Re: [fpc-devel] Sucesfull (mostly) experiment to build CHROMuLAN application with FPC main (3.3.1-git) and Lazarus GIT on RISC-V 64 target

2024-08-12 Thread Jeppe Johansen via fpc-devel
On 8/8/24 09:23, Pavel Pisa via fpc-devel wrote: The problem with keeping virtual registers instead of mapping to RV64 hardware ones in full Lazarus make bigide build seems to be exactly reproducible on real hardware same as in user space QEMU RISC-V Debian in chroot install. Assembling formed

Re: [fpc-devel] Help with inline assembler code

2023-03-29 Thread Jeppe Johansen via fpc-devel
Hi, A much more deterministic version would do it as this: function HiInteger(value: LongInt): LongInt; nostackframe; assembler; asm   mov r0, r0, asr #16 end; If you do it with an inline ASM block and without stackframes it becomes much harder to predict how you need to address variables by n

Re: [fpc-devel] Thoughts: Make FillChar etc. an intrinsic for specialised performance potential

2022-04-16 Thread Jeppe Johansen via fpc-devel
On 4/16/22 16:41, Sven Barth via fpc-devel wrote: Benito van der Zander via fpc-devel schrieb am Sa., 16. Apr. 2022, 15:43: Hi, it could always inline it. For small sizes do that mov and for large sizes do rep stosb on x86. It is very fast nowadays. Faster than FillChar on my