3.10.64-stable review

This is the start of the stable review cycle for the 3.10.64 release. There are 38 patches in this series, all will be posted as a response to this one. If anyone has any issues with these being applied, please let me know.

Responses should be made by Fri Jan 9 01:49:29 UTC 2015. Anything received after that time might be too late.

The whole patch series can be found in one patch at:

	kernel.org/pub/linux/kernel/v3.0/stable-review/patch-3.10.64-rc1.gz

and the diffstat can be found below.

thanks,

greg k-h

-------------
Pseudo-Shortlog of commits:

Greg Kroah-Hartman <[email protected]>

    Linux 3.10.64-rc1

Filipe Manana <[email protected]>
Btrfs: fix fs corruption on transaction abort if device supports discard

Josef Bacik <[email protected]>
Btrfs: do not move em to modified list when unpinning

Michael Halcrow <[email protected]>
eCryptfs: Remove buggy and unnecessary write in file name decode routine

Tyler Hicks <[email protected]>
eCryptfs: Force RO mount when encrypted view is enabled

Jan Kara <[email protected]>
udf: Verify symlink size before loading it

Oleg Nesterov <[email protected]>
exit: pidns: alloc_pid() leaks pid_namespace if child_reaper is exiting

Jan Kara <[email protected]>
ncpfs: return proper error from NCP_IOC_SETROOT ioctl

Rabin Vincent <[email protected]>
crypto: af_alg - fix backlog handling

Eric W. Biederman <[email protected]>
userns: Unbreak the unprivileged remount tests

Eric W. Biederman <[email protected]>
userns: Allow setting gid_maps without privilege when setgroups is disabled

Eric W. Biederman <[email protected]>
userns: Add a knob to disable setgroups on a per user namespace basis

Eric W. Biederman <[email protected]>
userns: Rename id_map_mutex to userns_state_mutex

Eric W. Biederman <[email protected]>
userns: Only allow the creator of the userns unprivileged mappings

Eric W. Biederman <[email protected]>
userns: Check euid no fsuid when establishing an unprivileged uid mapping

Eric W. Biederman <[email protected]>
userns: Don't allow unprivileged creation of gid mappings

Eric W. Biederman <[email protected]>
userns: Don't allow setgroups until a gid mapping has been setablished

Eric W. Biederman <[email protected]>
userns: Document what the invariant required for safe unprivileged mappings.

Eric W. Biederman <[email protected]>
groups: Consolidate the setgroups permission checks

Eric W. Biederman <[email protected]>
umount: Disallow unprivileged mount force

Eric W. Biederman <[email protected]>
mnt: Update unprivileged remount test

Eric W. Biederman <[email protected]>
mnt: Implicitly add MNT_NODEV on remount when it was implicitly added by mount

Johannes Berg <[email protected]>
mac80211: free management frame keys when removing station

Andreas Müller <[email protected]>
mac80211: fix multicast LED blinking and counter

Takashi Iwai <[email protected]>
KEYS: Fix stale key registration at error path

Jan Kara <[email protected]>
isofs: Fix unchecked printing of ER records

Andy Lutomirski <[email protected]>
x86/tls: Don't validate lm in set_thread_area() after all

Dan Carpenter <[email protected]>
dm space map metadata: fix sm_bootstrap_get_nr_blocks()

Darrick J. Wong <[email protected]>
dm bufio: fix memleak when using a dm_buffer's inline bio

Peng Tao <[email protected]>
nfs41: fix nfs4_proc_layoutget error handling

[email protected] <[email protected]>
megaraid_sas: corrected return of wait_event from abort frame path

Baruch Siach <[email protected]>
mmc: block: add newline to sysfs display of force_ro

Dmitry Eremin-Solenikov <[email protected]>
mfd: tc6393xb: Fail ohci suspend if full state restore is required

NeilBrown <[email protected]>
md/bitmap: always wait for writes on unplug.

Andy Lutomirski <[email protected]>
x86, kvm: Clear paravirt_enabled on KVM guests for espfix32's benefit

Andy Lutomirski <[email protected]>
x86_64, switch_to(): Load TLS descriptors before switching DS and ES

Andy Lutomirski <[email protected]>
x86/tls: Disallow unusual TLS segments

Andy Lutomirski <[email protected]>
x86/tls: Validate TLS entries to protect espfix

Jan Kara <[email protected]>
isofs: Fix infinite looping over CE entries

-------------

Diffstat:

 Makefile                                           |   4 +-
 arch/s390/kernel/compat_linux.c                    |   2 +-
 arch/x86/include/uapi/asm/ldt.h                    |   7 +
 arch/x86/kernel/kvm.c                              |   9 +-
 arch/x86/kernel/kvmclock.c                         |   1 -
 arch/x86/kernel/process_64.c                       | 101 +++++++---
 arch/x86/kernel/tls.c                              |  39 ++++
 crypto/af_alg.c                                    |   3 +
 drivers/md/bitmap.c                                |  16 +-
 drivers/md/dm-bufio.c                              |  20 +-
 drivers/md/persistent-data/dm-space-map-metadata.c |   4 +-
 drivers/mfd/tc6393xb.c                             |  13 +-
 drivers/mmc/card/block.c                           |   2 +-
 drivers/scsi/megaraid/megaraid_sas_base.c          |   2 +-
 fs/btrfs/disk-io.c                                 |   6 -
 fs/btrfs/extent-tree.c                             |  10 +-
 fs/btrfs/extent_map.c                              |   2 -
 fs/ecryptfs/crypto.c                               |   1 -
 fs/ecryptfs/file.c                                 |  12 --
 fs/ecryptfs/main.c                                 |  16 +-
 fs/isofs/rock.c                                    |   9 +
 fs/namespace.c                                     |  11 +-
 fs/ncpfs/ioctl.c                                   |   1 -
 fs/nfs/nfs4proc.c                                  |   6 +-
 fs/proc/base.c                                     |  53 ++++++
 fs/udf/symlink.c                                   |  17 +-
 include/linux/cred.h                               |   1 +
 include/linux/user_namespace.h                     |  12 ++
 kernel/groups.c                                    |  11 +-
 kernel/pid.c                                       |   2 +
 kernel/uid16.c                                     |   2 +-
 kernel/user.c                                      |   1 +
 kernel/user_namespace.c                            | 125 +++++++++++--
 net/mac80211/key.c                                 |   2 +-
 net/mac80211/rx.c                                  |  11 +-
 security/keys/encrypted-keys/encrypted.c           |   5 +-
 .../selftests/mount/unprivileged-remount-test.c    | 204 +++++++++++++++++----
 37 files changed, 599 insertions(+), 144 deletions(-)

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: Jan Kara <[email protected]>

commit f54e18f1b831c92f6512d2eedb224cd63d607d3d upstream.

Rock Ridge extensions define so called Continuation Entries (CE) which define where is further space with Rock Ridge data. Corrupted isofs image can contain arbitrarily long chain of these, including a one containing loop and thus causing kernel to end in an infinite loop when traversing these entries.

Limit the traversal to 32 entries which should be more than enough space to store all the Rock Ridge data.

---
 fs/isofs/rock.c |    6 ++++++
 1 file changed, 6 insertions(+)

@@ -30,6 +30,7 @@ struct rock_state {
 	int cont_size;
 	int cont_extent;
 	int cont_offset;
+	int cont_loops;
 	struct inode *inode;
 };
 

@@ -73,6 +74,9 @@ static void init_rock_state(struct rock_
 	rs->inode = inode;
 }
 
+/* Maximum number of Rock Ridge continuation entries */
+#define RR_MAX_CE_ENTRIES 32
+
 /*
  * Returns 0 if the caller should continue scanning, 1 if the scan must end
  * and -ve on error.

@@ -105,6 +109,8 @@ static int rock_continue(struct rock_sta
 			goto out;
 		}
 		ret = -EIO;
+		if (++rs->cont_loops >= RR_MAX_CE_ENTRIES)
+			goto out;
 		bh = sb_bread(rs->inode->i_sb, rs->cont_extent);
 		if (bh) {
 			memcpy(rs->buffer, bh->b_data + rs->cont_offset,

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: Andy Lutomirski <[email protected]>

commit 41bdc78544b8a93a9c6814b8bbbfef966272abbe upstream.

Installing a 16-bit RW data segment into the GDT defeats espfix. AFAICT this will not affect glibc, Wine, or dosemu at all.

---
 arch/x86/kernel/tls.c |   23 +++++++++++++++++++++++
 1 file changed, 23 insertions(+)

@@ -27,6 +27,21 @@ static int get_free_idx(void)
 	return -ESRCH;
 }
 
+static bool tls_desc_okay(const struct user_desc *info)
+{
+	if (LDT_empty(info))
+		return true;
+
+	/*
+	 * espfix is required for 16-bit data segments, but espfix
+	 * only works for LDT segments.
+	 */
+	if (!info->seg_32bit)
+		return false;
+
+	return true;
+}
+
 static void set_tls_desc(struct task_struct *p, int idx,
 			 const struct user_desc *info, int n)
 {

@@ -66,6 +81,9 @@ int do_set_thread_area(struct task_struc
 	if (copy_from_user(&info, u_info, sizeof(info)))
 		return -EFAULT;
 
+	if (!tls_desc_okay(&info))
+		return -EINVAL;
+
 	if (idx == -1)
 		idx = info.entry_number;
 

@@ -192,6 +210,7 @@ int regset_tls_set(struct task_struct *t
 {
 	struct user_desc infobuf[GDT_ENTRY_TLS_ENTRIES];
 	const struct user_desc *info;
+	int i;
 
 	if (pos >= GDT_ENTRY_TLS_ENTRIES * sizeof(struct user_desc) ||
 	    (pos % sizeof(struct user_desc)) != 0 ||

@@ -205,6 +224,10 @@ int regset_tls_set(struct task_struct *t
 	else
 		info = infobuf;
 
+	for (i = 0; i < count / sizeof(struct user_desc); i++)
+		if (!tls_desc_okay(info + i))
+			return -EINVAL;
+
 	set_tls_desc(target,
 		     GDT_ENTRY_TLS_MIN + (pos / sizeof(struct user_desc)),
 		     info, count / sizeof(struct user_desc));

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: Andy Lutomirski <[email protected]>

commit 0e58af4e1d2166e9e33375a0f121e4867010d4f8 upstream.

Users have no business installing custom code segments into the GDT, and segments that are not present but are otherwise valid are a historical source of interesting attacks.

For completeness, block attempts to set the L bit. (Prior to this patch, the L bit would have been silently dropped.)

This is an ABI break. I've checked glibc, musl, and Wine, and none of them look like they'll have any trouble.

Note to stable maintainers: this is a hardening patch that fixes no known bugs. Given the possibility of ABI issues, this probably shouldn't be backported quickly.

---
 arch/x86/kernel/tls.c |   22 ++++++++++++++++++++++
 1 file changed, 22 insertions(+)

@@ -39,6 +39,28 @@ static bool tls_desc_okay(const struct u
 	if (!info->seg_32bit)
 		return false;
 
+	/* Only allow data segments in the TLS array. */
+	if (info->contents > 1)
+		return false;
+
+	/*
+	 * Non-present segments with DPL 3 present an interesting attack
+	 * surface.  The kernel should handle such segments correctly,
+	 * but TLS is very difficult to protect in a sandbox, so prevent
+	 * such segments from being created.
+	 *
+	 * If userspace needs to remove a TLS entry, it can still delete
+	 * it outright.
+	 */
+	if (info->seg_not_present)
+		return false;
+
+#ifdef CONFIG_X86_64
+	/* The L bit makes no sense for data. */
+	if (info->lm)
+		return false;
+#endif
+
 	return true;
 }
 

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: Andy Lutomirski <[email protected]>

commit f647d7c155f069c1a068030255c300663516420e upstream.

