| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
The ff00::/8 multicast route is created without specifying the fc_protocol
field, so the default RTPROT_BOOT value is used:
$ ip -6 -d route
unicast ::1 dev lo proto kernel scope global metric 256 pref medium
unicast fe80::/64 dev eth0 proto kernel scope global metric 256 pref medium
unicast ff00::/8 dev eth0 proto boot scope global metric 256 pref medium
As the documentation says, this value identifies routes installed during
boot, but the route is created when interface is set up.
Change the value to RTPROT_KERNEL which is a better value.
Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
Signed-off-by: Matteo Croce <[email protected]>
Signed-off-by: Jakub Kicinski <[email protected]>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/jberg/mac80211
Johannes Berg says:
====================
Various fixes:
* kernel-doc parsing fixes
* incorrect debugfs string checks
* locking fix in regulatory
* some encryption-related fixes
* tag 'mac80211-for-net-2021-01-18.2' of git://git.kernel.org/pub/scm/linux/kernel/git/jberg/mac80211:
mac80211: check if atf has been disabled in __ieee80211_schedule_txq
mac80211: do not drop tx nulldata packets on encrypted links
mac80211: fix encryption key selection for 802.3 xmit
mac80211: fix fast-rx encryption check
mac80211: fix incorrect strlen of .write in debugfs
cfg80211: fix a kerneldoc markup
cfg80211: Save the regulatory domain with a lock
cfg80211/mac80211: fix kernel-doc for SAR APIs
====================
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>
|
|
Since main() does not return a value explicitly, the
return values from FAIL_IF() conditions are ignored
and the tests can still pass irrespective of failures.
This makes sure that we always explicitly return the
correct test exit status.
Fixes: 1addb6444791 ("selftests/powerpc: Add test for execute-disabled pkeys")
Fixes: c27f2fd1705a ("selftests/powerpc: Add test for pkey siginfo verification")
Reported-by: Eirik Fuller <[email protected]>
Signed-off-by: Sandipan Das <[email protected]>
Signed-off-by: Michael Ellerman <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
|
|
mv88e6xxx_port_vlan_join checks whether the VTU already contains an
entry for the given vid (via mv88e6xxx_vtu_getnext), and if so, merely
changes the relevant .member[] element and loads the updated entry
into the VTU.
However, at least for the mv88e6250, the on-stack struct
mv88e6xxx_vtu_entry vlan never has its .state[] array explicitly
initialized, neither in mv88e6xxx_port_vlan_join() nor inside the
getnext implementation. So the new entry has random garbage for the
STU bits, breaking VLAN filtering.
When the VTU entry is initially created, those bits are all zero, and
we should make sure to keep them that way when the entry is updated.
Fixes: 92307069a96c (net: dsa: mv88e6xxx: Avoid VTU corruption on 6097)
Signed-off-by: Rasmus Villemoes <[email protected]>
Reviewed-by: Florian Fainelli <[email protected]>
Reviewed-by: Tobias Waldekranz <[email protected]>
Tested-by: Tobias Waldekranz <[email protected]>
Signed-off-by: Jakub Kicinski <[email protected]>
|
|
fan control
This commit enables dual fan control for the new Lenovo P53 and P73
laptop models.
Signed-off-by: Jeannie Stevenson <[email protected]>
Link: https://lore.kernel.org/r/Pn_Xii4XYpQRFtgkf4PbNgieE89BAkHgLI1kWIq-zFudwh2A1DY5J_DJVHK06rMW_hGPHx_mPE33gd8mg9-8BxqJTaSC6hhPqAsfZlcNGH0=@protonmail.com
Signed-off-by: Hans de Goede <[email protected]>
|
|
The recently added thermal policy support makes a
hp_wmi_perform_query(0x4c, ...) call on older devices which do not
support thermal policies this causes the following warning to be
logged (seen on a HP Stream x360 Convertible PC 11):
[ 26.805305] hp_wmi: query 0x4c returned error 0x3
Error 0x3 is HPWMI_RET_UNKNOWN_COMMAND error. This commit silences
the warning for unknown-command errors, silencing the new warning.
Cc: Elia Devito <[email protected]>
Fixes: 81c93798ef3e ("platform/x86: hp-wmi: add support for thermal policy")
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Hans de Goede <[email protected]>
|
|
The blamed commit was too aggressive, and it made ocelot_netdevice_event
react only to network interface events emitted for the ocelot switch
ports.
In fact, only the PRECHANGEUPPER should have had that check.
When we ignore all events that are not for us, we miss the fact that the
upper of the LAG changes, and the bonding interface gets enslaved to a
bridge. This is an operation we could offload under certain conditions.
Fixes: 7afb3e575e5a ("net: mscc: ocelot: don't handle netdev events for other netdevs")
Signed-off-by: Vladimir Oltean <[email protected]>
Reviewed-by: Alexandre Belloni <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jakub Kicinski <[email protected]>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi
Pull spi fixes from Mark Brown:
"A few more bug fixes for SPI, both driver specific ones. The caching
in the Cadence driver is to avoid a deadlock trying to retrieve the
cached value later at runtime"
* tag 'spi-fix-v5.11-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/broonie/spi:
spi: cadence: cache reference clock rate during probe
spi: fsl: Fix driver breakage when SPI_CS_HIGH is not set in spi->mode
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/rppt/memblock
Pull ia64 build fix from Mike Rapoport:
"Fix an ia64 build failure caused by memory model changes"
* tag 'fixes-2021-01-18' of git://git.kernel.org/pub/scm/linux/kernel/git/rppt/memblock:
ia64: fix build failure caused by memory model changes
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6
Pull crypto fixes from Herbert Xu:
"A Kconfig dependency issue with omap-sham and a divide by zero in xor
on some platforms"
* 'linus' of git://git.kernel.org/pub/scm/linux/kernel/git/herbert/crypto-2.6:
crypto: omap-sham - Fix link error without crypto-engine
crypto: xor - Fix divide error in do_xor_speed()
|
|
https://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound into for-linus
ASoC: Fixes for v5.11
A few more fixes for v5.11, mostly around HDA jack detection, plus
a couple of updates to the MAINTAINERS entries.
|
|
THe HP Stream x360 Convertible PC 11 DSDT has the following VGBS function:
Method (VGBS, 0, Serialized)
{
If ((^^PCI0.LPCB.EC0.ROLS == Zero))
{
VBDS = Zero
}
Else
{
VBDS = Zero
}
Return (VBDS) /* \_SB_.VGBI.VBDS */
}
Which is obviously wrong, because it always returns 0 independent of the
2-in-1 being in laptop or tablet mode. This causes the intel-vbtn driver
to initially report SW_TABLET_MODE = 1 to userspace, which is known to
cause problems when the 2-in-1 is actually in laptop mode.
