On Wed, Jun 27, 2018 at 11:36:03AM +0200, Rasmus Villemoes wrote: > On 26 June 2018 at 19:54, Jason Gunthorpe <jgg@xxxxxxxxxxxx> wrote: > > > On Tue, Jun 26, 2018 at 10:07:07AM +0200, Rasmus Villemoes wrote: > > > On 25 June 2018 at 19:11, Jason Gunthorpe <[1]jgg@xxxxxxxxxxxx> > > wrote: > > > > > > > When thinking about signed cases.. The explicit u64 cast, and > > implict promotion to typeof(d), produce something counter intuitive, > > eg: > > > > (u64)(s32)-1 == 0xffffffffffffffff > > > > Which would result in a shift oucome that is not what anyone would > > expect, IMHO... So the first version isn't what I'd expect either.. > > > > Wouldn't check_shift_overflow(-1, 4, &someint) just put -16 in someint and > report no overflow? That's what I'd expect, if negative values are to be > supported at all. Most if not all the times we don't do shifts on negative values, so I don't think that we should support them. Thanks
Attachment:
signature.asc
Description: PGP signature
