| Message ID | b2693bf80cdfcd9fd6980b16e2093b3f14b08bff.1630511334.git.wschmidt@linux.ibm.com |
|---|---|
| State | New |
| Headers | show |
| Series | Replace the Power target-specific builtin machinery | expand |
On Wed, 2021-09-01 at 11:13 -0500, Bill Schmidt via Gcc-patches wrote: > I over-restricted use of __builtin_mffsl, since I was unaware that it > automatically uses mffs when mffsl is not available. Paul Clarke > pointed > this out in discussion of his SSE 4.1 compatibility patches. > > 2021-08-31 Bill Schmidt <wschmidt@linux.ibm.com> > > gcc/ > * config/rs6000/rs6000-call.c (__builtin_mffsl): Move from > [power9] > to [always]. > --- > gcc/config/rs6000/rs6000-builtin-new.def | 9 ++++++--- > 1 file changed, 6 insertions(+), 3 deletions(-) > > diff --git a/gcc/config/rs6000/rs6000-builtin-new.def > b/gcc/config/rs6000/rs6000-builtin-new.def > index 6a28d5189f8..a8c6b9e988f 100644 > --- a/gcc/config/rs6000/rs6000-builtin-new.def > +++ b/gcc/config/rs6000/rs6000-builtin-new.def > @@ -208,6 +208,12 @@ > double __builtin_mffs (); > MFFS rs6000_mffs {} > > +; Although the mffsl instruction is only available on POWER9 and > later > +; processors, this builtin automatically falls back to mffs on older > +; platforms. Thus it appears here in the [always] stanza. > + double __builtin_mffsl (); > + MFFSL rs6000_mffsl {} > + > ; This thing really assumes long double == __ibm128, and I'm told it > has > ; been used as such within libgcc. Given that __builtin_pack_ibm128 > ; exists for the same purpose, this should really not be used at > all. > @@ -2784,9 +2790,6 @@ > signed long long __builtin_darn_raw (); > DARN_RAW darn_raw {} > > - double __builtin_mffsl (); > - MFFSL rs6000_mffsl {} > - > const signed int __builtin_dtstsfi_eq_dd (const int<6>, > _Decimal64); > TSTSFI_EQ_DD dfptstsfi_eq_dd {} > Looks reasonable, Thanks -Will
On Wed, Sep 01, 2021 at 11:13:38AM -0500, Bill Schmidt wrote: > I over-restricted use of __builtin_mffsl, since I was unaware that it > automatically uses mffs when mffsl is not available. Paul Clarke pointed > this out in discussion of his SSE 4.1 compatibility patches. Right. Do we need to document this better? There are more builtins that can generate code for older archs than you might expect (like, set_fpscr_rn). Hrm, it *is* documented, but in a big wall of text. Not sure we can do much better though, there simply are this many builtins, but maybe you have an idea how to arrange things better? Anyway: okay for trunk. Thanks! Segher
diff --git a/gcc/config/rs6000/rs6000-builtin-new.def b/gcc/config/rs6000/rs6000-builtin-new.def index 6a28d5189f8..a8c6b9e988f 100644 --- a/gcc/config/rs6000/rs6000-builtin-new.def +++ b/gcc/config/rs6000/rs6000-builtin-new.def @@ -208,6 +208,12 @@ double __builtin_mffs (); MFFS rs6000_mffs {} +; Although the mffsl instruction is only available on POWER9 and later +; processors, this builtin automatically falls back to mffs on older +; platforms. Thus it appears here in the [always] stanza. + double __builtin_mffsl (); + MFFSL rs6000_mffsl {} + ; This thing really assumes long double == __ibm128, and I'm told it has ; been used as such within libgcc. Given that __builtin_pack_ibm128 ; exists for the same purpose, this should really not be used at all. @@ -2784,9 +2790,6 @@ signed long long __builtin_darn_raw (); DARN_RAW darn_raw {} - double __builtin_mffsl (); - MFFSL rs6000_mffsl {} - const signed int __builtin_dtstsfi_eq_dd (const int<6>, _Decimal64); TSTSFI_EQ_DD dfptstsfi_eq_dd {}