On Thu, Nov 3, 2016 at 1:34 AM, Jinpu Wang <jinpu.wang@xxxxxxxxxxxxxxxx> wrote: > Hi Duc, > > On Thu, Nov 3, 2016 at 2:02 AM, Duc Dang <dhdang@xxxxxxx> wrote: >> Hi Jack, Lindar and All, >> >> I was testing Adaptec PMC-Sierra PM8018 SAS HBA [Series 7H] card >> [9005:8088] with our ARM64 server SoC and occasionally saw a hang in >> pm8001_pci_probe. It looks like the driver was trying to read nvmd >> information (using get_nvmd_req) and wait for completion but the >> completion never happens, so I got a kernel timeout trace: >> >> "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message. >> swapper/0 D ffffffc000086ec0 0 1 0 0x00000000 >> Call trace: >> [<ffffffc000086ec0>] __switch_to+0x90/0xb0 >> [<ffffffc0009e2214>] __schedule+0x204/0x670 >> [<ffffffc0009e26c0>] schedule+0x40/0xb0 >> [<ffffffc0009e549c>] schedule_timeout+0x15c/0x1b0 >> [<ffffffc0009e32e0>] wait_for_common+0xa0/0x150 >> [<ffffffc0009e33a4>] wait_for_completion+0x14/0x20 >> [<ffffffc00059a9cc>] pm8001_pci_probe+0xb7c/0xd10 >> [<ffffffc00046c7f0>] pci_device_probe+0xa0/0x120 >> [<ffffffc00053ff7c>] driver_probe_device+0x12c/0x2e0 >> [<ffffffc0005401cc>] __driver_attach+0x9c/0xa0 >> [<ffffffc00053e0ac>] bus_for_each_dev+0x5c/0xa0 >> [<ffffffc00053f8e0>] driver_attach+0x20/0x30 >> [<ffffffc00053f428>] bus_add_driver+0x108/0x230 >> [<ffffffc0005409e0>] driver_register+0x60/0x100 >> [<ffffffc00046b5a0>] __pci_register_driver+0x40/0x50 >> [<ffffffc000de3610>] pm8001_init+0x70/0xa8 >> [<ffffffc000082904>] do_one_initcall+0x94/0x1b0 >> [<ffffffc000dbcab8>] kernel_init_freeable+0x150/0x1f4 >> [<ffffffc0009df2c0>] kernel_init+0x10/0xf0 >> >> The current driver in kernel.org is 0.1.38, is this already the latest one? >> >> I saw in the Readme file of BIOS package in Microsemi website >> (http://storage.microsemi.com/en-us/downloads/bios_fw/bios_fw_ver/productid=asa-71605h&dn=adaptec+71605h.php) >> mentions below issue. Is the hang symptom similar to the one that I >> saw? >> >> 5.14 FreeBSD Hangs when No Devices on HBA >> On FreeBSD systems, the OS fails to boot when no devices are >> connected to the HBAs. >> >> WORKAROUND: Connect at least one device to the HBA to complete >> the boot process. >> >> >> Regards, >> Duc Dang. > > Likely it related to your hardware, maybe you don't have NVM device to > save VPD related info, you can tweak that part, eg comment out > #define PM8001_READ_VPD. Thanks, Jack. I will do more tests with PM8001_READ_VPD disabled. Will let you know if I run into any issue. By the way, the 'NVM device' you mentioned is a separate external device that can be connect to the card via some cables (I see a small header in the card), not in the card itself? > > > > -- > Mit freundlichen Grüßen, > Best Regards, > > Jack Wang > > Linux Kernel Developer Storage > ProfitBricks GmbH The IaaS-Company. Regards, Duc Dang. -- To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
