Hi! On Fri, Dec 30, 2022 at 12:30:04AM -0800, Andrew Pinski wrote: > On Thu, Dec 29, 2022 at 11:45 PM Segher Boessenkool > <[email protected]> wrote: > > Ah! This simply shows rs6000_modes_tieable_p is decidedly non-optimal: > > it does not allow tying a scalar float to anything else. No such thing > > is required, or good apparently. I wonder why we have such restrictions > > at all in rs6000; is it just unfortunate history, was it good at one > > point in time? > > The documentation for TARGET_MODES_TIEABLE_P says the following: > If TARGET_HARD_REGNO_MODE_OK (r, mode1) and TARGET_HARD_REGNO_MODE_OK > (r, mode2) are always the same for any r, then TARGET_MODES_TIEABLE_P > (mode1, mode2) should be true. If they differ for any r, you should > define this hook to return false unless some other mechanism ensures > the accessibility of the value in a narrower mode. > > even though rs6000_hard_regno_mode_ok_uncached's comment has the following: > /* The float registers (except for VSX vector modes) can only hold floating > modes and DImode. */
That comment is incorrect. See fctiw for example, which defines only the SImode part of the result (the other bits are undefined). > TARGET_P8_VECTOR and TARGET_P9_VECTOR has special cased different modes now: > if (TARGET_P8_VECTOR && (mode == SImode)) > return 1; > > if (TARGET_P9_VECTOR && (mode == QImode || mode == HImode)) > return 1; > Which I suspect that means rs6000_modes_tieable_p should return true > for SImode and SFmode if TARGET_P8_VECTOR is true. Likewise for > TARGET_P9_VECTOR and SFmode and QImode/HImode too. It means that older CPUs do not have as many instructions to do scalar integer operations in vector registers, making it (almost) always a losing proposition to put scalar integers there. On newer CPUs it is not quite as bad, there is a full(er) complement of instructions to do such things in vector regs, just a bit slower than on GPRs. But yeah we might need to fix hard_regno_mode_ok if we change tieable. Segher
