Re: [PATCH v5 03/78] xarray: Add the xa_lock to the radix_tree_root

[Kirill A. Shutemov]

On Tue, Dec 26, 2017 at 07:58:15PM -0800, Matthew Wilcox wrote:
> On Tue, Dec 26, 2017 at 07:43:40PM -0800, Matthew Wilcox wrote:
> >     Also add the xa_lock() and xa_unlock() family of wrappers to make it
> >     easier to use the lock.  If we could rely on -fplan9-extensions in
> >     the compiler, we could avoid all of this syntactic sugar, but that
> >     wasn't added until gcc 4.6.
> 
> Oh, in case anyone's wondering, here's how I'd do it with plan9 extensions:
> 
> struct xarray {
>         spinlock_t;
>         int xa_flags;
>         void *xa_head;
> };
> 
> ...
>         spin_lock_irqsave(&mapping->pages, flags);
>         __delete_from_page_cache(page, NULL);
>         spin_unlock_irqrestore(&mapping->pages, flags);
> ...
> 
> The plan9 extensions permit passing a pointer to a struct which has an
> unnamed element to a function which is expecting a pointer to the type
> of that element.  The compiler does any necessary arithmetic to produce 
> a pointer.  It's exactly as if I had written:
> 
>         spin_lock_irqsave(&mapping->pages.xa_lock, flags);
>         __delete_from_page_cache(page, NULL);
>         spin_unlock_irqrestore(&mapping->pages.xa_lock, flags);
> 
> More details here: https://9p.io/sys/doc/compiler.html

Yeah, that's neat.

Dealing with old compilers is frustrating...

-- 
 Kirill A. Shutemov