On 23 January 2012 16:26, Helmut Schaa <helmut.sc...@googlemail.com> wrote: > On Mon, Jan 23, 2012 at 3:13 PM, Roman Yeryomin <leroi.li...@gmail.com> wrote: >> On 23 January 2012 14:38, Helmut Schaa <helmut.sc...@googlemail.com> wrote: >>> On Mon, Jan 23, 2012 at 11:37 AM, Roman Yeryomin <leroi.li...@gmail.com> >>> wrote: >>>> I will test more thoroughly but I noticed about 0.7 MB/s speed drop >>>> comparing with those spinlocks commented out. >>> >>> Hmm, ok, might as well just be a wifi interference/noise issue? Maybe just >>> try on eth only with iperf or something similar? With a dedicated peer and >>> a simple crossover cable maybe. >> >> I don't think it's interference/noise issue as I've tried several >> times with and without spinlocks before writing to the list. >> The interesting part was that on eth only (bridged or routed) the >> speed was the same (...I think so, I'm not sure now) with and without >> spinlocks - 11.2 MB/s. > > Maybe because the CPU wasn't exhausted with "only" ethernet throughput > while adding bridge + wifi will bring the CPU further to its limits > and ksoftirqd > kicks in ...
I guess bridge is always there (with br-lan)? > And you really just removed the spin_lock in the housekeeping tasklet > and nothing else? So, you haven't reverted the whole patch, right? Yes, I just commented out those two spinlock lines. Regards, Roman _______________________________________________ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/mailman/listinfo/openwrt-devel
