[DragonFlyBSD - Bug #3253] (New) HAMMER2 on NFS server. "(nlock & MTX_MASK) != MTX_MASK"

Tue Nov 3 14:46:18 PST 2020

Issue #3253 has been reported by yellowrabbit2010.

----------------------------------------
Bug #3253: HAMMER2 on NFS server. "(nlock & MTX_MASK) != MTX_MASK"
http://bugs.dragonflybsd.org/issues/3253

* Author: yellowrabbit2010
* Status: New
* Priority: Normal
* Assignee: 
* Category: 
* Target version: 
----------------------------------------
Hello,
the client machine writes (not very fast, like 150Kb/s) data to the DragonFlyBSD NFS server  with hammer2 partition.
The server dies periodically (3-4 days). These crashes do not depend on the time of continuous work (I tried to reboot several times a day), but rather just a coincidence of some circumstances.
I am using the master branch, radeon graphics driver with hardware acceleration is disabled. Crashes occur both when idle at night and when working during the day.

Here are the last three files, they refer to different commits on the master, of course, but the error is the same.

I have all the vmcore.? and kern.? files and I can put them wherever you specify.

======== kgdb -n 6 =========
Loaded symbols for /boot/kernel/radeonkmsfw_BONAIRE_vce.ko
_get_mycpu () at ./machine/thread.h:69
69          __asm ("movq %%gs:globaldata,%0" : "=r" (gd) : "m"(__mycpu__dummy));
(kgdb) bt
#0  _get_mycpu () at ./machine/thread.h:69
#1  panic (
    fmt=fmt at entry=0xffffffff80c25920 "assertion \"%s\" failed in %s at %s:%u")
    at /usr/src/sys/kern/kern_shutdown.c:868
#2  0xffffffff8066cf23 in _mtx_unlock (mtx=mtx at entry=0xfffff8039daaa040)
    at /usr/src/sys/kern/kern_mutex.c:725
#3  0xffffffff809464b0 in mtx_unlock (mtx=0xfffff8039daaa040)
    at /usr/src/sys/sys/mutex2.h:358
#4  mtx_lock_temp_release (mtx=0xfffff8039daaa040)
    at /usr/src/sys/sys/mutex2.h:491
#5  hammer2_inode_lock_temp_release (ip=0xfffff8039daa9fc0)
    at /usr/src/sys/vfs/hammer2/hammer2_inode.c:496
#6  hammer2_igetv (ip=ip at entry=0xfffff8039daa9fc0,
    errorp=errorp at entry=0xfffff8038047926c)
    at /usr/src/sys/vfs/hammer2/hammer2_inode.c:685
#7  0xffffffff8094fc24 in hammer2_vfs_vget (mp=<optimized out>,
    dvp=dvp at entry=0x0, ino=<optimized out>, ino at entry=14263250,
    vpp=vpp at entry=0xfffff80380479388)
    at /usr/src/sys/vfs/hammer2/hammer2_vfsops.c:1911
#8  0xffffffff80950596 in hammer2_vfs_fhtovp (mp=<optimized out>,
    rootvp=<optimized out>, fhp=<optimized out>, vpp=0xfffff80380479388)
    at /usr/src/sys/vfs/hammer2/hammer2_vfsops.c:2871
#9  0xffffffff807fc64f in nfsrv_fhtovp (fhp=fhp at entry=0xfffff803804793c8,
    lockflag=lockflag at entry=1, mpp=mpp at entry=0xfffff80380479390,
    vpp=vpp at entry=0xfffff80380479388, cred=cred at entry=0xfffff8039b2614e0,
    slp=slp at entry=0xfffff8038057f240, nam=0xfffff80157724070,
    rdonlyp=0xfffff80380479384, kerbflag=0, pubflag=1)
    at /usr/src/sys/vfs/nfs/nfs_subs.c:1247
#10 0xffffffff807ef326 in nfsrv_write (nfsd=0xfffff8039b261420,
    slp=0xfffff8038057f240, td=<optimized out>, mrq=0xfffff80380479708)
    at /usr/src/sys/vfs/nfs/nfs_serv.c:1089
#11 0xffffffff807fe23f in nfssvc_nfsd (td=<optimized out>,
    argp=<optimized out>, nsd=<optimized out>)
    at /usr/src/sys/vfs/nfs/nfs_syscalls.c:678
#12 sys_nfssvc (sysmsg=<optimized out>, uap=<optimized out>)
    at /usr/src/sys/vfs/nfs/nfs_syscalls.c:300
#13 0xffffffff80bc347e in syscall2 (frame=0xfffff803804799f8)
    at /usr/src/sys/platform/pc64/x86_64/trap.c:1284
#14 0xffffffff80b8664d in Xfast_syscall ()
    at /usr/src/sys/platform/pc64/x86_64/exception.S:448
#15 0x000000000000002b in ?? ()
Backtrace stopped: previous frame inner to this frame (corrupt stack?)
(kgdb)
====================

