4.14.149-stable review

This is the start of the stable review cycle for the 4.14.149 release. There are 61 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 Sat 12 Oct 2019 08:29:51 AM UTC. Anything received after that time might be too late.

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

	https://www.kernel.org/pub/linux/kernel/v4.x/stable-review/patch-4.14.149-rc1.gz

or in the git tree and branch at:
git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git linux-4.14.y
and the diffstat can be found below.

thanks,

greg k-h

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

Greg Kroah-Hartman <[email protected]>

    Linux 4.14.149-rc1

Oleksandr Suvorov <[email protected]>
ASoC: sgtl5000: Improve VAG power and mute control

Johannes Berg <[email protected]>
nl80211: validate beacon head

Jouni Malinen <[email protected]>
cfg80211: Use const more consistently in for_each_element macros

Johannes Berg <[email protected]>
cfg80211: add and use strongly typed element iteration macros

Andrew Murray <[email protected]>
coresight: etm4x: Use explicit barriers on enable/disable

Eric Sandeen <[email protected]>
vfs: Fix EOVERFLOW testing in put_compat_statfs64

Chris Wilson <[email protected]>
drm/i915/userptr: Acquire the page lock around set_page_dirty()

Srikar Dronamraju <[email protected]>
perf stat: Reset previous counts on repeat with interval

Srikar Dronamraju <[email protected]>
perf stat: Fix a segmentation fault when using repeat forever

Jiri Olsa <[email protected]>
perf tools: Fix segfault in cpu_cache_level__read()

Balasubramani Vivekanandan <[email protected]>
tick: broadcast-hrtimer: Fix a race in bc_set_next

Steven Rostedt (VMware) <[email protected]>

    tools lib traceevent: Do not free tep->cmdlines in add_new_comm() on failure

Gautham R. Shenoy <[email protected]>
powerpc/pseries: Fix cpu_hotplug_lock acquisition in resize_hpt()

Mike Christie <[email protected]>
nbd: fix max number of supported devs

Dan Melnic <[email protected]>
block/ndb: add WQ_UNBOUND to the knbd-recv workqueue

Xiubo Li <[email protected]>
nbd: fix crash when the blksize is zero

Cédric Le Goater <[email protected]>
KVM: PPC: Book3S HV: XIVE: Free escalation interrupts before disabling the VP

Arnaldo Carvalho de Melo <[email protected]>
perf unwind: Fix libunwind build failure on i386 systems

Valdis Kletnieks <[email protected]>
kernel/elfcore.c: include proper prototypes

Thomas Richter <[email protected]>
perf build: Add detection of java-11-openjdk-devel package

KeMeng Shi <[email protected]>
sched/core: Fix migration to invalid CPU in __set_cpus_allowed_ptr()

zhengbin <[email protected]>
fuse: fix memleak in cuse_channel_open

Ido Schimmel <[email protected]>
thermal: Fix use-after-free when unregistering thermal zone device

Fabrice Gasnier <[email protected]>
pwm: stm32-lp: Add check in case requested period cannot be achieved

Trond Myklebust <[email protected]>
pNFS: Ensure we do clear the return-on-close layout stateid on fatal errors

Trek <[email protected]>
drm/amdgpu: Check for valid number of registers to read

Florian Westphal <[email protected]>
netfilter: nf_tables: allow lookups in dynamic sets

Ryan Chen <[email protected]>
watchdog: aspeed: Add support for AST2600

Erqi Chen <[email protected]>
ceph: reconnect connection if session hang in opening state

Luis Henriques <[email protected]>
ceph: fix directories inode i_blkbits initialization

Igor Druzhinin <[email protected]>
xen/pci: reserve MCFG areas earlier

Chengguang Xu <[email protected]>
9p: avoid attaching writeback_fid on mmap with type PRIVATE

Jia-Ju Bai <[email protected]>
fs: nfs: Fix possible null-pointer dereferences in encode_attrs()

Sascha Hauer <[email protected]>
ima: always return negative code for error

Johannes Berg <[email protected]>
cfg80211: initialize on-stack chandefs

Johan Hovold <[email protected]>
ieee802154: atusb: fix use-after-free at disconnect

Juergen Gross <[email protected]>
xen/xenbus: fix self-deadlock after killing user process

Wanpeng Li <[email protected]>
Revert "locking/pvqspinlock: Don't wait if vCPU is preempted"

Russell King <[email protected]>
mmc: sdhci-of-esdhc: set DMA snooping based on DMA coherence

Russell King <[email protected]>
mmc: sdhci: improve ADMA error reporting

Tomi Valkeinen <[email protected]>
drm/omap: fix max fclk divider for omap36xx

Rasmus Villemoes <[email protected]>
watchdog: imx2_wdt: fix min() calculation in imx2_wdt_set_timeout

Li RongQing <[email protected]>
timer: Read jiffies once when forwarding base clk

Kees Cook <[email protected]>
usercopy: Avoid HIGHMEM pfn warning

Jiaxun Yang <[email protected]>
MIPS: Treat Loongson Extensions as ASEs

Horia Geantă <[email protected]>
crypto: caam - fix concurrency issue in givencrypt descriptor

Wei Yongjun <[email protected]>
crypto: cavium/zip - Add missing single_release()

Herbert Xu <[email protected]>
crypto: skcipher - Unmap pages after an external error

Alexander Sverdlin <[email protected]>
crypto: qat - Silence smp_processor_id() warning

Steven Rostedt (VMware) <[email protected]>
tools lib traceevent: Fix "robust" test of do_generate_dynamic_list_file

Marc Kleine-Budde <[email protected]>
can: mcp251x: mcp251x_hw_reset(): allow more time after a reset

Andrew Donnellan <[email protected]>
powerpc/powernv: Restrict OPAL symbol map to only be readable by root

Oleksandr Suvorov <[email protected]>
ASoC: Define a set of DAPM pre/post-up events

Dmitry Osipenko <[email protected]>
PM / devfreq: tegra: Fix kHz to Hz conversion

Jack Wang <[email protected]>
KVM: nVMX: handle page fault in vmread fix

Paul Mackerras <[email protected]>
KVM: PPC: Book3S HV: Don't lose pending doorbell request on migration on P9

Vasily Gorbik <[email protected]>
s390/cio: exclude subchannels with no parent from pseudo check

Vasily Gorbik <[email protected]>
s390/cio: avoid calling strlen on null pointer

Vasily Gorbik <[email protected]>
s390/topology: avoid firing events before kobjs are created

Thomas Huth <[email protected]>
KVM: s390: Test for bad access register and size at the start of S390_MEM_OP

Vasily Gorbik <[email protected]>
s390/process: avoid potential reading of freed stack

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

Diffstat:

 Makefile                                       |   4 +-
 arch/mips/include/asm/cpu-features.h           |  16 ++
 arch/mips/include/asm/cpu.h                    |   4 +
 arch/mips/kernel/cpu-probe.c                   |   6 +
 arch/mips/kernel/proc.c                        |   4 +
 arch/powerpc/kvm/book3s_hv.c                   |   9 +-
 arch/powerpc/kvm/book3s_xive.c                 |  18 +-
 arch/powerpc/mm/hash_utils_64.c                |   9 +-
 arch/powerpc/platforms/powernv/opal.c          |  11 +-
 arch/powerpc/platforms/pseries/lpar.c          |   8 +-
 arch/s390/kernel/process.c                     |  22 ++-
 arch/s390/kernel/topology.c                    |   3 +-
 arch/s390/kvm/kvm-s390.c                       |   2 +-
 arch/x86/kvm/vmx.c                             |   2 +-
 crypto/skcipher.c                              |  42 +++--
 drivers/block/nbd.c                            |  61 +++++--
 drivers/crypto/caam/caamalg_desc.c             |   9 +
 drivers/crypto/caam/caamalg_desc.h             |   2 +-
 drivers/crypto/cavium/zip/zip_main.c           |   3 +
 drivers/crypto/qat/qat_common/adf_common_drv.h |   2 +-
 drivers/devfreq/tegra-devfreq.c                |  12 +-
 drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c        |   3 +
 drivers/gpu/drm/i915/i915_gem_userptr.c        |  10 +-
 drivers/gpu/drm/omapdrm/dss/dss.c              |   2 +-
 drivers/hwtracing/coresight/coresight-etm4x.c  |  14 +-
 drivers/mmc/host/sdhci-of-esdhc.c              |   7 +-
 drivers/mmc/host/sdhci.c                       |  15 +-
 drivers/net/can/spi/mcp251x.c                  |  19 +-
 drivers/net/ieee802154/atusb.c                 |   3 +-
 drivers/pwm/pwm-stm32-lp.c                     |   6 +
 drivers/s390/cio/ccwgroup.c                    |   2 +-
 drivers/s390/cio/css.c                         |   2 +
 drivers/thermal/thermal_core.c                 |   2 +-
 drivers/watchdog/aspeed_wdt.c                  |   4 +-
 drivers/watchdog/imx2_wdt.c                    |   4 +-
 drivers/xen/pci.c                              |  21 ++-
 drivers/xen/xenbus/xenbus_dev_frontend.c       |  20 ++-
 fs/9p/vfs_file.c                               |   3 +
 fs/ceph/inode.c                                |   7 +-
 fs/ceph/mds_client.c                           |   4 +-
 fs/fuse/cuse.c                                 |   1 +
 fs/nfs/nfs4xdr.c                               |   2 +-
 fs/nfs/pnfs.c                                  |   9 +-
 fs/statfs.c                                    |  17 +-
 include/linux/ieee80211.h                      |  53 ++++++
 include/sound/soc-dapm.h                       |   2 +
 kernel/elfcore.c                               |   1 +
 kernel/locking/qspinlock_paravirt.h            |   2 +-
 kernel/sched/core.c                            |   4 +-
 kernel/time/tick-broadcast-hrtimer.c           |  57 +++---
 kernel/time/timer.c                            |   8 +-
 mm/usercopy.c                                  |   8 +-
 net/netfilter/nf_tables_api.c                  |   7 +-
 net/netfilter/nft_lookup.c                     |   3 -
 net/wireless/nl80211.c                         |  42 ++++-
 net/wireless/reg.c                             |   2 +-
 net/wireless/scan.c                            |  14 +-
 net/wireless/wext-compat.c                     |   2 +-
 security/integrity/ima/ima_crypto.c            |   5 +-
 sound/soc/codecs/sgtl5000.c                    | 232 +++++++++++++++++++++----
 tools/lib/traceevent/Makefile                  |   4 +-
 tools/lib/traceevent/event-parse.c             |   3 +-
 tools/perf/Makefile.config                     |   2 +-
 tools/perf/arch/x86/util/unwind-libunwind.c    |   2 +-
 tools/perf/builtin-stat.c                      |   5 +-
 tools/perf/util/header.c                       |   2 +-
 tools/perf/util/stat.c                         |  17 ++
 tools/perf/util/stat.h                         |   1 +
 68 files changed, 696 insertions(+), 208 deletions(-)

From: Vasily Gorbik <[email protected]>

commit 8769f610fe6d473e5e8e221709c3ac402037da6c upstream.

With THREAD_INFO_IN_TASK (which is selected on s390) task's stack usage is refcounted and should always be protected by get/put when touching other task's stack to avoid race conditions with task's destruction code.

---
 arch/s390/kernel/process.c |   22 ++++++++++++++++------
 1 file changed, 16 insertions(+), 6 deletions(-)

@@ -185,20 +185,30 @@ unsigned long get_wchan(struct task_stru
 
 	if (!p || p == current || p->state == TASK_RUNNING || !task_stack_page(p))
 		return 0;
+
+	if (!try_get_task_stack(p))
+		return 0;
+
 	low = task_stack_page(p);
 	high = (struct stack_frame *) task_pt_regs(p);
 	sf = (struct stack_frame *) p->thread.ksp;
-	if (sf <= low || sf > high)
-		return 0;
+	if (sf <= low || sf > high) {
+		return_address = 0;
+		goto out;
+	}
 	for (count = 0; count < 16; count++) {
 		sf = (struct stack_frame *) sf->back_chain;
-		if (sf <= low || sf > high)
-			return 0;
+		if (sf <= low || sf > high) {
+			return_address = 0;
+			goto out;
+		}
 		return_address = sf->gprs[8];
 		if (!in_sched_functions(return_address))
-			return return_address;
+			goto out;
 	}
-	return 0;
+out:
+	put_task_stack(p);
+	return return_address;
 }
 
 unsigned long arch_align_stack(unsigned long sp)

From: Thomas Huth <[email protected]>

commit a13b03bbb4575b350b46090af4dfd30e735aaed1 upstream.

If the KVM_S390_MEM_OP ioctl is called with an access register >= 16, then there is certainly a bug in the calling userspace application. We check for wrong access registers, but only if the vCPU was already in the access register mode before (i.e. the SIE block has recorded it). The check is also buried somewhere deep in the calling chain (in the function ar_translation()), so this is somewhat hard to find.

It's better to always report an error to the userspace in case this field is set wrong, and it's safer in the KVM code if we block wrong values here early instead of relying on a check somewhere deep down the calling chain, so let's add another check to kvm_s390_guest_mem_op() directly.

We also should check that the "size" is non-zero here (thanks to Janosch Frank for the hint!). If we do not check the size, we could call vmalloc() with this 0 value, and this will cause a kernel warning.

---
 arch/s390/kvm/kvm-s390.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -3658,7 +3658,7 @@ static long kvm_s390_guest_mem_op(struct
 	const u64 supported_flags = KVM_S390_MEMOP_F_INJECT_EXCEPTION
 				    | KVM_S390_MEMOP_F_CHECK_ONLY;
 
-	if (mop->flags & ~supported_flags)
+	if (mop->flags & ~supported_flags || mop->ar >= NUM_ACRS || !mop->size)
 		return -EINVAL;
 
 	if (mop->size > MEM_OP_MAX_SIZE)

From: Vasily Gorbik <[email protected]>

commit f3122a79a1b0a113d3aea748e0ec26f2cb2889de upstream.

arch_update_cpu_topology is first called from:
kernel_init_freeable->sched_init_smp->sched_init_domains

even before cpus has been registered in:
kernel_init_freeable->do_one_initcall->s390_smp_init

Do not trigger kobject_uevent change events until cpu devices are actually created. Fixes the following kasan findings:

BUG: KASAN: global-out-of-bounds in kobject_uevent_env+0xb40/0xee0
Read of size 8 at addr 0000000000000020 by task swapper/0/1

BUG: KASAN: global-out-of-bounds in kobject_uevent_env+0xb36/0xee0
Read of size 8 at addr 0000000000000018 by task swapper/0/1

CPU: 0 PID: 1 Comm: swapper/0 Tainted: G    B
Hardware name: IBM 3906 M04 704 (LPAR)
Call Trace:
([<0000000143c6db7e>] show_stack+0x14e/0x1a8)
 [<0000000145956498>] dump_stack+0x1d0/0x218
 [<000000014429fb4c>] print_address_description+0x64/0x380
 [<000000014429f630>] __kasan_report+0x138/0x168
 [<0000000145960b96>] kobject_uevent_env+0xb36/0xee0
 [<0000000143c7c47c>] arch_update_cpu_topology+0x104/0x108
 [<0000000143df9e22>] sched_init_domains+0x62/0xe8
 [<000000014644c94a>] sched_init_smp+0x3a/0xc0
 [<0000000146433a20>] kernel_init_freeable+0x558/0x958
 [<000000014599002a>] kernel_init+0x22/0x160
 [<00000001459a71d4>] ret_from_fork+0x28/0x30
 [<00000001459a71dc>] kernel_thread_starter+0x0/0x10

---
 arch/s390/kernel/topology.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

@@ -300,7 +300,8 @@ int arch_update_cpu_topology(void)
 	rc = __arch_update_cpu_topology();
 	for_each_online_cpu(cpu) {
 		dev = get_cpu_device(cpu);
-		kobject_uevent(&dev->kobj, KOBJ_CHANGE);
+		if (dev)
+			kobject_uevent(&dev->kobj, KOBJ_CHANGE);
 	}
 	return rc;
 }

From: Vasily Gorbik <[email protected]>

commit ea298e6ee8b34b3ed4366be7eb799d0650ebe555 upstream.

Fix the following kasan finding:

BUG: KASAN: global-out-of-bounds in ccwgroup_create_dev+0x850/0x1140
Read of size 1 at addr 0000000000000000 by task systemd-udevd.r/561

CPU: 30 PID: 561 Comm: systemd-udevd.r Tainted: G    B
Hardware name: IBM 3906 M04 704 (LPAR)
Call Trace:
([<0000000231b3db7e>] show_stack+0x14e/0x1a8)
 [<0000000233826410>] dump_stack+0x1d0/0x218
 [<000000023216fac4>] print_address_description+0x64/0x380
 [<000000023216f5a8>] __kasan_report+0x138/0x168
 [<00000002331b8378>] ccwgroup_create_dev+0x850/0x1140
 [<00000002332b618a>] group_store+0x3a/0x50
 [<00000002323ac706>] kernfs_fop_write+0x246/0x3b8
 [<00000002321d409a>] vfs_write+0x132/0x450
 [<00000002321d47da>] ksys_write+0x122/0x208
 [<0000000233877102>] system_call+0x2a6/0x2c8

Triggered by:
openat(AT_FDCWD, "/sys/bus/ccwgroup/drivers/qeth/group",

		O_WRONLY|O_CREAT|O_TRUNC|O_CLOEXEC, 0666) = 16

write(16, "0.0.bd00,0.0.bd01,0.0.bd02", 26) = 26

The problem is that __get_next_id in ccwgroup_create_dev might set "buf" buffer pointer to NULL and explicit check for that is required.

