| Title: | Alpha Developer Support |
| Notice: | [email protected], 800-332-4786 |
| Moderator: | HYDRA::SYSTEM |
| Created: | Mon Jun 06 1994 |
| Last Modified: | Fri Jun 06 1997 |
| Last Successful Update: | Fri Jun 06 1997 |
| Number of topics: | 3722 |
| Total number of notes: | 11359 |
Company Name : National Information System Inc
Contact Name : Arden Scott
Phone : 408-985-7100
Fax :
Email : [email protected]
Date/Time in : 1-MAY-1997 16:16:01
Entered by : L. Narayan
SPE center : MRO
Category : VMS
OS Version : V7.1
System H/W : Alpha
Brief Description of Problem:
-----------------------------
Arden has two questions
1. FORTRAN question
Passing a constant variable from a FORTRAN main program to a subroutine changes the
type to floating point variable in the subroutine.
I have requested Arden to send a sample code.
2. C library question
the funcation string to double ( strtod() ) while passing the value "12.25"
returns different values different architecture. On VAX it returns 1.4 and
on Alpha 1.3. The correct value should be 1.4
| T.R | Title | User | Personal Name | Date | Lines |
|---|---|---|---|---|---|
| 3575.1 | suggested compiler options | HYDRA::LNARAYAN | Thu May 01 1997 16:33 | 3 | |
For the second question I suggested him to try with the compiler
options assume_accuracy_sensitive, i16 etc. I'll also test this options
and let him know
| |||||
| 3575.2 | status | HYDRA::LNARAYAN | Thu May 29 1997 11:51 | 19 | |
For the FORTRAN question I have answered according to the replies to my
entry in the TURRIS::fortran note #1285
Regarding the accuracy, I understand ( thanks Chuck ) that
on VAX
D_FLOAT is default and 56 bit precision
G_FLOAT is 53 bit precision
ON ALPHA
G_FLOAT is default and 53 bit precision
IEEE 53 bit precision
The partner has developed his application using D_FLOAT on VAX and now
is porting this to Alpha. Because of this D_FLOAT to G_FLOAT conversion
he loses accuracy. That is his concern.
| |||||