| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Date: Tue, 26 Dec 2023 04:05:38 +0800
From: kernel test robot <lkp@...el.com>
To: Waiman Long <longman@...hat.com>
Cc: oe-kbuild-all@...ts.linux.dev, linux-kernel@...r.kernel.org,
Ingo Molnar <mingo@...nel.org>
Subject: kernel/locking/qspinlock_paravirt.h:298:14: warning: variable
'wait_early' set but not used
tree: https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git master
head: 861deac3b092f37b2c5e6871732f3e11486f7082
commit: ad53fa10fa9e816067bbae7109845940f5e6df50 locking/qspinlock_stat: Introduce generic lockevent_*() counting APIs
date: 4 years, 9 months ago
config: x86_64-buildonly-randconfig-006-20230826 (https://download.01.org/0day-ci/archive/20231226/202312260422.f4pK3f9m-lkp@intel.com/config)
compiler: gcc-12 (Debian 12.2.0-14) 12.2.0
reproduce (this is a W=1 build): (https://download.01.org/0day-ci/archive/20231226/202312260422.f4pK3f9m-lkp@intel.com/reproduce)
If you fix the issue in a separate patch/commit (i.e. not just a new version of
the same patch/commit), kindly add following tags
| Reported-by: kernel test robot <lkp@...el.com>
| Closes: https://lore.kernel.org/oe-kbuild-all/202312260422.f4pK3f9m-lkp@intel.com/
All warnings (new ones prefixed by >>):
In file included from kernel/locking/qspinlock.c:589:
kernel/locking/qspinlock_paravirt.h: In function 'pv_wait_node':
>> kernel/locking/qspinlock_paravirt.h:298:14: warning: variable 'wait_early' set but not used [-Wunused-but-set-variable]
298 | bool wait_early;
| ^~~~~~~~~~
kernel/locking/qspinlock_paravirt.h: At top level:
kernel/locking/qspinlock_paravirt.h:493:1: warning: no previous prototype for '__pv_queued_spin_unlock_slowpath' [-Wmissing-prototypes]
493 | __pv_queued_spin_unlock_slowpath(struct qspinlock *lock, u8 locked)
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
vim +/wait_early +298 kernel/locking/qspinlock_paravirt.h
a23db284fe0d18 Waiman Long 2015-04-24 287
a23db284fe0d18 Waiman Long 2015-04-24 288 /*
a23db284fe0d18 Waiman Long 2015-04-24 289 * Wait for node->locked to become true, halt the vcpu after a short spin.
75d2270280686b Waiman Long 2015-07-11 290 * pv_kick_node() is used to set _Q_SLOW_VAL and fill in hash table on its
75d2270280686b Waiman Long 2015-07-11 291 * behalf.
a23db284fe0d18 Waiman Long 2015-04-24 292 */
cd0272fab78507 Waiman Long 2015-11-09 293 static void pv_wait_node(struct mcs_spinlock *node, struct mcs_spinlock *prev)
a23db284fe0d18 Waiman Long 2015-04-24 294 {
a23db284fe0d18 Waiman Long 2015-04-24 295 struct pv_node *pn = (struct pv_node *)node;
cd0272fab78507 Waiman Long 2015-11-09 296 struct pv_node *pp = (struct pv_node *)prev;
a23db284fe0d18 Waiman Long 2015-04-24 297 int loop;
cd0272fab78507 Waiman Long 2015-11-09 @298 bool wait_early;
a23db284fe0d18 Waiman Long 2015-04-24 299
08be8f63c40c03 Waiman Long 2016-05-31 300 for (;;) {
cd0272fab78507 Waiman Long 2015-11-09 301 for (wait_early = false, loop = SPIN_THRESHOLD; loop; loop--) {
a23db284fe0d18 Waiman Long 2015-04-24 302 if (READ_ONCE(node->locked))
a23db284fe0d18 Waiman Long 2015-04-24 303 return;
cd0272fab78507 Waiman Long 2015-11-09 304 if (pv_wait_early(pp, loop)) {
cd0272fab78507 Waiman Long 2015-11-09 305 wait_early = true;
cd0272fab78507 Waiman Long 2015-11-09 306 break;
cd0272fab78507 Waiman Long 2015-11-09 307 }
a23db284fe0d18 Waiman Long 2015-04-24 308 cpu_relax();
a23db284fe0d18 Waiman Long 2015-04-24 309 }
a23db284fe0d18 Waiman Long 2015-04-24 310
a23db284fe0d18 Waiman Long 2015-04-24 311 /*
a23db284fe0d18 Waiman Long 2015-04-24 312 * Order pn->state vs pn->locked thusly:
a23db284fe0d18 Waiman Long 2015-04-24 313 *
a23db284fe0d18 Waiman Long 2015-04-24 314 * [S] pn->state = vcpu_halted [S] next->locked = 1
a23db284fe0d18 Waiman Long 2015-04-24 315 * MB MB
75d2270280686b Waiman Long 2015-07-11 316 * [L] pn->locked [RmW] pn->state = vcpu_hashed
a23db284fe0d18 Waiman Long 2015-04-24 317 *
75d2270280686b Waiman Long 2015-07-11 318 * Matches the cmpxchg() from pv_kick_node().