---
 drivers/s390/cio/ccwgroup.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -369,7 +369,7 @@ int ccwgroup_create_dev(struct device *p
 		goto error;
 	}
 	/* Check for trailing stuff. */
-	if (i == num_devices && strlen(buf) > 0) {
+	if (i == num_devices && buf && strlen(buf) > 0) {
 		rc = -EINVAL;
 		goto error;
 	}

From: Vasily Gorbik <[email protected]>

commit ab5758848039de9a4b249d46e4ab591197eebaf2 upstream.

ccw console is created early in start_kernel and used before css is initialized or ccw console subchannel is registered. Until then console subchannel does not have a parent. For that reason assume subchannels with no parent are not pseudo subchannels. This fixes the following kasan finding:

BUG: KASAN: global-out-of-bounds in sch_is_pseudo_sch+0x8e/0x98
Read of size 8 at addr 00000000000005e8 by task swapper/0/0

CPU: 0 PID: 0 Comm: swapper/0 Not tainted 5.3.0-rc8-07370-g6ac43dd12538 #2
Hardware name: IBM 2964 NC9 702 (z/VM 6.4.0)
Call Trace:
([<000000000012cd76>] show_stack+0x14e/0x1e0)
 [<0000000001f7fb44>] dump_stack+0x1a4/0x1f8
 [<00000000007d7afc>] print_address_description+0x64/0x3c8
 [<00000000007d75f6>] __kasan_report+0x14e/0x180
 [<00000000018a2986>] sch_is_pseudo_sch+0x8e/0x98
 [<000000000189b950>] cio_enable_subchannel+0x1d0/0x510
 [<00000000018cac7c>] ccw_device_recognition+0x12c/0x188
 [<0000000002ceb1a8>] ccw_device_enable_console+0x138/0x340
 [<0000000002cf1cbe>] con3215_init+0x25e/0x300
 [<0000000002c8770a>] console_init+0x68a/0x9b8
 [<0000000002c6a3d6>] start_kernel+0x4fe/0x728
 [<0000000000100070>] startup_continue+0x70/0xd0

---
 drivers/s390/cio/css.c |    2 ++
 1 file changed, 2 insertions(+)

@@ -1178,6 +1178,8 @@ device_initcall(cio_settle_init);
 
 int sch_is_pseudo_sch(struct subchannel *sch)
 {
+	if (!sch->dev.parent)
+		return 0;
 	return sch == to_css(sch->dev.parent)->pseudo_subchannel;
 }
 

From: Paul Mackerras <[email protected]>

commit ff42df49e75f053a8a6b4c2533100cdcc23afe69 upstream.

On POWER9, when userspace reads the value of the DPDES register on a vCPU, it is possible for 0 to be returned although there is a doorbell interrupt pending for the vCPU. This can lead to a doorbell interrupt being lost across migration. If the guest kernel uses doorbell interrupts for IPIs, then it could malfunction because of the lost interrupt.

This happens because a newly-generated doorbell interrupt is signalled by setting vcpu->arch.doorbell_request to 1; the DPDES value in vcpu->arch.vcore->dpdes is not updated, because it can only be updated when holding the vcpu mutex, in order to avoid races.

To fix this, we OR in vcpu->arch.doorbell_request when reading the DPDES value.

---
 arch/powerpc/kvm/book3s_hv.c |    9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)

@@ -1356,7 +1356,14 @@ static int kvmppc_get_one_reg_hv(struct
 		*val = get_reg_val(id, vcpu->arch.pspb);
 		break;
 	case KVM_REG_PPC_DPDES:
-		*val = get_reg_val(id, vcpu->arch.vcore->dpdes);
+		/*
+		 * On POWER9, where we are emulating msgsndp etc.,
+		 * we return 1 bit for each vcpu, which can come from
+		 * either vcore->dpdes or doorbell_request.
+		 * On POWER8, doorbell_request is 0.
+		 */
+		*val = get_reg_val(id, vcpu->arch.vcore->dpdes |
+				   vcpu->arch.doorbell_request);
 		break;
 	case KVM_REG_PPC_VTB:
 		*val = get_reg_val(id, vcpu->arch.vcore->vtb);

From: Jack Wang <[email protected]>

During backport f7eea636c3d5 ("KVM: nVMX: handle page fault in vmread"), there was a mistake the exception reference should be passed to function kvm_write_guest_virt_system, instead of NULL, other wise, we will get NULL pointer deref, eg

kvm-unit-test triggered a NULL pointer deref below:

[  948.518437] kvm [24114]: vcpu0, guest rIP: 0x407ef9 kvm_set_msr_common: MSR_IA32_DEBUGCTLMSR 0x3, nop
[  949.106464] BUG: unable to handle kernel NULL pointer dereference at 0000000000000000
[  949.106707] PGD 0 P4D 0
[  949.106872] Oops: 0002 [#1] SMP
[  949.107038] CPU: 2 PID: 24126 Comm: qemu-2.7 Not tainted 4.19.77-pserver #4.19.77-1+feature+daily+update+20191005.1625+a4168bb~deb9
[  949.107283] Hardware name: Dell Inc. Precision Tower 3620/09WH54, BIOS 2.7.3 01/31/2018
[  949.107549] RIP: 0010:kvm_write_guest_virt_system+0x12/0x40 [kvm]
[  949.107719] Code: c0 5d 41 5c 41 5d 41 5e 83 f8 03 41 0f 94 c0 41 c1 e0 02 e9 b0 ed ff ff 0f 1f 44 00 00 48 89 f0 c6 87 59 56 00 00 01 48 89 d6 <49> c7 00 00 00 00 00 89 ca 49 c7 40 08 00 00 00 00 49 c7 40 10 00
[  949.108044] RSP: 0018:ffffb31b0a953cb0 EFLAGS: 00010202
[  949.108216] RAX: 000000000046b4d8 RBX: ffff9e9f415b0000 RCX: 0000000000000008
[  949.108389] RDX: ffffb31b0a953cc0 RSI: ffffb31b0a953cc0 RDI: ffff9e9f415b0000
[  949.108562] RBP: 00000000d2e14928 R08: 0000000000000000 R09: 0000000000000000
[  949.108733] R10: 0000000000000000 R11: 0000000000000000 R12: ffffffffffffffc8
[  949.108907] R13: 0000000000000002 R14: ffff9e9f4f26f2e8 R15: 0000000000000000
[  949.109079] FS:  00007eff8694c700(0000) GS:ffff9e9f51a80000(0000) knlGS:0000000031415928
[  949.109318] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[  949.109495] CR2: 0000000000000000 CR3: 00000003be53b002 CR4: 00000000003626e0
[  949.109671] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[  949.109845] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[  949.110017] Call Trace:
[  949.110186]  handle_vmread+0x22b/0x2f0 [kvm_intel]
[  949.110356]  ? vmexit_fill_RSB+0xc/0x30 [kvm_intel]
[  949.110549]  kvm_arch_vcpu_ioctl_run+0xa98/0x1b30 [kvm]
[  949.110725]  ? kvm_vcpu_ioctl+0x388/0x5d0 [kvm]
[  949.110901]  kvm_vcpu_ioctl+0x388/0x5d0 [kvm]
[  949.111072]  do_vfs_ioctl+0xa2/0x620

---
 arch/x86/kvm/vmx.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -8026,7 +8026,7 @@ static int handle_vmread(struct kvm_vcpu
 		/* _system ok, nested_vmx_check_permission has verified cpl=0 */
 		if (kvm_write_guest_virt_system(vcpu, gva, &field_value,
 						(is_long_mode(vcpu) ? 8 : 4),
-						NULL))
+						&e))
 			kvm_inject_page_fault(vcpu, &e);
 	}
 

From: Dmitry Osipenko <[email protected]>

commit 62bacb06b9f08965c4ef10e17875450490c948c0 upstream.

The kHz to Hz is incorrectly converted in a few places in the code, this results in a wrong frequency being calculated because devfreq core uses OPP frequencies that are given in Hz to clamp the rate, while tegra-devfreq gives to the core value in kHz and then it also expects to receive value in kHz from the core. In a result memory freq is always set to a value which is close to ULONG_MAX because of the bug. Hence the EMC frequency is always capped to the maximum and the driver doesn't do anything useful. This patch was tested on Tegra30 and Tegra124 SoC's, EMC frequency scaling works properly now.

---
 drivers/devfreq/tegra-devfreq.c |   12 +++++-------
 1 file changed, 5 insertions(+), 7 deletions(-)

@@ -485,11 +485,11 @@ static int tegra_devfreq_target(struct d
 {
 	struct tegra_devfreq *tegra = dev_get_drvdata(dev);
 	struct dev_pm_opp *opp;
-	unsigned long rate = *freq * KHZ;
+	unsigned long rate;
 
-	opp = devfreq_recommended_opp(dev, &rate, flags);
+	opp = devfreq_recommended_opp(dev, freq, flags);
 	if (IS_ERR(opp)) {
-		dev_err(dev, "Failed to find opp for %lu KHz\n", *freq);
+		dev_err(dev, "Failed to find opp for %lu Hz\n", *freq);
 		return PTR_ERR(opp);
 	}
 	rate = dev_pm_opp_get_freq(opp);

@@ -498,8 +498,6 @@ static int tegra_devfreq_target(struct d
 	clk_set_min_rate(tegra->emc_clock, rate);
 	clk_set_rate(tegra->emc_clock, 0);
 
-	*freq = rate;
-
 	return 0;
 }
 

@@ -509,7 +507,7 @@ static int tegra_devfreq_get_dev_status(
 	struct tegra_devfreq *tegra = dev_get_drvdata(dev);
 	struct tegra_devfreq_device *actmon_dev;
 
-	stat->current_frequency = tegra->cur_freq;
+	stat->current_frequency = tegra->cur_freq * KHZ;
 
 	/* To be used by the tegra governor */
 	stat->private_data = tegra;

@@ -564,7 +562,7 @@ static int tegra_governor_get_target(str
 		target_freq = max(target_freq, dev->target_freq);
 	}
 
-	*freq = target_freq;
+	*freq = target_freq * KHZ;
 
 	return 0;
 }

From: Oleksandr Suvorov <[email protected]>

commit cfc8f568aada98f9608a0a62511ca18d647613e2 upstream.

Prepare to use SND_SOC_DAPM_PRE_POST_PMU definition to reduce coming code size and make it more readable.

---
 include/sound/soc-dapm.h |    2 ++
 1 file changed, 2 insertions(+)

@@ -349,6 +349,8 @@ struct device;
 #define SND_SOC_DAPM_WILL_PMD   0x80    /* called at start of sequence */
 #define SND_SOC_DAPM_PRE_POST_PMD \
 				(SND_SOC_DAPM_PRE_PMD | SND_SOC_DAPM_POST_PMD)
+#define SND_SOC_DAPM_PRE_POST_PMU \
+				(SND_SOC_DAPM_PRE_PMU | SND_SOC_DAPM_POST_PMU)
 
 /* convenience event type detection */
 #define SND_SOC_DAPM_EVENT_ON(e)	\

From: Andrew Donnellan <[email protected]>

commit e7de4f7b64c23e503a8c42af98d56f2a7462bd6d upstream.

Currently the OPAL symbol map is globally readable, which seems bad as it contains physical addresses.

Restrict it to root.

---
 arch/powerpc/platforms/powernv/opal.c |   11 +++++++----
 1 file changed, 7 insertions(+), 4 deletions(-)

@@ -617,7 +617,10 @@ static ssize_t symbol_map_read(struct fi
 				       bin_attr->size);
 }
 
-static BIN_ATTR_RO(symbol_map, 0);
+static struct bin_attribute symbol_map_attr = {
+	.attr = {.name = "symbol_map", .mode = 0400},
+	.read = symbol_map_read
+};
 
 static void opal_export_symmap(void)
 {

@@ -634,10 +637,10 @@ static void opal_export_symmap(void)
 		return;
 
 	/* Setup attributes */
-	bin_attr_symbol_map.private = __va(be64_to_cpu(syms[0]));
-	bin_attr_symbol_map.size = be64_to_cpu(syms[1]);
+	symbol_map_attr.private = __va(be64_to_cpu(syms[0]));
+	symbol_map_attr.size = be64_to_cpu(syms[1]);
 
-	rc = sysfs_create_bin_file(opal_kobj, &bin_attr_symbol_map);
+	rc = sysfs_create_bin_file(opal_kobj, &symbol_map_attr);
 	if (rc)
 		pr_warn("Error %d creating OPAL symbols file\n", rc);
 }

From: Marc Kleine-Budde <[email protected]>

commit d84ea2123f8d27144e3f4d58cd88c9c6ddc799de upstream.

Some boards take longer than 5ms to power up after a reset, so allow some retries attempts before giving up.

---
 drivers/net/can/spi/mcp251x.c |   19 ++++++++++++++-----
 1 file changed, 14 insertions(+), 5 deletions(-)

@@ -627,7 +627,7 @@ static int mcp251x_setup(struct net_devi
 static int mcp251x_hw_reset(struct spi_device *spi)
 {
 	struct mcp251x_priv *priv = spi_get_drvdata(spi);
-	u8 reg;
+	unsigned long timeout;
 	int ret;
 
 	/* Wait for oscillator startup timer after power up */

@@ -641,10 +641,19 @@ static int mcp251x_hw_reset(struct spi_d
 	/* Wait for oscillator startup timer after reset */
 	mdelay(MCP251X_OST_DELAY_MS);
 
-	reg = mcp251x_read_reg(spi, CANSTAT);
-	if ((reg & CANCTRL_REQOP_MASK) != CANCTRL_REQOP_CONF)
-		return -ENODEV;
-
+	/* Wait for reset to finish */
+	timeout = jiffies + HZ;
+	while ((mcp251x_read_reg(spi, CANSTAT) & CANCTRL_REQOP_MASK) !=
+	       CANCTRL_REQOP_CONF) {
+		usleep_range(MCP251X_OST_DELAY_MS * 1000,
+			     MCP251X_OST_DELAY_MS * 1000 * 2);
+
+		if (time_after(jiffies, timeout)) {
+			dev_err(&spi->dev,
+				"MCP251x didn't enter in conf mode after reset\n");
+			return -EBUSY;
+		}
+	}
 	return 0;
 }
 

From: Steven Rostedt (VMware) <[email protected]>

commit 82a2f88458d70704be843961e10b5cef9a6e95d3 upstream.

The tools/lib/traceevent/Makefile had a test added to it to detect a failure of the "nm" when making the dynamic list file (whatever that is). The problem is that the test sorts the values "U W w" and some versions of sort will place "w" ahead of "W" (even though it has a higher ASCII value, and break the test.

Add 'tr "w" "W"' to merge the two and not worry about the ordering.

---
 tools/lib/traceevent/Makefile |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

@@ -259,8 +259,8 @@ endef
 
 define do_generate_dynamic_list_file
 	symbol_type=`$(NM) -u -D $1 | awk 'NF>1 {print $$1}' | \
-	xargs echo "U W w" | tr ' ' '\n' | sort -u | xargs echo`;\
-	if [ "$$symbol_type" = "U W w" ];then				\
+	xargs echo "U w W" | tr 'w ' 'W\n' | sort -u | xargs echo`;\
+	if [ "$$symbol_type" = "U W" ];then				\
 		(echo '{';						\
 		$(NM) -u -D $1 | awk 'NF>1 {print "\t"$$2";"}' | sort -u;\
 		echo '};';						\

From: Alexander Sverdlin <[email protected]>

commit 1b82feb6c5e1996513d0fb0bbb475417088b4954 upstream.

It seems that smp_processor_id() is only used for a best-effort load-balancing, refer to qat_crypto_get_instance_node(). It's not feasible to disable preemption for the duration of the crypto requests. Therefore, just silence the warning. This commit is similar to e7a9b05ca4 ("crypto: cavium - Fix smp_processor_id() warnings").

Silences the following splat:

BUG: using smp_processor_id() in preemptible [00000000] code: cryptomgr_test/2904
caller is qat_alg_ablkcipher_setkey+0x300/0x4a0 [intel_qat]
CPU: 1 PID: 2904 Comm: cryptomgr_test Tainted: P           O    4.14.69 #1
...
Call Trace:
 dump_stack+0x5f/0x86
 check_preemption_disabled+0xd3/0xe0
 qat_alg_ablkcipher_setkey+0x300/0x4a0 [intel_qat]
 skcipher_setkey_ablkcipher+0x2b/0x40
 __test_skcipher+0x1f3/0xb20
 ? cpumask_next_and+0x26/0x40
 ? find_busiest_group+0x10e/0x9d0
 ? preempt_count_add+0x49/0xa0
 ? try_module_get+0x61/0xf0
 ? crypto_mod_get+0x15/0x30
 ? __kmalloc+0x1df/0x1f0
 ? __crypto_alloc_tfm+0x116/0x180
 ? crypto_skcipher_init_tfm+0xa6/0x180
 ? crypto_create_tfm+0x4b/0xf0
 test_skcipher+0x21/0xa0
 alg_test_skcipher+0x3f/0xa0
 alg_test.part.6+0x126/0x2a0
 ? finish_task_switch+0x21b/0x260
 ? __schedule+0x1e9/0x800
 ? __wake_up_common+0x8d/0x140
 cryptomgr_test+0x40/0x50
 kthread+0xff/0x130
 ? cryptomgr_notify+0x540/0x540
 ? kthread_create_on_node+0x70/0x70
 ret_from_fork+0x24/0x50

---
 drivers/crypto/qat/qat_common/adf_common_drv.h |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -95,7 +95,7 @@ struct service_hndl {
 
 static inline int get_current_node(void)
 {
-	return topology_physical_package_id(smp_processor_id());
+	return topology_physical_package_id(raw_smp_processor_id());
 }
 
 int adf_service_register(struct service_hndl *service);

From: Herbert Xu <[email protected]>

commit 0ba3c026e685573bd3534c17e27da7c505ac99c4 upstream.

skcipher_walk_done may be called with an error by internal or external callers. For those internal callers we shouldn't unmap pages but for external callers we must unmap any pages that are in use.

This patch distinguishes between the two cases by checking whether walk->nbytes is zero or not. For internal callers, we now set walk->nbytes to zero prior to the call. For external callers, walk->nbytes has always been non-zero (as zero is used to indicate the termination of a walk).

