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
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <202211291341.VgUcDmXr-lkp@intel.com>
Date:   Tue, 29 Nov 2022 13:16:27 +0800
From:   kernel test robot <lkp@...el.com>
To:     Sebastian Andrzej Siewior <bigeasy@...utronix.de>,
        Will Deacon <will@...nel.org>, Jan Kara <jack@...e.cz>
Cc:     llvm@...ts.linux.dev, oe-kbuild-all@...ts.linux.dev,
        Waiman Long <longman@...hat.com>,
        LKML <linux-kernel@...r.kernel.org>,
        Thomas Gleixner <tglx@...utronix.de>,
        Steven Rostedt <rostedt@...dmis.org>,
        Mel Gorman <mgorman@...e.de>,
        Peter Zijlstra <peterz@...radead.org>,
        Ingo Molnar <mingo@...hat.com>,
        Catalin Marinas <catalin.marinas@....com>
Subject: Re: Re: Crash with PREEMPT_RT on aarch64 machine

Hi Sebastian,

Thank you for the patch! Yet something to improve:

[auto build test ERROR on tip/locking/core]
[also build test ERROR on linus/master v6.1-rc7 next-20221128]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch#_base_tree_information]

url:    https://github.com/intel-lab-lkp/linux/commits/Sebastian-Andrzej-Siewior/Re-Crash-with-PREEMPT_RT-on-aarch64-machine/20221129-000000
patch link:    https://lore.kernel.org/r/Y4Tapja2qq8HiHBZ%40linutronix.de
patch subject: Re: Crash with PREEMPT_RT on aarch64 machine
config: hexagon-randconfig-r041-20221128
compiler: clang version 16.0.0 (https://github.com/llvm/llvm-project 6e4cea55f0d1104408b26ac574566a0e4de48036)
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://github.com/intel-lab-lkp/linux/commit/d4aea399f0b5472dc15a4c623d99b4ae79005050
        git remote add linux-review https://github.com/intel-lab-lkp/linux
        git fetch --no-tags linux-review Sebastian-Andrzej-Siewior/Re-Crash-with-PREEMPT_RT-on-aarch64-machine/20221129-000000
        git checkout d4aea399f0b5472dc15a4c623d99b4ae79005050
        # save the config file
        mkdir build_dir && cp config build_dir/.config
        COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross W=1 O=build_dir ARCH=hexagon SHELL=/bin/bash kernel/

If you fix the issue, kindly add following tag where applicable
| Reported-by: kernel test robot <lkp@...el.com>

All error/warnings (new ones prefixed by >>):

   In file included from kernel/locking/rtmutex_api.c:9:
>> kernel/locking/rtmutex.c:100:15: error: incompatible integer to pointer conversion initializing 'typeof (*(__ai_ptr))' (aka 'struct task_struct *') with an expression of type 'unsigned long' [-Wint-conversion]
           WARN_ON_ONCE(cmpxchg_acquire(&lock->owner, RT_MUTEX_HAS_WAITERS, val) != RT_MUTEX_HAS_WAITERS);
           ~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
   include/linux/atomic/atomic-instrumented.h:1923:2: note: expanded from macro 'cmpxchg_acquire'
           arch_cmpxchg_acquire(__ai_ptr, __VA_ARGS__); \
           ^
   include/linux/atomic/atomic-arch-fallback.h:35:30: note: expanded from macro 'arch_cmpxchg_acquire'
   #define arch_cmpxchg_acquire arch_cmpxchg
                                ^
   arch/hexagon/include/asm/cmpxchg.h:57:21: note: expanded from macro 'arch_cmpxchg'
           __typeof__(*(ptr)) __old = (old);                       \
                              ^
   include/asm-generic/bug.h:180:41: note: expanded from macro 'WARN_ON_ONCE'
   #define WARN_ON_ONCE(condition) WARN_ON(condition)
                                   ~~~~~~~~^~~~~~~~~~
   include/asm-generic/bug.h:167:25: note: expanded from macro 'WARN_ON'
           int __ret_warn_on = !!(condition);                              \
                                  ^~~~~~~~~
   In file included from kernel/locking/rtmutex_api.c:9:
>> kernel/locking/rtmutex.c:100:15: error: incompatible integer to pointer conversion initializing 'typeof (*(__ai_ptr))' (aka 'struct task_struct *') with an expression of type 'unsigned long' [-Wint-conversion]
           WARN_ON_ONCE(cmpxchg_acquire(&lock->owner, RT_MUTEX_HAS_WAITERS, val) != RT_MUTEX_HAS_WAITERS);
           ~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
   include/linux/atomic/atomic-instrumented.h:1923:2: note: expanded from macro 'cmpxchg_acquire'
           arch_cmpxchg_acquire(__ai_ptr, __VA_ARGS__); \
           ^
   include/linux/atomic/atomic-arch-fallback.h:35:30: note: expanded from macro 'arch_cmpxchg_acquire'
   #define arch_cmpxchg_acquire arch_cmpxchg
                                ^
   arch/hexagon/include/asm/cmpxchg.h:58:21: note: expanded from macro 'arch_cmpxchg'
           __typeof__(*(ptr)) __new = (new);                       \
                              ^
   include/asm-generic/bug.h:180:41: note: expanded from macro 'WARN_ON_ONCE'
   #define WARN_ON_ONCE(condition) WARN_ON(condition)
                                   ~~~~~~~~^~~~~~~~~~
   include/asm-generic/bug.h:167:25: note: expanded from macro 'WARN_ON'
           int __ret_warn_on = !!(condition);                              \
                                  ^~~~~~~~~
   In file included from kernel/locking/rtmutex_api.c:9:
>> kernel/locking/rtmutex.c:100:72: warning: comparison between pointer and integer ('typeof (*(__ai_ptr))' (aka 'struct task_struct *') and 'unsigned long') [-Wpointer-integer-compare]
           WARN_ON_ONCE(cmpxchg_acquire(&lock->owner, RT_MUTEX_HAS_WAITERS, val) != RT_MUTEX_HAS_WAITERS);
           ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~
   include/asm-generic/bug.h:180:41: note: expanded from macro 'WARN_ON_ONCE'
   #define WARN_ON_ONCE(condition) WARN_ON(condition)
                                   ~~~~~~~~^~~~~~~~~~
   include/asm-generic/bug.h:167:25: note: expanded from macro 'WARN_ON'
           int __ret_warn_on = !!(condition);                              \
                                  ^~~~~~~~~
   In file included from kernel/locking/rtmutex_api.c:9:
>> kernel/locking/rtmutex.c:117:2: error: incompatible integer to pointer conversion assigning to 'volatile typeof (lock->owner)' (aka 'struct task_struct *volatile') from 'unsigned long' [-Wint-conversion]
           WRITE_ONCE(lock->owner, val);
           ^~~~~~~~~~~~~~~~~~~~~~~~~~~~
   include/asm-generic/rwonce.h:61:2: note: expanded from macro 'WRITE_ONCE'
           __WRITE_ONCE(x, val);                                           \
           ^~~~~~~~~~~~~~~~~~~~
   include/asm-generic/rwonce.h:55:30: note: expanded from macro '__WRITE_ONCE'
           *(volatile typeof(x) *)&(x) = (val);                            \
                                       ^ ~~~~~
   1 warning and 3 errors generated.


