Should I add a sysctl option to print the module list upon every oops, to
catch the cases where the oops is fatal?
This is a fairly simple patch which shouldn't have adverse effects, so it'd be
nice if it went into 2.2. Especially as I was provoked into writing it by
having to debug problems in the pre-2.2 kernels.
Index: drivers/char/sysrq.c
===================================================================
RCS file: /cvs/linux/drivers/char/sysrq.c,v
retrieving revision 1.1.1.5
diff -u -w -u -r1.1.1.5 sysrq.c
--- drivers/char/sysrq.c 1998/10/29 00:07:14 1.1.1.5
+++ drivers/char/sysrq.c 1999/01/07 11:20:47
@@ -70,6 +70,12 @@
printk("Keyboard mode set to XLATE\n");
}
break;
+#ifdef CONFIG_MODULES
+ case 'd': /* D -- moDule listing */
+ printk("Module list:\n");
+ list_modules();
+ break;
+#endif
#ifdef CONFIG_VT
case 'k': /* K -- SAK */
printk("SAK\n");
@@ -140,6 +146,9 @@
printk("Boot "
#ifdef CONFIG_APM
"Off "
+#endif
+#ifdef CONFIG_MODULES
+ "moDulelist "
#endif
"Sync Unmount showPc showTasks showMem loglevel0-8 tErm kIll killalL\n");
/* Don't use 'A' as it's handled specially on the Sparc */
Index: kernel/module.c
===================================================================
RCS file: /cvs/linux/kernel/module.c,v
retrieving revision 1.1.1.5
diff -u -w -r1.1.1.5 module.c
--- kernel/module.c 1999/01/04 10:21:05 1.1.1.5
+++ kernel/module.c 1999/01/07 11:24:52
@@ -971,6 +971,25 @@
return 0;
}
+/*
+ * List the load address of all the currently loaded modules, for debugging
+ * purposes. Perhaps the oops routines should all call this automatically?
+ */
+
+void list_modules()
+{
+ struct module *mlist = module_list;
+
+ printk(" Name Address\n");
+
+ while (mlist && mlist->next) /* Don't do the last one */
+ {
+ printk ("%-16s %p\n", mlist->name,
+ &((char *)mlist)[mlist->size_of_struct] );
+ mlist = mlist->next;
+ }
+}
+
#else /* CONFIG_MODULES */
/* Dummy syscalls for people who don't want modules */
Index: Documentation/sysrq.txt
===================================================================
RCS file: /cvs/linux/Documentation/sysrq.txt,v
retrieving revision 1.1.1.2
diff -u -w -r1.1.1.2 sysrq.txt
--- Documentation/sysrq.txt 1998/10/28 21:44:54 1.1.1.2
+++ Documentation/sysrq.txt 1999/01/07 11:30:40
@@ -58,6 +58,9 @@
'l' - Send a SIGKILL to all processes, INCLUDING init. (Your system
will be non-functional after this.)
+'d' - List all the currently-loaded modules, along with the address at
+ which they are loaded.
+
* Okay, so what can I use them for?
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Well, un'R'aw is very handy when your X server or a svgalib program crashes.
@@ -87,6 +90,10 @@
t'E'rm and k'I'll are useful if you have some sort of runaway process you
are unable to kill any other way, especially if it's spawning other
processes.
+
+The mo'D'ule listing function is useful if your kernel has just oopsed in a
+module - you need to know the address at which each module is loaded in order
+to have any more than a whelk's chance in a supernova of tracing the problem.
* Sometimes SysRQ seems to get 'stuck' after using it, what can I do?
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
---- ---- ----
David Woodhouse David.Woodhouse@mvhi.com Office: (+44) 1223 810302
Project Leader, Process Information Systems Mobile: (+44) 976 658355
Axiom (Cambridge) Ltd., Swaffham Bulbeck, Cambridge, CB5 0NA, UK.
finger dwmw2@ferret.lmh.ox.ac.uk for PGP key.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/