#GCC is rejecting all code generated by #AI #neural-networks, or anyway LLMs.
on 02026-07-30#GCC discussion on how the #C preprocessor should work, quoting Dave Prosser #toread
on 02025-11-07vectorizing RGB to grayscale conversion code to #SIMD in #GCC
on 02025-11-05#GCC vector extensions for portable #SIMD code in C.
on 02025-10-23incompatibilities in #C and #C++ when porting to #GCC 15, which updates the default version of C to C23, which requires prototypes.
on 02025-10-15#GCC lets you associate #asm registers with particular #C variables, either global or local.
on 02024-09-23apparently #GCC 14.1 includes a #Rust frontend from the #gccrs project
on 02024-06-23#GCC for i386 has function attributes including fastcall, no_caller_saved_registers, and interrupt
#GCC options for #ARM ABIs (-mabi=) are just apcs-gnu, atpcs, aapcs, aapcs-linux, and iwmmxt. But there's an option to reserve a register for PIC: -msingle-pic-base -mpic-register=r10.
on 02023-07-19Oh, no, the #ARM #EABI is from 02003 and RealView 2.0 and later #GCC 4.1.0 and Linux 2.6.16.
on 02023-07-19oh hey #GCC has -fverbose-asm which helps significantly in getting readable #asm for #ARM!
apparently on #ARM when #GCC only has to push LR, it pushes an extra unnecessary temp register onto the stack to ensure SP is 8-byte-aligned at all times, because that’s what the ABI requires, because the ARMv4 strd and ldrd instructions require 8-byte alignment. #asm
gcc -no-pie is how you get #GCC to not try to generate a #position-independent executable, which disables ASLR for the executable. -static also forces that.
#GCC moved to #C++ starting in 02008 and that’s why it takes so much time and space to compile now
on 02022-05-21The 2011 #C++ standard required #GCC to break #ABI compatibility in GCC 5.1, although they did it in a backward-compatible way. But it’s not for some little-used feature; it’s for std::string.