Re: [PATCH kernel] commit 4fbdf9cb (lpfc: Fix for lun discovery issue with saturn adapter.)
I wrote: Alexey Kardashevskiy wrote: On 05/06/2015 07:46 AM, Sebastian Herbszt wrote: James Smart wrote: Reviewed-By: James Smart james.sm...@emulex.com Alexey, Sebastian, Yes - this section needs to be reverted. This patch is good. -- james s Reviewed-by: Sebastian Herbszt herb...@gmx.de Unfortunately, just this revert is not enough, it fixed one of my machines but I have another machine with Emulex Corporation Saturn-X: LightPulse Fibre Channel Host Adapter (rev 03) which does not boot - the booting process stops at [ OK ] Reached target Basic System. Any quick idea what else to revert between 4.0 and current upstream before I dig further? Thanks! Is the other machine also POWER7? I tested the revert on x86 and it fixed my issue: 4.1-rc4 (lpfc 10.5.0.0) - broken 4.0 + lpfc 10.5.0.0 - broken 4.1-rc4 (lpfc 10.5.0.0) + revert - ok 4.0 + lpfc 10.5.0.0 + revert - ok James, are you sending this revert to Linus any time soon? Support for SLI-3 adapters is broken on all platforms for more than a month now [1] and POWER7 doesn't seem to like it at all [2]. [1] http://marc.info/?l=linux-scsim=142938432414173 [2] http://marc.info/?l=linux-scsim=143021737816241 Sebastian ___ Linuxppc-dev mailing list Linuxppc-dev@lists.ozlabs.org https://lists.ozlabs.org/listinfo/linuxppc-dev
Re: [PATCH kernel] commit 4fbdf9cb (lpfc: Fix for lun discovery issue with saturn adapter.)
Alexey Kardashevskiy wrote: On 05/06/2015 07:46 AM, Sebastian Herbszt wrote: James Smart wrote: Reviewed-By: James Smart james.sm...@emulex.com Alexey, Sebastian, Yes - this section needs to be reverted. This patch is good. -- james s Reviewed-by: Sebastian Herbszt herb...@gmx.de Unfortunately, just this revert is not enough, it fixed one of my machines but I have another machine with Emulex Corporation Saturn-X: LightPulse Fibre Channel Host Adapter (rev 03) which does not boot - the booting process stops at [ OK ] Reached target Basic System. Any quick idea what else to revert between 4.0 and current upstream before I dig further? Thanks! Is the other machine also POWER7? I tested the revert on x86 and it fixed my issue: 4.1-rc4 (lpfc 10.5.0.0) - broken 4.0 + lpfc 10.5.0.0 - broken 4.1-rc4 (lpfc 10.5.0.0) + revert - ok 4.0 + lpfc 10.5.0.0 + revert - ok Sebastian ___ Linuxppc-dev mailing list Linuxppc-dev@lists.ozlabs.org https://lists.ozlabs.org/listinfo/linuxppc-dev
Re: [PATCH kernel] commit 4fbdf9cb (lpfc: Fix for lun discovery issue with saturn adapter.)
Not sure - I'll go look. -- james On 5/19/2015 11:55 PM, Alexey Kardashevskiy wrote: On 05/06/2015 07:46 AM, Sebastian Herbszt wrote: James Smart wrote: Reviewed-By: James Smart james.sm...@emulex.com Alexey, Sebastian, Yes - this section needs to be reverted. This patch is good. -- james s Reviewed-by: Sebastian Herbszt herb...@gmx.de Unfortunately, just this revert is not enough, it fixed one of my machines but I have another machine with Emulex Corporation Saturn-X: LightPulse Fibre Channel Host Adapter (rev 03) which does not boot - the booting process stops at [ OK ] Reached target Basic System. Any quick idea what else to revert between 4.0 and current upstream before I dig further? Thanks! ___ Linuxppc-dev mailing list Linuxppc-dev@lists.ozlabs.org https://lists.ozlabs.org/listinfo/linuxppc-dev
Re: [PATCH kernel] commit 4fbdf9cb (lpfc: Fix for lun discovery issue with saturn adapter.)