Otherwise, if buggy user code points DS or ES into the TLS array, they would be corrupted after a context switch.

This also significantly improves the comments and documents some gotchas in the code.

Before this patch, the both tests below failed. With this patch, the es test passes, although the gsbase test still fails.

----- begin es test -----

/*
* Copyright (c) 2014 Andy Lutomirski
* GPL v2
*/

static unsigned short GDT3(int idx)

{
	return (idx << 3) | 3;
}

static int create_tls(int idx, unsigned int base)

{
	struct user_desc desc = {
		.entry_number    = idx,
		.base_addr       = base,
		.limit           = 0xfffff,
		.seg_32bit       = 1,
		.contents        = 0, /* Data, grow-up */
		.read_exec_only  = 0,
		.limit_in_pages  = 1,
		.seg_not_present = 0,
		.useable         = 0,
	};

	if (syscall(SYS_set_thread_area, &desc) != 0)
		err(1, "set_thread_area");

	return desc.entry_number;
}

int main()

{
	int idx = create_tls(-1, 0);
	printf("Allocated GDT index %d\n", idx);

	unsigned short orig_es;
	asm volatile ("mov %%es,%0" : "=rm" (orig_es));

	int errors = 0;
	int total = 1000;
	for (int i = 0; i < total; i++) {
		asm volatile ("mov %0,%%es" : : "rm" (GDT3(idx)));
		usleep(100);

		unsigned short es;
		asm volatile ("mov %%es,%0" : "=rm" (es));
		asm volatile ("mov %0,%%es" : : "rm" (orig_es));
		if (es != GDT3(idx)) {
			if (errors == 0)
				printf("[FAIL]\tES changed from 0x%hx to 0x%hx\n",
				       GDT3(idx), es);
			errors++;
		}
	}

	if (errors) {
		printf("[FAIL]\tES was corrupted %d/%d times\n", errors, total);
		return 1;
	} else {
		printf("[OK]\tES was preserved\n");
		return 0;
	}
}

----- end es test -----

----- begin gsbase test -----

/*
* gsbase.c, a gsbase test
* Copyright (c) 2014 Andy Lutomirski
* GPL v2
*/

static unsigned char *testptr, *testptr2;

static unsigned char read_gs_testvals(void)

{
	unsigned char ret;
	asm volatile ("movb %%gs:%1, %0" : "=r" (ret) : "m" (*testptr));
	return ret;
}

int main()

{
	int errors = 0;

	testptr = mmap((void *)0x200000000UL, 1, PROT_READ | PROT_WRITE,
		       MAP_PRIVATE | MAP_FIXED | MAP_ANONYMOUS, -1, 0);
	if (testptr == MAP_FAILED)
		err(1, "mmap");

	testptr2 = mmap((void *)0x300000000UL, 1, PROT_READ | PROT_WRITE,
		       MAP_PRIVATE | MAP_FIXED | MAP_ANONYMOUS, -1, 0);
	if (testptr2 == MAP_FAILED)
		err(1, "mmap");

	*testptr = 0;
	*testptr2 = 1;

	if (syscall(SYS_arch_prctl, ARCH_SET_GS,
		    (unsigned long)testptr2 - (unsigned long)testptr) != 0)
		err(1, "ARCH_SET_GS");

	usleep(100);

	if (read_gs_testvals() == 1) {
		printf("[OK]\tARCH_SET_GS worked\n");
	} else {
		printf("[FAIL]\tARCH_SET_GS failed\n");
		errors++;
	}

	asm volatile ("mov %0,%%gs" : : "r" (0));

	if (read_gs_testvals() == 0) {
		printf("[OK]\tWriting 0 to gs worked\n");
	} else {
		printf("[FAIL]\tWriting 0 to gs failed\n");
		errors++;
	}

	usleep(100);

	if (read_gs_testvals() == 0) {
		printf("[OK]\tgsbase is still zero\n");
	} else {
		printf("[FAIL]\tgsbase was corrupted\n");
		errors++;
	}

	return errors == 0 ? 0 : 1;
}

----- end gsbase test -----

---
 arch/x86/kernel/process_64.c |  101 +++++++++++++++++++++++++++++++------------
 1 file changed, 73 insertions(+), 28 deletions(-)

@@ -279,24 +279,9 @@ __switch_to(struct task_struct *prev_p,
 
 	fpu = switch_fpu_prepare(prev_p, next_p, cpu);
 
-	/*
-	 * Reload esp0, LDT and the page table pointer:
-	 */
+	/* Reload esp0 and ss1. */
 	load_sp0(tss, next);
 
-	/*
-	 * Switch DS and ES.
-	 * This won't pick up thread selector changes, but I guess that is ok.
-	 */
-	savesegment(es, prev->es);
-	if (unlikely(next->es | prev->es))
-		loadsegment(es, next->es);
-
-	savesegment(ds, prev->ds);
-	if (unlikely(next->ds | prev->ds))
-		loadsegment(ds, next->ds);
-
-
 	/* We must save %fs and %gs before load_TLS() because
 	 * %fs and %gs may be cleared by load_TLS().
 	 *

@@ -305,41 +290,101 @@ __switch_to(struct task_struct *prev_p,
 	savesegment(fs, fsindex);
 	savesegment(gs, gsindex);
 
+	/*
+	 * Load TLS before restoring any segments so that segment loads
+	 * reference the correct GDT entries.
+	 */
 	load_TLS(next, cpu);
 
 	/*
-	 * Leave lazy mode, flushing any hypercalls made here.
-	 * This must be done before restoring TLS segments so
-	 * the GDT and LDT are properly updated, and must be
-	 * done before math_state_restore, so the TS bit is up
-	 * to date.
+	 * Leave lazy mode, flushing any hypercalls made here.  This
+	 * must be done after loading TLS entries in the GDT but before
+	 * loading segments that might reference them, and and it must
+	 * be done before math_state_restore, so the TS bit is up to
+	 * date.
 	 */
 	arch_end_context_switch(next_p);
 
+	/* Switch DS and ES.
+	 *
+	 * Reading them only returns the selectors, but writing them (if
+	 * nonzero) loads the full descriptor from the GDT or LDT.  The
+	 * LDT for next is loaded in switch_mm, and the GDT is loaded
+	 * above.
+	 *
+	 * We therefore need to write new values to the segment
+	 * registers on every context switch unless both the new and old
+	 * values are zero.
+	 *
+	 * Note that we don't need to do anything for CS and SS, as
+	 * those are saved and restored as part of pt_regs.
+	 */
+	savesegment(es, prev->es);
+	if (unlikely(next->es | prev->es))
+		loadsegment(es, next->es);
+
+	savesegment(ds, prev->ds);
+	if (unlikely(next->ds | prev->ds))
+		loadsegment(ds, next->ds);
+
 	/*
 	 * Switch FS and GS.
 	 *
-	 * Segment register != 0 always requires a reload.  Also
-	 * reload when it has changed.  When prev process used 64bit
-	 * base always reload to avoid an information leak.
+	 * These are even more complicated than FS and GS: they have
+	 * 64-bit bases are that controlled by arch_prctl.  Those bases
+	 * only differ from the values in the GDT or LDT if the selector
+	 * is 0.
+	 *
+	 * Loading the segment register resets the hidden base part of
+	 * the register to 0 or the value from the GDT / LDT.  If the
+	 * next base address zero, writing 0 to the segment register is
+	 * much faster than using wrmsr to explicitly zero the base.
+	 *
+	 * The thread_struct.fs and thread_struct.gs values are 0
+	 * if the fs and gs bases respectively are not overridden
+	 * from the values implied by fsindex and gsindex.  They
+	 * are nonzero, and store the nonzero base addresses, if
+	 * the bases are overridden.
+	 *
+	 * (fs != 0 && fsindex != 0) || (gs != 0 && gsindex != 0) should
+	 * be impossible.
+	 *
+	 * Therefore we need to reload the segment registers if either
+	 * the old or new selector is nonzero, and we need to override
+	 * the base address if next thread expects it to be overridden.
+	 *
+	 * This code is unnecessarily slow in the case where the old and
+	 * new indexes are zero and the new base is nonzero -- it will
+	 * unnecessarily write 0 to the selector before writing the new
+	 * base address.
+	 *
+	 * Note: This all depends on arch_prctl being the only way that
+	 * user code can override the segment base.  Once wrfsbase and
+	 * wrgsbase are enabled, most of this code will need to change.
 	 */
 	if (unlikely(fsindex | next->fsindex | prev->fs)) {
 		loadsegment(fs, next->fsindex);
+
 		/*
-		 * Check if the user used a selector != 0; if yes
-		 *  clear 64bit base, since overloaded base is always
-		 *  mapped to the Null selector
+		 * If user code wrote a nonzero value to FS, then it also
+		 * cleared the overridden base address.
+		 *
+		 * XXX: if user code wrote 0 to FS and cleared the base
+		 * address itself, we won't notice and we'll incorrectly
+		 * restore the prior base address next time we reschdule
+		 * the process.
 		 */
 		if (fsindex)
 			prev->fs = 0;
 	}
-	/* when next process has a 64bit base use it */
 	if (next->fs)
 		wrmsrl(MSR_FS_BASE, next->fs);
 	prev->fsindex = fsindex;
 
 	if (unlikely(gsindex | next->gsindex | prev->gs)) {
 		load_gs_index(next->gsindex);
+
+		/* This works (and fails) the same way as fsindex above. */
 		if (gsindex)
 			prev->gs = 0;
 	}

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: Andy Lutomirski <[email protected]>

commit 29fa6825463c97e5157284db80107d1bfac5d77b upstream.

paravirt_enabled has the following effects:

- Disables the F00F bug workaround warning. There is no F00F bug workaround any more because Linux's standard IDT handling already works around the F00F bug, but the warning still exists. This is only cosmetic, and, in any event, there is no such thing as KVM on a CPU with the F00F bug.

- Disables 32-bit APM BIOS detection. On a KVM paravirt system, there should be no APM BIOS anyway.

- Disables tboot. I think that the tboot code should check the CPUID hypervisor bit directly if it matters.

- paravirt_enabled disables espfix32. espfix32 should *not* be disabled under KVM paravirt.

The last point is the purpose of this patch. It fixes a leak of the high 16 bits of the kernel stack address on 32-bit KVM paravirt guests. Fixes CVE-2014-8134.

---
 arch/x86/kernel/kvm.c      |    9 ++++++++-
 arch/x86/kernel/kvmclock.c |    1 -
 2 files changed, 8 insertions(+), 2 deletions(-)

@@ -279,7 +279,14 @@ do_async_page_fault(struct pt_regs *regs
 static void __init paravirt_ops_setup(void)
 {
 	pv_info.name = "KVM";
-	pv_info.paravirt_enabled = 1;
+
+	/*
+	 * KVM isn't paravirt in the sense of paravirt_enabled.  A KVM
+	 * guest kernel works like a bare metal kernel with additional
+	 * features, and paravirt_enabled is about features that are
+	 * missing.
+	 */
+	pv_info.paravirt_enabled = 0;
 
 	if (kvm_para_has_feature(KVM_FEATURE_NOP_IO_DELAY))
 		pv_cpu_ops.io_delay = kvm_io_delay;

@@ -265,7 +265,6 @@ void __init kvmclock_init(void)
 #endif
 	kvm_get_preset_lpj();
 	clocksource_register_hz(&kvm_clock, NSEC_PER_SEC);
-	pv_info.paravirt_enabled = 1;
 	pv_info.name = "KVM";
 
 	if (kvm_para_has_feature(KVM_FEATURE_CLOCKSOURCE_STABLE_BIT))

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: NeilBrown <[email protected]>

commit 4b5060ddae2b03c5387321fafc089d242225697a upstream.

