* How to fix the ptrace flaw without rebooting
@ 2003-04-03 12:05 Chuck Ebbert
2003-04-03 12:30 ` Erik Hensema
0 siblings, 1 reply; 6+ messages in thread
From: Chuck Ebbert @ 2003-04-03 12:05 UTC (permalink / raw)
To: linux-kernel
#define PROG_MSG "pt_fix.c -- disable linux ptrace system call"
#define VERSION "0.50"
/*
Author: Chuck Ebbert
Adapted from: Phrack Inc., Volume 0x0b, Issue 0x3a, Phile #0x07
Directions: Run this program as root on an x86 machine.
It will disable the ptrace system call, thus
fixing the Linux 'ptrace flaw'. (It will also
break strace, debugging tools and User Mode Linux.)
WARNING: Your computer may crash or do other strange things
if you run this program as root. No warranty.
Tested on:
PPro Uni XT-PIC 2.2.19 pgcc-1.1.3 (sig #2)
PPro SMP IO-APIC 2.5.66 gcc-2.96-110 (#1)
PPro SMP IO_APIC 2.4.20aa1 gcc-2.96-110 (#1)
PII Uni IO-APIC 2.5.66 gcc-2.96-110 (#1)
K7 Uni APIC 2.4.20aa1 gcc-2.96-110 (#1)
*/
#include <stdio.h>
#include <fcntl.h>
#include <sys/types.h>
struct {
unsigned short limit;
unsigned int base;
} __attribute__ ((packed)) idtr;
struct {
unsigned short off1;
unsigned short sel;
unsigned char none,flags;
unsigned short off2;
} __attribute__ ((packed)) desc;
/* read from kmem */
static int rkm(int fd, int offset, void *buf, int size) {
if (lseek(fd, offset, 0) != offset) return 0;
if (read(fd, buf, size) != size) return 0;
return size;
}
/* write to kmem */
static int wkm(int fd, int offset, void *buf, int size) {
if (lseek(fd, offset, 0) != offset) return 0;
if (write(fd, buf, size) != size) return 0;
return size;
}
void perr(char *err) { /* exit with err msg */
perror(err), exit(1);
}
/* read/write wrapper with err handling */
#define xkm(f, p1, p2, p3, p4, msg) ({ \
int _p4 = (p4); if ((f)((p1), (p2), (p3), _p4) != _p4) perr(msg); })
#define IRQ_VECTOR 80 /* syscall interrupt */
#define CALLNR 26 /* ptrace syscall */
/* valid old code signatures */
/* 1 */ unsigned char v1code[1] = { 0x55 /* push ebp */ };
/* 2 */ unsigned char v2code[4] = {
0x83,0xec,0x10 /* sub esp,10 */, 0x55 /* push ebp */ };
/* new code for syscall */
unsigned char ncode[4] = { 0x31,0xc0 /* xor eax,eax */,
0x48 /* dec eax */, 0xc3 /* ret */ };
unsigned char ocode[16]; /* buffer for old ptrace code */
unsigned char sc_asm[100]; /* holds first 100 bytes of IRQ handler code */
main() {
unsigned sys_call_off, sct, fn;
int kmem, i, found = 0;
unsigned char *p;
printf(PROG_MSG ", version " VERSION "\n");
asm("sidt %0" : "=m" (idtr));
printf("idt is at %08x\n", idtr.base);
kmem = open("/dev/kmem", O_RDWR);
if (kmem < 0) perr("open kmem");
/* read-in desc for int vector */
xkm(rkm, kmem, idtr.base+8*0x80, &desc, sizeof(desc), "rkm desc");
sys_call_off = (desc.off2 << 16) | desc.off1;
printf("idt entry %d: flags=%02hhx sel=%04x off=%08x\n",
IRQ_VECTOR, desc.flags, desc.sel, sys_call_off);
/* we have syscall routine address now, look for syscall table
dispatch (indirect call) */
xkm(rkm, kmem, sys_call_off, sc_asm, sizeof(sc_asm), "rkm sc_asm");
p = (char *)memmem(sc_asm, sizeof(sc_asm), "\xff\x14\x85", 3);
sct = *(unsigned *)(p + 3); /* will fault if p is bad */
printf("sys_call_table is at %08x\n", sct);
xkm(rkm, kmem, sct+4*CALLNR, &fn, sizeof(fn), "rkm fn");
printf("entry %d points to %08x\n", CALLNR, fn);
xkm(rkm, kmem, fn, ocode, sizeof(ocode), "rkm ocode");
printf("code:");
for (i = 0; i < sizeof(ocode); i++) printf(" %02hhx", ocode[i]);
printf("\n");
if (!memcmp(ocode, v1code, sizeof(v1code))) found = 1;
if (!memcmp(ocode, v2code, sizeof(v2code))) found = 2;
if (!found) {
printf(!memcmp(ocode, ncode, sizeof(ncode)) ?
