John Fastabend says:
====================
Fix a splat introduced by recent changes to avoid skipping ingress policy
when kTLS is enabled. The RCU splat was introduced because in the non-TLS
case the caller is wrapped in an rcu_read_lock/unlock. But, in the TLS
case we have a reference to the psock and the caller did not wrap its
call in rcu_read_lock/unlock.
To fix extend the RCU section to include the redirect case which was
missed. From v1->v2 I changed the location a bit to simplify the code
some. See patch 1.
But, then Martin asked why it was not needed in the non-TLS case. The
answer for patch 1 was, as stated above, because the caller has the
rcu read lock. However, there was still a missing case where a BPF
user could in-theory line up a set of parameters to hit a case
where the code was entered from strparser side from a different context
then the initial caller. To hit this user would need a parser program
to return value greater than skb->len then an ENOMEM error could happen
in the strparser codepath triggering strparser to retry from a workqueue
and without rcu_read_lock original caller used. See patch 2 for details.
Finally, we don't actually have any selftests for parser returning a
value geater than skb->len so add one in patch 3. This is especially
needed because at least I don't have any code that uses the parser
to return value greater than skb->len. So I wouldn't have caught any
errors here in my own testing.
Thanks, John
v1->v2: simplify code in patch 1 some and add patches 2 and 3.
====================
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
If an ingress verdict program specifies message sizes greater than
skb->len and there is an ENOMEM error due to memory pressure we
may call the rcv_msg handler outside the strp_data_ready() caller
context. This is because on an ENOMEM error the strparser will
retry from a workqueue. The caller currently protects the use of
psock by calling the strp_data_ready() inside a rcu_read_lock/unlock
block.
But, in above workqueue error case the psock is accessed outside
the read_lock/unlock block of the caller. So instead of using
psock directly we must do a look up against the sk again to
ensure the psock is available.
There is an an ugly piece here where we must handle
the case where we paused the strp and removed the psock. On
psock removal we first pause the strparser and then remove
the psock. If the strparser is paused while an skb is
scheduled on the workqueue the skb will be dropped on the
flow and kfree_skb() is called. If the workqueue manages
to get called before we pause the strparser but runs the rcvmsg
callback after the psock is removed we will hit the unlikely
case where we run the sockmap rcvmsg handler but do not have
a psock. For now we will follow strparser logic and drop the
skb on the floor with skb_kfree(). This is ugly because the
data is dropped. To date this has not caused problems in practice
because either the application controlling the sockmap is
coordinating with the datapath so that skbs are "flushed"
before removal or we simply wait for the sock to be closed before
removing it.
This patch fixes the describe RCU bug and dropping the skb doesn't
make things worse. Future patches will improve this by allowing
the normal case where skbs are not merged to skip the strparser
altogether. In practice many (most?) use cases have no need to
merge skbs so its both a code complexity hit as seen above and
a performance issue. For example, in the Cilium case we always
set the strparser up to return sbks 1:1 without any merging and
have avoided above issues.
Fixes: e91de6afa8 ("bpf: Fix running sk_skb program types with ktls")
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Link: https://lore.kernel.org/bpf/159312679888.18340.15248924071966273998.stgit@john-XPS-13-9370
There are two paths to generate the below RCU splat the first and
most obvious is the result of the BPF verdict program issuing a
redirect on a TLS socket (This is the splat shown below). Unlike
the non-TLS case the caller of the *strp_read() hooks does not
wrap the call in a rcu_read_lock/unlock. Then if the BPF program
issues a redirect action we hit the RCU splat.
However, in the non-TLS socket case the splat appears to be
relatively rare, because the skmsg caller into the strp_data_ready()
is wrapped in a rcu_read_lock/unlock. Shown here,
static void sk_psock_strp_data_ready(struct sock *sk)
{
struct sk_psock *psock;
rcu_read_lock();
psock = sk_psock(sk);
if (likely(psock)) {
if (tls_sw_has_ctx_rx(sk)) {
psock->parser.saved_data_ready(sk);
} else {
write_lock_bh(&sk->sk_callback_lock);
strp_data_ready(&psock->parser.strp);
write_unlock_bh(&sk->sk_callback_lock);
}
}
rcu_read_unlock();
}
If the above was the only way to run the verdict program we
would be safe. But, there is a case where the strparser may throw an
ENOMEM error while parsing the skb. This is a result of a failed
skb_clone, or alloc_skb_for_msg while building a new merged skb when
the msg length needed spans multiple skbs. This will in turn put the
skb on the strp_wrk workqueue in the strparser code. The skb will
later be dequeued and verdict programs run, but now from a
different context without the rcu_read_lock()/unlock() critical
section in sk_psock_strp_data_ready() shown above. In practice
I have not seen this yet, because as far as I know most users of the
verdict programs are also only working on single skbs. In this case no
merge happens which could trigger the above ENOMEM errors. In addition
the system would need to be under memory pressure. For example, we
can't hit the above case in selftests because we missed having tests
to merge skbs. (Added in later patch)
To fix the below splat extend the rcu_read_lock/unnlock block to
include the call to sk_psock_tls_verdict_apply(). This will fix both
TLS redirect case and non-TLS redirect+error case. Also remove
psock from the sk_psock_tls_verdict_apply() function signature its
not used there.
[ 1095.937597] WARNING: suspicious RCU usage
[ 1095.940964] 5.7.0-rc7-02911-g463bac5f1ca79 #1 Tainted: G W
[ 1095.944363] -----------------------------
[ 1095.947384] include/linux/skmsg.h:284 suspicious rcu_dereference_check() usage!