If two threads call bitmap_unplug at the same time, then one might schedule all the writes, and the other might decide that it doesn't need to wait. But really it does.

It rarely hurts to wait when it isn't absolutely necessary, and the current code doesn't really focus on 'absolutely necessary' anyway. So just wait always.

This can potentially lead to data corruption if a crash happens at an awkward time and data was written before the bitmap was updated. It is very unlikely, but this should go to -stable just to be safe. Appropriate for any -stable.

---
 drivers/md/bitmap.c |   16 ++++++----------
 1 file changed, 6 insertions(+), 10 deletions(-)

@@ -883,7 +883,6 @@ void bitmap_unplug(struct bitmap *bitmap
 {
 	unsigned long i;
 	int dirty, need_write;
-	int wait = 0;
 
 	if (!bitmap || !bitmap->storage.filemap ||
 	    test_bit(BITMAP_STALE, &bitmap->flags))

@@ -901,16 +900,13 @@ void bitmap_unplug(struct bitmap *bitmap
 			clear_page_attr(bitmap, i, BITMAP_PAGE_PENDING);
 			write_page(bitmap, bitmap->storage.filemap[i], 0);
 		}
-		if (dirty)
-			wait = 1;
-	}
-	if (wait) { /* if any writes were performed, we need to wait on them */
-		if (bitmap->storage.file)
-			wait_event(bitmap->write_wait,
-				   atomic_read(&bitmap->pending_writes)==0);
-		else
-			md_super_wait(bitmap->mddev);
 	}
+	if (bitmap->storage.file)
+		wait_event(bitmap->write_wait,
+			   atomic_read(&bitmap->pending_writes)==0);
+	else
+		md_super_wait(bitmap->mddev);
+
 	if (test_bit(BITMAP_WRITE_ERROR, &bitmap->flags))
 		bitmap_file_kick(bitmap);
 }

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: Dmitry Eremin-Solenikov <[email protected]>

commit 1a5fb99de4850cba710d91becfa2c65653048589 upstream.

Some boards with TC6393XB chip require full state restore during system resume thanks to chip's VCC being cut off during suspend (Sharp SL-6000 tosa is one of them). Failing to do so would result in ohci Oops on resume due to internal memory contentes being changed. Fail ohci suspend on tc6393xb is full state restore is required.

Recommended workaround is to unbind tmio-ohci driver before suspend and rebind it after resume.

---
 drivers/mfd/tc6393xb.c |   13 ++++++++++++-
 1 file changed, 12 insertions(+), 1 deletion(-)

@@ -263,6 +263,17 @@ static int tc6393xb_ohci_disable(struct
 	return 0;
 }
 
+static int tc6393xb_ohci_suspend(struct platform_device *dev)
+{
+	struct tc6393xb_platform_data *tcpd = dev_get_platdata(dev->dev.parent);
+
+	/* We can't properly store/restore OHCI state, so fail here */
+	if (tcpd->resume_restore)
+		return -EBUSY;
+
+	return tc6393xb_ohci_disable(dev);
+}
+
 static int tc6393xb_fb_enable(struct platform_device *dev)
 {
 	struct tc6393xb *tc6393xb = dev_get_drvdata(dev->dev.parent);

@@ -403,7 +414,7 @@ static struct mfd_cell tc6393xb_cells[]
 		.num_resources = ARRAY_SIZE(tc6393xb_ohci_resources),
 		.resources = tc6393xb_ohci_resources,
 		.enable = tc6393xb_ohci_enable,
-		.suspend = tc6393xb_ohci_disable,
+		.suspend = tc6393xb_ohci_suspend,
 		.resume = tc6393xb_ohci_enable,
 		.disable = tc6393xb_ohci_disable,
 	},

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: Baruch Siach <[email protected]>

commit 0031a98a85e9fca282624bfc887f9531b2768396 upstream.

Make force_ro consistent with other sysfs entries.

---
 drivers/mmc/card/block.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -257,7 +257,7 @@ static ssize_t force_ro_show(struct devi
 	int ret;
 	struct mmc_blk_data *md = mmc_blk_get(dev_to_disk(dev));
 
-	ret = snprintf(buf, PAGE_SIZE, "%d",
+	ret = snprintf(buf, PAGE_SIZE, "%d\n",
 		       get_disk_ro(dev_to_disk(dev)) ^
 		       md->read_only);
 	mmc_blk_put(md);

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: "[email protected]" <[email protected]>

commit 170c238701ec38b1829321b17c70671c101bac55 upstream.

Corrected wait_event() call which was waiting for wrong completion status (0xFF).

---
 drivers/scsi/megaraid/megaraid_sas_base.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -933,7 +933,7 @@ megasas_issue_blocked_abort_cmd(struct m
 	abort_fr->abort_mfi_phys_addr_hi = 0;
 
 	cmd->sync_cmd = 1;
-	cmd->cmd_status = 0xFF;
+	cmd->cmd_status = ENODATA;
 
 	instance->instancet->issue_dcmd(instance, cmd);
 

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: Peng Tao <[email protected]>

commit 4bd5a980de87d2b5af417485bde97b8eb3d6cf6a upstream.

nfs4_layoutget_release() drops layout hdr refcnt. Grab the refcnt early so that it is safe to call .release in case nfs4_alloc_pages fails.

---
 fs/nfs/nfs4proc.c |    6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

@@ -6418,6 +6418,9 @@ nfs4_proc_layoutget(struct nfs4_layoutge
 
 	dprintk("--> %s\n", __func__);
 
+	/* nfs4_layoutget_release calls pnfs_put_layout_hdr */
+	pnfs_get_layout_hdr(NFS_I(inode)->layout);
+
 	lgp->args.layout.pages = nfs4_alloc_pages(max_pages, gfp_flags);
 	if (!lgp->args.layout.pages) {
 		nfs4_layoutget_release(lgp);

@@ -6430,9 +6433,6 @@ nfs4_proc_layoutget(struct nfs4_layoutge
 	lgp->res.seq_res.sr_slot = NULL;
 	nfs41_init_sequence(&lgp->args.seq_args, &lgp->res.seq_res, 0);
 
-	/* nfs4_layoutget_release calls pnfs_put_layout_hdr */
-	pnfs_get_layout_hdr(NFS_I(inode)->layout);
-
 	task = rpc_run_task(&task_setup_data);
 	if (IS_ERR(task))
 		return ERR_CAST(task);

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: "Darrick J. Wong" <[email protected]>

commit 445559cdcb98a141f5de415b94fd6eaccab87e6d upstream.

When dm-bufio sets out to use the bio built into a struct dm_buffer to issue an IO, it needs to call bio_reset after it's done with the bio so that we can free things attached to the bio such as the integrity payload. Therefore, inject our own endio callback to take care of the bio_reset after calling submit_io's end_io callback.

Test case:
1. modprobe scsi_debug delay=0 dif=1 dix=199 ato=1 dev_size_mb=300
2. Set up a dm-bufio client, e.g. dm-verity, on the scsi_debug device
3. Repeatedly read metadata and watch kmalloc-192 leak!

---
 drivers/md/dm-bufio.c |   20 +++++++++++++++++++-
 1 file changed, 19 insertions(+), 1 deletion(-)

@@ -529,6 +529,19 @@ static void use_dmio(struct dm_buffer *b
 		end_io(&b->bio, r);
 }
 
+static void inline_endio(struct bio *bio, int error)
+{
+	bio_end_io_t *end_fn = bio->bi_private;
+
+	/*
+	 * Reset the bio to free any attached resources
+	 * (e.g. bio integrity profiles).
+	 */
+	bio_reset(bio);
+
+	end_fn(bio, error);
+}
+
 static void use_inline_bio(struct dm_buffer *b, int rw, sector_t block,
 			   bio_end_io_t *end_io)
 {

@@ -540,7 +553,12 @@ static void use_inline_bio(struct dm_buf
 	b->bio.bi_max_vecs = DM_BUFIO_INLINE_VECS;
 	b->bio.bi_sector = block << b->c->sectors_per_block_bits;
 	b->bio.bi_bdev = b->c->bdev;
-	b->bio.bi_end_io = end_io;
+	b->bio.bi_end_io = inline_endio;
+	/*
+	 * Use of .bi_private isn't a problem here because
+	 * the dm_buffer's inline bio is local to bufio.
+	 */
+	b->bio.bi_private = end_io;
 
 	/*
 	 * We assume that if len >= PAGE_SIZE ptr is page-aligned.

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: Dan Carpenter <[email protected]>

commit c1c6156fe4d4577444b769d7edd5dd503e57bbc9 upstream.

This function isn't right and it causes a static checker warning:

	drivers/md/dm-thin.c:3016 maybe_resize_data_dev()
	error: potentially using uninitialized 'sb_data_size'.

It should set "*count" and return zero on success the same as the sm_metadata_get_nr_blocks() function does earlier.

---
 drivers/md/persistent-data/dm-space-map-metadata.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

@@ -493,7 +493,9 @@ static int sm_bootstrap_get_nr_blocks(st
 {
 	struct sm_metadata *smm = container_of(sm, struct sm_metadata, sm);
 
-	return smm->ll.nr_blocks;
+	*count = smm->ll.nr_blocks;
+
+	return 0;
 }
 
 static int sm_bootstrap_get_nr_free(struct dm_space_map *sm, dm_block_t *count)

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: Andy Lutomirski <[email protected]>

commit 3fb2f4237bb452eb4e98f6a5dbd5a445b4fed9d0 upstream.

It turns out that there's a lurking ABI issue. GCC, when compiling this in a 32-bit program:

struct user_desc desc = {
	.entry_number    = idx,
	.base_addr       = base,
	.limit           = 0xfffff,
	.seg_32bit       = 1,
	.contents        = 0, /* Data, grow-up */
	.read_exec_only  = 0,
	.limit_in_pages  = 1,
	.seg_not_present = 0,
	.useable         = 0,
};

will leave .lm uninitialized. This means that anything in the kernel that reads user_desc.lm for 32-bit tasks is unreliable.

Revert the .lm check in set_thread_area(). The value never did anything in the first place.

---
 arch/x86/include/uapi/asm/ldt.h |    7 +++++++
 arch/x86/kernel/tls.c           |    6 ------
 2 files changed, 7 insertions(+), 6 deletions(-)

@@ -28,6 +28,13 @@ struct user_desc {
 	unsigned int  seg_not_present:1;
 	unsigned int  useable:1;
 #ifdef __x86_64__
+	/*
+	 * Because this bit is not present in 32-bit user code, user
+	 * programs can pass uninitialized values here.  Therefore, in
+	 * any context in which a user_desc comes from a 32-bit program,
+	 * the kernel must act as though lm == 0, regardless of the
+	 * actual value.
+	 */
 	unsigned int  lm:1;
 #endif
 };

@@ -55,12 +55,6 @@ static bool tls_desc_okay(const struct u
 	if (info->seg_not_present)
 		return false;
 
-#ifdef CONFIG_X86_64
-	/* The L bit makes no sense for data. */
-	if (info->lm)
-		return false;
-#endif
-
 	return true;
 }
 

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: Jan Kara <[email protected]>

commit 4e2024624e678f0ebb916e6192bd23c1f9fdf696 upstream.

