105521b547
Merge branch 'master' into vec4-stack
2014-12-06 08:24:46 -08:00
fd3086f0a9
Fix compile warnings on RHEL5
2014-12-05 19:03:07 -08:00
d62a307c34
Fix 32-bit issue in vector4_to_value()
2014-12-05 18:00:08 -08:00
409f8c5823
Add the vec4 %subi instruction
2014-12-05 09:45:29 -08:00
03198356a5
vec4-stack stack manipulation improvements.
2014-12-04 17:15:27 -08:00
b1d2393789
Optimize the %sub instruction by integrating it with vvp_vector4_t class
2014-12-04 17:01:16 -08:00
0f740289e9
Optimize %mul instructions by integrating with vvp_vector4_t class.
2014-12-04 16:00:57 -08:00
46ce236cfb
Optimize the %add and %addi instructions
...
Tightly integrate with the vvp_vector4_t class to get much
better add performance.
2014-12-04 12:38:08 -08:00
86139c855d
Optimize the vec4-stack %cmp/s and %cmpi/s instructions.
...
Magnitude compare is called a LOT, so it is worth putting some
special effort into it.
2014-12-04 10:42:48 -08:00
eb070b061b
Optimize the vec4_to_index, which implements %ix/vec4 instructions.
2014-12-03 17:53:45 -08:00
2b1393e7b9
Reduce some vector copies in %load/vec4 and %concat/vec4 instructions.
...
By clever stack manipulations, we can eliminate some vector copies,
which can improve performance.
2014-12-03 13:12:06 -08:00
85c7b07a9b
Implement %cmp/ne and %cmpi/ne
...
These pull in the inversion of the output flags so that they more
efficiently implement != and !==, without %flag_inv instructions.
2014-12-03 11:06:11 -08:00
a57e0a1e3f
vvp implementation of copy from vec4 to vec2 can use subarray method.
2014-12-02 14:52:37 -08:00
b83144c552
The __vpiArray::set_word method should take a reference.
...
Seem to have missed a case where a vvp_vector4_t can be passed by
const reference instead of by value.
2014-12-02 13:02:46 -08:00
58fb80aec4
Implement and put to use the %muli instruction.
2014-12-02 12:46:17 -08:00
38f277d81b
Merge branch 'master' into vec4-stack
...
Conflicts:
vvp/array.cc
vvp/vthread.cc
2014-12-02 11:21:58 -08:00
1834d470cb
vvp: Refactored __vpiArrayVthrA & __vpiArrayVthrAPV.
2014-11-28 14:14:42 +01:00
9bc463aac0
vvp: Moved array_word_change(), array_attach_word(), array_alias_word() to __vpiArray.
2014-11-28 14:14:42 +01:00
7c77097f71
vvp: array_[set/get]_word changed to __vpiArray::[set/get]_word.
2014-11-28 14:14:42 +01:00
cc4c5f4998
vvp: Changed get_array_word_size(vvp_array_t) to __vpiArray::get_word_size().
2014-11-28 14:14:42 +01:00
230c435ae7
vvp: Added missing functions for __vpiDarray.
2014-11-28 14:14:42 +01:00
4ddef32631
vvp: __vpiArray::get_word_value handles StrVal variants.
2014-11-28 14:14:42 +01:00
a236c274f3
vvp: Minor changes.
2014-11-28 14:14:42 +01:00
21a8cb71ee
vvp: Handles for dynamic array ranges.
2014-11-28 14:14:42 +01:00
b4ccaa1b0c
vvp: Fixed iterators for dynamic arrays.
2014-11-28 14:14:42 +01:00
ca2ef5c956
vvp: Cleaning.
2014-11-28 14:14:39 +01:00
6015aceda2
vvp: vpiArrayIterator::vpi_index core moved to vpiArrayBase::get_iter_index.
2014-11-28 14:14:37 +01:00
9538501b54
vvp: Moved __vpiArrayBase::vpi_iterate to __vpiArayBase::vpi_array_base_iterate.
