Problem with ?orgqr for full Q

[Fernando_Moura]

Hey Guys,

This is a 'Complete the reference manual' thread.

There is a very important information missing in ?orgqr reference manual.

If we are interested in computing all 'm' columns of Q using

?orgqr(m, m, p, a, lda, tau, work, lwork, info)

the second dimension of array 'a' -->MUST<-- be at least 'm', so it can store a 'm' by 'm' matrix. Otherwise the result will be wrong and there will be some serious memory leaks.

Please add this information to your reference manual, both html and pdf.

Cheers

[Aleksandr_Z_Intel]

Dear Fernando,

Let me quotethe manual for ?orgqr:

"The second dimension ofamust be at least max(1,n)"

And n = m in the case.

So, information is already in reference manual.