[ 1095.950866]
[ 1095.950866] other info that might help us debug this:
[ 1095.950866]
[ 1095.957146]
[ 1095.957146] rcu_scheduler_active = 2, debug_locks = 1
[ 1095.961482] 1 lock held by test_sockmap/15970:
[ 1095.964501] #0: ffff9ea6b25de660 (sk_lock-AF_INET){+.+.}-{0:0}, at: tls_sw_recvmsg+0x13a/0x840 [tls]
[ 1095.968568]
[ 1095.968568] stack backtrace:
[ 1095.975001] CPU: 1 PID: 15970 Comm: test_sockmap Tainted: G W 5.7.0-rc7-02911-g463bac5f1ca79 #1
[ 1095.977883] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.12.0-1 04/01/2014
[ 1095.980519] Call Trace:
[ 1095.982191] dump_stack+0x8f/0xd0
[ 1095.984040] sk_psock_skb_redirect+0xa6/0xf0
[ 1095.986073] sk_psock_tls_strp_read+0x1d8/0x250
[ 1095.988095] tls_sw_recvmsg+0x714/0x840 [tls]
v2: Improve commit message to identify non-TLS redirect plus error case
condition as well as more common TLS case. In the process I decided
doing the rcu_read_unlock followed by the lock/unlock inside branches
was unnecessarily complex. We can just extend the current rcu block
and get the same effeective without the shuffling and branching.
Thanks Martin!
Fixes: e91de6afa8 ("bpf: Fix running sk_skb program types with ktls")
Reported-by: Jakub Sitnicki <jakub@cloudflare.com>
Reported-by: kernel test robot <rong.a.chen@intel.com>
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Acked-by: Jakub Sitnicki <jakub@cloudflare.com>
Link: https://lore.kernel.org/bpf/159312677907.18340.11064813152758406626.stgit@john-XPS-13-9370
Some performance regression on reaim benchmark have been raised with
commit 070f5e860e ("sched/fair: Take into account runnable_avg to classify group")
The problem comes from the init value of runnable_avg which is initialized
with max value. This can be a problem if the newly forked task is finally
a short task because the group of CPUs is wrongly set to overloaded and
tasks are pulled less agressively.
Set initial value of runnable_avg equals to util_avg to reflect that there
is no waiting time so far.
Fixes: 070f5e860e ("sched/fair: Take into account runnable_avg to classify group")
Reported-by: kernel test robot <rong.a.chen@intel.com>
Signed-off-by: Vincent Guittot <vincent.guittot@linaro.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20200624154422.29166-1-vincent.guittot@linaro.org
Paul reported rcutorture occasionally hitting a NULL deref:
sched_ttwu_pending()
ttwu_do_wakeup()
check_preempt_curr() := check_preempt_wakeup()
find_matching_se()
is_same_group()
if (se->cfs_rq == pse->cfs_rq) <-- *BOOM*
Debugging showed that this only appears to happen when we take the new
code-path from commit:
2ebb177175 ("sched/core: Offload wakee task activation if it the wakee is descheduling")
and only when @cpu == smp_processor_id(). Something which should not
be possible, because p->on_cpu can only be true for remote tasks.
Similarly, without the new code-path from commit:
c6e7bd7afa ("sched/core: Optimize ttwu() spinning on p->on_cpu")
this would've unconditionally hit:
smp_cond_load_acquire(&p->on_cpu, !VAL);
and if: 'cpu == smp_processor_id() && p->on_cpu' is possible, this
would result in an instant live-lock (with IRQs disabled), something
that hasn't been reported.
The NULL deref can be explained however if the task_cpu(p) load at the
beginning of try_to_wake_up() returns an old value, and this old value
happens to be smp_processor_id(). Further assume that the p->on_cpu
load accurately returns 1, it really is still running, just not here.
Then, when we enqueue the task locally, we can crash in exactly the
observed manner because p->se.cfs_rq != rq->cfs_rq, because p's cfs_rq
is from the wrong CPU, therefore we'll iterate into the non-existant
parents and NULL deref.
The closest semi-plausible scenario I've managed to contrive is
somewhat elaborate (then again, actual reproduction takes many CPU
hours of rcutorture, so it can't be anything obvious):
X->cpu = 1
rq(1)->curr = X
CPU0 CPU1 CPU2
// switch away from X
LOCK rq(1)->lock
smp_mb__after_spinlock
dequeue_task(X)
X->on_rq = 9
switch_to(Z)
X->on_cpu = 0
UNLOCK rq(1)->lock
// migrate X to cpu 0
LOCK rq(1)->lock
dequeue_task(X)
set_task_cpu(X, 0)
X->cpu = 0
UNLOCK rq(1)->lock
LOCK rq(0)->lock
enqueue_task(X)
X->on_rq = 1
UNLOCK rq(0)->lock
// switch to X
LOCK rq(0)->lock
smp_mb__after_spinlock
switch_to(X)
X->on_cpu = 1
UNLOCK rq(0)->lock
// X goes sleep
X->state = TASK_UNINTERRUPTIBLE
smp_mb(); // wake X
ttwu()
LOCK X->pi_lock
smp_mb__after_spinlock
if (p->state)
cpu = X->cpu; // =? 1
smp_rmb()
// X calls schedule()
LOCK rq(0)->lock
smp_mb__after_spinlock
dequeue_task(X)
X->on_rq = 0
if (p->on_rq)
smp_rmb();
if (p->on_cpu && ttwu_queue_wakelist(..)) [*]
smp_cond_load_acquire(&p->on_cpu, !VAL)
cpu = select_task_rq(X, X->wake_cpu, ...)
if (X->cpu != cpu)
switch_to(Y)
X->on_cpu = 0
UNLOCK rq(0)->lock
However I'm having trouble convincing myself that's actually possible
on x86_64 -- after all, every LOCK implies an smp_mb() there, so if ttwu
observes ->state != RUNNING, it must also observe ->cpu != 1.
(Most of the previous ttwu() races were found on very large PowerPC)
Nevertheless, this fully explains the observed failure case.
