Login directly as root.

sked for root's password in single mode. Hmm, if I got psyhical access to machine and ttyvX are in "insecure" mode and I know root's password I can just reboot machine and log in as root. So if "insecure" mode is a security feature, shouldn't this be in that way

Re: Hmm, sysctlfs.

On Tue, Dec 24, 2002 at 08:35:55AM +0100, Pawel Jakub Dawidek wrote: +> Hello hackers... +> +> I've wrote file system that operate on sysctl tree. +> Most of work is done, but some cosmetic changes are needed. +> If somebody think that this can be usefull... +> +>

Hmm, sysctlfs.

Hello hackers... I've wrote file system that operate on sysctl tree. Most of work is done, but some cosmetic changes are needed. If somebody think that this can be usefull... http://garage.freebsd.pl/sysctlfs.README http://garage.freebsd.pl/sysctlfs.tbz -- Pawel Jakub Da

Re: Syscall number.

I may catch even _all_ system call, so making separate functions for every syscall isn't good idea (if I understand You well). -- Pawel Jakub Dawidek UNIX Systems Administrator http://garage.freebsd.pl Am I Evil? Yes, I Am. msg38804/pgp0.pgp Description: PGP signature

Re: Syscall number.

return (tmp(p, uap)); } -- Pawel Jakub Dawidek UNIX Systems Administrator http://garage.freebsd.pl Am I Evil? Yes, I Am. msg38801/pgp0.pgp Description: PGP signature

Re: Syscall number.

On Fri, Dec 20, 2002 at 07:43:37PM +0100, Pawel Jakub Dawidek wrote: +> Simple example (from kld module): +> +> sysent[SYS_chmod].sy_call = myfunction; +> sysent[SYS_open].sy_call = myfunction; +> sysent[SYS_execve].sy_call = myfunction; +> +> int +> myfunction(register

Syscall number.

and I don't catch it). So is there some diffrent way to get syscall number (in -STABLE or in -CURRENT)? Anyone? -- Pawel Jakub Dawidek UNIX Systems Administrator http://garage.freebsd.pl Am I Evil? Yes, I Am. msg38792/pgp0.pgp Description: PGP signature

Re: tty/pty devices not safe in jail?

ay if You have not-jailed users. So this isn't a complete solution. But if there are no free ptys, I log in via: % ssh -vC /bin/sh -- Pawel Jakub Dawidek UNIX Systems Administrator http://garage.freebsd.pl Am I Evil? Yes, I Am. msg38115/pgp0.pgp Description: PGP signature

Re: Shared files within a jail

here (line 571 in /sys/kern/vfs_cache.c): if (vp->v_dd->v_id != vp->v_ddid) { numcwdfail1++; free(buf, M_TEMP); return (ENOTDIR); } If "dir3" is for example NFS mount-point t

Re: Shared files within a jail

ith jail-things. I'm wrong? -- Pawel Jakub Dawidek UNIX Systems Administrator http://garage.freebsd.pl Am I Evil? Yes, I Am. msg38042/pgp0.pgp Description: PGP signature

Re: Shared files within a jail

l stable? I'm using mount_null(8) for my jails for a long time and everything works fine. milla:root:~# mount | grep null | wc -l 22 -- Pawel Jakub Dawidek UNIX Systems Administrator http://garage.freebsd.pl Am I Evil? Yes, I Am. msg38016/pgp0.pgp Description: PGP signature

SMP and race conditions.

ere I could find some more information about it? For now I've wrote only comments in /sys/sys/lock.h, etc. There are any papers about programming in SMP? -- Pawel Jakub Dawidek UNIX Systems Administrator http://garage.freebsd.pl Am I Evil? Yes, I Am. msg37494/pgp0.pgp Description: PGP signature

Re: tracing exec system call

th, etc., has been resolved? +> +> This is useful if one wants to monitor a process and all the system calls it makes +and then disallow a few of them if suspicious. Take a look at: http://cerber.sourceforge.net If You want monitor only execve(), then rexec project should be

Re: Getfsstat(2), statfs(2) and fstatfs(2) jailed.

On Wed, Oct 02, 2002 at 12:45:36PM +0200, Oliver Fromme wrote: +> Pawel Jakub Dawidek <[EMAIL PROTECTED]> wrote: +> > With this kernel module for FreeBSD 4.x process in jail can see only +> > file systems that are mounted in jail's chroot directory. +> > +>

Bug in nullfs?

if (vp->v_dd->v_id != vp->v_ddid) { numcwdfail1++; free(buf, M_TEMP); return (ENOTDIR); } If "dir3" is for example NFS mount-point there are no problems. Any ideas? -- Pawel Jakub

Getfsstat(2), statfs(2) and fstatfs(2) jailed.

ot directory is mounted on. Statfs(2) and fstatfs(2) also returns cuted paths. Why this isn't a part of jail(2) implementation? Maybe I'm missing something? -- Pawel Jakub Dawidek UNIX Systems Administrator http://garage.freebsd.pl Am I Evil? Yes, I Am. msg37181/pgp0.pgp Description: PGP signature

Re: Changing process informations.

e crcopy() and setting P_SUGID. */ If "anything" really means "anything". -- Pawel Jakub Dawidek UNIX Systems Administrator http://garage.freebsd.pl Am I Evil? Yes, I Am. msg36926/pgp0.pgp Description: PGP signature

Syscalls numbers in -CURRENT.

number is stored) for i386, alpha, ia64, sparc64 and powerpc in -CURRENT? Thanks. -- Pawel Jakub Dawidek UNIX Systems Administrator http://garage.freebsd.pl Am I Evil? Yes, I Am. msg36646/pgp0.pgp Description: PGP signature

Re: chown() vs. setfown() prototype in vfs_syscalls.c

r and/or group of the file. */ } And: leila:root:~# egrep "#define.*VNOVAL" /sys/sys/vnode.h /sys/sys/vnode.h:#defineVNOVAL (-1) -- Pawel Jakub Dawidek UNIX Systems Administrator http://garage.freebsd.pl Am I Evil? Yes, I Am. msg36579/pgp0.pgp Description: PGP signature

Re: Replacing kernel functions.

