shm: handle separate PID namespaces case

[pandora-kernel.git] / ipc / shm.c

diff --git a/ipc/shm.c b/ipc/shm.c
index ab3385a..fdaf8be 100644 (file)
--- a/ipc/shm.c
+++ b/ipc/shm.c
@@ -74,6 +74,7 @@ void shm_init_ns(struct ipc_namespace *ns)
        ns->shm_ctlmax = SHMMAX;
        ns->shm_ctlall = SHMALL;
        ns->shm_ctlmni = SHMMNI;
+       ns->shm_rmid_forced = 0;
        ns->shm_tot = 0;
        ipc_init_ids(&shm_ids(ns));
 }
@@ -186,6 +187,23 @@ static void shm_destroy(struct ipc_namespace *ns, struct shmid_kernel *shp)
        ipc_rcu_putref(shp);
 }
 
+/*
+ * shm_may_destroy - identifies whether shm segment should be destroyed now
+ *
+ * Returns true if and only if there are no active users of the segment and
+ * one of the following is true:
+ *
+ * 1) shmctl(id, IPC_RMID, NULL) was called for this shp
+ *
+ * 2) sysctl kernel.shm_rmid_forced is set to 1.
+ */
+static bool shm_may_destroy(struct ipc_namespace *ns, struct shmid_kernel *shp)
+{
+       return (shp->shm_nattch == 0) &&
+              (ns->shm_rmid_forced ||
+               (shp->shm_perm.mode & SHM_DEST));
+}
+
 /*
  * remove the attach descriptor vma.
  * free memory for segment if it is marked destroyed.
@@ -206,11 +224,95 @@ static void shm_close(struct vm_area_struct *vma)
        shp->shm_lprid = task_tgid_vnr(current);
        shp->shm_dtim = get_seconds();
        shp->shm_nattch--;
-       if(shp->shm_nattch == 0 &&
-          shp->shm_perm.mode & SHM_DEST)
+       if (shm_may_destroy(ns, shp))
+               shm_destroy(ns, shp);
+       else
+               shm_unlock(shp);
+       up_write(&shm_ids(ns).rw_mutex);
+}
+
+static int shm_try_destroy_current(int id, void *p, void *data)
+{
+       struct ipc_namespace *ns = data;
+       struct shmid_kernel *shp = shm_lock(ns, id);
+
+       if (IS_ERR(shp))
+               return 0;
+
+       if (shp->shm_creator != current) {
+               shm_unlock(shp);
+               return 0;
+       }
+
+       /*
+        * Mark it as orphaned to destroy the segment when
+        * kernel.shm_rmid_forced is changed.
+        * It is noop if the following shm_may_destroy() returns true.
+        */
+       shp->shm_creator = NULL;
+
+       /*
+        * Don't even try to destroy it.  If shm_rmid_forced=0 and IPC_RMID
+        * is not set, it shouldn't be deleted here.
+        */
+       if (!ns->shm_rmid_forced) {
+               shm_unlock(shp);
+               return 0;
+       }
+
+       if (shm_may_destroy(ns, shp))
                shm_destroy(ns, shp);
        else
                shm_unlock(shp);
+       return 0;
+}
+
+static int shm_try_destroy_orphaned(int id, void *p, void *data)
+{
+       struct ipc_namespace *ns = data;
+       struct shmid_kernel *shp = shm_lock(ns, id);
+
+       if (IS_ERR(shp))
+               return 0;
+
+       /*
+        * We want to destroy segments without users and with already
+        * exit'ed originating process.
+        */
+       if (shp->shm_creator != NULL) {
+               shm_unlock(shp);
+               return 0;
+       }
+
+       if (shm_may_destroy(ns, shp))
+               shm_destroy(ns, shp);
+       else
+               shm_unlock(shp);
+       return 0;
+}
+
+void shm_destroy_orphaned(struct ipc_namespace *ns)
+{
+       down_write(&shm_ids(ns).rw_mutex);
+       idr_for_each(&shm_ids(ns).ipcs_idr, &shm_try_destroy_orphaned, ns);
+       up_write(&shm_ids(ns).rw_mutex);
+}
+
+
+void exit_shm(struct task_struct *task)
+{
+       struct nsproxy *nsp = task->nsproxy;
+       struct ipc_namespace *ns;
+
+       if (!nsp)
+               return;
+       ns = nsp->ipc_ns;
+       if (!ns)
+               return;
+
+       /* Destroy all already created segments, but not mapped yet */
+       down_write(&shm_ids(ns).rw_mutex);
+       idr_for_each(&shm_ids(ns).ipcs_idr, &shm_try_destroy_current, ns);
        up_write(&shm_ids(ns).rw_mutex);
 }
 
@@ -277,13 +379,13 @@ static int shm_release(struct inode *ino, struct file *file)
        return 0;
 }
 
-static int shm_fsync(struct file *file, int datasync)
+static int shm_fsync(struct file *file, loff_t start, loff_t end, int datasync)
 {
        struct shm_file_data *sfd = shm_file_data(file);
 
        if (!sfd->file->f_op->fsync)
                return -EINVAL;
-       return sfd->file->f_op->fsync(sfd->file, datasync);
+       return sfd->file->f_op->fsync(sfd->file, start, end, datasync);
 }
 
 static unsigned long shm_get_unmapped_area(struct file *file,
@@ -404,6 +506,7 @@ static int newseg(struct ipc_namespace *ns, struct ipc_params *params)
        shp->shm_segsz = size;
        shp->shm_nattch = 0;
        shp->shm_file = file;
+       shp->shm_creator = current;
        /*
         * shmid gets reported as "inode#" in /proc/pid/maps.
         * proc-ps tools use this. Changing this will break them.
@@ -950,8 +1053,7 @@ out_nattch:
        shp = shm_lock(ns, shmid);
        BUG_ON(IS_ERR(shp));
        shp->shm_nattch--;
-       if(shp->shm_nattch == 0 &&
-          shp->shm_perm.mode & SHM_DEST)
+       if (shm_may_destroy(ns, shp))
                shm_destroy(ns, shp);
        else
                shm_unlock(shp);