On 22.03.2016 15:00, Peter Krempa wrote: > While refactoring the old way to store iothread scheduler info I've added an > algorithm that isn't entirely optimal but allows to store the scheduler info in > a sane way. Unfortunately when you specify an insane number of iothreads the > code takes ages to execute. > > To avoid this series being completely useless except for the one corner case > I've opted to finally add support for self expanding bitmaps, which might become > useful in the future. The self-expanding bitmap is then used instead of one of > the loops that was necessary to determine the maximum iothread ID. > > Peter Krempa (3): > util: bitmap: Intoduce self-expanding bitmap APIs > conf: decrease iterations complexity when formatting iothreads > conf: Remove now unused virDomainIOThreadIDMap > > src/conf/domain_conf.c | 51 ++++++++++---------------- > src/conf/domain_conf.h | 3 -- > src/libvirt_private.syms | 4 ++- > src/util/virbitmap.c | 93 ++++++++++++++++++++++++++++++++++++++++++++++++ > src/util/virbitmap.h | 8 +++++ > tests/virbitmaptest.c | 51 ++++++++++++++++++++++++++ > 6 files changed, 173 insertions(+), 37 deletions(-) > ACK series. Michal -- libvir-list mailing list libvir-list@xxxxxxxxxx https://www.redhat.com/mailman/listinfo/libvir-list