Messages in this thread | ![/](/images/icornerl.gif) | | Date | Wed, 9 Oct 2019 10:16:02 +0200 | From | Peter Zijlstra <> | Subject | Re: [PATCH 3/3] KVM: x86/vPMU: Add lazy mechanism to release perf_event per vPMC |
| |
On Wed, Oct 09, 2019 at 09:15:03AM +0200, Paolo Bonzini wrote: > For stuff like hardware registers, bitfields are probably a bad idea > anyway, so let's only consider the case of space optimization.
Except for hardware registers? I actually like bitfields to describe hardware registers.
> bool:2 would definitely cause an eyebrow raise, but I don't see why > bool:1 bitfields are a problem. An integer type large enough to store > the values 0 and 1 can be of any size bigger than one bit.
Consider:
bool foo:1; bool bar:1;
Will bar use the second bit of _Bool? Does it have one? (yes it does, but it's still weird).
But worse, as used in the parent thread:
u8 count:7; bool flag:1;
Who says the @flag thing will even be the msb of the initial u8 and not a whole new variable due to change in base type?
> bool bitfields preserve the magic behavior where something like this: > > foo->x = y; > > (x is a bool bitfield) would be compiled as > > foo->x = (y != 0);
This is confusion; if y is a single bit bitfield, then there is absolutely _NO_ difference between these two expressions.
The _only_ thing about _Bool is that it magically casts values to 0,1. Single bit bitfield variables have no choice but to already be in that range.
So expressions where it matters are:
x = (7&2) // x == 2 vs x = !!(7&2) // x == 1
But it is impossible for int:1 and _Bool to behave differently.
> However, in this patch bitfields are unnecessary and they result in > worse code from the compiler.
Fully agreed :-)
| ![\](/images/icornerr.gif) |