Re: [Qemu-devel] [PATCH] docs/atomics.txt: Update pointer to linux macro

[Marc-André Lureau]

Hi

On Mon, Apr 11, 2016 at 7:30 PM, Pranith Kumar <address@hidden> wrote:
> Add a missing end brace and update doc to point to the latest access
> macro. ACCESS_ONE() is deprecated.

ONE/ONCE

>
> Signed-off-by: Pranith Kumar <address@hidden>
> ---
>  docs/atomics.txt | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
>
> diff --git a/docs/atomics.txt b/docs/atomics.txt
> index ef285e3..bba771e 100644
> --- a/docs/atomics.txt
> +++ b/docs/atomics.txt
> @@ -62,7 +62,7 @@ operations:
>      typeof(*ptr) atomic_fetch_sub(ptr, val)
>      typeof(*ptr) atomic_fetch_and(ptr, val)
>      typeof(*ptr) atomic_fetch_or(ptr, val)
> -    typeof(*ptr) atomic_xchg(ptr, val
> +    typeof(*ptr) atomic_xchg(ptr, val)

I was going to send the same fix ;)

>      typeof(*ptr) atomic_cmpxchg(ptr, old, new)
>
>  all of which return the old value of *ptr.  These operations are
> @@ -328,7 +328,7 @@ and memory barriers, and the equivalents in QEMU:
>
>  - atomic_read and atomic_set in Linux give no guarantee at all;
>    atomic_read and atomic_set in QEMU include a compiler barrier
> -  (similar to the ACCESS_ONCE macro in Linux).
> +  (similar to the READ_ONCE/WRITE_ONCE macros in Linux).
>

That looks accurate to me, but I am not very familiar with these

>  - most atomic read-modify-write operations in Linux return void;

why removing that line?

>    in QEMU, all of them return the old value of the variable.
> --
> 2.8.1
>
>





-- 
Marc-André Lureau