From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1753912AbdJaSH7 (ORCPT ); Tue, 31 Oct 2017 14:07:59 -0400 Received: from mga09.intel.com ([134.134.136.24]:36629 "EHLO mga09.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1753428AbdJaSH5 (ORCPT ); Tue, 31 Oct 2017 14:07:57 -0400 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.44,325,1505804400"; d="scan'208";a="1031746819" Subject: [PATCH] x86, mm: make alternatives code do stronger TLB flush To: linux-kernel@vger.kernel.org Cc: linux-mm@kvack.org, Dave Hansen , x86@kernel.org, luto@kernel.org From: Dave Hansen Date: Tue, 31 Oct 2017 11:07:57 -0700 Message-Id: <20171031180757.8B5DA496@viggo.jf.intel.com> Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org From: Dave Hansen local_flush_tlb() does a CR3 write. But, that kind of TLB flush is not guaranteed to invalidate global pages. The entire kernel is mapped with global pages. Also, now that we have PCIDs, local_flush_tlb() will only flush the *current* PCID. It would not flush the entries for all PCIDs. At the moment, this is a moot point because all kernel pages are _PAGE_GLOBAL which do not really *have* a particular PCID. Use the stronger __flush_tlb_all() which does flush global pages. This was found because of a warning I added to __native_flush_tlb() to look for calls to it when PCIDs are enabled. This patch does not fix any bug known to be hit in practice. Signed-off-by: Dave Hansen Cc: x86@kernel.org Cc: Andy Lutomirski --- b/arch/x86/kernel/alternative.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff -puN arch/x86/kernel/alternative.c~x86-mm-text-poke-misses-global-pages arch/x86/kernel/alternative.c --- a/arch/x86/kernel/alternative.c~x86-mm-text-poke-misses-global-pages 2017-10-31 10:28:44.306557256 -0700 +++ b/arch/x86/kernel/alternative.c 2017-10-31 10:28:44.309557393 -0700 @@ -722,7 +722,8 @@ void *text_poke(void *addr, const void * clear_fixmap(FIX_TEXT_POKE0); if (pages[1]) clear_fixmap(FIX_TEXT_POKE1); - local_flush_tlb(); + /* Make sure to flush Global pages: */ + __flush_tlb_all(); sync_core(); /* Could also do a CLFLUSH here to speed up CPU recovery; but that causes hangs on some VIA CPUs. */ _