Re: kern/54023: assertion panic during shutdown

[Hauke Fath <hf%spg.tu-darmstadt.de@localhost>]

The following reply was made to PR kern/54023; it has been noted by GNATS.

From: Hauke Fath <hf%spg.tu-darmstadt.de@localhost>
To: gnats-bugs%NetBSD.org@localhost
Cc: kern-bug-people%NetBSD.org@localhost, gnats-admin%NetBSD.org@localhost
Subject: Re: kern/54023: assertion panic during shutdown
Date: Thu, 7 Mar 2019 10:02:07 +0100

 On 3/1/19 10:30 PM, matthew green wrote:
 > set ddb.commandonenter
 >   to something like "show kernhist usbhist", then we may even get a
 >   useful usb log of what it was doing before the crash.  it may
 >   scroll a lot -- but the really useful stuff is likely in the last
 >   part [...]
 
 Interesting...
 
 This morning, the pinebook had rebooted. The message buffer starts with
 
 = 0xffff0000bf79f8a0, flags = 0x4, pipe = 0xffff0000bf67ac10, running = 1
 1551890685.395526 usb_insert_transfer#5404@2: called!
 1551890685.395527 usb_insert_transfer#5404@2: xfer = 0xffff0000bf79f8a0 
 pipe = 0xffff0000bf67ac10 running = 1 timeout = 0
 1551890685.395528 usb_insert_transfer#5404@2: <- done xfer 
 0xffff0000bf79f8a0, err 1
 1551890685.395529 usbd_transfer#5404@2: <- done transfer 
 0xffff0000bf79f8a0, err = 1
 1551890685.395529 usbd_transfer#5404@2: <- done xfer 0xffff0000bf79f8a0, 
 not sync (err 1)
 1551890685.395530 usbd_start_next#5868@2: called!
 1551890685.395531 usbd_start_next#5868@2: pipe = 0xffff0000bf67ac10, 
 xfer = 0xffff0000bf79f8a0
 db{2}> 
 bdeletematchsearchwbreakdmesgnextsetwatchbtdwatchpshowwhatiscexamineprintsiftingwritecallexitpsstepxcallouthelpquitsynccontinuekillreboottracedmachinesuntildb{2}> 
 db{2}> [  62
 .0496787] Bad character in number
 db{2}> [  62.0496787] Bad character
 [    62.049679] ?
 db{2}> [  62.0496787] Bad character in number
 db{2}> [  62.0496787] Bad character in number
 db{2}> [  62.0496787] Bad character
 
 followed by more random input (I had set 'options DDB_TEE_MSGBUF=1') and 
 several panics like
 
 fp ffffffc044d8f840 cpu_Debugger() at ffffffc00006eacc netbsd:cpu_Debugger
 fp ffffffc044d8f8a0 kern_assert() at ffffffc000569bc4 
 netbsd:kern_assert+0x5c
 fp ffffffc044d8f930 ehci_abort_xfer() at ffffffc000178ac4 
 netbsd:ehci_abort_xfer+0x2c4
 fp ffffffc044d8f990 usbd_ar_pipe() at ffffffc0000c1000 
 netbsd:usbd_ar_pipe+0x1c8
 fp ffffffc044d8f9e0 usbd_abort_pipe() at ffffffc0000c1390 
 netbsd:usbd_abort_pipe+0x28
 fp ffffffc044d8fa00 uhidev_stop() at ffffffc0000de8e8 
 netbsd:uhidev_stop+0x30
 fp ffffffc044d8fa20 uhidclose() at ffffffc0000df374 netbsd:uhidclose+0x64
 fp ffffffc044d8fa50 cdev_close() at ffffffc000425e58 netbsd:cdev_close+0xb8
 fp ffffffc044d8fa90 spec_close() at ffffffc0004ae990 netbsd:spec_close+0x1a0
 fp ffffffc044d8fae0 VOP_CLOSE() at ffffffc0004a2084 netbsd:VOP_CLOSE+0x3c
 fp ffffffc044d8fb20 vn_close() at ffffffc00049a67c netbsd:vn_close+0x34
 fp ffffffc044d8fb50 closef() at ffffffc0003d7ea8 netbsd:closef+0x50
 fp ffffffc044d8fb90 fd_free() at ffffffc0003db3b4 netbsd:fd_free+0x1ac
 fp ffffffc044d8fc00 exit1() at ffffffc0003e2d18 netbsd:exit1+0xe8
 fp ffffffc044d8fcc0 sigexit() at ffffffc000407e78 netbsd:sigexit+0x1d8
 fp ffffffc044d8fd10 postsig() at ffffffc00040822c netbsd:postsig+0x26c
 fp ffffffc044d8fdb0 lwp_userret() at ffffffc0003eccf0 
 netbsd:lwp_userret+0x1a0
 fp ffffffc044d8fe00 userret() at ffffffc00007063c netbsd:userret+0x6c
 fp ffffffc044d8fe20 syscall() at ffffffc0000703cc netbsd:syscall+0x104
 tf ffffffc044d8fed0 el0_trap() at ffffffc00006ea10 netbsd:el0_trap
 
 till the machine eventually rebooted:
 
 [    62.049679] dump to dev 92,17 not possible
 [    62.049679] rebooting...
 
 Is there any way to limit the output of ddb's 'show kernhhist'? The man 
 page entry is kind of obscure...
 
 Cheerio,
 hauke