Re: Signedness of boolean types (and Ada)

2021-05-18 Thread Richard Biener via Gcc
On Tue, May 18, 2021 at 11:41 AM Eric Botcazou wrote: > > > I noticed while debugging why my "A?CST1:CST0" patch was broken for > > Ada, I noticed the following ranges for boolean types: > > # RANGE [0, 1] NONZERO 1 > > _14 = checks__saved_checks_tos.482_2 > 0; > > # RANGE [0, 255] NONZERO 1

Re: Signedness of boolean types (and Ada)

2021-05-18 Thread Eric Botcazou
> I noticed while debugging why my "A?CST1:CST0" patch was broken for > Ada, I noticed the following ranges for boolean types: > # RANGE [0, 1] NONZERO 1 > _14 = checks__saved_checks_tos.482_2 > 0; > # RANGE [0, 255] NONZERO 1 > _18 = _14 == 0; > _19 = ~_18; The '~' looks problematic if

Re: Signedness of boolean types (and Ada)

2021-05-18 Thread Richard Biener via Gcc
On Tue, May 18, 2021 at 4:51 AM Andrew Pinski via Gcc wrote: > > On Mon, May 17, 2021 at 6:52 PM Andrew Pinski wrote: > > > > I noticed while debugging why my "A?CST1:CST0" patch was broken for > > Ada, I noticed the following ranges for boolean types: > > # RANGE [0, 1] NONZERO 1 > > _14 = c

Re: Signedness of boolean types (and Ada)

2021-05-17 Thread Andrew Pinski via Gcc
On Mon, May 17, 2021 at 6:52 PM Andrew Pinski wrote: > > I noticed while debugging why my "A?CST1:CST0" patch was broken for > Ada, I noticed the following ranges for boolean types: > # RANGE [0, 1] NONZERO 1 > _14 = checks__saved_checks_tos.482_2 > 0; > # RANGE [0, 255] NONZERO 1 > _18 =

Signedness of boolean types (and Ada)

2021-05-17 Thread Andrew Pinski via Gcc
I noticed while debugging why my "A?CST1:CST0" patch was broken for Ada, I noticed the following ranges for boolean types: # RANGE [0, 1] NONZERO 1 _14 = checks__saved_checks_tos.482_2 > 0; # RANGE [0, 255] NONZERO 1 _18 = _14 == 0; _19 = ~_18; # RANGE [0, 1] NONZERO 1 _15 = _19; if