We didn't check length of rock ridge ER records before printing them. Thus corrupted isofs image can cause us to access and print some memory behind the buffer with obvious consequences.

---
 fs/isofs/rock.c |    3 +++
 1 file changed, 3 insertions(+)

@@ -362,6 +362,9 @@ repeat:
 			rs.cont_size = isonum_733(rr->u.CE.size);
 			break;
 		case SIG('E', 'R'):
+			/* Invalid length of ER tag id? */
+			if (rr->u.ER.len_id + offsetof(struct rock_ridge, u.ER.data) > rr->len)
+				goto out;
 			ISOFS_SB(inode->i_sb)->s_rock = 1;
 			printk(KERN_DEBUG "ISO 9660 Extensions: ");
 			{

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: Takashi Iwai <[email protected]>

commit b26bdde5bb27f3f900e25a95e33a0c476c8c2c48 upstream.

When loading encrypted-keys module, if the last check of aes_get_sizes() in init_encrypted() fails, the driver just returns an error without unregistering its key type. This results in the stale entry in the list. In addition to memory leaks, this leads to a kernel crash when registering a new key type later.

This patch fixes the problem by swapping the calls of aes_get_sizes() and register_key_type(), and releasing resources properly at the error paths.

Bugzilla: https://bugzilla.opensuse.org/show_bug.cgi?id=908163

---
 security/keys/encrypted-keys/encrypted.c |    5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

@@ -1018,10 +1018,13 @@ static int __init init_encrypted(void)
 	ret = encrypted_shash_alloc();
 	if (ret < 0)
 		return ret;
+	ret = aes_get_sizes();
+	if (ret < 0)
+		goto out;
 	ret = register_key_type(&key_type_encrypted);
 	if (ret < 0)
 		goto out;
-	return aes_get_sizes();
+	return 0;
 out:
 	encrypted_shash_release();
 	return ret;

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: Johannes Berg <[email protected]>

commit 28a9bc68124c319b2b3dc861e80828a8865fd1ba upstream.

When writing the code to allow per-station GTKs, I neglected to take into account the management frame keys (index 4 and 5) when freeing the station and only added code to free the first four data frame keys.

Fix this by iterating the array of keys over the right length.

---
 net/mac80211/key.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -607,7 +607,7 @@ void ieee80211_free_sta_keys(struct ieee
 	int i;
 
 	mutex_lock(&local->key_mtx);
-	for (i = 0; i < NUM_DEFAULT_KEYS; i++) {
+	for (i = 0; i < ARRAY_SIZE(sta->gtk); i++) {
 		key = key_mtx_dereference(local, sta->gtk[i]);
 		if (!key)
 			continue;

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: "Eric W. Biederman" <[email protected]>

commit 3e1866410f11356a9fd869beb3e95983dc79c067 upstream.

Now that remount is properly enforcing the rule that you can't remove nodev at least sandstorm.io is breaking when performing a remount.

It turns out that there is an easy intuitive solution implicitly add nodev on remount when nodev was implicitly added on mount.

---
 fs/namespace.c |    8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

@@ -1816,7 +1816,13 @@ static int do_remount(struct path *path,
 	}
 	if ((mnt->mnt.mnt_flags & MNT_LOCK_NODEV) &&
 	    !(mnt_flags & MNT_NODEV)) {
-		return -EPERM;
+		/* Was the nodev implicitly added in mount? */
+		if ((mnt->mnt_ns->user_ns != &init_user_ns) &&
+		    !(sb->s_type->fs_flags & FS_USERNS_DEV_MOUNT)) {
+			mnt_flags |= MNT_NODEV;
+		} else {
+			return -EPERM;
+		}
 	}
 	if ((mnt->mnt.mnt_flags & MNT_LOCK_NOSUID) &&
 	    !(mnt_flags & MNT_NOSUID)) {

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: "Eric W. Biederman" <[email protected]>

commit 4a44a19b470a886997d6647a77bb3e38dcbfa8c5 upstream.

- MNT_NODEV should be irrelevant except when reading back mount flags, no longer specify MNT_NODEV on remount.

- Test MNT_NODEV on devpts where it is meaningful even for unprivileged mounts.

- Add a test to verify that remount of a prexisting mount with the same flags is allowed and does not change those flags.

- Cleanup up the definitions of MS_REC, MS_RELATIME, MS_STRICTATIME that are used when the code is built in an environment without them.

- Correct the test error messages when tests fail. There were not 5 tests that tested MS_RELATIME.

---
 tools/testing/selftests/mount/unprivileged-remount-test.c |  172 +++++++++++---
 1 file changed, 142 insertions(+), 30 deletions(-)

@@ -6,6 +6,8 @@
 #include <sys/types.h>
 #include <sys/mount.h>
 #include <sys/wait.h>
+#include <sys/vfs.h>
+#include <sys/statvfs.h>
 #include <stdlib.h>
 #include <unistd.h>
 #include <fcntl.h>

@@ -32,11 +34,14 @@
 # define CLONE_NEWPID 0x20000000
 #endif
 
+#ifndef MS_REC
+# define MS_REC 16384
+#endif
 #ifndef MS_RELATIME
-#define MS_RELATIME (1 << 21)
+# define MS_RELATIME (1 << 21)
 #endif
 #ifndef MS_STRICTATIME
-#define MS_STRICTATIME (1 << 24)
+# define MS_STRICTATIME (1 << 24)
 #endif
 
 static void die(char *fmt, ...)

@@ -87,6 +92,45 @@ static void write_file(char *filename, c
 	}
 }
 
+static int read_mnt_flags(const char *path)
+{
+	int ret;
+	struct statvfs stat;
+	int mnt_flags;
+
+	ret = statvfs(path, &stat);
+	if (ret != 0) {
+		die("statvfs of %s failed: %s\n",
+			path, strerror(errno));
+	}
+	if (stat.f_flag & ~(ST_RDONLY | ST_NOSUID | ST_NODEV | \
+			ST_NOEXEC | ST_NOATIME | ST_NODIRATIME | ST_RELATIME | \
+			ST_SYNCHRONOUS | ST_MANDLOCK)) {
+		die("Unrecognized mount flags\n");
+	}
+	mnt_flags = 0;
+	if (stat.f_flag & ST_RDONLY)
+		mnt_flags |= MS_RDONLY;
+	if (stat.f_flag & ST_NOSUID)
+		mnt_flags |= MS_NOSUID;
+	if (stat.f_flag & ST_NODEV)
+		mnt_flags |= MS_NODEV;
+	if (stat.f_flag & ST_NOEXEC)
+		mnt_flags |= MS_NOEXEC;
+	if (stat.f_flag & ST_NOATIME)
+		mnt_flags |= MS_NOATIME;
+	if (stat.f_flag & ST_NODIRATIME)
+		mnt_flags |= MS_NODIRATIME;
+	if (stat.f_flag & ST_RELATIME)
+		mnt_flags |= MS_RELATIME;
+	if (stat.f_flag & ST_SYNCHRONOUS)
+		mnt_flags |= MS_SYNCHRONOUS;
+	if (stat.f_flag & ST_MANDLOCK)
+		mnt_flags |= ST_MANDLOCK;
+
+	return mnt_flags;
+}
+
 static void create_and_enter_userns(void)
 {
 	uid_t uid;

@@ -118,7 +162,8 @@ static void create_and_enter_userns(void
 }
 
 static
-bool test_unpriv_remount(int mount_flags, int remount_flags, int invalid_flags)
+bool test_unpriv_remount(const char *fstype, const char *mount_options,
+			 int mount_flags, int remount_flags, int invalid_flags)
 {
 	pid_t child;
 

@@ -151,9 +196,11 @@ bool test_unpriv_remount(int mount_flags
 			strerror(errno));
 	}
 
-	if (mount("testing", "/tmp", "ramfs", mount_flags, NULL) != 0) {
-		die("mount of /tmp failed: %s\n",
-			strerror(errno));
+	if (mount("testing", "/tmp", fstype, mount_flags, mount_options) != 0) {
+		die("mount of %s with options '%s' on /tmp failed: %s\n",
+		    fstype,
+		    mount_options? mount_options : "",
+		    strerror(errno));
 	}
 
 	create_and_enter_userns();

@@ -181,62 +228,127 @@ bool test_unpriv_remount(int mount_flags
 
 static bool test_unpriv_remount_simple(int mount_flags)
 {
-	return test_unpriv_remount(mount_flags, mount_flags, 0);
+	return test_unpriv_remount("ramfs", NULL, mount_flags, mount_flags, 0);
 }
 
 static bool test_unpriv_remount_atime(int mount_flags, int invalid_flags)
 {
-	return test_unpriv_remount(mount_flags, mount_flags, invalid_flags);
+	return test_unpriv_remount("ramfs", NULL, mount_flags, mount_flags,
+				   invalid_flags);
+}
+
+static bool test_priv_mount_unpriv_remount(void)
+{
+	pid_t child;
+	int ret;
+	const char *orig_path = "/dev";
+	const char *dest_path = "/tmp";
+	int orig_mnt_flags, remount_mnt_flags;
+
+	child = fork();
+	if (child == -1) {
+		die("fork failed: %s\n",
+			strerror(errno));
+	}
+	if (child != 0) { /* parent */
+		pid_t pid;
+		int status;
+		pid = waitpid(child, &status, 0);
+		if (pid == -1) {
+			die("waitpid failed: %s\n",
+				strerror(errno));
+		}
+		if (pid != child) {
+			die("waited for %d got %d\n",
+				child, pid);
+		}
+		if (!WIFEXITED(status)) {
+			die("child did not terminate cleanly\n");
+		}
+		return WEXITSTATUS(status) == EXIT_SUCCESS ? true : false;
+	}
+
+	orig_mnt_flags = read_mnt_flags(orig_path);
+
+	create_and_enter_userns();
+	ret = unshare(CLONE_NEWNS);
+	if (ret != 0) {
+		die("unshare(CLONE_NEWNS) failed: %s\n",
+			strerror(errno));
+	}
+
+	ret = mount(orig_path, dest_path, "bind", MS_BIND | MS_REC, NULL);
+	if (ret != 0) {
+		die("recursive bind mount of %s onto %s failed: %s\n",
+			orig_path, dest_path, strerror(errno));
+	}
+
+	ret = mount(dest_path, dest_path, "none",
+		    MS_REMOUNT | MS_BIND | orig_mnt_flags , NULL);
+	if (ret != 0) {
+		/* system("cat /proc/self/mounts"); */
+		die("remount of /tmp failed: %s\n",
+		    strerror(errno));
+	}
+
+	remount_mnt_flags = read_mnt_flags(dest_path);
+	if (orig_mnt_flags != remount_mnt_flags) {
+		die("Mount flags unexpectedly changed during remount of %s originally mounted on %s\n",
+			dest_path, orig_path);
+	}
+	exit(EXIT_SUCCESS);
 }
 
