* [wrecked] x86-add-failure-injection-to-get-put-clear_user.patch removed from -mm tree
@ 2020-10-13 18:20 akpm
[not found] ` <CAG_fn=U+zDi-Z7FmkhUWsMjVALvG-jkOgvmjiz6_HfjzLcF4UQ@mail.gmail.com>
0 siblings, 1 reply; 2+ messages in thread
From: akpm @ 2020-10-13 18:20 UTC (permalink / raw)
To: akinobu.mita, alinde, andreyknvl, arnd, bp, corbet, dvyukov,
elver, glider, hch, hpa, mingo, mm-commits, peterz, tglx, viro
The patch titled
Subject: x86: add failure injection to get/put/clear_user
has been removed from the -mm tree. Its filename was
x86-add-failure-injection-to-get-put-clear_user.patch
This patch was dropped because other changes were merged, which wrecked this patch
------------------------------------------------------
From: Albert van der Linde <alinde@google.com>
Subject: x86: add failure injection to get/put/clear_user
To test fault-tolerance of user memory acceses in x86, add support for
fault injection.
Make both put_user() and get_user() fail with -EFAULT, and clear_user()
fail by not clearing any bytes.
Link: http://lkml.kernel.org/r/20200831171733.955393-4-alinde@google.com
Signed-off-by: Albert van der Linde <alinde@google.com>
Reviewed-by: Akinobu Mita <akinobu.mita@gmail.com>
Reviewed-by: Alexander Potapenko <glider@google.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Andrey Konovalov <andreyknvl@google.com>
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Marco Elver <elver@google.com>
Cc: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Christoph Hellwig <hch@lst.de>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---
arch/x86/include/asm/uaccess.h | 68 +++++++++++++++++--------------
arch/x86/lib/usercopy_64.c | 3 +
2 files changed, 42 insertions(+), 29 deletions(-)
--- a/arch/x86/include/asm/uaccess.h~x86-add-failure-injection-to-get-put-clear_user
+++ a/arch/x86/include/asm/uaccess.h
@@ -5,6 +5,7 @@
* User space memory access functions
*/
#include <linux/compiler.h>
+#include <linux/fault-inject-usercopy.h>
#include <linux/kasan-checks.h>
#include <linux/string.h>
#include <asm/asm.h>
@@ -175,11 +176,16 @@ extern int __get_user_bad(void);
register __inttype(*(ptr)) __val_gu asm("%"_ASM_DX); \
__chk_user_ptr(ptr); \
might_fault(); \
- asm volatile("call __get_user_%P4" \
- : "=a" (__ret_gu), "=r" (__val_gu), \
+ if (should_fail_usercopy()) { \
+ (x) = 0; \
+ __ret_gu = -EFAULT; \
+ } else { \
+ asm volatile("call __get_user_%P4" \
+ : "=a" (__ret_gu), "=r" (__val_gu), \
ASM_CALL_CONSTRAINT \
- : "0" (ptr), "i" (sizeof(*(ptr)))); \
- (x) = (__force __typeof__(*(ptr))) __val_gu; \
+ : "0" (ptr), "i" (sizeof(*(ptr)))); \
+ (x) = (__force __typeof__(*(ptr))) __val_gu; \
+ } \
__builtin_expect(__ret_gu, 0); \
})
@@ -236,31 +242,35 @@ extern void __put_user_8(void);
*
* Return: zero on success, or -EFAULT on error.
