1. 26 Jan, 2009 31 commits
  2. 25 Jan, 2009 2 commits
  3. 23 Jan, 2009 7 commits
    • Ian Campbell's avatar
      xen: handle highmem pages correctly when shrinking a domain · ff4ce8c3
      Ian Campbell authored
      Commit 1058a75f ("xen: actually release
      memory when shrinking domain") causes a crash if the page being released
      is a highmem page.
      If a page is highmem then there is no need to unmap it.
      Signed-off-by: default avatarIan Campbell <ian.campbell@citrix.com>
      Acked-by: default avatarJeremy Fitzhardinge <jeremy@goop.org>
      Signed-off-by: default avatarIngo Molnar <mingo@elte.hu>
    • Peter Zijlstra's avatar
      x86, mm: fix pte_free() · 42ef73fe
      Peter Zijlstra authored
      On -rt we were seeing spurious bad page states like:
      Bad page state in process 'firefox'
      page:c1bc2380 flags:0x40000000 mapping:c1bc2390 mapcount:0 count:0
      Trying to fix it up, but a reboot is needed
      Pid: 503, comm: firefox Not tainted #3
      [<c043d0f3>] ? printk+0x14/0x19
      [<c0272d4e>] bad_page+0x4e/0x79
      [<c0273831>] free_hot_cold_page+0x5b/0x1d3
      [<c02739f6>] free_hot_page+0xf/0x11
      [<c0273a18>] __free_pages+0x20/0x2b
      [<c027d170>] __pte_alloc+0x87/0x91
      [<c027d25e>] handle_mm_fault+0xe4/0x733
      [<c043f680>] ? rt_mutex_down_read_trylock+0x57/0x63
      [<c043f680>] ? rt_mutex_down_read_trylock+0x57/0x63
      [<c0218875>] do_page_fault+0x36f/0x88a
      This is the case where a concurrent fault already installed the PTE and
      we get to free the newly allocated one.
      This is due to pgtable_page_ctor() doing the spin_lock_init(&page->ptl)
      which is overlaid with the {private, mapping} struct.
      union {
          struct {
              unsigned long private;
              struct address_space *mapping;
          spinlock_t ptl;
          struct kmem_cache *slab;
          struct page *first_page;
      Normally the spinlock is small enough to not stomp on page->mapping, but
      PREEMPT_RT=y has huge 'spin'locks.
      But lockdep kernels should also be able to trigger this splat, as the
      lock tracking code grows the spinlock to cover page->mapping.
      The obvious fix is calling pgtable_page_dtor() like the regular pte free
      path __pte_free_tlb() does.
      It seems all architectures except x86 and nm10300 already do this, and
      nm10300 doesn't seem to use pgtable_page_ctor(), which suggests it
      doesn't do SMP or simply doesnt do MMU at all or something.
      Signed-off-by: default avatarPeter Zijlstra <a.p.zijlsta@chello.nl>
      Signed-off-by: default avatarIngo Molnar <mingo@elte.hu>
      Cc: <stable@kernel.org>
    • Takashi Iwai's avatar
      Merge branch 'fix/asoc' into for-linus · 0f0779b1
      Takashi Iwai authored
    • Michael Holzheu's avatar
      [S390] Add missing compat system call wrappers. · e34a6280
      Michael Holzheu authored
      Add wrapper functions for the following compat system calls:
      * readahead
      * sendfile64
      * tkill
      * tgkill
      * keyctl
      This ensures that the high order bits of the parameter registers are correctly
      sign extended.
      Signed-off-by: default avatarMichael Holzheu <holzheu@linux.vnet.ibm.com>
      Signed-off-by: default avatarHeiko Carstens <heiko.carstens@de.ibm.com>
    • Heiko Carstens's avatar
      [S390] etr/stp: fix possible deadlock · 179cb81a
      Heiko Carstens authored
      Precreate stop_machine threads in case the machine supports ETR/STP.
      Otherwise we might deadlock if a time sync operation gets scheduled
      and the creation of stop_machine threads would cause disk I/O.
      This is just the minimal fix.
      The real fix would be to only precreate stop_machine threads if
      ETR/STP is actually used. But that would be a rather large and
      complicated patch.
      Signed-off-by: default avatarHeiko Carstens <heiko.carstens@de.ibm.com>
    • Heiko Carstens's avatar
      [S390] cputime: fix lowcore initialization on cpu hotplug · f9a2f797
      Heiko Carstens authored
      On (initial) cpu hotplug the lowcore values for user_timer and
      system_timer don't get initialized like they would get on each
      process schedule.
      On initial start of secondary cpus this leads to the situation
      where per thread user/system_timer values are larger than the
      corresponding contents of the lowcore. When later calculating
      time spent in user/system context the result can be negative.
      So for cpu hotplug we should manually initialize lowcore values.
      Fixes this bug:
      Kernel BUG at 000ec080 [verbose debug info unavailable]
      fixpoint divide exception: 0009 [#1] PREEMPT SMP
      Modules linked in:
      CPU: 10 Not tainted 2.6.28 #4
      Process sysctl (pid: 975, task: 3fa752e0, ksp: 3fbebca0)
      Krnl PSW : 070c1000 800ec080 (show_stat+0x390/0x5fc)
                 R:0 T:1 IO:1 EX:1 Key:0 M:1 W:0 P:0 AS:0 CC:1 PM:0
      Krnl GPRS: 7fffffff fefc7ce5 3faec080 003879ae
                 00000001 01388000 7fffffff 01388000
                 00000000 00000000 0049ad50 3fbebcf8
                 01388000 002f51a8 800ec1fe 3fbebcf8
      Krnl Code: 800ec076: 9001b188           stm     %r0,%r1,392(%r11)
                 800ec07a: 9801b0c0           lm      %r0,%r1,192(%r11)
                 800ec07e: 1d05               dr      %r0,%r5
                >800ec080: 9001b0c0           stm     %r0,%r1,192(%r11)
                 800ec084: 5860b0c4           l       %r6,196(%r11)
                 800ec088: 1806               lr      %r0,%r6
                 800ec08a: 8c800001           srdl    %r8,1
                 800ec08e: 1d87               dr      %r8,%r7
      Call Trace:
      ([<00000000000ec1ee>] show_stat+0x4fe/0x5fc)
       [<00000000000c13e8>] seq_read+0xc4/0x3ac
       [<00000000000e4796>] proc_reg_read+0x6e/0x9c
       [<00000000000a6a44>] vfs_read+0x78/0x100
       [<00000000000a6ba8>] sys_read+0x40/0x80
       [<00000000000234a8>] sysc_do_restart+0x1a/0x1e
      Signed-off-by: default avatarHeiko Carstens <heiko.carstens@de.ibm.com>
    • Michael Holzheu's avatar
      [S390] fix compat sigaltstack syscall table entry · e9a4e9d5
      Michael Holzheu authored
      When 31 bit user space programs call sigaltstack on a 64 bit Linux
      OS, the system call returns -1 with errno=EFAULT. The 31 bit pointer passed
      to the system call is extended to 64 bit, but the high order bits are not
      set to zero. The kernel detects the invalid user space pointer and
      returns -EFAULT. To solve the problem, sys32_sigaltstack_wrapper()
      instead of sys32_sigaltstack() has to be called. The wrapper function sets
      the high order bits to zero.
      Signed-off-by: default avatarMichael Holzheu <holzheu@linux.vnet.ibm.com>
      Signed-off-by: default avatarHeiko Carstens <heiko.carstens@de.ibm.com>