On Tue, 2025-10-14 at 07:51 +0200, Thomas Weißschuh wrote:
> Reactors can be called from any context through tracepoints.
> When developing reactors care needs to be taken to only call APIs which
> are safe. As the tracepoints used during testing may not actually be
> called from restrictive contexts lockdep may not be helpful.
> 
> Add explicit overrides to help lockdep find invalid code patterns.
> 
> The usage of LD_WAIT_FREE will trigger lockdep warnings in the panic
> reactor. These are indeed valid warnings but they are out of scope for
> RV and will instead be fixed by the printk subsystem.

Looks like a nice addition!
If I get it correctly, this patch does trigger a lockdep warning with the
current state of the kernel. Is there a plan of fixing the warning in printk?
I assume this series would need to wait for that or did you have other ideas?

Thanks,
Gabriele

> 
> Signed-off-by: Thomas Weißschuh <[email protected]>
> ---
>  kernel/trace/rv/rv_reactors.c | 4 ++++
>  1 file changed, 4 insertions(+)
> 
> diff --git a/kernel/trace/rv/rv_reactors.c b/kernel/trace/rv/rv_reactors.c
> index
> 8c02426bc3bd944265f809e431283d1a20d56a8c..d9d335ae9badaa320f1d35dd159a033c3a30
> eb1a 100644
> --- a/kernel/trace/rv/rv_reactors.c
> +++ b/kernel/trace/rv/rv_reactors.c
> @@ -61,6 +61,7 @@
>   *      printk
>   */
>  
> +#include <linux/lockdep.h>
>  #include <linux/slab.h>
>  
>  #include "rv.h"
> @@ -480,6 +481,7 @@ int init_rv_reactors(struct dentry *root_dir)
>  
>  void rv_react(struct rv_monitor *monitor, const char *msg, ...)
>  {
> +     static DEFINE_WAIT_OVERRIDE_MAP(rv_react_map, LD_WAIT_FREE);
>       va_list args;
>  
>       if (!rv_reacting_on() || !monitor->react)
> @@ -487,7 +489,9 @@ void rv_react(struct rv_monitor *monitor, const char *msg,
> ...)
>  
>       va_start(args, msg);
>  
> +     lock_map_acquire_try(&rv_react_map);
>       monitor->react(msg, args);
> +     lock_map_release(&rv_react_map);
>  
>       va_end(args);
>  }


Reply via email to