| 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
| ||
|
Message-ID: <201604130658.n9ttDvqp%fengguang.wu@intel.com>
Date: Wed, 13 Apr 2016 06:39:02 +0800
From: kbuild test robot <lkp@...el.com>
To: Jason Low <jason.low2@...com>
Cc: kbuild-all@...org, Peter Zijlstra <peterz@...radead.org>,
Will Deacon <will.deacon@....com>,
Linus Torvalds <torvalds@...ux-foundation.org>,
linux-kernel@...r.kernel.org, mingo@...hat.com,
paulmck@...ux.vnet.ibm.com, terry.rudd@....com,
waiman.long@....com, boqun.feng@...il.com, dave@...olabs.net,
jason.low2@...com
Subject: Re: [PATCH] MCS spinlock: Use smp_cond_load_acquire()
Hi Jason,
[auto build test ERROR on v4.6-rc3]
[also build test ERROR on next-20160412]
[cannot apply to tip/core/locking]
[if your patch is applied to the wrong git tree, please drop us a note to help improving the system]
url: https://github.com/0day-ci/linux/commits/Jason-Low/MCS-spinlock-Use-smp_cond_load_acquire/20160413-053726
config: i386-randconfig-s0-201615 (attached as .config)
reproduce:
# save the attached .config to linux build tree
make ARCH=i386
All error/warnings (new ones prefixed by >>):
In file included from kernel/locking/qspinlock.c:70:0:
kernel/locking/mcs_spinlock.h: In function 'mcs_spin_lock':
>> kernel/locking/mcs_spinlock.h:31:2: error: implicit declaration of function 'smp_cond_load_acquire' [-Werror=implicit-function-declaration]
smp_cond_load_acquire(&l, VAL); \
^
>> kernel/locking/mcs_spinlock.h:91:2: note: in expansion of macro 'arch_mcs_spin_lock_contended'
arch_mcs_spin_lock_contended(&node->locked);
^
>> kernel/locking/mcs_spinlock.h:31:24: error: lvalue required as unary '&' operand
smp_cond_load_acquire(&l, VAL); \
^
>> kernel/locking/mcs_spinlock.h:91:2: note: in expansion of macro 'arch_mcs_spin_lock_contended'
arch_mcs_spin_lock_contended(&node->locked);
^
>> kernel/locking/mcs_spinlock.h:31:28: error: 'VAL' undeclared (first use in this function)
smp_cond_load_acquire(&l, VAL); \
^
>> kernel/locking/mcs_spinlock.h:91:2: note: in expansion of macro 'arch_mcs_spin_lock_contended'
arch_mcs_spin_lock_contended(&node->locked);
^
kernel/locking/mcs_spinlock.h:31:28: note: each undeclared identifier is reported only once for each function it appears in
smp_cond_load_acquire(&l, VAL); \
^
>> kernel/locking/mcs_spinlock.h:91:2: note: in expansion of macro 'arch_mcs_spin_lock_contended'
arch_mcs_spin_lock_contended(&node->locked);
^
kernel/locking/qspinlock.c: In function 'native_queued_spin_lock_slowpath':
>> kernel/locking/mcs_spinlock.h:31:24: error: lvalue required as unary '&' operand
smp_cond_load_acquire(&l, VAL); \
^
>> kernel/locking/qspinlock.c:411:3: note: in expansion of macro 'arch_mcs_spin_lock_contended'
arch_mcs_spin_lock_contended(&node->locked);
^
>> kernel/locking/mcs_spinlock.h:31:28: error: 'VAL' undeclared (first use in this function)
smp_cond_load_acquire(&l, VAL); \
^
>> kernel/locking/qspinlock.c:411:3: note: in expansion of macro 'arch_mcs_spin_lock_contended'
arch_mcs_spin_lock_contended(&node->locked);
^
kernel/locking/qspinlock.c: In function '__pv_queued_spin_lock_slowpath':
>> kernel/locking/mcs_spinlock.h:31:24: error: lvalue required as unary '&' operand
smp_cond_load_acquire(&l, VAL); \
^
>> kernel/locking/qspinlock.c:411:3: note: in expansion of macro 'arch_mcs_spin_lock_contended'
arch_mcs_spin_lock_contended(&node->locked);
^
>> kernel/locking/mcs_spinlock.h:31:28: error: 'VAL' undeclared (first use in this function)
smp_cond_load_acquire(&l, VAL); \
^
>> kernel/locking/qspinlock.c:411:3: note: in expansion of macro 'arch_mcs_spin_lock_contended'
arch_mcs_spin_lock_contended(&node->locked);
^
cc1: some warnings being treated as errors
vim +/smp_cond_load_acquire +31 kernel/locking/mcs_spinlock.h
25 * Using smp_cond_load_acquire() provides the acquire semantics
26 * required so that subsequent operations happen after the
27 * lock is acquired.
28 */
29 #define arch_mcs_spin_lock_contended(l) \
30 do { \
> 31 smp_cond_load_acquire(&l, VAL); \
32 } while (0)
33 #endif
34
35 #ifndef arch_mcs_spin_unlock_contended
36 /*
37 * smp_store_release() provides a memory barrier to ensure all
38 * operations in the critical section has been completed before
39 * unlocking.
40 */
41 #define arch_mcs_spin_unlock_contended(l) \
42 smp_store_release((l), 1)
43 #endif
44
45 /*
46 * Note: the smp_load_acquire/smp_store_release pair is not
47 * sufficient to form a full memory barrier across
48 * cpus for many architectures (except x86) for mcs_unlock and mcs_lock.
49 * For applications that need a full barrier across multiple cpus
50 * with mcs_unlock and mcs_lock pair, smp_mb__after_unlock_lock() should be
51 * used after mcs_lock.
52 */
53
54 /*
55 * In order to acquire the lock, the caller should declare a local node and
56 * pass a reference of the node to this function in addition to the lock.
57 * If the lock has already been acquired, then this will proceed to spin
58 * on this node->locked until the previous lock holder sets the node->locked
59 * in mcs_spin_unlock().
60 */
61 static inline
62 void mcs_spin_lock(struct mcs_spinlock **lock, struct mcs_spinlock *node)
63 {
64 struct mcs_spinlock *prev;
65
66 /* Init node */
67 node->locked = 0;
68 node->next = NULL;
69
70 /*
71 * We rely on the full barrier with global transitivity implied by the
72 * below xchg() to order the initialization stores above against any
73 * observation of @node. And to provide the ACQUIRE ordering associated
74 * with a LOCK primitive.
75 */
76 prev = xchg(lock, node);
77 if (likely(prev == NULL)) {
78 /*
79 * Lock acquired, don't need to set node->locked to 1. Threads
80 * only spin on its own node->locked value for lock acquisition.
81 * However, since this thread can immediately acquire the lock
82 * and does not proceed to spin on its own node->locked, this
83 * value won't be used. If a debug mode is needed to
84 * audit lock status, then set node->locked value here.
85 */
86 return;
87 }
88 WRITE_ONCE(prev->next, node);
89
90 /* Wait until the lock holder passes the lock down. */
> 91 arch_mcs_spin_lock_contended(&node->locked);
92 }
93
94 /*
---
0-DAY kernel test infrastructure Open Source Technology Center
https://lists.01.org/pipermail/kbuild-all Intel Corporation
Download attachment ".config.gz" of type "application/octet-stream" (32276 bytes)
Powered by blists - more mailing lists