Date: Sun, 15 May 2011 01:30:28 +0200 From: Marius Strobl <marius@alchemy.franken.de> To: Attilio Rao <attilio@freebsd.org> Cc: svn-src-projects@freebsd.org, src-committers@freebsd.org Subject: Re: svn commit: r221842 - projects/largeSMP/sys/sparc64/sparc64 Message-ID: <20110514233028.GF92688@alchemy.franken.de> In-Reply-To: <BANLkTi=zi4D0RAUb2RLpGc9PcKm=bKPVvw@mail.gmail.com> References: <201105131521.p4DFLVKs074711@svn.freebsd.org> <BANLkTi=zi4D0RAUb2RLpGc9PcKm=bKPVvw@mail.gmail.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Fri, May 13, 2011 at 05:28:59PM +0200, Attilio Rao wrote: > 2011/5/13 Marius Strobl <marius@freebsd.org>: > > Author: marius > > Date: Fri May 13 15:21:31 2011 > > New Revision: 221842 > > URL: http://svn.freebsd.org/changeset/base/221842 > > > > Log: > > ??When setting up pc_other_cpus for APs based on pc_allcpu clear pc_cpuid > > ??in the former rather than the latter. > > ??This gets this branch working on at least Jalapeno-class CPUs. > > > > Modified: > > ??projects/largeSMP/sys/sparc64/sparc64/mp_machdep.c > > > > Modified: projects/largeSMP/sys/sparc64/sparc64/mp_machdep.c > > ============================================================================== > > --- projects/largeSMP/sys/sparc64/sparc64/mp_machdep.c ??Fri May 13 15:20:57 2011 ?? ?? ?? ??(r221841) > > +++ projects/largeSMP/sys/sparc64/sparc64/mp_machdep.c ??Fri May 13 15:21:31 2011 ?? ?? ?? ??(r221842) > > @@ -470,7 +470,7 @@ cpu_mp_bootstrap(struct pcpu *pc) > > ?? ?? ?? ??smp_cpus++; > > ?? ?? ?? ??KASSERT(curthread != NULL, ("%s: curthread", __func__)); > > ?? ?? ?? ??ocpus = all_cpus; > > - ?? ?? ?? CPU_CLR(curcpu, &all_cpus); > > + ?? ?? ?? CPU_CLR(curcpu, &ocpus); > > ?? ?? ?? ??PCPU_SET(other_cpus, ocpus); > > ?? ?? ?? ??printf("SMP: AP CPU #%d Launched!\n", curcpu); > > > > > > How idiot can I be? > I watched that path like 10 times yesterday... > FYI, as of r221932 this branch now works MP on all three classes of Cheetah, Jalapeno and Spitfire CPUs. While working on this I noticed that there's a bit of room for improvement in the sparc64 IPI code, which I'll commit once this branch has been merged into head. Then I'll also add support for more than 32 CPUs in the MD part, both of which really are orthogonal to the cpuset_t conversion. Marius
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20110514233028.GF92688>