2014-11-28 14:12:49 +01:00
ea7fa24f9e
vvp: Further extraction of common code pieces from ArrayBase.
2014-11-28 14:12:46 +01:00
48d3701c02
vvp: get_word_index() & get_word_parent() moved to __vpiArrayWord.
2014-11-28 14:12:43 +01:00
c1164dcc33
vvp: Initial work on support for dynamic arrays in VPI.
...
Added array_common.[ch] to store shared code.
2014-11-28 14:12:37 +01:00
1166cd932a
vpi: Handle new formats in vpi_get_value() for unpacked arrays.
...
It is possible to use vpiIntVal, vpiVectorVal, vpiRealVal,
vpiStringVal and vpiObjTypeVal to retrieve data from unpacked
array cells.
2014-11-27 17:36:23 +01:00
679021a1b5
Optimize special cases of immediate value extraction.
2014-11-22 10:38:16 -08:00
f3392561ed
Optimize the vvp pad implementation.
2014-11-22 09:59:02 -08:00
7a9a022744
Remove some dead code.
2014-11-21 17:36:22 -08:00
c222b3b6a4
Minor improvements to the %split/vec4 instruction.
2014-11-21 16:51:59 -08:00
663c79d4af
Add the %cmp/e instructions, and put them to use.
...
When testing for == and ===, there is no need to also calculate <,
so it makes sense to have a special instruction for these cases.
2014-11-21 16:45:27 -08:00
bea03db25d
Minor vec4 stack manipulation optimizations.
2014-11-21 14:41:31 -08:00
b96f04ccce
Implement the %parti/X instructions
...
This allows part select with constant base to be handled optimally.
Also update some more instructions to more optimally work with
the vec4 stack.
2014-11-20 18:43:24 -08:00
a9db765f98
Generate better code for condition expressions.
2014-11-20 14:47:44 -08:00
04bdfbccee
Add %cmpi/s and %cmpi/u instructions for performance
...
These bypass the vec4 stack in some common cases, saving instructions
and vec4 manipulations.
Also, minor improvement to the %flag/set/vec4 statement.
Kill a few warnings.
2014-11-19 16:38:43 -08:00
2acc9fbdee
Remove dead instructsion %set/qb and %set/qf / Kill some warnings.
2014-11-19 09:15:01 -08:00
725ed869ba
Remove dead %load/vp0 instructions and related infrastructure
...
Also remove some &A<> and &PV<> symbols that use this now dead
infrastructure.
2014-11-19 09:02:21 -08:00
301edf69d3
Add and use %concati/vec4 and %addi instructions.
...
Also, clean up some warnings, and optimize some existing opcodes.
2014-11-18 12:27:55 -08:00
1612c6d638
Instruction %concat/vec4, %pad/u and %shiftl manipulate stack in place
2014-11-17 16:32:18 -08:00
aadd67cd3b
Some instructions can do stack manipultations in place.
...
By doing some stack manipulations in place, certain instructions
can eliminate, or optimize, vector copies.
2014-11-14 18:38:15 -08:00
0601b4e43b
Optimizations for %pushi/vec4 and %load/vec4
...
- Have %pushi/vec4 handle some special cases optimally.
- Eliminate some duplicated method calls in %load/vec4.
- Optimize the vvp_vector4_t::copy_from_ method by inlining
some parts.
2014-11-14 14:41:04 -08:00
8aca66b109
Performance tweaks
...
Better performance for %cmp/s and the vvp_fun_part_sa node.
2014-11-14 11:48:36 -08:00
c2e53f0d63
Fix for br963 - vvp qpop instructions need to call thr_check_addr.
2014-11-02 15:36:52 +00:00
8ac06fb63f
Support vpiObjTypeVal for vec4Stack values.
...
Also allow vpi_get_value outside of scope. (Return 'x)
2014-10-25 20:38:39 -07:00
Stephen Williams
Cary R
Maciej Suminski
Martin Whitaker