lkml.org 
[lkml]   [2003]   [Aug]   [2]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
Date
From
SubjectRe: [PATCH] [1/2] random: SMP locking
On Sat, Aug 02, 2003 at 08:35:22AM -0400, Zwane Mwaikambo wrote:
> On Sat, 2 Aug 2003, Andrew Morton wrote:
>
> > Cannot perform userspace access while holding a lock - a pagefault could
> > occur, perform IO, schedule away and the same CPU tries to take the same
> > lock via a different process.
>
> Perhaps might_sleep() in *_user, copy_* etc is in order?

I think it's been suggested before, but I threw a patch together for
i386 anyway.

This only checks in the non-__ versions, as those are occassionally
called inside things like kmap_atomic pairs which take a spinlock in
with highmem. It's all conditional on CONFIG_DEBUG_SPINLOCK_SLEEP
(which isn't quite the right name) so there's no overhead for normal
builds.

diff -urN -X dontdiff orig/arch/i386/lib/usercopy.c work/arch/i386/lib/usercopy.c
--- orig/arch/i386/lib/usercopy.c 2003-07-13 22:28:54.000000000 -0500
+++ work/arch/i386/lib/usercopy.c 2003-08-02 13:33:43.000000000 -0500
@@ -149,6 +149,7 @@
unsigned long
clear_user(void __user *to, unsigned long n)
{
+ might_sleep();
if (access_ok(VERIFY_WRITE, to, n))
__do_clear_user(to, n);
return n;
@@ -188,6 +189,8 @@
unsigned long mask = -__addr_ok(s);
unsigned long res, tmp;

+ might_sleep();
+
__asm__ __volatile__(
" testl %0, %0\n"
" jz 3f\n"
diff -urN -X dontdiff orig/include/asm-i386/uaccess.h work/include/asm-i386/uaccess.h
--- orig/include/asm-i386/uaccess.h 2003-07-13 22:30:48.000000000 -0500
+++ work/include/asm-i386/uaccess.h 2003-08-02 13:15:42.000000000 -0500
@@ -260,6 +260,7 @@
({ \
long __pu_err = -EFAULT; \
__typeof__(*(ptr)) *__pu_addr = (ptr); \
+ might_sleep(); \
if (access_ok(VERIFY_WRITE,__pu_addr,size)) \
__put_user_size((x),__pu_addr,(size),__pu_err,-EFAULT); \
__pu_err; \
@@ -469,6 +470,7 @@
static inline unsigned long
copy_to_user(void __user *to, const void *from, unsigned long n)
{
+ might_sleep();
if (access_ok(VERIFY_WRITE, to, n))
n = __copy_to_user(to, from, n);
return n;
@@ -493,6 +495,7 @@
static inline unsigned long
copy_from_user(void *to, const void __user *from, unsigned long n)
{
+ might_sleep();
if (access_ok(VERIFY_READ, from, n))
n = __copy_from_user(to, from, n);
else

--
Matt Mackall : http://www.selenic.com : of or relating to the moon
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-03-22 13:47    [W:0.060 / U:0.360 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site