 int main(int argc, char **argv)
 {
-	if (!test_unpriv_remount_simple(MS_RDONLY|MS_NODEV)) {
+	if (!test_unpriv_remount_simple(MS_RDONLY)) {
 		die("MS_RDONLY malfunctions\n");
 	}
-	if (!test_unpriv_remount_simple(MS_NODEV)) {
+	if (!test_unpriv_remount("devpts", "newinstance", MS_NODEV, MS_NODEV, 0)) {
 		die("MS_NODEV malfunctions\n");
 	}
-	if (!test_unpriv_remount_simple(MS_NOSUID|MS_NODEV)) {
+	if (!test_unpriv_remount_simple(MS_NOSUID)) {
 		die("MS_NOSUID malfunctions\n");
 	}
-	if (!test_unpriv_remount_simple(MS_NOEXEC|MS_NODEV)) {
+	if (!test_unpriv_remount_simple(MS_NOEXEC)) {
 		die("MS_NOEXEC malfunctions\n");
 	}
-	if (!test_unpriv_remount_atime(MS_RELATIME|MS_NODEV,
-				       MS_NOATIME|MS_NODEV))
+	if (!test_unpriv_remount_atime(MS_RELATIME,
+				       MS_NOATIME))
 	{
 		die("MS_RELATIME malfunctions\n");
 	}
-	if (!test_unpriv_remount_atime(MS_STRICTATIME|MS_NODEV,
-				       MS_NOATIME|MS_NODEV))
+	if (!test_unpriv_remount_atime(MS_STRICTATIME,
+				       MS_NOATIME))
 	{
 		die("MS_STRICTATIME malfunctions\n");
 	}
-	if (!test_unpriv_remount_atime(MS_NOATIME|MS_NODEV,
-				       MS_STRICTATIME|MS_NODEV))
+	if (!test_unpriv_remount_atime(MS_NOATIME,
+				       MS_STRICTATIME))
 	{
-		die("MS_RELATIME malfunctions\n");
+		die("MS_NOATIME malfunctions\n");
 	}
-	if (!test_unpriv_remount_atime(MS_RELATIME|MS_NODIRATIME|MS_NODEV,
-				       MS_NOATIME|MS_NODEV))
+	if (!test_unpriv_remount_atime(MS_RELATIME|MS_NODIRATIME,
+				       MS_NOATIME))
 	{
-		die("MS_RELATIME malfunctions\n");
+		die("MS_RELATIME|MS_NODIRATIME malfunctions\n");
 	}
-	if (!test_unpriv_remount_atime(MS_STRICTATIME|MS_NODIRATIME|MS_NODEV,
-				       MS_NOATIME|MS_NODEV))
+	if (!test_unpriv_remount_atime(MS_STRICTATIME|MS_NODIRATIME,
+				       MS_NOATIME))
 	{
-		die("MS_RELATIME malfunctions\n");
+		die("MS_STRICTATIME|MS_NODIRATIME malfunctions\n");
 	}
-	if (!test_unpriv_remount_atime(MS_NOATIME|MS_NODIRATIME|MS_NODEV,
-				       MS_STRICTATIME|MS_NODEV))
+	if (!test_unpriv_remount_atime(MS_NOATIME|MS_NODIRATIME,
+				       MS_STRICTATIME))
 	{
-		die("MS_RELATIME malfunctions\n");
+		die("MS_NOATIME|MS_DIRATIME malfunctions\n");
 	}
-	if (!test_unpriv_remount(MS_STRICTATIME|MS_NODEV, MS_NODEV,
-				 MS_NOATIME|MS_NODEV))
+	if (!test_unpriv_remount("ramfs", NULL, MS_STRICTATIME, 0, MS_NOATIME))
 	{
 		die("Default atime malfunctions\n");
 	}
+	if (!test_priv_mount_unpriv_remount()) {
+		die("Mount flags unexpectedly changed after remount\n");
+	}
 	return EXIT_SUCCESS;
 }

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: "Eric W. Biederman" <[email protected]>

commit b2f5d4dc38e034eecb7987e513255265ff9aa1cf upstream.

Forced unmount affects not just the mount namespace but the underlying superblock as well. Restrict forced unmount to the global root user for now. Otherwise it becomes possible a user in a less privileged mount namespace to force the shutdown of a superblock of a filesystem in a more privileged mount namespace, allowing a DOS attack on root.

---
 fs/namespace.c |    3 +++
 1 file changed, 3 insertions(+)

@@ -1342,6 +1342,9 @@ SYSCALL_DEFINE2(umount, char __user *, n
 		goto dput_and_out;
 	if (!check_mnt(mnt))
 		goto dput_and_out;
+	retval = -EPERM;
+	if (flags & MNT_FORCE && !capable(CAP_SYS_ADMIN))
+		goto dput_and_out;
 
 	retval = do_umount(mnt, flags);
 dput_and_out:

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: "Eric W. Biederman" <[email protected]>

commit 7ff4d90b4c24a03666f296c3d4878cd39001e81e upstream.

Today there are 3 instances of setgroups and due to an oversight their permission checking has diverged. Add a common function so that they may all share the same permission checking code.

This corrects the current oversight in the current permission checks and adds a helper to avoid this in the future.

A user namespace security fix will update this new helper, shortly.

---
 arch/s390/kernel/compat_linux.c |    2 +-
 include/linux/cred.h            |    1 +
 kernel/groups.c                 |    9 ++++++++-
 kernel/uid16.c                  |    2 +-
 4 files changed, 11 insertions(+), 3 deletions(-)

@@ -248,7 +248,7 @@ asmlinkage long sys32_setgroups16(int gi
 	struct group_info *group_info;
 	int retval;
 
-	if (!capable(CAP_SETGID))
+	if (!may_setgroups())
 		return -EPERM;
 	if ((unsigned)gidsetsize > NGROUPS_MAX)
 		return -EINVAL;

@@ -68,6 +68,7 @@ extern void groups_free(struct group_inf
 extern int set_current_groups(struct group_info *);
 extern int set_groups(struct cred *, struct group_info *);
 extern int groups_search(const struct group_info *, kgid_t);
+extern bool may_setgroups(void);
 
 /* access the groups "array" with this macro */
 #define GROUP_AT(gi, i) \

@@ -223,6 +223,13 @@ out:
 	return i;
 }
 
+bool may_setgroups(void)
+{
+	struct user_namespace *user_ns = current_user_ns();
+
+	return ns_capable(user_ns, CAP_SETGID);
+}
+
 /*
  *	SMP: Our groups are copy-on-write. We can set them safely
  *	without another task interfering.

@@ -233,7 +240,7 @@ SYSCALL_DEFINE2(setgroups, int, gidsetsi
 	struct group_info *group_info;
 	int retval;
 
-	if (!nsown_capable(CAP_SETGID))
+	if (!may_setgroups())
 		return -EPERM;
 	if ((unsigned)gidsetsize > NGROUPS_MAX)
 		return -EINVAL;

@@ -176,7 +176,7 @@ SYSCALL_DEFINE2(setgroups16, int, gidset
 	struct group_info *group_info;
 	int retval;
 
-	if (!nsown_capable(CAP_SETGID))
+	if (!may_setgroups())
 		return -EPERM;
 	if ((unsigned)gidsetsize > NGROUPS_MAX)
 		return -EINVAL;

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: "Eric W. Biederman" <[email protected]>

commit 0542f17bf2c1f2430d368f44c8fcf2f82ec9e53e upstream.

The rule is simple. Don't allow anything that wouldn't be allowed without unprivileged mappings.

It was previously overlooked that establishing gid mappings would allow dropping groups and potentially gaining permission to files and directories that had lesser permissions for a specific group than for all other users.

This is the rule needed to fix CVE-2014-8989 and prevent any other security issues with new_idmap_permitted.

The reason for this rule is that the unix permission model is old and there are programs out there somewhere that take advantage of every little corner of it. So allowing a uid or gid mapping to be established without privielge that would allow anything that would not be allowed without that mapping will result in expectations from some code somewhere being violated. Violated expectations about the behavior of the OS is a long way to say a security issue.

---
 kernel/user_namespace.c |    4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

@@ -800,7 +800,9 @@ static bool new_idmap_permitted(const st
 				struct user_namespace *ns, int cap_setid,
 				struct uid_gid_map *new_map)
 {
-	/* Allow mapping to your own filesystem ids */
+	/* Don't allow mappings that would allow anything that wouldn't
+	 * be allowed without the establishment of unprivileged mappings.
+	 */
 	if ((new_map->nr_extents == 1) && (new_map->extent[0].count == 1)) {
 		u32 id = new_map->extent[0].lower_first;
 		if (cap_setid == CAP_SETUID) {

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: "Eric W. Biederman" <[email protected]>

commit 273d2c67c3e179adb1e74f403d1e9a06e3f841b5 upstream.

setgroups is unique in not needing a valid mapping before it can be called, in the case of setgroups(0, NULL) which drops all supplemental groups.

The design of the user namespace assumes that CAP_SETGID can not actually be used until a gid mapping is established. Therefore add a helper function to see if the user namespace gid mapping has been established and call that function in the setgroups permission check.

This is part of the fix for CVE-2014-8989, being able to drop groups without privilege using user namespaces.

---
 include/linux/user_namespace.h |    5 +++++
 kernel/groups.c                |    4 +++-
 kernel/user_namespace.c        |   14 ++++++++++++++
 3 files changed, 22 insertions(+), 1 deletion(-)

@@ -59,6 +59,7 @@ extern struct seq_operations proc_projid
 extern ssize_t proc_uid_map_write(struct file *, const char __user *, size_t, loff_t *);
 extern ssize_t proc_gid_map_write(struct file *, const char __user *, size_t, loff_t *);
 extern ssize_t proc_projid_map_write(struct file *, const char __user *, size_t, loff_t *);
+extern bool userns_may_setgroups(const struct user_namespace *ns);
 #else
 
 static inline struct user_namespace *get_user_ns(struct user_namespace *ns)

@@ -83,6 +84,10 @@ static inline void put_user_ns(struct us
 {
 }
 
+static inline bool userns_may_setgroups(const struct user_namespace *ns)
+{
+	return true;
+}
 #endif
 
 void update_mnt_policy(struct user_namespace *userns);

@@ -6,6 +6,7 @@
 #include <linux/slab.h>
 #include <linux/security.h>
 #include <linux/syscalls.h>
+#include <linux/user_namespace.h>
 #include <asm/uaccess.h>
 
 /* init to 2 - one for init_task, one to ensure it is never freed */

@@ -227,7 +228,8 @@ bool may_setgroups(void)
 {
 	struct user_namespace *user_ns = current_user_ns();
 
-	return ns_capable(user_ns, CAP_SETGID);
+	return ns_capable(user_ns, CAP_SETGID) &&
+		userns_may_setgroups(user_ns);
 }
 
 /*

@@ -832,6 +832,20 @@ static bool new_idmap_permitted(const st
 	return false;
 }
 
+bool userns_may_setgroups(const struct user_namespace *ns)
+{
+	bool allowed;
+
+	mutex_lock(&id_map_mutex);
+	/* It is not safe to use setgroups until a gid mapping in
+	 * the user namespace has been established.
+	 */
+	allowed = ns->gid_map.nr_extents != 0;
+	mutex_unlock(&id_map_mutex);
+
+	return allowed;
+}
+
 static void *userns_get(struct task_struct *task)
 {
 	struct user_namespace *user_ns;

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: "Eric W. Biederman" <[email protected]>

commit be7c6dba2332cef0677fbabb606e279ae76652c3 upstream.

As any gid mapping will allow and must allow for backwards compatibility dropping groups don't allow any gid mappings to be established without CAP_SETGID in the parent user namespace.

