On 05/08/2017 12:29 PM, Thomas Koenig wrote: > Am 08.05.2017 um 18:58 schrieb Jerry DeLisle: > > he attached patch reduces the stack usage by the blocked >>> version of matmul for cases where we don't need the full buffer. >>> This should improve stack usage. >>> >>> OK for trunk? >>> >> >> OK, thanks. > > Is this something we should consider for backporting to gcc-7? > I think the large block size caused > https://gcc.gnu.org/bugzilla/show_bug.cgi?id=79876 . > > That one was fixed by adjusting the ridiculously low > stack size for multi-threaded applications on OSX, but > the underlying problem could still bite somewhere or > somebody else. > > Opinions? >
I think it should be back ported since we really changed the matmul and this was a fallout and using a to much stack might end up being a regression. OK to back port. Jerry
