Skip to contentSkip to conversation

2.6.24-rc3-mm2 - add-64-bit-capability-support-to-the-kernel

DISCUSSION [email protected] 3 participants 3 messages Dec 2, 2007 – Dec 6, 2007 Capabilities

lore
Share & tools
lore
#0 lore raw

Question:

The patch does the semantic equivalent of:

-#define cap_clear(c) do { cap_t(c) = 0; } while(0)
-#define cap_set_full(c) do { cap_t(c) = ~0; } while(0)

+# define cap_clear(c)         do { (c) = __cap_empty_set; } while (0)
+# define cap_set_full(c)      do { (c) = __cap_full_set; } while (0)
+# define cap_set_init_eff(c)  do { (c) = __cap_init_eff_set; } while (0)

Was it intentional, or an oversight, that this blows chunks in modules that try to use cap_clear() or cap_set_full() because the __cap_* symbols don't get an EXPORT_SYMBOL() attached to them?

#1 lore raw
Quoted 11 lines · Quoting [email protected] ([email protected]):

Question:

The patch does the semantic equivalent of:

(quoted diff) +3 -2
-#define cap_clear(c)         do { cap_t(c) =  0; } while(0)
-#define cap_set_full(c)      do { cap_t(c) = ~0; } while(0)
 
+# define cap_clear(c)         do { (c) = __cap_empty_set; } while (0)
+# define cap_set_full(c)      do { (c) = __cap_full_set; } while (0)
+# define cap_set_init_eff(c)  do { (c) = __cap_init_eff_set; } while (0)
Quoted 3 lines

Was it intentional, or an oversight, that this blows chunks in modules that try to use cap_clear() or cap_set_full() because the __cap_* symbols don't get an EXPORT_SYMBOL() attached to them?

I think that's one big oops, and the following is needed.

Quoted 1 line

From 9c217a6280b559cca28edab621d2d22288ce0ac6 Mon Sep 17 00:00:00 2001

From: Serge E. Hallyn <[email protected]>
Date: Mon, 3 Dec 2007 09:59:50 -0500
Subject: [PATCH 1/1] capabilities: export __cap_ symbols

Export __cap_empty_set and friends so that the cap_clear(), cap_set_full() and cap_set_init_eff() defines can be used by modules.

---
 kernel/capability.c |    4 ++++
 1 files changed, 4 insertions(+), 0 deletions(-)
kernel/capability.c +4 -0
index 9f2db55..064eb8a 100644
@@ -29,6 +29,10 @@ const kernel_cap_t __cap_empty_set = CAP_EMPTY_SET;
 const kernel_cap_t __cap_full_set = CAP_FULL_SET;
 const kernel_cap_t __cap_init_eff_set = CAP_INIT_EFF_SET;
 
+EXPORT_SYMBOL(__cap_empty_set);
+EXPORT_SYMBOL(__cap_full_set);
+EXPORT_SYMBOL(__cap_init_eff_set);
+
 /*
  * For sys_getproccap() and sys_setproccap(), any of the three
  * capability set pointers may be NULL -- indicating that that set is
Signature · 1 line
1.5.1.1.GIT
#2 lore raw

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Quoted 8 lines · Serge E. Hallyn
Quoted 7 lines · Quoting [email protected] ([email protected]):

Question:

The patch does the semantic equivalent of:

(quoted diff) +0 -2
-#define cap_clear(c)         do { cap_t(c) =  0; } while(0)
-#define cap_set_full(c)      do { cap_t(c) = ~0; } while(0)
Quoted 7 lines
Quoted 7 lines
(quoted diff) +3 -0
+# define cap_clear(c)         do { (c) = __cap_empty_set; } while (0)
+# define cap_set_full(c)      do { (c) = __cap_full_set; } while (0)
+# define cap_set_init_eff(c)  do { (c) = __cap_init_eff_set; } while (0)

Was it intentional, or an oversight, that this blows chunks in modules that try to use cap_clear() or cap_set_full() because the __cap_* symbols don't get an EXPORT_SYMBOL() attached to them?

Definitely an oversight. Thanks Serge for the fix!

Cheers

Andrew
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.7 (Darwin)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iD8DBQFHV1SP+bHCR3gb8jsRAtikAJ9FEm6ETGa9qrr82kLI4ZtsACZ43gCgqIcY GbuHBsxpaZtQDEgU/3OCgA0=
=A9+O
-----END PGP SIGNATURE-----