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: <feae7a189afa69091816074dcf66a496b11bdd38.camel@redhat.com>
Date: Mon, 21 Jul 2025 12:29:24 +0200
From: Gabriele Monaco <gmonaco@...hat.com>
To: Nam Cao <namcao@...utronix.de>
Cc: Steven Rostedt <rostedt@...dmis.org>, Masami Hiramatsu
 <mhiramat@...nel.org>,  Mathieu Desnoyers <mathieu.desnoyers@...icios.com>,
 linux-trace-kernel@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 5/6] rv: Remove the nop reactor

On Mon, 2025-07-21 at 11:47 +0200, Nam Cao wrote:
> As suggested by the name, the nop reactor does not do anything. It is
> the
> default reactor when nothing else is selected.
> 
> However, the monitors already null-check the reactor function
> pointers.
> Thus, instead of a nop reactor, just set the react function pointer
> to
> NULL. The nop reactor can then be removed.
> 
> Signed-off-by: Nam Cao <namcao@...utronix.de>

Thanks for the patch, I'd need to go through this a bit more in detail.

As far as I remember, the only way to disable reaction is to set it to
the nop reactor.
With your patch the behaviour changes and, to disable the reactor, you
now need to write an empty string, this should be documented somewhere,
at the very least. Perhaps userspace tools (tools/verification/rv)
might break and would need adaptation.

We could still remove the kernel side implementation, but from
userspace (tracefs) we might want to keep the nop reactor available,
setting it would set the reactor to NULL under the hood.

If you really want to change also the user space interface, we might
want to imitate other tracefs features and use something like printk /
!printk to enable/disable a reactor.

What do you think? Did I miss anything here?

Thanks,
Gabriele

> ---
>  kernel/trace/rv/rv_reactors.c | 63 ++++++---------------------------
> --
>  1 file changed, 11 insertions(+), 52 deletions(-)
> 
> diff --git a/kernel/trace/rv/rv_reactors.c
> b/kernel/trace/rv/rv_reactors.c
> index a8e849e6cd85..aee622e4b833 100644
> --- a/kernel/trace/rv/rv_reactors.c
> +++ b/kernel/trace/rv/rv_reactors.c
> @@ -70,17 +70,6 @@
>   */
>  static LIST_HEAD(rv_reactors_list);
>  
> -static struct rv_reactor *get_reactor_rdef_by_name(char *name)
> -{
> -	struct rv_reactor *r;
> -
> -	list_for_each_entry(r, &rv_reactors_list, list) {
> -		if (strcmp(name, r->name) == 0)
> -			return r;
> -	}
> -	return NULL;
> -}
> -
>  /*
>   * Available reactors seq functions.
>   */
> @@ -174,7 +163,7 @@ static void monitor_swap_reactors_single(struct
> rv_monitor *mon,
>  
>  	mon->reactor = reactor;
>  	mon->reacting = reacting;
> -	mon->react = reactor->react;
> +	mon->react = reactor ? reactor->react : NULL;
>  
>  	/* enable only once if iterating through a container */
>  	if (monitor_enabled && !nested)
> @@ -210,10 +199,15 @@ monitor_reactors_write(struct file *file, const
> char __user *user_buf,
>  	struct rv_reactor *reactor;
>  	struct seq_file *seq_f;
>  	int retval = -EINVAL;
> -	bool enable;
>  	char *ptr;
>  	int len;
>  
> +	/*
> +	 * See monitor_reactors_open()
> +	 */
> +	seq_f = file->private_data;
> +	mon = seq_f->private;
> +
>  	if (count < 1 || count > MAX_RV_REACTOR_NAME_SIZE + 1)
>  		return -EINVAL;
>  
> @@ -226,14 +220,10 @@ monitor_reactors_write(struct file *file, const
> char __user *user_buf,
>  	ptr = strim(buff);
>  
>  	len = strlen(ptr);
> -	if (!len)
> +	if (!len) {
> +		monitor_swap_reactors(mon, NULL, false);
>  		return count;
> -
> -	/*
> -	 * See monitor_reactors_open()
> -	 */
> -	seq_f = file->private_data;
> -	mon = seq_f->private;
> +	}
>  
>  	mutex_lock(&rv_interface_lock);
>  
> @@ -243,12 +233,7 @@ monitor_reactors_write(struct file *file, const
> char __user *user_buf,
>  		if (strcmp(ptr, reactor->name) != 0)
>  			continue;
>  
> -		if (strcmp(reactor->name, "nop"))
> -			enable = false;
> -		else
> -			enable = true;
> -
> -		monitor_swap_reactors(mon, reactor, enable);
> +		monitor_swap_reactors(mon, reactor, true);
>  
>  		retval = count;
>  		break;
> @@ -435,32 +420,12 @@ int reactor_populate_monitor(struct rv_monitor
> *mon)
>  	if (!tmp)
>  		return -ENOMEM;
>  
> -	/*
> -	 * Configure as the rv_nop reactor.
> -	 */
> -	mon->reactor = get_reactor_rdef_by_name("nop");
> -	mon->reacting = false;
> -
>  	return 0;
>  }
>  
> -/*
> - * Nop reactor register
> - */
> -__printf(1, 2) static void rv_nop_reaction(const char *msg, ...)
> -{
> -}
> -
> -static struct rv_reactor rv_nop = {
> -	.name = "nop",
> -	.description = "no-operation reactor: do nothing.",
> -	.react = rv_nop_reaction
> -};
> -
>  int init_rv_reactors(struct dentry *root_dir)
>  {
>  	struct dentry *available, *reacting;
> -	int retval;
>  
>  	available = rv_create_file("available_reactors",
> RV_MODE_READ, root_dir, NULL,
>  				   &available_reactors_ops);
> @@ -471,16 +436,10 @@ int init_rv_reactors(struct dentry *root_dir)
>  	if (!reacting)
>  		goto rm_available;
>  
> -	retval = __rv_register_reactor(&rv_nop);
> -	if (retval)
> -		goto rm_reacting;
> -
>  	turn_reacting_on();
>  
>  	return 0;
>  
> -rm_reacting:
> -	rv_remove(reacting);
>  rm_available:
>  	rv_remove(available);
>  out_err:


Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