Fix it by ordering the task_cpu(p) load after the p->on_cpu load,
which is easy since nothing actually uses @cpu before this.
Fixes: c6e7bd7afa ("sched/core: Optimize ttwu() spinning on p->on_cpu")
Reported-by: Paul E. McKenney <paulmck@kernel.org>
Tested-by: Paul E. McKenney <paulmck@kernel.org>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lkml.kernel.org/r/20200622125649.GC576871@hirez.programming.kicks-ass.net
syzbot reported the following warning:
WARNING: CPU: 1 PID: 6351 at kernel/sched/deadline.c:628
enqueue_task_dl+0x22da/0x38a0 kernel/sched/deadline.c:1504
At deadline.c:628 we have:
623 static inline void setup_new_dl_entity(struct sched_dl_entity *dl_se)
624 {
625 struct dl_rq *dl_rq = dl_rq_of_se(dl_se);
626 struct rq *rq = rq_of_dl_rq(dl_rq);
627
628 WARN_ON(dl_se->dl_boosted);
629 WARN_ON(dl_time_before(rq_clock(rq), dl_se->deadline));
[...]
}
Which means that setup_new_dl_entity() has been called on a task
currently boosted. This shouldn't happen though, as setup_new_dl_entity()
is only called when the 'dynamic' deadline of the new entity
is in the past w.r.t. rq_clock and boosted tasks shouldn't verify this
condition.
Digging through the PI code I noticed that what above might in fact happen
if an RT tasks blocks on an rt_mutex hold by a DEADLINE task. In the
first branch of boosting conditions we check only if a pi_task 'dynamic'
deadline is earlier than mutex holder's and in this case we set mutex
holder to be dl_boosted. However, since RT 'dynamic' deadlines are only
initialized if such tasks get boosted at some point (or if they become
DEADLINE of course), in general RT 'dynamic' deadlines are usually equal
to 0 and this verifies the aforementioned condition.
Fix it by checking that the potential donor task is actually (even if
temporary because in turn boosted) running at DEADLINE priority before
using its 'dynamic' deadline value.
Fixes: 2d3d891d33 ("sched/deadline: Add SCHED_DEADLINE inheritance logic")
Reported-by: syzbot+119ba87189432ead09b4@syzkaller.appspotmail.com
Signed-off-by: Juri Lelli <juri.lelli@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Reviewed-by: Daniel Bristot de Oliveira <bristot@redhat.com>
Tested-by: Daniel Wagner <dwagner@suse.de>
Link: https://lkml.kernel.org/r/20181119153201.GB2119@localhost.localdomain
syzbot reported the following warning triggered via SYSC_sched_setattr():
WARNING: CPU: 0 PID: 6973 at kernel/sched/deadline.c:593 setup_new_dl_entity /kernel/sched/deadline.c:594 [inline]
WARNING: CPU: 0 PID: 6973 at kernel/sched/deadline.c:593 enqueue_dl_entity /kernel/sched/deadline.c:1370 [inline]
WARNING: CPU: 0 PID: 6973 at kernel/sched/deadline.c:593 enqueue_task_dl+0x1c17/0x2ba0 /kernel/sched/deadline.c:1441
This happens because the ->dl_boosted flag is currently not initialized by
__dl_clear_params() (unlike the other flags) and setup_new_dl_entity()
rightfully complains about it.
Initialize dl_boosted to 0.
Fixes: 2d3d891d33 ("sched/deadline: Add SCHED_DEADLINE inheritance logic")
Reported-by: syzbot+5ac8bac25f95e8b221e7@syzkaller.appspotmail.com
Signed-off-by: Juri Lelli <juri.lelli@redhat.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Tested-by: Daniel Wagner <dwagner@suse.de>
Link: https://lkml.kernel.org/r/20200617072919.818409-1-juri.lelli@redhat.com
This function is concerned with the long-term CPU mask, not the
transitory mask the task might have while migrate disabled. Before
this patch, if a task was migrate-disabled at the time
__set_cpus_allowed_ptr() was called, and the new mask happened to be
equal to the CPU that the task was running on, then the mask update
would be lost.
Signed-off-by: Scott Wood <swood@redhat.com>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Link: https://lkml.kernel.org/r/20200617121742.cpxppyi7twxmpin7@linutronix.de
i.MX fixes for 5.8:
- Fix LDO1 and LDO2 voltage range for a couple of i.MX8M board device
trees.
- Fix i.MX8MP UID fuse offset in i.MX8M SoC driver.
- Fix watchdog configuration in imx6ul-kontron device tree.
- Fix one build warning seen on building soc-imx8m driver with
x86_64-randconfig.
- Add missing put_device() call for a couple of mach-imx PM functions.