---
 crypto/skcipher.c |   42 +++++++++++++++++++++++-------------------
 1 file changed, 23 insertions(+), 19 deletions(-)

@@ -95,7 +95,7 @@ static inline u8 *skcipher_get_spot(u8 *
 	return max(start, end_page);
 }
 
-static void skcipher_done_slow(struct skcipher_walk *walk, unsigned int bsize)
+static int skcipher_done_slow(struct skcipher_walk *walk, unsigned int bsize)
 {
 	u8 *addr;
 

@@ -103,19 +103,21 @@ static void skcipher_done_slow(struct sk
 	addr = skcipher_get_spot(addr, bsize);
 	scatterwalk_copychunks(addr, &walk->out, bsize,
 			       (walk->flags & SKCIPHER_WALK_PHYS) ? 2 : 1);
+	return 0;
 }
 
 int skcipher_walk_done(struct skcipher_walk *walk, int err)
 {
-	unsigned int n; /* bytes processed */
-	bool more;
+	unsigned int n = walk->nbytes;
+	unsigned int nbytes = 0;
 
-	if (unlikely(err < 0))
+	if (!n)
 		goto finish;
 
-	n = walk->nbytes - err;
-	walk->total -= n;
-	more = (walk->total != 0);
+	if (likely(err >= 0)) {
+		n -= err;
+		nbytes = walk->total - n;
+	}
 
 	if (likely(!(walk->flags & (SKCIPHER_WALK_PHYS |
 				    SKCIPHER_WALK_SLOW |

@@ -131,7 +133,7 @@ unmap_src:
 		memcpy(walk->dst.virt.addr, walk->page, n);
 		skcipher_unmap_dst(walk);
 	} else if (unlikely(walk->flags & SKCIPHER_WALK_SLOW)) {
-		if (err) {
+		if (err > 0) {
 			/*
 			 * Didn't process all bytes.  Either the algorithm is
 			 * broken, or this was the last step and it turned out

@@ -139,27 +141,29 @@ unmap_src:
 			 * the algorithm requires it.
 			 */
 			err = -EINVAL;
-			goto finish;
-		}
-		skcipher_done_slow(walk, n);
-		goto already_advanced;
+			nbytes = 0;
+		} else
+			n = skcipher_done_slow(walk, n);
 	}
 
+	if (err > 0)
+		err = 0;
+
+	walk->total = nbytes;
+	walk->nbytes = 0;
+
 	scatterwalk_advance(&walk->in, n);
 	scatterwalk_advance(&walk->out, n);
-already_advanced:
-	scatterwalk_done(&walk->in, 0, more);
-	scatterwalk_done(&walk->out, 1, more);
+	scatterwalk_done(&walk->in, 0, nbytes);
+	scatterwalk_done(&walk->out, 1, nbytes);
 
-	if (more) {
+	if (nbytes) {
 		crypto_yield(walk->flags & SKCIPHER_WALK_SLEEP ?
 			     CRYPTO_TFM_REQ_MAY_SLEEP : 0);
 		return skcipher_walk_next(walk);
 	}
-	err = 0;
-finish:
-	walk->nbytes = 0;
 
+finish:
 	/* Short-circuit for the common/fast path. */
 	if (!((unsigned long)walk->buffer | (unsigned long)walk->page))
 		goto out;

From: Wei Yongjun <[email protected]>

commit c552ffb5c93d9d65aaf34f5f001c4e7e8484ced1 upstream.

When using single_open() for opening, single_release() should be used instead of seq_release(), otherwise there is a memory leak.

---
 drivers/crypto/cavium/zip/zip_main.c |    3 +++
 1 file changed, 3 insertions(+)

@@ -595,6 +595,7 @@ static const struct file_operations zip_
 	.owner = THIS_MODULE,
 	.open  = zip_stats_open,
 	.read  = seq_read,
+	.release = single_release,
 };
 
 static int zip_clear_open(struct inode *inode, struct file *file)

@@ -606,6 +607,7 @@ static const struct file_operations zip_
 	.owner = THIS_MODULE,
 	.open  = zip_clear_open,
 	.read  = seq_read,
+	.release = single_release,
 };
 
 static int zip_regs_open(struct inode *inode, struct file *file)

@@ -617,6 +619,7 @@ static const struct file_operations zip_
 	.owner = THIS_MODULE,
 	.open  = zip_regs_open,
 	.read  = seq_read,
+	.release = single_release,
 };
 
 /* Root directory for thunderx_zip debugfs entry */

From: Horia Geantă <[email protected]>

commit 48f89d2a2920166c35b1c0b69917dbb0390ebec7 upstream.

IV transfer from ofifo to class2 (set up at [29][30]) is not guaranteed to be scheduled before the data transfer from ofifo to external memory (set up at [38]:

[29] 10FA0004 ld: ind-nfifo (len=4) imm
[30] 81F00010 <nfifo_entry: ofifo->class2 type=msg len=16>
[31] 14820004 ld: ccb2-datasz len=4 offs=0 imm
[32] 00000010 data:0x00000010
[33] 8210010D operation: cls1-op aes cbc init-final enc
[34] A8080B04 math: (seqin + math0)->vseqout len=4
[35] 28000010 seqfifold: skip len=16
[36] A8080A04 math: (seqin + math0)->vseqin len=4
[37] 2F1E0000 seqfifold: both msg1->2-last2-last1 len=vseqinsz
[38] 69300000 seqfifostr: msg len=vseqoutsz
[39] 5C20000C seqstr: ccb2 ctx len=12 offs=0

If ofifo -> external memory transfer happens first, DECO will hang (issuing a Watchdog Timeout error, if WDOG is enabled) waiting for data availability in ofifo for the ofifo -> c2 ififo transfer.

Make sure IV transfer happens first by waiting for all CAAM internal transfers to end before starting payload transfer.

New descriptor with jump command inserted at [37]:

[..]
[36] A8080A04 math: (seqin + math0)->vseqin len=4
[37] A1000401 jump: jsl1 all-match[!nfifopend] offset=[01] local->[38]
[38] 2F1E0000 seqfifold: both msg1->2-last2-last1 len=vseqinsz
[39] 69300000 seqfifostr: msg len=vseqoutsz
[40] 5C20000C seqstr: ccb2 ctx len=12 offs=0

[Note: the issue is present in the descriptor from the very beginning (cf. Fixes tag). However I've marked it v4.19+ since it's the oldest maintained kernel that the patch applies clean against.]

---
 drivers/crypto/caam/caamalg_desc.c |    9 +++++++++
 drivers/crypto/caam/caamalg_desc.h |    2 +-
 2 files changed, 10 insertions(+), 1 deletion(-)

@@ -476,6 +476,7 @@ void cnstr_shdsc_aead_givencap(u32 * con
 			       const bool is_qi)
 {
 	u32 geniv, moveiv;
+	u32 *wait_cmd;
 
 	/* Note: Context registers are saved. */
 	init_sh_desc_key_aead(desc, cdata, adata, is_rfc3686, nonce);

@@ -566,6 +567,14 @@ copy_iv:
 
 	/* Will read cryptlen */
 	append_math_add(desc, VARSEQINLEN, SEQINLEN, REG0, CAAM_CMD_SZ);
+
+	/*
+	 * Wait for IV transfer (ofifo -> class2) to finish before starting
+	 * ciphertext transfer (ofifo -> external memory).
+	 */
+	wait_cmd = append_jump(desc, JUMP_JSL | JUMP_TEST_ALL | JUMP_COND_NIFP);
+	set_jump_tgt_here(desc, wait_cmd);
+
 	append_seq_fifo_load(desc, 0, FIFOLD_CLASS_BOTH | KEY_VLF |
 			     FIFOLD_TYPE_MSG1OUT2 | FIFOLD_TYPE_LASTBOTH);
 	append_seq_fifo_store(desc, 0, FIFOST_TYPE_MESSAGE_DATA | KEY_VLF);

@@ -12,7 +12,7 @@
 #define DESC_AEAD_BASE			(4 * CAAM_CMD_SZ)
 #define DESC_AEAD_ENC_LEN		(DESC_AEAD_BASE + 11 * CAAM_CMD_SZ)
 #define DESC_AEAD_DEC_LEN		(DESC_AEAD_BASE + 15 * CAAM_CMD_SZ)
-#define DESC_AEAD_GIVENC_LEN		(DESC_AEAD_ENC_LEN + 7 * CAAM_CMD_SZ)
+#define DESC_AEAD_GIVENC_LEN		(DESC_AEAD_ENC_LEN + 8 * CAAM_CMD_SZ)
 #define DESC_QI_AEAD_ENC_LEN		(DESC_AEAD_ENC_LEN + 3 * CAAM_CMD_SZ)
 #define DESC_QI_AEAD_DEC_LEN		(DESC_AEAD_DEC_LEN + 3 * CAAM_CMD_SZ)
 #define DESC_QI_AEAD_GIVENC_LEN		(DESC_AEAD_GIVENC_LEN + 3 * CAAM_CMD_SZ)

From: Jiaxun Yang <[email protected]>

commit d2f965549006acb865c4638f1f030ebcefdc71f6 upstream.

Recently, binutils had split Loongson-3 Extensions into four ASEs:
MMI, CAM, EXT, EXT2. This patch do the samething in kernel and expose them in cpuinfo so applications can probe supported ASEs at runtime.

---
 arch/mips/include/asm/cpu-features.h |   16 ++++++++++++++++
 arch/mips/include/asm/cpu.h          |    4 ++++
 arch/mips/kernel/cpu-probe.c         |    6 ++++++
 arch/mips/kernel/proc.c              |    4 ++++
 4 files changed, 30 insertions(+)

@@ -348,6 +348,22 @@
 #define cpu_has_dsp3		(cpu_data[0].ases & MIPS_ASE_DSP3)
 #endif
 
+#ifndef cpu_has_loongson_mmi
+#define cpu_has_loongson_mmi		__ase(MIPS_ASE_LOONGSON_MMI)
+#endif
+
+#ifndef cpu_has_loongson_cam
+#define cpu_has_loongson_cam		__ase(MIPS_ASE_LOONGSON_CAM)
+#endif
+
+#ifndef cpu_has_loongson_ext
+#define cpu_has_loongson_ext		__ase(MIPS_ASE_LOONGSON_EXT)
+#endif
+
+#ifndef cpu_has_loongson_ext2
+#define cpu_has_loongson_ext2		__ase(MIPS_ASE_LOONGSON_EXT2)
+#endif
+
 #ifndef cpu_has_mipsmt
 #define cpu_has_mipsmt		(cpu_data[0].ases & MIPS_ASE_MIPSMT)
 #endif

@@ -433,5 +433,9 @@ enum cpu_type_enum {
 #define MIPS_ASE_MSA		0x00000100 /* MIPS SIMD Architecture */
 #define MIPS_ASE_DSP3		0x00000200 /* Signal Processing ASE Rev 3*/
 #define MIPS_ASE_MIPS16E2	0x00000400 /* MIPS16e2 */
+#define MIPS_ASE_LOONGSON_MMI	0x00000800 /* Loongson MultiMedia extensions Instructions */
+#define MIPS_ASE_LOONGSON_CAM	0x00001000 /* Loongson CAM */
+#define MIPS_ASE_LOONGSON_EXT	0x00002000 /* Loongson EXTensions */
+#define MIPS_ASE_LOONGSON_EXT2	0x00004000 /* Loongson EXTensions R2 */
 
 #endif /* _ASM_CPU_H */

@@ -1478,6 +1478,8 @@ static inline void cpu_probe_legacy(stru
 			__cpu_name[cpu] = "ICT Loongson-3";
 			set_elf_platform(cpu, "loongson3a");
 			set_isa(c, MIPS_CPU_ISA_M64R1);
+			c->ases |= (MIPS_ASE_LOONGSON_MMI | MIPS_ASE_LOONGSON_CAM |
+				MIPS_ASE_LOONGSON_EXT);
 			break;
 		case PRID_REV_LOONGSON3B_R1:
 		case PRID_REV_LOONGSON3B_R2:

@@ -1485,6 +1487,8 @@ static inline void cpu_probe_legacy(stru
 			__cpu_name[cpu] = "ICT Loongson-3";
 			set_elf_platform(cpu, "loongson3b");
 			set_isa(c, MIPS_CPU_ISA_M64R1);
+			c->ases |= (MIPS_ASE_LOONGSON_MMI | MIPS_ASE_LOONGSON_CAM |
+				MIPS_ASE_LOONGSON_EXT);
 			break;
 		}
 

@@ -1845,6 +1849,8 @@ static inline void cpu_probe_loongson(st
 		decode_configs(c);
 		c->options |= MIPS_CPU_FTLB | MIPS_CPU_TLBINV | MIPS_CPU_LDPTE;
 		c->writecombine = _CACHE_UNCACHED_ACCELERATED;
+		c->ases |= (MIPS_ASE_LOONGSON_MMI | MIPS_ASE_LOONGSON_CAM |
+			MIPS_ASE_LOONGSON_EXT | MIPS_ASE_LOONGSON_EXT2);
 		break;
 	default:
 		panic("Unknown Loongson Processor ID!");

@@ -124,6 +124,10 @@ static int show_cpuinfo(struct seq_file
 	if (cpu_has_eva)	seq_printf(m, "%s", " eva");
 	if (cpu_has_htw)	seq_printf(m, "%s", " htw");
 	if (cpu_has_xpa)	seq_printf(m, "%s", " xpa");
+	if (cpu_has_loongson_mmi)	seq_printf(m, "%s", " loongson-mmi");
+	if (cpu_has_loongson_cam)	seq_printf(m, "%s", " loongson-cam");
+	if (cpu_has_loongson_ext)	seq_printf(m, "%s", " loongson-ext");
+	if (cpu_has_loongson_ext2)	seq_printf(m, "%s", " loongson-ext2");
 	seq_printf(m, "\n");
 
 	if (cpu_has_mmips) {

From: Kees Cook <[email protected]>

commit 314eed30ede02fa925990f535652254b5bad6b65 upstream.

When running on a system with >512MB RAM with a 32-bit kernel built with:

CONFIG_DEBUG_VIRTUAL=y
CONFIG_HIGHMEM=y
CONFIG_HARDENED_USERCOPY=y

all execve()s will fail due to argv copying into kmap()ed pages, and on usercopy checking the calls ultimately of virt_to_page() will be looking for "bad" kmap (highmem) pointers due to CONFIG_DEBUG_VIRTUAL=y:

 ------------[ cut here ]------------
 kernel BUG at ../arch/x86/mm/physaddr.c:83!
 invalid opcode: 0000 [#1] PREEMPT SMP DEBUG_PAGEALLOC
 CPU: 1 PID: 1 Comm: swapper/0 Not tainted 5.3.0-rc8 #6
 Hardware name: Dell Inc. Inspiron 1318/0C236D, BIOS A04 01/15/2009
 EIP: __phys_addr+0xaf/0x100
 ...
 Call Trace:
  __check_object_size+0xaf/0x3c0
  ? __might_sleep+0x80/0xa0
  copy_strings+0x1c2/0x370
  copy_strings_kernel+0x2b/0x40
  __do_execve_file+0x4ca/0x810
  ? kmem_cache_alloc+0x1c7/0x370
  do_execve+0x1b/0x20
  ...

The check is from arch/x86/mm/physaddr.c:

	VIRTUAL_BUG_ON((phys_addr >> PAGE_SHIFT) > max_low_pfn);

Due to the kmap() in fs/exec.c:

		kaddr = kmap(kmapped_page);
	...
	if (copy_from_user(kaddr+offset, str, bytes_to_copy)) ...

Now we can fetch the correct page to avoid the pfn check. In both cases, hardened usercopy will need to walk the page-span checker (if enabled) to do sanity checking.

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

@@ -15,6 +15,7 @@
 #define pr_fmt(fmt) KBUILD_MODNAME ": " fmt
 
 #include <linux/mm.h>
+#include <linux/highmem.h>
 #include <linux/slab.h>
 #include <linux/sched.h>
 #include <linux/sched/task.h>

@@ -203,7 +204,12 @@ static inline const char *check_heap_obj
 	if (!virt_addr_valid(ptr))
 		return NULL;
 
-	page = virt_to_head_page(ptr);
+	/*
+	 * When CONFIG_HIGHMEM=y, kmap_to_page() will give either the
+	 * highmem page or fallback to virt_to_page(). The following
+	 * is effectively a highmem-aware virt_to_head_page().
+	 */
+	page = compound_head(kmap_to_page((void *)ptr));
 
 	/* Check slab allocator for flags and size. */
 	if (PageSlab(page))

From: Li RongQing <[email protected]>

commit e430d802d6a3aaf61bd3ed03d9404888a29b9bf9 upstream.

The timer delayed for more than 3 seconds warning was triggered during testing.

  Workqueue: events_unbound sched_tick_remote
  RIP: 0010:sched_tick_remote+0xee/0x100
  ...
  Call Trace:
   process_one_work+0x18c/0x3a0
   worker_thread+0x30/0x380
   kthread+0x113/0x130
   ret_from_fork+0x22/0x40

The reason is that the code in collect_expired_timers() uses jiffies unprotected:

if (next_event > jiffies)
base->clk = jiffies;

As the compiler is allowed to reload the value base->clk can advance between the check and the store and in the worst case advance farther than next event. That causes the timer expiry to be delayed until the wheel pointer wraps around.

Convert the code to use READ_ONCE()

---
 kernel/time/timer.c |    8 +++++---
 1 file changed, 5 insertions(+), 3 deletions(-)

@@ -1545,21 +1545,23 @@ void timer_clear_idle(void)
 static int collect_expired_timers(struct timer_base *base,
 				  struct hlist_head *heads)
 {
+	unsigned long now = READ_ONCE(jiffies);
+
 	/*
 	 * NOHZ optimization. After a long idle sleep we need to forward the
 	 * base to current jiffies. Avoid a loop by searching the bitfield for
 	 * the next expiring timer.
 	 */
-	if ((long)(jiffies - base->clk) > 2) {
+	if ((long)(now - base->clk) > 2) {
 		unsigned long next = __next_timer_interrupt(base);
 
 		/*
 		 * If the next timer is ahead of time forward to current
 		 * jiffies, otherwise forward to the next expiry time:
 		 */
-		if (time_after(next, jiffies)) {
+		if (time_after(next, now)) {
 			/* The call site will increment clock! */
-			base->clk = jiffies - 1;
+			base->clk = now - 1;
 			return 0;
 		}
 		base->clk = next;

From: Rasmus Villemoes <[email protected]>

commit 144783a80cd2cbc45c6ce17db649140b65f203dd upstream.

Converting from ms to s requires dividing by 1000, not multiplying. So this is currently taking the smaller of new_timeout and 1.28e8, i.e. effectively new_timeout.

The driver knows what it set max_hw_heartbeat_ms to, so use that value instead of doing a division at run-time.

FWIW, this can easily be tested by booting into a busybox shell and doing "watchdog -t 5 -T 130 /dev/watchdog" - without this patch, the watchdog fires after 130&127 == 2 seconds.

---
 drivers/watchdog/imx2_wdt.c |    4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

@@ -58,7 +58,7 @@
 
 #define IMX2_WDT_WMCR		0x08		/* Misc Register */
 
-#define IMX2_WDT_MAX_TIME	128
+#define IMX2_WDT_MAX_TIME	128U
 #define IMX2_WDT_DEFAULT_TIME	60		/* in seconds */
 
 #define WDOG_SEC_TO_COUNT(s)	((s * 2 - 1) << 8)

@@ -183,7 +183,7 @@ static int imx2_wdt_set_timeout(struct w
 {
 	unsigned int actual;
 
-	actual = min(new_timeout, wdog->max_hw_heartbeat_ms * 1000);
+	actual = min(new_timeout, IMX2_WDT_MAX_TIME);
 	__imx2_wdt_set_timeout(wdog, actual);
 	wdog->timeout = new_timeout;
 	return 0;

From: Tomi Valkeinen <[email protected]>

commit e2c4ed148cf3ec8669a1d90dc66966028e5fad70 upstream.

