Date: Mon, 6 Feb 2012 10:35:02 -0500 From: Ryan Stone <rysto32@gmail.com> To: Ivan Voras <ivoras@freebsd.org> Cc: svn-src-head@freebsd.org, Ryan Stone <rstone@freebsd.org>, src-committers@freebsd.org, svn-src-all@freebsd.org Subject: Re: svn commit: r230984 - head/sys/kern Message-ID: <CAFMmRNyk2ftJiwZ7vtCPjjbuo%2BQ_OAZhnKn4AqqZ0R0Tj%2BOOGw@mail.gmail.com> In-Reply-To: <CAF-QHFXQXqFskmMU%2BYxqmFEVKOoiHGMdnHBkSjqaEVZmjs_ovQ@mail.gmail.com> References: <201202041649.q14GnUnI043572@svn.freebsd.org> <CAF-QHFXQXqFskmMU%2BYxqmFEVKOoiHGMdnHBkSjqaEVZmjs_ovQ@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, Feb 6, 2012 at 6:34 AM, Ivan Voras <ivoras@freebsd.org> wrote: > Shouldn't new threads inherit affinity from the threads which spawned them? Kernel threads were not inheriting affinity from the thread that spawned them. They were inheriting affinity from an arbitrary sibling thread in the same kernel process. For example, if you set net.isr.bindthreads=1, all threads in the intr process that don't set affinity(which is most of them) would end up bound to CPU 0, including every single softclock thread.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAFMmRNyk2ftJiwZ7vtCPjjbuo%2BQ_OAZhnKn4AqqZ0R0Tj%2BOOGw>