* tag 'imx-fixes-5.8' of git://git.kernel.org/pub/scm/linux/kernel/git/shawnguo/linux:
soc: imx8m: fix build warning
ARM: imx6: add missing put_device() call in imx6q_suspend_init()
ARM: imx5: add missing put_device() call in imx_suspend_alloc_ocram()
soc: imx8m: Correct i.MX8MP UID fuse offset
ARM: dts: imx6ul-kontron: Change WDOG_ANY signal from push-pull to open-drain
ARM: dts: imx6ul-kontron: Move watchdog from Kontron i.MX6UL/ULL board to SoM
arm64: dts: imx8mm-beacon: Fix voltages on LDO1 and LDO2
arm64: dts: imx8mn-ddr4-evk: correct ldo1/ldo2 voltage range
arm64: dts: imx8mm-evk: correct ldo1/ldo2 voltage range
Link: https://lore.kernel.org/r/20200624111725.GA24312@dragon
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
This pull request contains Broadcom ARM/ARM64/MIPS SoCs drivers fixes
for 5.8, please pull the following:
- Andy provides a fix for the Raspberry Pi firmware driver to print the
correct time upon boot. This is a fallout from a converstion to use
the ptT format
* tag 'arm-soc/for-5.8/drivers-fixes' of https://github.com/Broadcom/stblinux:
ARM: bcm2835: Fix integer overflow in rpi_firmware_print_firmware_revision()
Link: https://lore.kernel.org/r/20200619202250.19029-2-f.fainelli@gmail.com
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
This pull request contains Broadcom ARM-based SoCs Device Tree fixes for
5.8, please pull the following:
- Rafal adds a missing 'device_type' property to the Luxul XWC-2000
required for the memory nodes to be correctly parsed by Linux
- Matthew provides two fixes for the NSP SoCs, one to disable the PL330
DMA controller by default since it can be left in reset by the
bootloader and the second to correct the flow accelerator mailbox node
* tag 'arm-soc/for-5.8/devicetree-fixes' of https://github.com/Broadcom/stblinux:
ARM: dts: NSP: Correct FA2 mailbox node
ARM: dts: NSP: Disable PL330 by default, add dma-coherent property
ARM: dts: BCM5301X: Add missing memory "device_type" for Luxul XWC-2000
Link: https://lore.kernel.org/r/20200619202250.19029-1-f.fainelli@gmail.com
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
This reverts commit 7b8e0188fa.
Initially, STiH410-B2260 was supposed to be secured, that's why
l2c_write_sec was stubbed to avoid secure register access from
non secure world.
But by default, STiH410-B2260 is running in non secure mode,
so L2 cache register accesses are authorized, l2c_write_sec stub
is not needed.
With this patch, L2 cache is configured and performance are enhanced.
Link: https://lore.kernel.org/r/20200618172456.29475-1-patrice.chotard@st.com
Signed-off-by: Patrice Chotard <patrice.chotard@st.com>
Cc: Alain Volmat <alain.volmat@st.com>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Few dts fixes for omaps for v5.8
Few fixes for various devices:
- Prevent pocketgeagle header line signal from accidentally setting
micro-SD write protection signal by removing the default mux
- Fix NFSroot flakeyness after resume for duover by switching the
smsc911x gpio interrupt to back to level sensitive
- Fix regression for omap4 clockevent source after recent system
timer changes
- Yet another ethernet regression fix for the "rgmii" vs "rgmii-rxid"
phy-mode
* tag 'omap-for-v5.8/fixes-rc1-signed' of git://git.kernel.org/pub/scm/linux/kernel/git/tmlind/linux-omap:
ARM: dts: am5729: beaglebone-ai: fix rgmii phy-mode
ARM: dts: Fix omap4 system timer source clocks
ARM: dts: Fix duovero smsc interrupt for suspend
ARM: dts: am335x-pocketbeagle: Fix mmc0 Write Protect
Link: https://lore.kernel.org/r/pull-1592499282-121092@atomide.com
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Missed sdhci patch for am3 and am4
I forgot to send a pull request earlier for converting am3 and am4 to
use sdhci-omap driver instead of the old omap_hsmmc driver.
There was a display subsystem related suspend and resume regression found
recently and looks like I forgot to send a pull request for this patch
while debugging the regression. This patch has been tested without the
display subsystem, and has been in Linux next for several weeks now, so
would be good to have merged for v5.8.
* tag 'omap-for-v5.8/dt-missed-signed' of git://git.kernel.org/pub/scm/linux/kernel/git/tmlind/linux-omap:
ARM: dts: Move am33xx and am43xx mmc nodes to sdhci-omap driver
Link: https://lore.kernel.org/r/pull-1591637467-607254@atomide.com
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Fixes for omaps for v5.8
The recent display subsystem (DSS) related platform data changes caused
display related regressions for suspend and resume. Looks like I only
tested suspend and resume before dropping the legacy platform data, and
forgot to test it after dropping it. Turns out the main issue was that
we no longer have platform code calling pm_runtime_suspend for DSS like
we did for the legacy platform data case, and that fix is still being
discussed on the dri-devel list and will get merged separately. The DSS
related testing exposed a pile other other display related issues that
also need fixing though:
- Fix ti-sysc optional clock handling and reset status checks
for devices that reset automatically in idle like DSS
- Ignore ti-sysc clockactivity bit unless separately requested
to avoid unexpected performance issues
- Init ti-sysc framedonetv_irq to true and disable for am4
- Avoid duplicate DSS reset for legacy mode with dts data
- Remove LCD timings for am4 as they cause warnings now that we're
using generic panels
Then there is a pile of other fixes not related to the DSS:
- Fix omap_prm reset deassert as we still have drivers setting the
pm_runtime_irq_safe() flag
- Flush posted write for ti-sysc enable and disable
- Fix droid4 spi related errors with spi flags
- Fix am335x USB range and a typo for softreset
- Fix dra7 timer nodes for clocks for IPU and DSP
- Drop duplicate mailboxes after mismerge for dra7
* tag 'omap-for-v5.8/fixes-merge-window-signed' of git://git.kernel.org/pub/scm/linux/kernel/git/tmlind/linux-omap:
Revert "bus: ti-sysc: Increase max softreset wait"
ARM: dts: am437x-epos-evm: remove lcd timings
ARM: dts: am437x-gp-evm: remove lcd timings
ARM: dts: am437x-sk-evm: remove lcd timings
ARM: dts: dra7-evm-common: Fix duplicate mailbox nodes
ARM: dts: dra7: Fix timer nodes properly for timer_sys_ck clocks
ARM: dts: Fix am33xx.dtsi ti,sysc-mask wrong softreset flag
ARM: dts: Fix am33xx.dtsi USB ranges length
bus: ti-sysc: Increase max softreset wait
ARM: OMAP2+: Fix legacy mode dss_reset
bus: ti-sysc: Fix uninitialized framedonetv_irq
bus: ti-sysc: Ignore clockactivity unless specified as a quirk
bus: ti-sysc: Use optional clocks on for enable and wait for softreset bit
ARM: dts: omap4-droid4: Fix spi configuration and increase rate
bus: ti-sysc: Flush posted write on enable and disable
soc: ti: omap-prm: use atomic iopoll instead of sleeping one
Link: https://lore.kernel.org/r/pull-1591889257-410830@atomide.com
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
The cell name stored in the afs_cell struct is a 64-char + NUL buffer -
when it needs to be able to handle up to AFS_MAXCELLNAME (256 chars) + NUL.
