Date: Mon, 30 May 2011 15:22:56 -0500 From: Nathan Whitehorn <nwhitehorn@freebsd.org> To: John Baldwin <jhb@freebsd.org> Cc: svn-src-projects@freebsd.org, src-committers@freebsd.org Subject: Re: svn commit: r222356 - projects/pseries/kern Message-ID: <4DE3FCA0.2000400@freebsd.org> In-Reply-To: <201105271202.04421.jhb@freebsd.org> References: <201105271427.p4RERSUb039281@svn.freebsd.org> <201105271202.04421.jhb@freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On 05/27/11 11:02, John Baldwin wrote: > On Friday, May 27, 2011 10:27:28 am Nathan Whitehorn wrote: >> Author: nwhitehorn >> Date: Fri May 27 14:27:28 2011 >> New Revision: 222356 >> URL: http://svn.freebsd.org/changeset/base/222356 >> >> Log: >> On multi-core, multi-threaded PPC systems, it is important that the threads >> be brought up in the order they are enumerated in the device tree (in >> particular, that thread 0 on each core be brought up first). The SLIST >> through which we loop to start the CPUs has all of its entries added with >> SLIST_INSERT_HEAD(), which means it is in reverse order of enumeration >> and so AP startup would always fail in such situation (causing a machine >> check or RTAS failure). >> >> The best fix is probably to change this from a LIST to a TAILQ, but fix >> this by looping through to add new cpus to the end of the list. > Just make it a STAILQ. That has a STAILQ_INSERT_TAIL() method and doesn't > change the ABI of 'struct pcpu'. It just adds an extra pointer to the head. I did that, and it works well. A patch against HEAD can be found at http://people.freebsd.org/~nwhitehorn/pcpu_tailq.diff. Let me know if it looks OK (or if there are any objections to committing it). -Nathan
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?4DE3FCA0.2000400>