Re: [PATCH 5/7] Ksplice: Add functions for walking kallsymssymbols

[Anders Kaseorg]

On Wed, 4 Feb 2009, Rusty Russell wrote:
> Seems reasonable.  Did you really not want a name-filtering version?  That
> might be generally useful.

We donât actually want a name-filtering version for Ksplice.  We process 
all the symbols we need in one pass, in order to avoid traversing the 
kallsyms list separately for each symbol.

> But does no locking at all.  It either needs a comment that it can only be
> called from inside stop_machine, or that it needs preempt disabled, or
> whatever.

It might as well use preempt_disable() and RCU itself (patch below).

> Is the !CONFIG_KALLSYMS version useful to you?  If not, please don't
> implement the noop version.  I want someone who *does* expect it to work to
> have to think about it if they use it...

No, we donât have a particular use for it; we added it for consistency 
with kallsyms_lookup_name(), etc.

Anders

diff --git a/kernel/module.c b/kernel/module.c
index 556112c..6bb5814 100644
--- a/kernel/module.c
+++ b/kernel/module.c
@@ -2577,17 +2577,20 @@ int module_kallsyms_on_each_symbol(int (*fn)(void *, const char *,
 {
 	struct module *mod;
 	unsigned int i;
-	int ret;
+	int ret = 0;

-	list_for_each_entry(mod, &modules, list) {
+	preempt_disable();
+	list_for_each_entry_rcu(mod, &modules, list) {
 		for (i = 0; i < mod->num_symtab; i++) {
 			ret = fn(data, mod->strtab + mod->symtab[i].st_name,
 				 mod, mod->symtab[i].st_value);
 			if (ret != 0)
-				return ret;
+				goto out;
 		}
 	}
-	return 0;
+out:
+	preempt_enable();
+	return ret;
 }
 #endif /* CONFIG_KALLSYMS */

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/