Ceph-fuse fails to remount the fs

When mounting ceph via ceph-fuse, it fails to remount the FS to clear the dentries:

hut# ceph-fuse -f -s -n client.user -m 10.0.40.40 /ceph2
2023-08-29T11:24:27.614+0200 7feabf5873c0 -1 init, newargv = 0x56548f82ddc0 newargc=16
ceph-fuse[2155186]: starting ceph client
ceph-fuse[2155186]: starting fuse
mount: /ceph2: mount point not mounted or bad option.
       dmesg(1) may have more information after failed mount system call.
2023-08-29T11:24:27.636+0200 7feaa17fa6c0 -1 client.254236 failed to remount (to trim kernel dentries): return code = 32
mount: /ceph2: mount point not mounted or bad option.
       dmesg(1) may have more information after failed mount system call.
2023-08-29T11:24:28.647+0200 7feaa17fa6c0 -1 client.254236 failed to remount (to trim kernel dentries): return code = 32
mount: /ceph2: mount point not mounted or bad option.
       dmesg(1) may have more information after failed mount system call.
2023-08-29T11:24:29.657+0200 7feaa17fa6c0 -1 client.254236 failed to remount (to trim kernel dentries): return code = 32
mount: /ceph2: mount point not mounted or bad option.
       dmesg(1) may have more information after failed mount system call.
2023-08-29T11:24:30.667+0200 7feaa17fa6c0 -1 client.254236 failed to remount (to trim kernel dentries): return code = 32
mount: /ceph2: mount point not mounted or bad option.
       dmesg(1) may have more information after failed mount system call.
2023-08-29T11:24:31.678+0200 7feaa17fa6c0 -1 client.254236 failed to remount (to trim kernel dentries): return code = 32

However, setting client_die_on_failed_dentry_invalidate=false allows the mount command to succeed and the filesystem seems to work fine in some preliminary tests with fio and dd.