Fix this by changing the array to a pointer and allocating the string.
Found using Coverity.
Fixes: 989782dcdc ("afs: Overhaul cell database management")
Reported-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Commit cd238effef ("docs: kbuild: convert docs to ReST and rename to
*.rst") missed a ReST header and a verbatim file content area.
Signed-off-by: Dov Murik <dovmurik@linux.vnet.ibm.com>
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
We can't cast sk_buff to rtable by (struct rtable *)hint. Use skb_rtable().
Fixes: 02b2494161 ("ipv4: use dst hint for ipv4 list receive")
Signed-off-by: Miaohe Lin <linmiaohe@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
intel-pinctrl for v5.8-2
* Fix output pin value handling on Intel Baytrail
The following is an automated git shortlog grouped by driver:
baytrail:
- Fix pin being driven low for a while on gpiod_get(..., GPIOD_OUT_HIGH)
Pull cifs fixes from Steve French:
"Six cifs/smb3 fixes, three of them for stable.
Fixes xfstests 451, 313 and 316"
* tag '5.8-rc2-smb3-fixes' of git://git.samba.org/sfrench/cifs-2.6:
cifs: misc: Use array_size() in if-statement controlling expression
cifs: update ctime and mtime during truncate
cifs/smb3: Fix data inconsistent when punch hole
cifs/smb3: Fix data inconsistent when zero file range
cifs: Fix double add page to memcg when cifs_readpages
cifs: Fix cached_fid refcnt leak in open_shroot
Pull SCSI fixes from James Bottomley:
"Six small fixes, five in drivers and one to correct another minor
regression from cc97923a5b ("block: move dma drain handling to
scsi") where we still need the drain stub to be built in to the kernel
for the modular libata, non-modular SAS driver case"
* tag 'scsi-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/jejb/scsi:
scsi: mptscsih: Fix read sense data size
scsi: zfcp: Fix panic on ERP timeout for previously dismissed ERP action
scsi: lpfc: Avoid another null dereference in lpfc_sli4_hba_unset()
scsi: libata: Fix the ata_scsi_dma_need_drain stub
scsi: qla2xxx: Keep initiator ports after RSCN
scsi: qla2xxx: Set NVMe status code for failed NVMe FCP request
Pull VFIO fixes from Alex Williamson:
- Fix double free of eventfd ctx (Alex Williamson)
- Fix duplicate use of capability ID (Alex Williamson)
- Fix SR-IOV VF memory enable handling (Alex Williamson)
* tag 'vfio-v5.8-rc3' of git://github.com/awilliam/linux-vfio:
vfio/pci: Fix SR-IOV VF handling with MMIO blocking
vfio/type1: Fix migration info capability ID
vfio/pci: Clear error and request eventfd ctx after releasing
Pull i2c fixes from Wolfram Sang:
"This contains a 5.8 regression fix for the Designware driver, a
register bitfield fix for the fsi driver, and a missing sanity check
for the I2C core"
* 'i2c/for-next' of git://git.kernel.org/pub/scm/linux/kernel/git/wsa/linux:
i2c: core: check returned size of emulated smbus block read
i2c: fsi: Fix the port number field in status register
i2c: designware: Adjust bus speed independently of ACPI
Pull staging driver fixes from Greg KH:
"Here are a small number of tiny staging driver fixes for 5.8-rc3.
Not much here, but there were some reported problems to be fixed:
- three wfx driver fixes
- rtl8723bs driver fix
All of these have been in linux-next with no reported issues"
* tag 'staging-5.8-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/staging:
Staging: rtl8723bs: prevent buffer overflow in update_sta_support_rate()
staging: wfx: fix coherency of hif_scan() prototype
staging: wfx: drop useless loop
staging: wfx: fix AC priority
Pull USB fixes from Greg KH:
"Here are some small USB fixes for 5.8-rc3 to resolve some reported
issues.
Nothing major here:
- gadget driver fixes
- cdns3 driver fixes
- xhci fixes
- renesas_usbhs driver fixes
- some new device support with ids
- documentation update
All of these have been in linux-next with no reported issues"
* tag 'usb-5.8-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb: (27 commits)
usb: renesas_usbhs: getting residue from callback_result
Revert "usb: dwc3: exynos: Add support for Exynos5422 suspend clk"
xhci: Poll for U0 after disabling USB2 LPM
xhci: Return if xHCI doesn't support LPM
usb: host: xhci-mtk: avoid runtime suspend when removing hcd
xhci: Fix enumeration issue when setting max packet size for FS devices.
xhci: Fix incorrect EP_STATE_MASK
usb: cdns3: ep0: add spinlock for cdns3_check_new_setup
usb: cdns3: trace: using correct dir value
usb: cdns3: ep0: fix the test mode set incorrectly
Revert "usb: dwc3: exynos: Add support for Exynos5422 suspend clk"
usb: gadget: udc: Potential Oops in error handling code
usb: phy: tegra: Fix unnecessary check in tegra_usb_phy_probe()
usb: dwc3: pci: Fix reference count leak in dwc3_pci_resume_work
usb: cdns3: ep0: add spinlock for cdns3_check_new_setup
usb: cdns3: trace: using correct dir value
usb: cdns3: ep0: fix the test mode set incorrectly
usb: typec: tcpci_rt1711h: avoid screaming irq causing boot hangs
USB: ohci-sm501: Add missed iounmap() in remove
cdc-acm: Add DISABLE_ECHO quirk for Microchip/SMSC chip
...