Alexey, Can you send me boot messages so I can see everything the lpfc driver may have done ? And you're using 10.5.0.0 verbatim ? -- james On 5/20/2015 9:11 AM, James Smart wrote: Not sure - I'll go look. -- james On 5/19/2015 11:55 PM, Alexey Kardashevskiy wrote: On 05/06/2015 07:46 AM, Sebastian Herbszt wrote: James Smart wrote: Reviewed-By: James Smart james.sm...@emulex.com Alexey, Sebastian, Yes - this section needs to be reverted. This patch is good. -- james s Reviewed-by: Sebastian Herbszt herb...@gmx.de Unfortunately, just this revert is not enough, it fixed one of my machines but I have another machine with Emulex Corporation Saturn-X: LightPulse Fibre Channel Host Adapter (rev 03) which does not boot - the booting process stops at [ OK ] Reached target Basic System. Any quick idea what else to revert between 4.0 and current upstream before I dig further? Thanks! ___ Linuxppc-dev mailing list Linuxppc-dev@lists.ozlabs.org https://lists.ozlabs.org/listinfo/linuxppc-dev
Re: [PATCH kernel] commit 4fbdf9cb (lpfc: Fix for lun discovery issue with saturn adapter.)
On 05/06/2015 07:46 AM, Sebastian Herbszt wrote: James Smart wrote: Reviewed-By: James Smart james.sm...@emulex.com Alexey, Sebastian, Yes - this section needs to be reverted. This patch is good. -- james s Reviewed-by: Sebastian Herbszt herb...@gmx.de Unfortunately, just this revert is not enough, it fixed one of my machines but I have another machine with Emulex Corporation Saturn-X: LightPulse Fibre Channel Host Adapter (rev 03) which does not boot - the booting process stops at [ OK ] Reached target Basic System. Any quick idea what else to revert between 4.0 and current upstream before I dig further? Thanks! -- Alexey ___ Linuxppc-dev mailing list Linuxppc-dev@lists.ozlabs.org https://lists.ozlabs.org/listinfo/linuxppc-dev
Re: [PATCH kernel] commit 4fbdf9cb (lpfc: Fix for lun discovery issue with saturn adapter.)
Reviewed-By: James Smart james.sm...@emulex.com Alexey, Sebastian, Yes - this section needs to be reverted. This patch is good. -- james s On 4/28/2015 4:26 AM, Alexey Kardashevskiy wrote: This reverts 4fbdf9cb is breaks LPFC on POWER7 machine, big endian kernel. This is the hardware used for verification: 0005:01:00.0 Fibre Channel [0c04]: Emulex Corporation Saturn-X: LightPulse Fibre Channel Host Adapter [10df:f100] (rev 03) 0005:01:00.1 Fibre Channel [0c04]: Emulex Corporation Saturn-X: LightPulse Fibre Channel Host Adapter [10df:f100] (rev 03) Signed-off-by: Alexey Kardashevskiy a...