Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 27 Aug 2019 07:18:22 +0200
From:      Mateusz Guzik <mjguzik@gmail.com>
To:        Cy Schubert <Cy.Schubert@cschubert.com>
Cc:        src-committers@freebsd.org, svn-src-all@freebsd.org,  svn-src-head@freebsd.org
Subject:   Re: svn commit: r351472 - head/sys/fs/nullfs
Message-ID:  <CAGudoHGaCEpOyd6c7UpVEM%2BFVEW0gGL46b0A5mTN2ES=mA_h9w@mail.gmail.com>
In-Reply-To: <201908270417.x7R4HcfE024857@slippy.cwsent.com>
References:  <201908250513.x7P5DFFx017824@repo.freebsd.org> <6C6BF3A4-DC07-4D9A-9A5F-FA075BF535A8@cschubert.com> <201908261411.x7QEBMWi075354@slippy.cwsent.com> <A961554B-2B0B-4382-BA73-D69FC0129D0B@cschubert.com> <C3B4FE65-FC10-4248-AB00-0E30532C77EB@cschubert.com> <CAGudoHHGLguWZZMh%2BTOCBwANKKp7-6z5JvP1NgRo3jAr4vZ%2BPw@mail.gmail.com> <201908270025.x7R0PtNT008594@slippy.cwsent.com> <CAGudoHHucKGxYKz9L46w-KAhbAKQTXZD6syBPer2aHwh37vx0Q@mail.gmail.com> <201908270341.x7R3fJne062385@slippy.cwsent.com> <CAGudoHE=L8MQy=Sb5h04BZvehiSaxmek-kYmEdZ1PpCM2kM24A@mail.gmail.com> <201908270417.x7R4HcfE024857@slippy.cwsent.com>

next in thread | previous in thread | raw e-mail | index | archive | help
This survived light testing for me:

https://people.freebsd.org/~mjg/unionfs-interlock.diff

note unionfs instantly panics with DEBUG_VFS_LOCKS regardless
of any of this work.

On 8/27/19, Cy Schubert <cy.schubert@gmail.com> wrote:
> In message
> <CAGudoHE=L8MQy=Sb5h04BZvehiSaxmek-kYmEdZ1PpCM2kM24A@mail.gmail.c
> om>
> , Mateusz Guzik writes:
>> On 8/27/19, Cy Schubert <cy.schubert@gmail.com> wrote:
>> > Making progress:
>> >
>> > Unread portion of the kernel message buffer:
>> >
>> > panic: null_unlock: LK_INTERLOCK passed
>> > cpuid = 2
>> > time = 1566868081
>> > KDB: stack backtrace:
>> > db_trace_self_wrapper() at db_trace_self_wrapper+0x2b/frame
>> > 0xfffffe0054a954e0
>> > vpanic() at vpanic+0x19d/frame 0xfffffe0054a95530
>> > panic() at panic+0x43/frame 0xfffffe0054a95590
>> > null_unlock() at null_unlock+0x8a/frame 0xfffffe0054a955d0
>> > VOP_UNLOCK_APV() at VOP_UNLOCK_APV+0x56/frame 0xfffffe0054a955f0
>> > unionfs_unlock() at unionfs_unlock+0x124/frame 0xfffffe0054a95650
>> > VOP_UNLOCK_APV() at VOP_UNLOCK_APV+0x56/frame 0xfffffe0054a95670
>> > vfs_domount() at vfs_domount+0xc3f/frame 0xfffffe0054a958a0
>> > vfs_donmount() at vfs_donmount+0x934/frame 0xfffffe0054a95940
>> > sys_nmount() at sys_nmount+0x69/frame 0xfffffe0054a95980
>> > amd64_syscall() at amd64_syscall+0x3ae/frame 0xfffffe0054a95ab0
>> > fast_syscall_common() at fast_syscall_common+0x101/frame
>> > 0xfffffe0054a95ab0
>> > --- syscall (378, FreeBSD ELF64, sys_nmount), rip = 0x2bed6a4a, rsp =
>> > 0x7fffffffe2e8, rbp = 0x7fffffffec70 ---
>> > Uptime: 1m35s
>> >
>>
>> Ok, this explains everything. I did not think anyone uses unionfs,
>> let alone stacked on top of nullfs. I'll have a patch later.
>
> It's for a few jails. Thanks for persevering.
>
>
> --
> Cheers,
> Cy Schubert <Cy.Schubert@cschubert.com>
> FreeBSD UNIX:  <cy@FreeBSD.org>   Web:  http://www.FreeBSD.org
>
> 	The need of the many outweighs the greed of the few.
>
>
>


-- 
Mateusz Guzik <mjguzik gmail.com>



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAGudoHGaCEpOyd6c7UpVEM%2BFVEW0gGL46b0A5mTN2ES=mA_h9w>