Pull char/misc fixes from Greg KH:
"Some tiny char/misc driver fixes for 5.8-rc3.
The "largest" changes are in the mei driver, to resolve some reported
problems and add some new device ids. There's also a binder bugfix, an
fpga driver build fix, and some assorted habanalabs fixes.
All of these, except for the habanalabs fixes, have been in linux-next
with no reported issues. The habanalabs driver changes showed up in my
tree on Friday, but as they are totally self-contained, all should be
good there"
* tag 'char-misc-5.8-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc:
habanalabs: increase h/w timer when checking idle
habanalabs: Correct handling when failing to enqueue CB
habanalabs: increase GAUDI QMAN ARB WDT timeout
habanalabs: rename mmu_write() to mmu_asid_va_write()
habanalabs: use PI in MMU cache invalidation
habanalabs: block scalar load_and_exe on external queue
mei: me: add tiger lake point device ids for H platforms.
mei: me: disable mei interface on Mehlow server platforms
binder: fix null deref of proc->context
fpga: zynqmp: fix modular build
Pull EDAC fix from Borislav Petkov:
"A single fix for amd64_edac restoring the reporting of the DRAM scrub
rate on family 0x15 CPUs"
* tag 'edac_urgent_for_5.8' of git://git.kernel.org/pub/scm/linux/kernel/git/ras/ras:
EDAC/amd64: Read back the scrub rate PCI register on F15h
Pull dma-mapping fixes from Christoph Hellwig:
- fix dma coherent mmap in nommu (me)
- more AMD SEV fallout (David Rientjes, me)
- fix alignment in dma_common_*_remap (Eric Auger)
* tag 'dma-mapping-5.8-4' of git://git.infradead.org/users/hch/dma-mapping:
dma-remap: align the size in dma_common_*_remap()
dma-mapping: DMA_COHERENT_POOL should select GENERIC_ALLOCATOR
dma-direct: add missing set_memory_decrypted() for coherent mapping
dma-direct: check return value when encrypting or decrypting memory
dma-direct: re-encrypt memory if dma_direct_alloc_pages() fails
dma-direct: always align allocation size in dma_direct_alloc_pages()
dma-direct: mark __dma_direct_alloc_pages static
dma-direct: re-enable mmap for !CONFIG_MMU
Pull NFS client bugfixes from Anna Schumaker:
"Stable Fixes:
- xprtrdma: Fix handling of RDMA_ERROR replies
- sunrpc: Fix rollback in rpc_gssd_dummy_populate()
- pNFS/flexfiles: Fix list corruption if the mirror count changes
- NFSv4: Fix CLOSE not waiting for direct IO completion
- SUNRPC: Properly set the @subbuf parameter of xdr_buf_subsegment()
Other Fixes:
- xprtrdma: Fix a use-after-free with r_xprt->rx_ep
- Fix other xprtrdma races during disconnect
- NFS: Fix memory leak of export_path"
* tag 'nfs-for-5.8-2' of git://git.linux-nfs.org/projects/anna/linux-nfs:
SUNRPC: Properly set the @subbuf parameter of xdr_buf_subsegment()
NFSv4 fix CLOSE not waiting for direct IO compeletion
pNFS/flexfiles: Fix list corruption if the mirror count changes
nfs: Fix memory leak of export_path
sunrpc: fixed rollback in rpc_gssd_dummy_populate()
xprtrdma: Fix handling of RDMA_ERROR replies
xprtrdma: Clean up disconnect
xprtrdma: Clean up synopsis of rpcrdma_flush_disconnect()
xprtrdma: Use re_connect_status safely in rpcrdma_xprt_connect()
xprtrdma: Prevent dereferencing r_xprt->rx_ep after it is freed
Pull io_uring fixes from Jens Axboe:
"Three small fixes:
- Close a corner case for polled IO resubmission (Pavel)
- Toss commands when exiting (Pavel)
- Fix SQPOLL conditional reschedule on perpetually busy submit
(Xuan)"
* tag 'io_uring-5.8-2020-06-26' of git://git.kernel.dk/linux-block:
io_uring: fix current->mm NULL dereference on exit
io_uring: fix hanging iopoll in case of -EAGAIN
io_uring: fix io_sq_thread no schedule when busy
Pull block fixes from Jens Axboe:
- NVMe pull request from Christoph:
- multipath deadlock fixes (Anton)
- NUMA fixes (Max)
- RDMA completion vector fix (Max)
- IO deadlock fix (Sagi)
- multipath reference fix (Sagi)
- NS mutation fix (Sagi)
- Use right allocator when freeing bip in error path (Chengguang)
* tag 'block-5.8-2020-06-26' of git://git.kernel.dk/linux-block:
nvme-multipath: fix bogus request queue reference put
nvme-multipath: fix deadlock due to head->lock
nvme: don't protect ns mutation with ns->head->lock
nvme-multipath: fix deadlock between ana_work and scan_work
nvme: fix possible deadlock when I/O is blocked
nvme-rdma: assign completion vector correctly
nvme-loop: initialize tagset numa value to the value of the ctrl
nvme-tcp: initialize tagset numa value to the value of the ctrl
nvme-pci: initialize tagset numa value to the value of the ctrl
nvme-pci: override the value of the controller's numa node
nvme: set initial value for controller's numa node
block: release bip in a right way in error path
Pull device mapper fixes from Mike Snitzer:
- Quite a few DM zoned target fixes and a Zone append fix in DM core.
Considering the amount of dm-zoned changes that went in during the
5.8 merge window these fixes are not that surprising.
