From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1752394AbcCKLTg (ORCPT ); Fri, 11 Mar 2016 06:19:36 -0500 Received: from mail-wm0-f49.google.com ([74.125.82.49]:34287 "EHLO mail-wm0-f49.google.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751534AbcCKLT1 (ORCPT ); Fri, 11 Mar 2016 06:19:27 -0500 From: Matt Fleming To: Ingo Molnar , "H . Peter Anvin" , Thomas Gleixner Cc: Ard Biesheuvel , Matt Fleming , linux-kernel@vger.kernel.org, linux-efi@vger.kernel.org, Alexis Murzeau , Ben Hutchings , Borislav Petkov , Maarten Lankhorst , Matthew Garrett , Raphael Hertzog , Roger Shimizu Subject: [GIT PULL] EFI urgent fix for v4.6 queue Date: Fri, 11 Mar 2016 11:19:22 +0000 Message-Id: <1457695163-29632-1-git-send-email-matt@codeblueprint.co.uk> X-Mailer: git-send-email 2.6.2 Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Folks, the following patch fixes a bug that is triggered with the new EFI page table code sitting in tip/efi/core and queued up for the merge window. It turns out that we were relying on the kernel's mappings for those regions that are marked E820_RESERVED, etc in the e820 map. The specific example in the bug report was page zero which was always "special" on legacy BIOS. On EFI, legitimate code/data can be placed there and so if it's part of the EFI memory map, we must make sure we map it into the EFI page tables. EFI runtime drivers may need to access it. The following changes since commit 2ad510dc372c2caac9aada9ff6dd10e787616e1d: x86/efi: Only map kernel text for EFI mixed mode (2016-02-22 08:26:28 +0100) are available in the git repository at: git://git.kernel.org/pub/scm/linux/kernel/git/mfleming/efi.git tags/efi-urgent for you to fetch changes up to c7af05d993ef60102eea6e0056de7eb5f6d4e52d: x86/efi: Always map boot service regions into new EFI page tables (2016-03-11 11:08:36 +0000) ---------------------------------------------------------------- * Fix a bug triggered with the new separate EFI page table code where EFI boot services regions that are marked E820_RESERVED are not mapped into the page tables, leading to an oops during SetVirtualAddressMap() - Matt Fleming ---------------------------------------------------------------- Matt Fleming (1): x86/efi: Always map boot service regions into new EFI page tables arch/x86/platform/efi/quirks.c | 79 +++++++++++++++++++++++++++++++++--------- 1 file changed, 62 insertions(+), 17 deletions(-)