Date: Tue, 1 Jul 2003 14:28:36 -0400 (EDT) From: Jeff Roberson <jroberson@chesapeake.net> To: Andrew Gallatin <gallatin@cs.duke.edu> Cc: current@freebsd.org Subject: Re: ULE problems on HTT SMP Message-ID: <20030701142802.G17881-100000@mail.chesapeake.net> In-Reply-To: <16128.16880.618174.525346@grasshopper.cs.duke.edu>
next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, 30 Jun 2003, Andrew Gallatin wrote: > > Jeff Roberson writes: > > On Fri, 27 Jun 2003, John Baldwin wrote: > > > > > > > > On 27-Jun-2003 Andrew Gallatin wrote: > > > > > > > > Jeff Roberson writes: > > > > > > > > > > Can you call kseq_print(0) and kseq_print(1) from ddb? > > > > > > > > > > > > > I found a different problem which is nearly as interesting. > > > > Note that ps thinks sysctl is on cpu 255... > > > > > > #define NOCPU 0xff /* For when we aren't on a CPU. (SMP) */ > > > > > > So that isn't but so interesting. :) > > > > The problem is that the logical cpu halting code does not put the halted > > CPU in the stopped cpus set. ULE has no way of knowing that it can not > > migrate a thread to this cpu. I'd prefer it if you could make this change > > John, but I can certainly do it if you're busy. > > > > Does this mean that if, as a temporary measure, I disable > machdep.cpu_idle_hlt, ULE should work for me? > Yes, but it needs to be disabled before booting so you'll have to adjust it in the code. See i386/i386/mp_machdep.c Cheers, Jeff
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20030701142802.G17881-100000>