Patch "powerpc: Avoid code patching freed init sections" has been added to the 4.4-stable tree

gregkh at linuxfoundation.org gregkh at linuxfoundation.org
Mon Apr 29 19:51:25 AEST 2019


This is a note to let you know that I've just added the patch titled

    powerpc: Avoid code patching freed init sections

to the 4.4-stable tree which can be found at:
    http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary

The filename of the patch is:
     powerpc-avoid-code-patching-freed-init-sections.patch
and it can be found in the queue-4.4 subdirectory.

If you, or anyone else, feels it should not be added to the stable tree,
please let <stable at vger.kernel.org> know about it.


>From foo at baz Mon 29 Apr 2019 11:38:37 AM CEST
From: Michael Ellerman <mpe at ellerman.id.au>
Date: Mon, 22 Apr 2019 00:20:29 +1000
Subject: powerpc: Avoid code patching freed init sections
To: stable at vger.kernel.org, gregkh at linuxfoundation.org
Cc: linuxppc-dev at ozlabs.org, diana.craciun at nxp.com, msuchanek at suse.de, npiggin at gmail.com, christophe.leroy at c-s.fr
Message-ID: <20190421142037.21881-45-mpe at ellerman.id.au>

From: Michael Neuling <mikey at neuling.org>

commit 51c3c62b58b357e8d35e4cc32f7b4ec907426fe3 upstream.

This stops us from doing code patching in init sections after they've
been freed.

In this chain:
  kvm_guest_init() ->
    kvm_use_magic_page() ->
      fault_in_pages_readable() ->
	 __get_user() ->
	   __get_user_nocheck() ->
	     barrier_nospec();

We have a code patching location at barrier_nospec() and
kvm_guest_init() is an init function. This whole chain gets inlined,
so when we free the init section (hence kvm_guest_init()), this code
goes away and hence should no longer be patched.

We seen this as userspace memory corruption when using a memory
checker while doing partition migration testing on powervm (this
starts the code patching post migration via
/sys/kernel/mobility/migration). In theory, it could also happen when
using /sys/kernel/debug/powerpc/barrier_nospec.

Signed-off-by: Michael Neuling <mikey at neuling.org>
Reviewed-by: Nicholas Piggin <npiggin at gmail.com>
Reviewed-by: Christophe Leroy <christophe.leroy at c-s.fr>
Signed-off-by: Michael Ellerman <mpe at ellerman.id.au>
Signed-off-by: Greg Kroah-Hartman <gregkh at linuxfoundation.org>
---
 arch/powerpc/include/asm/setup.h |    1 +
 arch/powerpc/lib/code-patching.c |   13 +++++++++++++
 arch/powerpc/mm/mem.c            |    2 ++
 3 files changed, 16 insertions(+)

--- a/arch/powerpc/include/asm/setup.h
+++ b/arch/powerpc/include/asm/setup.h
@@ -8,6 +8,7 @@ extern void ppc_printk_progress(char *s,
 
 extern unsigned int rtas_data;
 extern unsigned long long memory_limit;
+extern bool init_mem_is_free;
 extern unsigned long klimit;
 extern void *zalloc_maybe_bootmem(size_t size, gfp_t mask);
 
--- a/arch/powerpc/lib/code-patching.c
+++ b/arch/powerpc/lib/code-patching.c
@@ -14,12 +14,25 @@
 #include <asm/page.h>
 #include <asm/code-patching.h>
 #include <asm/uaccess.h>
+#include <asm/setup.h>
+#include <asm/sections.h>
 
 
+static inline bool is_init(unsigned int *addr)
+{
+	return addr >= (unsigned int *)__init_begin && addr < (unsigned int *)__init_end;
+}
+
 int patch_instruction(unsigned int *addr, unsigned int instr)
 {
 	int err;
 
+	/* Make sure we aren't patching a freed init section */
+	if (init_mem_is_free && is_init(addr)) {
+		pr_debug("Skipping init section patching addr: 0x%px\n", addr);
+		return 0;
+	}
+
 	__put_user_size(instr, addr, 4, err);
 	if (err)
 		return err;
--- a/arch/powerpc/mm/mem.c
+++ b/arch/powerpc/mm/mem.c
@@ -62,6 +62,7 @@
 #endif
 
 unsigned long long memory_limit;
+bool init_mem_is_free;
 
 #ifdef CONFIG_HIGHMEM
 pte_t *kmap_pte;
@@ -381,6 +382,7 @@ void __init mem_init(void)
 void free_initmem(void)
 {
 	ppc_md.progress = ppc_printk_progress;
+	init_mem_is_free = true;
 	free_initmem_default(POISON_FREE_INITMEM);
 }
 