For a small class of applications this change breaks userspace and removes useful functionality. This small class of applications includes tools/testing/selftests/mount/unprivilged-remount-test.c

Most of the removed functionality will be added back with the addition of a one way knob to disable setgroups. Once setgroups is disabled setting the gid_map becomes as safe as setting the uid_map.

For more common applications that set the uid_map and the gid_map with privilege this change will have no affect.

This is part of a fix for CVE-2014-8989.

---
 kernel/user_namespace.c |    5 -----
 1 file changed, 5 deletions(-)

@@ -810,11 +810,6 @@ static bool new_idmap_permitted(const st
 			if (uid_eq(uid, file->f_cred->fsuid))
 				return true;
 		}
-		else if (cap_setid == CAP_SETGID) {
-			kgid_t gid = make_kgid(ns->parent, id);
-			if (gid_eq(gid, file->f_cred->fsgid))
-				return true;
-		}
 	}
 
 	/* Allow anyone to set a mapping that doesn't require privilege */

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: "Eric W. Biederman" <[email protected]>

commit 80dd00a23784b384ccea049bfb3f259d3f973b9d upstream.

setresuid allows the euid to be set to any of uid, euid, suid, and fsuid. Therefor it is safe to allow an unprivileged user to map their euid and use CAP_SETUID privileged with exactly that uid, as no new credentials can be obtained.

I can not find a combination of existing system calls that allows setting uid, euid, suid, and fsuid from the fsuid making the previous use of fsuid for allowing unprivileged mappings a bug.

This is part of a fix for CVE-2014-8989.

---
 kernel/user_namespace.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -807,7 +807,7 @@ static bool new_idmap_permitted(const st
 		u32 id = new_map->extent[0].lower_first;
 		if (cap_setid == CAP_SETUID) {
 			kuid_t uid = make_kuid(ns->parent, id);
-			if (uid_eq(uid, file->f_cred->fsuid))
+			if (uid_eq(uid, file->f_cred->euid))
 				return true;
 		}
 	}

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: "Eric W. Biederman" <[email protected]>

commit f95d7918bd1e724675de4940039f2865e5eec5fe upstream.

If you did not create the user namespace and are allowed to write to uid_map or gid_map you should already have the necessary privilege in the parent user namespace to establish any mapping you want so this will not affect userspace in practice.

Limiting unprivileged uid mapping establishment to the creator of the user namespace makes it easier to verify all credentials obtained with the uid mapping can be obtained without the uid mapping without privilege.

Limiting unprivileged gid mapping establishment (which is temporarily absent) to the creator of the user namespace also ensures that the combination of uid and gid can already be obtained without privilege.

This is part of the fix for CVE-2014-8989.

---
 kernel/user_namespace.c |    6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

@@ -800,14 +800,16 @@ static bool new_idmap_permitted(const st
 				struct user_namespace *ns, int cap_setid,
 				struct uid_gid_map *new_map)
 {
+	const struct cred *cred = file->f_cred;
 	/* Don't allow mappings that would allow anything that wouldn't
 	 * be allowed without the establishment of unprivileged mappings.
 	 */
-	if ((new_map->nr_extents == 1) && (new_map->extent[0].count == 1)) {
+	if ((new_map->nr_extents == 1) && (new_map->extent[0].count == 1) &&
+	    uid_eq(ns->owner, cred->euid)) {
 		u32 id = new_map->extent[0].lower_first;
 		if (cap_setid == CAP_SETUID) {
 			kuid_t uid = make_kuid(ns->parent, id);
-			if (uid_eq(uid, file->f_cred->euid))
+			if (uid_eq(uid, cred->euid))
 				return true;
 		}
 	}

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: "Eric W. Biederman" <[email protected]>

commit f0d62aec931e4ae3333c797d346dc4f188f454ba upstream.

Generalize id_map_mutex so it can be used for more state of a user namespace.

---
 kernel/user_namespace.c |   14 ++++++--------
 1 file changed, 6 insertions(+), 8 deletions(-)

@@ -24,6 +24,7 @@
 #include <linux/fs_struct.h>
 
 static struct kmem_cache *user_ns_cachep __read_mostly;
+static DEFINE_MUTEX(userns_state_mutex);
 
 static bool new_idmap_permitted(const struct file *file,
 				struct user_namespace *ns, int cap_setid,

@@ -577,9 +578,6 @@ static bool mappings_overlap(struct uid_
 	return false;
 }
 
-
-static DEFINE_MUTEX(id_map_mutex);
-
 static ssize_t map_write(struct file *file, const char __user *buf,
 			 size_t count, loff_t *ppos,
 			 int cap_setid,

@@ -596,7 +594,7 @@ static ssize_t map_write(struct file *fi
 	ssize_t ret = -EINVAL;
 
 	/*
-	 * The id_map_mutex serializes all writes to any given map.
+	 * The userns_state_mutex serializes all writes to any given map.
 	 *
 	 * Any map is only ever written once.
 	 *

@@ -614,7 +612,7 @@ static ssize_t map_write(struct file *fi
 	 * order and smp_rmb() is guaranteed that we don't have crazy
 	 * architectures returning stale data.
 	 */
-	mutex_lock(&id_map_mutex);
+	mutex_lock(&userns_state_mutex);
 
 	ret = -EPERM;
 	/* Only allow one successful write to the map */

@@ -741,7 +739,7 @@ static ssize_t map_write(struct file *fi
 	*ppos = count;
 	ret = count;
 out:
-	mutex_unlock(&id_map_mutex);
+	mutex_unlock(&userns_state_mutex);
 	if (page)
 		free_page(page);
 	return ret;

@@ -833,12 +831,12 @@ bool userns_may_setgroups(const struct u
 {
 	bool allowed;
 
-	mutex_lock(&id_map_mutex);
+	mutex_lock(&userns_state_mutex);
 	/* It is not safe to use setgroups until a gid mapping in
 	 * the user namespace has been established.
 	 */
 	allowed = ns->gid_map.nr_extents != 0;
-	mutex_unlock(&id_map_mutex);
+	mutex_unlock(&userns_state_mutex);
 
 	return allowed;
 }

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: "Eric W. Biederman" <[email protected]>

commit 9cc46516ddf497ea16e8d7cb986ae03a0f6b92f8 upstream.

- Expose the knob to user space through a proc file /proc/<pid>/setgroups

A value of "deny" means the setgroups system call is disabled in the current processes user namespace and can not be enabled in the future in this user namespace.

A value of "allow" means the segtoups system call is enabled.

- Descendant user namespaces inherit the value of setgroups from their parents.

- A proc file is used (instead of a sysctl) as sysctls currently do not allow checking the permissions at open time.

- Writing to the proc file is restricted to before the gid_map for the user namespace is set.

This ensures that disabling setgroups at a user namespace level will never remove the ability to call setgroups from a process that already has that ability.

A process may opt in to the setgroups disable for itself by creating, entering and configuring a user namespace or by calling setns on an existing user namespace with setgroups disabled. Processes without privileges already can not call setgroups so this is a noop. Prodcess with privilege become processes without privilege when entering a user namespace and as with any other path to dropping privilege they would not have the ability to call setgroups. So this remains within the bounds of what is possible without a knob to disable setgroups permanently in a user namespace.

---
 fs/proc/base.c                 |   53 +++++++++++++++++++++++++
 include/linux/user_namespace.h |    7 +++
 kernel/user.c                  |    1 
 kernel/user_namespace.c        |   85 +++++++++++++++++++++++++++++++++++++++++
 4 files changed, 146 insertions(+)

@@ -2612,6 +2612,57 @@ static const struct file_operations proc
 	.llseek		= seq_lseek,
 	.release	= proc_id_map_release,
 };
+
+static int proc_setgroups_open(struct inode *inode, struct file *file)
+{
+	struct user_namespace *ns = NULL;
+	struct task_struct *task;
+	int ret;
+
+	ret = -ESRCH;
+	task = get_proc_task(inode);
+	if (task) {
+		rcu_read_lock();
+		ns = get_user_ns(task_cred_xxx(task, user_ns));
+		rcu_read_unlock();
+		put_task_struct(task);
+	}
+	if (!ns)
+		goto err;
+
+	if (file->f_mode & FMODE_WRITE) {
+		ret = -EACCES;
+		if (!ns_capable(ns, CAP_SYS_ADMIN))
+			goto err_put_ns;
+	}
+
+	ret = single_open(file, &proc_setgroups_show, ns);
+	if (ret)
+		goto err_put_ns;
+
+	return 0;
+err_put_ns:
+	put_user_ns(ns);
+err:
+	return ret;
+}
+
+static int proc_setgroups_release(struct inode *inode, struct file *file)
+{
+	struct seq_file *seq = file->private_data;
+	struct user_namespace *ns = seq->private;
+	int ret = single_release(inode, file);
+	put_user_ns(ns);
+	return ret;
+}
+
+static const struct file_operations proc_setgroups_operations = {
+	.open		= proc_setgroups_open,
+	.write		= proc_setgroups_write,
+	.read		= seq_read,
+	.llseek		= seq_lseek,
+	.release	= proc_setgroups_release,
+};
 #endif /* CONFIG_USER_NS */
 
 static int proc_pid_personality(struct seq_file *m, struct pid_namespace *ns,

@@ -2720,6 +2771,7 @@ static const struct pid_entry tgid_base_
 	REG("uid_map",    S_IRUGO|S_IWUSR, proc_uid_map_operations),
 	REG("gid_map",    S_IRUGO|S_IWUSR, proc_gid_map_operations),
 	REG("projid_map", S_IRUGO|S_IWUSR, proc_projid_map_operations),
+	REG("setgroups",  S_IRUGO|S_IWUSR, proc_setgroups_operations),
 #endif
 #ifdef CONFIG_CHECKPOINT_RESTORE
 	REG("timers",	  S_IRUGO, proc_timers_operations),

@@ -3073,6 +3125,7 @@ static const struct pid_entry tid_base_s
 	REG("uid_map",    S_IRUGO|S_IWUSR, proc_uid_map_operations),
 	REG("gid_map",    S_IRUGO|S_IWUSR, proc_gid_map_operations),
 	REG("projid_map", S_IRUGO|S_IWUSR, proc_projid_map_operations),
+	REG("setgroups",  S_IRUGO|S_IWUSR, proc_setgroups_operations),
 #endif
 };
 

@@ -17,6 +17,10 @@ struct uid_gid_map {	/* 64 bytes -- 1 ca
 	} extent[UID_GID_MAP_MAX_EXTENTS];
 };
 
+#define USERNS_SETGROUPS_ALLOWED 1UL
+
+#define USERNS_INIT_FLAGS USERNS_SETGROUPS_ALLOWED
+
 struct user_namespace {
 	struct uid_gid_map	uid_map;
 	struct uid_gid_map	gid_map;

@@ -27,6 +31,7 @@ struct user_namespace {
 	kuid_t			owner;
 	kgid_t			group;
 	unsigned int		proc_inum;
+	unsigned long		flags;
 	bool			may_mount_sysfs;
 	bool			may_mount_proc;
 };

@@ -59,6 +64,8 @@ extern struct seq_operations proc_projid
 extern ssize_t proc_uid_map_write(struct file *, const char __user *, size_t, loff_t *);
 extern ssize_t proc_gid_map_write(struct file *, const char __user *, size_t, loff_t *);
 extern ssize_t proc_projid_map_write(struct file *, const char __user *, size_t, loff_t *);
+extern ssize_t proc_setgroups_write(struct file *, const char __user *, size_t, loff_t *);
+extern int proc_setgroups_show(struct seq_file *m, void *v);
 extern bool userns_may_setgroups(const struct user_namespace *ns);
 #else
 

@@ -51,6 +51,7 @@ struct user_namespace init_user_ns = {
 	.owner = GLOBAL_ROOT_UID,
 	.group = GLOBAL_ROOT_GID,
 	.proc_inum = PROC_USER_INIT_INO,
+	.flags = USERNS_INIT_FLAGS,
 	.may_mount_sysfs = true,
 	.may_mount_proc = true,
 };

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: "Eric W. Biederman" <[email protected]>

commit 66d2f338ee4c449396b6f99f5e75cd18eb6df272 upstream.

