Lines Matching refs:underflow
457 set unfl_bit, 3 # underflow
464 set aunfl_bit, 5 # accrued underflow bit
485 set unfl_mask, 0x00000800 # underflow exception mask
492 set aunfl_mask, 0x00000020 # accrued underflow
829 # _fpsp_unfl(): 060FPSP entry point for FP Underflow exception. #
832 # FP Underflow exception in an operating system. #
853 # Underflow Exception enabled: #
856 # Underflow Exception disabled: #
861 # On the 060, if an FP underflow is present as the result of any #
862 # instruction, the 060 will take an underflow exception whether the #
875 # Two other conditions exist. First, if underflow was disabled #
880 # underflow was disabled and the trace exception was enabled, this #
973 # underflow exception. Since this is incorrect, we need to check
975 # no underflow was called for. We do these checks only in
1006 # underflow exception. Since this is incorrect, we check here to see
1008 # no underflow was called for.
1034 # underflow exception.
1399 btst &unfl_bit,FPSR_EXCEPT(%a6) # did disabled underflow occur?
1731 # underflow can happen for extended precision. extended precision opclass
1977 btst &unfl_bit,FPSR_EXCEPT(%a6) # did disabled underflow occur?
2680 # or underflow that was disabled, then we have to force an overflow or
2681 # underflow frame.
2684 btst &unfl_bit,FPSR_EXCEPT(%a6) # did underflow occur?
4214 # or underflow that was disabled, then we have to force an overflow or
4215 # underflow frame. the eventual overflow or underflow handler will see that
4218 # a disabled overflow or underflow has also happened.
4221 btst &unfl_bit,FPSR_EXCEPT(%a6) # did underflow occur?
4244 # insert an underflow frame
5318 # here, the operation may underflow iff the precision is sgl or dbl.
6934 # neither overflow nor underflow. If AdjFlag = 1, that #
6937 # Hence, exp(X) may overflow or underflow or neither. #
6940 # over/underflow. Possible exception in 6.4 is overflow #
6941 # or underflow. The inexact exception is not generated in #
6978 # Notes: Exp(X) will surely overflow or underflow, depending on #
6980 # extended-precision numbers whose square over/underflow #
6982 # inexact together with either overflow or underflow. #
8991 # generate underflow by Tiny * Tiny. #
9173 #--|X| IS LARGE, GENERATE OVERFLOW IF X > 0; ELSE GENERATE UNDERFLOW
9477 # exponents out of range for single or double DO NOT cause underflow
9575 # >= 2^14, an overflow or underflow is returned. #
9687 bmi t_unfl # yes; underflow
10070 fmul.x Scale(%pc),%fp0 # may cause underflow
10171 # - we only return the EXOP here if either underflow or #
10352 # routine may have caused an underflow or overflow. #
10390 # an underflow or overflow exception occurred.
10416 # unf_res(): underflow default result calculation for transcendentals #
11365 # Overflow, Underflow, and Unsupported Data Type. #
11539 # unf_res() - return default underflow result #
11559 # an exception. If so, return the default overflow/underflow result #
11621 bgt.w fmul_unfl # result will underflow
11625 # - the result of the multiply operation will neither overflow nor underflow.
11770 # UNDERFLOW:
11771 # - the result of the multiply operation is an underflow.
11857 # MAY UNDERFLOW:
11859 # that do not underflow.
11875 fbgt.w fmul_normal_exit # no; no underflow occurred
11876 fblt.w fmul_unfl # yes; underflow occurred
11879 # we still don't know if underflow occurred. result is ~ equal to 2. but,
11880 # we don't know if the result was an underflow that rounded up to a 2 or
11899 fbge.w fmul_normal_exit # no; no underflow occurred
11900 bra.w fmul_unfl # yes, underflow occurred
12036 # unf_res() - return default underflow result #
12086 # precision selected is extended. so...we cannot get an underflow
12116 # the input is an extended DENORM and underflow is enabled in the FPCR.
12152 cmpi.l %d0,&0x3fff-0x3f80 # will move in underflow?