Patches currently in stable-queue which might be from mpe at ellerman.id.au are

queue-4.4/powerpc-64s-add-support-for-a-store-forwarding-barrier-at-kernel-entry-exit.patch
queue-4.4/powerpc-64-make-stf-barrier-ppc_book3s_64-specific.patch
queue-4.4/powerpc-pseries-set-or-clear-security-feature-flags.patch
queue-4.4/powerpc-fsl-fix-spectre_v2-mitigations-reporting.patch
queue-4.4/powerpc-64s-patch-barrier_nospec-in-modules.patch
queue-4.4/powerpc-pseries-support-firmware-disable-of-rfi-flush.patch
queue-4.4/powerpc-rfi-flush-call-setup_rfi_flush-after-lpm-migration.patch
queue-4.4/powerpc-pseries-query-hypervisor-for-count-cache-flush-settings.patch
queue-4.4/powerpc-powernv-set-or-clear-security-feature-flags.patch
queue-4.4/powerpc-64s-add-support-for-software-count-cache-flush.patch
queue-4.4/powerpc64s-show-ori31-availability-in-spectre_v1-sysfs-file-not-v2.patch
queue-4.4/powerpc-fsl-flush-the-branch-predictor-at-each-kernel-entry-64bit.patch
queue-4.4/powerpc-fsl-update-spectre-v2-reporting.patch
queue-4.4/powerpc-64s-wire-up-cpu_show_spectre_v2.patch
queue-4.4/powerpc-64-make-meltdown-reporting-book3s-64-specific.patch
queue-4.4/powerpc-rfi-flush-make-it-possible-to-call-setup_rfi_flush-again.patch
queue-4.4/powerpc-64s-add-support-for-ori-barrier_nospec-patching.patch
queue-4.4/powerpc-use-barrier_nospec-in-copy_from_user.patch
queue-4.4/powerpc-64s-fix-section-mismatch-warnings-from-setup_rfi_flush.patch
queue-4.4/powerpc-avoid-code-patching-freed-init-sections.patch
queue-4.4/powerpc-fsl-add-macro-to-flush-the-branch-predictor.patch
queue-4.4/powerpc-xmon-add-rfi-flush-related-fields-to-paca-dump.patch
queue-4.4/powerpc-fsl-add-barrier_nospec-implementation-for-nxp-powerpc-book3e.patch
queue-4.4/powerpc-security-fix-spectre_v2-reporting.patch
queue-4.4/powerpc-add-security-feature-flags-for-spectre-meltdown.patch
queue-4.4/powerpc-powernv-use-the-security-flags-in-pnv_setup_rfi_flush.patch
queue-4.4/powerpc-64-disable-the-speculation-barrier-from-the-command-line.patch
queue-4.4/powerpc-fsl-fix-the-flush-of-branch-predictor.patch
queue-4.4/powerpc-pseries-use-the-security-flags-in-pseries_setup_rfi_flush.patch
queue-4.4/powerpc-64-add-config_ppc_barrier_nospec.patch
queue-4.4/powerpc-64s-move-cpu_show_meltdown.patch
queue-4.4/powerpc-64-use-barrier_nospec-in-syscall-entry.patch
queue-4.4/powerpc-fsl-add-nospectre_v2-command-line-argument.patch
queue-4.4/powerpc-64s-add-new-security-feature-flags-for-count-cache-flush.patch
queue-4.4/powerpc-fsl-add-infrastructure-to-fixup-branch-predictor-flush.patch
queue-4.4/powerpc-rfi-flush-differentiate-enabled-and-patched-flush-types.patch
queue-4.4/powerpc-64s-enhance-the-information-in-cpu_show_spectre_v1.patch
queue-4.4/powerpc-64-call-setup_barrier_nospec-from-setup_arch.patch
queue-4.4/powerpc-rfi-flush-always-enable-fallback-flush-on-pseries.patch
queue-4.4/powerpc-64s-improve-rfi-l1-d-cache-flush-fallback.patch
queue-4.4/powerpc-asm-add-a-patch_site-macro-helpers-for-patching-instructions.patch
queue-4.4/powerpc-pseries-add-new-h_get_cpu_characteristics-flags.patch
queue-4.4/powerpc-64s-enable-barrier_nospec-based-on-firmware-settings.patch
queue-4.4/powerpc-powernv-support-firmware-disable-of-rfi-flush.patch
queue-4.4/powerpc-rfi-flush-move-the-logic-to-avoid-a-redo-into-the-debugfs-code.patch
queue-4.4/powerpc-powernv-query-firmware-for-count-cache-flush-settings.patch
queue-4.4/powerpc-64s-wire-up-cpu_show_spectre_v1.patch
queue-4.4/powerpc-64s-add-barrier_nospec.patch
queue-4.4/powerpc-64s-enhance-the-information-in-cpu_show_meltdown.patch
queue-4.4/powerpc-move-default-security-feature-flags.patch
queue-4.4/powerpc-pseries-fix-clearing-of-security-feature-flags.patch
queue-4.4/powerpc-pseries-restore-default-security-feature-flags-on-setup.patch


More information about the Linuxppc-dev mailing list