summaryrefslogtreecommitdiffhomepage
path: root/ir/tv
Commit message (Collapse)AuthorAge
* Recognize quad precision floats as used on AArch64.Manuel Mohr2021-03-04
|
* Reorganize include directivesMatthias Braun2017-02-17
|
* Fix indentationSebastian Buchwald2016-05-12
|
* tv: Add 'tarval_is_minus_null()'.Christoph Mallon2016-04-22
|
* tv: Add get_tarval_magnitude().Christoph Mallon2016-02-26
|
* fltcalc: Report negative overflow intead of bad when converting negative ↵Matthias Braun2016-02-14
| | | | number to unsigned int
* Rename is_po2() to is_po2_or_zero()Matthias Braun2016-02-13
| | | | | The function also returns true for zero even though that is not a power of two.
* tv: Allow irma_twos_complement everywhere where mode_is_int was allowedMatthias Braun2016-01-24
| | | | | pointer modes have irma_tows_complement as well and I don't see why we shouldn't allow performing tarval operations on them like this.
* tv: Only allow tarval_popcount on twos_complement modesMatthias Braun2016-01-24
|
* tv: More constMatthias Braun2016-01-24
|
* Use more const modifiers for tarval operationsMatthias Braun2016-01-24
|
* adt: Add and use 'streq()'.Christoph Mallon2015-12-07
|
* Avoid unnecessary inclusion of strings.hMatthias Braun2015-12-06
|
* tv: Remove mode parameter from tarval_subMatthias Braun2015-10-18
| | | | | We can get a proper mode for pointer-pointer from get_reference_offset_mode().
* Always include irnode_t.h from inside libfirm (or firm_types.h)Matthias Braun2015-09-10
|
* Move ir_node struct from irtypes.h to irnode_t.h, remove irtypes.h, cleanupMatthias Braun2015-09-10
|
* Move ir_mode struct from irtypes.h to irmode_t.h, cleanupMatthias Braun2015-09-10
|
* Constify tarval parameter.Sebastian Buchwald2015-06-05
|
* Constify mode parameter.Sebastian Buchwald2015-06-05
|
* cleanup: Move the implementation of bitset_fprint() into .c, to not spill ↵Christoph Mallon2015-05-29
| | | | #include "irprintf.h" everywhere.
* Fix fc_castAndreas Zwinkau2015-05-18
| | | | | | | | | | | | | | Matze's cleanup broke things 8ebe31d226843d6f2cdfcc1ee100ca607c79ba65 Apparently, value==result shall be explicitly allowed, but in the case of NaN, result was modified and afterwards a read from value returns the wrong results. Fix was to move the reads before the writes. Also, a cast from implicit one to explicit one forgot to set the one. Likewise, the size comparison was oblivious of explicit ones. Adds few more asserts. Adds a unittest to check NaN payload stuff.
* add assertsAndreas Zwinkau2015-05-18
|
* refactor assertsAndreas Zwinkau2015-05-18
|
* util: Add and use is_digit().Christoph Mallon2015-04-27
|
* tv: When checking for lossless conversion, examine the mantissa of NaNs.Christoph Mallon2015-02-01
|
* tv: Correct converting NaNs.Christoph Mallon2015-02-01
| | | | Now the mantissa is correctly transferred, i.e. it is aligned at the most significant bit.
* tv: NaN can be negated, too.Christoph Mallon2015-02-01
|
* tv: Print the sign of NaN and zero, too.Christoph Mallon2015-02-01
|
* cleanup: Remove unnecessary #include <limits.h>.Christoph Mallon2015-01-25
|
* tv: Allow tarval_get_bit() for tarvals with any data mode.Christoph Mallon2015-01-18
| | | | In particular this unbreaks dumping boolean bitinfo.
* tv: cleanup; remove some unused functionsMatthias Braun2014-12-31
|
* assert for float mode in get_tarval_{small|epsilon}; remove unused ↵Matthias Braun2014-12-31
| | | | get_tarval_minus_inf
* restrict new_tarval_from_long to two_comeplement modesMatthias Braun2014-12-31
|
* tv: remove unused possibility to parse boolean values and 'null'Matthias Braun2014-12-31
|
* fltcalc: assert that there is enough precisionMatthias Braun2014-12-31
|
* fltcalc: remove unreachable codeMatthias Braun2014-12-31
|
* strcalc: remove unused SC_hex, SC_OCT and SC_BIN printing supportMatthias Braun2014-12-31
|
* tv: split get_tarval() into get_int_tarval() and get_fp_tarval()Matthias Braun2014-12-31
|
* fltcalc: cleanupMatthias Braun2014-12-31
|
* fltcalc: cleanup, remove printing in FC_PACKED(unused) and FC_BIN(unimplemented)Matthias Braun2014-12-31
|
* fltcalc: remove the global accumulation bufferMatthias Braun2014-12-31
| | | | | | This avoids constant testing whether an output buffer was specified or the global accumulation buffer should be used. Also remove pointless return values.
* fltcalc: change normalize() to always work in placeMatthias Braun2014-12-31
|
* strcalc: change sc_inc to always work in-placeMatthias Braun2014-12-31
|
* tv: rework NaN handlingMatthias Braun2014-12-30
| | | | | | | | - remove get_mode_nan() as there is not just a single NaN values but several different ones. - introduce new_tarval_nan() which allows to construct signaling and quiet NaNs with payload. - introduce tarval_is_snan() and tarval_is_qnan() to help unittests.
* fltcalc: fix some NaN handling bugs.Matthias Braun2014-12-30
|
* fltcalc: take the "smaller NaN" for binops with 2 NaN inputsMatthias Braun2014-12-30
| | | | | | | | | | ieee754 specifies that we have to take the payload of one of the input NaN. We choose the "smaller" one so commutativity of mal/add is preserved. This is still somewhat problematic as we really should do what our actual compilation target host would do, but have no information about that...
* tv: remove space skipping feature and len == 0 handling from ↵Matthias Braun2014-12-30
| | | | | | | | new_tarval_from_str_int() The tarval module is not the place to validate your input, leave that to frontends and others... This way the code is simpler and slightly faster.
* tv: introduce tarval_to_bytes()Matthias Braun2014-12-30
|
* tv: remove endianess paramter from new_tarval_from_bytes()Matthias Braun2014-12-30
| | | | | | Most inputs are little endian anyway and the few remaining users can easily swap the bytes themselfes instead of having the function swap them.
* strcalc: cleanup and fix print_buf for mode_size%8 != 0Matthias Braun2014-12-30
|