linux-kernel - Re: [PATCH] module interface improvement for kprobes

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-Id: <20060804090018.0bd4020d.rdunlap@xenotime.net>
Date:	Fri, 4 Aug 2006 09:00:18 -0700
From:	"Randy.Dunlap" <rdunlap@...otime.net>
To:	David Smith <dsmith@...hat.com>
Cc:	linux-kernel@...r.kernel.org, rusty@...tcorp.com.au,
	prasanna@...ibm.com, ananth@...ibm.com,
	anil.s.keshavamurthy@...el.com, davem@...emloft.net
Subject: Re: [PATCH] module interface improvement for kprobes

On Fri, 04 Aug 2006 10:17:32 -0500 David Smith wrote:

> When inserting a kprobes probe into a loadable module, there isn't a way
> for the kprobes module to get a module reference (in order to find the
> base address of the module among perhaps other things).
> 
> A kprobes probe needs the base address of the module in order to
> "relocate" the addresses of probe points (since the load address of the
> module can change from run to run of the kprobe).
> 
> I've added a new function, module_get_byname(), that looks up a module
> by name and returns the module reference.  Note that module_get_byname()
> also increments the module reference count.  It does this so that the
> module won't be unloaded between the time that module_get_byname() is
> called and register_kprobe() is called.
> 
> Here's an example of how it would be used from a kprobe:
> 
> ====
> static struct module *mod = NULL;
> 
> int init_module(void)
> {
> 	/* grab the module, making sure it won't get unloaded until
> 	 * we're done */
> 	const char *mod_name = "joydev";
> 	if (module_get_byname(mod_name, &mod) != 0)
> 		return 1;
> 
> 	/* Specify the address/offset where you want to insert
> 	 * probe.  If this were a real kprobe module, we'd "relocate"
> 	 * our probe address based on the load address of the module
> 	 * we're interested in. */
> 	kp.addr = (kprobe_opcode_t *) mod->module_core + 0;
> 
> 	/* All set to register with Kprobes */
>         register_kprobe(&kp);
> 	return 0;
> }
> 
> void cleanup_module(void)
> {
> 	if (kp.addr != NULL)
> 		unregister_kprobe(&kp);
> 
> 	/* allow the module to get unloaded, if needed */
> 	if (mod != NULL)
> 		module_put(mod);
> }
> ====
> 
> I've included a patch that implements the new function. 
> 
> Signed-off-by: David Smith <dsmith@...hat.com>
> ---
> 
> diff --git a/include/linux/module.h b/include/linux/module.h
> index 0dfb794..7464e29 100644
> --- a/include/linux/module.h
> +++ b/include/linux/module.h
> @@ -374,6 +374,8 @@ extern void __module_put_and_exit(struct
>  	__attribute__((noreturn));
>  #define module_put_and_exit(code) __module_put_and_exit(THIS_MODULE,
> code);
>  
> +long module_get_byname(const char *mod_name, struct module **mod);

Why long instead of int?

> +
>  #ifdef CONFIG_MODULE_UNLOAD
>  unsigned int module_refcount(struct module *mod);
>  void __symbol_put(const char *symbol);
> @@ -549,6 +551,11 @@ static inline int is_exported(const char
>  	return 0;
>  }
>  
> +static inline long module_get_byname(const char *mod_name, struct
> module **mod)
> +{
> +	return 1;
> +}
> +
>  static inline int register_module_notifier(struct notifier_block * nb)
>  {
>  	/* no events will happen anyway, so this can always succeed */
> diff --git a/kernel/module.c b/kernel/module.c
> index 2a19cd4..30449ce 100644
> --- a/kernel/module.c
> +++ b/kernel/module.c
> @@ -291,7 +291,27 @@ static struct module *find_module(const 
>  	}
>  	return NULL;
>  }
> +  
> +long module_get_byname(const char *mod_name, struct module **mod)
> +{
> +	*mod = NULL;
>  
> +	/* We must hold module_mutex before calling find_module(). */
> +	if (mutex_lock_interruptible(&module_mutex) != 0)
> +		return -EINTR;
> +
> +	*mod = find_module(mod_name);
> +	mutex_unlock(&module_mutex);
> +	if (*mod) {
> +		if (! strong_try_module_get(*mod)) {

no space following the '!'.

> +			*mod = NULL;
> +			return -EINVAL;
> +		}
> +	}
> +	return 0;
> +}
> +EXPORT_SYMBOL(module_get_byname);
> + 
>  #ifdef CONFIG_SMP
>  /* Number of blocks used and allocated. */
>  static unsigned int pcpu_num_used, pcpu_num_allocated;


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