Node:Usage, Next:, Previous:Selftest, Up:Glucas basics



SourceForgeLogo
 

Glucas Configure files.

At the moment, Glucas only does Lucas-Lehmer tests. Glucas manages the work to do in a queue file we can select. In addition to some features selectable on the command line, there is an ini file we can also use to configure more important aspects of Glucas and a results file where Glucas writes the most important results (not the verbose-output). (See Configure files.)

The usage of Glucas:

Glucas [-o][-n][-f][-s option][-i inifile][-q queuefile]
       [-r resultfile][-W Work_directory][file1][file2]

The optional [ ] items are:

-d
Use initial random shift. With this option, Glucas performs an initial random shift. The first Lucas-Lehmer iteration is (See Lucas-Lehmer test.) L(0)=4<<(initial_random_shift), with the default set to L(0)=4. With this initial random shift algorithm, two different Lucas-Lehmer tests made with same program on the same platform (but different random shifts) can be considered as independent. This option is used when doing double-checks.
-o
Write the save files in old style format. By default, Glucas writes the save files using the compact integer format for residues. When this flag is enabled, the save files are compatible with MERS rw() format. The save files in this form are three times bigger than the default format. Most important, the default uses the compatible Mersenne residue format, making the save files compatible among most platforms and processors out there. So, if you don't need it, don't use the -o flag.
-n
No nice flag. When selected, Glucas will run at normal user priority. This can slowdown other tasks. Don't use it unless you are going to do a short test, especially if you are not the owner of the system.
-f
Force this work to be done first. When Glucas is called with a worktodo file as argument on the commandline, by default this files work is appended at the end of the queue file. Using this flag, the the work is prepended at the the beginning of the queue file instead and will be done first.
-s options
selftest. (See Selftest.)
-i inifile
The inifile is inifile. Here you can set the ini file. The default file is glucas.ini. An existing ini file in the working directory is optional.
-q queuefile
The queue file is queuefile. Here you can set the queue file. If you don't select any, the default queue file will be glucas.que. NOTE: It is important to note that the queue files are for internal use of Glucas. Some users have tried to use files from PrimeNet as queue files and reported getting errors. To import files in PrimeNet format, simply add their name as last argument on the commandline (see below). Anyway the format of queue files is easy. A plain ascii line per exponent with four numbers. First is the exponent, second is the task (0 = First L-L, 1 = Double Check) and then other two parameters (at the moment 0).
-D initial_shift
Force initial shift. This option is similar to the -d option. Instead of using a random initial shift, this uses the commandline argument initial_shift. Used in some QA work.
-W work_directory
The working directory. Except binary, Glucas will read from/write to this work_directory. Don't use this option with selftest flag before. Write first the -W option and then -s option instead.
file1 file2 ...
Do the valid work in files: file1,file2 ... First, these arguments will be tried as file names. If existing, the valid work will be extracted and queued. Otherwise the arguments will be considered as an exponents to test. If there is no valid work, the queuefile remains unaltered.

Glucas is prepared to import the worktodo files from PrimeNet manual test forms. Paste the work obtained from PrimeNet in a file and restart Glucas:

Glucas [options] file

the work will be appended to the queue file. Warning: A queue file can only have up to 128 exponents to test. Do not make larger queue files.

Obviously, Glucas will stop when there is no other job in the queue file. When a job is interrupted, Glucas will store the work already done to save files. When restarting, Glucas will continue at the point it left the work for a given exponent. Remember Glucas will continue the task of the queue file we enter (or the default glucas.que).