On Thu, 13 Jul 2017, Andrew Pinski wrote: > Hi, > This patch adds Decimal floating point support to aarch64. It is > the base support in that since there is no hardware support for DFP, > it just defines the ABI. The ABI I chose is that _Decimal32 is > treated like float, _Decimal64 is treated like double and _Decimal128 > is treated like long double. In that they are passed via the floating > registers (sN, dN, qN). > Is this ok an ABI?
Shouldn't you also change config/dfp.m4 and regenerate affected configure scripts? Are you using DPD or BID encoding? What's the rationale for your choice between them? (x86 uses BID, software only; POWER and S/390 use DPD, and have some hardware support.) I'd then expect that choice to be documented in install.texi (though that seems out-of-date regarding S/390). Any plans for other software such as libdfp? (FWIW, whereas GCC defaults to DPD for architectures not otherwise configured in dfp.m4, libdfp defaults to BID for architectures other than POWER / S/390.) Are you working with the AAPCS64 document maintainers to get DFP support added there? -- Joseph S. Myers jos...@codesourcery.com