*/
-#define put_user(x, ptr) \
-({ \
- int __ret_pu; \
- __typeof__(*(ptr)) __pu_val; \
- __chk_user_ptr(ptr); \
- might_fault(); \
- __pu_val = x; \
- switch (sizeof(*(ptr))) { \
- case 1: \
- __put_user_x(1, __pu_val, ptr, __ret_pu); \
- break; \
- case 2: \
- __put_user_x(2, __pu_val, ptr, __ret_pu); \
- break; \
- case 4: \
- __put_user_x(4, __pu_val, ptr, __ret_pu); \
- break; \
- case 8: \
- __put_user_x8(__pu_val, ptr, __ret_pu); \
- break; \
- default: \
- __put_user_x(X, __pu_val, ptr, __ret_pu); \
- break; \
- } \
- __builtin_expect(__ret_pu, 0); \
+#define put_user(x, ptr) \
+({ \
+ int __ret_pu; \
+ __typeof__(*(ptr)) __pu_val; \
+ __chk_user_ptr(ptr); \
+ might_fault(); \
+ __pu_val = x; \
+ if (should_fail_usercopy()) { \
+ __ret_pu = -EFAULT; \
+ } else { \
+ switch (sizeof(*(ptr))) { \
+ case 1: \
+ __put_user_x(1, __pu_val, ptr, __ret_pu); \
+ break; \
+ case 2: \
+ __put_user_x(2, __pu_val, ptr, __ret_pu); \
+ break; \
+ case 4: \
+ __put_user_x(4, __pu_val, ptr, __ret_pu); \
+ break; \
+ case 8: \
+ __put_user_x8(__pu_val, ptr, __ret_pu); \
+ break; \
+ default: \
+ __put_user_x(X, __pu_val, ptr, __ret_pu); \
+ break; \
+ } \
+ } \
+ __builtin_expect(__ret_pu, 0); \
})
#define __put_user_size(x, ptr, size, label) \
--- a/arch/x86/lib/usercopy_64.c~x86-add-failure-injection-to-get-put-clear_user
+++ a/arch/x86/lib/usercopy_64.c
@@ -7,6 +7,7 @@
* Copyright 2002 Andi Kleen <ak@suse.de>
*/
#include <linux/export.h>
+#include <linux/fault-inject-usercopy.h>
#include <linux/uaccess.h>
#include <linux/highmem.h>
@@ -50,6 +51,8 @@ EXPORT_SYMBOL(__clear_user);
unsigned long clear_user(void __user *to, unsigned long n)
{
+ if (should_fail_usercopy())
+ return n;
if (access_ok(to, n))
return __clear_user(to, n);
return n;
_
Patches currently in -mm which might be from alinde@google.com are
lib-include-linux-add-usercopy-failure-capability.patch
lib-uaccess-add-failure-injection-to-usercopy-functions.patch
^ permalink raw reply [flat|nested] 2+ messages in thread
* Re: [wrecked] x86-add-failure-injection-to-get-put-clear_user.patch removed from -mm tree
[not found] ` <CAG_fn=U+zDi-Z7FmkhUWsMjVALvG-jkOgvmjiz6_HfjzLcF4UQ@mail.gmail.com>
@ 2020-10-14 21:00 ` Andrew Morton
0 siblings, 0 replies; 2+ messages in thread
From: Andrew Morton @ 2020-10-14 21:00 UTC (permalink / raw)
To: Alexander Potapenko
Cc: Akinobu Mita, Albert van der Linde, Andrey Konovalov,
Arnd Bergmann, Borislav Petkov, Jonathan Corbet, Dmitriy Vyukov,
Marco Elver, Christoph Hellwig, H. Peter Anvin, Ingo Molnar,
mm-commits, Peter Zijlstra, Thomas Gleixner, Al Viro
On Wed, 14 Oct 2020 16:00:56 +0200 Alexander Potapenko <glider@google.com> wrote:
> Hi Andrew,
>
> On Tue, Oct 13, 2020 at 8:20 PM <akpm@linux-foundation.org> wrote:
> >
> >
> > The patch titled
> > Subject: x86: add failure injection to get/put/clear_user
> > has been removed from the -mm tree. Its filename was
> > x86-add-failure-injection-to-get-put-clear_user.patch
> >
> > This patch was dropped because other changes were merged, which wrecked this patch
>
> On top of which tree should we rebase the patches to see/resolve these
> conflicts?
> linux-next/akpm and linux-mm/master seem to still contain all three
> patches in question.
The problem is ec6347bb43395 ("x86, powerpc: Rename memcpy_mcsafe() to
copy_mc_to_{user, kernel}()"), which is in Linus's tree. It seems to
have got slammed in at the last minute (with a cc:stable so I guess
it's important). Please rebase and retest against current -linus and
resend?
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2020-10-14 21:01 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-10-13 18:20 [wrecked] x86-add-failure-injection-to-get-put-clear_user.patch removed from -mm tree akpm
[not found] ` <CAG_fn=U+zDi-Z7FmkhUWsMjVALvG-jkOgvmjiz6_HfjzLcF4UQ@mail.gmail.com>
2020-10-14 21:00 ` Andrew Morton
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).