Recreate SHM locks when renumbering on count mismatch

When we're renumbering locks, we're destroying all existing
allocations anyways, so destroying the old lock struct is not a
particularly big deal. Existing long-lived libpod instances will
continue to use the old locks, but that will be solved in a
followon.

Also, solve an issue with returning error values in the C code.
There were a few places where we return ERRNO where it was not
set, so make them return actual error codes).

Signed-off-by: Matthew Heon <matthew.heon@pm.me>

2 files changed, 24 insertions, 8 deletions

diff --git a/libpod/lock/shm/shm_lock.c b/libpod/lock/shm/shm_lock.c
index 367055823..d11fce71a 100644
--- a/libpod/lock/shm/shm_lock.c
+++ b/libpod/lock/shm/shm_lock.c
@@ -203,6 +203,8 @@ shm_struct_t *setup_lock_shm(char *path, uint32_t num_locks, int *error_code) {
 // terminating NULL byte.
 // Returns a valid pointer on success or NULL on error.
 // If an error occurs, negative ERRNO values will be written to error_code.
+// ERANGE is returned for a mismatch between num_locks and the number of locks
+// available in the the SHM lock struct.
 shm_struct_t *open_lock_shm(char *path, uint32_t num_locks, int *error_code) {
   int shm_fd;
   shm_struct_t *shm;
@@ -255,11 +257,11 @@ shm_struct_t *open_lock_shm(char *path, uint32_t num_locks, int *error_code) {
 
   // Need to check the SHM to see if it's actually our locks
   if (shm->magic != MAGIC) {
-    *error_code = -1 * errno;
+    *error_code = -1 * EBADF;
     goto CLEANUP;
   }
   if (shm->num_locks != (num_bitmaps * BITMAP_SIZE)) {
-    *error_code = -1 * errno;
+    *error_code = -1 * ERANGE;
     goto CLEANUP;
   }
 
diff --git a/libpod/runtime.go b/libpod/runtime.go
index bc7c061c4..850df4fc9 100644
--- a/libpod/runtime.go
+++ b/libpod/runtime.go
@@ -7,6 +7,7 @@ import (
 	"os/exec"
 	"path/filepath"
 	"sync"
+	"syscall"
 
 	"github.com/BurntSushi/toml"
 	is "github.com/containers/image/storage"
@@ -757,12 +758,6 @@ func makeRuntime(runtime *Runtime) (err error) {
 			aliveLock.Unlock()
 		}
 	}()
-	// If we're renumbering locks, do it now.
-	// It breaks out of normal runtime init, and will not return a valid
-	// runtime.
-	if runtime.doRenumber {
-		return runtime.renumberLocks()
-	}
 
 	_, err = os.Stat(runtimeAliveFile)
 	if err != nil {
@@ -789,12 +784,31 @@ func makeRuntime(runtime *Runtime) (err error) {
 			if err != nil {
 				return err
 			}
+		} else if err == syscall.ERANGE && runtime.doRenumber {
+			// ERANGE indicates a lock numbering mismatch.
+			// Since we're renumbering, this is not fatal.
+			// Remove the earlier set of locks and recreate.
+			if err := os.Remove(filepath.Join("/dev/shm", lockPath)); err != nil {
+				return errors.Wrapf(err, "error removing libpod locks file %s", lockPath)
+			}
+
+			manager, err = lock.NewSHMLockManager(lockPath, runtime.config.NumLocks)
+			if err != nil {
+				return err
+			}
 		} else {
 			return err
 		}
 	}
 	runtime.lockManager = manager
 
+	// If we're renumbering locks, do it now.
+	// It breaks out of normal runtime init, and will not return a valid
+	// runtime.
+	if runtime.doRenumber {
+		return runtime.renumberLocks()
+	}
+
 	// If we need to refresh the state, do it now - things are guaranteed to
 	// be set up by now.
 	if doRefresh {