[paulmck-rcu:dev.2021.12.07a 56/57] kernel/rcu/tree.c:4312:45: warning: variable 'flags' is uninitialized when used here
6:30 p.m.
tree: https://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu.git
dev.2021.12.07a
head: 89ddc546d3ad541bc89ec553c92036c98aef57a6
commit: 342f2d0f88f3c2d7ab2ae206be5846579319e3bf [56/57] rcu: Kill rnp->ofl_seq and use
only rcu_state.ofl_lock for exclusion
config: hexagon-randconfig-r041-20211210
(https://download.01.org/0day-ci/archive/20211210/202112101543.wHvqhgYk-lk...)
compiler: clang version 14.0.0 (https://github.com/llvm/llvm-project
097a1cb1d5ebb3a0ec4bcaed8ba3ff6a8e33c00a)
reproduce (this is a W=1 build):
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O
~/bin/make.cross
chmod +x ~/bin/make.cross
#
https://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu.git/com...
git remote add paulmck-rcu
https://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu.git
git fetch --no-tags paulmck-rcu dev.2021.12.07a
git checkout 342f2d0f88f3c2d7ab2ae206be5846579319e3bf
# save the config file to linux build tree
mkdir build_dir
COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross W=1 O=build_dir
ARCH=hexagon SHELL=/bin/bash
If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <lkp(a)intel.com>
All warnings (new ones prefixed by >>):
> kernel/rcu/tree.c:4312:45: warning: variable 'flags' is
uninitialized when used here [-Wuninitialized]
rcu_report_qs_rnp(mask, rnp, rnp->gp_seq, flags);
^~~~~
kernel/rcu/tree.c:4280:21: note: initialize the variable 'flags' to silence
this warning
unsigned long flags, seq_flags;
^
= 0
1 warning generated.
vim +/flags +4312 kernel/rcu/tree.c
4df8374254ea929 Thomas Gleixner 2016-07-13 4266
7ec99de36f40261 Paul E. McKenney 2016-06-30 4267 /*
7ec99de36f40261 Paul E. McKenney 2016-06-30 4268 * Mark the specified CPU as
being online so that subsequent grace periods
7ec99de36f40261 Paul E. McKenney 2016-06-30 4269 * (both expedited and normal)
will wait on it. Note that this means that
7ec99de36f40261 Paul E. McKenney 2016-06-30 4270 * incoming CPUs are not allowed
to use RCU read-side critical sections
7ec99de36f40261 Paul E. McKenney 2016-06-30 4271 * until this function is
called. Failing to observe this restriction
7ec99de36f40261 Paul E. McKenney 2016-06-30 4272 * will result in lockdep
splats.
deb34f364398083 Paul E. McKenney 2017-03-23 4273 *
deb34f364398083 Paul E. McKenney 2017-03-23 4274 * Note that this function is
special in that it is invoked directly
deb34f364398083 Paul E. McKenney 2017-03-23 4275 * from the incoming CPU rather
than from the cpuhp_step mechanism.
deb34f364398083 Paul E. McKenney 2017-03-23 4276 * This is because this function
must be invoked at a precise location.
7ec99de36f40261 Paul E. McKenney 2016-06-30 4277 */
7ec99de36f40261 Paul E. McKenney 2016-06-30 4278 void rcu_cpu_starting(unsigned
int cpu)
7ec99de36f40261 Paul E. McKenney 2016-06-30 4279 {
342f2d0f88f3c2d David Woodhouse 2021-12-09 4280 unsigned long flags,
seq_flags;
7ec99de36f40261 Paul E. McKenney 2016-06-30 4281 unsigned long mask;
7ec99de36f40261 Paul E. McKenney 2016-06-30 4282 struct rcu_data *rdp;
7ec99de36f40261 Paul E. McKenney 2016-06-30 4283 struct rcu_node *rnp;
abfce0414814149 Wei Yang 2020-04-19 4284 bool newcpu;
7ec99de36f40261 Paul E. McKenney 2016-06-30 4285
c0f97f20e5d97a1 Paul E. McKenney 2020-07-24 4286 rdp =
per_cpu_ptr(&rcu_data, cpu);
c0f97f20e5d97a1 Paul E. McKenney 2020-07-24 4287 if (rdp->cpu_started)
f64c6013a202931 Peter Zijlstra 2018-05-22 4288 return;
c0f97f20e5d97a1 Paul E. McKenney 2020-07-24 4289 rdp->cpu_started = true;
f64c6013a202931 Peter Zijlstra 2018-05-22 4290
7ec99de36f40261 Paul E. McKenney 2016-06-30 4291 rnp = rdp->mynode;
7ec99de36f40261 Paul E. McKenney 2016-06-30 4292 mask = rdp->grpmask;
342f2d0f88f3c2d David Woodhouse 2021-12-09 4293 local_irq_save(seq_flags);
342f2d0f88f3c2d David Woodhouse 2021-12-09 4294
arch_spin_lock(&rcu_state.ofl_lock);
2caebefb00f03b5 Paul E. McKenney 2021-07-28 4295 rcu_dynticks_eqs_online();
4d60b475f858ebd Paul E. McKenney 2020-10-13 4296 smp_mb(); // Pair with
rcu_gp_cleanup()'s ->ofl_seq barrier().