The OMAP36xx and AM/DM37x TRMs say that the maximum divider for DSS fclk (in CM_CLKSEL_DSS) is 32. Experimentation shows that this is not correct, and using divider of 32 breaks DSS with a flood or underflows and sync losts. Dividers up to 31 seem to work fine.

There is another patch to the DT files to limit the divider correctly, but as the DSS driver also needs to know the maximum divider to be able to iteratively find good rates, we also need to do the fix in the DSS driver.

---
 drivers/gpu/drm/omapdrm/dss/dss.c |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -1100,7 +1100,7 @@ static const struct dss_features omap34x
 
 static const struct dss_features omap3630_dss_feats = {
 	.model			=	DSS_MODEL_OMAP3,
-	.fck_div_max		=	32,
+	.fck_div_max		=	31,
 	.fck_freq_max		=	173000000,
 	.dss_fck_multiplier	=	1,
 	.parent_clk_name	=	"dpll4_ck",

From: Russell King <[email protected]>

commit d1c536e3177390da43d99f20143b810c35433d1f upstream.

ADMA errors are potentially data corrupting events; although we print the register state, we do not usefully print the ADMA descriptors. Worse than that, we print them by referencing their virtual address which is meaningless when the register state gives us the DMA address of the failing descriptor.

Print the ADMA descriptors giving their DMA addresses rather than their virtual addresses, and print them using SDHCI_DUMP() rather than DBG().

We also do not show the correct value of the interrupt status register; the register dump shows the current value, after we have cleared the pending interrupts we are going to service. What is more useful is to print the interrupts that _were_ pending at the time the ADMA error was encountered. Fix that too.

---
 drivers/mmc/host/sdhci.c |   15 ++++++++++-----
 1 file changed, 10 insertions(+), 5 deletions(-)

@@ -2638,6 +2638,7 @@ static void sdhci_cmd_irq(struct sdhci_h
 static void sdhci_adma_show_error(struct sdhci_host *host)
 {
 	void *desc = host->adma_table;
+	dma_addr_t dma = host->adma_addr;
 
 	sdhci_dumpregs(host);
 

@@ -2645,18 +2646,21 @@ static void sdhci_adma_show_error(struct
 		struct sdhci_adma2_64_desc *dma_desc = desc;
 
 		if (host->flags & SDHCI_USE_64_BIT_DMA)
-			DBG("%p: DMA 0x%08x%08x, LEN 0x%04x, Attr=0x%02x\n",
-			    desc, le32_to_cpu(dma_desc->addr_hi),
+			SDHCI_DUMP("%08llx: DMA 0x%08x%08x, LEN 0x%04x, Attr=0x%02x\n",
+			    (unsigned long long)dma,
+			    le32_to_cpu(dma_desc->addr_hi),
 			    le32_to_cpu(dma_desc->addr_lo),
 			    le16_to_cpu(dma_desc->len),
 			    le16_to_cpu(dma_desc->cmd));
 		else
-			DBG("%p: DMA 0x%08x, LEN 0x%04x, Attr=0x%02x\n",
-			    desc, le32_to_cpu(dma_desc->addr_lo),
+			SDHCI_DUMP("%08llx: DMA 0x%08x, LEN 0x%04x, Attr=0x%02x\n",
+			    (unsigned long long)dma,
+			    le32_to_cpu(dma_desc->addr_lo),
 			    le16_to_cpu(dma_desc->len),
 			    le16_to_cpu(dma_desc->cmd));
 
 		desc += host->desc_sz;
+		dma += host->desc_sz;
 
 		if (dma_desc->cmd & cpu_to_le16(ADMA2_END))
 			break;

@@ -2732,7 +2736,8 @@ static void sdhci_data_irq(struct sdhci_
 			!= MMC_BUS_TEST_R)
 		host->data->error = -EILSEQ;
 	else if (intmask & SDHCI_INT_ADMA_ERROR) {
-		pr_err("%s: ADMA error\n", mmc_hostname(host->mmc));
+		pr_err("%s: ADMA error: 0x%08x\n", mmc_hostname(host->mmc),
+		       intmask);
 		sdhci_adma_show_error(host);
 		host->data->error = -EIO;
 		if (host->ops->adma_workaround)

From: Russell King <[email protected]>

commit 121bd08b029e03404c451bb237729cdff76eafed upstream.

We must not unconditionally set the DMA snoop bit; if the DMA API is assuming that the device is not DMA coherent, and the device snoops the CPU caches, the device can see stale cache lines brought in by speculative prefetch.

This leads to the device seeing stale data, potentially resulting in corrupted data transfers. Commonly, this results in a descriptor fetch error such as:

mmc0: ADMA error
mmc0: sdhci: ============ SDHCI REGISTER DUMP ===========
mmc0: sdhci: Sys addr: 0x00000000 | Version: 0x00002202
mmc0: sdhci: Blk size: 0x00000008 | Blk cnt: 0x00000001
mmc0: sdhci: Argument: 0x00000000 | Trn mode: 0x00000013
mmc0: sdhci: Present: 0x01f50008 | Host ctl: 0x00000038
mmc0: sdhci: Power: 0x00000003 | Blk gap: 0x00000000
mmc0: sdhci: Wake-up: 0x00000000 | Clock: 0x000040d8
mmc0: sdhci: Timeout: 0x00000003 | Int stat: 0x00000001
mmc0: sdhci: Int enab: 0x037f108f | Sig enab: 0x037f108b
mmc0: sdhci: ACmd stat: 0x00000000 | Slot int: 0x00002202
mmc0: sdhci: Caps: 0x35fa0000 | Caps_1: 0x0000af00
mmc0: sdhci: Cmd: 0x0000333a | Max curr: 0x00000000
mmc0: sdhci: Resp[0]: 0x00000920 | Resp[1]: 0x001d8a33
mmc0: sdhci: Resp[2]: 0x325b5900 | Resp[3]: 0x3f400e00
mmc0: sdhci: Host ctl2: 0x00000000
mmc0: sdhci: ADMA Err: 0x00000009 | ADMA Ptr: 0x000000236d43820c
mmc0: sdhci: ============================================
mmc0: error -5 whilst initialising SD card

but can lead to other errors, and potentially direct the SDHCI controller to read/write data to other memory locations (e.g. if a valid descriptor is visible to the device in a stale cache line.)

Fix this by ensuring that the DMA snoop bit corresponds with the behaviour of the DMA API. Since the driver currently only supports DT, use of_dma_is_coherent(). Note that device_get_dma_attr() can not be used as that risks re-introducing this bug if/when the driver is converted to ACPI.

---
 drivers/mmc/host/sdhci-of-esdhc.c |    7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

@@ -435,7 +435,12 @@ static int esdhc_of_enable_dma(struct sd
 		dma_set_mask_and_coherent(dev, DMA_BIT_MASK(40));
 
 	value = sdhci_readl(host, ESDHC_DMA_SYSCTL);
-	value |= ESDHC_DMA_SNOOP;
+
+	if (of_dma_is_coherent(dev->of_node))
+		value |= ESDHC_DMA_SNOOP;
+	else
+		value &= ~ESDHC_DMA_SNOOP;
+
 	sdhci_writel(host, value, ESDHC_DMA_SYSCTL);
 	return 0;
 }

From: Wanpeng Li <[email protected]>

commit 89340d0935c9296c7b8222b6eab30e67cb57ab82 upstream.

This patch reverts commit 75437bb304b20 (locking/pvqspinlock: Don't wait if vCPU is preempted). A large performance regression was caused by this commit. on over-subscription scenarios.

The test was run on a Xeon Skylake box, 2 sockets, 40 cores, 80 threads, with three VMs of 80 vCPUs each. The score of ebizzy -M is reduced from 13000-14000 records/s to 1700-1800 records/s:

Host Guest score

vanilla w/o kvm optimizations     upstream    1700-1800 records/s
vanilla w/o kvm optimizations     revert      13000-14000 records/s
vanilla w/ kvm optimizations      upstream    4500-5000 records/s
vanilla w/ kvm optimizations      revert      14000-15500 records/s

Exit from aggressive wait-early mechanism can result in premature yield and extra scheduling latency.

Actually, only 6% of wait_early events are caused by vcpu_is_preempted() being true. However, when one vCPU voluntarily releases its vCPU, all the subsequently waiters in the queue will do the same and the cascading effect leads to bad performance.

kvm optimizations:
[1] commit d73eb57b80b (KVM: Boost vCPUs that are delivering interrupts)
[2] commit 266e85a5ec9 (KVM: X86: Boost queue head vCPU to mitigate lock waiter preemption)

---
 kernel/locking/qspinlock_paravirt.h |    2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

@@ -247,7 +247,7 @@ pv_wait_early(struct pv_node *prev, int
 	if ((loop & PV_PREV_CHECK_MASK) != 0)
 		return false;
 
-	return READ_ONCE(prev->state) != vcpu_running || vcpu_is_preempted(prev->cpu);
+	return READ_ONCE(prev->state) != vcpu_running;
 }
 
 /*

From: Juergen Gross <[email protected]>

commit a8fabb38525c51a094607768bac3ba46b3f4a9d5 upstream.

In case a user process using xenbus has open transactions and is killed e.g. via ctrl-C the following cleanup of the allocated resources might result in a deadlock due to trying to end a transaction in the xenbus worker thread:

[ 2551.474706] INFO: task xenbus:37 blocked for more than 120 seconds.
[ 2551.492215]       Tainted: P           OE     5.0.0-29-generic #5
[ 2551.510263] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
[ 2551.528585] xenbus          D    0    37      2 0x80000080
[ 2551.528590] Call Trace:
[ 2551.528603]  __schedule+0x2c0/0x870
[ 2551.528606]  ? _cond_resched+0x19/0x40
[ 2551.528632]  schedule+0x2c/0x70
[ 2551.528637]  xs_talkv+0x1ec/0x2b0
[ 2551.528642]  ? wait_woken+0x80/0x80
[ 2551.528645]  xs_single+0x53/0x80
[ 2551.528648]  xenbus_transaction_end+0x3b/0x70
[ 2551.528651]  xenbus_file_free+0x5a/0x160
[ 2551.528654]  xenbus_dev_queue_reply+0xc4/0x220
[ 2551.528657]  xenbus_thread+0x7de/0x880
[ 2551.528660]  ? wait_woken+0x80/0x80
[ 2551.528665]  kthread+0x121/0x140
[ 2551.528667]  ? xb_read+0x1d0/0x1d0
[ 2551.528670]  ? kthread_park+0x90/0x90
[ 2551.528673]  ret_from_fork+0x35/0x40

Fix this by doing the cleanup via a workqueue instead.

---
 drivers/xen/xenbus/xenbus_dev_frontend.c |   20 ++++++++++++++++++--
 1 file changed, 18 insertions(+), 2 deletions(-)

@@ -55,6 +55,7 @@
 #include <linux/string.h>
 #include <linux/slab.h>
 #include <linux/miscdevice.h>
+#include <linux/workqueue.h>
 
 #include <xen/xenbus.h>
 #include <xen/xen.h>

@@ -113,6 +114,8 @@ struct xenbus_file_priv {
 	wait_queue_head_t read_waitq;
 
 	struct kref kref;
+
+	struct work_struct wq;
 };
 
 /* Read out any raw xenbus messages queued up. */

@@ -297,14 +300,14 @@ static void watch_fired(struct xenbus_wa
 	mutex_unlock(&adap->dev_data->reply_mutex);
 }
 
-static void xenbus_file_free(struct kref *kref)
+static void xenbus_worker(struct work_struct *wq)
 {
 	struct xenbus_file_priv *u;
 	struct xenbus_transaction_holder *trans, *tmp;
 	struct watch_adapter *watch, *tmp_watch;
 	struct read_buffer *rb, *tmp_rb;
 
-	u = container_of(kref, struct xenbus_file_priv, kref);
+	u = container_of(wq, struct xenbus_file_priv, wq);
 
 	/*
 	 * No need for locking here because there are no other users,

@@ -330,6 +333,18 @@ static void xenbus_file_free(struct kref
 	kfree(u);
 }
 
+static void xenbus_file_free(struct kref *kref)
+{
+	struct xenbus_file_priv *u;
+
+	/*
+	 * We might be called in xenbus_thread().
+	 * Use workqueue to avoid deadlock.
+	 */
+	u = container_of(kref, struct xenbus_file_priv, kref);
+	schedule_work(&u->wq);
+}
+
 static struct xenbus_transaction_holder *xenbus_get_transaction(
 	struct xenbus_file_priv *u, uint32_t tx_id)
 {

@@ -626,6 +641,7 @@ static int xenbus_file_open(struct inode
 	INIT_LIST_HEAD(&u->watches);
 	INIT_LIST_HEAD(&u->read_buffers);
 	init_waitqueue_head(&u->read_waitq);
+	INIT_WORK(&u->wq, xenbus_worker);
 
 	mutex_init(&u->reply_mutex);
 	mutex_init(&u->msgbuffer_mutex);

From: Johan Hovold <[email protected]>

commit 7fd25e6fc035f4b04b75bca6d7e8daa069603a76 upstream.

The disconnect callback was accessing the hardware-descriptor private data after having having freed it.

---
 drivers/net/ieee802154/atusb.c |    3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

@@ -907,10 +907,11 @@ static void atusb_disconnect(struct usb_
 
 	ieee802154_unregister_hw(atusb->hw);
 
+	usb_put_dev(atusb->usb_dev);
+
 	ieee802154_free_hw(atusb->hw);
 
 	usb_set_intfdata(interface, NULL);
-	usb_put_dev(atusb->usb_dev);
 
 	pr_debug("atusb_disconnect done\n");
 }

From: Johannes Berg <[email protected]>

commit f43e5210c739fe76a4b0ed851559d6902f20ceb1 upstream.

In a few places we don't properly initialize on-stack chandefs, resulting in EDMG data to be non-zero, which broke things.

Additionally, in a few places we rely on the driver to init the data completely, but perhaps we shouldn't as non-EDMG drivers may not initialize the EDMG data, also initialize it there.

---
 net/wireless/nl80211.c     |    4 +++-
 net/wireless/reg.c         |    2 +-
 net/wireless/wext-compat.c |    2 +-
 3 files changed, 5 insertions(+), 3 deletions(-)

@@ -2111,6 +2111,8 @@ static int nl80211_parse_chandef(struct
 
 	control_freq = nla_get_u32(info->attrs[NL80211_ATTR_WIPHY_FREQ]);
 
+	memset(chandef, 0, sizeof(*chandef));
+
 	chandef->chan = ieee80211_get_channel(&rdev->wiphy, control_freq);
 	chandef->width = NL80211_CHAN_WIDTH_20_NOHT;
 	chandef->center_freq1 = control_freq;

@@ -2580,7 +2582,7 @@ static int nl80211_send_iface(struct sk_
 
 	if (rdev->ops->get_channel) {
 		int ret;
-		struct cfg80211_chan_def chandef;
+		struct cfg80211_chan_def chandef = {};
 
 		ret = rdev_get_channel(rdev, wdev, &chandef);
 		if (ret == 0) {

@@ -1567,7 +1567,7 @@ static void reg_call_notifier(struct wip
 
 static bool reg_wdev_chan_valid(struct wiphy *wiphy, struct wireless_dev *wdev)
 {
-	struct cfg80211_chan_def chandef;
+	struct cfg80211_chan_def chandef = {};
 	struct cfg80211_registered_device *rdev = wiphy_to_rdev(wiphy);
 	enum nl80211_iftype iftype;
 

@@ -800,7 +800,7 @@ static int cfg80211_wext_giwfreq(struct
 {
 	struct wireless_dev *wdev = dev->ieee80211_ptr;
 	struct cfg80211_registered_device *rdev = wiphy_to_rdev(wdev->wiphy);
-	struct cfg80211_chan_def chandef;
+	struct cfg80211_chan_def chandef = {};
 	int ret;
 
 	switch (wdev->iftype) {

From: Sascha Hauer <[email protected]>

[ Upstream commit f5e1040196dbfe14c77ce3dfe3b7b08d2d961e88 ]

integrity_kernel_read() returns the number of bytes read. If this is a short read then this positive value is returned from ima_calc_file_hash_atfm(). Currently this is only indirectly called from ima_calc_file_hash() and this function only tests for the return value being zero or nonzero and also doesn't forward the return value. Nevertheless there's no point in returning a positive value as an error, so translate a short read into -EINVAL.

---
 security/integrity/ima/ima_crypto.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

index af680b5b678a4..06b0ee75f34fb 100644
@@ -293,8 +293,11 @@ static int ima_calc_file_hash_atfm(struct file *file,
 		rbuf_len = min_t(loff_t, i_size - offset, rbuf_size[active]);
 		rc = integrity_kernel_read(file, offset, rbuf[active],
 					   rbuf_len);
-		if (rc != rbuf_len)
+		if (rc != rbuf_len) {
+			if (rc >= 0)
+				rc = -EINVAL;
 			goto out3;
+		}
 
 		if (rbuf[1] && offset) {
 			/* Using two buffers, and it is not the first

2.20.1

From: Jia-Ju Bai <[email protected]>

[ Upstream commit e2751463eaa6f9fec8fea80abbdc62dbc487b3c5 ]

In encode_attrs(), there is an if statement on line 1145 to check whether label is NULL:

    if (label && (attrmask[2] & FATTR4_WORD2_SECURITY_LABEL))

When label is NULL, it is used on lines 1178-1181:

    *p++ = cpu_to_be32(label->lfs);
    *p++ = cpu_to_be32(label->pi);
    *p++ = cpu_to_be32(label->len);
    p = xdr_encode_opaque_fixed(p, label->label, label->len);

To fix these bugs, label is checked before being used.

These bugs are found by a static analysis tool STCheck written by us.

---
 fs/nfs/nfs4xdr.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

index 549c916d28599..525684b0056fc 100644
@@ -1132,7 +1132,7 @@ static void encode_attrs(struct xdr_stream *xdr, const struct iattr *iap,
 		} else
 			*p++ = cpu_to_be32(NFS4_SET_TO_SERVER_TIME);
 	}
-	if (bmval[2] & FATTR4_WORD2_SECURITY_LABEL) {
+	if (label && (bmval[2] & FATTR4_WORD2_SECURITY_LABEL)) {
 		*p++ = cpu_to_be32(label->lfs);
 		*p++ = cpu_to_be32(label->pi);
 		*p++ = cpu_to_be32(label->len);

2.20.1

From: Chengguang Xu <[email protected]>

[ Upstream commit c87a37ebd40b889178664c2c09cc187334146292 ]

Currently on mmap cache policy, we always attach writeback_fid whether mmap type is SHARED or PRIVATE. However, in the use case of kata-container which combines 9p(Guest OS) with overlayfs(Host OS), this behavior will trigger overlayfs' copy-up when excute command inside container.

