Compiler capability defines and more

[Fr0sT.Brutal 904]

Include files for checking compiler versions and capabilities. Useful for developers of libs/components that have to support multiple compilers.

 

https://github.com/Fr0sT-Brutal/Delphi_Compilers

 

Additionally, there are files for checking current compiler settings, some cross-compiler and cross-platform defines and cross-compiler declarations.

[Stefan Glienke 2148]

The general hints in the readme are very valuable (and made me revisit some of my ifdef checks) - apart from that it reminds me a bit of https://xkcd.com/927/

 

Yes, jedi.inc is long and unwieldy at times but it got everything commonly needed - in case it does not - wouldn't it be wiser to add that instead of rolling a new thing?

Apart from that the prefix CAPS is kinda misleading imo - why not go with the HAS_ or SUPPORTS_ prefixes?

 

Oh a note on your comment for bitness checking - the example does not entirely match what you describe, it specifically checks for x64 - so x86 and arm would go in the else.

Bitness checks would be CPU32BITS and CPU64BITS or whats also common to check SizeOf(Pointer) = 4 for 32bit.

Historically the checks are rather for x86 specifically to implement something using asm and then pure pascal in else.

Edited August 10, 2020 by Stefan Glienke

[Bill Meyer 339]

17 minutes ago, Stefan Glienke said:

The general hints in the readme are very valuable (and made me revisit some of my ifdef checks) - apart from that it reminds me a bit of https://xkcd.com/927/

As a former colleague was fond of saying, the nice thing about standards is that we have so many from which to choose.

[Fr0sT.Brutal 904]

15 hours ago, Stefan Glienke said:

The general hints in the readme are very valuable (and made me revisit some of my ifdef checks) 

I'm glad to help 🙂

15 hours ago, Stefan Glienke said:

Yes, jedi.inc is long and unwieldy at times but it got everything commonly needed - in case it does not - wouldn't it be wiser to add that instead of rolling a new thing?

Well, I've never used JEDI as is right because of its monsterosity. Moreover their basic define system has the critical weakness - it must be updated for every new compiler version; and it's DAMN long while I wanted short and universal solution.

15 hours ago, Stefan Glienke said:

Apart from that the prefix CAPS is kinda misleading imo - why not go with the HAS_ or SUPPORTS_ prefixes?

Caps are compiler capabilities... I don't remember why I called them so. SUPPORTS is longer but acceptable as well, agreed.

15 hours ago, Stefan Glienke said:

Oh a note on your comment for bitness checking - the example does not entirely match what you describe, it specifically checks for x64 - so x86 and arm would go in the else.

Bitness checks would be CPU32BITS and CPU64BITS or whats also common to check SizeOf(Pointer) = 4 for 32bit.

Historically the checks are rather for x86 specifically to implement something using asm and then pure pascal in else.

You're right, that was previous, quite weak, sample. I already changed it to check for 64 and 32 bitness explicitly. Oh and there's another bug indeed - I used "X" letter in defines which means Intel arch. In fact, it was meant to be CPU32/64 (FPC-style) which are defined for Delphi in XPlatformCompat.inc file.

[Stefan Glienke 2148]

8 minutes ago, Fr0sT.Brutal said:

critical weakness - it must be updated for every new compiler version

No, it's upwards compatible - an unknown version identifies as latest known one including all feature defines

[Fr0sT.Brutal 904]

7 minutes ago, Stefan Glienke said:

No, it's upwards compatible - an unknown version identifies as latest known one including all feature defines

Likely it is. I looked at that file and got lost in these loooong nested defines so won't argue