"Already installed.\n" :
"Code mismatch, aborting.\n");
exit(1);
}
printf("Valid entry code signature #%d found.\n", found);
xkm(wkm, kmem, fn, ncode, sizeof(ncode), "wkm ncode");
printf("Kernel patch succeeded.\n"
"NOTE: patch will only work until next reboot.\n");
close(kmem);
exit(0);
}
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: How to fix the ptrace flaw without rebooting
2003-04-03 12:05 How to fix the ptrace flaw without rebooting Chuck Ebbert
@ 2003-04-03 12:30 ` Erik Hensema
2003-04-04 6:19 ` Nuno Silva
0 siblings, 1 reply; 6+ messages in thread
From: Erik Hensema @ 2003-04-03 12:30 UTC (permalink / raw)
To: linux-kernel
Chuck Ebbert (76306.1226@compuserve.com) wrote:
> Author: Chuck Ebbert
> Adapted from: Phrack Inc., Volume 0x0b, Issue 0x3a, Phile #0x07
> Directions: Run this program as root on an x86 machine.
> It will disable the ptrace system call, thus
> fixing the Linux 'ptrace flaw'. (It will also
> break strace, debugging tools and User Mode Linux.)
> WARNING: Your computer may crash or do other strange things
> if you run this program as root. No warranty.
If you can't reboot to apply a security fix, you've got a serious problem.
A better fix in a running system is to simply disable dynamic module
loading: echo /no/such/file > /proc/sys/kernel/modprobe
At the very least you can be sure your machine won't crash this way ;-)
--
Erik Hensema <erik@hensema.net>
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: How to fix the ptrace flaw without rebooting
2003-04-03 12:30 ` Erik Hensema
@ 2003-04-04 6:19 ` Nuno Silva
0 siblings, 0 replies; 6+ messages in thread
From: Nuno Silva @ 2003-04-04 6:19 UTC (permalink / raw)
To: erik; +Cc: linux-kernel
Hello!
Erik Hensema wrote:
>
> If you can't reboot to apply a security fix, you've got a serious problem.
>
> A better fix in a running system is to simply disable dynamic module
> loading: echo /no/such/file > /proc/sys/kernel/modprobe
> At the very least you can be sure your machine won't crash this way ;-)
>
IIRC, dynamic module loading is not required to exploit all the bugs
present in ptrace. Luckly all the exploits floating around require kmod :)
Regards,
Nuno Silva
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: How to fix the ptrace flaw without rebooting
2003-04-04 11:18 Chuck Ebbert
@ 2003-04-04 11:15 ` Alan Cox
2003-04-06 1:28 ` Paul Jakma
0 siblings, 1 reply; 6+ messages in thread
From: Alan Cox @ 2003-04-04 11:15 UTC (permalink / raw)
To: Chuck Ebbert; +Cc: Linux Kernel Mailing List
On Gwe, 2003-04-04 at 12:18, Chuck Ebbert wrote:
> Erik Hensema wrote:
>
>
> > A better fix in a running system is to simply disable dynamic module
> > loading: echo /no/such/file > /proc/sys/kernel/modprobe
>
>
> You mean like this?
>
> # echo 'x'>/proc/sys/kernel/modprobe
> bash: /proc/sys/kernel/modprobe: No such file or directory
Thats not a sufficient fix except for people blindly running the
example exploit
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: How to fix the ptrace flaw without rebooting
@ 2003-04-04 11:18 Chuck Ebbert
2003-04-04 11:15 ` Alan Cox
0 siblings, 1 reply; 6+ messages in thread
From: Chuck Ebbert @ 2003-04-04 11:18 UTC (permalink / raw)
To: linux-kernel
Erik Hensema wrote:
> A better fix in a running system is to simply disable dynamic module
> loading: echo /no/such/file > /proc/sys/kernel/modprobe
You mean like this?
# echo 'x'>/proc/sys/kernel/modprobe
bash: /proc/sys/kernel/modprobe: No such file or directory
:)
--
Chuck
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: How to fix the ptrace flaw without rebooting
2003-04-04 11:15 ` Alan Cox
@ 2003-04-06 1:28 ` Paul Jakma
0 siblings, 0 replies; 6+ messages in thread
From: Paul Jakma @ 2003-04-06 1:28 UTC (permalink / raw)
To: Alan Cox; +Cc: Chuck Ebbert, Linux Kernel Mailing List
On 4 Apr 2003, Alan Cox wrote:
> Thats not a sufficient fix except for people blindly running the
> example exploit
Also, removing CAP_SYS_PTRACE from the capability bounded set does
not close the hole (well, as shown by the exploit) either. It seems
CAP_SYS_PTRACE only closes the case of PTRACE_ATTACH, all other
ptrace()'s work, eg PTRACE_TRACEME -> strace echo foo and the exploit
works too.
Should CAP_SYS_PTRACE not do as the name suggests and completely
disallow ptrace()? (not just PTRACE_ATTACH).
regards,
--
Paul Jakma paul@clubi.ie paul@jakma.org Key ID: 64A2FF6A
warning: do not ever send email to spam@dishone.st
Fortune:
An egghead is one who stands firmly on both feet, in mid-air, on both
sides of an issue.
-- Homer Ferguson
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2003-04-06 1:16 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2003-04-03 12:05 How to fix the ptrace flaw without rebooting Chuck Ebbert
2003-04-03 12:30 ` Erik Hensema
2003-04-04 6:19 ` Nuno Silva
-- strict thread matches above, loose matches on Subject: below --
2003-04-04 11:18 Chuck Ebbert
2003-04-04 11:15 ` Alan Cox
2003-04-06 1:28 ` Paul Jakma
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox