GMP Problem with % under C++

Thu Nov 5 18:38:49 UTC 2020

Dear Sirs.

I have installed and started to use the GMP library. I wanted to use it
on Part B of the problem presented in Advent of Code 2019 on Day 22.
Previously, I was able to solve that problem only in Perl, using the
bigint library.

I am using Fedora 33. I downloaded and installed the GMP package by the
instruction 'sudo dnf install gmp gmp-devel'. The installation passed
without problems. The command 'sudo dnf info gmd' gives the following
answer:

   Name : gmp
   Epoch : 1
   Version : 6.2.0
   Release : 5.fc33
   Architecture : x86_64
   Size : 799 k
   Source : gmp-6.2.0-5.fc33.src.rpm
   Repository : @System
   From repo : fedora
   Summary : A GNU arbitrary precision library
   URL : http://gmplib.org/
   License : LGPLv3+ or GPLv2+

The command 'sudo dnf info gmd-devel' gives the following answer:

   Name : gmp-devel
   Epoch : 1
   Version : 6.2.0
   Release : 5.fc33
   Architecture : x86_64
   Size : 351 k
   Source : gmp-6.2.0-5.fc33.src.rpm
   Repository : @System
   From repo : fedora
   Summary : Development tools for the GNU MP arbitrary precision library
   URL : http://gmplib.org/
   License : LGPLv3+ or GPLv2+

The command 'uname -a' gives the following answer:

Linux kupka.lix.polytechnique.fr 5.8.17-300.fc33.x86_64 #1 SMP Thu Oct
29 15:55:40 UTC 2020 x86_64 x86_64 x86_64 GNU/Linux

I compiled the C++ sources with the command

   g++ -o 22b-gmp 22b-gmp.cpp -lgmpxx -lgmp

The last line of the output of the command 'g++ -v' is

   gcc version 10.2.1 20201016 (Red Hat 10.2.1-6) (GCC)

The result of the command '22b-gmp < 22input.txt' is:

   *** b = 62010736820046
   f(1) - b = -15681174147849
   a = -15681174147849
   On position 2020 is the card 102220661749926

For comparison, the result of the same computation in Perl with the
bigint package, by means of the command '22b-verbose.pl < 22input.txt"
is

   *** b = 62010736820046
   f(1) - b = -15681174147849
   a = 103634543366198
   On position 2020 is the card 93750418158025

My Perl program 22b-verbose.pl gives the right answer, as it was
approved by the Advent of Code website.

As you can see, the modulo operation % on Line 90 of the source file
22b-gmp.cpp has not been executed properly.

I am not sure if I used the GMP package properly or if I did everything
right, but it seems to me that very probably there is a problem
somewhere in the GMP package and/or its interface with C++. Please,
find enclosed the source of the C++ program (22b-gmp.cpp), the Perl
program (22-verbose.pl), and the input to both (22input.txt).

Thanks you very much in advance with dealing with my request.

Yous sincerely,
Nicolas (Miki) Hermann

-------------- next part --------------
A non-text attachment was scrubbed...
Name: 22b-gmp.cpp
Type: text/x-c++src
Size: 2370 bytes
Desc: not available
URL: <https://gmplib.org/list-archives/gmp-bugs/attachments/20201105/8a1bcd12/attachment.bin>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: 22b-verbose.pl
Type: application/x-perl
Size: 1855 bytes
Desc: not available
URL: <https://gmplib.org/list-archives/gmp-bugs/attachments/20201105/8a1bcd12/attachment-0001.bin>
-------------- next part --------------
cut -7890
deal with increment 17
cut 3131
deal with increment 60
cut -5009
deal into new stack
deal with increment 42
cut 9853
deal with increment 7
cut 9796
deal into new stack
deal with increment 3
deal into new stack
deal with increment 20
cut -8192
deal with increment 23
cut 9641
deal with increment 72
cut 7468
deal with increment 69
cut -7959
deal into new stack
deal with increment 54
cut -3345
deal with increment 48
cut 105
deal with increment 67
deal into new stack
deal with increment 66
cut 877
deal into new stack
cut 2789
deal into new stack
deal with increment 15
cut 8475
deal with increment 30
cut -7360
deal with increment 47
cut -4555
deal with increment 17
cut 1907
deal into new stack
cut -1934
deal with increment 53
deal into new stack
deal with increment 11
cut 8486
deal with increment 46
cut -4167
deal with increment 51
cut 3407
deal into new stack
cut 7797
deal with increment 9
cut 6983
deal with increment 13
cut 4408
deal into new stack
deal with increment 37
cut 422
deal with increment 18
cut 6433
deal with increment 11
cut 4689
deal into new stack
cut 6073
deal with increment 25
cut -3262
deal with increment 21
cut -2736
deal into new stack
deal with increment 53
cut -8515
deal with increment 9
cut 8314
deal with increment 41
deal into new stack
deal with increment 34
cut 5004
deal with increment 53
cut 3004
deal into new stack
deal with increment 6
cut -8757
deal with increment 38
cut -9534
deal with increment 27
cut 6052
deal with increment 54
cut 3907
deal into new stack
cut 7995
deal into new stack
cut 7291
deal with increment 15
cut 3884
deal with increment 4
cut -9637
deal with increment 59
cut -1150