Re: New warnings from WITNESS

From: Michael Tuexen <tuexen_at_freebsd.org>
Date: Sun, 6 Nov 2016 20:19:21 +0100
> On 6 Nov 2016, at 19:41, Scott Long <scottl_at_samsco.org> wrote:
> 
> 
>> On Nov 6, 2016, at 11:01 AM, Michael Tuexen <tuexen_at_FreeBSD.org> wrote:
>> 
>>> On 6 Nov 2016, at 15:39, Konstantin Belousov <kostikbel_at_gmail.com> wrote:
>>> 
>>> On Sun, Nov 06, 2016 at 02:17:45PM +0100, Michael Tuexen wrote:
>>>>> On 6 Nov 2016, at 13:28, Konstantin Belousov <kostikbel_at_gmail.com> wrote:
>>>>> 
>>>>> On Sun, Nov 06, 2016 at 12:50:12PM +0100, Michael Tuexen wrote:
>>>>>> bus_dmamap_create with the following non-sleepable locks held:
>>>>>> exclusive sleep mutex mpt (mpt) r = 0 (0xfffffe0000e2f008) locked _at_ dev/mpt/mpt.c:2287
>>>>>> stack backtrace:
>>>>>> #0 0xffffffff80ac0300 at witness_debugger+0x70
>>>>>> #1 0xffffffff80ac15e7 at witness_warn+0x3d7
>>>>>> #2 0xffffffff81055fef at bus_dmamap_create+0x2f
>>>>>> #3 0xffffffff80678a25 at mpt_configure_ioc+0x3a5
>>>>>> #4 0xffffffff80677476 at mpt_attach+0x226
>>>>>> #5 0xffffffff80683299 at mpt_pci_attach+0x9c9
>>>>>> #6 0xffffffff80a9478d at device_attach+0x41d
>>>>>> #7 0xffffffff80a9595a at bus_generic_attach+0x4a
>>>>>> #8 0xffffffff806ebe75 at pci_attach+0xd5
>>>>>> #9 0xffffffff80a9478d at device_attach+0x41d
>>>>>> #10 0xffffffff80a9595a at bus_generic_attach+0x4a
>>>>>> #11 0xffffffff803c11a2 at acpi_pcib_acpi_attach+0x402
>>>>>> #12 0xffffffff80a9478d at device_attach+0x41d
>>>>>> #13 0xffffffff80a9595a at bus_generic_attach+0x4a
>>>>>> #14 0xffffffff803b4c8f at acpi_attach+0xdbf
>>>>>> #15 0xffffffff80a9478d at device_attach+0x41d
>>>>>> #16 0xffffffff80a9595a at bus_generic_attach+0x4a
>>>>>> #17 0xffffffff80ee03e3 at nexus_acpi_attach+0x73
>>>>>> 
>>>>>> ... and so on. Not sure which revision introduced it...
>>>>> r308268
>>>>> 
>>>>> I believe that this is an mpt(4) driver issue, which calls
>>>>> bus_dmamap_create(9) with the mpt mutex held.
>>>> OK. Whom to contact? Or are you willing to look into it?
>>>> I haven't worked in that area...
>>> 
>>> I am really not sure.  Looking at the svn history is not very encouraging.
>> Hmm. Time to change my setup, I guess...
>>> 
>>> Might be try freebsd-scsi_at_ as well.
>> I dropped them an e-mail...
>> 
>> 
> 
> Please see my response on that list.
Will do once it shows up in the archive since I'm not subscribed
to that list...

Best regards
Michael
> 
> Scott
> 
> 
Received on Sun Nov 06 2016 - 18:19:28 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:41:08 UTC