a23db284fe0d18 Waiman Long 2015-04-24 319 */
b92b8b35a2e38b Peter Zijlstra 2015-05-12 320 smp_store_mb(pn->state, vcpu_halted);
a23db284fe0d18 Waiman Long 2015-04-24 321
45e898b735620f Waiman Long 2015-11-09 322 if (!READ_ONCE(node->locked)) {
ad53fa10fa9e81 Waiman Long 2019-04-04 323 lockevent_inc(pv_wait_node);
ad53fa10fa9e81 Waiman Long 2019-04-04 324 lockevent_cond_inc(pv_wait_early, wait_early);
a23db284fe0d18 Waiman Long 2015-04-24 325 pv_wait(&pn->state, vcpu_halted);
45e898b735620f Waiman Long 2015-11-09 326 }
a23db284fe0d18 Waiman Long 2015-04-24 327
a23db284fe0d18 Waiman Long 2015-04-24 328 /*
45e898b735620f Waiman Long 2015-11-09 329 * If pv_kick_node() changed us to vcpu_hashed, retain that
1c4941fd53afb4 Waiman Long 2015-11-10 330 * value so that pv_wait_head_or_lock() knows to not also try
1c4941fd53afb4 Waiman Long 2015-11-10 331 * to hash this lock.
a23db284fe0d18 Waiman Long 2015-04-24 332 */
75d2270280686b Waiman Long 2015-07-11 333 cmpxchg(&pn->state, vcpu_halted, vcpu_running);
a23db284fe0d18 Waiman Long 2015-04-24 334
a23db284fe0d18 Waiman Long 2015-04-24 335 /*
a23db284fe0d18 Waiman Long 2015-04-24 336 * If the locked flag is still not set after wakeup, it is a
a23db284fe0d18 Waiman Long 2015-04-24 337 * spurious wakeup and the vCPU should wait again. However,
a23db284fe0d18 Waiman Long 2015-04-24 338 * there is a pretty high overhead for CPU halting and kicking.
a23db284fe0d18 Waiman Long 2015-04-24 339 * So it is better to spin for a while in the hope that the
a23db284fe0d18 Waiman Long 2015-04-24 340 * MCS lock will be released soon.
a23db284fe0d18 Waiman Long 2015-04-24 341 */
ad53fa10fa9e81 Waiman Long 2019-04-04 342 lockevent_cond_inc(pv_spurious_wakeup,
ad53fa10fa9e81 Waiman Long 2019-04-04 343 !READ_ONCE(node->locked));
a23db284fe0d18 Waiman Long 2015-04-24 344 }
75d2270280686b Waiman Long 2015-07-11 345
a23db284fe0d18 Waiman Long 2015-04-24 346 /*
a23db284fe0d18 Waiman Long 2015-04-24 347 * By now our node->locked should be 1 and our caller will not actually
a23db284fe0d18 Waiman Long 2015-04-24 348 * spin-wait for it. We do however rely on our caller to do a
a23db284fe0d18 Waiman Long 2015-04-24 349 * load-acquire for us.
a23db284fe0d18 Waiman Long 2015-04-24 350 */
a23db284fe0d18 Waiman Long 2015-04-24 351 }
a23db284fe0d18 Waiman Long 2015-04-24 352
:::::: The code at line 298 was first introduced by commit
:::::: cd0272fab785077c121aa91ec2401090965bbc37 locking/pvqspinlock: Queue node adaptive spinning
:::::: TO: Waiman Long <Waiman.Long@....com>
:::::: CC: Ingo Molnar <mingo@...nel.org>
--
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests/wiki
Powered by blists - more mailing lists