342f2d0f88f3c2d David Woodhouse 2021-12-09 4297 raw_spin_lock_rcu_node(rnp);
105abf82b0a6266 Paul E. McKenney 2020-01-03 4298
WRITE_ONCE(rnp->qsmaskinitnext, rnp->qsmaskinitnext | mask);
abfce0414814149 Wei Yang 2020-04-19 4299 newcpu =
!(rnp->expmaskinitnext & mask);
7ec99de36f40261 Paul E. McKenney 2016-06-30 4300 rnp->expmaskinitnext |=
mask;
313517fc44fb2d8 Paul E. McKenney 2017-06-08 4301 /* Allow lockless access for
expedited grace periods. */
abfce0414814149 Wei Yang 2020-04-19 4302
smp_store_release(&rcu_state.ncpus, rcu_state.ncpus + newcpu); /* ^^^ */
2f08469563550d1 Paul E. McKenney 2020-02-10 4303
ASSERT_EXCLUSIVE_WRITER(rcu_state.ncpus);
e05121ba5b81e2f Paul E. McKenney 2018-05-07 4304 rcu_gpnum_ovf(rnp, rdp); /*
Offline-induced counter wrap? */
eb7a6653887b540 Paul E. McKenney 2018-07-05 4305 rdp->rcu_onl_gp_seq =
READ_ONCE(rcu_state.gp_seq);
eb7a6653887b540 Paul E. McKenney 2018-07-05 4306 rdp->rcu_onl_gp_flags =
READ_ONCE(rcu_state.gp_flags);
9f866dac94292f9 Joel Fernandes (Google 2020-09-29 4307)
9f866dac94292f9 Joel Fernandes (Google 2020-09-29 4308) /* An incoming CPU should never
be blocking a grace period. */
9f866dac94292f9 Joel Fernandes (Google 2020-09-29 4309) if (WARN_ON_ONCE(rnp->qsmask
& mask)) { /* RCU waiting on incoming CPU? */
516e5ae0c940162 Joel Fernandes (Google 2019-09-05 4310)
rcu_disable_urgency_upon_qs(rdp);
99990da1b3c00f6 Paul E. McKenney 2018-05-03 4311 /* Report QS -after- changing
->qsmaskinitnext! */
b50912d0b5e03f1 Paul E. McKenney 2018-07-03 @4312 rcu_report_qs_rnp(mask, rnp,
rnp->gp_seq, flags);
99990da1b3c00f6 Paul E. McKenney 2018-05-03 4313 } else {
7ec99de36f40261 Paul E. McKenney 2016-06-30 4314
raw_spin_unlock_irqrestore_rcu_node(rnp, flags);
7ec99de36f40261 Paul E. McKenney 2016-06-30 4315 }
342f2d0f88f3c2d David Woodhouse 2021-12-09 4316
arch_spin_unlock(&rcu_state.ofl_lock);
342f2d0f88f3c2d David Woodhouse 2021-12-09 4317 local_irq_restore(seq_flags);
313517fc44fb2d8 Paul E. McKenney 2017-06-08 4318 smp_mb(); /* Ensure RCU
read-side usage follows above initialization. */
7ec99de36f40261 Paul E. McKenney 2016-06-30 4319 }
7ec99de36f40261 Paul E. McKenney 2016-06-30 4320
:::::: The code at line 4312 was first introduced by commit
:::::: b50912d0b5e03f11004fec1e2b50244de9e2fa41 rcu: Remove rsp parameter from
rcu_report_qs_rnp()
:::::: TO: Paul E. McKenney <paulmck(a)linux.vnet.ibm.com>
:::::: CC: Paul E. McKenney <paulmck(a)linux.vnet.ibm.com>
---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org
12:57 a.m.
