lkml.org 
[lkml]   [2011]   [Dec]   [19]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
Date
From
Subject[PATCH v2] oprofile: Fix uninitialized memory access when writing to
On 19.12.11 15:17:03, Robert Richter wrote:
> On 19.12.11 15:00:49, Ingo Molnar wrote:
>
> > + if (!count)
> > + return 0;
> > +
> > retval = oprofilefs_ulong_from_user(&value, buf, count);
> > if (retval)
> > return retval;
> >
> > See the ugly and fragile pattern?
> >
> > This should *really* be solved via the
> > oprofilefs_ulong_from_user() helper function, not by sprinkling
> > the !count checks in half a dozen places ...
>
> Hmm, I thought there was no way to leave the code path with count == 0
> and retval. But thinking about it it would be possible with returning
> count or errors < 0. Will improve the patch.

Ingo, see my updated version below. I am fine with appling it directly
on tip/perf/urgent.

Thanks,

-Robert

--


From f07214bc05560f63da8404d31aa83034a16c6229 Mon Sep 17 00:00:00 2001
From: Robert Richter <robert.richter@amd.com>
Date: Fri, 16 Dec 2011 15:45:31 +0100
Subject: [PATCH] oprofile: Fix uninitialized memory access when writing to
oprofilefs

If oprofilefs_ulong_from_user() is called with count equals zero, *val
remains unchanged. Depending on the implementation it might be
uninitialized.

Change oprofilefs_ulong_from_user()'s interface to return count on
success. Thus, we are able to return early if count equals zero which
avoids using *val uninitialized. Fixing all users of oprofilefs_ulong_
from_user().

This follows write syscall implementation when count is zero: "If
count is zero ... [and if] no errors are detected, 0 will be returned
without causing any other effect." (man 2 write)

Reported-By: Mike Waychison <mikew@google.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: stable@vger.kernel.org
Signed-off-by: Robert Richter <robert.richter@amd.com>
---
arch/s390/oprofile/init.c | 2 +-
drivers/oprofile/oprofile_files.c | 7 ++++---
drivers/oprofile/oprofilefs.c | 11 +++++++++--
3 files changed, 14 insertions(+), 6 deletions(-)

diff --git a/arch/s390/oprofile/init.c b/arch/s390/oprofile/init.c
index 6efc18b..bd58b72 100644
--- a/arch/s390/oprofile/init.c
+++ b/arch/s390/oprofile/init.c
@@ -88,7 +88,7 @@ static ssize_t hwsampler_write(struct file *file, char const __user *buf,
return -EINVAL;

retval = oprofilefs_ulong_from_user(&val, buf, count);
- if (retval)
+ if (retval <= 0)
return retval;

if (oprofile_started)
diff --git a/drivers/oprofile/oprofile_files.c b/drivers/oprofile/oprofile_files.c
index 89f6345..84a208d 100644
--- a/drivers/oprofile/oprofile_files.c
+++ b/drivers/oprofile/oprofile_files.c
@@ -45,7 +45,7 @@ static ssize_t timeout_write(struct file *file, char const __user *buf,
return -EINVAL;

retval = oprofilefs_ulong_from_user(&val, buf, count);
- if (retval)
+ if (retval <= 0)
return retval;

retval = oprofile_set_timeout(val);
@@ -84,7 +84,7 @@ static ssize_t depth_write(struct file *file, char const __user *buf, size_t cou
return -EINVAL;

retval = oprofilefs_ulong_from_user(&val, buf, count);
- if (retval)
+ if (retval <= 0)
return retval;

retval = oprofile_set_ulong(&oprofile_backtrace_depth, val);
@@ -141,9 +141,10 @@ static ssize_t enable_write(struct file *file, char const __user *buf, size_t co
return -EINVAL;

retval = oprofilefs_ulong_from_user(&val, buf, count);
- if (retval)
+ if (retval <= 0)
return retval;

+ retval = 0;
if (val)
retval = oprofile_start();
else
diff --git a/drivers/oprofile/oprofilefs.c b/drivers/oprofile/oprofilefs.c
index d0de6cc..2f0aa0f 100644
--- a/drivers/oprofile/oprofilefs.c
+++ b/drivers/oprofile/oprofilefs.c
@@ -60,6 +60,13 @@ ssize_t oprofilefs_ulong_to_user(unsigned long val, char __user *buf, size_t cou
}


+/*
+ * Note: If oprofilefs_ulong_from_user() returns 0, then *val remains
+ * unchanged and might be uninitialized. This follows write syscall
+ * implementation when count is zero: "If count is zero ... [and if]
+ * no errors are detected, 0 will be returned without causing any
+ * other effect." (man 2 write)
+ */
int oprofilefs_ulong_from_user(unsigned long *val, char const __user *buf, size_t count)
{
char tmpbuf[TMPBUFSIZE];
@@ -79,7 +86,7 @@ int oprofilefs_ulong_from_user(unsigned long *val, char const __user *buf, size_
raw_spin_lock_irqsave(&oprofilefs_lock, flags);
*val = simple_strtoul(tmpbuf, NULL, 0);
raw_spin_unlock_irqrestore(&oprofilefs_lock, flags);
- return 0;
+ return count;
}


@@ -99,7 +106,7 @@ static ssize_t ulong_write_file(struct file *file, char const __user *buf, size_
return -EINVAL;

retval = oprofilefs_ulong_from_user(&value, buf, count);
- if (retval)
+ if (retval <= 0)
return retval;

retval = oprofile_set_ulong(file->private_data, value);
--
1.7.7



--
Advanced Micro Devices, Inc.
Operating System Research Center



\
 
 \ /
  Last update: 2011-12-19 16:41    [W:0.064 / U:0.400 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site