kernel-packages team mailing list archive

[Tim Gardner <tim.gardner@xxxxxxxxxxxxx>]

Several kernel updates have occurred since this bug report. Can you re-
run your stress testing against the current 15.10 kernel version ?

-- 
You received this bug notification because you are a member of Kernel
Packages, which is subscribed to linux in Ubuntu.
https://bugs.launchpad.net/bugs/1496989

Title:
  ISST-LTE: system crashes at lpfc_sli4_scmd_to_wqidx_distr

Status in linux package in Ubuntu:
  New

Bug description:
  -- Problem Description --
  We have Ubuntu 15.10 installed on our system and run stress test for around 24 hrs then it crashes at lpfc_sli4_scmd_to_wqidx_distr+0x30/0x100

  0:mon> e
  cpu 0x0: Vector: 300 (Data Access) at [c00000000a0575a0]
      pc: d000000003115b30: lpfc_sli4_scmd_to_wqidx_distr+0x30/0x100 [lpfc]
      lr: d0000000030b749c: lpfc_sli_calc_ring.part.20+0xdc/0x100 [lpfc]
      sp: c00000000a057820
     msr: 8000000100009033
     dar: 0
   dsisr: 40000000
    current = 0xc000000272dbbcf0
    paca    = 0xc00000000e7f0000	 softe: 0	 irq_happened: 0x01
      pid   = 246, comm = scsi_eh_0
  0:mon> t
  [c00000000a057850] d0000000030b749c lpfc_sli_calc_ring.part.20+0xdc/0x100 [lpfc]
  [c00000000a057890] d0000000030bf680 lpfc_sli_issue_iocb+0xf0/0x320 [lpfc]
  [c00000000a0578f0] d0000000030c3804 lpfc_sli_issue_iocb_wait+0x264/0x680 [lpfc]
  [c00000000a0579d0] d000000003110fd4 lpfc_send_taskmgmt+0x2d4/0x7d0 [lpfc]
  [c00000000a057aa0] d000000003111bf4 lpfc_device_reset_handler+0x114/0x210 [lpfc]
  [c00000000a057b60] c00000000071254c scsi_eh_ready_devs+0x68c/0xee0
  [c00000000a057c50] c00000000071474c scsi_error_handler+0x6ac/0x9e0
  [c00000000a057d80] c0000000000e1e20 kthread+0x110/0x130
  [c00000000a057e30] c000000000009530 ret_from_kernel_thread+0x5c/0xac
  0:mon> di d000000003115b30
  d000000003115b30  e92a0000	ld	r9,0(r10)
  d000000003115b34  e9290000	ld	r9,0(r9)
  d000000003115b38  e92901a8	ld	r9,424(r9)
  d000000003115b3c  7928b7e3	rldicl.	r8,r9,54,63
  d000000003115b40  40820090	bne	d000000003115bd0	# lpfc_sli4_scmd_to_wqidx_distr+0xd0/0x100 [lpfc]
  d000000003115b44  813f0ae0	lwz	r9,2784(r31)
  d000000003115b48  2f890001	cmpwi	cr7,r9,1
  d000000003115b4c  419e0054	beq	cr7,d000000003115ba0	# lpfc_sli4_scmd_to_wqidx_distr+0xa0/0x100 [lpfc]
  d000000003115b50  395f0d58	addi	r10,r31,3416
  d000000003115b54  39200001	li	r9,1
  d000000003115b58  7c2004ac	lwsync
  d000000003115b5c  7c605028	lwarx	r3,0,r10
  d000000003115b60  7c691a14	add	r3,r9,r3
  d000000003115b64  7c60512d	stwcx.	r3,0,r10
  d000000003115b68  40c2fff4	bne-	d000000003115b5c	# lpfc_sli4_scmd_to_wqidx_distr+0x5c/0x100 [lpfc]
  d000000003115b6c  7c0004ac	sync

  
  0:mon> d c000000000ab00e0
  c000000000ab00e0 4c696e7578207665 7273696f6e20342e  |Linux version 4.|
  c000000000ab00f0 322e302d372d6765 6e65726963202862  |2.0-7-generic (b|
  c000000000ab0100 75696c6464406465 6e6e656564303429  |uildd@denneed04)|
  c000000000ab0110 2028676363207665 7273696f6e20352e  | (gcc version 5.|

  lpfc_sli4_scmd_to_wqidx_distr() got moved around and changed a bit to
  lpfc_scsi.c with commit 8b0dff14164d3f43eba8365950b506d898e0e1e6 and
  the crash appears to be due to an invalid address of 0x0 for struct
  scsi_cmnd *cmnd