On Fri, Dec 10, 2021 at 03:30:22PM +0800, kernel test robot wrote:
tree:
https://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu.git dev.2021.12.07a
head: 89ddc546d3ad541bc89ec553c92036c98aef57a6
commit: 342f2d0f88f3c2d7ab2ae206be5846579319e3bf [56/57] rcu: Kill rnp->ofl_seq and
use only rcu_state.ofl_lock for exclusion
config: hexagon-randconfig-r041-20211210
(https://download.01.org/0day-ci/archive/20211210/202112101543.wHvqhgYk-lk...)
compiler: clang version 14.0.0 (https://github.com/llvm/llvm-project
097a1cb1d5ebb3a0ec4bcaed8ba3ff6a8e33c00a)
reproduce (this is a W=1 build):
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O
~/bin/make.cross
chmod +x ~/bin/make.cross
#
https://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu.git/com...
git remote add paulmck-rcu
https://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu.git
git fetch --no-tags paulmck-rcu dev.2021.12.07a
git checkout 342f2d0f88f3c2d7ab2ae206be5846579319e3bf
# save the config file to linux build tree
mkdir build_dir
COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross W=1 O=build_dir
ARCH=hexagon SHELL=/bin/bash
If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <lkp(a)intel.com>
All warnings (new ones prefixed by >>):
>> kernel/rcu/tree.c:4312:45: warning: variable 'flags' is uninitialized
when used here [-Wuninitialized]
rcu_report_qs_rnp(mask, rnp, rnp->gp_seq, flags);
^~~~~
kernel/rcu/tree.c:4280:21: note: initialize the variable 'flags' to silence
this warning
unsigned long flags, seq_flags;
^
= 0
1 warning generated.
Good catch, but Neeraj and David beat you to this. The updated commit
is defd1f8a1f4e ("rcu: Kill rnp->ofl_seq and use only rcu_state.ofl_lock
for exclusion").
One thing that might be helpful to you during times of backlock... In the
-rcu tree, if you have several branches with the same prefix (for example,
dev.2021.12.03a, dev.2021.12.07a, and dev.2021.12.09a), you can safely
ignore all but the last one in lexicographic order. In this example,
dev.2021.12.09a obsoletes dev.2021.12.03a and dev.2021.12.07a, so you
would lose nothing by just ignoring the earlier two branches.
Thanx, Paul
> vim +/flags +4312 kernel/rcu/tree.c
>
> 4df8374254ea929 Thomas Gleixner 2016-07-13 4266
> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4267 /*
> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4268 * Mark the specified CPU
as being online so that subsequent grace periods
> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4269 * (both expedited and
normal) will wait on it. Note that this means that
> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4270 * incoming CPUs are not
allowed to use RCU read-side critical sections
> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4271 * until this function is
called. Failing to observe this restriction
> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4272 * will result in lockdep
splats.
> deb34f364398083 Paul E. McKenney 2017-03-23 4273 *
> deb34f364398083 Paul E. McKenney 2017-03-23 4274 * Note that this function
is special in that it is invoked directly
> deb34f364398083 Paul E. McKenney 2017-03-23 4275 * from the incoming CPU
rather than from the cpuhp_step mechanism.
> deb34f364398083 Paul E. McKenney 2017-03-23 4276 * This is because this
function must be invoked at a precise location.
> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4277 */
> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4278 void
rcu_cpu_starting(unsigned int cpu)
> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4279 {
> 342f2d0f88f3c2d David Woodhouse 2021-12-09 4280 unsigned long flags,
seq_flags;
> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4281 unsigned long mask;
> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4282 struct rcu_data *rdp;
> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4283 struct rcu_node *rnp;
> abfce0414814149 Wei Yang 2020-04-19 4284 bool newcpu;
> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4285
> c0f97f20e5d97a1 Paul E. McKenney 2020-07-24 4286 rdp =
per_cpu_ptr(&rcu_data, cpu);
> c0f97f20e5d97a1 Paul E. McKenney 2020-07-24 4287 if (rdp->cpu_started)
> f64c6013a202931 Peter Zijlstra 2018-05-22 4288 return;
> c0f97f20e5d97a1 Paul E. McKenney 2020-07-24 4289 rdp->cpu_started =
true;
> f64c6013a202931 Peter Zijlstra 2018-05-22 4290
> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4291 rnp = rdp->mynode;
> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4292 mask = rdp->grpmask;
> 342f2d0f88f3c2d David Woodhouse 2021-12-09 4293
local_irq_save(seq_flags);
> 342f2d0f88f3c2d David Woodhouse 2021-12-09 4294
arch_spin_lock(&rcu_state.ofl_lock);
> 2caebefb00f03b5 Paul E. McKenney 2021-07-28 4295
rcu_dynticks_eqs_online();
> 4d60b475f858ebd Paul E. McKenney 2020-10-13 4296 smp_mb(); // Pair with
rcu_gp_cleanup()'s ->ofl_seq barrier().