On Wed, Aug 28, 2002 at 08:30:18PM +0200, Pawel Jakub Dawidek wrote: +> On Wed, Aug 28, 2002 at 05:03:23AM -0700, Terry Lambert wrote: +> +> -- Terry +> +> Thanks to Your help I've found maybe better way to do this. [...] +> But this should works, I'm testing it at th

Re: Replacing kernel functions.

complicated, because if I will change address of this function, it changes for every descriptor on this file system. So I need cache original address of fo_close() functions, etc. But this should works, I'm testing it at the moment. -- Pawel Jakub Dawidek UNIX Systems Administrator http://gara

Re: Replacing kernel functions.

ode to add a null_name pointer: [...] Yes, but I don't want to patch kernel at all. Everythings should be in one kld module and should works for all file systems (ufs/ffs, fat32, procfs, etc.). File system shouldn't be important. -- Pawel Jakub Dawidek UNIX Systems Administrator http://

Re: Replacing kernel functions.

(nice) way. +> Personally, I've used this technique myself, for the purposes +> of code instrumentation for profiling on an OD for which I +> lacked source code, but I would *never* consider exporting +> this as a general API. So when all kernel functions addresses will be in

Replacing kernel functions.

of archs supported by FreeBSD. So if You know how to port them, fell free to send me Your version:) Thanks! -- Pawel Jakub Dawidek UNIX Systems Administrator http://garage.freebsd.pl Am I Evil? Yes, I Am. msg36531/pgp0.pgp Description: PGP signature

JailNG.

Hello there... When jailNG will be commited? -- Pawel Jakub Dawidek UNIX Systems Administrator http://garage.freebsd.pl Am I Evil? Yes, I Am. msg36152/pgp0.pgp Description: PGP signature

Re: kernel hacking environment

#x27;t +> chain a 'dummy' machine off of a work machine with ethernet and serial +> debugging :) Just looking for ideas and maybe good urls or docs +> I got FreeBSD 4.6 STABLE in vmware2 on FreeBSD 4.6 STABLE and it is great. I got lots of panics every day and this is really supe

What for we need set-uid-root on passwd/chpass/etc.?:)

wd_parser will be broken, then users can't run it directly so I think if he wnt to use it, it should be security hole in passwd/chpass too. I know that passwd/chpass are safe (no security holes for long time or never), but what You think about this solution? -- Pawel Jakub Dawidek UNIX

Re: Kernel space: MALLOC() & TAILQ_*()

On Fri, Jul 12, 2002 at 01:36:22PM -0700, Julian Elischer wrote: +> did you get answers to this? +> Here You got example how to use SLIST_* macros, sysctls, etc.: http://garage.freebsd.pl/slist.tgz more examples on http://garage.freebsd.pl -- Pawel Jakub Dawidek UNIX Systems Administrato

Re: No suid crontab(1).

.securelevel' > 0 ? +> Yes, it does. I'm using uchg, not schg flags. -- Pawel Jakub Dawidek UNIX Systems Administrator http://garage.freebsd.pl Am I Evil? Yes, I Am. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message

Re: Non suid crontab(1).

On Thu, Jul 11, 2002 at 01:03:56PM +0200, Pawel Jakub Dawidek wrote: http://garage.freebsd.pl/crontab.diff http://garage.freebsd.pl/crontab.README.eng There You got updated patch, added chroot() to /var/cron as soon as possbile and fixed note about permissions, for users files should be

Re: No suid crontab(1).

I forgot, that should be always an empty file "root" in /var/cron/tabs and files of all users if we don't want to gives attacker their rights. But I think the best way is to change cron(8) that it will be check file owner. -- Pawel Jakub Dawidek UNIX Systems Adm

No suid crontab(1).

(chflags(n, UF_IMMUTABLE) < OK) + { + warn("chflags"); + unlink(n); + return (-2); + } + log_it(RealUser, Pid, "REPLACE", User); poke_daemon();  -- Pawel Jakub Dawidek UNIX Systems Administrator http://garage.freebsd.pl Am I Evil? Yes, I Am. To Unsubscribe: send mail to [EMAIL PROTECTED] with "unsubscribe freebsd-hackers" in the body of the message

Re: Procfs patch (for FreeBSD 4.x).

On Thu, Jun 20, 2002 at 11:13:02PM +0200, Pawel Jakub Dawidek wrote: +> What You think about something like that: +> [17:33:39] [ttyv8] [61] leila:root:/sys/miscfs/procfs# diff procfs_subr.c.orig +procfs_subr.c [...] +> This gives us new sysctl (vfs.procfs.umask) and with this we ca

Procfs patch (for FreeBSd 4.x).

What You think about something like that: [17:33:39] [ttyv8] [61] leila:root:/sys/miscfs/procfs# diff procfs_subr.c.orig procfs_subr.c 42a43 > #include 80a82,119 > > SYSCTL_NODE(_vfs, OID_AUTO, procfs, CTLFLAG_RW, 0, "vfs-procfs-level"); > > int procfsumask = 077; > > static int > sysctl_vfs_

How to catch closef()?

his, because in cerb (http://cerber.sourceforge.net) I need to have function that will give me filename from file descryptor and pid. Any ideas? I don't want to patch kernel directly. -- Pawel Jakub Dawidek Network Administrator. Am I Evil? Yes, I Am. msg33508/pgp0.pgp Description: PGP signature

<    1   2   3