On 6/24/19 10:03 AM, Tim Düsterhus wrote:
Willy,
Am 24.06.19 um 07:33 schrieb Willy Tarreau:
Hi Tim,
[snipped]
If it emits an error when using sizeof(int) instead of sizeof(*foo) where
foo is an pointer to unsigned int, it is bogus. I challenge you to present
me a relevant architecture where unsigned changes the allocated size!r
For C11 it is guaranteed that there is none according to this:
https://stackoverflow.com/a/13169473/782822
This is guaranteed since the beginning of C. Not only for C11. :)
Fred.
