If you have a lot of RAM or your genome fasta file is fragmented in small pieces, you can just use a python script (requires biopython). Something like this will work:

import sys
from Bio import SeqIO

query = sys.argv[2].lower()

for record in SeqIO.parse(open(sys.argv[1],'r'),'fasta'):
    indeces = []
    i = str(record.seq).lower().find(query)
    while i >= 0:
        indeces.append(str(i + 1))
        i =  str(record.seq).lower().find(query, i + 1)

    if len(indeces) > 0:
        print record.id + "\t" + ','.join(indeces)

Use by:

python script.py genome.fasta 'AGATAGATATATAGATAGA'

Took a minute or two on my computer with a ~800 megabyte genome in around 40,000 pieces.

You can use Biopieces in particular patscan_seq:

read_fasta -i genome.fna | patscan_seq -p AGATAGATATATAGATAGA | write_bed -x

Another solution in Python:

import sys

query = raw_input('Enter the query:\n')

subject = raw_input('Enter the subject:\n')

if query not in subject:
    sys.exit('Query not in subject')

n = len(query)
m = len(subject)
j = 0

while j <= m:
    if query == subject[j:j+n]:
        print 'Found match at positions', j+1, '-', j+n
    j += 1