During earlier testing this turned out to not be a problem because the
2-in-1 would do a Notify(..., 0xCC) or Notify(..., 0xCD) soon after
the intel-vbtn driver loaded, correcting the SW_TABLET_MODE state.
Further testing however has shown that this Notify() soon after the
intel-vbtn driver loads, does not always happen. When the Notify
does not happen, then intel-vbtn reports SW_TABLET_MODE = 1 resulting in
a non-working touchpad.
IOW the tablet-mode reporting is not reliable on this device, so it
should be dropped from the allow-list, fixing the touchpad sometimes
not working.
Fixes: 8169bd3e6e19 ("platform/x86: intel-vbtn: Switch to an allow-list for SW_TABLET_MODE reporting")
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Hans de Goede <[email protected]>
|
|
As of the "arm64: expose FAR_EL1 tag bits in siginfo" patch, the address
that is passed to report_tag_fault has pointer tags in the format of 0x0X,
while KASAN uses 0xFX format (note the difference in the top 4 bits).
Fix up the pointer tag for kernel pointers in do_tag_check_fault by
setting them to the same value as bit 55. Explicitly use __untagged_addr()
instead of untagged_addr(), as the latter doesn't affect TTBR1 addresses.
Fixes: dceec3ff7807 ("arm64: expose FAR_EL1 tag bits in siginfo")
Fixes: 4291e9ee6189 ("kasan, arm64: print report from tag fault handler")
Signed-off-by: Andrey Konovalov <[email protected]>
Reviewed-by: Catalin Marinas <[email protected]>
Reviewed-by: Vincenzo Frascino <[email protected]>
Link: https://linux-review.googlesource.com/id/I9ced973866036d8679e8f4ae325de547eb969649
Link: https://lore.kernel.org/r/ff30b0afe6005fd046f9ac72bfb71822aedccd89.1610731872.git.andreyknvl@google.com
Signed-off-by: Catalin Marinas <[email protected]>
|
|
The function nvmet_execute_identify_ns() doesn't set the status if call
to nvmet_find_namespace() fails. In that case we set the status of the
request to the value return by the nvmet_copy_sgl().
Set the status to NVME_SC_INVALID_NS and adjust the code such that
request will have the right status on nvmet_find_namespace() failure.
Without this patch :-
NVME Identify Namespace 3:
nsze : 0
ncap : 0
nuse : 0
nsfeat : 0
nlbaf : 0
flbas : 0
mc : 0
dpc : 0
dps : 0
nmic : 0
rescap : 0
fpi : 0
dlfeat : 0
nawun : 0
nawupf : 0
nacwu : 0
nabsn : 0
nabo : 0
nabspf : 0
noiob : 0
nvmcap : 0
mssrl : 0
mcl : 0
msrc : 0
nsattr : 0
nvmsetid: 0
anagrpid: 0
endgid : 0
nguid : 00000000000000000000000000000000
eui64 : 0000000000000000
lbaf 0 : ms:0 lbads:0 rp:0 (in use)
With this patch-series :-
feb3b88b501e (HEAD -> nvme-5.11) nvmet: remove extra variable in identify ns
6302aa67210a nvmet: remove extra variable in id-desclist
ed57951da453 nvmet: remove extra variable in smart log nsid
be384b8c24dc nvmet: set right status on error in id-ns handler
NVMe status: INVALID_NS: The namespace or the format of that namespace is invalid(0xb)
Signed-off-by: Chaitanya Kulkarni <[email protected]>
Reviewed-by: Sagi Grimberg <[email protected]>
Signed-off-by: Christoph Hellwig <[email protected]>
|
|
Since NVMe v1.4 the Controller Memory Buffer must be explicitly enabled
by the host.
Signed-off-by: Klaus Jensen <[email protected]>
[hch: avoid a local variable and add a comment]
Signed-off-by: Christoph Hellwig <[email protected]>
|
|
Each name space has a request queue, if complete request long time,
multi request queues may have time out requests at the same time,
nvme_tcp_timeout will execute concurrently. Multi requests in different
request queues may be queued in the same tcp queue, multi
nvme_tcp_timeout may call nvme_tcp_stop_queue at the same time.
The first nvme_tcp_stop_queue will clear NVME_TCP_Q_LIVE and continue
stopping the tcp queue(cancel io_work), but the others check
NVME_TCP_Q_LIVE is already cleared, and then directly complete the
requests, complete request before the io work is completely canceled may
lead to a use-after-free condition.
Add a multex lock to serialize nvme_tcp_stop_queue.
Signed-off-by: Chao Leng <[email protected]>
Signed-off-by: Christoph Hellwig <[email protected]>
|
|
A crash happens when inject completing request long time(nearly 30s).
Each name space has a request queue, when inject completing request long
time, multi request queues may have time out requests at the same time,
nvme_rdma_timeout will execute concurrently. Multi requests in different
request queues may be queued in the same rdma queue, multi
nvme_rdma_timeout may call nvme_rdma_stop_queue at the same time.
The first nvme_rdma_timeout will clear NVME_RDMA_Q_LIVE and continue
stopping the rdma queue(drain qp), but the others check NVME_RDMA_Q_LIVE
is already cleared, and then directly complete the requests, complete
request before the qp is fully drained may lead to a use-after-free
condition.
Add a multex lock to serialize nvme_rdma_stop_queue.
Signed-off-by: Chao Leng <[email protected]>
Tested-by: Israel Rukshin <[email protected]>
Reviewed-by: Israel Rukshin <[email protected]>
Signed-off-by: Christoph Hellwig <[email protected]>
|
|
According to NVMe spec v1.4, section 8.3.1, the PRINFO bit and
the metadata size play a vital role in deteriming the host buffer size.
If PRIFNO bit is set and MS==8, the host doesn't add the metadata buffer,
instead the controller adds it.
Signed-off-by: Revanth Rajashekar <[email protected]>
Signed-off-by: Christoph Hellwig <[email protected]>
|
|
In Linux, if a driver does disable_irq() and later does enable_irq()
on its interrupt, I believe it's expecting these properties:
* If an interrupt was pending when the driver disabled then it will
still be pending after the driver re-enables.
