Welcome, Guest
Username: Password: Remember me
  • Page:
  • 1
  • 2

TOPIC: Final merged result file appears corrupted.

Final merged result file appears corrupted. 8 years 9 months ago #19528

  • j_floyd
  • j_floyd's Avatar
After declaring py_pos to be 8byte integer throughout, there was still a problem. The calc of my_pos in add_data was being done as a int4 precission. Needed to force irec to int8 at line 2258 to get the correct numbers.

Need to attach full util_selafin.f file
The administrator has disabled public write access.

Final merged result file appears corrupted ----SOLVED 8 years 9 months ago #19529

  • j_floyd
  • j_floyd's Avatar
Hello

Full corrected utils_selafin.f file. Might have to check if there are other occurences of using my_pos and stream type write or read functions. Calc and variable for position need to be bigger than int4.

Cheers
John
Attachments:
The administrator has disabled public write access.

Final merged result file appears corrupted. SOLVED Part 2 8 years 9 months ago #19530

  • j_floyd
  • j_floyd's Avatar
Oops too quick to claim success,

missed a py_pos calc for the first variable in a time step in add_data.

New Source attached

Cheers
John
Attachments:
The administrator has disabled public write access.

Final merged result file appears corrupted. SOLVED Part 2 8 years 9 months ago #19533

  • yugi
  • yugi's Avatar
  • OFFLINE
  • openTELEMAC Guru
  • Posts: 851
  • Thank you received: 244
Hi,

Thanks for your work.
I indeed forgot to take into account that in big files the position could get bigger than a 4-byte-integer.
I will get your modification validated and tested and i think we might release a 7.1.1 with that correction and a couple others.

Cheers.
There are 10 types of people in the world: those who understand binary, and those who don't.
The administrator has disabled public write access.
The following user(s) said Thank You: Svensmolders, j_floyd

Final merged result file appears corrupted. SOLVED Part 2 8 years 9 months ago #19544

  • j_floyd
  • j_floyd's Avatar
One area that this doesnt cover, it assumes that each record will still fit within a int4 size. I think that this is quite adequate as that gives us around 10^9 elements/nodes - seems plenty.

I think we are safe ...
The administrator has disabled public write access.
  • Page:
  • 1
  • 2
Moderators: pham

The open TELEMAC-MASCARET template for Joomla!2.5, the HTML 4 version.