Re: Another kvm user can go away?

[Paul Goyette <paul%whooppee.com@localhost>]

On Sat, 31 Dec 2016, Simon Burge wrote:

> Paul Goyette wrote:
>
> > Any suggestions on how to ensure proper alignment?  There doesn't seem
> > to be any way to kmem_allocate() what alignment is needed...  :)  It
> > seems to currently always provide 64-bit (8-byte) alignment, but the man
> > page makes no guarantees.
> >
> > I guess I could allocate an extra 8 bytes, and then manipulate the
> > returned address... But that just "feels wrong".  :)
>
> I meant alignment within the structures.  Maybe "packing" might be
> a better word.  This helps for example with mixed 32-bit and 64-bit
> emulations.  If you look at your original struct sysctl_history_event:
>
> 	struct sysctl_history_event {
> 		struct timeval	she_tv;
> 		uint64_t	she_callnumber;
> 		uint64_t	she_values[4];
>
> The timeval is a structure with a 64-bit int for seconds and and
> a 32-bit int for microseconds.  With the following uint64_t some
> architectures might pad out she_callnumber in the structure so that it's
> on a 64-bit boundary while others will by happy with a 32-bit alignment
> and not use any padding.
>
> In short, just make sure everything in the struct packs nicely. :)

Yeah, I figured that out after I replied (with the help of Christos's
suggestion.  My latest revision uses a 'struct timespec' which has two 
64-bit members, so preserves alignment of subsequent members!


+------------------+--------------------------+------------------------+
| Paul Goyette     | PGP Key fingerprint:     | E-mail addresses:      |
| (Retired)        | FA29 0E3B 35AF E8AE 6651 | paul at whooppee.com   |
| Kernel Developer | 0786 F758 55DE 53BA 7731 | pgoyette at netbsd.org |
+------------------+--------------------------+------------------------+