* If an edge-triggered interrupt comes in while an interrupt is
disabled it should assert when the interrupt is re-enabled.
If you think that the above sounds a lot like the disable_irq() and
enable_irq() are supposed to be masking/unmasking the interrupt
instead of disabling/enabling it then you've made an astute
observation. Specifically when talking about interrupts, "mask"
usually means to stop posting interrupts but keep tracking them and
"disable" means to fully shut off interrupt detection. It's
unfortunate that this is so confusing, but presumably this is all the
way it is for historical reasons.
Perhaps more confusing than the above is that, even though clients of
IRQs themselves don't have a way to request mask/unmask
vs. disable/enable calls, IRQ chips themselves can implement both.
...and yet more confusing is that if an IRQ chip implements
disable/enable then they will be called when a client driver calls
disable_irq() / enable_irq().
It does feel like some of the above could be cleared up. However,
without any other core interrupt changes it should be clear that when
an IRQ chip gets a request to "disable" an IRQ that it has to treat it
like a mask of that IRQ.
In any case, after that long interlude you can see that the "unmask
and clear" can break things. Maulik tried to fix it so that we no
longer did "unmask and clear" in commit 71266d9d3936 ("pinctrl: qcom:
Move clearing pending IRQ to .irq_request_resources callback"), but it
only handled the PDC case and it had problems (it caused
sc7180-trogdor devices to fail to suspend). Let's fix.
>From my understanding the source of the phantom interrupt in the
were these two things:
1. One that could have been introduced in msm_gpio_irq_set_type()
(only for the non-PDC case).
2. Edges could have been detected when a GPIO was muxed away.
Fixing case #1 is easy. We can just add a clear in
msm_gpio_irq_set_type().
Fixing case #2 is harder. Let's use a concrete example. In
sc7180-trogdor.dtsi we configure the uart3 to have two pinctrl states,
sleep and default, and mux between the two during runtime PM and
system suspend (see geni_se_resources_{on,off}() for more
details). The difference between the sleep and default state is that
the RX pin is muxed to a GPIO during sleep and muxed to the UART
otherwise.
As per Qualcomm, when we mux the pin over to the UART function the PDC
(or the non-PDC interrupt detection logic) is still watching it /
latching edges. These edges don't cause interrupts because the
current code masks the interrupt unless we're entering suspend.
However, as soon as we enter suspend we unmask the interrupt and it's
counted as a wakeup.
Let's deal with the problem like this:
* When we mux away, we'll mask our interrupt. This isn't necessary in
the above case since the client already masked us, but it's a good
idea in general.
* When we mux back will clear any interrupts and unmask.
Fixes: 4b7618fdc7e6 ("pinctrl: qcom: Add irq_enable callback for msm gpio")
Fixes: 71266d9d3936 ("pinctrl: qcom: Move clearing pending IRQ to .irq_request_resources callback")
Signed-off-by: Douglas Anderson <[email protected]>
Reviewed-by: Maulik Shah <[email protected]>
Tested-by: Maulik Shah <[email protected]>
Reviewed-by: Stephen Boyd <[email protected]>
Link: https://lore.kernel.org/r/20210114191601.v7.4.I7cf3019783720feb57b958c95c2b684940264cd1@changeid
Signed-off-by: Linus Walleij <[email protected]>
|
|
In commit 4b7618fdc7e6 ("pinctrl: qcom: Add irq_enable callback for
msm gpio") we tried to Ack interrupts during unmask. However, that
patch forgot to check "intr_ack_high" so, presumably, it only worked
for a certain subset of SoCs.
Let's add a small accessor so we don't need to open-code the logic in
both places.
This was found by code inspection. I don't have any access to the
hardware in question nor software that needs the Ack during unmask.
Fixes: 4b7618fdc7e6 ("pinctrl: qcom: Add irq_enable callback for msm gpio")
Signed-off-by: Douglas Anderson <[email protected]>
Reviewed-by: Maulik Shah <[email protected]>
Tested-by: Maulik Shah <[email protected]>
Reviewed-by: Stephen Boyd <[email protected]>
Reviewed-by: Bjorn Andersson <[email protected]>
Link: https://lore.kernel.org/r/20210114191601.v7.3.I32d0f4e174d45363b49ab611a13c3da8f1e87d0f@changeid
Signed-off-by: Linus Walleij <[email protected]>
|
|
When the Qualcomm pinctrl driver wants to Ack an interrupt, it does a
read-modify-write on the interrupt status register. On some SoCs it
makes sure that the status bit is 1 to "Ack" and on others it makes
sure that the bit is 0 to "Ack". Presumably the first type of
interrupt controller is a "write 1 to clear" type register and the
second just let you directly set the interrupt status register.
As far as I can tell from scanning structure definitions, the
interrupt status bit is always in a register by itself. Thus with
both types of interrupt controllers it is safe to "Ack" interrupts
without doing a read-modify-write. We can do a simple write.
It should be noted that if the interrupt status bit _was_ ever in a
register with other things (like maybe status bits for other GPIOs):
a) For "write 1 clear" type controllers then read-modify-write would
be totally wrong because we'd accidentally end up clearing
interrupts we weren't looking at.
b) For "direct set" type controllers then read-modify-write would also
be wrong because someone setting one of the other bits in the
register might accidentally clear (or set) our interrupt.
I say this simply to show that the current read-modify-write doesn't
provide any sort of "future proofing" of the code. In fact (for
"write 1 clear" controllers) the new code is slightly more "future
proof" since it would allow more than one interrupt status bits to
share a register.
NOTE: this code fixes no bugs--it simply avoids an extra register
read.
Signed-off-by: Douglas Anderson <[email protected]>
Reviewed-by: Maulik Shah <[email protected]>
Tested-by: Maulik Shah <[email protected]>
Reviewed-by: Stephen Boyd <[email protected]>
Reviewed-by: Bjorn Andersson <[email protected]>
Link: https://lore.kernel.org/r/20210114191601.v7.2.I3635de080604e1feda770591c5563bd6e63dd39d@changeid
Signed-off-by: Linus Walleij <[email protected]>
|
|
There's currently a comment in the code saying function 0 is GPIO.
Instead of hardcoding it, let's add a member where an SoC can specify
it. No known SoCs use a number other than 0, but this just makes the
code clearer. NOTE: no SoC code needs to be updated since we can rely
on zero-initialization.