@ozlabs.ru --- drivers/scsi/lpfc/lpfc_scsi.c | 41 + 1 file changed, 21 insertions(+), 20 deletions(-) diff --git a/drivers/scsi/lpfc/lpfc_scsi.c b/drivers/scsi/lpfc/lpfc_scsi.c index cb73cf9..c140f99 100644 --- a/drivers/scsi/lpfc/lpfc_scsi.c +++ b/drivers/scsi/lpfc/lpfc_scsi.c @@ -1130,25 +1130,6 @@ lpfc_release_scsi_buf(struct lpfc_hba *phba, struct lpfc_scsi_buf *psb) } /** - * lpfc_fcpcmd_to_iocb - copy the fcp_cmd data into the IOCB - * @data: A pointer to the immediate command data portion of the IOCB. - * @fcp_cmnd: The FCP Command that is provided by the SCSI layer. - * - * The routine copies the entire FCP command from @fcp_cmnd to @data while - * byte swapping the data to big endian format for transmission on the wire. - **/ -static void -lpfc_fcpcmd_to_iocb(uint8_t *data, struct fcp_cmnd *fcp_cmnd) -{ - int i, j; - - for (i = 0, j = 0; i sizeof(struct fcp_cmnd); -i += sizeof(uint32_t), j++) { - ((uint32_t *)data)[j] = cpu_to_be32(((uint32_t *)fcp_cmnd)[j]); - } -} - -/** * lpfc_scsi_prep_dma_buf_s3 - DMA mapping for scsi buffer to SLI3 IF spec * @phba: The Hba for which this call is being executed. * @lpfc_cmd: The scsi buffer which is going to be mapped. @@ -1283,7 +1264,6 @@ lpfc_scsi_prep_dma_buf_s3(struct lpfc_hba *phba, struct lpfc_scsi_buf *lpfc_cmd) * we need to set word 4 of IOCB here */ iocb_cmd-un.fcpi.fcpi_parm = scsi_bufflen(scsi_cmnd); - lpfc_fcpcmd_to_iocb(iocb_cmd-unsli3.fcp_ext.icd, fcp_cmnd); return 0; } @@ -4147,6 +4127,24 @@ lpfc_scsi_cmd_iocb_cmpl(struct lpfc_hba *phba, struct lpfc_iocbq *pIocbIn, } /** + * lpfc_fcpcmd_to_iocb - copy the fcp_cmd data into the IOCB + * @data: A pointer to the immediate command data portion of the IOCB. + * @fcp_cmnd: The FCP Command that is provided by the SCSI layer. + * + * The routine copies the entire FCP command from @fcp_cmnd to @data while + * byte swapping the data to big endian format for transmission on the wire. + **/ +static void +lpfc_fcpcmd_to_iocb(uint8_t *data, struct fcp_cmnd *fcp_cmnd) +{ + int i, j; + for (i = 0, j = 0; i sizeof(struct fcp_cmnd); +i += sizeof(uint32_t), j++) { + ((uint32_t *)data)[j] = cpu_to_be32(((uint32_t *)fcp_cmnd)[j]); + } +} + +/** * lpfc_scsi_prep_cmnd - Wrapper func for convert scsi cmnd to FCP info unit * @vport: The virtual port for which this call is being executed. * @lpfc_cmd: The scsi command which needs to send. @@ -4225,6 +4223,9 @@ lpfc_scsi_prep_cmnd(struct lpfc_vport *vport, struct lpfc_scsi_buf *lpfc_cmd, fcp_cmnd-fcpCntl3 = 0; phba-fc4ControlRequests++; } + if (phba-sli_rev == 3 + !(phba-sli3_options LPFC_SLI3_BG_ENABLED)) + lpfc_fcpcmd_to_iocb(iocb_cmd-unsli3.fcp_ext.icd, fcp_cmnd); /* * Finish initializing those IOCB fields that are independent * of the scsi_cmnd request_buffer ___ Linuxppc-dev mailing list Linuxppc-dev@lists.ozlabs.org https://lists.ozlabs.org/listinfo/linuxppc-dev
Re: [PATCH kernel] commit 4fbdf9cb (lpfc: Fix for lun discovery issue with saturn adapter.)
James Smart wrote: Reviewed-By: James Smart james.sm...@emulex.com Alexey, Sebastian, Yes - this section needs to be reverted. This patch is good. -- james s Reviewed-by: Sebastian Herbszt herb...@gmx.de Sebastian ___ Linuxppc-dev mailing list Linuxppc-dev@lists.ozlabs.org https://lists.ozlabs.org/listinfo/linuxppc-dev
Re: [PATCH kernel] commit 4fbdf9cb (lpfc: Fix for lun discovery issue with saturn adapter.)
