1. 28 Oct, 2008 9 commits
    • Heiko Carstens's avatar
      [S390] No more 4kb stacks. · 7f5a8ba6
      Heiko Carstens authored
      We got a stack overflow with a small stack configuration on a 32 bit
      system. It just looks like as 4kb isn't enough and too dangerous.
      So lets get rid of 4kb stacks on 32 bit.
      
      But one thing I completely dislike about the call trace below is that
      just for debugging or tracing purposes sprintf gets called (cio_start_key):
      
      	/* process condition code */
      	sprintf(dbf_txt, "ccode:%d", ccode);
      	CIO_TRACE_EVENT(4, dbf_txt);
      
      But maybe its just me who thinks that this could be done better.
      
          <4>Kernel stack overflow.
          <4>Modules linked in: dm_multipath sunrpc bonding qeth_l2 dm_mod qeth ccwgroup vmur
          <4>CPU: 1 Not tainted 2.6.27-30.x.20081015-s390default #1
          <4>Process httpd (pid: 3807, task: 20ae2df8, ksp: 1666fb78)
          <4>Krnl PSW : 040c0000 8027098a (number+0xe/0x348)
          <4>           R:0 T:1 IO:0 EX:0 Key:0 M:1 W:0 P:0 AS:0 CC:0 PM:0
          <4>Krnl GPRS: 00d43318 0027097c 1666f277 9666f270
          <4>           00000000 00000000 0000000a ffffffff
          <4>           9666f270 1666f228 1666f277 1666f098
          <4>           00000002 80270982 80271016 1666f098
          <4>Krnl Code: 8027097e: f0340dd0a7f1	srp	3536(4,%r0),2033(%r10),4
          <4>           80270984: 0f00		clcl	%r0,%r0
          <4>           80270986: a7840001		brc	8,80270988
          <4>          >8027098a: 18ef		lr	%r14,%r15
          <4>           8027098c: a7faff68		ahi	%r15,-152
          <4>           80270990: 18bf		lr	%r11,%r15
          <4>           80270992: 18a2		lr	%r10,%r2
          <4>           80270994: 1893		lr	%r9,%r3
      
      Modified calltrace with annotated stackframe size of each function:
      
      stackframe size
          |
       0 304 vsnprintf+850 [0x271016]
       1  72 sprintf+74 [0x271522]
       2  56 cio_start_key+262 [0x2d4c16]
       3  56 ccw_device_start_key+222 [0x2dfe92]
       4  56 ccw_device_start+40 [0x2dff28]
       5  48 raw3215_start_io+104 [0x30b0f8]
       6  56 raw3215_write+494 [0x30ba0a]
       7  40 con3215_write+68 [0x30bafc]
       8  40 __call_console_drivers+146 [0x12b0fa]
       9  32 _call_console_drivers+102 [0x12b192]
      10  64 release_console_sem+268 [0x12b614]
      11 168 vprintk+462 [0x12bca6]
      12  72 printk+68 [0x12bfd0]
      13 256 __print_symbol+50 [0x15a882]
      14  56 __show_trace+162 [0x103d06]
      15  32 show_trace+224 [0x103e70]
      16  48 show_stack+152 [0x103f20]
      17  56 dump_stack+126 [0x104612]
      18  96 __alloc_pages_internal+592 [0x175004]
      19  80 cache_alloc_refill+776 [0x196f3c]
      20  40 __kmalloc+258 [0x1972ae]
      21  40 __alloc_skb+94 [0x328086]
      22  32 pskb_copy+50 [0x328252]
      23  32 skb_realloc_headroom+110 [0x328a72]
      24 104 qeth_l2_hard_start_xmit+378 [0x7803bfde]
      25  56 dev_hard_start_xmit+450 [0x32ef6e]
      26  56 __qdisc_run+390 [0x3425d6]
      27  48 dev_queue_xmit+410 [0x331e06]
      28  40 ip_finish_output+308 [0x354ac8]
      29  56 ip_output+218 [0x355b6e]
      30  24 ip_local_out+56 [0x354584]
      31 120 ip_queue_xmit+300 [0x355cec]
      32  96 tcp_transmit_skb+812 [0x367da8]
      33  40 tcp_push_one+158 [0x369fda]
      34 112 tcp_sendmsg+852 [0x35d5a0]
      35 240 sock_sendmsg+164 [0x32035c]
      36  56 kernel_sendmsg+86 [0x32064a]
      37  88 sock_no_sendpage+98 [0x322b22]
      38 104 tcp_sendpage+70 [0x35cc1e]
      39  48 sock_sendpage+74 [0x31eb66]
      40  64 pipe_to_sendpage+102 [0x1c4b2e]
      41  64 __splice_from_pipe+120 [0x1c5340]
      42  72 splice_from_pipe+90 [0x1c57e6]
      43  56 generic_splice_sendpage+38 [0x1c5832]
      44  48 do_splice_from+104 [0x1c4c38]
      45  48 direct_splice_actor+52 [0x1c4c88]
      46  80 splice_direct_to_actor+180 [0x1c4f80]
      47  72 do_splice_direct+70 [0x1c5112]
      48  64 do_sendfile+360 [0x19de18]
      49  72 sys_sendfile64+126 [0x19df32]
      50 336 sysc_do_restart+18 [0x111a1a]
      Signed-off-by: default avatarHeiko Carstens <heiko.carstens@de.ibm.com>
      Signed-off-by: default avatarMartin Schwidefsky <schwidefsky@de.ibm.com>
      7f5a8ba6
    • Heiko Carstens's avatar
      [S390] Change default IPL method to IPL_VM. · 46e7951f
      Heiko Carstens authored
      allyesconfig and allmodconfig built kernels have a tape IPL record.
      A the vmreader record makes much more sense, since hardly anybody will
      ever IPL a kernel from tape. So change the default.
      As I side effect I can test these kernels without fiddling around with
      the kernel config ;)
      Signed-off-by: default avatarHeiko Carstens <heiko.carstens@de.ibm.com>
      Signed-off-by: default avatarMartin Schwidefsky <schwidefsky@de.ibm.com>
      46e7951f
    • Frank Munzert's avatar
      [S390] tape: disable interrupts in tape_open and tape_release · b3c21e49
      Frank Munzert authored
      Get tape device lock with interrupts disabled. Otherwise lockdep will issue a
      warning similar to:
      
