Re: [PATCH] drivers/char/hpet.c: integer constant is too large for "long" type

[Andrew Morton <akpm@linux-foundation.org>]

On Sat, 21 Jul 2007 14:19:44 +0300
"S.__a__lar Onur" <caglar@pardus.org.tr> wrote:

> Following trivial patch silents
> 
> drivers/char/hpet.c:76: warning: integer constant is too large for "long" type
> 
> warning (with gcc-3.4.6)
> 
> Signed-off-by: S.__a__lar Onur <caglar@pardus.org.tr>
>  
>  drivers/char/hpet.c |    2 +-
>  1 files changed, 1 insertions(+), 1 deletions(-)
> 
> diff --git a/drivers/char/hpet.c b/drivers/char/hpet.c
> index ba0e74a..23fadf5 100644
> --- a/drivers/char/hpet.c
> +++ b/drivers/char/hpet.c
> @@ -73,7 +73,7 @@ static struct clocksource clocksource_hpet = {
>          .name           = "hpet",
>          .rating         = 250,
>          .read           = read_hpet,
> -        .mask           = 0xffffffffffffffff,
> +        .mask           = 0xffffffffffffffffLL,
>          .mult           = 0, /*to be caluclated*/
>          .shift          = 10,
>          .flags          = CLOCK_SOURCE_IS_CONTINUOUS,

hm, well, this does assume that cycle_t is implemented as long long (or
unsigned long long, actually).  Which in turn assumes that u64 is
implemented as [unsigned] long long.

Not all architectures use unsigned long long for u64: some use unsigned
long.  They'll probably be OK with this change but I think I'll convert
that into plain old "-1", which just works everywhere.