Re: [PATCH 2/2] uprobes/core: counter to optimize probe hits.

[Ingo Molnar <mingo@kernel.org>]

* Srikar Dronamraju <srikar@linux.vnet.ibm.com> wrote:

> diff --git a/include/linux/mm_types.h b/include/linux/mm_types.h
> index 9ade86e..62d5aeb 100644
> --- a/include/linux/mm_types.h
> +++ b/include/linux/mm_types.h
> @@ -390,6 +390,7 @@ struct mm_struct {
>  	struct cpumask cpumask_allocation;
>  #endif
>  #ifdef CONFIG_UPROBES
> +	atomic_t mm_uprobes_count;
>  	struct uprobes_xol_area *uprobes_xol_area;
>  #endif

Since mm_types.h includes uprobes.h already it's much better to 
stick this into a 'struct uprobes_state' and thus keep the main 
'struct mm_struct' definition as simple as possible.

Also, your patch titles suck:

  - no proper capitalization like you can observe with previous 
    uprobes commits

  - extra period at the end

  - missing verb from the sentence. Check existing uprobes 
    commits to see the kind of sentences that commit titles are 
    expected to be.

> +	if (!atomic_read(&uprobe_events) || !valid_vma(vma, false))
> +		return;		/* Bail-out */
> +
> +	if (!atomic_read(&vma->vm_mm->mm_uprobes_count))
> +		return;
> +
> +	inode = vma->vm_file->f_mapping->host;
> +	if (!inode)
> +		return;

The 'Bail-out' comment tacked on to one of the returns seems 
entirely superfluous.

> +		if (vaddr >= vma->vm_start && vaddr < vma->vm_end) {
> +
> +			/*

That newline looks superfluous too.

> +	return;
> +}

Please read your own patches more carefully ... this should 
stick out like a sore thumb.

Some of the above comments apply to your other patch as well.

The structure and granularity of the patches looks good 
otherwise.

Thanks,

	Ingo