---
 fs/9p/vfs_file.c | 3 +++
 1 file changed, 3 insertions(+)

index 89e69904976a5..2651192f01667 100644
@@ -528,6 +528,7 @@ v9fs_mmap_file_mmap(struct file *filp, struct vm_area_struct *vma)
 	v9inode = V9FS_I(inode);
 	mutex_lock(&v9inode->v_mutex);
 	if (!v9inode->writeback_fid &&
+	    (vma->vm_flags & VM_SHARED) &&
 	    (vma->vm_flags & VM_WRITE)) {
 		/*
 		 * clone a fid and add it to writeback_fid

@@ -629,6 +630,8 @@ static void v9fs_mmap_vm_close(struct vm_area_struct *vma)
 			(vma->vm_end - vma->vm_start - 1),
 	};
 
+	if (!(vma->vm_flags & VM_SHARED))
+		return;
 
 	p9_debug(P9_DEBUG_VFS, "9p VMA close, %p, flushing", vma);
 

2.20.1

From: Igor Druzhinin <[email protected]>

[ Upstream commit a4098bc6eed5e31e0391bcc068e61804c98138df ]

If MCFG area is not reserved in E820, Xen by default will defer its usage until Dom0 registers it explicitly after ACPI parser recognizes it as a reserved resource in DSDT. Having it reserved in E820 is not mandatory according to "PCI Firmware Specification, rev 3.2" (par. 4.1.2) and firmware is free to keep a hole in E820 in that place. Xen doesn't know what exactly is inside this hole since it lacks full ACPI view of the platform therefore it's potentially harmful to access MCFG region without additional checks as some machines are known to provide inconsistent information on the size of the region.

Now xen_mcfg_late() runs after acpi_init() which is too late as some basic PCI enumeration starts exactly there as well. Trying to register a device prior to MCFG reservation causes multiple problems with PCIe extended capability initializations in Xen (e.g. SR-IOV VF BAR sizing). There are no convenient hooks for us to subscribe to so register MCFG areas earlier upon the first invocation of xen_add_device(). It should be safe to do once since all the boot time buses must have their MCFG areas in MCFG table already and we don't support PCI bus hot-plug.

---
 drivers/xen/pci.c | 21 +++++++++++++++------
 1 file changed, 15 insertions(+), 6 deletions(-)

index 7494dbeb4409c..db58aaa4dc598 100644
@@ -29,6 +29,8 @@
 #include "../pci/pci.h"
 #ifdef CONFIG_PCI_MMCONFIG
 #include <asm/pci_x86.h>
+
+static int xen_mcfg_late(void);
 #endif
 
 static bool __read_mostly pci_seg_supported = true;

@@ -40,7 +42,18 @@ static int xen_add_device(struct device *dev)
 #ifdef CONFIG_PCI_IOV
 	struct pci_dev *physfn = pci_dev->physfn;
 #endif
-
+#ifdef CONFIG_PCI_MMCONFIG
+	static bool pci_mcfg_reserved = false;
+	/*
+	 * Reserve MCFG areas in Xen on first invocation due to this being
+	 * potentially called from inside of acpi_init immediately after
+	 * MCFG table has been finally parsed.
+	 */
+	if (!pci_mcfg_reserved) {
+		xen_mcfg_late();
+		pci_mcfg_reserved = true;
+	}
+#endif
 	if (pci_seg_supported) {
 		struct {
 			struct physdev_pci_device_add add;

@@ -213,7 +226,7 @@ static int __init register_xen_pci_notifier(void)
 arch_initcall(register_xen_pci_notifier);
 
 #ifdef CONFIG_PCI_MMCONFIG
-static int __init xen_mcfg_late(void)
+static int xen_mcfg_late(void)
 {
 	struct pci_mmcfg_region *cfg;
 	int rc;

@@ -252,8 +265,4 @@ static int __init xen_mcfg_late(void)
 	}
 	return 0;
 }
-/*
- * Needs to be done after acpi_init which are subsys_initcall.
- */
-subsys_initcall_sync(xen_mcfg_late);
 #endif

2.20.1

From: Luis Henriques <[email protected]>

[ Upstream commit 750670341a24cb714e624e0fd7da30900ad93752 ]

When filling an inode with info from the MDS, i_blkbits is being initialized using fl_stripe_unit, which contains the stripe unit in bytes. Unfortunately, this doesn't make sense for directories as they have fl_stripe_unit set to '0'. This means that i_blkbits will be set to 0xff, causing an UBSAN undefined behaviour in i_blocksize():

  UBSAN: Undefined behaviour in ./include/linux/fs.h:731:12
  shift exponent 255 is too large for 32-bit type 'int'

Fix this by initializing i_blkbits to CEPH_BLOCK_SHIFT if fl_stripe_unit is zero.

---
 fs/ceph/inode.c | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)

index 9bda8c7a80a05..879bc08250931 100644
@@ -789,7 +789,12 @@ static int fill_inode(struct inode *inode, struct page *locked_page,
 	ci->i_version = le64_to_cpu(info->version);
 	inode->i_version++;
 	inode->i_rdev = le32_to_cpu(info->rdev);
-	inode->i_blkbits = fls(le32_to_cpu(info->layout.fl_stripe_unit)) - 1;
+	/* directories have fl_stripe_unit set to zero */
+	if (le32_to_cpu(info->layout.fl_stripe_unit))
+		inode->i_blkbits =
+			fls(le32_to_cpu(info->layout.fl_stripe_unit)) - 1;
+	else
+		inode->i_blkbits = CEPH_BLOCK_SHIFT;
 
 	if ((new_version || (new_issued & CEPH_CAP_AUTH_SHARED)) &&
 	    (issued & CEPH_CAP_AUTH_EXCL) == 0) {

2.20.1

From: Erqi Chen <[email protected]>

[ Upstream commit 71a228bc8d65900179e37ac309e678f8c523f133 ]

If client mds session is evicted in CEPH_MDS_SESSION_OPENING state, mds won't send session msg to client, and delayed_work skip CEPH_MDS_SESSION_OPENING state session, the session hang forever.

Allow ceph_con_keepalive to reconnect a session in OPENING to avoid session hang. Also, ensure that we skip sessions in RESTARTING and REJECTED states since those states can't be resurrected by issuing a keepalive.

---
 fs/ceph/mds_client.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

index e1ded4bd61154..b968334f841e8 100644
@@ -3543,7 +3543,9 @@ static void delayed_work(struct work_struct *work)
 				pr_info("mds%d hung\n", s->s_mds);
 			}
 		}
-		if (s->s_state < CEPH_MDS_SESSION_OPEN) {
+		if (s->s_state == CEPH_MDS_SESSION_NEW ||
+		    s->s_state == CEPH_MDS_SESSION_RESTARTING ||
+		    s->s_state == CEPH_MDS_SESSION_REJECTED) {
 			/* this mds is failed or recovering, just wait */
 			ceph_put_mds_session(s);
 			continue;

2.20.1

From: Ryan Chen <[email protected]>

[ Upstream commit b3528b4874480818e38e4da019d655413c233e6a ]

The ast2600 can be supported by the same code as the ast2500.

---
 drivers/watchdog/aspeed_wdt.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

index fd91007b4e41b..cee7334b2a000 100644
@@ -38,6 +38,7 @@ static const struct aspeed_wdt_config ast2500_config = {
 static const struct of_device_id aspeed_wdt_of_table[] = {
 	{ .compatible = "aspeed,ast2400-wdt", .data = &ast2400_config },
 	{ .compatible = "aspeed,ast2500-wdt", .data = &ast2500_config },
+	{ .compatible = "aspeed,ast2600-wdt", .data = &ast2500_config },
 	{ },
 };
 MODULE_DEVICE_TABLE(of, aspeed_wdt_of_table);

@@ -257,7 +258,8 @@ static int aspeed_wdt_probe(struct platform_device *pdev)
 		set_bit(WDOG_HW_RUNNING, &wdt->wdd.status);
 	}
 
-	if (of_device_is_compatible(np, "aspeed,ast2500-wdt")) {
+	if ((of_device_is_compatible(np, "aspeed,ast2500-wdt")) ||
+		(of_device_is_compatible(np, "aspeed,ast2600-wdt"))) {
 		u32 reg = readl(wdt->base + WDT_RESET_WIDTH);
 
 		reg &= config->ext_pulse_width_mask;

2.20.1

From: Florian Westphal <[email protected]>

[ Upstream commit acab713177377d9e0889c46bac7ff0cfb9a90c4d ]

This un-breaks lookups in sets that have the 'dynamic' flag set. Given this active example configuration:

table filter {

  set set1 {
    type ipv4_addr
    size 64
    flags dynamic,timeout
    timeout 1m

}

  chain input {
     type filter hook input priority 0; policy accept;

}
}

... this works:
nft add rule ip filter input add @set1 { ip saddr }

-> whenever rule is triggered, the source ip address is inserted into the set (if it did not exist).

This won't work:
nft add rule ip filter input ip saddr @set1 counter
Error: Could not process rule: Operation not supported

In other words, we can add entries to the set, but then can't make matching decision based on that set.

That is just wrong -- all set backends support lookups (else they would not be very useful).
The failure comes from an explicit rejection in nft_lookup.c.

Looking at the history, it seems like NFT_SET_EVAL used to mean 'set contains expressions' (aka. "is a meter"), for instance something like

nft add rule ip filter input meter example { ip saddr limit rate 10/second } or
nft add rule ip filter input meter example { ip saddr counter }

The actual meaning of NFT_SET_EVAL however, is
'set can be updated from the packet path'.

'meters' and packet-path insertions into sets, such as
'add @set { ip saddr }' use exactly the same kernel code (nft_dynset.c) and thus require a set backend that provides the ->update() function.

The only set that provides this also is the only one that has the NFT_SET_EVAL feature flag.

Removing the wrong check makes the above example work. While at it, also fix the flag check during set instantiation to allow supported combinations only.

---
 net/netfilter/nf_tables_api.c | 7 +++++--
 net/netfilter/nft_lookup.c    | 3 ---
 2 files changed, 5 insertions(+), 5 deletions(-)

index b149a72190846..7ef126489d4ed 100644
@@ -3131,8 +3131,11 @@ static int nf_tables_newset(struct net *net, struct sock *nlsk,
 			      NFT_SET_OBJECT))
 			return -EINVAL;
 		/* Only one of these operations is supported */
-		if ((flags & (NFT_SET_MAP | NFT_SET_EVAL | NFT_SET_OBJECT)) ==
-			     (NFT_SET_MAP | NFT_SET_EVAL | NFT_SET_OBJECT))
+		if ((flags & (NFT_SET_MAP | NFT_SET_OBJECT)) ==
+			     (NFT_SET_MAP | NFT_SET_OBJECT))
+			return -EOPNOTSUPP;
+		if ((flags & (NFT_SET_EVAL | NFT_SET_OBJECT)) ==
+			     (NFT_SET_EVAL | NFT_SET_OBJECT))
 			return -EOPNOTSUPP;
 	}
 

index 475570e89ede7..44015a151ad69 100644
@@ -76,9 +76,6 @@ static int nft_lookup_init(const struct nft_ctx *ctx,
 	if (IS_ERR(set))
 		return PTR_ERR(set);
 
-	if (set->flags & NFT_SET_EVAL)
-		return -EOPNOTSUPP;
-
 	priv->sreg = nft_parse_register(tb[NFTA_LOOKUP_SREG]);
 	err = nft_validate_register_load(priv->sreg, set->klen);
 	if (err < 0)

2.20.1

From: Trek <[email protected]>

[ Upstream commit 73d8e6c7b841d9bf298c8928f228fb433676635c ]

Do not try to allocate any amount of memory requested by the user. Instead limit it to 128 registers. Actually the longest series of consecutive allowed registers are 48, mmGB_TILE_MODE0-31 and mmGB_MACROTILE_MODE0-15 (0x2644-0x2673).

---
 drivers/gpu/drm/amd/amdgpu/amdgpu_kms.c | 3 +++
 1 file changed, 3 insertions(+)

index e16229000a983..884ed359f2493 100644
@@ -540,6 +540,9 @@ static int amdgpu_info_ioctl(struct drm_device *dev, void *data, struct drm_file
 		if (sh_num == AMDGPU_INFO_MMR_SH_INDEX_MASK)
 			sh_num = 0xffffffff;
 
+		if (info->read_mmr_reg.count > 128)
+			return -EINVAL;
+
 		regs = kmalloc_array(info->read_mmr_reg.count, sizeof(*regs), GFP_KERNEL);
 		if (!regs)
 			return -ENOMEM;

2.20.1

From: Trond Myklebust <[email protected]>

[ Upstream commit 9c47b18cf722184f32148784189fca945a7d0561 ]

IF the server rejected our layout return with a state error such as NFS4ERR_BAD_STATEID, or even a stale inode error, then we do want to clear out all the remaining layout segments and mark that stateid as invalid.

---
 fs/nfs/pnfs.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)

index 96867fb159bf7..ec04cce31814b 100644
@@ -1319,10 +1319,15 @@ void pnfs_roc_release(struct nfs4_layoutreturn_args *args,
 	const nfs4_stateid *res_stateid = NULL;
 	struct nfs4_xdr_opaque_data *ld_private = args->ld_private;
 
-	if (ret == 0) {
-		arg_stateid = &args->stateid;
+	switch (ret) {
+	case -NFS4ERR_NOMATCHING_LAYOUT:
+		break;
+	case 0:
 		if (res->lrs_present)
 			res_stateid = &res->stateid;
+		/* Fallthrough */
+	default:
+		arg_stateid = &args->stateid;
 	}
 	pnfs_layoutreturn_free_lsegs(lo, arg_stateid, &args->range,
 			res_stateid);

2.20.1

From: Fabrice Gasnier <[email protected]>

[ Upstream commit c91e3234c6035baf5a79763cb4fcd5d23ce75c2b ]

LPTimer can use a 32KHz clock for counting. It depends on clock tree configuration. In such a case, PWM output frequency range is limited. Although unlikely, nothing prevents user from requesting a PWM frequency above counting clock (32KHz for instance):
- This causes (prd - 1) = 0xffff to be written in ARR register later in the apply() routine.
This results in badly configured PWM period (and also duty_cycle). Add a check to report an error is such a case.

---
 drivers/pwm/pwm-stm32-lp.c | 6 ++++++
 1 file changed, 6 insertions(+)

index 9793b296108ff..3f2e4ef695d75 100644
@@ -59,6 +59,12 @@ static int stm32_pwm_lp_apply(struct pwm_chip *chip, struct pwm_device *pwm,
 	/* Calculate the period and prescaler value */
 	div = (unsigned long long)clk_get_rate(priv->clk) * state->period;
 	do_div(div, NSEC_PER_SEC);
+	if (!div) {
+		/* Clock is too slow to achieve requested period. */
+		dev_dbg(priv->chip.dev, "Can't reach %u ns\n",	state->period);
+		return -EINVAL;
+	}
+
 	prd = div;
 	while (div > STM32_LPTIM_MAX_ARR) {
 		presc++;

2.20.1

From: Ido Schimmel <[email protected]>

[ Upstream commit 1851799e1d2978f68eea5d9dff322e121dcf59c1 ]

thermal_zone_device_unregister() cancels the delayed work that polls the thermal zone, but it does not wait for it to finish. This is racy with respect to the freeing of the thermal zone device, which can result in a use-after-free [1].

Fix this by waiting for the delayed work to finish before freeing the thermal zone device. Note that thermal_zone_device_set_polling() is never invoked from an atomic context, so it is safe to call cancel_delayed_work_sync() that can block.