Signed-off-by: Douglas Anderson <[email protected]>
Reviewed-by: Stephen Boyd <[email protected]>
Reviewed-by: Maulik Shah <[email protected]>
Tested-by: Maulik Shah <[email protected]>
Reviewed-by: Bjorn Andersson <[email protected]>
Link: https://lore.kernel.org/r/20210114191601.v7.1.I3ad184e3423d8e479bc3e86f5b393abb1704a1d1@changeid
Signed-off-by: Linus Walleij <[email protected]>
|
|
If we fail to update a block group item in the loop we'll break, however
we'll do btrfs_run_delayed_refs and lose our error value in ret, and
thus not clean up properly. Fix this by only running the delayed refs
if there was no failure.
CC: [email protected] # 4.4+
Reviewed-by: Qu Wenruo <[email protected]>
Reviewed-by: Johannes Thumshirn <[email protected]>
Signed-off-by: Josef Bacik <[email protected]>
Reviewed-by: David Sterba <[email protected]>
Signed-off-by: David Sterba <[email protected]>
|
|
While testing the error paths of relocation I hit the following lockdep
splat:
======================================================
WARNING: possible circular locking dependency detected
5.10.0-rc6+ #217 Not tainted
------------------------------------------------------
mount/779 is trying to acquire lock:
ffffa0e676945418 (&fs_info->balance_mutex){+.+.}-{3:3}, at: btrfs_recover_balance+0x2f0/0x340
but task is already holding lock:
ffffa0e60ee31da8 (btrfs-root-00){++++}-{3:3}, at: __btrfs_tree_read_lock+0x27/0x100
which lock already depends on the new lock.
the existing dependency chain (in reverse order) is:
-> #2 (btrfs-root-00){++++}-{3:3}:
down_read_nested+0x43/0x130
__btrfs_tree_read_lock+0x27/0x100
btrfs_read_lock_root_node+0x31/0x40
btrfs_search_slot+0x462/0x8f0
btrfs_update_root+0x55/0x2b0
btrfs_drop_snapshot+0x398/0x750
clean_dirty_subvols+0xdf/0x120
btrfs_recover_relocation+0x534/0x5a0
btrfs_start_pre_rw_mount+0xcb/0x170
open_ctree+0x151f/0x1726
btrfs_mount_root.cold+0x12/0xea
legacy_get_tree+0x30/0x50
vfs_get_tree+0x28/0xc0
vfs_kern_mount.part.0+0x71/0xb0
btrfs_mount+0x10d/0x380
legacy_get_tree+0x30/0x50
vfs_get_tree+0x28/0xc0
path_mount+0x433/0xc10
__x64_sys_mount+0xe3/0x120
do_syscall_64+0x33/0x40
entry_SYSCALL_64_after_hwframe+0x44/0xa9
-> #1 (sb_internal#2){.+.+}-{0:0}:
start_transaction+0x444/0x700
insert_balance_item.isra.0+0x37/0x320
btrfs_balance+0x354/0xf40
btrfs_ioctl_balance+0x2cf/0x380
__x64_sys_ioctl+0x83/0xb0
do_syscall_64+0x33/0x40
entry_SYSCALL_64_after_hwframe+0x44/0xa9
-> #0 (&fs_info->balance_mutex){+.+.}-{3:3}:
__lock_acquire+0x1120/0x1e10
lock_acquire+0x116/0x370
__mutex_lock+0x7e/0x7b0
btrfs_recover_balance+0x2f0/0x340
open_ctree+0x1095/0x1726
btrfs_mount_root.cold+0x12/0xea
legacy_get_tree+0x30/0x50
vfs_get_tree+0x28/0xc0
vfs_kern_mount.part.0+0x71/0xb0
btrfs_mount+0x10d/0x380
legacy_get_tree+0x30/0x50
vfs_get_tree+0x28/0xc0
path_mount+0x433/0xc10
__x64_sys_mount+0xe3/0x120
do_syscall_64+0x33/0x40
entry_SYSCALL_64_after_hwframe+0x44/0xa9
other info that might help us debug this:
Chain exists of:
&fs_info->balance_mutex --> sb_internal#2 --> btrfs-root-00
Possible unsafe locking scenario:
CPU0 CPU1
---- ----
lock(btrfs-root-00);
lock(sb_internal#2);
lock(btrfs-root-00);
lock(&fs_info->balance_mutex);
*** DEADLOCK ***
2 locks held by mount/779:
#0: ffffa0e60dc040e0 (&type->s_umount_key#47/1){+.+.}-{3:3}, at: alloc_super+0xb5/0x380
#1: ffffa0e60ee31da8 (btrfs-root-00){++++}-{3:3}, at: __btrfs_tree_read_lock+0x27/0x100
stack backtrace:
CPU: 0 PID: 779 Comm: mount Not tainted 5.10.0-rc6+ #217
Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.13.0-2.fc32 04/01/2014
Call Trace:
dump_stack+0x8b/0xb0
check_noncircular+0xcf/0xf0
? trace_call_bpf+0x139/0x260
__lock_acquire+0x1120/0x1e10
lock_acquire+0x116/0x370
? btrfs_recover_balance+0x2f0/0x340
__mutex_lock+0x7e/0x7b0
? btrfs_recover_balance+0x2f0/0x340
? btrfs_recover_balance+0x2f0/0x340
? rcu_read_lock_sched_held+0x3f/0x80
? kmem_cache_alloc_trace+0x2c4/0x2f0
? btrfs_get_64+0x5e/0x100
btrfs_recover_balance+0x2f0/0x340
open_ctree+0x1095/0x1726
btrfs_mount_root.cold+0x12/0xea
? rcu_read_lock_sched_held+0x3f/0x80
legacy_get_tree+0x30/0x50
vfs_get_tree+0x28/0xc0
vfs_kern_mount.part.0+0x71/0xb0
btrfs_mount+0x10d/0x380
? __kmalloc_track_caller+0x2f2/0x320
legacy_get_tree+0x30/0x50
vfs_get_tree+0x28/0xc0
? capable+0x3a/0x60
path_mount+0x433/0xc10
__x64_sys_mount+0xe3/0x120
do_syscall_64+0x33/0x40
entry_SYSCALL_64_after_hwframe+0x44/0xa9
This is straightforward to fix, simply release the path before we setup
the balance_ctl.
