The command  \l creates a log file called by default  pari.log

In the next two lines I generate two large primes as factors of random numbers. The answers are given as matrices.

(09:10) gp > factor(9873487593794857872759730457827983857893794857972985704523)

%1 =

[3 1]

[3457849 1]

[31668164536226264201041 1]

[30055254729705346546970152849 1]

(09:11) gp > factor(8394789579270578937498572508475082734857958475811111)

%2 =

[3 1]

[70879 1]

[39479439510859252799364985907791601343406643603 1]

Next I create a composite integer   n  which is a product of two large primes.

(09:12) gp > n=%1[4,1]*%2[3,1]

%3 = 1186564611084868690886822575074728078567619925083753057867407227894978074947

I ask PARI if  n  is prime. The answer is instantenous NO (given as truth value 0 in less than a second)

(09:12) gp > isprime(n)

%4 = 0

For comparison: it takes PARI 52 minutes to actually factor  n.

(09:12) gp > factor(n)

%5 =

[30055254729705346546970152849 1]

[39479439510859252799364985907791601343406643603 1]

Here I do the primality test by hand.  a  is a randomly chosen number mod  n  and a^(n-1) is not one mod  n  so  n  is not prime.

(10:04) gp > a=Mod(938745,n)

%6 = Mod(938745, 1186564611084868690886822575074728078567619925083753057867407227894978074947)

(10:04) gp > a^(n-1)

%7 = Mod(948108788665754975270179520312202099971058880986156262640240616277220083349, 1186564611084868690886822575074728078567619925083753057867407227894978074947)

The following command closes the log file.

(10:04) gp > \l

log = 0 (off)

[logfile was "pari.log"]