      =================================
      [ INFO: inconsistent lock state ]
      2.6.27 #1
      ---------------------------------
      inconsistent {in-hardirq-W} -> {hardirq-on-W} usage.
      vol_id/2903 [HC0[0]:SC0[0]:HE1:SE1] takes:
       (sch->lock){++..}, at: [<000003e00004c7a2>] tape_open+0x42/0x1a4 [tape]
      {in-hardirq-W} state was registered at:
        [<000000000007ce5c>] __lock_acquire+0x894/0xa74
        [<000000000007d0ce>] lock_acquire+0x92/0xb8
        [<0000000000345154>] _spin_lock+0x5c/0x9c
        [<0000000000202264>] do_IRQ+0x124/0x1f0
        [<0000000000026610>] io_return+0x0/0x8
      
      irq event stamp: 847
      hardirqs last  enabled at (847): [<000000000007aca6>] trace_hardirqs_on+0x2a/0x38
      hardirqs last disabled at (846): [<0000000000076ca2>] trace_hardirqs_off+0x2a/0x38
      softirqs last  enabled at (0): [<000000000004909e>] copy_process+0x43e/0x11f4
      softirqs last disabled at (0): [<0000000000000000>] 0x0
      
      other info that might help us debug this:
      1 lock held by vol_id/2903:
       #0:  (&bdev->bd_mutex){--..}, at: [<000000000010e0f4>] do_open+0x78/0x358
      