- A few DM writecache target fixes.
- A fix to Documentation index to include DM ebs target docs.
- Small cleanup to use struct_size() in DM core's retrieve_deps().
* tag 'for-5.8/dm-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/device-mapper/linux-dm:
dm writecache: add cond_resched to loop in persistent_memory_claim()
dm zoned: Fix reclaim zone selection
dm zoned: Fix random zone reclaim selection
dm: update original bio sector on Zone Append
dm zoned: Fix metadata zone size check
docs: device-mapper: add dm-ebs.rst to an index file
dm ioctl: use struct_size() helper in retrieve_deps()
dm writecache: skip writecache_wait when using pmem mode
dm writecache: correct uncommitted_block when discarding uncommitted entry
dm zoned: assign max_io_len correctly
dm zoned: fix uninitialized pointer dereference
Pull kgdb fixes from Daniel Thompson:
"The main change here is a fix for a number of unsafe interactions
between kdb and the console system. The fixes are specific to kdb
(pure kgdb debugging does not use the console system at all). On
systems with an NMI then kdb, if it is enabled, must get messages to
the user despite potentially running from some "difficult" calling
contexts. These fixes avoid using the console system where we have
been provided an alternative (safer) way to interact with the user
and, if using the console system in unavoidable, use oops_in_progress
for deadlock avoidance. These fixes also ensure kdb honours the
console enable flag.
Also included is a fix that wraps kgdb trap handling in an RCU read
lock to avoids triggering diagnostic warnings. This is a wide lock
scope but this is OK because kgdb is a stop-the-world debugger. When
we stop the world we put all the CPUs into holding pens and this
inhibits RCU update anyway"
* tag 'kgdb-5.8-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/danielt/linux:
kgdb: Avoid suspicious RCU usage warning
kdb: Switch to use safer dbg_io_ops over console APIs
kdb: Make kdb_printf() console handling more robust
kdb: Check status of console prior to invoking handlers
kdb: Re-factor kdb_printf() message write code
Pull powerpc fixes from Michael Ellerman:
- A fix for a crash in nested KVM when CONFIG_DEBUG_VIRTUAL=y.
- Two minor build fixes.
Thanks to: Aneesh Kumar K.V, Arseny Solokha, Harish.
* tag 'powerpc-5.8-4' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:
selftests/powerpc: Fix build failure in ebb tests
powerpc/kvm/book3s64: Fix kernel crash with nested kvm & DEBUG_VIRTUAL
powerpc/fsl_booke/32: Fix build with CONFIG_RANDOMIZE_BASE
Pull RISC-V fixes from Palmer Dabbelt:
"This contains a handful of fixes I'd like to target for rc3.
Most of them fix issues with the conversion of our vDSO to C. There is
also one fix to the SiFive PRCI driver that I picked up as it's
causing boot issues on the hardware.
- A fix to allow kernels with dynamic ftrace to use the vDSO.
- Some build fixes for the C vDSO functions.
- A fix to the PRCI driver's memory allocation, which was the cause
of some boot panics with FREELIST_RANDOM"
* tag 'riscv-for-linus-5.8-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/riscv/linux:
riscv: Fixup __vdso_gettimeofday broke dynamic ftrace
riscv: Add extern declarations for vDSO time-related functions
clk: sifive: allocate sufficient memory for struct __prci_data
riscv: Add -fPIC option to CFLAGS_vgettimeofday.o
Pull arm64 fixes from Will Deacon:
"The big fix here is to our vDSO sigreturn trampoline as, after a
painfully long stint of debugging, it turned out that fixing some of
our CFI directives in the merge window lit up a bunch of logic in
libgcc which has been shown to SEGV in some cases during asynchronous
pthread cancellation.
It looks like we can fix this by extending the directives to restore
most of the interrupted register state from the sigcontext, but it's
risky and hard to test so we opted to remove the CFI directives for
now and rely on the unwinder fallback path like we used to.
- Fix unwinding through vDSO sigreturn trampoline
- Fix build warnings by raising minimum LD version for PAC
- Whitelist some Kryo Cortex-A55 derivatives for Meltdown and SSB
- Fix perf register PC reporting for compat tasks
- Fix 'make clean' warning for arm64 signal selftests
- Fix ftrace when BTI is compiled in
- Avoid building the compat vDSO using GCC plugins"
* tag 'arm64-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/arm64/linux:
arm64: Add KRYO{3,4}XX silver CPU cores to SSB safelist
arm64: perf: Report the PC value in REGS_ABI_32 mode
kselftest: arm64: Remove redundant clean target
arm64: kpti: Add KRYO{3, 4}XX silver CPU cores to kpti safelist
arm64: Don't insert a BTI instruction at inner labels
arm64: vdso: Don't use gcc plugins for building vgettimeofday.c
arm64: vdso: Only pass --no-eh-frame-hdr when linker supports it
arm64: Depend on newer binutils when building PAC
arm64: compat: Remove 32-bit sigreturn code from the vDSO
arm64: compat: Always use sigpage for sigreturn trampoline
arm64: compat: Allow 32-bit vdso and sigpage to co-exist
arm64: vdso: Disable dwarf unwinding through the sigreturn trampoline
The mpt fusion driver still uses the legacy PCI DMA API which hardcodes
atomic allocations. This caused the driver to fail to load on some powerpc
VMs with incoherent DMA and small memory sizes. Switch to use the modern
DMA API and sleeping allocations for large allocations instead. This is
not a full cleanup of the PCI DMA API usage yet, but just enough to fix the
regression caused by reducing the default atomic pool size.