> 342f2d0f88f3c2d David Woodhouse 2021-12-09 4297
raw_spin_lock_rcu_node(rnp);
> 105abf82b0a6266 Paul E. McKenney 2020-01-03 4298
WRITE_ONCE(rnp->qsmaskinitnext, rnp->qsmaskinitnext | mask);
> abfce0414814149 Wei Yang 2020-04-19 4299 newcpu =
!(rnp->expmaskinitnext & mask);
> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4300 rnp->expmaskinitnext |=
mask;
> 313517fc44fb2d8 Paul E. McKenney 2017-06-08 4301 /* Allow lockless access
for expedited grace periods. */
> abfce0414814149 Wei Yang 2020-04-19 4302
smp_store_release(&rcu_state.ncpus, rcu_state.ncpus + newcpu); /* ^^^ */
> 2f08469563550d1 Paul E. McKenney 2020-02-10 4303
ASSERT_EXCLUSIVE_WRITER(rcu_state.ncpus);
> e05121ba5b81e2f Paul E. McKenney 2018-05-07 4304 rcu_gpnum_ovf(rnp, rdp);
/* Offline-induced counter wrap? */
> eb7a6653887b540 Paul E. McKenney 2018-07-05 4305 rdp->rcu_onl_gp_seq =
READ_ONCE(rcu_state.gp_seq);
> eb7a6653887b540 Paul E. McKenney 2018-07-05 4306 rdp->rcu_onl_gp_flags =
READ_ONCE(rcu_state.gp_flags);
> 9f866dac94292f9 Joel Fernandes (Google 2020-09-29 4307)
> 9f866dac94292f9 Joel Fernandes (Google 2020-09-29 4308) /* An incoming CPU should
never be blocking a grace period. */
> 9f866dac94292f9 Joel Fernandes (Google 2020-09-29 4309) if
(WARN_ON_ONCE(rnp->qsmask & mask)) { /* RCU waiting on incoming CPU? */
> 516e5ae0c940162 Joel Fernandes (Google 2019-09-05 4310)
rcu_disable_urgency_upon_qs(rdp);
> 99990da1b3c00f6 Paul E. McKenney 2018-05-03 4311 /* Report QS -after-
changing ->qsmaskinitnext! */
> b50912d0b5e03f1 Paul E. McKenney 2018-07-03 @4312 rcu_report_qs_rnp(mask,
rnp, rnp->gp_seq, flags);
> 99990da1b3c00f6 Paul E. McKenney 2018-05-03 4313 } else {
> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4314
raw_spin_unlock_irqrestore_rcu_node(rnp, flags);
> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4315 }
> 342f2d0f88f3c2d David Woodhouse 2021-12-09 4316
arch_spin_unlock(&rcu_state.ofl_lock);
> 342f2d0f88f3c2d David Woodhouse 2021-12-09 4317
local_irq_restore(seq_flags);
> 313517fc44fb2d8 Paul E. McKenney 2017-06-08 4318 smp_mb(); /* Ensure RCU
read-side usage follows above initialization. */
> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4319 }
> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4320
>
> :::::: The code at line 4312 was first introduced by commit
> :::::: b50912d0b5e03f11004fec1e2b50244de9e2fa41 rcu: Remove rsp parameter from
rcu_report_qs_rnp()
>
> :::::: TO: Paul E. McKenney <paulmck(a)linux.vnet.ibm.com>
> :::::: CC: Paul E. McKenney <paulmck(a)linux.vnet.ibm.com>
>
> ---
> 0-DAY CI Kernel Test Service, Intel Corporation
> https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org
7:36 p.m.
On 12/10/2021 9:57 PM, Paul E. McKenney wrote:
On Fri, Dec 10, 2021 at 03:30:22PM +0800, kernel test robot wrote:
> tree: https://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu.git
dev.2021.12.07a
> head: 89ddc546d3ad541bc89ec553c92036c98aef57a6
> commit: 342f2d0f88f3c2d7ab2ae206be5846579319e3bf [56/57] rcu: Kill rnp->ofl_seq
and use only rcu_state.ofl_lock for exclusion
> config: hexagon-randconfig-r041-20211210
(https://download.01.org/0day-ci/archive/20211210/202112101543.wHvqhgYk-lk...)
> compiler: clang version 14.0.0 (https://github.com/llvm/llvm-project
097a1cb1d5ebb3a0ec4bcaed8ba3ff6a8e33c00a)
> reproduce (this is a W=1 build):
> wget
https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O
~/bin/make.cross
> chmod +x ~/bin/make.cross
> #
https://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu.git/com...
> git remote add paulmck-rcu
https://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu.git
> git fetch --no-tags paulmck-rcu dev.2021.12.07a
> git checkout 342f2d0f88f3c2d7ab2ae206be5846579319e3bf
> # save the config file to linux build tree
> mkdir build_dir
> COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross W=1 O=build_dir
ARCH=hexagon SHELL=/bin/bash
>
> If you fix the issue, kindly add following tag as appropriate
> Reported-by: kernel test robot <lkp(a)intel.com>
>
> All warnings (new ones prefixed by >>):
>
>>> kernel/rcu/tree.c:4312:45: warning: variable 'flags' is uninitialized
when used here [-Wuninitialized]
> rcu_report_qs_rnp(mask, rnp, rnp->gp_seq, flags);
> ^~~~~
> kernel/rcu/tree.c:4280:21: note: initialize the variable 'flags' to
silence this warning
> unsigned long flags, seq_flags;
> ^
> = 0
> 1 warning generated.
