Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 6 Jul 2018 09:33:06 -0700
From:      John Baldwin <jhb@FreeBSD.org>
To:        Stefan Parvu <sparvu@kronometrix.org>, freebsd-arm@freebsd.org
Subject:   Re: RPI3B+ UFS filesystem issues
Message-ID:  <3ac73628-4174-1277-67d9-fadd16e50891@FreeBSD.org>
In-Reply-To: <B58A2AD4-D74F-4D17-9E5E-BF1A802DB71C@kronometrix.org>
References:  <B58A2AD4-D74F-4D17-9E5E-BF1A802DB71C@kronometrix.org>

next in thread | previous in thread | raw e-mail | index | archive | help
On 7/3/18 10:07 PM, Stefan Parvu wrote:
> Are these signal of troubles form the filesystem layer ? Im using a Transcend 64GB 
> https://fi.transcend-info.com/Products/No-697. Dont know if I can benefit from this
> card of its full speed on RPI3B+ .
> 
> Im using FreeBSD 12.0 current latest 28 June snapshot:
> FreeBSD generic 12.0-CURRENT FreeBSD 12.0-CURRENT #0 r335760: Thu Jun 28 16:32:04 UTC 2018     root@releng3.nyi.freebsd.org:/usr/obj/usr/src/arm64.aarch64/sys/GENERIC  arm64
> 
> lock order reversal:
>  1st 0xffff000040d299f0 bufwait (bufwait) @ /usr/src/sys/kern/vfs_bio.c:3916
>  2nd 0xfffffd001582ba00 dirhash (dirhash) @ /usr/src/sys/ufs/ufs/ufs_dirhash.c:289
> stack backtrace:
> #0 0xffff0000003eb5e4 at witness_debugger+0x64
> #1 0xffff00000038f2cc at _sx_xlock+0x7c
> #2 0xffff0000006124d4 at ufsdirhash_add+0x38
> #3 0xffff000000614f68 at ufs_direnter+0x3b0
> #4 0xffff00000061d43c at ufs_makeinode+0x4b8
> #5 0xffff000000619b14 at ufs_create+0x38
> #6 0xffff0000006ccf10 at VOP_CREATE_APV+0xac
> #7 0xffff00000045d9c4 at vn_open_cred+0x264
> #8 0xffff000000456d00 at kern_openat+0x208
> #9 0xffff000000696128 at do_el0_sync+0x4b0
> #10 0xffff00000067c9f0 at handle_el0_sync+0x74
> lock order reversal:
>  1st 0xfffffd00038a39c8 ufs (ufs) @ /usr/src/sys/ufs/ffs/ffs_vfsops.c:1677
>  2nd 0xffff000040d299f0 bufwait (bufwait) @ /usr/src/sys/ufs/ffs/ffs_vnops.c:282
>  3rd 0xfffffd00038a37e8 ufs (ufs) @ /usr/src/sys/kern/vfs_subr.c:2572
> stack backtrace:
> #0 0xffff0000003eb5e4 at witness_debugger+0x64
> #1 0xffff00000035c340 at lockmgr_xlock_hard+0x6c
> #2 0xffff00000035cbbc at __lockmgr_args+0x4ac
> #3 0xffff00000060da28 at ffs_lock+0x88
> #4 0xffff0000006cfc58 at VOP_LOCK1_APV+0xac
> #5 0xffff00000045e050 at _vn_lock+0x64
> #6 0xffff00000044e5b8 at vget+0x78
> #7 0xffff0000004413bc at vfs_hash_get+0xec
> #8 0xffff000000609a58 at ffs_vgetf+0x44
> #9 0xffff000000600698 at softdep_sync_buf+0x9f4
> #10 0xffff00000060e764 at ffs_syncvnode+0x26c
> #11 0xffff0000005e8738 at ffs_truncate+0x6b4
> #12 0xffff000000615330 at ufs_direnter+0x778
> #13 0xffff00000061d43c at ufs_makeinode+0x4b8
> #14 0xffff000000619b14 at ufs_create+0x38
> #15 0xffff0000006ccf10 at VOP_CREATE_APV+0xac
> #16 0xffff00000045d9c4 at vn_open_cred+0x264
> #17 0xffff000000456d00 at kern_openat+0x208

No these are known false positives in the lock order verifier.

-- 
John Baldwin



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?3ac73628-4174-1277-67d9-fadd16e50891>