Found via `codespell -q 3 -w -L ith,als,dum,nd,amin,nto,wis,ba -S ./relapack,./kernel,./lapack-netlib`

Correct length of name string in xerbla call

Provide CBLAS interfaces to I?MIN and I?MAX

revert workaround introduced for issue #1332 as the actual cause appears to be my incorrect fix from #1262 (see #1388)

See #1886 and JuliaRobotics issue 500. trsm benchmarks on Haswell and Zen showed that with these values performance is roughly doubled for matrix sizes between 8x8 and 14x14, and still 10 to 20 percent better near the new cutoff at 32x32.

OpenBLAS has a fancy algorithm for copying the input data while laying
it out in a more CPU friendly memory layout.

This is great for large matrixes; the cost of the copy is easily
ammortized by the gains from the better memory layout.

But for small matrixes (on CPUs that can do efficient unaligned loads) this
copy can be a net loss.

This patch adds (for SKYLAKEX initially) a "sgemm direct" mode, that bypasses
the whole copy machinary for ALPHA=1/BETA=0/... standard arguments,
for small matrixes only.

What is small? For the non-threaded case this has been measured to be
in the M*N*K = 28 * 512 * 512 range, while in the threaded case it's
less, around M*N*K = 1 * 512 * 512

Handle special case INCX=0,INCY=0 in the axpy interface

disable threading in C/ZSWAP copying from S/DSWAP

Delay _hemv threading in attempt to address #1820

Currently the generic ARMV8 target uses C implementations
for many routines. Replace these with the neon implementations
written for THUNDERX2T99 target which are upto 6x faster for
certain routines.

Add explicit casts to silence compiler warnings

for #1710

for #1710

Introduce blasabs macro and use it to switch between abs and labs for INTERFACE64

Fixes two calls that were using `fabs` on a `long double` argument rather than `fabsl`, which looks like it is doing an unintentional truncation to `double` precision.

I don't have as many benchmarks for these as for gemm, but it should still
make a difference for small matrices.

Always checking num_cpu_avail() regardless of whether threading will actually
be used adds noticeable overhead for small matrices.  Most other uses of
num_cpu_avail() do so only if threading will be used, so do the same here.

copies daxpy improvement from #27, see #1560

Disable multithreading in ztrmv

BLAS-Tester shows that the same problem exists as with DTRMV (issue #1332)

* Rewrite ROTMG based on the new implementation in GONUM based on the algorithm proposed by Tim Hopkins, see issue 1452 for the reference
* Correct ROTMG utest for issue1452 and add another from gonum, also correct transposition of expected and observed values in error messages

Fixes #1452 and is more in line with how ATLAS does it. The earlier fix from #356 only moved the bug elsewhere, but we will never want the iterative rescaling to change the dflag setting and variable associations with each cycle.

as a (hopefully temporary) workaround for #1332

Eliminate mode variable where not needed in syrk interface

Modify complex CBLAS functions to take void pointers instead of float or double arguments (to bring the prototypes in line with netlib and other implementations' cblas.h)

Fix cols/rows mixup in omatcopy 2nd step for BlasTrans cases

Equivalent of #1244 (issue #899) for the non-complex cases. Fixes #1289

Optimized standard Blas Level-1,2 (excluding nrm2 functions) for z13 (double precision). Issue 884

Use in-place transform shortcut only if matrix is square