removexattr()` probe on the host directory (no new
+  library dependency — Linux POSIX ACLs are themselves stored as xattrs, so this
+  one probe stands in for both, logging a single `spdlog::warn` if unsupported).
+  `-T`/`--no-target-directory` is required on that second `cp` — confirmed by
+  direct testing: without it, since the host directory already exists, plain
+  `cp SRC DST` copies `SRC` *into* `DST` as a nested `DST/basename(SRC)`
+  subdirectory instead of reconciling `DST`'s own attributes, which is exactly
+  the bug this fix closes. A nonzero `cp` exit is only ever a warning, never
+  fatal — often just an ownership-preservation shortfall when not running as
+  root. Verified end-to-end against `images/gitea.tar`'s real declared
+  `/etc/gitea`/`/var/lib/gitea` volumes under a real rootless mount: the
+  resulting host directories' mode/ownership matched the image's own declared
+  values in both cases, and no mounts/layers were left behind afterward.
 
 Errors are logged via `spdlog::error`; every external command is also traced at debug
 level in `run_process()`/`run_process_foreground()` (`src/process.cpp`) — visible via