Re: document the fact that "Known Maximum Errors" might not be maximal

[Vincent Lefevre <vincent@vinc17.net>]

On 2024-02-16 09:48:07 +0100, Paul Zimmermann wrote:
> the attached patch documents the fact that the "Known Maximum Errors"
> given in the reference manual might not be maximal.

 This section lists the known errors of the functions in the math
-library.  Errors are measured in ``units of the last place''.  This is a
+library.  However, for some functions like the Bessel functions,
+larger errors are known, but are documented elsewhere (bugzilla).
+Errors are measured in ``units of the last place''.  This is a

Instead of saying that they are "documented elsewhere", shouldn't the
manual be updated?

BTW, the manual contains

fmaf	-	-	-	-	-
fma	-	-	-	-	-
fmal	-	-	-	-	-
fmaf128	-	-	-	-	-
fma_ldoublef	-	-	-	-	-
fma_ldouble	-	-	-	-	-
fma_ldoublel	-	-	-	-	-
fma_ldoublef128	-	-	-	-	-

However, correct rounding is required for these functions.

And I can't see any reference to f128, ldoublef, etc. in the manual
except in this table.

-- 
Vincent Lefèvre <vincent@vinc17.net> - Web: <https://www.vinc17.net/>
100% accessible validated (X)HTML - Blog: <https://www.vinc17.net/blog/>
Work: CR INRIA - computer arithmetic / AriC project (LIP, ENS-Lyon)