Date: Sat, 3 Sep 2016 23:34:55 +0300 From: Konstantin Belousov <kostikbel@gmail.com> To: Vladislav Prodan <universite@ukr.net> Cc: freebsd-hackers@freebsd.org, freebsd-acpi@freebsd.org Subject: Re: Freeze during boot process on Intel S1200RP Board Message-ID: <20160903203455.GB83214@kib.kiev.ua> In-Reply-To: <1472932913.386457492.0z8b0wn5@frv35.fwdcdn.com> References: <1472927832.254600230.0917rnnm@frv35.fwdcdn.com> <20160903190950.GA83214@kib.kiev.ua> <1472932913.386457492.0z8b0wn5@frv35.fwdcdn.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, Sep 03, 2016 at 11:08:41PM +0300, Vladislav Prodan wrote: > > > > --- Original message --- > From: "Konstantin Belousov" <kostikbel@gmail.com> > Date: 3 September 2016, 22:10:13 > > > > > On Sat, Sep 03, 2016 at 09:46:30PM +0300, Vladislav Prodan wrote: > > > > > > I encountered a similar problem on Intel S1200RP motherboard. > > > > > > Attached is the output dmesg: > > > http://pastebin.com/raw/ut11rb2T > > Is this from a succesful boot ? I.e., userland did started normally. > > Yes. > > > > > Please use 11.0 and provide me with the verbose dmesg of both succesful > > and hanging boot. > > There is no opportunity and time to install FreeBSD 11.0. > C client approved installation FreeBSD 10.3-STABLE. It will be much less opportunity and much more time required to track it down with debugging done on 10 and not HEAD or 11. Also, there were some significant changes in 11/HEAD WRT LAPIC timer, which might directly affect this machine. Note that you really do not need to overwrite the 10.3 userspace, I only need 11 or HEAD kernel. > > There is a screenshot at the time of freezing: > http://s017.radikal.ru/i412/1609/cf/f24d5d5a61c0.jpg So now I am confused. The system hangs at boot, according to the screenshot. Also, you stated that changing eventtimer choice with sysctl (from LAPIC ?) to HPET helps. Sysctl cannot be issued unless system booted to at least single-user mode. What is the exact scenario for the hang ? Regardless of the answer to the question above, apply the following debugging patch to the kernel and show the screenshot of the hang. Index: sys/amd64/amd64/mp_machdep.c =================================================================== --- sys/amd64/amd64/mp_machdep.c (revision 305361) +++ sys/amd64/amd64/mp_machdep.c (working copy) @@ -1007,6 +1007,7 @@ start_ap(int apic_id) int vector, ms; int cpus; +printf("start_ap apic_id %d\n", apic_id); /* calculate the vector */ vector = (boot_address >> 12) & 0xff;
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20160903203455.GB83214>