From mboxrd@z Thu Jan 1 00:00:00 1970 From: Toshi Kani Subject: Re: [BUG] can't boot up: unable to handle kernel paging request at ffffffffff340003 Date: Wed, 12 Feb 2014 09:15:03 -0700 Message-ID: <1392221703.5612.164.camel@misato.fc.hp.com> References: <8738jr5t3e.fsf@redhat.com> <20140210175246.GB2887@console-pimps.org> <87mwhy8iaj.fsf@redhat.com> <20140211131640.GD2887@console-pimps.org> <87lhxh69u0.fsf@redhat.com> <20140211141846.GA1725@console-pimps.org> Mime-Version: 1.0 Content-Type: text/plain; charset="UTF-8" Content-Transfer-Encoding: 7bit Return-path: In-Reply-To: <20140211141846.GA1725-HNK1S37rvNbeXh+fF434Mdi2O/JbrIOy@public.gmane.org> Sender: linux-efi-owner-u79uwXL29TY76Z2rM5mHXA@public.gmane.org To: Matt Fleming Cc: Madper Xie , Matt Fleming , "linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org" , "linux-efi-u79uwXL29TY76Z2rM5mHXA@public.gmane.org" , Matthew Garrett , "Rafael J. Wysocki" , Josh Triplett List-Id: linux-efi@vger.kernel.org On Tue, 2014-02-11 at 14:18 +0000, Matt Fleming wrote: > (Entering the realm of ACPI, pulling in more folks) > > On Tue, 11 Feb, at 09:47:57PM, Madper Xie wrote: > > Thanks for your reply. I attached the all the dmesg outputs for you. > > [ 0.000000] ACPI: BGRT 00000000d9bb5100 00003C (v00 HPQOEM SLIC-BPC 01072009 AMI 00010013) > > Your BGRT table has an extra 4 bytes for some reason, (0x3C should be > 0x38) and doesn't match the defintion in include/acpi/actbl3.h, > > > struct acpi_table_bgrt { > struct acpi_table_header header; /* Common ACPI table header */ > u16 version; > u8 status; > u8 image_type; > u64 image_address; > u32 image_offset_x; > u32 image_offset_y; > }; > > > which explains why when we dereference ->image_address we don't get a > sensible value. > > Guys, what do you make of this? Firmware bug? Full thread here, > > http://article.gmane.org/gmane.linux.kernel.efi/3499 > Hi Matt, Yes, I agree that the table size should be 0x38. However, ACPI spec states that bit0 of status indicates if the boot image graphic is valid. This bit is set to 0 (invalid) on the system. Can you check this bit and return when invalid? Thanks, -Toshi