Now that setgroups can be disabled and not reenabled, setting gid_map without privielge can now be enabled when setgroups is disabled.

This restores most of the functionality that was lost when unprivileged setting of gid_map was removed. Applications that use this functionality will need to check to see if they use setgroups or init_groups, and if they don't they can be fixed by simply disabling setgroups before writing to gid_map.

---
 kernel/user_namespace.c |    5 +++++
 1 file changed, 5 insertions(+)

@@ -814,6 +814,11 @@ static bool new_idmap_permitted(const st
 			kuid_t uid = make_kuid(ns->parent, id);
 			if (uid_eq(uid, cred->euid))
 				return true;
+		} else if (cap_setid == CAP_SETGID) {
+			kgid_t gid = make_kgid(ns->parent, id);
+			if (!(ns->flags & USERNS_SETGROUPS_ALLOWED) &&
+			    gid_eq(gid, cred->egid))
+				return true;
 		}
 	}
 

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: "Eric W. Biederman" <[email protected]>

commit db86da7cb76f797a1a8b445166a15cb922c6ff85 upstream.

A security fix in caused the way the unprivileged remount tests were using user namespaces to break. Tweak the way user namespaces are being used so the test works again.

---
 tools/testing/selftests/mount/unprivileged-remount-test.c |   32 ++++++++++----
 1 file changed, 24 insertions(+), 8 deletions(-)

@@ -53,17 +53,14 @@ static void die(char *fmt, ...)
 	exit(EXIT_FAILURE);
 }
 
-static void write_file(char *filename, char *fmt, ...)
+static void vmaybe_write_file(bool enoent_ok, char *filename, char *fmt, va_list ap)
 {
 	char buf[4096];
 	int fd;
 	ssize_t written;
 	int buf_len;
-	va_list ap;
 
-	va_start(ap, fmt);
 	buf_len = vsnprintf(buf, sizeof(buf), fmt, ap);
-	va_end(ap);
 	if (buf_len < 0) {
 		die("vsnprintf failed: %s\n",
 		    strerror(errno));

@@ -74,6 +71,8 @@ static void write_file(char *filename, c
 
 	fd = open(filename, O_WRONLY);
 	if (fd < 0) {
+		if ((errno == ENOENT) && enoent_ok)
+			return;
 		die("open of %s failed: %s\n",
 		    filename, strerror(errno));
 	}

@@ -92,6 +91,26 @@ static void write_file(char *filename, c
 	}
 }
 
+static void maybe_write_file(char *filename, char *fmt, ...)
+{
+	va_list ap;
+
+	va_start(ap, fmt);
+	vmaybe_write_file(true, filename, fmt, ap);
+	va_end(ap);
+
+}
+
+static void write_file(char *filename, char *fmt, ...)
+{
+	va_list ap;
+
+	va_start(ap, fmt);
+	vmaybe_write_file(false, filename, fmt, ap);
+	va_end(ap);
+
+}
+
 static int read_mnt_flags(const char *path)
 {
 	int ret;

@@ -144,13 +163,10 @@ static void create_and_enter_userns(void
 			strerror(errno));
 	}
 
+	maybe_write_file("/proc/self/setgroups", "deny");
 	write_file("/proc/self/uid_map", "0 %d 1", uid);
 	write_file("/proc/self/gid_map", "0 %d 1", gid);
 
-	if (setgroups(0, NULL) != 0) {
-		die("setgroups failed: %s\n",
-			strerror(errno));
-	}
 	if (setgid(0) != 0) {
 		die ("setgid(0) failed %s\n",
 			strerror(errno));

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: Rabin Vincent <[email protected]>

commit 7e77bdebff5cb1e9876c561f69710b9ab8fa1f7e upstream.

If a request is backlogged, it's complete() handler will get called
twice: once with -EINPROGRESS, and once with the final error code.

af_alg's complete handler, unlike other users, does not handle the -EINPROGRESS but instead always completes the completion that recvmsg() is waiting on. This can lead to a return to user space while the request is still pending in the driver. If userspace closes the sockets before the requests are handled by the driver, this will lead to use-after-frees (and potential crashes) in the kernel due to the tfm having been freed.

The crashes can be easily reproduced (for example) by reducing the max queue length in cryptod.c and running the following (from
http://www.chronox.de/libkcapi.html) on AES-NI capable hardware:

 $ while true; do kcapi -x 1 -e -c '__ecb-aes-aesni' \
    -k 00000000000000000000000000000000 \
    -p 00000000000000000000000000000000 >/dev/null & done

---
 crypto/af_alg.c |    3 +++
 1 file changed, 3 insertions(+)

@@ -449,6 +449,9 @@ void af_alg_complete(struct crypto_async
 {
 	struct af_alg_completion *completion = req->data;
 
+	if (err == -EINPROGRESS)
+		return;
+
 	completion->err = err;
 	complete(&completion->completion);
 }

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: Jan Kara <[email protected]>

commit a682e9c28cac152e6e54c39efcf046e0c8cfcf63 upstream.

If some error happens in NCP_IOC_SETROOT ioctl, the appropriate error return value is then (in most cases) just overwritten before we return. This can result in reporting success to userspace although error happened.

This bug was introduced by commit 2e54eb96e2c8 ("BKL: Remove BKL from ncpfs"). Propagate the errors correctly.

Coverity id: 1226925.

---
 fs/ncpfs/ioctl.c |    1 -
 1 file changed, 1 deletion(-)

@@ -448,7 +448,6 @@ static long __ncp_ioctl(struct inode *in
 						result = -EIO;
 					}
 				}
-				result = 0;
 			}
 			mutex_unlock(&server->root_setup_lock);
 

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: Oleg Nesterov <[email protected]>

commit 24c037ebf5723d4d9ab0996433cee4f96c292a4d upstream.

alloc_pid() does get_pid_ns() beforehand but forgets to put_pid_ns() if it fails because disable_pid_allocation() was called by the exiting child_reaper.

We could simply move get_pid_ns() down to successful return, but this fix tries to be as trivial as possible.

---
 kernel/pid.c |    2 ++
 1 file changed, 2 insertions(+)

@@ -335,6 +335,8 @@ out:
 
 out_unlock:
 	spin_unlock_irq(&pidmap_lock);
+	put_pid_ns(ns);
+
 out_free:
 	while (++i <= ns->level)
 		free_pidmap(pid->numbers + i);

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: Jan Kara <[email protected]>

commit a1d47b262952a45aae62bd49cfaf33dd76c11a2c upstream.

UDF specification allows arbitrarily large symlinks. However we support only symlinks at most one block large. Check the length of the symlink so that we don't access memory beyond end of the symlink block.

---
 fs/udf/symlink.c |   17 +++++++++++++----
 1 file changed, 13 insertions(+), 4 deletions(-)

@@ -80,11 +80,17 @@ static int udf_symlink_filler(struct fil
 	struct inode *inode = page->mapping->host;
 	struct buffer_head *bh = NULL;
 	unsigned char *symlink;
-	int err = -EIO;
+	int err;
 	unsigned char *p = kmap(page);
 	struct udf_inode_info *iinfo;
 	uint32_t pos;
 
+	/* We don't support symlinks longer than one block */
+	if (inode->i_size > inode->i_sb->s_blocksize) {
+		err = -ENAMETOOLONG;
+		goto out_unmap;
+	}
+
 	iinfo = UDF_I(inode);
 	pos = udf_block_map(inode, 0);
 

@@ -94,8 +100,10 @@ static int udf_symlink_filler(struct fil
 	} else {
 		bh = sb_bread(inode->i_sb, pos);
 
-		if (!bh)
-			goto out;
+		if (!bh) {
+			err = -EIO;
+			goto out_unlock_inode;
+		}
 
 		symlink = bh->b_data;
 	}

@@ -109,9 +117,10 @@ static int udf_symlink_filler(struct fil
 	unlock_page(page);
 	return 0;
 
-out:
+out_unlock_inode:
 	up_read(&iinfo->i_data_sem);
 	SetPageError(page);
+out_unmap:
 	kunmap(page);
 	unlock_page(page);
 	return err;

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: Tyler Hicks <[email protected]>

commit 332b122d39c9cbff8b799007a825d94b2e7c12f2 upstream.

The ecryptfs_encrypted_view mount option greatly changes the functionality of an eCryptfs mount. Instead of encrypting and decrypting lower files, it provides a unified view of the encrypted files in the lower filesystem. The presence of the ecryptfs_encrypted_view mount option is intended to force a read-only mount and modifying files is not supported when the feature is in use. See the following commit for more information:

e77a56d [PATCH] eCryptfs: Encrypted passthrough

This patch forces the mount to be read-only when the ecryptfs_encrypted_view mount option is specified by setting the MS_RDONLY flag on the superblock. Additionally, this patch removes some broken logic in ecryptfs_open() that attempted to prevent modifications of files when the encrypted view feature was in use. The check in ecryptfs_open() was not sufficient to prevent file modifications using system calls that do not operate on a file descriptor.

---
 fs/ecryptfs/file.c |   12 ------------
 fs/ecryptfs/main.c |   16 +++++++++++++---
 2 files changed, 13 insertions(+), 15 deletions(-)

@@ -196,23 +196,11 @@ static int ecryptfs_open(struct inode *i
 {
 	int rc = 0;
 	struct ecryptfs_crypt_stat *crypt_stat = NULL;
-	struct ecryptfs_mount_crypt_stat *mount_crypt_stat;
 	struct dentry *ecryptfs_dentry = file->f_path.dentry;
 	/* Private value of ecryptfs_dentry allocated in
 	 * ecryptfs_lookup() */
 	struct ecryptfs_file_info *file_info;
 
-	mount_crypt_stat = &ecryptfs_superblock_to_private(
-		ecryptfs_dentry->d_sb)->mount_crypt_stat;
-	if ((mount_crypt_stat->flags & ECRYPTFS_ENCRYPTED_VIEW_ENABLED)
-	    && ((file->f_flags & O_WRONLY) || (file->f_flags & O_RDWR)
-		|| (file->f_flags & O_CREAT) || (file->f_flags & O_TRUNC)
-		|| (file->f_flags & O_APPEND))) {
-		printk(KERN_WARNING "Mount has encrypted view enabled; "
-		       "files may only be read\n");
-		rc = -EPERM;
-		goto out;
-	}
 	/* Released in ecryptfs_release or end of function if failure */
 	file_info = kmem_cache_zalloc(ecryptfs_file_info_cache, GFP_KERNEL);
 	ecryptfs_set_file_private(file, file_info);

@@ -494,6 +494,7 @@ static struct dentry *ecryptfs_mount(str
 {
 	struct super_block *s;
 	struct ecryptfs_sb_info *sbi;
+	struct ecryptfs_mount_crypt_stat *mount_crypt_stat;
 	struct ecryptfs_dentry_info *root_info;
 	const char *err = "Getting sb failed";
 	struct inode *inode;