  3860 int lpfc_sli4_scmd_to_wqidx_distr(struct lpfc_hba *phba,
  3861                                   struct lpfc_scsi_buf *lpfc_cmd)
  3862 {
  3863         struct scsi_cmnd *cmnd = lpfc_cmd->pCmd;
  3864         struct lpfc_vector_map_info *cpup;
  3865         int chann, cpu;
  3866         uint32_t tag;
  3867         uint16_t hwq;
  3868 
  3869         if (shost_use_blk_mq(cmnd->device->host)) {
  3870                 tag = blk_mq_unique_tag(cmnd->request);
  3871                 hwq = blk_mq_unique_tag_to_hwq(tag);
  3872 
  3873                 return hwq;
  3874         }

  0:mon> r
  R00 = d0000000030b749c   R16 = c00000000a057cd0
  R01 = c00000000a057820   R17 = c00000000a057cb8
  R02 = d000000003163d28   R18 = c00000000a52a088
  R03 = c00000027e9fe000   R19 = c00000000a057cb0
  R04 = c00000027139a400   R20 = 000000000000001e
  R05 = c00000027139a470   R21 = 0000000000000001
  R06 = 0000000000000001   R22 = c00000000180c268
  R07 = d000000003163d28   R23 = c00000027139a470
  R08 = d00000000310de90   R24 = c00000027139a400
  R09 = 0000000000000004   R25 = c00000000a057978
  R10 = 0000000000000000   R26 = 0000000000000001
  R11 = d000000003137e20   R27 = 0000000000000000
  R12 = 0000000028641824   R28 = c00000000a528000
  R13 = c00000000e7f0000   R29 = c00000027e9fe000
  R14 = c00000000a057cb8   R30 = c00000027139a400
  R15 = 0000000000000000   R31 = c00000027e9fe000
  pc  = d000000003115b30 lpfc_sli4_scmd_to_wqidx_distr+0x30/0x100 [lpfc]
  cfar= c000000000008468 slb_miss_realmode+0x50/0x78
  lr  = d0000000030b749c lpfc_sli_calc_ring.part.20+0xdc/0x100 [lpfc]
  msr = 8000000100009033   cr  = 28648828
  ctr = c000000000a95a70   xer = 0000000020000000   trap =  300
  dar = 0000000000000000   dsisr = 40000000
  0:mon> di $lpfc_sli4_scmd_to_wqidx_distr
  d000000003115b00  3c4c0005	addis	r2,r12,5
  d000000003115b04  3842e228	addi	r2,r2,-7640
  d000000003115b08  7c0802a6	mflr	r0
  d000000003115b0c  fbc1fff0	std	r30,-16(r1)
  d000000003115b10  fbe1fff8	std	r31,-8(r1)
  d000000003115b14  f8010010	std	r0,16(r1)
  d000000003115b18  f821ffd1	stdu	r1,-48(r1)
  d000000003115b1c  7c9e2378	mr	r30,r4
  d000000003115b20  7c7f1b78	mr	r31,r3
  d000000003115b24  48000008	b	d000000003115b2c	# lpfc_sli4_scmd_to_wqidx_distr+0x2c/0x100 [lpfc]
  d000000003115b28  e8410018	ld	r2,24(r1)
  d000000003115b2c  e95e0010	ld	r10,16(r30)
  d000000003115b30  e92a0000	ld	r9,0(r10)
  d000000003115b34  e9290000	ld	r9,0(r9)
  d000000003115b38  e92901a8	ld	r9,424(r9)
  d000000003115b3c  7928b7e3	rldicl.	r8,r9,54,63
  0:mon> d c00000027139a400
  c00000027139a400 0001100000000000 0002200000000000  |.......... .....|
  c00000027139a410 0000000000000000 a8649972020000c0  |.........d.r....|
  c00000027139a420 3c00000000000000 0000000000000000  |<...............|
  c00000027139a430 0000000000000000 0000000000000000  |................|
  0:mon> 

  There were some I/O errors and failed paths before the kernel
  faults...

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/1496989/+subscriptions