On 04/28/2015 07:18 PM, Sebastian Herbszt wrote: Alexey Kardashevskiy wrote: This reverts 4fbdf9cb is breaks LPFC on POWER7 machine, big endian kernel. This is the hardware used for verification: 0005:01:00.0 Fibre Channel [0c04]: Emulex Corporation Saturn-X: LightPulse Fibre Channel Host Adapter [10df:f100] (rev 03) 0005:01:00.1 Fibre Channel [0c04]: Emulex Corporation Saturn-X: LightPulse Fibre Channel Host Adapter [10df:f100] (rev 03) Signed-off-by: Alexey Kardashevskiy a...@ozlabs.ru This issue is not specific to POWER7. I hit it on x86 [1] and James promised to look at it. [1] http://marc.info/?l=linux-scsim=142938432414173 Sebastian Well, I hope so, I just wanted to be more specific and the fault looks much different (and much cooler! :) ) on my hardware (it actually enters an infinite loop of oops'es): Welcome to Fedora 20 (Heisenbug)! INFO: rcu_sched self-detected stall on CPU INFO: rcu_sched self-detected stall on CPU INFO: rcu_sched self-detected stall on CPU 1: (2100 ticks this GP) idle=981/141/0 softirq=234/234 fqs =2083 2: (2100 ticks this GP) idle=c3d/141/0 softirq=259/259 fqs =2083 (t=2100 jiffies g=-7 c=-8 q=11820) (t=2100 jiffies g=-7 c=-8 q=11820) Task dump for CPU 0: kworker/u97:0 R running task 8192 7 2 0x0804 Workqueue: events_unbound .async_run_entry_fn Call Trace: [c00ffa29ef80] [c00ffa29f060] 0xc00ffa29f060 (unreliable) Task dump for CPU 1: kworker/u97:2 R running task10304 1636 2 0x0804 Workqueue: events_unbound .async_run_entry_fn Call Trace: [c00ff2fd2f80] [c00ff2fd3060] 0xc00ff2fd3060 (unreliable) Task dump for CPU 2: kworker/u97:1 R running task 8288 1633 2 0x0804 Workqueue: events_unbound .async_run_entry_fn Call Trace: [c00ff2f92eb0] [c00cf610] .sched_show_task+0xf0/0x180 (unreliable) [c00ff2f92f30] [c01041d8] .rcu_dump_cpu_stacks+0xd8/0x150 [c00ff2f92fd0] [c0108794] .rcu_check_callbacks+0x674/0x990 [c00ff2f93110] [c010e994] .update_process_times+0x44/0x90 [c00ff2f93190] [c01223f0] .tick_sched_handle.isra.16+0x20/0xa0 [c00ff2f93210] [c01224cc] .tick_sched_timer+0x5c/0xb0 [c00ff2f932b0] [c010f108] .__run_hrtimer+0x98/0x260 [c00ff2f93350] [c010fff8] .hrtimer_interrupt+0x138/0x2f0 [c00ff2f93460] [c001be1c] .__timer_interrupt+0x8c/0x230 [c00ff2f93500] [c001c488] .timer_interrupt+0x98/0xd0 [c00ff2f93580] [c00025d0] decrementer_common+0x150/0x180 --- interrupt: 901 at .string_get_size+0x120/0x250 LR = .sd_revalidate_disk+0x57c/0x1c10 [c00ff2f93870] [c048f84c] .string_get_size+0x18c/0x250 (unreliable ) [c00ff2f93940] [c05e7c1c] .sd_revalidate_disk+0x57c/0x1c10 [c00ff2f93a70] [c05e951c] .sd_probe_async+0xac/0x230 [c00ff2f93b00] [c00c28ec] .async_run_entry_fn+0x6c/0x180 [c00ff2f93ba0] [c00b7b78] .process_one_work+0x1a8/0x4a0 [c00ff2f93c40] [c00b7ff0] .worker_thread+0x180/0x5a0 [c00ff2f93d30] [c00bee08] .kthread+0x108/0x130 [c00ff2f93e30] [c0009590] .ret_from_kernel_thread+0x58/0xc8 Task dump for CPU 0: kworker/u97:0 R running task 8192 7 2 0x0804 Workqueue: events_unbound .async_run_entry_fn Call Trace: [c00ffa29ef80] [c00ffa29f060] 0xc00ffa29f060 (unreliable) Task dump for CPU 1: kworker/u97:2 R running task 9488 1636 2 0x0804 Workqueue: events_unbound .async_run_entry_fn Call Trace: [c00ff2fd2eb0] [c00cf610] .sched_show_task+0xf0/0x180 (unreliable) [c00ff2fd2f30] [c01041d8] .rcu_dump_cpu_stacks+0xd8/0x150 [c00ff2fd2fd0] [c0108794] .rcu_check_callbacks+0x674/0x990 [c00ff2fd3110] [c010e994] .update_process_times+0x44/0x90 [c00ff2fd3190] [c01223f0] .tick_sched_handle.isra.16+0x20/0xa0 [c00ff2fd3210] [c01224cc] .tick_sched_timer+0x5c/0xb0 [c00ff2fd32b0] [c010f108] .__run_hrtimer+0x98/0x260 [c00ff2fd3350] [c010fff8] .hrtimer_interrupt+0x138/0x2f0 [c00ff2fd3460] [c001be1c] .__timer_interrupt+0x8c/0x230 [c00ff2fd3500] [c001c488] .timer_interrupt+0x98/0xd0 [c00ff2fd3580] [c00025d0] decrementer_common+0x150/0x180 --- interrupt: 901 at .string_get_size+0x110/0x250 LR = .sd_revalidate_disk+0x57c/0x1c10 [c00ff2fd3870] [c048f84c] .string_get_size+0x18c/0x250 (unreliable ) [c00ff2fd3940] [c05e7c1c] .sd_revalidate_disk+0x57c/0x1c10 [c00ff2fd3a70] [c05e951c] .sd_probe_async+0xac/0x230 [c00ff2fd3b00] [c00c28ec] .async_run_entry_fn+0x6c/0x180 [c00ff2fd3ba0] [c00b7b78] .process_one_work+0x1a8/0x4a0 [c00ff2fd3c40] [c00b7ff0] .worker_thread+0x180/0x5a0 [c00ff2fd3d30] [c00bee08] .kthread+0x108/0x130 [c00ff2fd3e30] [c0009590]
Re: [PATCH kernel] commit 4fbdf9cb (lpfc: Fix for lun discovery issue with saturn adapter.)
Alexey Kardashevskiy wrote: This reverts 4fbdf9cb is breaks LPFC on POWER7 machine, big endian kernel. This is the hardware used for verification: 0005:01:00.0 Fibre Channel [0c04]: Emulex Corporation Saturn-X: LightPulse Fibre Channel Host Adapter [10df:f100] (rev 03) 0005:01:00.1 Fibre Channel [0c04]: Emulex Corporation Saturn-X: LightPulse Fibre Channel Host Adapter [10df:f100] (rev 03) Signed-off-by: Alexey Kardashevskiy a...@ozlabs.ru This issue is not specific to POWER7. I hit it on x86 [1] and James promised to look at it. [1] http://marc.info/?l=linux-scsim=142938432414173 Sebastian ___ Linuxppc-dev mailing list Linuxppc-dev@lists.ozlabs.org https://lists.ozlabs.org/listinfo/linuxppc-dev
[PATCH kernel] commit 4fbdf9cb (lpfc: Fix for lun discovery issue with saturn adapter.)