[1]

[  +0.002221] ==================================================================
[  +0.000064] BUG: KASAN: use-after-free in __mutex_lock+0x1076/0x11c0
[  +0.000016] Read of size 8 at addr ffff8881e48e0450 by task kworker/1:0/17

[  +0.000023] CPU: 1 PID: 17 Comm: kworker/1:0 Not tainted 5.2.0-rc6-custom-02495-g8e73ca3be4af #1701
[  +0.000010] Hardware name: Mellanox Technologies Ltd. MSN2100-CB2FO/SA001017, BIOS 5.6.5 06/07/2016
[  +0.000016] Workqueue: events_freezable_power_ thermal_zone_device_check
[  +0.000012] Call Trace:
[  +0.000021]  dump_stack+0xa9/0x10e
[  +0.000020]  print_address_description.cold.2+0x9/0x25e
[  +0.000018]  __kasan_report.cold.3+0x78/0x9d
[  +0.000016]  kasan_report+0xe/0x20
[  +0.000016]  __mutex_lock+0x1076/0x11c0
[  +0.000014]  step_wise_throttle+0x72/0x150
[  +0.000018]  handle_thermal_trip+0x167/0x760
[  +0.000019]  thermal_zone_device_update+0x19e/0x5f0
[  +0.000019]  process_one_work+0x969/0x16f0
[  +0.000017]  worker_thread+0x91/0xc40
[  +0.000014]  kthread+0x33d/0x400
[  +0.000015]  ret_from_fork+0x3a/0x50

[  +0.000020] Allocated by task 1:
[  +0.000015]  save_stack+0x19/0x80
[  +0.000015]  __kasan_kmalloc.constprop.4+0xc1/0xd0
[  +0.000014]  kmem_cache_alloc_trace+0x152/0x320
[  +0.000015]  thermal_zone_device_register+0x1b4/0x13a0
[  +0.000015]  mlxsw_thermal_init+0xc92/0x23d0
[  +0.000014]  __mlxsw_core_bus_device_register+0x659/0x11b0
[  +0.000013]  mlxsw_core_bus_device_register+0x3d/0x90
[  +0.000013]  mlxsw_pci_probe+0x355/0x4b0
[  +0.000014]  local_pci_probe+0xc3/0x150
[  +0.000013]  pci_device_probe+0x280/0x410
[  +0.000013]  really_probe+0x26a/0xbb0
[  +0.000013]  driver_probe_device+0x208/0x2e0
[  +0.000013]  device_driver_attach+0xfe/0x140
[  +0.000013]  __driver_attach+0x110/0x310
[  +0.000013]  bus_for_each_dev+0x14b/0x1d0
[  +0.000013]  driver_register+0x1c0/0x400
[  +0.000015]  mlxsw_sp_module_init+0x5d/0xd3
[  +0.000014]  do_one_initcall+0x239/0x4dd
[  +0.000013]  kernel_init_freeable+0x42b/0x4e8
[  +0.000012]  kernel_init+0x11/0x18b
[  +0.000013]  ret_from_fork+0x3a/0x50

[  +0.000015] Freed by task 581:
[  +0.000013]  save_stack+0x19/0x80
[  +0.000014]  __kasan_slab_free+0x125/0x170
[  +0.000013]  kfree+0xf3/0x310
[  +0.000013]  thermal_release+0xc7/0xf0
[  +0.000014]  device_release+0x77/0x200
[  +0.000014]  kobject_put+0x1a8/0x4c0
[  +0.000014]  device_unregister+0x38/0xc0
[  +0.000014]  thermal_zone_device_unregister+0x54e/0x6a0
[  +0.000014]  mlxsw_thermal_fini+0x184/0x35a
[  +0.000014]  mlxsw_core_bus_device_unregister+0x10a/0x640
[  +0.000013]  mlxsw_devlink_core_bus_device_reload+0x92/0x210
[  +0.000015]  devlink_nl_cmd_reload+0x113/0x1f0
[  +0.000014]  genl_family_rcv_msg+0x700/0xee0
[  +0.000013]  genl_rcv_msg+0xca/0x170
[  +0.000013]  netlink_rcv_skb+0x137/0x3a0
[  +0.000012]  genl_rcv+0x29/0x40
[  +0.000013]  netlink_unicast+0x49b/0x660
[  +0.000013]  netlink_sendmsg+0x755/0xc90
[  +0.000013]  __sys_sendto+0x3de/0x430
[  +0.000013]  __x64_sys_sendto+0xe2/0x1b0
[  +0.000013]  do_syscall_64+0xa4/0x4d0
[  +0.000013]  entry_SYSCALL_64_after_hwframe+0x49/0xbe

[ +0.000017] The buggy address belongs to the object at ffff8881e48e0008
which belongs to the cache kmalloc-2k of size 2048
[ +0.000012] The buggy address is located 1096 bytes inside of

               2048-byte region [ffff8881e48e0008, ffff8881e48e0808)
[  +0.000007] The buggy address belongs to the page:
[  +0.000012] page:ffffea0007923800 refcount:1 mapcount:0 mapping:ffff88823680d0c0 index:0x0 compound_mapcount: 0
[  +0.000020] flags: 0x200000000010200(slab|head)
[  +0.000019] raw: 0200000000010200 ffffea0007682008 ffffea00076ab808 ffff88823680d0c0
[  +0.000016] raw: 0000000000000000 00000000000d000d 00000001ffffffff 0000000000000000
[  +0.000007] page dumped because: kasan: bad access detected

[  +0.000012] Memory state around the buggy address:
[  +0.000012]  ffff8881e48e0300: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
[  +0.000012]  ffff8881e48e0380: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
[  +0.000012] >ffff8881e48e0400: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
[  +0.000008]                                                  ^
[  +0.000012]  ffff8881e48e0480: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
[  +0.000012]  ffff8881e48e0500: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
[  +0.000007] ==================================================================

---
 drivers/thermal/thermal_core.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

index 17d6079c76429..456ef213dc141 100644
@@ -299,7 +299,7 @@ static void thermal_zone_device_set_polling(struct thermal_zone_device *tz,
 		mod_delayed_work(system_freezable_wq, &tz->poll_queue,
 				 msecs_to_jiffies(delay));
 	else
-		cancel_delayed_work(&tz->poll_queue);
+		cancel_delayed_work_sync(&tz->poll_queue);
 }
 
 static void monitor_thermal_zone(struct thermal_zone_device *tz)

2.20.1

From: zhengbin <[email protected]>

[ Upstream commit 9ad09b1976c562061636ff1e01bfc3a57aebe56b ]

If cuse_send_init fails, need to fuse_conn_put cc->fc.

cuse_channel_open->fuse_conn_init->refcount_set(&fc->count, 1)

                 ->fuse_dev_alloc->fuse_conn_get
                 ->fuse_dev_free->fuse_conn_put

---
 fs/fuse/cuse.c | 1 +
 1 file changed, 1 insertion(+)

index e9e97803442a6..55db06c7c587e 100644
@@ -513,6 +513,7 @@ static int cuse_channel_open(struct inode *inode, struct file *file)
 	rc = cuse_send_init(cc);
 	if (rc) {
 		fuse_dev_free(fud);
+		fuse_conn_put(&cc->fc);
 		return rc;
 	}
 	file->private_data = fud;

2.20.1

From: KeMeng Shi <[email protected]>

[ Upstream commit 714e501e16cd473538b609b3e351b2cc9f7f09ed ]

An oops can be triggered in the scheduler when running qemu on arm64:

 Unable to handle kernel paging request at virtual address ffff000008effe40
 Internal error: Oops: 96000007 [#1] SMP
 Process migration/0 (pid: 12, stack limit = 0x00000000084e3736)
 pstate: 20000085 (nzCv daIf -PAN -UAO)
 pc : __ll_sc___cmpxchg_case_acq_4+0x4/0x20
 lr : move_queued_task.isra.21+0x124/0x298
 ...
 Call trace:
  __ll_sc___cmpxchg_case_acq_4+0x4/0x20
  __migrate_task+0xc8/0xe0
  migration_cpu_stop+0x170/0x180
  cpu_stopper_thread+0xec/0x178
  smpboot_thread_fn+0x1ac/0x1e8
  kthread+0x134/0x138
  ret_from_fork+0x10/0x18

__set_cpus_allowed_ptr() will choose an active dest_cpu in affinity mask to migrage the process if process is not currently running on any one of the CPUs specified in affinity mask. __set_cpus_allowed_ptr() will choose an invalid dest_cpu (dest_cpu >= nr_cpu_ids, 1024 in my virtual machine) if CPUS in an affinity mask are deactived by cpu_down after cpumask_intersects check. cpumask_test_cpu() of dest_cpu afterwards is overflown and may pass if corresponding bit is coincidentally set. As a consequence, kernel will access an invalid rq address associate with the invalid CPU in migration_cpu_stop->__migrate_task->move_queued_task and the Oops occurs.

The reproduce the crash:

1) A process repeatedly binds itself to cpu0 and cpu1 in turn by calling sched_setaffinity.

2) A shell script repeatedly does "echo 0 > /sys/devices/system/cpu/cpu1/online" and "echo 1 > /sys/devices/system/cpu/cpu1/online" in turn.

3) Oops appears if the invalid CPU is set in memory after tested cpumask.

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

index 3d24d401b9d42..32ba789c544ca 100644
@@ -1111,7 +1111,8 @@ static int __set_cpus_allowed_ptr(struct task_struct *p,
 	if (cpumask_equal(&p->cpus_allowed, new_mask))
 		goto out;
 
-	if (!cpumask_intersects(new_mask, cpu_valid_mask)) {
+	dest_cpu = cpumask_any_and(cpu_valid_mask, new_mask);
+	if (dest_cpu >= nr_cpu_ids) {
 		ret = -EINVAL;
 		goto out;
 	}

@@ -1132,7 +1133,6 @@ static int __set_cpus_allowed_ptr(struct task_struct *p,
 	if (cpumask_test_cpu(task_cpu(p), new_mask))
 		goto out;
 
-	dest_cpu = cpumask_any_and(cpu_valid_mask, new_mask);
 	if (task_running(rq, p) || p->state == TASK_WAKING) {
 		struct migration_arg arg = { p, dest_cpu };
 		/* Need help from migration thread: drop lock and wait. */

2.20.1

From: Thomas Richter <[email protected]>

[ Upstream commit 815c1560bf8fd522b8d93a1d727868b910c1cc24 ]

With Java 11 there is no seperate JRE anymore.

Details:

https://coderanch.com/t/701603/java/JRE-JDK

Therefore the detection of the JRE needs to be adapted.

This change works for s390 and x86. I have not tested other platforms.

Committer testing:

Continues to work with the OpenJDK 8:

  $ rm -f ~acme/lib64/libperf-jvmti.so
  $ rpm -qa | grep jdk-devel
  java-1.8.0-openjdk-devel-1.8.0.222.b10-0.fc30.x86_64
  $ git log --oneline -1
  a51937170f33 (HEAD -> perf/core) perf build: Add detection of java-11-openjdk-devel package
  $ rm -rf /tmp/build/perf ; mkdir -p /tmp/build/perf ; make -C tools/perf O=/tmp/build/perf install > /dev/null 2>1
  $ ls -la ~acme/lib64/libperf-jvmti.so
  -rwxr-xr-x. 1 acme acme 230744 Sep 24 16:46 /home/acme/lib64/libperf-jvmti.so
  $

---
 tools/perf/Makefile.config | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

index f362ee46506ad..b97e31498ff76 100644
@@ -795,7 +795,7 @@ ifndef NO_JVMTI
     JDIR=$(shell /usr/sbin/update-java-alternatives -l | head -1 | awk '{print $$3}')
   else
     ifneq (,$(wildcard /usr/sbin/alternatives))
-      JDIR=$(shell /usr/sbin/alternatives --display java | tail -1 | cut -d' ' -f 5 | sed 's%/jre/bin/java.%%g')
+      JDIR=$(shell /usr/sbin/alternatives --display java | tail -1 | cut -d' ' -f 5 | sed -e 's%/jre/bin/java.%%g' -e 's%/bin/java.%%g')
     endif
   endif
   ifndef JDIR

2.20.1

From: Valdis Kletnieks <[email protected]>

[ Upstream commit 0f74914071ab7e7b78731ed62bf350e3a344e0a5 ]

When building with W=1, gcc properly complains that there's no prototypes:

CC kernel/elfcore.o
kernel/elfcore.c:7:17: warning: no previous prototype for 'elf_core_extra_phdrs' [-Wmissing-prototypes]
7 | Elf_Half __weak elf_core_extra_phdrs(void)
| ^~~~~~~~~~~~~~~~~~~~
kernel/elfcore.c:12:12: warning: no previous prototype for 'elf_core_write_extra_phdrs' [-Wmissing-prototypes]
12 | int __weak elf_core_write_extra_phdrs(struct coredump_params *cprm, loff_t offset)

      |            ^~~~~~~~~~~~~~~~~~~~~~~~~~
kernel/elfcore.c:17:12: warning: no previous prototype for 'elf_core_write_extra_data' [-Wmissing-prototypes]
   17 | int __weak elf_core_write_extra_data(struct coredump_params *cprm)
      |            ^~~~~~~~~~~~~~~~~~~~~~~~~
kernel/elfcore.c:22:15: warning: no previous prototype for 'elf_core_extra_data_size' [-Wmissing-prototypes]
   22 | size_t __weak elf_core_extra_data_size(void)
      |               ^~~~~~~~~~~~~~~~~~~~~~~~

Provide the include file so gcc is happy, and we don't have potential code drift

---
 kernel/elfcore.c | 1 +
 1 file changed, 1 insertion(+)

index fc482c8e0bd88..57fb4dcff4349 100644
@@ -3,6 +3,7 @@
 #include <linux/fs.h>
 #include <linux/mm.h>
 #include <linux/binfmts.h>
+#include <linux/elfcore.h>
 
 Elf_Half __weak elf_core_extra_phdrs(void)
 {

2.20.1

From: Arnaldo Carvalho de Melo <[email protected]>

[ Upstream commit 26acf400d2dcc72c7e713e1f55db47ad92010cc2 ]

Naresh Kamboju reported, that on the i386 build pr_err() doesn't get defined properly due to header ordering:

perf-in.o: In function `libunwind__x86_reg_id':
tools/perf/util/libunwind/../../arch/x86/util/unwind-libunwind.c:109:
undefined reference to `pr_err'

---
 tools/perf/arch/x86/util/unwind-libunwind.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

index 05920e3edf7a7..47357973b55b2 100644
@@ -1,11 +1,11 @@
 // SPDX-License-Identifier: GPL-2.0
 
 #include <errno.h>
+#include "../../util/debug.h"
 #ifndef REMOTE_UNWIND_LIBUNWIND
 #include <libunwind.h>
 #include "perf_regs.h"
 #include "../../util/unwind.h"
-#include "../../util/debug.h"
 #endif
 
 #ifdef HAVE_ARCH_X86_64_SUPPORT

2.20.1

From: Cédric Le Goater <[email protected]>

[ Upstream commit 237aed48c642328ff0ab19b63423634340224a06 ]

When a vCPU is brought done, the XIVE VP (Virtual Processor) is first disabled and then the event notification queues are freed. When freeing the queues, we check for possible escalation interrupts and free them also.

But when a XIVE VP is disabled, the underlying XIVE ENDs also are disabled in OPAL. When an END (Event Notification Descriptor) is disabled, its ESB pages (ESn and ESe) are disabled and loads return all 1s. Which means that any access on the ESB page of the escalation interrupt will return invalid values.

When an interrupt is freed, the shutdown handler computes a 'saved_p' field from the value returned by a load in xive_do_source_set_mask(). This value is incorrect for escalation interrupts for the reason described above.

This has no impact on Linux/KVM today because we don't make use of it but we will introduce in future changes a xive_get_irqchip_state() handler. This handler will use the 'saved_p' field to return the state of an interrupt and 'saved_p' being incorrect, softlockup will occur.

Fix the vCPU cleanup sequence by first freeing the escalation interrupts if any, then disable the XIVE VP and last free the queues.

---
 arch/powerpc/kvm/book3s_xive.c | 18 ++++++++++--------
 1 file changed, 10 insertions(+), 8 deletions(-)

index 3c75eee45edf9..46f99fc1901c8 100644
@@ -1001,20 +1001,22 @@ void kvmppc_xive_cleanup_vcpu(struct kvm_vcpu *vcpu)
 	/* Mask the VP IPI */
 	xive_vm_esb_load(&xc->vp_ipi_data, XIVE_ESB_SET_PQ_01);
 
-	/* Disable the VP */
-	xive_native_disable_vp(xc->vp_id);
-
-	/* Free the queues & associated interrupts */
+	/* Free escalations */
 	for (i = 0; i < KVMPPC_XIVE_Q_COUNT; i++) {
-		struct xive_q *q = &xc->queues[i];
-
-		/* Free the escalation irq */
 		if (xc->esc_virq[i]) {
 			free_irq(xc->esc_virq[i], vcpu);
 			irq_dispose_mapping(xc->esc_virq[i]);
 			kfree(xc->esc_virq_names[i]);
 		}
-		/* Free the queue */
+	}
+
+	/* Disable the VP */
+	xive_native_disable_vp(xc->vp_id);
+
+	/* Free the queues */
+	for (i = 0; i < KVMPPC_XIVE_Q_COUNT; i++) {
+		struct xive_q *q = &xc->queues[i];
+
 		xive_native_disable_queue(xc->vp_id, q, i);
 		if (q->qpage) {
 			free_pages((unsigned long)q->qpage,

2.20.1

From: Xiubo Li <[email protected]>

[ Upstream commit 553768d1169a48c0cd87c4eb4ab57534ee663415 ]

This will allow the blksize to be set zero and then use 1024 as default.

---
 drivers/block/nbd.c | 23 ++++++++++++++++++++---
 1 file changed, 20 insertions(+), 3 deletions(-)

index e4b049f281f50..a65e4ed6c9372 100644
@@ -132,6 +132,8 @@ static struct dentry *nbd_dbg_dir;
 
 #define NBD_MAGIC 0x68797548
 
+#define NBD_DEF_BLKSIZE 1024
+
 static unsigned int nbds_max = 16;
 static int max_part = 16;
 static struct workqueue_struct *recv_workqueue;

@@ -1216,6 +1218,14 @@ static void nbd_clear_sock_ioctl(struct nbd_device *nbd,
 		nbd_config_put(nbd);
 }
 
+static bool nbd_is_valid_blksize(unsigned long blksize)
+{
+	if (!blksize || !is_power_of_2(blksize) || blksize < 512 ||
+	    blksize > PAGE_SIZE)
+		return false;
+	return true;
+}
+
 /* Must be called with config_lock held */
 static int __nbd_ioctl(struct block_device *bdev, struct nbd_device *nbd,
 		       unsigned int cmd, unsigned long arg)

@@ -1231,8 +1241,9 @@ static int __nbd_ioctl(struct block_device *bdev, struct nbd_device *nbd,
 	case NBD_SET_SOCK:
 		return nbd_add_socket(nbd, arg, false);
 	case NBD_SET_BLKSIZE:
-		if (!arg || !is_power_of_2(arg) || arg < 512 ||
-		    arg > PAGE_SIZE)
+		if (!arg)
+			arg = NBD_DEF_BLKSIZE;
+		if (!nbd_is_valid_blksize(arg))
 			return -EINVAL;
 		nbd_size_set(nbd, arg,
 			     div_s64(config->bytesize, arg));

@@ -1312,7 +1323,7 @@ static struct nbd_config *nbd_alloc_config(void)
 	atomic_set(&config->recv_threads, 0);
 	init_waitqueue_head(&config->recv_wq);
 	init_waitqueue_head(&config->conn_wait);
-	config->blksize = 1024;
+	config->blksize = NBD_DEF_BLKSIZE;
 	atomic_set(&config->live_connections, 0);
 	try_module_get(THIS_MODULE);
 	return config;

@@ -1744,6 +1755,12 @@ static int nbd_genl_connect(struct sk_buff *skb, struct genl_info *info)
 	if (info->attrs[NBD_ATTR_BLOCK_SIZE_BYTES]) {
 		u64 bsize =
 			nla_get_u64(info->attrs[NBD_ATTR_BLOCK_SIZE_BYTES]);
+		if (!bsize)
+			bsize = NBD_DEF_BLKSIZE;
+		if (!nbd_is_valid_blksize(bsize)) {
+			ret = -EINVAL;
+			goto out;
+		}
 		nbd_size_set(nbd, bsize, div64_u64(config->bytesize, bsize));
 	}
 	if (info->attrs[NBD_ATTR_TIMEOUT]) {

2.20.1

From: Dan Melnic <[email protected]>

[ Upstream commit 2189c97cdbed630d5971ab22f05dc998774e354e ]

Add WQ_UNBOUND to the knbd-recv workqueue so we're not bound to a single CPU that is selected at device creation time.