Link: https://lore.kernel.org/r/20200624165724.1818496-1-hch@lst.de
Fixes: 3ee06a6d53 ("dma-pool: fix too large DMA pools on medium memory size systems")
Reported-by: Guenter Roeck <linux@roeck-us.net>
Tested-by: Guenter Roeck <linux@roeck-us.net>
Signed-off-by: Christoph Hellwig <hch@lst.de>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
When an rport event (RPORT_EV_READY) is updated without work being queued,
avoid taking an additional reference.
This issue was leading to memory leak. Trace from KMEMLEAK tool:
unreferenced object 0xffff8888259e8780 (size 512):
comm "kworker/2:1", jiffies 4433237386 (age 113021.971s)
hex dump (first 32 bytes):
58 0a ec cf 83 88 ff ff 00 00 00 00 00 00 00 00
01 00 00 00 08 00 00 00 13 7d f0 1e 0e 00 00 10
backtrace:
[<000000006b25760f>] fc_rport_recv_req+0x3c6/0x18f0 [libfc]
[<00000000f208d994>] fc_lport_recv_els_req+0x120/0x8a0 [libfc]
[<00000000a9c437b8>] fc_lport_recv+0xb9/0x130 [libfc]
[<00000000a9c437b8>] fc_lport_recv+0xb9/0x130 [libfc]
[<00000000ad5be37b>] qedf_ll2_process_skb+0x73d/0xad0 [qedf]
[<00000000e0eb6893>] process_one_work+0x382/0x6c0
[<000000002dfd9e21>] worker_thread+0x57/0x5c0
[<00000000b648204f>] kthread+0x1a0/0x1c0
[<0000000072f5ab20>] ret_from_fork+0x35/0x40
[<000000001d5c05d8>] 0xffffffffffffffff
Below is the log sequence which leads to memory leak. Here we get the
RPORT_EV_READY and RPORT_EV_STOP back to back, which lead to overwrite the
event RPORT_EV_READY by event RPORT_EV_STOP. Because of this, kref_count
gets incremented by 1.
kernel: host0: rport fffce5: Received PLOGI request
kernel: host0: rport fffce5: Received PLOGI in INIT state
kernel: host0: rport fffce5: Port is Ready
kernel: host0: rport fffce5: Received PRLI request while in state Ready
kernel: host0: rport fffce5: PRLI rspp type 8 active 1 passive 0
kernel: host0: rport fffce5: Received LOGO request while in state Ready
kernel: host0: rport fffce5: Delete port
kernel: host0: rport fffce5: Received PLOGI request
kernel: host0: rport fffce5: Received PLOGI in state Delete - send busy
kernel: host0: rport fffce5: work event 3
kernel: host0: rport fffce5: lld callback ev 3
kernel: host0: rport fffce5: work delete
Link: https://lore.kernel.org/r/20200626094959.32151-1-jhasan@marvell.com
Reviewed-by: Girish Basrur <gbasrur@marvell.com>
Reviewed-by: Saurav Kashyap <skashyap@marvell.com>
Reviewed-by: Shyam Sundar <ssundar@marvell.com>
Signed-off-by: Javed Hasan <jhasan@marvell.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Handling of extra kref which is done by lookup table in case rdata is
already present in list.
This issue was leading to memory leak. Trace from KMEMLEAK tool:
unreferenced object 0xffff8888259e8780 (size 512):
comm "kworker/2:1", pid 182614, jiffies 4433237386 (age 113021.971s)
hex dump (first 32 bytes):
58 0a ec cf 83 88 ff ff 00 00 00 00 00 00 00 00
01 00 00 00 08 00 00 00 13 7d f0 1e 0e 00 00 10
backtrace:
[<000000006b25760f>] fc_rport_recv_req+0x3c6/0x18f0 [libfc]
[<00000000f208d994>] fc_lport_recv_els_req+0x120/0x8a0 [libfc]
[<00000000a9c437b8>] fc_lport_recv+0xb9/0x130 [libfc]
[<00000000ad5be37b>] qedf_ll2_process_skb+0x73d/0xad0 [qedf]
[<00000000e0eb6893>] process_one_work+0x382/0x6c0
[<000000002dfd9e21>] worker_thread+0x57/0x5c0
[<00000000b648204f>] kthread+0x1a0/0x1c0
[<0000000072f5ab20>] ret_from_fork+0x35/0x40
[<000000001d5c05d8>] 0xffffffffffffffff
Below is the log sequence which leads to memory leak. Here we get the
nested "Received PLOGI request" for same port and this request leads to
call the fc_rport_create() twice for the same rport.
kernel: host1: rport fffce5: Received PLOGI request
kernel: host1: rport fffce5: Received PLOGI in INIT state
kernel: host1: rport fffce5: Port is Ready
kernel: host1: rport fffce5: Received PRLI request while in state Ready
kernel: host1: rport fffce5: PRLI rspp type 8 active 1 passive 0
kernel: host1: rport fffce5: Received LOGO request while in state Ready
kernel: host1: rport fffce5: Delete port
kernel: host1: rport fffce5: Received PLOGI request
kernel: host1: rport fffce5: Received PLOGI in state Delete - send busy
Link: https://lore.kernel.org/r/20200622101212.3922-2-jhasan@marvell.com
Reviewed-by: Girish Basrur <gbasrur@marvell.com>
Reviewed-by: Saurav Kashyap <skashyap@marvell.com>
Reviewed-by: Shyam Sundar <ssundar@marvell.com>
Signed-off-by: Javed Hasan <jhasan@marvell.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
Disable the plane if it's not visible. Otherwise mtk_ovl_layer_config()
would proceed with invalid plane and we may see vblank timeout.
Fixes: 119f517362 ("drm/mediatek: Add DRM Driver for Mediatek SoC MT8173.")
Signed-off-by: Hsin-Yi Wang <hsinyi@chromium.org>
Reviewed-by: Tomasz Figa <tfiga@chromium.org>
Signed-off-by: Chun-Kuang Hu <chunkuang.hu@kernel.org>
