Skip site navigation (1)Skip section navigation (2)
Date:      Sun, 23 Jul 2017 09:47:14 +0300
From:      Konstantin Belousov <kostikbel@gmail.com>
To:        Don Lewis <truckman@FreeBSD.org>
Cc:        pz-freebsd-stable@ziemba.us, freebsd-stable@FreeBSD.org
Subject:   Re: stable/11 r321349 crashing immediately
Message-ID:  <20170723064714.GW1935@kib.kiev.ua>
In-Reply-To: <201707230551.v6N5pgdZ082546@gw.catspoiler.org>
References:  <201707230543.v6N5gwwH082362@gw.catspoiler.org> <201707230551.v6N5pgdZ082546@gw.catspoiler.org>

next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, Jul 22, 2017 at 10:51:42PM -0700, Don Lewis wrote:
> > The stack is aligned to a 4096 (0x1000) boundary.  The first access to a
> > local variable below 0xfffffe085cfa5000 is what triggered the trap.  The
> > other end of the stack must be at 0xfffffe085cfa9000 less a bit. I don't
> > know why the first stack pointer value in the trace is
> > 0xfffffe085cfa8a10. That would seem to indicate that amd64_syscall is
> > using ~1500 bytes of stack space.
> 
> Actually there could be quite a bit of CPU context that gets saved. That
> could be sizeable on amd64.

Yes, the usermode trap frame is located on the kernel stack.  Also, pcb
and usermode FPU save area (FPU == all non-general purpose x86 registers,
including XMM/AVX/AVX512 as implemented by CPU) are on the stack.



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20170723064714.GW1935>