Date: Sat, 4 Jan 2014 21:52:11 +0200 From: Guy Yur <guyyur@gmail.com> To: freebsd-net@freebsd.org, freebsd-arm@freebsd.org Subject: Re: 10.0-RC1, armv6: "pfctl -s state" crashes on BeagleBone Black due to unaligned access Message-ID: <CAC67Hz_9ts60gUE9xzB9nacO4F7kNznJ26LW6mPLm-8pT-f20w@mail.gmail.com> In-Reply-To: <E4255780-6E58-435D-BDE6-491915B642CF@bsdimp.com> References: <CAC67Hz_QXcHHSFOLLgUGqLWRQpzhRRv_b%2BWGMMQsfk-VQp74RA@mail.gmail.com> <E4255780-6E58-435D-BDE6-491915B642CF@bsdimp.com>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, Jan 4, 2014 at 6:50 PM, Warner Losh <imp@bsdimp.com> wrote: > I think this was changed in later RC versions. > > Warner > Do you mean r259308 in 10-STABLE? I compiled a new kernel with the change applied and still get SIGBUS on unaligned access. >From my reading of the ARM manual for Cortex-A it looks like setting CPU_CONTROL_AFLT_ENABLE to 1 means you want to get a trap on unaligned access. "1 = Strict alignment fault checking enabled." I see that dab_align delivers a SIGBUS on user-mode alignment fault. > On Jan 4, 2014, at 6:06 AM, Guy Yur wrote: > >> Hi, >> >> I am running 10.0-RC1 arm.armv6 on the BeagleBone Black. >> The "pfctl -s state" command is crashing when trying to print the >> second entry. >> >> struct pfsync_state has a size that is not divisiable by 4 or 8 leading to the >> second entry in the returned state array not being aligned and pfctl >> core dumps on Bus error when trying to access a uint32_t field. >> Regards, Guy
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?CAC67Hz_9ts60gUE9xzB9nacO4F7kNznJ26LW6mPLm-8pT-f20w>