Date: Wed, 27 Jun 2007 01:01:48 -0400 From: John Baldwin <jhb@freebsd.org> To: freebsd-amd64@freebsd.org Cc: Adam Soderlund <adamsod@nc.rr.com> Subject: Re: amd64 SMP AP init corrupting the BDA? (newbie) Message-ID: <200706270101.49313.jhb@freebsd.org> In-Reply-To: <200605120829.k4C8T1gp025713@ms-smtp-02.southeast.rr.com> References: <200605120829.k4C8T1gp025713@ms-smtp-02.southeast.rr.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Friday 12 May 2006 04:28:59 am Adam Soderlund wrote: > Hi All, > > Hope this is the right place for questions like this. I've got > something bad going on when booting 6.1release or 7-current amd64 SMP, the > bios data area on my machine is getting whacked (killing usb legacy, which > in turn causes a hang when kernel tries to take over the host > controller(s)). I've debugged it a little bit and here is what I see. > when setting up to launch the APs start_all_aps() calls into pmap_kenter() > which is writing a pte to 0x000004e0 (physical). > > > > 0x04e0 in the BDA prior to pmap_kenter(va, boot_address) > > 0x0000002B > > 0x04e0 in the BDA after pmap_kenter(va, boot_address) > > 0x03C10900 (boot_address on the way in 0x9C000) The pmap_kenter() should not be changing that address. The page tables actually are located after the kernel itself (just after kernend) in memory, so pmap_kenter() isn't going to be writing to a PTE at an address that low. -- John Baldwin
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200706270101.49313.jhb>