12153 bge.w fin_sd_unfl # yes; go handle underflow
12159 # operand will NOT overflow or underflow when moved into the fp reg file
12195 cmpi.l %d0,&0x3fff-0x3c00 # will move in underflow?
12196 bge.w fin_sd_unfl # yes; go handle underflow
12203 # operand WILL underflow when moved in to the fp register file
12212 # if underflow or inexact is enabled, then go calculate the EXOP first.
12227 # operand will underflow AND underflow or inexact is enabled.
12356 # unf_res() - return default underflow result #
12376 # an exception. If so, return the default overflow/underflow result #
12441 cmp.l %d0,(tbl_fdiv_unfl.w,%pc,%d1.w*4) # will result underflow?
12443 bgt.w fdiv_unfl # yes; go handle underflow
12626 # the divide operation MAY underflow:
12643 fbgt.w fdiv_normal_exit # no; no underflow occurred
12644 fblt.w fdiv_unfl # yes; underflow occurred
12647 # we still don't know if underflow occurred. result is ~ equal to 1. but,
12648 # we don't know if the result was an underflow that rounded up to a 1
12667 fbge.w fdiv_normal_exit # no; no underflow occurred
12668 bra.w fdiv_unfl # yes; underflow occurred
12811 # unf_res() - return default underflow result #
12827 # and an actual fneg performed to see if overflow/underflow would have #
12828 # occurred. If so, return default underflow/overflow result. Else, #
12859 # precision selected is extended. so...we can not get an underflow
12899 # the input is an extended DENORM and underflow is enabled in the FPCR.
12932 cmpi.l %d0,&0x3fff-0x3f80 # will move in underflow?
12933 bge.w fneg_sd_unfl # yes; go handle underflow
12939 # operand will NOT overflow or underflow when moved in to the fp reg file
12975 cmpi.l %d0,&0x3fff-0x3c00 # will move in underflow?
12976 bge.b fneg_sd_unfl # yes; go handle underflow
12983 # operand WILL underflow when moved in to the fp register file
12992 # if underflow or inexact is enabled, go calculate EXOP first.
13007 # operand will underflow AND underflow is enabled.
13083 # the move in MAY underflow. so...
13429 # unf_res() - calculate underflow result #
13445 # gets an EXOP created for it since it's an underflow. #
13446 # Double and single precision can overflow and underflow. First, #
13482 # precision selected is extended. so...we can not get an underflow
13517 # the input is an extended DENORM and underflow is enabled in the FPCR.
13550 cmpi.l %d0,&0x3fff-0x3f80 # will move in underflow?
13551 bge.w fabs_sd_unfl # yes; go handle underflow
13557 # operand will NOT overflow or underflow when moved in to the fp reg file
13593 cmpi.l %d0,&0x3fff-0x3c00 # will move in underflow?
13594 bge.b fabs_sd_unfl # yes; go handle underflow
13601 # operand WILL underflow when moved in to the fp register file
13608 # if underflow or inexact is enabled, go calculate EXOP first.
13622 # operand will underflow AND underflow is enabled.
13698 # the move in MAY underflow. so...
13941 # unf_res4() - return default underflow result for sglop #
13961 # an exception. If so, return the default overflow/underflow result #
14000 bgt.w fsglmul_unfl # result will underflow
14170 fbgt.w fsglmul_normal_exit # no; no underflow occurred
14171 fblt.w fsglmul_unfl # yes; underflow occurred
14174 # we still don't know if underflow occurred. result is ~ equal to 2. but,
14175 # we don't know if the result was an underflow that rounded up to a 2 or
14194 fbge.w fsglmul_normal_exit # no; no underflow occurred
14195 bra.w fsglmul_unfl # yes, underflow occurred
14282 # unf_res4() - return default underflow result for sglop #
14302 # an exception. If so, return the default overflow/underflow result #
14345 cmpi.l %d0,&0x3fff-0x0000 # will result underflow?