Good catch, but Neeraj and David beat you to this. The updated commit
is defd1f8a1f4e ("rcu: Kill rnp->ofl_seq and use only rcu_state.ofl_lock
for exclusion").
One thing that might be helpful to you during times of backlock... In the
-rcu tree, if you have several branches with the same prefix (for example,
dev.2021.12.03a, dev.2021.12.07a, and dev.2021.12.09a), you can safely
ignore all but the last one in lexicographic order. In this example,
dev.2021.12.09a obsoletes dev.2021.12.03a and dev.2021.12.07a, so you
would lose nothing by just ignoring the earlier two branches.
Thanx, Paul
Hi Paul,
Thanks for the advice, We'll investigate it in the near future to avoid
sending old issues that already fixed in the latest branch.
Best Regards,
Rong Chen
>
>> vim +/flags +4312 kernel/rcu/tree.c
>>
>> 4df8374254ea929 Thomas Gleixner 2016-07-13 4266
>> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4267 /*
>> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4268 * Mark the specified
CPU as being online so that subsequent grace periods
>> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4269 * (both expedited and
normal) will wait on it. Note that this means that
>> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4270 * incoming CPUs are
not allowed to use RCU read-side critical sections
>> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4271 * until this function
is called. Failing to observe this restriction
>> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4272 * will result in
lockdep splats.
>> deb34f364398083 Paul E. McKenney 2017-03-23 4273 *
>> deb34f364398083 Paul E. McKenney 2017-03-23 4274 * Note that this
function is special in that it is invoked directly
>> deb34f364398083 Paul E. McKenney 2017-03-23 4275 * from the incoming
CPU rather than from the cpuhp_step mechanism.
>> deb34f364398083 Paul E. McKenney 2017-03-23 4276 * This is because this
function must be invoked at a precise location.
>> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4277 */
>> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4278 void
rcu_cpu_starting(unsigned int cpu)
>> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4279 {
>> 342f2d0f88f3c2d David Woodhouse 2021-12-09 4280 unsigned long flags,
seq_flags;
>> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4281 unsigned long mask;
>> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4282 struct rcu_data *rdp;
>> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4283 struct rcu_node *rnp;
>> abfce0414814149 Wei Yang 2020-04-19 4284 bool newcpu;
>> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4285
>> c0f97f20e5d97a1 Paul E. McKenney 2020-07-24 4286 rdp =
per_cpu_ptr(&rcu_data, cpu);
>> c0f97f20e5d97a1 Paul E. McKenney 2020-07-24 4287 if
(rdp->cpu_started)
>> f64c6013a202931 Peter Zijlstra 2018-05-22 4288 return;
>> c0f97f20e5d97a1 Paul E. McKenney 2020-07-24 4289 rdp->cpu_started =
true;
>> f64c6013a202931 Peter Zijlstra 2018-05-22 4290
>> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4291 rnp = rdp->mynode;
>> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4292 mask =
rdp->grpmask;
>> 342f2d0f88f3c2d David Woodhouse 2021-12-09 4293
local_irq_save(seq_flags);
>> 342f2d0f88f3c2d David Woodhouse 2021-12-09 4294
arch_spin_lock(&rcu_state.ofl_lock);
>> 2caebefb00f03b5 Paul E. McKenney 2021-07-28 4295
rcu_dynticks_eqs_online();
>> 4d60b475f858ebd Paul E. McKenney 2020-10-13 4296 smp_mb(); // Pair with
rcu_gp_cleanup()'s ->ofl_seq barrier().
