-
Notifications
You must be signed in to change notification settings - Fork 14
/
TODO
66 lines (50 loc) · 1.95 KB
/
TODO
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
TODO before 1.0 final release
Serious bugs
------------
No known serious bug left so far. Exception raising should be improved.
New functions
-------------
Document the cases on which power is guaranteed correctly rounded values.
Possibly write 2^x and 10^x (in harmonious combination with power)
Proof stuff
-----------
Update chapter asin and write chapter acos.
Add the scripts generating the polynomials.
Write Gappa proofs for log2-td, log10-td, expm1, log1p
Check all the existing .gappa against last version of Gappa
+ set the paths right (where there are paths).
Write proper proofs of the DE integer rounding tests, or remove DE altogether.
Cleaning up stuff
-----------------
Check .mpl with underscored variables (some versions of Maple don't like them)
Purge obsolete .mpl
Ensure compilation on icc9 and SunStudio11
Solution: check recent icc first. Then add FMADE macros everywhere FMA intended
add make timings
Autotest stuff
--------------
Add worst cases for all which are currently written TODO
A lot of the Lefevre worst cases are still missing
You never have enough test cases. In particular asin and acos
More cases around the various boundaries would be welcome.
Optim stuff
-----------
Remove remaining SCS: in atan, in trigo.
Replace the current test for non-IEEE subnormal (on x86) in the code
with a test taking into account that x86-64 is OK
Minor improvements to atanpi as mentionned in the doc.
Fuse atan-pentium and atan-itanium into atan-de, and add in atanpi.
Or maybe remove all the DE stuff altogether for cleaning up.
===========================================================================
TODO some day for a later release
New functions
-------------
Write inverse hyperbolic functions
Optimisation stuff
------------------
Write TD versions of: trigo
Write DE versions of: most functions
Use technique in paper by Karp and Markstein for sqrt12 and sqrt13
Interval stuff
--------------
Write interval functions for each CR function