CC: [email protected] # 4.4+
Reviewed-by: Qu Wenruo <[email protected]>
Reviewed-by: Johannes Thumshirn <[email protected]>
Signed-off-by: Josef Bacik <[email protected]>
Reviewed-by: David Sterba <[email protected]>
Signed-off-by: David Sterba <[email protected]>
|
|
Zygo reported the following KASAN splat:
BUG: KASAN: use-after-free in btrfs_backref_cleanup_node+0x18a/0x420
Read of size 8 at addr ffff888112402950 by task btrfs/28836
CPU: 0 PID: 28836 Comm: btrfs Tainted: G W 5.10.0-e35f27394290-for-next+ #23
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.12.0-1 04/01/2014
Call Trace:
dump_stack+0xbc/0xf9
? btrfs_backref_cleanup_node+0x18a/0x420
print_address_description.constprop.8+0x21/0x210
? record_print_text.cold.34+0x11/0x11
? btrfs_backref_cleanup_node+0x18a/0x420
? btrfs_backref_cleanup_node+0x18a/0x420
kasan_report.cold.10+0x20/0x37
? btrfs_backref_cleanup_node+0x18a/0x420
__asan_load8+0x69/0x90
btrfs_backref_cleanup_node+0x18a/0x420
btrfs_backref_release_cache+0x83/0x1b0
relocate_block_group+0x394/0x780
? merge_reloc_roots+0x4a0/0x4a0
btrfs_relocate_block_group+0x26e/0x4c0
btrfs_relocate_chunk+0x52/0x120
btrfs_balance+0xe2e/0x1900
? check_flags.part.50+0x6c/0x1e0
? btrfs_relocate_chunk+0x120/0x120
? kmem_cache_alloc_trace+0xa06/0xcb0
? _copy_from_user+0x83/0xc0
btrfs_ioctl_balance+0x3a7/0x460
btrfs_ioctl+0x24c8/0x4360
? __kasan_check_read+0x11/0x20
? check_chain_key+0x1f4/0x2f0
? __asan_loadN+0xf/0x20
? btrfs_ioctl_get_supported_features+0x30/0x30
? kvm_sched_clock_read+0x18/0x30
? check_chain_key+0x1f4/0x2f0
? lock_downgrade+0x3f0/0x3f0
? handle_mm_fault+0xad6/0x2150
? do_vfs_ioctl+0xfc/0x9d0
? ioctl_file_clone+0xe0/0xe0
? check_flags.part.50+0x6c/0x1e0
? check_flags.part.50+0x6c/0x1e0
? check_flags+0x26/0x30
? lock_is_held_type+0xc3/0xf0
? syscall_enter_from_user_mode+0x1b/0x60
? do_syscall_64+0x13/0x80
? rcu_read_lock_sched_held+0xa1/0xd0
? __kasan_check_read+0x11/0x20
? __fget_light+0xae/0x110
__x64_sys_ioctl+0xc3/0x100
do_syscall_64+0x37/0x80
entry_SYSCALL_64_after_hwframe+0x44/0xa9
RIP: 0033:0x7f4c4bdfe427
Allocated by task 28836:
kasan_save_stack+0x21/0x50
__kasan_kmalloc.constprop.18+0xbe/0xd0
kasan_kmalloc+0x9/0x10
kmem_cache_alloc_trace+0x410/0xcb0
btrfs_backref_alloc_node+0x46/0xf0
btrfs_backref_add_tree_node+0x60d/0x11d0
build_backref_tree+0xc5/0x700
relocate_tree_blocks+0x2be/0xb90
relocate_block_group+0x2eb/0x780
btrfs_relocate_block_group+0x26e/0x4c0
btrfs_relocate_chunk+0x52/0x120
btrfs_balance+0xe2e/0x1900
btrfs_ioctl_balance+0x3a7/0x460
btrfs_ioctl+0x24c8/0x4360
__x64_sys_ioctl+0xc3/0x100
do_syscall_64+0x37/0x80
entry_SYSCALL_64_after_hwframe+0x44/0xa9
Freed by task 28836:
kasan_save_stack+0x21/0x50
kasan_set_track+0x20/0x30
kasan_set_free_info+0x1f/0x30
__kasan_slab_free+0xf3/0x140
kasan_slab_free+0xe/0x10
kfree+0xde/0x200
btrfs_backref_error_cleanup+0x452/0x530
build_backref_tree+0x1a5/0x700
relocate_tree_blocks+0x2be/0xb90
relocate_block_group+0x2eb/0x780
btrfs_relocate_block_group+0x26e/0x4c0
btrfs_relocate_chunk+0x52/0x120
btrfs_balance+0xe2e/0x1900
btrfs_ioctl_balance+0x3a7/0x460
btrfs_ioctl+0x24c8/0x4360
__x64_sys_ioctl+0xc3/0x100
do_syscall_64+0x37/0x80
entry_SYSCALL_64_after_hwframe+0x44/0xa9
This occurred because we freed our backref node in
btrfs_backref_error_cleanup(), but then tried to free it again in
btrfs_backref_release_cache(). This is because
btrfs_backref_release_cache() will cycle through all of the
cache->leaves nodes and free them up. However
btrfs_backref_error_cleanup() freed the backref node with
btrfs_backref_free_node(), which simply kfree()d the backref node
without unlinking it from the cache. Change this to a
btrfs_backref_drop_node(), which does the appropriate cleanup and
removes the node from the cache->leaves list, so when we go to free the
remaining cache we don't trip over items we've already dropped.
Fixes: 75bfb9aff45e ("Btrfs: cleanup error handling in build_backref_tree")
CC: [email protected] # 4.4+
Signed-off-by: Josef Bacik <[email protected]>
Reviewed-by: David Sterba <[email protected]>
Signed-off-by: David Sterba <[email protected]>
|
|
This was partially fixed by f3e3d9cc3525 ("btrfs: avoid possible signal
interruption of btrfs_drop_snapshot() on relocation tree"), however it
missed a spot when we restart a trans handle because we need to end the
transaction. The fix is the same, simply use btrfs_join_transaction()
instead of btrfs_start_transaction() when deleting reloc roots.
Fixes: f3e3d9cc3525 ("btrfs: avoid possible signal interruption of btrfs_drop_snapshot() on relocation tree")
CC: [email protected] # 5.4+
Signed-off-by: Josef Bacik <[email protected]>
Reviewed-by: David Sterba <[email protected]>
Signed-off-by: David Sterba <[email protected]>
|
|
The items of the 'maintainers' list are indented with three spaces.
Use the usual two spaces instead, for consistency and to silence
yamllint.
Signed-off-by: Jonathan Neuschäfer <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Mark Brown <[email protected]>
|
|
When switching between firmware types, the wrong control
can be selected when requesting control in kernel API.
