[PATCH 3/7] powerpc/32: Avoid risk of unrecoverable TLBmiss inside entry_32.S

Path	csiph.com!goblin3!goblin.stu.neva.ru!news.redatomik.org!aioe.org!bofh.it!news.nic.it!robomod
From	Christophe Leroy <christophe.leroy@c-s.fr>
Newsgroups	linux.kernel
Subject	[PATCH 3/7] powerpc/32: Avoid risk of unrecoverable TLBmiss inside entry_32.S
Date	Wed, 12 Jul 2017 12:20:01 +0200
Message-ID	<u2mUp-6WS-13@gated-at.bofh.it> (permalink)
References	<u2mKJ-6Tu-5@gated-at.bofh.it>
Sender	robomod@news.nic.it
List-ID	<linux-kernel.vger.kernel.org>
X-Mailing-List	linux-kernel@vger.kernel.org
Approved	robomod@news.nic.it
Lines	41
Organization	linux.* mail to news gateway
X-Original-Cc	linux-kernel@vger.kernel.org, linuxppc-dev@lists.ozlabs.org
X-Original-Date	Wed, 12 Jul 2017 12:08:49 +0200 (CEST)
X-Original-Message-ID	<e180ea05ab13a9e0099f09c2223aa509a2ba4329.1499633349.git.christophe.leroy@c-s.fr>
X-Original-References	<cover.1499633349.git.christophe.leroy@c-s.fr>
X-Original-Sender	linux-kernel-owner@vger.kernel.org
Xref	csiph.com linux.kernel:1685680

Show key headers only | View raw

By default, the 8xx pins an ITLB on the first 8M of memory in order
to avoid any ITLB miss on kernel code.
However, with some debug functions like DEBUG_PAGEALLOC and
DEBUG_RODATA, pinning TLBs is contradictory.

In order to avoid any ITLB miss in a critical section without pinning
TLBs, we have to ensure that there is no page boundary crossed between
the setup of a new value in SRR0/SRR1 and the associated RFI.

The functions modifying srr0/srr1 are all located in setup_32.S.
They are spread over almost 4kbytes.

The patch forces a 12 bits (4kbytes) alignment for those
functions. This garanties that the functions remain in a
single 4k page.

Signed-off-by: Christophe Leroy <christophe.leroy@c-s.fr>
---
 arch/powerpc/kernel/entry_32.S | 7 +++++++
 1 file changed, 7 insertions(+)

diff --git a/arch/powerpc/kernel/entry_32.S b/arch/powerpc/kernel/entry_32.S
index 8587059ad848..4e9a359ceff6 100644
--- a/arch/powerpc/kernel/entry_32.S
+++ b/arch/powerpc/kernel/entry_32.S
@@ -43,6 +43,13 @@
 #define LOAD_MSR_KERNEL(r, x)	li r,(x)
 #endif
 
+/*
+ * Align to 4k in order to ensure that all functions modyfing srr0/srr1
+ * fit into one page in order to not encounter a TLB miss between the
+ * modification of srr0/srr1 and the associated rfi.
+ */
+	.align	12
+
 #ifdef CONFIG_BOOKE
 	.globl	mcheck_transfer_to_handler
 mcheck_transfer_to_handler:
-- 
2.12.0

Back to linux.kernel | Previous | Next — Previous in thread | Find similar | Unroll thread

Thread

[PATCH 0/7] Prepare 8xx for CONFIG_STRICT_KERNEL_RWX Christophe Leroy <christophe.leroy@c-s.fr> - 2017-07-12 12:10 +0200
  [PATCH 5/7] powerpc/8xx: Do not allow Pinned TLBs with  STRICT_KERNEL_RWX or DEBUG_PAGEALLOC Christophe Leroy <christophe.leroy@c-s.fr> - 2017-07-12 12:10 +0200
  [PATCH 1/7] powerpc/8xx: Ensures RAM mapped with LTLB is seen as  block mapped on 8xx. Christophe Leroy <christophe.leroy@c-s.fr> - 2017-07-12 12:10 +0200
  [PATCH 2/7] powerpc/8xx: Remove macro that checks kernel address Christophe Leroy <christophe.leroy@c-s.fr> - 2017-07-12 12:10 +0200
  [PATCH 6/7] powerpc/8xx: mark init functions with __init Christophe Leroy <christophe.leroy@c-s.fr> - 2017-07-12 12:10 +0200
  [PATCH 3/7] powerpc/32: Avoid risk of unrecoverable TLBmiss inside  entry_32.S Christophe Leroy <christophe.leroy@c-s.fr> - 2017-07-12 12:20 +0200

csiph-web