======== kgdb -n 4 =========
Loaded symbols for /boot/kernel/radeonkmsfw_BONAIRE_vce.ko
_get_mycpu () at ./machine/thread.h:69
69          __asm ("movq %%gs:globaldata,%0" : "=r" (gd) : "m"(__mycpu__dummy));
(kgdb)
(kgdb) bt
#0  _get_mycpu () at ./machine/thread.h:69
#1  panic (
    fmt=fmt at entry=0xffffffff80c5ec30 "assertion \"%s\" failed in %s at %s:%u")
    at /usr/src/sys/kern/kern_shutdown.c:868
#2  0xffffffff8066b973 in _mtx_unlock (mtx=mtx at entry=0xfffff8031a395a40)
    at /usr/src/sys/kern/kern_mutex.c:725
#3  0xffffffff809447c0 in mtx_unlock (mtx=0xfffff8031a395a40)
    at /usr/src/sys/sys/mutex2.h:358
#4  mtx_lock_temp_release (mtx=0xfffff8031a395a40)
    at /usr/src/sys/sys/mutex2.h:491
#5  hammer2_inode_lock_temp_release (ip=0xfffff8031a3959c0)
    at /usr/src/sys/vfs/hammer2/hammer2_inode.c:496
#6  hammer2_igetv (ip=ip at entry=0xfffff8031a3959c0,
    errorp=errorp at entry=0xfffff8031312e22c)
    at /usr/src/sys/vfs/hammer2/hammer2_inode.c:685
#7  0xffffffff8094de84 in hammer2_vfs_vget (mp=<optimized out>,
    dvp=dvp at entry=0x0, ino=<optimized out>, ino at entry=14263012,
    vpp=vpp at entry=0xfffff8031312e348)
    at /usr/src/sys/vfs/hammer2/hammer2_vfsops.c:1922
#8  0xffffffff8094e7c6 in hammer2_vfs_fhtovp (mp=<optimized out>,
    rootvp=<optimized out>, fhp=<optimized out>, vpp=0xfffff8031312e348)
    at /usr/src/sys/vfs/hammer2/hammer2_vfsops.c:2895
#9  0xffffffff807fa90f in nfsrv_fhtovp (fhp=fhp at entry=0xfffff8031312e388,
    lockflag=lockflag at entry=1, mpp=mpp at entry=0xfffff8031312e350,
    vpp=vpp at entry=0xfffff8031312e348, cred=cred at entry=0xfffff80325a10660,
    slp=slp at entry=0xfffff8031316fac0, nam=0xfffff800e80a4130,
    rdonlyp=0xfffff8031312e344, kerbflag=0, pubflag=1)
    at /usr/src/sys/vfs/nfs/nfs_subs.c:1232
#10 0xffffffff807ed7d6 in nfsrv_write (nfsd=0xfffff80325a105a0,
    slp=0xfffff8031316fac0, td=<optimized out>, mrq=0xfffff8031312e6c8)
    at /usr/src/sys/vfs/nfs/nfs_serv.c:1089
#11 0xffffffff807fc51f in nfssvc_nfsd (td=<optimized out>,
    argp=<optimized out>, nsd=<optimized out>)
    at /usr/src/sys/vfs/nfs/nfs_syscalls.c:678
#12 sys_nfssvc (uap=<optimized out>) at /usr/src/sys/vfs/nfs/nfs_syscalls.c:300
#13 0xffffffff80babbeb in syscall2 (frame=0xfffff8031312e9f8)
    at /usr/src/sys/platform/pc64/x86_64/trap.c:1319
#14 0xffffffff80b85d5d in Xfast_syscall ()
    at /usr/src/sys/platform/pc64/x86_64/exception.S:450
#15 0x000000000000002b in ?? ()
Backtrace stopped: previous frame inner to this frame (corrupt stack?)
(kgdb)
=============================

---Files--------------------------------
core.txt.4 (454 KB)
core.txt.5 (489 KB)
core.txt.6 (431 KB)
ls-crash.png (112 KB)

-- 
You have received this notification because you have either subscribed to it, or are involved in it.
To change your notification preferences, please click here: http://bugs.dragonflybsd.org/my/account