Re: fsck segfaults on rpi3 running 13-stable (and on 14-CURRENT analyzing the same file system that resulted from the 13-STABLE crash)

bob prohaska wrote this message on Tue, Feb 14, 2023 at 08:14 -0800:
> > On Feb 13, 2023, at 15:25, John-Mark Gurney <[email protected]> wrote:
> > 
> [huge snip]
> 
> > > Ok, decided to run AFL on fsck, and this one was the first crash it
> > > discovered.  The problem is that ctime can return NULL, and the return
> > > value isn't checked, because it then immediately does &p[4] which
> > > results is printf and friends being passed 0x4.
> > > 
> > > Simple test program that demonstrates this problem:
> > > #include <time.h>
> > > #include <stdio.h>
> > > 
> > > int
> > > main()
> > > {
> > >        const char *p;
> > >        time_t t;
> > > 
> > >        t = -5098919203113507862;
> > > 
> > >        p = ctime(&t);
> > > 
> > >        printf("MTIME=%12.12s %4.4s ", &p[4], &p[20]);
> > > 
> > >        return 0;
> > > }
> > > 
> > > I'm not sure what the correct fix is for when times are wildly out of
> > > valid range.
> > > 
> 
> Is this a demonstration that the fsck segfault can be reproduced 
> independtly of my particular corrupt filesystem? AFL is new to me. 

Yes, it is.  It turns out that the FS to produce this failure is a LOT
smaller than I expected when compresed, I have included it later in the
email.  The constant above was taken directly from the failing FS.

AFL is a very useful tool, and found this crash and apparently 50+
other crashes in only 5-10 minutes of running... I'll be investigating
a few of the other crashes as well, as fsck does ocassionally deal w/
untrusted fs's.

Also, afl was a LOT easier to run than I expected (I had run it before,
just not insturmented a FreeBSD program before), installed
afl++-llvm from ports, went into the fsck_ffs dir, and ran:
CC=/usr/local/afl++-llvm/bin/afl-cc CXX=/usr/local/afl++-llvm/bin/afl-c++ make

I then created a simple starting file system:
mkdir testcase_dir; cd testcase_dir
touch -s 1m test.ufs; mdconfig -f test.ufs; newfs /dev/md0; mdconfig -d -u 0
cd ..

And then ran it against that test case:
mkdir findings_dir
/usr/local/afl++-llvm/bin/afl-fuzz -i testcase_dir/ -o findings_dir/ /usr/obj/usr/src/arm64.aarch64/sbin/fsck_ffs/fsck_ffs -y @@

To get back the FS, pipe in the base64 blob below through:
openssl enc -d -base64 | xz -d > somefile

And then run fsck_ffs on it.

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-- 
  John-Mark Gurney				Voice: +1 415 225 5579

     "All that I will do, has been done, All that I have, has not."