>> 342f2d0f88f3c2d David Woodhouse 2021-12-09 4297
raw_spin_lock_rcu_node(rnp);
>> 105abf82b0a6266 Paul E. McKenney 2020-01-03 4298
WRITE_ONCE(rnp->qsmaskinitnext, rnp->qsmaskinitnext | mask);
>> abfce0414814149 Wei Yang 2020-04-19 4299 newcpu =
!(rnp->expmaskinitnext & mask);
>> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4300
rnp->expmaskinitnext |= mask;
>> 313517fc44fb2d8 Paul E. McKenney 2017-06-08 4301 /* Allow lockless
access for expedited grace periods. */
>> abfce0414814149 Wei Yang 2020-04-19 4302
smp_store_release(&rcu_state.ncpus, rcu_state.ncpus + newcpu); /* ^^^ */
>> 2f08469563550d1 Paul E. McKenney 2020-02-10 4303
ASSERT_EXCLUSIVE_WRITER(rcu_state.ncpus);
>> e05121ba5b81e2f Paul E. McKenney 2018-05-07 4304 rcu_gpnum_ovf(rnp,
rdp); /* Offline-induced counter wrap? */
>> eb7a6653887b540 Paul E. McKenney 2018-07-05 4305 rdp->rcu_onl_gp_seq
= READ_ONCE(rcu_state.gp_seq);
>> eb7a6653887b540 Paul E. McKenney 2018-07-05 4306
rdp->rcu_onl_gp_flags = READ_ONCE(rcu_state.gp_flags);
>> 9f866dac94292f9 Joel Fernandes (Google 2020-09-29 4307)
>> 9f866dac94292f9 Joel Fernandes (Google 2020-09-29 4308) /* An incoming CPU
should never be blocking a grace period. */
>> 9f866dac94292f9 Joel Fernandes (Google 2020-09-29 4309) if
(WARN_ON_ONCE(rnp->qsmask & mask)) { /* RCU waiting on incoming CPU? */
>> 516e5ae0c940162 Joel Fernandes (Google 2019-09-05 4310)
rcu_disable_urgency_upon_qs(rdp);
>> 99990da1b3c00f6 Paul E. McKenney 2018-05-03 4311 /* Report QS -after-
changing ->qsmaskinitnext! */
>> b50912d0b5e03f1 Paul E. McKenney 2018-07-03 @4312
rcu_report_qs_rnp(mask, rnp, rnp->gp_seq, flags);
>> 99990da1b3c00f6 Paul E. McKenney 2018-05-03 4313 } else {
>> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4314
raw_spin_unlock_irqrestore_rcu_node(rnp, flags);
>> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4315 }
>> 342f2d0f88f3c2d David Woodhouse 2021-12-09 4316
arch_spin_unlock(&rcu_state.ofl_lock);
>> 342f2d0f88f3c2d David Woodhouse 2021-12-09 4317
local_irq_restore(seq_flags);
>> 313517fc44fb2d8 Paul E. McKenney 2017-06-08 4318 smp_mb(); /* Ensure
RCU read-side usage follows above initialization. */
>> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4319 }
>> 7ec99de36f40261 Paul E. McKenney 2016-06-30 4320
>>
>> :::::: The code at line 4312 was first introduced by commit
>> :::::: b50912d0b5e03f11004fec1e2b50244de9e2fa41 rcu: Remove rsp parameter from
rcu_report_qs_rnp()
>>
>> :::::: TO: Paul E. McKenney <paulmck(a)linux.vnet.ibm.com>
>> :::::: CC: Paul E. McKenney <paulmck(a)linux.vnet.ibm.com>
>>
>> ---
>> 0-DAY CI Kernel Test Service, Intel Corporation
>> https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org
> _______________________________________________
> kbuild-all mailing list -- kbuild-all(a)lists.01.org
> To unsubscribe send an email to kbuild-all-leave(a)lists.01.org
>
7:02 a.m.
On Mon, Dec 13, 2021 at 04:36:36PM +0800, Chen, Rong A wrote:
On 12/10/2021 9:57 PM, Paul E. McKenney wrote:
> On Fri, Dec 10, 2021 at 03:30:22PM +0800, kernel test robot wrote:
> > tree: https://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu.git
dev.2021.12.07a
> > head: 89ddc546d3ad541bc89ec553c92036c98aef57a6
> > commit: 342f2d0f88f3c2d7ab2ae206be5846579319e3bf [56/57] rcu: Kill
rnp->ofl_seq and use only rcu_state.ofl_lock for exclusion
> > config: hexagon-randconfig-r041-20211210
(https://download.01.org/0day-ci/archive/20211210/202112101543.wHvqhgYk-lk...)
> > compiler: clang version 14.0.0 (https://github.com/llvm/llvm-project
097a1cb1d5ebb3a0ec4bcaed8ba3ff6a8e33c00a)
> > reproduce (this is a W=1 build):
> > wget
https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O
~/bin/make.cross
> > chmod +x ~/bin/make.cross
> > #
https://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu.git/com...
> > git remote add paulmck-rcu
https://git.kernel.org/pub/scm/linux/kernel/git/paulmck/linux-rcu.git
> > git fetch --no-tags paulmck-rcu dev.2021.12.07a
> > git checkout 342f2d0f88f3c2d7ab2ae206be5846579319e3bf
> > # save the config file to linux build tree
> > mkdir build_dir
> > COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross W=1
O=build_dir ARCH=hexagon SHELL=/bin/bash
> >
> > If you fix the issue, kindly add following tag as appropriate
> > Reported-by: kernel test robot <lkp(a)intel.com>
> >
> > All warnings (new ones prefixed by >>):
> >
> > > > kernel/rcu/tree.c:4312:45: warning: variable 'flags' is
uninitialized when used here [-Wuninitialized]
> > rcu_report_qs_rnp(mask, rnp, rnp->gp_seq, flags);
> > ^~~~~
> > kernel/rcu/tree.c:4280:21: note: initialize the variable 'flags' to
silence this warning
> > unsigned long flags, seq_flags;
> > ^
> > = 0
> > 1 warning generated.