This reverts 4fbdf9cb is breaks LPFC on POWER7 machine, big endian kernel. This is the hardware used for verification: 0005:01:00.0 Fibre Channel [0c04]: Emulex Corporation Saturn-X: LightPulse Fibre Channel Host Adapter [10df:f100] (rev 03) 0005:01:00.1 Fibre Channel [0c04]: Emulex Corporation Saturn-X: LightPulse Fibre Channel Host Adapter [10df:f100] (rev 03) Signed-off-by: Alexey Kardashevskiy a...@ozlabs.ru --- drivers/scsi/lpfc/lpfc_scsi.c | 41 + 1 file changed, 21 insertions(+), 20 deletions(-) diff --git a/drivers/scsi/lpfc/lpfc_scsi.c b/drivers/scsi/lpfc/lpfc_scsi.c index cb73cf9..c140f99 100644 --- a/drivers/scsi/lpfc/lpfc_scsi.c +++ b/drivers/scsi/lpfc/lpfc_scsi.c @@ -1130,25 +1130,6 @@ lpfc_release_scsi_buf(struct lpfc_hba *phba, struct lpfc_scsi_buf *psb) } /** - * lpfc_fcpcmd_to_iocb - copy the fcp_cmd data into the IOCB - * @data: A pointer to the immediate command data portion of the IOCB. - * @fcp_cmnd: The FCP Command that is provided by the SCSI layer. - * - * The routine copies the entire FCP command from @fcp_cmnd to @data while - * byte swapping the data to big endian format for transmission on the wire. - **/ -static void -lpfc_fcpcmd_to_iocb(uint8_t *data, struct fcp_cmnd *fcp_cmnd) -{ - int i, j; - - for (i = 0, j = 0; i sizeof(struct fcp_cmnd); -i += sizeof(uint32_t), j++) { - ((uint32_t *)data)[j] = cpu_to_be32(((uint32_t *)fcp_cmnd)[j]); - } -} - -/** * lpfc_scsi_prep_dma_buf_s3 - DMA mapping for scsi buffer to SLI3 IF spec * @phba: The Hba for which this call is being executed. * @lpfc_cmd: The scsi buffer which is going to be mapped. @@ -1283,7 +1264,6 @@ lpfc_scsi_prep_dma_buf_s3(struct lpfc_hba *phba, struct lpfc_scsi_buf *lpfc_cmd) * we need to set word 4 of IOCB here */ iocb_cmd-un.fcpi.fcpi_parm = scsi_bufflen(scsi_cmnd); - lpfc_fcpcmd_to_iocb(iocb_cmd-unsli3.fcp_ext.icd, fcp_cmnd); return 0; } @@ -4147,6 +4127,24 @@ lpfc_scsi_cmd_iocb_cmpl(struct lpfc_hba *phba, struct lpfc_iocbq *pIocbIn, } /** + * lpfc_fcpcmd_to_iocb - copy the fcp_cmd data into the IOCB + * @data: A pointer to the immediate command data portion of the IOCB. + * @fcp_cmnd: The FCP Command that is provided by the SCSI layer. + * + * The routine copies the entire FCP command from @fcp_cmnd to @data while + * byte swapping the data to big endian format for transmission on the wire. + **/ +static void +lpfc_fcpcmd_to_iocb(uint8_t *data, struct fcp_cmnd *fcp_cmnd) +{ + int i, j; + for (i = 0, j = 0; i sizeof(struct fcp_cmnd); +i += sizeof(uint32_t), j++) { + ((uint32_t *)data)[j] = cpu_to_be32(((uint32_t *)fcp_cmnd)[j]); + } +} + +/** * lpfc_scsi_prep_cmnd - Wrapper func for convert scsi cmnd to FCP info unit * @vport: The virtual port for which this call is being executed. * @lpfc_cmd: The scsi command which needs to send. @@ -4225,6 +4223,9 @@ lpfc_scsi_prep_cmnd(struct lpfc_vport *vport, struct lpfc_scsi_buf *lpfc_cmd, fcp_cmnd-fcpCntl3 = 0; phba-fc4ControlRequests++; } + if (phba-sli_rev == 3 + !(phba-sli3_options LPFC_SLI3_BG_ENABLED)) + lpfc_fcpcmd_to_iocb(iocb_cmd-unsli3.fcp_ext.icd, fcp_cmnd); /* * Finish initializing those IOCB fields that are independent * of the scsi_cmnd request_buffer -- 2.0.0 ___ Linuxppc-dev mailing list Linuxppc-dev@lists.ozlabs.org https://lists.ozlabs.org/listinfo/linuxppc-dev