Lines Matching refs:we
954 # maybe we can make these entry points ONLY the OVFL entry points of each routine.
964 # underflow exception. Since this is incorrect, we need to check
970 # the exceptional possibilities we have left ourselves with are ONLY overflow
989 # overflow is enabled AND overflow, of course, occurred. so, we have the EXOP
1321 # we determine the highest priority exception(if any) set by the
1395 # here, we insert the correct fsave status value into the fsave frame for the
1484 # here, we count on norm not to alter a0...
1561 # it here. if it was used from supervisor mode, then we have to handle this
1587 # the extended precision result is still in fp0. but, we need to save it
1588 # somewhere on the stack until we can copy it to its final resting place.
1589 # here, we're counting on the top of the stack to be the old place-holders
1866 # we determine the highest priority exception(if any) set by the
1902 # addressing mode was (a7)+. if so, we'll need to shift the
1973 # here, we insert the correct fsave status value into the fsave frame for the
1976 # as a reminder for future predicted pain and agony, we are passing in fsave the
2133 # addressing mode was -(a7). if so, we'll need to shift the
2297 # if we're stuffing a source operand back into an fsave frame then we
2416 # FPU before exiting. In either the enabled or disabled cases, we #
2417 # must also check if a Trace exception is pending, in which case, we #
2419 # stack frame. If no Trace is pending, we simply exit through #
2484 # here, we will have:
2584 # now, we're left with ftst and fcmp. so, first let's tag them so that they don't
2618 # we determine the highest priority exception(if any) set by the
2623 # now, we save the result, unless, of course, the operation was ftst or fcmp.
2944 # The FPU is disabled and so we should really have taken the "Line
2945 # F Emulator" exception. So, here we create an 8-word stack frame
2946 # from our 4-word stack frame. This means we must calculate the length
2997 # here, we actually create the 8-word frame from the 4-word frame,
3134 # cause an operr so we don't need to check for them here.
3327 # here, we simply see if the operand in the fsave frame needs to be "unskewed".
3348 # snan exceptions. we do this here before passing control to
3353 # fpsp_unsupp(), then we need to do nothing else for them here.
3354 # for byte, word, and long, we simply need to test the sign of the src
3636 # emulate the whole instruction. Notice, though, that we don't kill the
3790 # here, we simply see if the operand in the fsave frame needs to be "unskewed".
3856 # the exception was an "F-Line Illegal" exception. we check to see
4126 # until I feel like fixing this, we'll sidestep the exception.
4185 # no exception that was enabled was set BUT if we got an exact overflow
4205 # or underflow that was disabled, then we have to force an overflow or
4216 # appropriate type. don't forget to "skew" the source operand in case we
4386 # The ftrap<cc>, fs<cc>, or fdb<cc> is to take an enabled bsun. we must convert
4426 # as usual, we have to check for trace mode being on here. since instructions
6109 #--we want P+p = W+w but |p| <= half ulp of P
6110 #--Then, we need to compute A := R-P and a := r-p
6120 #--Now we need to normalize (A,a) to "new (R,r)" where R+r = A+a but
7071 # inefficiency, we stay away from denormalized numbers #
7072 # the best we can. For |X| >= 2^(-16312), the #
7649 # here, we build the result in a tmp location so as not to disturb the input
9627 # for the multiply to work. Therefore, we're going to actually do a
9634 subi.l &-0x3fff,%d0 # how many should we shift?
10162 # - we only return the EXOP here if either underflow or #
10208 # we have a DENORM that needs to be converted into an EXOP.
10285 # dst op is a DENORM. we have to normalize the mantissa to see if the
10287 # dst so we don't screw up the version passed to us.
11655 # - if overflow or inexact is enabled, we need a multiply result rounded to
11656 # extended precision. if the original operation was extended, then we have this
11657 # result. if the original operation was single or double, we have to do another
11695 # - if precision is extended, then we have the EXOP. simply bias the exponent
11696 # with an extra -0x6000. if the precision is single or double, we need to
11757 # no, it didn't overflow; we have correct result
11766 # - if overflow or inexact is enabled, we need a multiply result rounded to
11767 # extended precision. if the original operation was extended, then we have this
11768 # result. if the original operation was single or double, we have to do another
11871 # we don't know if the result was an underflow that rounded up to a 2 or
11987 # strictly compatible w/ the 68881/882, we make sure to return an
12639 # we don't know if the result was an underflow that rounded up to a 1
13372 # so, we could either set these manually or force the DENORM
13473 # precision selected is extended. so...we can not get an underflow
13614 # Therefore, we must return the result rounded to extended precision.
13656 # OVFL is not enabled; therefore, we must create the default result by
13706 # no, it didn't overflow; we have correct result
13842 # 'N' bit for a negative QNAN or SNAN input so we must squelch it here.
14166 # we don't know if the result was an underflow that rounded up to a 2 or
14501 # we still don't know if underflow occurred. result is ~ equal to 1. but,
14863 # now, we must determine whether the pre-rounded result was an underflow
14865 # so, we do this be re-executing the add using RZ as the rounding mode and
14976 # - Therefore, we return +ZERO if the rounding modes are RN,RZ, or RP.
15036 bpl.b fadd_inf_done # yes; we're done
15316 # now, we must determine whether the pre-rounded result was an underflow
15318 # so, we do this be re-executing the add using RZ as the rounding mode and
15428 # - Therefore, we return +ZERO if the rounding mode is RN,RZ, or RP
15696 # Therefore, we must return the result rounded to extended precision.
15738 # OVFL is not enabled; therefore, we must create the default result by
15791 # no, it didn't overflow; we have correct result
16331 # counter. However, the true branch may set bsun so we check to see #
16379 bne.w fdbcc_bsun # yes; we have an exception
16398 bne.w fdbcc_bsun # yes; we have an exception
16414 bne.w fdbcc_bsun # yes; we have an exception
16421 bne.w fdbcc_bsun # yes; we have an exception
16439 bne.w fdbcc_bsun # yes; we have an exception
16455 bne.w fdbcc_bsun # yes; we have an exception
16474 bne.w fdbcc_bsun # yes; we have an exception
16490 bne.w fdbcc_bsun # yes; we have an exception
16497 bne.w fdbcc_bsun # yes; we have an exception
16515 bne.w fdbcc_bsun # yes; we have an exception
16531 bne.w fdbcc_bsun # yes; we have an exception
16550 bne.w fdbcc_bsun # yes; we have an exception
16564 bne.w fdbcc_bsun # yes; we have an exception
16581 bne.w fdbcc_bsun # yes; we have an exception
16618 bne.w fdbcc_bsun # yes; we have an exception
16631 bne.w fdbcc_bsun # yes; we have an exception
16647 bne.w fdbcc_bsun # yes; we have an exception
16654 bne.w fdbcc_bsun # yes; we have an exception
16670 bne.w fdbcc_bsun # yes; we have an exception
16677 bne.w fdbcc_bsun # yes; we have an exception