      stack backtrace:
      CPU: 1 Not tainted 2.6.27 #1},
      Process vol_id (pid: 2903, task: 000000003d4c0000, ksp: 000000003d4e3b10)
      0400000000000000 000000003d4e3830 0000000000000002 0000000000000000
             000000003d4e38d0 000000003d4e3848 000000003d4e3848 00000000000168a8
             0000000000000000 000000003d4e3b10 0000000000000000 0000000000000000
             000000003d4e3830 000000000000000c 000000003d4e3830 000000003d4e38a0
             000000000034aa98 00000000000168a8 000000003d4e3830 000000003d4e3880
      Call Trace:
      ([<000000000001681c>] show_trace+0x138/0x158)
       [<0000000000016902>] show_stack+0xc6/0xf8
       [<00000000000170d4>] dump_stack+0xb0/0xc0
       [<0000000000078810>] print_usage_bug+0x1e8/0x228
       [<000000000007a71c>] mark_lock+0xb14/0xd24
       [<000000000007cd5a>] __lock_acquire+0x792/0xa74
       [<000000000007d0ce>] lock_acquire+0x92/0xb8
       [<0000000000345154>] _spin_lock+0x5c/0x9c
       [<000003e00004c7a2>] tape_open+0x42/0x1a4 [tape]
       [<000003e00005185c>] tapeblock_open+0x98/0xd0 [tape]
      Signed-off-by: default avatarFrank Munzert <munzert@de.ibm.com>
      Signed-off-by: default avatarMartin Schwidefsky <schwidefsky@de.ibm.com>
      b3c21e49
    • Roel Kluin's avatar
      [S390] appldata: unsigned ops->size cannot be negative · 13f8b7c5
      Roel Kluin authored
      unsigned ops->size cannot be negative
      Signed-off-by: default avatarRoel Kluin <roel.kluin@gmail.com>
      Signed-off-by: default avatarMartin Schwidefsky <schwidefsky@de.ibm.com>
      13f8b7c5
    • Frank Munzert's avatar
      [S390] tape block: complete request with correct locking · 7a4a1ccd
      Frank Munzert authored
      __blk_end_request must be called with request queue lock held. We need to use
      blk_end_request rather than  __blk_end_request.
      Signed-off-by: default avatarFrank Munzert <munzert@de.ibm.com>
      Signed-off-by: default avatarMartin Schwidefsky <schwidefsky@de.ibm.com>
      7a4a1ccd
    • Heiko Carstens's avatar
      [S390] Fix sysdev class file creation. · da5aae70
      Heiko Carstens authored
      Use sysdev_class_create_file() to create create sysdev class attributes
      instead of sysfs_create_file(). Using sysfs_create_file() wasn't a very
      good idea since the show and store functions have a different amount of
      parameters for sysfs files and sysdev class files.
      In particular the pointer to the buffer is the last argument and
      therefore accesses to random memory regions happened.
      Still worked surprisingly well until we got a kernel panic.
      
      Cc: stable@kernel.org
      Signed-off-by: default avatarHeiko Carstens <heiko.carstens@de.ibm.com>
      Signed-off-by: default avatarMartin Schwidefsky <schwidefsky@de.ibm.com>
      da5aae70
    • Christian Borntraeger's avatar
      [S390] pgtables: Fix race in enable_sie vs. page table ops · 250cf776
      Christian Borntraeger authored
      The current enable_sie code sets the mm->context.pgstes bit to tell
      dup_mm that the new mm should have extended page tables. This bit is also
      used by the s390 specific page table primitives to decide about the page
      table layout - which means context.pgstes has two meanings. This can cause
      any kind of bugs. For example  - e.g. shrink_zone can call
      ptep_clear_flush_young while enable_sie is running. ptep_clear_flush_young
      will test for context.pgstes. Since enable_sie changed that value of the old
      struct mm without changing the page table layout ptep_clear_flush_young will
      do the wrong thing.
      The solution is to split pgstes into two bits
      - one for the allocation
      - one for the current state
      Signed-off-by: default avatarChristian Borntraeger <borntraeger@de.ibm.com>
      Signed-off-by: default avatarMartin Schwidefsky <schwidefsky@de.ibm.com>
      250cf776
    • Jan Glauber's avatar
      [S390] qdio: remove incorrect memset · 2c780914
      Jan Glauber authored
      Remove the memset since zeroing the string is not needed and use
      snprintf instead of sprintf.
      Signed-off-by: default avatarJan Glauber <jang@linux.vnet.ibm.com>
      Signed-off-by: default avatarMartin Schwidefsky <schwidefsky@de.ibm.com>
      2c780914
    • Jan Glauber's avatar
      [S390] qdio: prevent double qdio shutdown in case of I/O errors · 7c045aa2
      Jan Glauber authored
      In case of I/O errors on a qdio subchannel qdio_shutdown may be
      called twice by the qdio driver and by zfcp. Remove the
      superfluous shutdown from qdio and let the upper layer driver
      handle the error condition.
      Signed-off-by: default avatarJan Glauber <jang@linux.vnet.ibm.com>
      Signed-off-by: default avatarMartin Schwidefsky <schwidefsky@de.ibm.com>
      7c045aa2
  2. 27 Oct, 2008 31 commits