[PATCH] module support: record in vermagic ability to unload a module
authorPaolo 'Blaisorblade' Giarrusso <blaisorblade@yahoo.it>
Tue, 11 Apr 2006 05:54:28 +0000 (22:54 -0700)
committerLinus Torvalds <torvalds@g5.osdl.org>
Tue, 11 Apr 2006 13:18:45 +0000 (06:18 -0700)
An UML user reported (against 2.6.13.3/UML) he got kernel Oopses when
trying to rmmod (on a kernel with module unloading enabled) a module
compiled with module unloading disabled.  As crashing is a very correct
thing to do in that case, a solution is altering the vermagic string to
include this too.

Possibly, however, the code should not crash in this case, even if the
module didn't support unloading - it should simply abort the module
removal.  In this case, fixing that bug would be a better solution.  I've
not investigated though.

(akpm: a bit marginal - root screwed up and shot himself in the foot).

Cc: Hayim Shaul <hayim@post.tau.ac.il>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Signed-off-by: Paolo 'Blaisorblade' Giarrusso <blaisorblade@yahoo.it>
Signed-off-by: Andrew Morton <akpm@osdl.org>
Signed-off-by: Linus Torvalds <torvalds@osdl.org>
include/linux/vermagic.h

index fadc535..dc7c621 100644 (file)
 #else
 #define MODULE_VERMAGIC_PREEMPT ""
 #endif
+#ifdef CONFIG_MODULE_UNLOAD
+#define MODULE_VERMAGIC_MODULE_UNLOAD "mod_unload "
+#else
+#define MODULE_VERMAGIC_MODULE_UNLOAD ""
+#endif
 #ifndef MODULE_ARCH_VERMAGIC
 #define MODULE_ARCH_VERMAGIC ""
 #endif
@@ -19,5 +24,5 @@
 #define VERMAGIC_STRING                                                \
        UTS_RELEASE " "                                                 \
        MODULE_VERMAGIC_SMP MODULE_VERMAGIC_PREEMPT                     \
-       MODULE_ARCH_VERMAGIC                                            \
+       MODULE_VERMAGIC_MODULE_UNLOAD MODULE_ARCH_VERMAGIC              \
        "gcc-" __stringify(__GNUC__) "." __stringify(__GNUC_MINOR__)