Use the currently selected DSP firwmare type to select
the proper mixer control.
Signed-off-by: James Schulman <[email protected]>
Acked-by: Charles Keepax <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Mark Brown <[email protected]>
|
|
When the capacity of the disc is too large (assuming the 4.7G
specification), the disc (UDF file system) will be burned
multiple times in the windows (Multisession Usage). When the
remaining capacity of the CD is less than 300M (estimated
value, for reference only), open the CD in the Linux system,
the content of the CD is displayed as blank (the kernel will
say "No VRS found"). Windows can display the contents of the
CD normally.
Through analysis, in the "fs/udf/super.c": udf_check_vsd
function, the actual value of VSD_MAX_SECTOR_OFFSET may
be much larger than 0x800000. According to the current code
logic, it is found that the type of sbi->s_session is "__s32",
when the remaining capacity of the disc is less than 300M
(take a set of test values: sector=3154903040,
sbi->s_session=1540464, sb->s_blocksize_bits=11 ), the
calculation result of "sbi->s_session << sb->s_blocksize_bits"
will overflow. Therefore, it is necessary to convert the
type of s_session to "loff_t" (when udf_check_vsd starts,
assign a value to _sector, which is also converted in this
way), so that the result will not overflow, and then the
content of the disc can be displayed normally.
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: lianzhi chang <[email protected]>
Signed-off-by: Jan Kara <[email protected]>
|
|
Since we allow removing the timeline map at runtime, there is a risk
that rq->hwsp points into a stale page. To control that risk, we hold
the RCU read lock while reading *rq->hwsp, but we missed a couple of
important barriers. First, the unpinning / removal of the timeline map
must be after all RCU readers into that map are complete, i.e. after an
rcu barrier (in this case courtesy of call_rcu()). Secondly, we must
make sure that the rq->hwsp we are about to dereference under the RCU
lock is valid. In this case, we make the rq->hwsp pointer safe during
i915_request_retire() and so we know that rq->hwsp may become invalid
only after the request has been signaled. Therefore is the request is
not yet signaled when we acquire rq->hwsp under the RCU, we know that
rq->hwsp will remain valid for the duration of the RCU read lock.
This is a very small window that may lead to either considering the
request not completed (causing a delay until the request is checked
again, any wait for the request is not affected) or dereferencing an
invalid pointer.
Fixes: 3adac4689f58 ("drm/i915: Introduce concept of per-timeline (context) HWSP")
Signed-off-by: Chris Wilson <[email protected]>
Cc: Tvrtko Ursulin <[email protected]>
Cc: <[email protected]> # v5.1+
Reviewed-by: Tvrtko Ursulin <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
(cherry picked from commit 9bb36cf66091ddf2d8840e5aa705ad3c93a6279b)
Signed-off-by: Jani Nikula <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
|
|
Chris found a CI report which points out calling intel_runtime_pm_get from
inside i915_pmu_enable hook is not allowed since it can be invoked from
hard irq context. This is something we knew but forgot, so lets fix it
once again.
We do this by syncing the internal book keeping with hardware rc6 counter
on driver load.
v2:
* Always sync on parking and fully sync on init.
Signed-off-by: Tvrtko Ursulin <[email protected]>
Fixes: f4e9894b6952 ("drm/i915/pmu: Correct the rc6 offset upon enabling")
Cc: Chris Wilson <[email protected]>
Reviewed-by: Chris Wilson <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
(cherry picked from commit dbe13ae1d6abaab417edf3c37601c6a56594a4cd)
Signed-off-by: Jani Nikula <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
|
|
On error we unpin and free the wa_ctx.vma, but do not clear any of the
derived flags. During lrc_init, we look at the flags and attempt to
dereference the wa_ctx.vma if they are set. To protect the error path
where we try to limp along without the wa_ctx, make sure we clear those
flags!
Reported-by: Matt Roper <[email protected]>
Fixes: 604a8f6f1e33 ("drm/i915/lrc: Only enable per-context and per-bb buffers if set")
Signed-off-by: Chris Wilson <[email protected]>
Cc: Matt Roper <[email protected]>
Cc: Tvrtko Ursulin <[email protected]>
Cc: Mika Kuoppala <[email protected]>
Cc: <[email protected]> # v4.15+
Reviewed-by: Matt Roper <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
(cherry-picked from 5b4dc95cf7f573e927fbbd406ebe54225d41b9b2)
Signed-off-by: Jani Nikula <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
|
|
User-space ALSA matches a card's driver name against an internal list of
aliases in order to select the correct configuration for the system.
When the driver name isn't defined, the match is performed against the
card's name.
With the introduction of RPi4 we now have two HDMI ports with two
distinct audio cards. This is reflected in their names, making them
different from previous RPi versions. With this, ALSA ultimately misses
the board's configuration on RPi4.
In order to avoid this, set "card->driver_name" to "vc4-hdmi"
unanimously.
Signed-off-by: Nicolas Saenz Julienne <[email protected]>
Fixes: f437bc1ec731 ("drm/vc4: drv: Support BCM2711")
Reviewed-by: Takashi Iwai <[email protected]>
Signed-off-by: Maxime Ripard <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
|
|
Some platforms, such as mips64, don't map __u64 to long long unsigned
int so using %llu produces a warning:
gpio-watch.c: In function ‘main’:
gpio-watch.c:89:30: warning: format ‘%llu’ expects argument of type ‘long long unsigned int’, but argument 4 has type ‘__u64’ {aka ‘long unsigned int’} [-Wformat=]
89 | printf("line %u: %s at %llu\n",
| ~~~^
| |
| long long unsigned int
| %lu
90 | chg.info.offset, event, chg.timestamp_ns);
| ~~~~~~~~~~~~~~~~
| |
| __u64 {aka long unsigned int}
Replace the %llu with PRIu64 and cast the argument to uint64_t.
Fixes: 33f0c47b8fb4 ("tools: gpio: implement gpio-watch")
Signed-off-by: Kent Gibson <[email protected]>
Signed-off-by: Bartosz Golaszewski <[email protected]>
|
|
Some platforms, such as mips64, don't map __u64 to long long unsigned
int so using %llu produces a warning:
gpio-event-mon.c:110:37: warning: format ‘%llu’ expects argument of type ‘long long unsigned int’, but argument 3 has type ‘__u64’ {aka ‘long unsigned int’} [-Wformat=]
110 | fprintf(stdout, "GPIO EVENT at %llu on line %d (%d|%d) ",
| ~~~^
| |
| long long unsigned int
| %lu
111 | event.timestamp_ns, event.offset, event.line_seqno,
| ~~~~~~~~~~~~~~~~~~
| |
| __u64 {aka long unsigned int}
Replace the %llu with PRIu64 and cast the argument to uint64_t.