@@ -512,6 +513,7 @@ static struct dentry *ecryptfs_mount(str
 		err = "Error parsing options";
 		goto out;
 	}
+	mount_crypt_stat = &sbi->mount_crypt_stat;
 
 	s = sget(fs_type, NULL, set_anon_super, flags, NULL);
 	if (IS_ERR(s)) {

@@ -558,11 +560,19 @@ static struct dentry *ecryptfs_mount(str
 
 	/**
 	 * Set the POSIX ACL flag based on whether they're enabled in the lower
-	 * mount. Force a read-only eCryptfs mount if the lower mount is ro.
-	 * Allow a ro eCryptfs mount even when the lower mount is rw.
+	 * mount.
 	 */
 	s->s_flags = flags & ~MS_POSIXACL;
-	s->s_flags |= path.dentry->d_sb->s_flags & (MS_RDONLY | MS_POSIXACL);
+	s->s_flags |= path.dentry->d_sb->s_flags & MS_POSIXACL;
+
+	/**
+	 * Force a read-only eCryptfs mount when:
+	 *   1) The lower mount is ro
+	 *   2) The ecryptfs_encrypted_view mount option is specified
+	 */
+	if (path.dentry->d_sb->s_flags & MS_RDONLY ||
+	    mount_crypt_stat->flags & ECRYPTFS_ENCRYPTED_VIEW_ENABLED)
+		s->s_flags |= MS_RDONLY;
 
 	s->s_maxbytes = path.dentry->d_sb->s_maxbytes;
 	s->s_blocksize = path.dentry->d_sb->s_blocksize;

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: Michael Halcrow <[email protected]>

commit 942080643bce061c3dd9d5718d3b745dcb39a8bc upstream.

Dmitry Chernenkov used KASAN to discover that eCryptfs writes past the end of the allocated buffer during encrypted filename decoding. This fix corrects the issue by getting rid of the unnecessary 0 write when the current bit offset is 2.

---
 fs/ecryptfs/crypto.c |    1 -
 1 file changed, 1 deletion(-)

@@ -2102,7 +2102,6 @@ ecryptfs_decode_from_filename(unsigned c
 			break;
 		case 2:
 			dst[dst_byte_offset++] |= (src_byte);
-			dst[dst_byte_offset] = 0;
 			current_bit_offset = 0;
 			break;
 		}

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: Josef Bacik <[email protected]>

commit a28046956c71985046474283fa3bcd256915fb72 upstream.

We use the modified list to keep track of which extents have been modified so we know which ones are candidates for logging at fsync() time. Newly modified extents are added to the list at modification time, around the same time the ordered extent is created. We do this so that we don't have to wait for ordered extents to complete before we know what we need to log. The problem is when something like this happens

log extent 0-4k on inode 1
copy csum for 0-4k from ordered extent into log
sync log
commit transaction
log some other extent on inode 1
ordered extent for 0-4k completes and adds itself onto modified list again log changed extents
see ordered extent for 0-4k has already been logged
at this point we assume the csum has been copied
sync log
crash

On replay we will see the extent 0-4k in the log, drop the original 0-4k extent which is the same one that we are replaying which also drops the csum, and then we won't find the csum in the log for that bytenr. This of course causes us to have errors about not having csums for certain ranges of our inode. So remove the modified list manipulation in unpin_extent_cache, any modified extents should have been added well before now, and we don't want them re-logged. This fixes my test that I could reliably reproduce this problem with. Thanks,

---
 fs/btrfs/extent_map.c |    2 --
 1 file changed, 2 deletions(-)

@@ -263,8 +263,6 @@ int unpin_extent_cache(struct extent_map
 	if (!em)
 		goto out;
 
-	if (!test_bit(EXTENT_FLAG_LOGGING, &em->flags))
-		list_move(&em->list, &tree->modified_extents);
 	em->generation = gen;
 	clear_bit(EXTENT_FLAG_PINNED, &em->flags);
 	em->mod_start = em->start;

3.10-stable review patch. If anyone has any objections, please let me know.

------------------

From: Filipe Manana <[email protected]>

commit 678886bdc6378c1cbd5072da2c5a3035000214e3 upstream.

When we abort a transaction we iterate over all the ranges marked as dirty in fs_info->freed_extents[0] and fs_info->freed_extents[1], clear them from those trees, add them back (unpin) to the free space caches and, if the fs was mounted with "-o discard", perform a discard on those regions. Also, after adding the regions to the free space caches, a fitrim ioctl call can see those ranges in a block group's free space cache and perform a discard on the ranges, so the same issue can happen without "-o discard" as well.

This causes corruption, affecting one or multiple btree nodes (in the worst case leaving the fs unmountable) because some of those ranges (the ones in the fs_info->pinned_extents tree) correspond to btree nodes/leafs that are referred by the last committed super block - breaking the rule that anything that was committed by a transaction is untouched until the next transaction commits successfully.

I ran into this while running in a loop (for several hours) the fstest that I recently submitted:

[PATCH] fstests: add btrfs test to stress chunk allocation/removal and fstrim

The corruption always happened when a transaction aborted and then fsck complained like this:

_check_btrfs_filesystem: filesystem on /dev/sdc is inconsistent *** fsck.btrfs output ***
Check tree block failed, want=94945280, have=0
Check tree block failed, want=94945280, have=0
Check tree block failed, want=94945280, have=0
Check tree block failed, want=94945280, have=0
Check tree block failed, want=94945280, have=0
read block failed check_tree_block
Couldn't open file system

In this case 94945280 corresponded to the root of a tree. Using frace what I observed was the following sequence of steps happened:

1) transaction N started, fs_info->pinned_extents pointed to

      fs_info->freed_extents[0];

2) node/eb 94945280 is created;

3) eb is persisted to disk;

4) transaction N commit starts, fs_info->pinned_extents now points to

      fs_info->freed_extents[1], and transaction N completes;

5) transaction N + 1 starts;

6) eb is COWed, and btrfs_free_tree_block() called for this eb;

7) eb range (94945280 to 94945280 + 16Kb) is added to

      fs_info->pinned_extents (fs_info->freed_extents[1]);

8) Something goes wrong in transaction N + 1, like hitting ENOSPC

      for example, and the transaction is aborted, turning the fs into
      readonly mode. The stack trace I got for example:

      [112065.253935]  [<ffffffff8140c7b6>] dump_stack+0x4d/0x66
      [112065.254271]  [<ffffffff81042984>] warn_slowpath_common+0x7f/0x98
      [112065.254567]  [<ffffffffa0325990>] ? __btrfs_abort_transaction+0x50/0x10b [btrfs]
      [112065.261674]  [<ffffffff810429e5>] warn_slowpath_fmt+0x48/0x50
      [112065.261922]  [<ffffffffa032949e>] ? btrfs_free_path+0x26/0x29 [btrfs]
      [112065.262211]  [<ffffffffa0325990>] __btrfs_abort_transaction+0x50/0x10b [btrfs]
      [112065.262545]  [<ffffffffa036b1d6>] btrfs_remove_chunk+0x537/0x58b [btrfs]
      [112065.262771]  [<ffffffffa033840f>] btrfs_delete_unused_bgs+0x1de/0x21b [btrfs]
      [112065.263105]  [<ffffffffa0343106>] cleaner_kthread+0x100/0x12f [btrfs]
      (...)
      [112065.264493] ---[ end trace dd7903a975a31a08 ]---
      [112065.264673] BTRFS: error (device sdc) in btrfs_remove_chunk:2625: errno=-28 No space left
      [112065.264997] BTRFS info (device sdc): forced readonly

9) The clear kthread sees that the BTRFS_FS_STATE_ERROR bit is set in

      fs_info->fs_state and calls btrfs_cleanup_transaction(), which in
      turn calls btrfs_destroy_pinned_extent();

10) Then btrfs_destroy_pinned_extent() iterates over all the ranges marked as dirty in fs_info->freed_extents[], and for each one it calls discard, if the fs was mounted with "-o discard", and adds the range to the free space cache of the respective block group;

11) btrfs_trim_block_group(), invoked from the fitrim ioctl code path,

       sees the free space entries and performs a discard;

12) After an umount and mount (or fsck), our eb's location on disk was full of zeroes, and it should have been untouched, because it was marked as dirty in the fs_info->pinned_extents tree, and therefore used by the trees that the last committed superblock points to.

Fix this by not performing a discard and not adding the ranges to the free space caches - it's useless from this point since the fs is now in readonly mode and we won't write free space caches to disk anymore (otherwise we would leak space) nor any new superblock. By not adding the ranges to the free space caches, it prevents other code paths from allocating that space and write to it as well, therefore being safer and simpler.

This isn't a new problem, as it's been present since 2011 (git commit acce952b0263825da32cf10489413dec78053347).

---
 fs/btrfs/disk-io.c     |    6 ------
 fs/btrfs/extent-tree.c |   10 ++++++----
 2 files changed, 6 insertions(+), 10 deletions(-)

@@ -3857,12 +3857,6 @@ again:
 		if (ret)
 			break;
 
-		/* opt_discard */
-		if (btrfs_test_opt(root, DISCARD))
-			ret = btrfs_error_discard_extent(root, start,
-							 end + 1 - start,
-							 NULL);
-
 		clear_extent_dirty(unpin, start, end, GFP_NOFS);
 		btrfs_error_unpin_extent_range(root, start, end);
 		cond_resched();

@@ -5277,7 +5277,8 @@ void btrfs_prepare_extent_commit(struct
 	update_global_block_rsv(fs_info);
 }
 
-static int unpin_extent_range(struct btrfs_root *root, u64 start, u64 end)
+static int unpin_extent_range(struct btrfs_root *root, u64 start, u64 end,
+			      const bool return_free_space)
 {
 	struct btrfs_fs_info *fs_info = root->fs_info;
 	struct btrfs_block_group_cache *cache = NULL;

@@ -5301,7 +5302,8 @@ static int unpin_extent_range(struct btr
 
 		if (start < cache->last_byte_to_unpin) {
 			len = min(len, cache->last_byte_to_unpin - start);
-			btrfs_add_free_space(cache, start, len);
+			if (return_free_space)
+				btrfs_add_free_space(cache, start, len);
 		}
 
 		start += len;

@@ -5364,7 +5366,7 @@ int btrfs_finish_extent_commit(struct bt
 						   end + 1 - start, NULL);
 
 		clear_extent_dirty(unpin, start, end, GFP_NOFS);
-		unpin_extent_range(root, start, end);
+		unpin_extent_range(root, start, end, true);
 		cond_resched();
 	}
 

@@ -8564,7 +8566,7 @@ out:
 
 int btrfs_error_unpin_extent_range(struct btrfs_root *root, u64 start, u64 end)
 {
-	return unpin_extent_range(root, start, end);
+	return unpin_extent_range(root, start, end, false);
 }
 
 int btrfs_error_discard_extent(struct btrfs_root *root, u64 bytenr,

Build results:

	total: 137 pass: 137 fail: 0
Qemu tests:
	total: 27 pass: 27 fail: 0

Details are available at http://server.roeck-us.net:8010/builders.

Guenter

Compiled and booted on my test system. No dmesg regressions.

thanks,
-- Shuah

Shuah Khan
Sr. Linux Kernel Developer
Open Source Innovation Group
Samsung Research America (Silicon Valley)
[email protected] | (970) 217-8978