Date: Thu, 3 Jun 1999 12:38:04 -0700 (PDT) From: Matthew Dillon <dillon@apollo.backplane.com> To: "David E. Cross" <crossd@cs.rpi.edu> Cc: freebsd-hackers@FreeBSD.ORG, schimken@cs.rpi.edu, crossd@cs.rpi.edu Subject: Re: 3.2-stable, panic #12 Message-ID: <199906031938.MAA99463@apollo.backplane.com> References: <199906031735.NAA37037@cs.rpi.edu>
next in thread | previous in thread | raw e-mail | index | archive | help
I'm not surprised. I think this is similar to the other known NFS panics.
I haven't had a chance to reproduce it yet ( may not be able to until
after USENIX ).
I have to say, though, that in order to fix these bugs I really do need
my commit privs back. If people want these things fixed, complain to core.
I have the time to fix the bugs with commit privs, but I just don't have
the time or inclination to fix them without. It is just too much stress
keeping a patch that should be committed in a day or two on the table for
two weeks and then have to beg people to deal with it. I am getting
wholely sick and tired of it, and I have better things to do then try
to maintain a pipeline of fixes that constantly get stopped up. I *want*
to fix these and other bugs, but I am effectively being prevented from
doing so because some core members freak out over the speed at which I
program and the amount of rewriting I sometimes do. I will point out that
all the rewriting I've done so far has been to the ultimate benefit of
the project in hind sight, resulting in better commented, cleaner,
and more reliable code and catching more bugs.
-Matt
Matthew Dillon
<dillon@backplane.com>
:Our home directory NFS server went down again today, "same bat-panic".
:This time it went down on ".Maillock" (usually it goes down on a netscape
:cache file or .Xauthorit-c. Piecing some more together I modified my old
:"crash_patoot.c" file (which didn't cause any problems), to the new and
:improved version that does.
:
:This is our environment:
:FreeBSD NFS server running 3.2-STABLE from 1.5 to 2 weeks ago. Multiple
:client machines of multiple architectures (Solaris 2.6, Irix 6.5.2+, FreeBSB
:3.2+). These crashes were all reproduced with a Solaris client, I do not know
:if it is reproduceable with other clients. Below is the short code segment
:that will cause the crash, the additions I added to it to cause the crash
:were rename(2) and unlink(2), without those I could not get a crash.
:
:Also, available upon request is a packet dump of all traffic to/from that
:machine leading to the crash (it is only 198336 bytes long, it was
:captured with '-s 1500' with tcpdump).
:
:Without further ado crash_patoot.c:
:int main(int argc, char **argv)
:{
: int fd;
: int counter;
: char newfilename[1024];
:
: strcpy(newfilename,argv[1]);
: strcat(newfilename,".old");
: for(counter=0;counter<1000000;counter++) {
: fd=open(argv[1], O_CREAT, 600);
: write(fd, &counter,4);
: rename(argv[1], newfilename);
: close(fd);
: rename(newfilename, argv[1]);
: fd=open(newfilename, O_CREAT,600);
: unlink(newfilename);
: close(fd);
: fd=open(newfilename, O_CREAT,600);
: close(fd);
: unlink(newfilename);
: }
: return 0;
:}
:
:If you are able to reproduce this panic please let me know. I want to be
:assured I am not going out of my mind. I am attempting to dig through the NFS
:code to try to find the bug myself, but it is a daunting task.
:
:--
:David Cross | email: crossd@cs.rpi.edu
:Systems Administrator/Research Programmer | Web: http://www.cs.rpi.edu/~crossd
:Rensselaer Polytechnic Institute, | Ph: 518.276.2860
:Department of Computer Science | Fax: 518.276.4033
:I speak only for myself. | WinNT:Linux::Linux:FreeBSD
:
:
:To Unsubscribe: send mail to majordomo@FreeBSD.org
:with "unsubscribe freebsd-hackers" in the body of the message
:
To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-hackers" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199906031938.MAA99463>