Fixes: 03fd11b03362 ("tools/gpio/gpio-event-mon: fix warning")
Signed-off-by: Kent Gibson <[email protected]>
Signed-off-by: Bartosz Golaszewski <[email protected]>
|
|
For addressing the regression on Pioneer devices, we recently
corrected the quirk code to enable the implicit feedback mode on those
devices properly. However, the devices still showed problems with the
full duplex operations with JACK, and after debug sessions, we figured
out that the older kernels that had worked with JACK also didn't use
the implicit feedback mode at all although they had the quirk code to
enable it; instead, the old code worked just to skip the normal sync
endpoint setup that would have been detected without it. IOW, what
broke without the implicit-fb quirk in the past was the application of
the normal sync endpoint that is actually the capture data endpoint on
these devices.
This patch covers the overseen piece: it modifies the quirk code again
not to enable the implicit feedback mode but just to make the driver
skipping the sync endpoint detection. This made the driver working
with JACK full-duplex mode again.
Still it's not quite clear why the implicit feedback doesn't work on
those devices yet; maybe it's about some issues in the URB setup. But
at least, with this patch, the driver should work in the level of the
older kernels again.
Fixes: 167c9dc84ec3 ("ALSA: usb-audio: Fix implicit feedback sync setup for Pioneer devices")
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
The UAC2/3 sample rate setup is based on the clock node, which is
usually shared in the interface, and can't be re-setup without
deselecting the interface once, and that's how the current code
behaves. OTOH, the sample rate setup of UAC1 is per endpoint, hence
we basically need to call for each endpoint usage even if those share
the same interface.
This patch fixes the behavior of UAC1 to call always
snd_usb_init_sample_rate() in snd_usb_endpoint_configure().
Fixes: bf6313a0ff76 ("ALSA: usb-audio: Refactor endpoint management")
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
The current sample rate setup function for UAC1 assumes only the first
endpoint retrieved from the interface:altset pair, but the rate set up
may be needed also for the secondary endpoint. Also, retrieving the
endpoint number from the interface descriptor is redundant; we have
already the target endpoint in the given audioformat object.
This patch simplifies the code and corrects the target endpoint as
described in the above. It simply refers to fmt->endpoint directly.
Also, this patch drops the pioneer_djm_set_format_quirk() that is
caleld from snd_usb_set_format_quirk(); this function does the sample
rate setup but for the capture endpoint (0x82), and that's exactly
what the change above fixes.
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Takashi Iwai <[email protected]>
|
|
The only flag we really need is __GFP_NOMEMALLOC, highmem depends on
dma32 and moveable/compound should never be set in the first place.
Signed-off-by: Christian König <[email protected]>
Link: https://patchwork.freedesktop.org/patch/413812/
Link: https://patchwork.freedesktop.org/patch/413964/
Fixes: d099fc8f540a ("drm/ttm: new TT backend allocation pool v3")
Reported-by: Hans de Goede <[email protected]>
Reviewed-by: Daniel Vetter <[email protected]>
|
|
Get DRM connector reference count while scheduling a prop work
to avoid any possible destroy of DRM connector when it is in
DRM_CONNECTOR_REGISTERED state.
Fixes: a6597faa2d59 ("drm/i915: Protect workers against disappearing connectors")
Cc: Sean Paul <[email protected]>
Cc: Ramalingam C <[email protected]>
Reviewed-by: Uma Shankar <[email protected]>
Reviewed-by: Ramalingam C <[email protected]>
Tested-by: Karthik B S <[email protected]>
Signed-off-by: Anshuman Gupta <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
(cherry picked from commit b3c6661aad979ec3d4f5675cf3e6a35828607d6a)
Signed-off-by: Jani Nikula <[email protected]>
|
|
When crtc state need_modeset is true it is not necessary
it is going to be a real modeset, it can turns to be a
fastset instead of modeset.
This turns content protection property to be DESIRED and hdcp
update_pipe left with property to be in DESIRED state but
actual hdcp->value was ENABLED.
This issue is caught with DP MST setup, where we have multiple
connector in same DP_MST topology. When disabling HDCP on one of
DP MST connector leads to set the crtc state need_modeset to true
for all other crtc driving the other DP-MST topology connectors.
This turns up other DP MST connectors CP property to be DESIRED
despite the actual hdcp->value is ENABLED.
Above scenario fails the DP MST HDCP IGT test, disabling HDCP on
one MST stream should not cause to disable HDCP on another MST
stream on same DP MST topology.
v2:
- Fixed connector->base.registration_state == DRM_CONNECTOR_REGISTERED
WARN_ON.
v3:
- Commit log improvement. [Uma]
- Added a comment before scheduling prop_work. [Uma]
Fixes: 33f9a623bfc6 ("drm/i915/hdcp: Update CP as per the kernel internal state")
Cc: Ramalingam C <[email protected]>
Reviewed-by: Uma Shankar <[email protected]>
Reviewed-by: Ramalingam C <[email protected]>
Tested-by: Karthik B S <[email protected]>
Signed-off-by: Anshuman Gupta <[email protected]>
Link: https://patchwork.freedesktop.org/patch/msgid/[email protected]
(cherry picked from commit d276e16702e2d634094f75f69df3b493f359fe31)
Signed-off-by: Jani Nikula <[email protected]>
|
|
Fix build error in x86/xen/ when PARAVIRT_SPINLOCKS is not enabled.
Fixes this build error:
../arch/x86/xen/smp_hvm.c: In function ‘xen_hvm_smp_init’:
../arch/x86/xen/smp_hvm.c:77:3: error: ‘nopvspin’ undeclared (first use in this function)
nopvspin = true;
Fixes: 3d7746bea925 ("x86/xen: Fix xen_hvm_smp_init() when vector callback not available")
Signed-off-by: Randy Dunlap <[email protected]>
Reviewed-by: Juergen Gross <[email protected]>
Cc: David Woodhouse <[email protected]>
Cc: Juergen Gross <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Juergen Gross <[email protected]>
|
|
Fix the follow warnings:
./fs/cifs/connect.c: WARNING: Comparison of 0/1 to bool variable
Reported-by: Abaci Robot <[email protected]>
Signed-off-by: Jiapeng Zhong <[email protected]>
Signed-off-by: Steve French <[email protected]>
|
|
Fix the following coccicheck warnings:
./fs/cifs/connect.c:3386:2-21: WARNING: Assignment of 0/1 to
bool variable.
Reported-by: Abaci Robot <[email protected]>
Signed-off-by: Jiapeng Zhong <[email protected]>
Signed-off-by: Steve French <[email protected]>
|
|
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux
Pull perf tools fixes from Arnaldo Carvalho de Melo:
- Fix 'CPU too large' error in Intel PT
- Correct event attribute sizes in 'perf inject'
- Sync build_bug.h and kvm.h kernel copies
- Fix bpf.h header include directive in 5sec.c 'perf trace' bpf example
- libbpf tests fixes
- Fix shadow stat 'perf test' for non-bash shells
- Take cgroups into account for shadow stats in 'perf stat'
* tag 'perf-tools-fixes-2021-01-17' of git://git.kernel.org/pub/scm/linux/kernel/git/acme/linux:
perf inject: Correct event attribute sizes
perf intel-pt: Fix 'CPU too large' error
perf stat: Take cgroups into account for shadow stats
perf stat: Introduce struct runtime_stat_data
libperf tests: Fail when failing to get a tracepoint id
libperf tests: If a test fails return non-zero
libperf tests: Avoid uninitialized variable warning
perf test: Fix shadow stat test for non-bash shells
tools headers: Syncronize linux/build_bug.h with the kernel sources
tools headers UAPI: Sync kvm.h headers with the kernel sources
perf bpf examples: Fix bpf.h header include directive in 5sec.c example
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux
Pull powerpc fixes from Michael Ellerman:
"One fix for a lack of alignment in our linker script, that can lead to
crashes depending on configuration etc.
One fix for the 32-bit VDSO after the C VDSO conversion.
Thanks to Andreas Schwab, Ariel Marcovitch, and Christophe Leroy"
* tag 'powerpc-5.11-4' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:
powerpc/vdso: Fix clock_gettime_fallback for vdso32
powerpc: Fix alignment bug within the init sections
|
|
Pull misc vfs fixes from Al Viro:
"Several assorted fixes.
I still think that audit ->d_name race is better fixed this way for
the benefit of backports, with any possibly fancier variants done on
top of it"
* 'fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
dump_common_audit_data(): fix racy accesses to ->d_name
iov_iter: fix the uaccess area in copy_compat_iovec_from_user
umount(2): move the flag validity checks first
|
|
So technically there is nothing wrong with adding a pinned page to the
swap cache, but the pinning obviously means that the page can't actually
be free'd right now anyway, so it's a bit pointless.
However, the real problem is not with it being a bit pointless: the real
issue is that after we've added it to the swap cache, we'll try to unmap
the page. That will succeed, because the code in mm/rmap.c doesn't know
or care about pinned pages.
Even the unmapping isn't fatal per se, since the page will stay around
in memory due to the pinning, and we do hold the connection to it using
the swap cache. But when we then touch it next and take a page fault,
the logic in do_swap_page() will map it back into the process as a
possibly read-only page, and we'll then break the page association on
the next COW fault.
Honestly, this issue could have been fixed in any of those other places:
(a) we could refuse to unmap a pinned page (which makes conceptual
sense), or (b) we could make sure to re-map a pinned page writably in
do_swap_page(), or (c) we could just make do_wp_page() not COW the
pinned page (which was what we historically did before that "mm:
do_wp_page() simplification" commit).
But while all of them are equally valid models for breaking this chain,
not putting pinned pages into the swap cache in the first place is the
simplest one by far.
It's also the safest one: the reason why do_wp_page() was changed in the
first place was that getting the "can I re-use this page" wrong is so
fraught with errors. If you do it wrong, you end up with an incorrectly
shared page.
As a result, using "page_maybe_dma_pinned()" in either do_wp_page() or
do_swap_page() would be a serious bug since it is only a (very good)
heuristic. Re-using the page requires a hard black-and-white rule with
no room for ambiguity.
In contrast, saying "this page is very likely dma pinned, so let's not
add it to the swap cache and try to unmap it" is an obviously safe thing
to do, and if the heuristic might very rarely be a false positive, no
harm is done.
Fixes: 09854ba94c6a ("mm: do_wp_page() simplification")
Reported-and-tested-by: Martin Raiber <[email protected]>
Cc: Pavel Begunkov <[email protected]>
Cc: Jens Axboe <[email protected]>
Cc: Peter Xu <[email protected]>
Signed-off-by: Linus Torvalds <[email protected]>
|
|
With commit 4df4cb9e99f8, the Hyper-V direct-mode STIMER is actually
initialized before LAPIC is initialized: see
apic_intr_mode_init()
x86_platform.apic_post_init()
hyperv_init()
hv_stimer_alloc()
apic_bsp_setup()
setup_local_APIC()
setup_local_APIC() temporarily disables LAPIC, initializes it and
re-eanble it. The direct-mode STIMER depends on LAPIC, and when it's
registered, it can be programmed immediately and the timer can fire
very soon:
hv_stimer_init
clockevents_config_and_register
clockevents_register_device
tick_check_new_device
tick_setup_device
tick_setup_periodic(), tick_setup_oneshot()
clockevents_program_event
When the timer fires in the hypervisor, if the LAPIC is in the
disabled state, new versions of Hyper-V ignore the event and don't inject
the timer interrupt into the VM, and hence the VM hangs when it boots.
Note: when the VM starts/reboots, the LAPIC is pre-enabled by the
firmware, so the window of LAPIC being temporarily disabled is pretty
small, and the issue can only happen once out of 100~200 reboots for
a 40-vCPU VM on one dev host, and on another host the issue doesn't
reproduce after 2000 reboots.
The issue is more noticeable for kdump/kexec, because the LAPIC is
disabled by the first kernel, and stays disabled until the kdump/kexec
kernel enables it. This is especially an issue to a Generation-2 VM
(for which Hyper-V doesn't emulate the PIT timer) when CONFIG_HZ=1000
(rather than CONFIG_HZ=250) is used.
Fix the issue by moving hv_stimer_alloc() to a later place where the
LAPIC timer is initialized.
Fixes: 4df4cb9e99f8 ("x86/hyperv: Initialize clockevents earlier in CPU onlining")
Signed-off-by: Dexuan Cui <[email protected]>
Reviewed-by: Michael Kelley <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Wei Liu <[email protected]>
|