Merge android-4.4.194 (a749771) into msm-4.4

* refs/heads/tmp-a749771
  Linux 4.4.194
  net_sched: let qdisc_put() accept NULL pointer
  ARC: export "abort" for modules
  media: technisat-usb2: break out of loop at end of buffer
  floppy: fix usercopy direction
  keys: Fix missing null pointer check in request_key_auth_describe()
  dmaengine: ti: omap-dma: Add cleanup in omap_dma_probe()
  net: seeq: Fix the function used to release some memory in an error handling path
  tools/power turbostat: fix buffer overrun
  sky2: Disable MSI on yet another ASUS boards (P6Xxxx)
  cifs: Use kzfree() to zero out the password
  cifs: set domainName when a domain-key is used in multiuser
  NFSv2: Fix write regression
  NFSv2: Fix eof handling
  netfilter: nf_conntrack_ftp: Fix debug output
  x86/apic: Fix arch_dynirq_lower_bound() bug for DT enabled machines
  r8152: Set memory to all 0xFFs on failed reg reads
  ARM: 8874/1: mm: only adjust sections of valid mm structures
  Kconfig: Fix the reference to the IDT77105 Phy driver in the description of ATM_NICSTAR_USE_IDT77105
  NFS: Fix initialisation of I/O result struct in nfs_pgio_rpcsetup
  NFSv4: Fix return values for nfs4_file_open()
  s390/bpf: use 32-bit index for tail calls
  ARM: OMAP2+: Fix omap4 errata warning on other SoCs
  s390/bpf: fix lcgr instruction encoding
  mwifiex: Fix three heap overflow at parsing element in cfg80211_ap_settings
  tty/serial: atmel: reschedule TX after RX was started
  serial: sprd: correct the wrong sequence of arguments
  KVM: coalesced_mmio: add bounds checking
  xen-netfront: do not assume sk_buff_head list is empty in error handling
  x86/boot: Add missing bootparam that breaks boot on some platforms
  media: tm6000: double free if usb disconnect while streaming
  USB: usbcore: Fix slab-out-of-bounds bug during device reset
  ARC: configs: Remove CONFIG_INITRAMFS_SOURCE from defconfigs
  MIPS: netlogic: xlr: Remove erroneous check in nlm_fmn_send()
  x86/build: Add -Wnoaddress-of-packed-member to REALMODE_CFLAGS, to silence GCC9 build warning
  crypto: talitos - check data blocksize in ablkcipher.
  crypto: talitos - check AES key size
  driver core: Fix use-after-free and double free on glue directory
  clk: rockchip: Don't yell about bad mmc phases when getting
  MIPS: VDSO: Use same -m%-float cflag as the kernel proper
  MIPS: VDSO: Prevent use of smp_processor_id()
  KVM: nVMX: handle page fault in vmread
  KVM: x86: work around leak of uninitialized stack contents
  KVM: s390: Do not leak kernel stack data in the KVM_S390_INTERRUPT ioctl
  genirq: Prevent NULL pointer dereference in resend_irqs()
  Btrfs: fix assertion failure during fsync and use of stale transaction
  Revert "MIPS: SiByte: Enable swiotlb for SWARM, LittleSur and BigSur"
  tun: fix use-after-free when register netdev failed
  tipc: add NULL pointer check before calling kfree_rcu
  tcp: fix tcp_ecn_withdraw_cwr() to clear TCP_ECN_QUEUE_CWR
  sctp: use transport pf_retrans in sctp_do_8_2_transport_strike
  sctp: Fix the link time qualifier of 'sctp_ctrlsock_exit()'
  sch_hhf: ensure quantum and hhf_non_hh_weight are non-zero
  net: Fix null de-reference of device refcount
  isdn/capi: check message length in capi_write()
  ipv6: Fix the link time qualifier of 'ping_v6_proc_exit_net()'
  cdc_ether: fix rndis support for Mediatek based smartphones
  bridge/mdb: remove wrong use of NLM_F_MULTI

Change-Id: I950778c771159febb721a4ebc2656c57ef40ad83
Signed-off-by: Srinivasarao P <spathi@codeaurora.org>

1 files changed, 52 insertions, 1 deletions

diff --git a/drivers/base/core.c b/drivers/base/core.c
index cef0f5c1bf85..4e27b3abf5a5 100644
--- a/drivers/base/core.c
+++ b/drivers/base/core.c
@@ -862,12 +862,63 @@ static inline struct kobject *get_glue_dir(struct device *dev)
  */
 static void cleanup_glue_dir(struct device *dev, struct kobject *glue_dir)
 {
+	unsigned int ref;
+
 	/* see if we live in a "glue" directory */
 	if (!live_in_glue_dir(glue_dir, dev))
 		return;
 
 	mutex_lock(&gdp_mutex);
-	if (!kobject_has_children(glue_dir))
+	/**
+	 * There is a race condition between removing glue directory
+	 * and adding a new device under the glue directory.
+	 *
+	 * CPU1:                                         CPU2:
+	 *
+	 * device_add()
+	 *   get_device_parent()
+	 *     class_dir_create_and_add()
+	 *       kobject_add_internal()
+	 *         create_dir()    // create glue_dir
+	 *
+	 *                                               device_add()
+	 *                                                 get_device_parent()
+	 *                                                   kobject_get() // get glue_dir
+	 *
+	 * device_del()
+	 *   cleanup_glue_dir()
+	 *     kobject_del(glue_dir)
+	 *
+	 *                                               kobject_add()
+	 *                                                 kobject_add_internal()
+	 *                                                   create_dir() // in glue_dir
+	 *                                                     sysfs_create_dir_ns()
+	 *                                                       kernfs_create_dir_ns(sd)
+	 *
+	 *       sysfs_remove_dir() // glue_dir->sd=NULL
+	 *       sysfs_put()        // free glue_dir->sd
+	 *
+	 *                                                         // sd is freed
+	 *                                                         kernfs_new_node(sd)
+	 *                                                           kernfs_get(glue_dir)
+	 *                                                           kernfs_add_one()
+	 *                                                           kernfs_put()
+	 *
+	 * Before CPU1 remove last child device under glue dir, if CPU2 add
+	 * a new device under glue dir, the glue_dir kobject reference count
+	 * will be increase to 2 in kobject_get(k). And CPU2 has been called
+	 * kernfs_create_dir_ns(). Meanwhile, CPU1 call sysfs_remove_dir()
+	 * and sysfs_put(). This result in glue_dir->sd is freed.
+	 *
+	 * Then the CPU2 will see a stale "empty" but still potentially used
+	 * glue dir around in kernfs_new_node().
+	 *
+	 * In order to avoid this happening, we also should make sure that
+	 * kernfs_node for glue_dir is released in CPU1 only when refcount
+	 * for glue_dir kobj is 1.
+	 */
+	ref = atomic_read(&glue_dir->kref.refcount);
+	if (!kobject_has_children(glue_dir) && !--ref)
 		kobject_del(glue_dir);
 	kobject_put(glue_dir);
 	mutex_unlock(&gdp_mutex);