vim +100 kernel/locking/rtmutex.c

    64	
    65	/*
    66	 * lock->owner state tracking:
    67	 *
    68	 * lock->owner holds the task_struct pointer of the owner. Bit 0
    69	 * is used to keep track of the "lock has waiters" state.
    70	 *
    71	 * owner	bit0
    72	 * NULL		0	lock is free (fast acquire possible)
    73	 * NULL		1	lock is free and has waiters and the top waiter
    74	 *				is going to take the lock*
    75	 * taskpointer	0	lock is held (fast release possible)
    76	 * taskpointer	1	lock is held and has waiters**
    77	 *
    78	 * The fast atomic compare exchange based acquire and release is only
    79	 * possible when bit 0 of lock->owner is 0.
    80	 *
    81	 * (*) It also can be a transitional state when grabbing the lock
    82	 * with ->wait_lock is held. To prevent any fast path cmpxchg to the lock,
    83	 * we need to set the bit0 before looking at the lock, and the owner may be
    84	 * NULL in this small time, hence this can be a transitional state.
    85	 *
    86	 * (**) There is a small time when bit 0 is set but there are no
    87	 * waiters. This can happen when grabbing the lock in the slow path.
    88	 * To prevent a cmpxchg of the owner releasing the lock, we need to
    89	 * set this bit before looking at the lock.
    90	 */
    91	
    92	static __always_inline void
    93	rt_mutex_set_owner(struct rt_mutex_base *lock, struct task_struct *owner)
    94	{
    95		unsigned long val = (unsigned long)owner;
    96	
    97		if (rt_mutex_has_waiters(lock))
    98			val |= RT_MUTEX_HAS_WAITERS;
    99	
 > 100		WARN_ON_ONCE(cmpxchg_acquire(&lock->owner, RT_MUTEX_HAS_WAITERS, val) != RT_MUTEX_HAS_WAITERS);
   101	}
   102	
   103	static __always_inline void clear_rt_mutex_waiters(struct rt_mutex_base *lock)
   104	{
   105		lock->owner = (struct task_struct *)
   106				((unsigned long)lock->owner & ~RT_MUTEX_HAS_WAITERS);
   107	}
   108	
   109	static __always_inline void
   110	rt_mutex_set_owner_pi(struct rt_mutex_base *lock, struct task_struct *owner)
   111	{
   112		unsigned long val = (unsigned long)owner;
   113	
   114		if (rt_mutex_has_waiters(lock))
   115			val |= RT_MUTEX_HAS_WAITERS;
   116	
 > 117		WRITE_ONCE(lock->owner, val);
   118	}
   119	

-- 
0-DAY CI Kernel Test Service
https://01.org/lkp

View attachment "config" of type "text/plain" (137698 bytes)

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