On Tuesday, 12 June 2007 23:55, Rafael J. Wysocki wrote: > On Monday, 11 June 2007 09:17, Tejun Heo wrote: > > Hello, Rafael. > > > > Rafael J. Wysocki wrote: > > > gregkh-driver-sysfs-use-singly-linked-list-for-sysfs_dirent-tree.patch > > > breaks > > > suspend to RAM on HPC nx6325 (x86_64). > > > > > > With this patch applied I get a general protection fault in > > > mutex_lock+0x15 > > > (kernel/mutex.c:91), called by sysfs_hash_and_remove() > > > (fs/sysfs/inode.c:298), > > > called by threshold_cpu_callback(), called from _cpu_down(). > > > > I'm not sure whether this is bug in sysfs or in sysfs handling code in > > mce_amd and I can't test mce_amd here. Can you please apply the > > attached patch and post the resulting dmesg including oops? > > I've applied the patch, but the oops is a kernel panic, so I can't generate a > dmesg including it. ;-) > > Here's the dmesg output from a fresh boot (runlevel 2): > > http://www.sisk.pl/kernel/debug/2.6.22-rc4-mm2/dmesg.log > > Here's a picture of the oops taken after a failed attempt to suspend: > > http://www.sisk.pl/kernel/debug/2.6.22-rc4-mm2/during_suspend.jpg > > Here's a picture of the oops taken after a failed attempt to offline CPU1 > using 'echo 1 > /sys/devices/system/cpu/cpu1/online': > > http://www.sisk.pl/kernel/debug/2.6.22-rc4-mm2/after_offlining_cpu1.jpg > > [Sorry for the quality of pictures, I couldn't get anything better.]
More information: With the debug patch applied the oops is a NULL pointer dereference at sysfs_hash_and_remove+0x16, which according to gdb is (gdb) l *sysfs_hash_and_remove+0x16 0xffffffff802d4bff is in sysfs_hash_and_remove (/home/rafael/src/mm/linux-2.6.22-rc4-mm2/fs/sysfs/inode.c:294). 289 int found = 0; 290 291 if (!dir) 292 return -ENOENT; 293 294 if (dir->d_inode == NULL) 295 /* no inode means this hasn't been made visible yet */ 296 return -ENOENT; 297 298 mutex_lock_nested(&dir->d_inode->i_mutex, I_MUTEX_PARENT); That doesn't make much sense to me, but it's 100% reproducible. Greetings, Rafael -- "Premature optimization is the root of all evil." - Donald Knuth - To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html Please read the FAQ at http://www.tux.org/lkml/
