Hello,
I use cblas_dtrsm heavily as part of my application and in some cases need to solve a system where the right hand side is a column of the matrix itself and I would like to get the result stored within the matrix itself i.e. Rx=x where x=R_{k} i.e. x is the column vector k of the matrix. Is this possible without getting race conditions or corrupted matrix R? In my case k is always the last column of the matrix. This will help me squeeze every bit of performance by avoiding an extra copy and allocation for each invocation of this use-case.
Many TIA,
Best regards,
Giovanni
