================
@@ -1169,7 +1169,11 @@
DEFINE_LIBUNWIND_FUNCTION(_ZN9libunwind15Registers_riscv6jumptoEv)
ILOAD x\i, (RISCV_ISIZE * \i)(a0)
.endr
// skip a0 for now
+#if defined(__riscv_abi_rve)
----------------
ArcaneNibble wrote:
In general, libunwind seems to be somewhat inconsistent with when it checks ISA
vs ABI defines?
* i386 doesn't touch any FPU registers at all. not sure what happens if
`vectorcall` gets involved?
* x86_64 only restores XMM registers for `_WIN64` which is an ABI-related define
* PPC32/PPC64 checks `__VSX__`/`__NO_FPRS__`/`__ALTIVEC__` which are
ISA-related defines
* ARM has its own special unique logic for handling FPU registers involving
runtime flags
* MIPS checks `__mips_hard_float` which is an ABI-related define but also
`__mips_isa_rev` which is an ISA-related define
* SPARC has separate functions for two different ABIs
* RISC-V checks `__riscv_flen` which is an ISA-related define
* LoongArch checks `__loongarch_frlen` which is an ISA-related define
* aarch64/or1k/hexagon/s390x do not have any variations to handle
https://github.com/llvm/llvm-project/pull/98855
_______________________________________________
cfe-commits mailing list
cfe-commits@lists.llvm.org
https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
