x86, ptrace: require admin privileges for ptrace BTS extension

PATCH In review Markus Metzger 2 participants 3 messages Dec 22, 2008

Subsystems (2)

x86 ptrace support

lore

Share & tools

lore

Markus Metzger

author

2008-12-22 12:02 UTC #0 lore raw

PATCH 1/1 x86, ptrace: require admin privileges for ptrace BTS extension

Require admin privileges for ptrace BTS extension.

Reported-by Ingo Molnar <[email protected]>

Signed-off-by Markus Metzger <[email protected]>

---

arch/x86/kernel/ptrace.c +5 -0

Index: gits/arch/x86/kernel/ptrace.c
===================================================================
@@ -21,6 +21,7 @@
 #include <linux/audit.h>
 #include <linux/seccomp.h>
 #include <linux/signal.h>
+#include <linux/capability.h>
 
 #include <asm/uaccess.h>
 #include <asm/pgtable.h>

@@ -742,6 +743,10 @@
 	struct ptrace_bts_config cfg;
 	int error = 0;
 
+	error = -EPERM;
+	if (!capable(CAP_SYS_ADMIN))
+		goto errout;
+
 	error = -EOPNOTSUPP;
 	if (!bts_cfg.sizeof_bts)
 		goto errout;

---------------------------------------------------------------------
Intel GmbH
Dornacher Strasse 1
85622 Feldkirchen/Muenchen Germany
Sitz der Gesellschaft: Feldkirchen bei Muenchen
Geschaeftsfuehrer: Douglas Lusk, Peter Gleissner, Hannes Schwaderer
Registergericht: Muenchen HRB 47456 Ust.-IdNr.
VAT Registration No.: DE129385895
Citibank Frankfurt (BLZ 502 109 00) 600119052

This e-mail and any attachments may contain confidential material for the sole use of the intended recipient(s). Any review or distribution by others is strictly prohibited. If you are not the intended recipient, please contact the sender and delete all copies.

stephane eranian

2008-12-22 13:23 UTC #1 lore raw

Markus,

Quoted 2 lines · Markus Metzger

Require admin privileges for ptrace BTS extension.

Can you explain the motivations for this?

I thought the BTS extension was a per-process functionality. So how come I can debug and single step my process without admin privilege and I cannot capture its own branches. Does this have to do with user vs. kernel execution of the proces (BTS captures everything if I recall)?

Thanks.

Quoted 44 lines

Reported-by Ingo Molnar <[email protected]>
Signed-off-by Markus Metzger <[email protected]>
---
Index: gits/arch/x86/kernel/ptrace.c
===================================================================
arch/x86/kernel/ptrace.c 2008-12-22 11:03:01.000000000 +0100 renamed from arch/x86/kernel/ptrace.c 2008-12-22 09:09:25.000000000 +0100 +5 -1
@@ -21,6 +21,7 @@
 #include <linux/audit.h>
 #include <linux/seccomp.h>
 #include <linux/signal.h>
+#include <linux/capability.h>
 
 #include <asm/uaccess.h>
 #include <asm/pgtable.h>
@@ -742,6 +743,10 @@
       struct ptrace_bts_config cfg;
       int error = 0;
 
+       error = -EPERM;
+       if (!capable(CAP_SYS_ADMIN))
+               goto errout;
+
       error = -EOPNOTSUPP;
       if (!bts_cfg.sizeof_bts)
               goto errout;
---------------------------------------------------------------------
Intel GmbH
Dornacher Strasse 1
85622 Feldkirchen/Muenchen Germany
Sitz der Gesellschaft: Feldkirchen bei Muenchen
Geschaeftsfuehrer: Douglas Lusk, Peter Gleissner, Hannes Schwaderer
Registergericht: Muenchen HRB 47456 Ust.-IdNr.
VAT Registration No.: DE129385895
Citibank Frankfurt (BLZ 502 109 00) 600119052
This e-mail and any attachments may contain confidential material for the sole use of the intended recipient(s). Any review or distribution by others is strictly prohibited. If you are not the intended recipient, please contact the sender and delete all copies.

Metzger, Markus T

author

2008-12-22 15:02 UTC #2 lore raw

Quoted 9 lines

-----Original Message-----
From: stephane eranian [mailto:[email protected]]
Sent: Montag, 22. Dezember 2008 14:24
To: Metzger, Markus T
Quoted 2 lines · Markus Metzger
Require admin privileges for ptrace BTS extension.

Quoted 7 lines

Can you explain the motivations for this?
I thought the BTS extension was a per-process functionality. So how come I can debug and single step my process without admin privilege and I cannot capture its own branches. Does this have to do with user vs. kernel execution of the proces (BTS captures everything if I recall)?

Ingo requested this as a precaution.
We found a bug: the DS context pointer is not cleared on fork.
The bug is fixed in tip/master but the fix is considered too intrusive to be promoted that late. The restriction will be removed once the fix is in.

regards,
markus.

Quoted 43 lines

Thanks.
Quoted 8 lines
Reported-by Ingo Molnar <[email protected]>
Signed-off-by Markus Metzger <[email protected]>
---
Index: gits/arch/x86/kernel/ptrace.c
===================================================================
--- gits.orig/arch/x86/kernel/ptrace.c 2008-12-22
09:09:25.000000000 +0100
Quoted 1 line
+++ gits/arch/x86/kernel/ptrace.c 2008-12-22
11:03:01.000000000 +0100
Quoted 29 lines
arch/x86/kernel/ptrace.c +5 -1
@@ -21,6 +21,7 @@
 #include <linux/audit.h>
 #include <linux/seccomp.h>
 #include <linux/signal.h>
+#include <linux/capability.h>
 
 #include <asm/uaccess.h>
 #include <asm/pgtable.h>
@@ -742,6 +743,10 @@
       struct ptrace_bts_config cfg;
       int error = 0;
 
+       error = -EPERM;
+       if (!capable(CAP_SYS_ADMIN))
+               goto errout;
+
       error = -EOPNOTSUPP;
       if (!bts_cfg.sizeof_bts)
               goto errout;
---------------------------------------------------------------------
Intel GmbH
Dornacher Strasse 1
85622 Feldkirchen/Muenchen Germany
Sitz der Gesellschaft: Feldkirchen bei Muenchen
Geschaeftsfuehrer: Douglas Lusk, Peter Gleissner, Hannes Schwaderer
Registergericht: Muenchen HRB 47456 Ust.-IdNr.
VAT Registration No.: DE129385895
Citibank Frankfurt (BLZ 502 109 00) 600119052

Quoted 7 lines

Quoted 6 lines
This e-mail and any attachments may contain confidential material for the sole use of the intended recipient(s). Any review or distribution by others is strictly prohibited. If you are not the intended recipient, please contact the sender and delete all copies.