---
 drivers/block/nbd.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

index a65e4ed6c9372..14b491c5cf7b6 100644
@@ -2217,7 +2217,8 @@ static int __init nbd_init(void)
 	if (nbds_max > 1UL << (MINORBITS - part_shift))
 		return -EINVAL;
 	recv_workqueue = alloc_workqueue("knbd-recv",
-					 WQ_MEM_RECLAIM | WQ_HIGHPRI, 0);
+					 WQ_MEM_RECLAIM | WQ_HIGHPRI |
+					 WQ_UNBOUND, 0);
 	if (!recv_workqueue)
 		return -ENOMEM;
 

2.20.1

From: Mike Christie <[email protected]>

[ Upstream commit e9e006f5fcf2bab59149cb38a48a4817c1b538b4 ]

This fixes a bug added in 4.10 with commit:

commit 9561a7ade0c205bc2ee035a2ac880478dcc1a024

Author: Josef Bacik <[email protected]>
Date:   Tue Nov 22 14:04:40 2016 -0500

nbd: add multi-connection support

that limited the number of devices to 256. Before the patch we could create 1000s of devices, but the patch switched us from using our own thread to using a work queue which has a default limit of 256 active works.

The problem is that our recv_work function sits in a loop until disconnection but only handles IO for one connection. The work is started when the connection is started/restarted, but if we end up creating 257 or more connections, the queue_work call just queues connection257+'s recv_work and that waits for connection 1 - 256's recv_work to be disconnected and that work instance completing.

Instead of reverting back to kthreads, this has us allocate a workqueue_struct per device, so we can block in the work.

---
 drivers/block/nbd.c | 39 +++++++++++++++++++++++++--------------
 1 file changed, 25 insertions(+), 14 deletions(-)

index 14b491c5cf7b6..a234600849558 100644
@@ -106,6 +106,7 @@ struct nbd_device {
 	struct nbd_config *config;
 	struct mutex config_lock;
 	struct gendisk *disk;
+	struct workqueue_struct *recv_workq;
 
 	struct list_head list;
 	struct task_struct *task_recv;

@@ -136,7 +137,6 @@ static struct dentry *nbd_dbg_dir;
 
 static unsigned int nbds_max = 16;
 static int max_part = 16;
-static struct workqueue_struct *recv_workqueue;
 static int part_shift;
 
 static int nbd_dev_dbg_init(struct nbd_device *nbd);

@@ -1015,7 +1015,7 @@ static int nbd_reconnect_socket(struct nbd_device *nbd, unsigned long arg)
 		/* We take the tx_mutex in an error path in the recv_work, so we
 		 * need to queue_work outside of the tx_mutex.
 		 */
-		queue_work(recv_workqueue, &args->work);
+		queue_work(nbd->recv_workq, &args->work);
 
 		atomic_inc(&config->live_connections);
 		wake_up(&config->conn_wait);

@@ -1120,6 +1120,10 @@ static void nbd_config_put(struct nbd_device *nbd)
 		kfree(nbd->config);
 		nbd->config = NULL;
 
+		if (nbd->recv_workq)
+			destroy_workqueue(nbd->recv_workq);
+		nbd->recv_workq = NULL;
+
 		nbd->tag_set.timeout = 0;
 		queue_flag_clear_unlocked(QUEUE_FLAG_DISCARD, nbd->disk->queue);
 

@@ -1145,6 +1149,14 @@ static int nbd_start_device(struct nbd_device *nbd)
 		return -EINVAL;
 	}
 
+	nbd->recv_workq = alloc_workqueue("knbd%d-recv",
+					  WQ_MEM_RECLAIM | WQ_HIGHPRI |
+					  WQ_UNBOUND, 0, nbd->index);
+	if (!nbd->recv_workq) {
+		dev_err(disk_to_dev(nbd->disk), "Could not allocate knbd recv work queue.\n");
+		return -ENOMEM;
+	}
+
 	blk_mq_update_nr_hw_queues(&nbd->tag_set, config->num_connections);
 	nbd->task_recv = current;
 

@@ -1175,7 +1187,7 @@ static int nbd_start_device(struct nbd_device *nbd)
 		INIT_WORK(&args->work, recv_work);
 		args->nbd = nbd;
 		args->index = i;
-		queue_work(recv_workqueue, &args->work);
+		queue_work(nbd->recv_workq, &args->work);
 	}
 	nbd_size_update(nbd);
 	return error;

@@ -1195,8 +1207,10 @@ static int nbd_start_device_ioctl(struct nbd_device *nbd, struct block_device *b
 	mutex_unlock(&nbd->config_lock);
 	ret = wait_event_interruptible(config->recv_wq,
 					 atomic_read(&config->recv_threads) == 0);
-	if (ret)
+	if (ret) {
 		sock_shutdown(nbd);
+		flush_workqueue(nbd->recv_workq);
+	}
 	mutex_lock(&nbd->config_lock);
 	bd_set_size(bdev, 0);
 	/* user requested, ignore socket errors */

@@ -1836,6 +1850,12 @@ static void nbd_disconnect_and_put(struct nbd_device *nbd)
 	mutex_lock(&nbd->config_lock);
 	nbd_disconnect(nbd);
 	mutex_unlock(&nbd->config_lock);
+	/*
+	 * Make sure recv thread has finished, so it does not drop the last
+	 * config ref and try to destroy the workqueue from inside the work
+	 * queue.
+	 */
+	flush_workqueue(nbd->recv_workq);
 	if (test_and_clear_bit(NBD_HAS_CONFIG_REF,
 			       &nbd->config->runtime_flags))
 		nbd_config_put(nbd);

@@ -2216,20 +2236,12 @@ static int __init nbd_init(void)
 
 	if (nbds_max > 1UL << (MINORBITS - part_shift))
 		return -EINVAL;
-	recv_workqueue = alloc_workqueue("knbd-recv",
-					 WQ_MEM_RECLAIM | WQ_HIGHPRI |
-					 WQ_UNBOUND, 0);
-	if (!recv_workqueue)
-		return -ENOMEM;
 
-	if (register_blkdev(NBD_MAJOR, "nbd")) {
-		destroy_workqueue(recv_workqueue);
+	if (register_blkdev(NBD_MAJOR, "nbd"))
 		return -EIO;
-	}
 
 	if (genl_register_family(&nbd_genl_family)) {
 		unregister_blkdev(NBD_MAJOR, "nbd");
-		destroy_workqueue(recv_workqueue);
 		return -EINVAL;
 	}
 	nbd_dbg_init();

@@ -2271,7 +2283,6 @@ static void __exit nbd_cleanup(void)
 
 	idr_destroy(&nbd_index_idr);
 	genl_unregister_family(&nbd_genl_family);
-	destroy_workqueue(recv_workqueue);
 	unregister_blkdev(NBD_MAJOR, "nbd");
 }
 

2.20.1

From: Gautham R. Shenoy <[email protected]>

[ Upstream commit c784be435d5dae28d3b03db31753dd7a18733f0c ]

The calls to arch_add_memory()/arch_remove_memory() are always made with the read-side cpu_hotplug_lock acquired via memory_hotplug_begin(). On pSeries, arch_add_memory()/arch_remove_memory() eventually call resize_hpt() which in turn calls stop_machine() which acquires the read-side cpu_hotplug_lock again, thereby resulting in the recursive acquisition of this lock.

In the absence of CONFIG_PROVE_LOCKING, we hadn't observed a system lockup during a memory hotplug operation because cpus_read_lock() is a per-cpu rwsem read, which, in the fast-path (in the absence of the writer, which in our case is a CPU-hotplug operation) simply increments the read_count on the semaphore. Thus a recursive read in the fast-path doesn't cause any problems.

However, we can hit this problem in practice if there is a concurrent CPU-Hotplug operation in progress which is waiting to acquire the write-side of the lock. This will cause the second recursive read to block until the writer finishes. While the writer is blocked since the first read holds the lock. Thus both the reader as well as the writers fail to make any progress thereby blocking both CPU-Hotplug as well as Memory Hotplug operations.

Memory-Hotplug CPU-Hotplug
CPU 0 CPU 1
------ ------

1. down_read(cpu_hotplug_lock.rw_sem)
[memory_hotplug_begin]

					2. down_write(cpu_hotplug_lock.rw_sem)
					[cpu_up/cpu_down]

3. down_read(cpu_hotplug_lock.rw_sem)
[stop_machine()]

Lockdep complains as follows in these code-paths.

swapper/0/1 is trying to acquire lock:
(____ptrval____) (cpu_hotplug_lock.rw_sem){++++}, at: stop_machine+0x2c/0x60

but task is already holding lock:
(____ptrval____) (cpu_hotplug_lock.rw_sem){++++}, at: mem_hotplug_begin+0x20/0x50

other info that might help us debug this:

  Possible unsafe locking scenario:

        CPU0
        ----

lock(cpu_hotplug_lock.rw_sem);
lock(cpu_hotplug_lock.rw_sem);

*** DEADLOCK ***

May be due to missing lock nesting notation

3 locks held by swapper/0/1:
#0: (____ptrval____) (&dev->mutex){....}, at: __driver_attach+0x12c/0x1b0 #1: (____ptrval____) (cpu_hotplug_lock.rw_sem){++++}, at: mem_hotplug_begin+0x20/0x50 #2: (____ptrval____) (mem_hotplug_lock.rw_sem){++++}, at: percpu_down_write+0x54/0x1a0

stack backtrace:
 CPU: 0 PID: 1 Comm: swapper/0 Not tainted 5.0.0-rc5-58373-gbc99402235f3-dirty #166
 Call Trace:
   dump_stack+0xe8/0x164 (unreliable)
   __lock_acquire+0x1110/0x1c70
   lock_acquire+0x240/0x290
   cpus_read_lock+0x64/0xf0
   stop_machine+0x2c/0x60
   pseries_lpar_resize_hpt+0x19c/0x2c0
   resize_hpt_for_hotplug+0x70/0xd0
   arch_add_memory+0x58/0xfc
   devm_memremap_pages+0x5e8/0x8f0
   pmem_attach_disk+0x764/0x830
   nvdimm_bus_probe+0x118/0x240
   really_probe+0x230/0x4b0
   driver_probe_device+0x16c/0x1e0
   __driver_attach+0x148/0x1b0
   bus_for_each_dev+0x90/0x130
   driver_attach+0x34/0x50
   bus_add_driver+0x1a8/0x360
   driver_register+0x108/0x170
   __nd_driver_register+0xd0/0xf0
   nd_pmem_driver_init+0x34/0x48
   do_one_initcall+0x1e0/0x45c
   kernel_init_freeable+0x540/0x64c
   kernel_init+0x2c/0x160
   ret_from_kernel_thread+0x5c/0x68

Fix this issue by
1) Requiring all the calls to pseries_lpar_resize_hpt() be made with cpu_hotplug_lock held.

2) In pseries_lpar_resize_hpt() invoke stop_machine_cpuslocked() as a consequence of 1)

3) To satisfy 1), in hpt_order_set(), call mmu_hash_ops.resize_hpt() with cpu_hotplug_lock held.

---
 arch/powerpc/mm/hash_utils_64.c       | 9 ++++++++-
 arch/powerpc/platforms/pseries/lpar.c | 8 ++++++--
 2 files changed, 14 insertions(+), 3 deletions(-)

index 87687e46b48bb..58c14749bb0c1 100644
@@ -35,6 +35,7 @@
 #include <linux/memblock.h>
 #include <linux/context_tracking.h>
 #include <linux/libfdt.h>
+#include <linux/cpu.h>
 
 #include <asm/debugfs.h>
 #include <asm/processor.h>

@@ -1852,10 +1853,16 @@ static int hpt_order_get(void *data, u64 *val)
 
 static int hpt_order_set(void *data, u64 val)
 {
+	int ret;
+
 	if (!mmu_hash_ops.resize_hpt)
 		return -ENODEV;
 
-	return mmu_hash_ops.resize_hpt(val);
+	cpus_read_lock();
+	ret = mmu_hash_ops.resize_hpt(val);
+	cpus_read_unlock();
+
+	return ret;
 }
 
 DEFINE_SIMPLE_ATTRIBUTE(fops_hpt_order, hpt_order_get, hpt_order_set, "%llu\n");

index 55e97565ed2dd..eb738ef577926 100644
@@ -643,7 +643,10 @@ static int pseries_lpar_resize_hpt_commit(void *data)
 	return 0;
 }
 
-/* Must be called in user context */
+/*
+ * Must be called in process context. The caller must hold the
+ * cpus_lock.
+ */
 static int pseries_lpar_resize_hpt(unsigned long shift)
 {
 	struct hpt_resize_state state = {

@@ -699,7 +702,8 @@ static int pseries_lpar_resize_hpt(unsigned long shift)
 
 	t1 = ktime_get();
 
-	rc = stop_machine(pseries_lpar_resize_hpt_commit, &state, NULL);
+	rc = stop_machine_cpuslocked(pseries_lpar_resize_hpt_commit,
+				     &state, NULL);
 
 	t2 = ktime_get();
 

2.20.1

From: Steven Rostedt (VMware) <[email protected]>

[ Upstream commit e0d2615856b2046c2e8d5bfd6933f37f69703b0b ]

If the re-allocation of tep->cmdlines succeeds, then the previous allocation of tep->cmdlines will be freed. If we later fail in add_new_comm(), we must not free cmdlines, and also should assign tep->cmdlines to the new allocation. Otherwise when freeing tep, the tep->cmdlines will be pointing to garbage.

---
 tools/lib/traceevent/event-parse.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

index 7989dd6289e7a..8211e8010e09b 100644
@@ -268,10 +268,10 @@ static int add_new_comm(struct pevent *pevent, const char *comm, int pid)
 		errno = ENOMEM;
 		return -1;
 	}
+	pevent->cmdlines = cmdlines;
 
 	cmdlines[pevent->cmdline_count].comm = strdup(comm);
 	if (!cmdlines[pevent->cmdline_count].comm) {
-		free(cmdlines);
 		errno = ENOMEM;
 		return -1;
 	}

@@ -282,7 +282,6 @@ static int add_new_comm(struct pevent *pevent, const char *comm, int pid)
 		pevent->cmdline_count++;
 
 	qsort(cmdlines, pevent->cmdline_count, sizeof(*cmdlines), cmdline_cmp);
-	pevent->cmdlines = cmdlines;
 
 	return 0;
 }

2.20.1

From: Balasubramani Vivekanandan <[email protected]>

[ Upstream commit b9023b91dd020ad7e093baa5122b6968c48cc9e0 ]

When a cpu requests broadcasting, before starting the tick broadcast hrtimer, bc_set_next() checks if the timer callback (bc_handler) is active using hrtimer_try_to_cancel(). But hrtimer_try_to_cancel() does not provide the required synchronization when the callback is active on other core.

The callback could have already executed tick_handle_oneshot_broadcast() and could have also returned. But still there is a small time window where the hrtimer_try_to_cancel() returns -1. In that case bc_set_next() returns without doing anything, but the next_event of the tick broadcast clock device is already set to a timeout value.

In the race condition diagram below, CPU #1 is running the timer callback and CPU #2 is entering idle state and so calls bc_set_next().

In the worst case, the next_event will contain an expiry time, but the hrtimer will not be started which happens when the racing callback returns HRTIMER_NORESTART. The hrtimer might never recover if all further requests from the CPUs to subscribe to tick broadcast have timeout greater than the next_event of tick broadcast clock device. This leads to cascading of failures and finally noticed as rcu stall warnings

Here is a depiction of the race condition

CPU #1 (Running timer callback) CPU #2 (Enter idle
and subscribe to tick broadcast)
--------------------- ---------------------

__run_hrtimer() tick_broadcast_enter()

bc_handler() __tick_broadcast_oneshot_control()

    tick_handle_oneshot_broadcast()

      raw_spin_lock(&tick_broadcast_lock);

      dev->next_event = KTIME_MAX;                  //wait for tick_broadcast_lock
      //next_event for tick broadcast clock
      set to KTIME_MAX since no other cores
      subscribed to tick broadcasting

      raw_spin_unlock(&tick_broadcast_lock);

    if (dev->next_event == KTIME_MAX)
      return HRTIMER_NORESTART
    // callback function exits without
       restarting the hrtimer                      //tick_broadcast_lock acquired
                                                   raw_spin_lock(&tick_broadcast_lock);

                                                   tick_broadcast_set_event()

                                                     clockevents_program_event()

                                                       dev->next_event = expires;

                                                       bc_set_next()

                                                         hrtimer_try_to_cancel()
                                                         //returns -1 since the timer
                                                         callback is active. Exits without
                                                         restarting the timer

cpu_base->running = NULL;

The comment that hrtimer cannot be armed from within the callback is wrong. It is fine to start the hrtimer from within the callback. Also it is safe to start the hrtimer from the enter/exit idle code while the broadcast handler is active. The enter/exit idle code and the broadcast handler are synchronized using tick_broadcast_lock. So there is no need for the existing try to cancel logic. All this can be removed which will eliminate the race condition as well.

---
 kernel/time/tick-broadcast-hrtimer.c | 57 ++++++++++++++--------------
 1 file changed, 29 insertions(+), 28 deletions(-)

index 58045eb976c38..c750c80570e88 100644
@@ -44,34 +44,39 @@ static int bc_shutdown(struct clock_event_device *evt)
  */
 static int bc_set_next(ktime_t expires, struct clock_event_device *bc)
 {
-	int bc_moved;
 	/*
-	 * We try to cancel the timer first. If the callback is on
-	 * flight on some other cpu then we let it handle it. If we
-	 * were able to cancel the timer nothing can rearm it as we
-	 * own broadcast_lock.
+	 * This is called either from enter/exit idle code or from the
+	 * broadcast handler. In all cases tick_broadcast_lock is held.
 	 *
-	 * However we can also be called from the event handler of
-	 * ce_broadcast_hrtimer itself when it expires. We cannot
-	 * restart the timer because we are in the callback, but we
-	 * can set the expiry time and let the callback return
-	 * HRTIMER_RESTART.
+	 * hrtimer_cancel() cannot be called here neither from the
+	 * broadcast handler nor from the enter/exit idle code. The idle
+	 * code can run into the problem described in bc_shutdown() and the
+	 * broadcast handler cannot wait for itself to complete for obvious
+	 * reasons.
 	 *
-	 * Since we are in the idle loop at this point and because
-	 * hrtimer_{start/cancel} functions call into tracing,
-	 * calls to these functions must be bound within RCU_NONIDLE.
+	 * Each caller tries to arm the hrtimer on its own CPU, but if the
+	 * hrtimer callbback function is currently running, then
+	 * hrtimer_start() cannot move it and the timer stays on the CPU on
+	 * which it is assigned at the moment.
+	 *
+	 * As this can be called from idle code, the hrtimer_start()
+	 * invocation has to be wrapped with RCU_NONIDLE() as
+	 * hrtimer_start() can call into tracing.
 	 */
-	RCU_NONIDLE({
-			bc_moved = hrtimer_try_to_cancel(&bctimer) >= 0;
-			if (bc_moved)
-				hrtimer_start(&bctimer, expires,
-					      HRTIMER_MODE_ABS_PINNED);});
-	if (bc_moved) {
-		/* Bind the "device" to the cpu */
-		bc->bound_on = smp_processor_id();
-	} else if (bc->bound_on == smp_processor_id()) {
-		hrtimer_set_expires(&bctimer, expires);
-	}
+	RCU_NONIDLE( {
+		hrtimer_start(&bctimer, expires, HRTIMER_MODE_ABS_PINNED);
+		/*
+		 * The core tick broadcast mode expects bc->bound_on to be set
+		 * correctly to prevent a CPU which has the broadcast hrtimer
+		 * armed from going deep idle.
+		 *
+		 * As tick_broadcast_lock is held, nothing can change the cpu
+		 * base which was just established in hrtimer_start() above. So
+		 * the below access is safe even without holding the hrtimer
+		 * base lock.
+		 */
+		bc->bound_on = bctimer.base->cpu_base->cpu;
+	} );
 	return 0;
 }
 

@@ -97,10 +102,6 @@ static enum hrtimer_restart bc_handler(struct hrtimer *t)
 {
 	ce_broadcast_hrtimer.event_handler(&ce_broadcast_hrtimer);
 
-	if (clockevent_state_oneshot(&ce_broadcast_hrtimer))
-		if (ce_broadcast_hrtimer.next_event != KTIME_MAX)
-			return HRTIMER_RESTART;
-
 	return HRTIMER_NORESTART;
 }
 

2.20.1

From: Srikar Dronamraju <[email protected]>

[ Upstream commit 443f2d5ba13d65ccfd879460f77941875159d154 ]

Observe a segmentation fault when 'perf stat' is asked to repeat forever with the interval option.

Without fix:

  # perf stat -r 0 -I 5000 -e cycles -a sleep 10
  #           time             counts unit events
       5.000211692  3,13,89,82,34,157      cycles
      10.000380119  1,53,98,52,22,294      cycles
      10.040467280       17,16,79,265      cycles
  Segmentation fault

This problem was only observed when we use forever option aka -r 0 and works with limited repeats. Calling print_counter with ts being set to NULL, is not a correct option when interval is set. Hence avoid print_counter(NULL,..) if interval is set.

With fix:

  # perf stat -r 0 -I 5000 -e cycles -a sleep 10
   #           time             counts unit events
       5.019866622  3,15,14,43,08,697      cycles
      10.039865756  3,15,16,31,95,261      cycles
      10.059950628     1,26,05,47,158      cycles
       5.009902655  3,14,52,62,33,932      cycles
      10.019880228  3,14,52,22,89,154      cycles
      10.030543876       66,90,18,333      cycles
       5.009848281  3,14,51,98,25,437      cycles
      10.029854402  3,15,14,93,04,918      cycles
       5.009834177  3,14,51,95,92,316      cycles

Committer notes:

Did the 'git bisect' to find the cset introducing the problem to add the Fixes tag below, and at that time the problem reproduced as:

(gdb) run stat -r0 -I500 sleep 1
<SNIP>
Program received signal SIGSEGV, Segmentation fault.
print_interval (prefix=prefix@entry=0x7fffffffc8d0 "", ts=ts@entry=0x0) at builtin-stat.c:866
866 sprintf(prefix, "%6lu.%09lu%s", ts->tv_sec, ts->tv_nsec, csv_sep);
(gdb) bt

  #0  print_interval (prefix=prefix@entry=0x7fffffffc8d0 "", ts=ts@entry=0x0) at builtin-stat.c:866
  #1  0x000000000041860a in print_counters (ts=ts@entry=0x0, argc=argc@entry=2, argv=argv@entry=0x7fffffffd640) at builtin-stat.c:938
  #2  0x0000000000419a7f in cmd_stat (argc=2, argv=0x7fffffffd640, prefix=<optimized out>) at builtin-stat.c:1411
  #3  0x000000000045c65a in run_builtin (p=p@entry=0x6291b8 <commands+216>, argc=argc@entry=5, argv=argv@entry=0x7fffffffd640) at perf.c:370
  #4  0x000000000045c893 in handle_internal_command (argc=5, argv=0x7fffffffd640) at perf.c:429
  #5  0x000000000045c8f1 in run_argv (argcp=argcp@entry=0x7fffffffd4ac, argv=argv@entry=0x7fffffffd4a0) at perf.c:473
  #6  0x000000000045cac9 in main (argc=<optimized out>, argv=<optimized out>) at perf.c:588

(gdb)

Mostly the same as just before this patch:

Program received signal SIGSEGV, Segmentation fault.
0x00000000005874a7 in print_interval (config=0xa1f2a0 <stat_config>, evlist=0xbc9b90, prefix=0x7fffffffd1c0 "`", ts=0x0) at util/stat-display.c:964
964 sprintf(prefix, "%6lu.%09lu%s", ts->tv_sec, ts->tv_nsec, config->csv_sep);
(gdb) bt
#0 0x00000000005874a7 in print_interval (config=0xa1f2a0 <stat_config>, evlist=0xbc9b90, prefix=0x7fffffffd1c0 "`", ts=0x0) at util/stat-display.c:964 #1 0x0000000000588047 in perf_evlist__print_counters (evlist=0xbc9b90, config=0xa1f2a0 <stat_config>, _target=0xa1f0c0 <target>, ts=0x0, argc=2, argv=0x7fffffffd670)

      at util/stat-display.c:1172
  #2  0x000000000045390f in print_counters (ts=0x0, argc=2, argv=0x7fffffffd670) at builtin-stat.c:656
  #3  0x0000000000456bb5 in cmd_stat (argc=2, argv=0x7fffffffd670) at builtin-stat.c:1960
  #4  0x00000000004dd2e0 in run_builtin (p=0xa30e00 <commands+288>, argc=5, argv=0x7fffffffd670) at perf.c:310
  #5  0x00000000004dd54d in handle_internal_command (argc=5, argv=0x7fffffffd670) at perf.c:362
  #6  0x00000000004dd694 in run_argv (argcp=0x7fffffffd4cc, argv=0x7fffffffd4c0) at perf.c:406
  #7  0x00000000004dda11 in main (argc=5, argv=0x7fffffffd670) at perf.c:531
  (gdb)

---
 tools/perf/builtin-stat.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

index 52486c90ab936..b6c1c9939c2f1 100644
@@ -2770,7 +2770,7 @@ int cmd_stat(int argc, const char **argv)
 				run_idx + 1);
 
 		status = run_perf_stat(argc, argv);
-		if (forever && status != -1) {
+		if (forever && status != -1 && !interval) {
 			print_counters(NULL, argc, argv);
 			perf_stat__reset_stats();
 		}

2.20.1

From: Srikar Dronamraju <[email protected]>

[ Upstream commit b63fd11cced17fcb8e133def29001b0f6aaa5e06 ]

When using 'perf stat' with repeat and interval option, it shows wrong values for events.

The wrong values will be shown for the first interval on the second and subsequent repetitions.

Without the fix:

  # perf stat -r 3 -I 2000 -e faults -e sched:sched_switch -a sleep 5

     2.000282489                 53      faults
     2.000282489                513      sched:sched_switch
     4.005478208              3,721      faults
     4.005478208              2,666      sched:sched_switch
     5.025470933                395      faults
     5.025470933              1,307      sched:sched_switch
     2.009602825 1,84,46,74,40,73,70,95,47,520      faults 		<------
     2.009602825 1,84,46,74,40,73,70,95,49,568      sched:sched_switch  <------
     4.019612206              4,730      faults
     4.019612206              2,746      sched:sched_switch
     5.039615484              3,953      faults
     5.039615484              1,496      sched:sched_switch
     2.000274620 1,84,46,74,40,73,70,95,47,520      faults		<------
     2.000274620 1,84,46,74,40,73,70,95,47,520      sched:sched_switch	<------
     4.000480342              4,282      faults
     4.000480342              2,303      sched:sched_switch
     5.000916811              1,322      faults
     5.000916811              1,064      sched:sched_switch

#

prev_raw_counts is allocated when using intervals. This is used when calculating the difference in the counts of events when using interval.

The current counts are stored in prev_raw_counts to calculate the differences in the next iteration.

On the first interval of the second and subsequent repetitions, prev_raw_counts would be the values stored in the last interval of the previous repetitions, while the current counts will only be for the first interval of the current repetition.

Hence there is a possibility of events showing up as big number.

Fix this by resetting prev_raw_counts whenever perf stat repeats the command.

With the fix:

  # perf stat -r 3 -I 2000 -e faults -e sched:sched_switch -a sleep 5

     2.019349347              2,597      faults
     2.019349347              2,753      sched:sched_switch
     4.019577372              3,098      faults
     4.019577372              2,532      sched:sched_switch
     5.019415481              1,879      faults
     5.019415481              1,356      sched:sched_switch
     2.000178813              8,468      faults
     2.000178813              2,254      sched:sched_switch
     4.000404621              7,440      faults
     4.000404621              1,266      sched:sched_switch
     5.040196079              2,458      faults
     5.040196079                556      sched:sched_switch
     2.000191939              6,870      faults
     2.000191939              1,170      sched:sched_switch
     4.000414103                541      faults
     4.000414103                902      sched:sched_switch
     5.000809863                450      faults
     5.000809863                364      sched:sched_switch

#

Committer notes:

This was broken since the cset introducing the --interval feature, i.e. --repeat + --interval wasn't tested at that point, add the Fixes tag so that automatic scripts can pick this up.

---
 tools/perf/builtin-stat.c |  3 +++
 tools/perf/util/stat.c    | 17 +++++++++++++++++
 tools/perf/util/stat.h    |  1 +
 3 files changed, 21 insertions(+)

index b6c1c9939c2f1..0801e0ffba4ae 100644
@@ -2769,6 +2769,9 @@ int cmd_stat(int argc, const char **argv)
 			fprintf(output, "[ perf stat: executing run #%d ... ]\n",
 				run_idx + 1);
 
+		if (run_idx != 0)
+			perf_evlist__reset_prev_raw_counts(evsel_list);
+
 		status = run_perf_stat(argc, argv);
 		if (forever && status != -1 && !interval) {
 			print_counters(NULL, argc, argv);

index c9bae5fb8b479..d028c2786802e 100644
@@ -154,6 +154,15 @@ static void perf_evsel__free_prev_raw_counts(struct perf_evsel *evsel)
 	evsel->prev_raw_counts = NULL;
 }
 
+static void perf_evsel__reset_prev_raw_counts(struct perf_evsel *evsel)
+{
+	if (evsel->prev_raw_counts) {
+		evsel->prev_raw_counts->aggr.val = 0;
+		evsel->prev_raw_counts->aggr.ena = 0;
+		evsel->prev_raw_counts->aggr.run = 0;
+       }
+}
+
 static int perf_evsel__alloc_stats(struct perf_evsel *evsel, bool alloc_raw)
 {
 	int ncpus = perf_evsel__nr_cpus(evsel);

@@ -204,6 +213,14 @@ void perf_evlist__reset_stats(struct perf_evlist *evlist)
 	}
 }
 
+void perf_evlist__reset_prev_raw_counts(struct perf_evlist *evlist)
+{
+	struct perf_evsel *evsel;
+
+	evlist__for_each_entry(evlist, evsel)
+		perf_evsel__reset_prev_raw_counts(evsel);
+}
+
 static void zero_per_pkg(struct perf_evsel *counter)
 {
 	if (counter->per_pkg_mask)

index 96326b1f94438..bdfbed8e2df28 100644
@@ -100,6 +100,7 @@ void perf_stat__collect_metric_expr(struct perf_evlist *);
 int perf_evlist__alloc_stats(struct perf_evlist *evlist, bool alloc_raw);
 void perf_evlist__free_stats(struct perf_evlist *evlist);
 void perf_evlist__reset_stats(struct perf_evlist *evlist);
+void perf_evlist__reset_prev_raw_counts(struct perf_evlist *evlist);
 
 int perf_stat_process_counter(struct perf_stat_config *config,
 			      struct perf_evsel *counter);

2.20.1

From: Chris Wilson <[email protected]>

[ Upstream commit cb6d7c7dc7ff8cace666ddec66334117a6068ce2 ]

set_page_dirty says:

For pages with a mapping this should be done under the page lock for the benefit of asynchronous memory errors who prefer a consistent dirty state. This rule can be broken in some special cases, but should be better not to.

Under those rules, it is only safe for us to use the plain set_page_dirty calls for shmemfs/anonymous memory. Userptr may be used with real mappings and so needs to use the locked version (set_page_dirty_lock).

Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=203317

---
 drivers/gpu/drm/i915/i915_gem_userptr.c | 10 +++++++++-
 1 file changed, 9 insertions(+), 1 deletion(-)

index 05ae8c4a8a1b6..9760b67dab28b 100644
@@ -691,7 +691,15 @@ i915_gem_userptr_put_pages(struct drm_i915_gem_object *obj,
 
 	for_each_sgt_page(page, sgt_iter, pages) {
 		if (obj->mm.dirty)
-			set_page_dirty(page);
+			/*
+			 * As this may not be anonymous memory (e.g. shmem)
+			 * but exist on a real mapping, we have to lock
+			 * the page in order to dirty it -- holding
+			 * the page reference is not sufficient to
+			 * prevent the inode from being truncated.
+			 * Play safe and take the lock.
+			 */
+			set_page_dirty_lock(page);
 
 		mark_page_accessed(page);
 		put_page(page);

2.20.1

From: Eric Sandeen <[email protected]>

commit cc3a7bfe62b947b423fcb2cfe89fcba92bf48fa3 upstream.

Today, put_compat_statfs64() disallows nearly any field value over 2^32 if f_bsize is only 32 bits, but that makes no sense. compat_statfs64 is there for the explicit purpose of providing 64-bit fields for f_files, f_ffree, etc. And f_bsize is always only 32 bits.

As a result, 32-bit userspace gets -EOVERFLOW for i.e. large file counts even with -D_FILE_OFFSET_BITS=64 set.

In reality, only f_bsize and f_frsize can legitimately overflow (fields like f_type and f_namelen should never be large), so test only those fields.

This bug was discussed at length some time ago, and this is the proposal Al suggested at https://lkml.org/lkml/2018/8/6/640. It seemed to get dropped amid the discussion of other related changes, but this part seems obviously correct on its own, so I've picked it up and sent it, for expediency.

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

@@ -304,19 +304,10 @@ COMPAT_SYSCALL_DEFINE2(fstatfs, unsigned
 static int put_compat_statfs64(struct compat_statfs64 __user *ubuf, struct kstatfs *kbuf)
 {
 	struct compat_statfs64 buf;
-	if (sizeof(ubuf->f_bsize) == 4) {
-		if ((kbuf->f_type | kbuf->f_bsize | kbuf->f_namelen |
-		     kbuf->f_frsize | kbuf->f_flags) & 0xffffffff00000000ULL)
-			return -EOVERFLOW;
-		/* f_files and f_ffree may be -1; it's okay
-		 * to stuff that into 32 bits */
-		if (kbuf->f_files != 0xffffffffffffffffULL
-		 && (kbuf->f_files & 0xffffffff00000000ULL))
-			return -EOVERFLOW;
-		if (kbuf->f_ffree != 0xffffffffffffffffULL
-		 && (kbuf->f_ffree & 0xffffffff00000000ULL))
-			return -EOVERFLOW;
-	}
+
+	if ((kbuf->f_bsize | kbuf->f_frsize) & 0xffffffff00000000ULL)
+		return -EOVERFLOW;
+
 	memset(&buf, 0, sizeof(struct compat_statfs64));
 	buf.f_type = kbuf->f_type;
 	buf.f_bsize = kbuf->f_bsize;

From: Andrew Murray <[email protected]>

commit 1004ce4c255fc3eb3ad9145ddd53547d1b7ce327 upstream.

Synchronization is recommended before disabling the trace registers to prevent any start or stop points being speculative at the point of disabling the unit (section 7.3.77 of ARM IHI 0064D).

Synchronization is also recommended after programming the trace registers to ensure all updates are committed prior to normal code resuming (section 4.3.7 of ARM IHI 0064D).

Let's ensure these syncronization points are present in the code and clearly commented.

Note that we could rely on the barriers in CS_LOCK and coresight_disclaim_device_unlocked or the context switch to user space - however coresight may be of use in the kernel.

On armv8 the mb macro is defined as dsb(sy) - Given that the etm4x is only used on armv8 let's directly use dsb(sy) instead of mb(). This removes some ambiguity and makes it easier to correlate the code with the TRM.

---
 drivers/hwtracing/coresight/coresight-etm4x.c |   14 ++++++++++++--
 1 file changed, 12 insertions(+), 2 deletions(-)

@@ -181,6 +181,12 @@ static void etm4_enable_hw(void *info)
 	if (coresight_timeout(drvdata->base, TRCSTATR, TRCSTATR_IDLE_BIT, 0))
 		dev_err(drvdata->dev,
 			"timeout while waiting for Idle Trace Status\n");
+	/*
+	 * As recommended by section 4.3.7 ("Synchronization when using the
+	 * memory-mapped interface") of ARM IHI 0064D
+	 */
+	dsb(sy);
+	isb();
 
 	CS_LOCK(drvdata->base);
 

@@ -331,8 +337,12 @@ static void etm4_disable_hw(void *info)
 	/* EN, bit[0] Trace unit enable bit */
 	control &= ~0x1;
 
-	/* make sure everything completes before disabling */
-	mb();
+	/*
+	 * Make sure everything completes before disabling, as recommended
+	 * by section 7.3.77 ("TRCVICTLR, ViewInst Main Control Register,
+	 * SSTATUS") of ARM IHI 0064D
+	 */
+	dsb(sy);
 	isb();
 	writel_relaxed(control, drvdata->base + TRCPRGCTLR);
 

From: Johannes Berg <[email protected]>

commit 0f3b07f027f87a38ebe5c436490095df762819be upstream.

Rather than always iterating elements from frames with pure u8 pointers, add a type "struct element" that encapsulates the id/datalen/data format of them.

Then, add the element iteration macros
* for_each_element
* for_each_element_id
* for_each_element_extid

which take, as their first 'argument', such a structure and iterate through a given u8 array interpreting it as elements.

While at it and since we'll need it, also add
* for_each_subelement
* for_each_subelement_id
* for_each_subelement_extid