On Tue, 16 Aug 2022, Juha Manninen via fpc-devel wrote:
On Thu, Aug 11, 2022 at 9:37 PM Martin Frb via fpc-devel <
fpc-devel@lists.freepascal.org> wrote:
fpc code is still "unexpected".
It takes into account the MinValue
type TMyEnum = (a=5, b=7);
GetEnumName(pt, 5) // a
GetEnumName(pt, 6
On 16/08/2022 09:37, Michael Van Canneyt via fpc-devel wrote:
On Tue, 16 Aug 2022, Juha Manninen via fpc-devel wrote:
And function GetEnumValue() is buggy.
It is not buggy. It does not support enumerateds with gaps, because
RTTI does not support enumerateds with gaps.
It never has.
...
On Tue, 16 Aug 2022, Martin Frb via fpc-devel wrote:
On 16/08/2022 09:37, Michael Van Canneyt via fpc-devel wrote:
On Tue, 16 Aug 2022, Juha Manninen via fpc-devel wrote:
And function GetEnumValue() is buggy.
It is not buggy. It does not support enumerateds with gaps, because
RTTI does n
Am 16.08.2022 um 09:37 schrieb Michael Van Canneyt via fpc-devel:
On Tue, 16 Aug 2022, Juha Manninen via fpc-devel wrote:
On Thu, Aug 11, 2022 at 9:37 PM Martin Frb via fpc-devel <
fpc-devel@lists.freepascal.org> wrote:
fpc code is still "unexpected".
It takes into account the MinValue
type T
On Tue, 16 Aug 2022, Ondrej Pokorny via fpc-devel wrote:
--- code end
Run it in Delphi and FPC and compare the outputs:
Delphi output:
PropCount: 0
Property info is null
FPC output:
PropCount: 1
Property name: Fruit
Property kind: tkEnumeration
As you can see FPC creates a valid PropInfo
Am 16.08.2022 um 13:19 schrieb Michael Van Canneyt via fpc-devel:
On Tue, 16 Aug 2022, Ondrej Pokorny via fpc-devel wrote:
--- code end
Run it in Delphi and FPC and compare the outputs:
Delphi output:
PropCount: 0
Property info is null
FPC output:
PropCount: 1
Property name: Fruit
Property ki
On Tue, 16 Aug 2022, Ondrej Pokorny via fpc-devel wrote:
Am 16.08.2022 um 13:19 schrieb Michael Van Canneyt via fpc-devel:
On Tue, 16 Aug 2022, Ondrej Pokorny via fpc-devel wrote:
--- code end
Run it in Delphi and FPC and compare the outputs:
Delphi output:
PropCount: 0
Property info is nu
Hi everyone,
Recently I've made some optimisations centred around the SHR instruction
on x86, and there was one pair of instructions that caught my attention:
movl (%rbx),%eax
shrl $24,%eax
Is it permissible to optimise this to (x86 is little-endian):
movzbl 3(%rbx),%eax?
(You could also op