>
> Good catch, but Neeraj and David beat you to this. The updated commit
> is defd1f8a1f4e ("rcu: Kill rnp->ofl_seq and use only rcu_state.ofl_lock
> for exclusion").
>
> One thing that might be helpful to you during times of backlock... In the
> -rcu tree, if you have several branches with the same prefix (for example,
> dev.2021.12.03a, dev.2021.12.07a, and dev.2021.12.09a), you can safely
> ignore all but the last one in lexicographic order. In this example,
> dev.2021.12.09a obsoletes dev.2021.12.03a and dev.2021.12.07a, so you
> would lose nothing by just ignoring the earlier two branches.
>
> Thanx, Paul
Hi Paul,
Thanks for the advice, We'll investigate it in the near future to avoid
sending old issues that already fixed in the latest branch.
If you have already done the testing, there is no problem sending me the
reports. But if you are behind, why needlessly get more behind? ;-)
Thanx, Paul
> Best Regards,
> Rong Chen
>
> >
> > > vim +/flags +4312 kernel/rcu/tree.c
> > >
> > > 4df8374254ea929 Thomas Gleixner 2016-07-13 4266
> > > 7ec99de36f40261 Paul E. McKenney 2016-06-30 4267 /*
> > > 7ec99de36f40261 Paul E. McKenney 2016-06-30 4268 * Mark the
specified CPU as being online so that subsequent grace periods
> > > 7ec99de36f40261 Paul E. McKenney 2016-06-30 4269 * (both
expedited and normal) will wait on it. Note that this means that
> > > 7ec99de36f40261 Paul E. McKenney 2016-06-30 4270 * incoming CPUs
are not allowed to use RCU read-side critical sections
> > > 7ec99de36f40261 Paul E. McKenney 2016-06-30 4271 * until this
function is called. Failing to observe this restriction
> > > 7ec99de36f40261 Paul E. McKenney 2016-06-30 4272 * will result in
lockdep splats.
> > > deb34f364398083 Paul E. McKenney 2017-03-23 4273 *
> > > deb34f364398083 Paul E. McKenney 2017-03-23 4274 * Note that this
function is special in that it is invoked directly
> > > deb34f364398083 Paul E. McKenney 2017-03-23 4275 * from the
incoming CPU rather than from the cpuhp_step mechanism.
> > > deb34f364398083 Paul E. McKenney 2017-03-23 4276 * This is
because this function must be invoked at a precise location.
> > > 7ec99de36f40261 Paul E. McKenney 2016-06-30 4277 */
> > > 7ec99de36f40261 Paul E. McKenney 2016-06-30 4278 void
rcu_cpu_starting(unsigned int cpu)
> > > 7ec99de36f40261 Paul E. McKenney 2016-06-30 4279 {
> > > 342f2d0f88f3c2d David Woodhouse 2021-12-09 4280 unsigned long
flags, seq_flags;
> > > 7ec99de36f40261 Paul E. McKenney 2016-06-30 4281 unsigned long
mask;
> > > 7ec99de36f40261 Paul E. McKenney 2016-06-30 4282 struct rcu_data
*rdp;
> > > 7ec99de36f40261 Paul E. McKenney 2016-06-30 4283 struct rcu_node
*rnp;
> > > abfce0414814149 Wei Yang 2020-04-19 4284 bool newcpu;
> > > 7ec99de36f40261 Paul E. McKenney 2016-06-30 4285
> > > c0f97f20e5d97a1 Paul E. McKenney 2020-07-24 4286 rdp =
per_cpu_ptr(&rcu_data, cpu);
> > > c0f97f20e5d97a1 Paul E. McKenney 2020-07-24 4287 if
(rdp->cpu_started)
> > > f64c6013a202931 Peter Zijlstra 2018-05-22 4288 return;
> > > c0f97f20e5d97a1 Paul E. McKenney 2020-07-24 4289
rdp->cpu_started = true;
> > > f64c6013a202931 Peter Zijlstra 2018-05-22 4290
> > > 7ec99de36f40261 Paul E. McKenney 2016-06-30 4291 rnp =
rdp->mynode;
> > > 7ec99de36f40261 Paul E. McKenney 2016-06-30 4292 mask =
rdp->grpmask;
> > > 342f2d0f88f3c2d David Woodhouse 2021-12-09 4293
local_irq_save(seq_flags);
> > > 342f2d0f88f3c2d David Woodhouse 2021-12-09 4294
arch_spin_lock(&rcu_state.ofl_lock);
> > > 2caebefb00f03b5 Paul E. McKenney 2021-07-28 4295
rcu_dynticks_eqs_online();
> > > 4d60b475f858ebd Paul E. McKenney 2020-10-13 4296 smp_mb(); //
Pair with rcu_gp_cleanup()'s ->ofl_seq barrier().