14347 bgt.w fsgldiv_unfl # yes; go handle underflow
14489 # the divide operation MAY underflow:
14506 fbgt.w fsgldiv_normal_exit # no; no underflow occurred
14507 fblt.w fsgldiv_unfl # yes; underflow occurred
14510 # we still don't know if underflow occurred. result is ~ equal to 1. but,
14511 # we don't know if the result was an underflow that rounded up to a 1
14529 underflow occurred
14530 bra.w fsgldiv_unfl # yes; underflow occurred
14619 # unf_res() - return default underflow result #
14701 cmp.l %d2,(tbl_fadd_unfl.b,%pc,%d1.w*4) # is it an underflow?
14850 # underflow that rounded up.
14855 beq.w fadd_normal # yes; no underflow occurred
14859 bne.w fadd_normal # no; no underflow occurred
14862 bne.w fadd_normal # no; no underflow occurred
14865 beq.w fadd_normal # no; no underflow occurred
14872 # now, we must determine whether the pre-rounded result was an underflow
14893 fbgt.w fadd_unfl # yes; it's an underflow
14894 bra.w fadd_normal # no; it's not an underflow
15072 # unf_res() - return default underflow result #
15154 cmp.l %d2,(tbl_fsub_unfl.b,%pc,%d1.w*4) # is it an underflow?
15303 # underflow that rounded up.
15308 beq.w fsub_normal # yes; no underflow occurred
15312 bne.w fsub_normal # no; no underflow occurred
15315 bne.w fsub_normal # no; no underflow occurred
15318 beq.w fsub_normal # no; no underflow occurred
15325 # now, we must determine whether the pre-rounded result was an underflow
15346 fbgt.w fsub_unfl # yes; it's an underflow
15347 bra.w fsub_normal # no; it's not an underflow
15517 # unf_res() - return default underflow result #
15536 # an exception. If so, return the default overflow/underflow result #
15612 cmpi.l %d0,&0x3fff-0x3f81 # will move in underflow?
15614 bgt.w fsqrt_sd_unfl # yes; go handle underflow
15620 # operand will NOT overflow or underflow when moved in to the fp reg file
15657 cmpi.l %d0,&0x3fff-0x3c01 # will move in underflow?
15659 bgt.b fsqrt_sd_unfl # yes; go handle underflow
15667 # elsewise fall through to underflow.
15670 bne.w fsqrt_sd_normal # yes, so no underflow
15673 # operand WILL underflow when moved in to the fp register file
15688 # if underflow or inexact is enabled, go calculate EXOP first.
15704 # operand will underflow AND underflow is enabled.
15780 # the move in MAY underflow. so...
20165 # unf_res() - create default underflow result for sgl/dbl prec. #
20181 # fp0 : intermediate underflow or overflow result if #
20192 # For sgl or dbl precision, overflow or underflow can occur. If #
20196 # the source is a denorm and if underflow is enabled, an EXOP must be #
20370 # The DENORM causes an Underflow exception.
20404 # the number is a DENORM. must set the underflow exception bit
20406 bset &unfl_bit,FPSR_EXCEPT(%a6) # set underflow exc bit
20449 # would cause either an underflow or overflow. these cases are handled
20459 cmpi.w %d0,&SGL_LO # will operand underflow?
20460 blt.w fout_sgl_unfl # yes; go handle underflow
20522 bsr.l unf_res # calc default underflow result
20687 # would cause either an underflow or overflow. these cases are handled
20697 cmpi.w %d0,&DBL_LO # will operand underflow?
20698 blt.w fout_dbl_unfl # yes; go handle underflow
20751 bsr.l unf_res # calc default underflow result
21700 # According to the exponent underflow threshold for the given #
21706 # (1) _denorm() is called by the underflow routines #
21781 # check to see how much less than the underflow threshold the operand
22676 # unf_res(): routine to produce default underflow result of a #
22692 # a0 = pointer to default underflow result in extended precision #
22761 # we know that underflow has occurred. aunfl should be set if INEX2 is also set.
22821 # we know that underflow has occurred. aunfl should be set if INEX2 is also set.
23952 # in this way, we avoid underflow on intermediate stages of the
Indexes created Wed Oct 15 07:09:58 GMT 2025