Merge branch 'akpm' (second patchbomb from Andrew Morton)

Merge more incoming from Andrew Morton:
 "Two new syscalls:

     memfd_create in "shm: add memfd_create() syscall"
     kexec_file_load in "kexec: implementation of new syscall kexec_file_load"

  And:

   - Most (all?) of the rest of MM

   - Lots of the usual misc bits

   - fs/autofs4

   - drivers/rtc

   - fs/nilfs

   - procfs

   - fork.c, exec.c

   - more in lib/

   - rapidio

   - Janitorial work in filesystems: fs/ufs, fs/reiserfs, fs/adfs,
     fs/cramfs, fs/romfs, fs/qnx6.

   - initrd/initramfs work

   - "file sealing" and the memfd_create() syscall, in tmpfs

   - add pci_zalloc_consistent, use it in lots of places

   - MAINTAINERS maintenance

   - kexec feature work"

* emailed patches from Andrew Morton <akpm@linux-foundation.org: (193 commits)
  MAINTAINERS: update nomadik patterns
  MAINTAINERS: update usb/gadget patterns
  MAINTAINERS: update DMA BUFFER SHARING patterns
  kexec: verify the signature of signed PE bzImage
  kexec: support kexec/kdump on EFI systems
  kexec: support for kexec on panic using new system call
  kexec-bzImage64: support for loading bzImage using 64bit entry
  kexec: load and relocate purgatory at kernel load time
  purgatory: core purgatory functionality
  purgatory/sha256: provide implementation of sha256 in purgaotory context
  kexec: implementation of new syscall kexec_file_load
  kexec: new syscall kexec_file_load() declaration
  kexec: make kexec_segment user buffer pointer a union
  resource: provide new functions to walk through resources
  kexec: use common function for kimage_normal_alloc() and kimage_crash_alloc()
  kexec: move segment verification code in a separate function
  kexec: rename unusebale_pages to unusable_pages
  kernel: build bin2c based on config option CONFIG_BUILD_BIN2C
  bin2c: move bin2c in scripts/basic
  shm: wait for pins to be released when sealing
  ...

1 files changed, 39 insertions, 36 deletions

diff --git a/ipc/shm.c b/ipc/shm.c
index 89fc354156cb..7fc9f9f3a26b 100644
--- a/ipc/shm.c
+++ b/ipc/shm.c
@@ -178,6 +178,7 @@ static void shm_rcu_free(struct rcu_head *head)
 
 static inline void shm_rmid(struct ipc_namespace *ns, struct shmid_kernel *s)
 {
+	list_del(&s->shm_clist);
 	ipc_rmid(&shm_ids(ns), &s->shm_perm);
 }
 
@@ -268,37 +269,6 @@ static void shm_close(struct vm_area_struct *vma)
 }
 
 /* Called with ns->shm_ids(ns).rwsem locked */
-static int shm_try_destroy_current(int id, void *p, void *data)
-{
-	struct ipc_namespace *ns = data;
-	struct kern_ipc_perm *ipcp = p;
-	struct shmid_kernel *shp = container_of(ipcp, struct shmid_kernel, shm_perm);
-
-	if (shp->shm_creator != current)
-		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)
-		return 0;
-
-	if (shm_may_destroy(ns, shp)) {
-		shm_lock_by_ptr(shp);
-		shm_destroy(ns, shp);
-	}
-	return 0;
-}
-
-/* Called with ns->shm_ids(ns).rwsem locked */
 static int shm_try_destroy_orphaned(int id, void *p, void *data)
 {
 	struct ipc_namespace *ns = data;
@@ -329,18 +299,50 @@ void shm_destroy_orphaned(struct ipc_namespace *ns)
 	up_write(&shm_ids(ns).rwsem);
 }
 
-
+/* Locking assumes this will only be called with task == current */
 void exit_shm(struct task_struct *task)
 {
 	struct ipc_namespace *ns = task->nsproxy->ipc_ns;
+	struct shmid_kernel *shp, *n;
 
-	if (shm_ids(ns).in_use == 0)
+	if (list_empty(&task->sysvshm.shm_clist))
 		return;
 
-	/* Destroy all already created segments, but not mapped yet */
+	/*
+	 * If kernel.shm_rmid_forced is not set then only keep track of
+	 * which shmids are orphaned, so that a later set of the sysctl
+	 * can clean them up.
+	 */
+	if (!ns->shm_rmid_forced) {
+		down_read(&shm_ids(ns).rwsem);
+		list_for_each_entry(shp, &task->sysvshm.shm_clist, shm_clist)
+			shp->shm_creator = NULL;
+		/*
+		 * Only under read lock but we are only called on current
+		 * so no entry on the list will be shared.
+		 */
+		list_del(&task->sysvshm.shm_clist);
+		up_read(&shm_ids(ns).rwsem);
+		return;
+	}
+
+	/*
+	 * Destroy all already created segments, that were not yet mapped,
+	 * and mark any mapped as orphan to cover the sysctl toggling.
+	 * Destroy is skipped if shm_may_destroy() returns false.
+	 */
 	down_write(&shm_ids(ns).rwsem);
-	if (shm_ids(ns).in_use)
-		idr_for_each(&shm_ids(ns).ipcs_idr, &shm_try_destroy_current, ns);
+	list_for_each_entry_safe(shp, n, &task->sysvshm.shm_clist, shm_clist) {
+		shp->shm_creator = NULL;
+
+		if (shm_may_destroy(ns, shp)) {
+			shm_lock_by_ptr(shp);
+			shm_destroy(ns, shp);
+		}
+	}
+
+	/* Remove the list head from any segments still attached. */
+	list_del(&task->sysvshm.shm_clist);
 	up_write(&shm_ids(ns).rwsem);
 }
 
@@ -561,6 +563,7 @@ static int newseg(struct ipc_namespace *ns, struct ipc_params *params)
 	shp->shm_nattch = 0;
 	shp->shm_file = file;
 	shp->shm_creator = current;
+	list_add(&shp->shm_clist, &current->sysvshm.shm_clist);
 
 	/*
 	 * shmid gets reported as "inode#" in /proc/pid/maps.