SQL & PL/SQL :: SELECT With Count Of Rows Depending On Sum Of A Field?

I want to create a SELECT, that shall give back only a special amount of rows, depending on the sum of one of the selected fields.

At first a code sample of the complete selection:

SELECT DISTINCT mnr, ktxt,
(SELECT Sum(meng_4)FROM reldb d1 WHERE d1.mnr=d.mnr)qty
FROM reldb d
WHERE mnr IN (SELECT mnr FROM relac WHERE Lower(rlnr) NOT LIKE 'platte geprägt%')
AND saext='M'
ORDER BY qty DESC,ktxt;

This selection produces some lines of output (in my case i.e. like 300). What I want to see is only that much lines that the condition 'sum of all items listed below meng_4<=sum of all items meng_4 of the whole selection * 0.9' is fulfilled.

So, if the whole selection produces a total of 10000 as sum for all items meng_4, I want to see only that amount of rows that sums a total of at least 9000 for all items meng_4.
I hope, this specification is exactly enough to understand my intent.

1. Can I do this in a query?

2. If yes, what would this query look like?