I think your regular expression here:

my %row = m/\s*(\w+)\S*\s+(\S.*)/g;
[download]

... doesn't extract everything because it's in "dot-does-not-match-newline" mode.

The /s switch makes "." match newlines ("treat variable as 's'ingle line") and thus could fix your problem:

$_ = <<DATA;
    Detail:              Some really nice infos these are 
    Info:              This is a problem
                       but there is a solution
DATA
my %row = m/\s*(\w+):\s+(\S[^:]+)/gs;
[download]

But then, that means that Info gets gobbled up again into the description part.

Personally, I would do manual line-by-line parsing instead of using one regular expression to capture everything:

$_ = <<DATA;
    Detail:              Some really nice infos these are 
    Info:              This is a problem
                       but there is a solution
DATA
my %row;

my $curr;
for (split /\n/) {
    if( /^\s*(\w+):\s+(.*)/ ) {
        $curr = $1;
        $row{ $curr } = $2;
    } elsif( $curr and /^\s*(.*)/ ) {
        $row{ $curr } .= ' ' . $1;
    } else {
        die "Unknown input data [$_]";
    };
};
[download]