> > > 342f2d0f88f3c2d David Woodhouse 2021-12-09 4297
raw_spin_lock_rcu_node(rnp);
> > > 105abf82b0a6266 Paul E. McKenney 2020-01-03 4298
WRITE_ONCE(rnp->qsmaskinitnext, rnp->qsmaskinitnext | mask);
> > > abfce0414814149 Wei Yang 2020-04-19 4299 newcpu =
!(rnp->expmaskinitnext & mask);
> > > 7ec99de36f40261 Paul E. McKenney 2016-06-30 4300
rnp->expmaskinitnext |= mask;
> > > 313517fc44fb2d8 Paul E. McKenney 2017-06-08 4301 /* Allow
lockless access for expedited grace periods. */
> > > abfce0414814149 Wei Yang 2020-04-19 4302
smp_store_release(&rcu_state.ncpus, rcu_state.ncpus + newcpu); /* ^^^ */
> > > 2f08469563550d1 Paul E. McKenney 2020-02-10 4303
ASSERT_EXCLUSIVE_WRITER(rcu_state.ncpus);
> > > e05121ba5b81e2f Paul E. McKenney 2018-05-07 4304
rcu_gpnum_ovf(rnp, rdp); /* Offline-induced counter wrap? */
> > > eb7a6653887b540 Paul E. McKenney 2018-07-05 4305
rdp->rcu_onl_gp_seq = READ_ONCE(rcu_state.gp_seq);
> > > eb7a6653887b540 Paul E. McKenney 2018-07-05 4306
rdp->rcu_onl_gp_flags = READ_ONCE(rcu_state.gp_flags);
> > > 9f866dac94292f9 Joel Fernandes (Google 2020-09-29 4307)
> > > 9f866dac94292f9 Joel Fernandes (Google 2020-09-29 4308) /* An incoming
CPU should never be blocking a grace period. */
> > > 9f866dac94292f9 Joel Fernandes (Google 2020-09-29 4309) if
(WARN_ON_ONCE(rnp->qsmask & mask)) { /* RCU waiting on incoming CPU? */
> > > 516e5ae0c940162 Joel Fernandes (Google 2019-09-05 4310)
rcu_disable_urgency_upon_qs(rdp);
> > > 99990da1b3c00f6 Paul E. McKenney 2018-05-03 4311 /* Report QS
-after- changing ->qsmaskinitnext! */
> > > b50912d0b5e03f1 Paul E. McKenney 2018-07-03 @4312
rcu_report_qs_rnp(mask, rnp, rnp->gp_seq, flags);
> > > 99990da1b3c00f6 Paul E. McKenney 2018-05-03 4313 } else {
> > > 7ec99de36f40261 Paul E. McKenney 2016-06-30 4314
raw_spin_unlock_irqrestore_rcu_node(rnp, flags);
> > > 7ec99de36f40261 Paul E. McKenney 2016-06-30 4315 }
> > > 342f2d0f88f3c2d David Woodhouse 2021-12-09 4316
arch_spin_unlock(&rcu_state.ofl_lock);
> > > 342f2d0f88f3c2d David Woodhouse 2021-12-09 4317
local_irq_restore(seq_flags);
> > > 313517fc44fb2d8 Paul E. McKenney 2017-06-08 4318 smp_mb(); /*
Ensure RCU read-side usage follows above initialization. */
> > > 7ec99de36f40261 Paul E. McKenney 2016-06-30 4319 }
> > > 7ec99de36f40261 Paul E. McKenney 2016-06-30 4320
> > >
> > > :::::: The code at line 4312 was first introduced by commit
> > > :::::: b50912d0b5e03f11004fec1e2b50244de9e2fa41 rcu: Remove rsp parameter
from rcu_report_qs_rnp()
> > >
> > > :::::: TO: Paul E. McKenney <paulmck(a)linux.vnet.ibm.com>
> > > :::::: CC: Paul E. McKenney <paulmck(a)linux.vnet.ibm.com>
> > >
> > > ---
> > > 0-DAY CI Kernel Test Service, Intel Corporation
> > > https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org
> > _______________________________________________
> > kbuild-all mailing list -- kbuild-all(a)lists.01.org
> > To unsubscribe send an email to kbuild-all-leave(a)lists.01.org
> >
281
days inactive
284
days old
3 comments
3 participants
participants (3)
-
Chen, Rong A -
kernel test robot -
Paul E. McKenney