* [patch 3/5] vfs: introduce FMODE_NEG_OFFSET for allowing negative f_pos
@ 2010-05-24 19:24 akpm
0 siblings, 0 replies; 6+ messages in thread
From: akpm @ 2010-05-24 19:24 UTC (permalink / raw)
To: viro
Cc: linux-fsdevel, akpm, kamezawa.hiroyu, fengguang.wu,
heiko.carstens, viro
From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Now, rw_verify_area() checsk f_pos is negative or not. And if negative,
returns -EINVAL.
But, some special files as /dev/(k)mem and /proc/<pid>/mem etc.. has
negative offsets. And we can't do any access via read/write to the
file(device).
So introduce FMODE_NEG_OFFSET to allow negative file offsets.
Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Al Viro <viro@ZenIV.linux.org.uk>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---
drivers/char/mem.c | 4 ++++
fs/proc/base.c | 2 ++
fs/read_write.c | 21 +++++++++++++++++++--
include/linux/fs.h | 3 +++
4 files changed, 28 insertions(+), 2 deletions(-)
diff -puN drivers/char/mem.c~vfs-introduce-fmode_neg_offset-for-allowing-negative-f_pos drivers/char/mem.c
--- a/drivers/char/mem.c~vfs-introduce-fmode_neg_offset-for-allowing-negative-f_pos
+++ a/drivers/char/mem.c
@@ -873,6 +873,10 @@ static int memory_open(struct inode *ino
if (dev->dev_info)
filp->f_mapping->backing_dev_info = dev->dev_info;
+ /* Is /dev/mem or /dev/kmem ? */
+ if (dev->dev_info == &directly_mappable_cdev_bdi)
+ filp->f_mode |= FMODE_NEG_OFFSET;
+
if (dev->fops->open)
return dev->fops->open(inode, filp);
diff -puN fs/proc/base.c~vfs-introduce-fmode_neg_offset-for-allowing-negative-f_pos fs/proc/base.c
--- a/fs/proc/base.c~vfs-introduce-fmode_neg_offset-for-allowing-negative-f_pos
+++ a/fs/proc/base.c
@@ -775,6 +775,8 @@ static const struct file_operations proc
static int mem_open(struct inode* inode, struct file* file)
{
file->private_data = (void*)((long)current->self_exec_id);
+ /* OK to pass negative loff_t, we can catch out-of-range */
+ file->f_mode |= FMODE_NEG_OFFSET;
return 0;
}
diff -puN fs/read_write.c~vfs-introduce-fmode_neg_offset-for-allowing-negative-f_pos fs/read_write.c
--- a/fs/read_write.c~vfs-introduce-fmode_neg_offset-for-allowing-negative-f_pos
+++ a/fs/read_write.c
@@ -205,6 +205,20 @@ bad:
}
#endif
+static int
+__negative_fpos_check(struct file *file, loff_t pos, size_t count)
+{
+ /*
+ * pos or pos+count is negative here, check overflow.
+ * too big "count" will be caught in rw_verify_area().
+ */
+ if ((pos < 0) && (pos + count < pos))
+ return -EOVERFLOW;
+ if (file->f_mode & FMODE_NEG_OFFSET)
+ return 0;
+ return -EINVAL;
+}
+
/*
* rw_verify_area doesn't like huge counts. We limit
* them to something that fits in "int" so that others
@@ -222,8 +236,11 @@ int rw_verify_area(int read_write, struc
if (unlikely((ssize_t) count < 0))
return retval;
pos = *ppos;
- if (unlikely((pos < 0) || (loff_t) (pos + count) < 0))
- return retval;
+ if (unlikely((pos < 0) || (loff_t) (pos + count) < 0)) {
+ retval = __negative_fpos_check(file, pos, count);
+ if (retval)
+ return retval;
+ }
if (unlikely(inode->i_flock && mandatory_lock(inode))) {
retval = locks_mandatory_area(
diff -puN include/linux/fs.h~vfs-introduce-fmode_neg_offset-for-allowing-negative-f_pos include/linux/fs.h
--- a/include/linux/fs.h~vfs-introduce-fmode_neg_offset-for-allowing-negative-f_pos
+++ a/include/linux/fs.h
@@ -90,6 +90,9 @@ struct inodes_stat_t {
/* Expect random access pattern */
#define FMODE_RANDOM ((__force fmode_t)0x1000)
+/* File is huge (eg. /dev/kmem): treat loff_t as unsigned */
+#define FMODE_NEG_OFFSET ((__force fmode_t)0x2000)
+
/* File was opened by fanotify and shouldn't generate fanotify events */
/* File was opened by fanotify and shouldn't generate fanotify events */
#define FMODE_NONOTIFY ((__force fmode_t)0x1000000)
_
^ permalink raw reply [flat|nested] 6+ messages in thread
* [patch 3/5] vfs: introduce FMODE_NEG_OFFSET for allowing negative f_pos
@ 2010-07-20 22:29 akpm
2010-07-21 2:54 ` Nick Piggin
0 siblings, 1 reply; 6+ messages in thread
From: akpm @ 2010-07-20 22:29 UTC (permalink / raw)
To: viro
Cc: linux-fsdevel, akpm, kamezawa.hiroyu, fengguang.wu,
heiko.carstens, viro
From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Now, rw_verify_area() checsk f_pos is negative or not. And if negative,
returns -EINVAL.
But, some special files as /dev/(k)mem and /proc/<pid>/mem etc.. has
negative offsets. And we can't do any access via read/write to the
file(device).
So introduce FMODE_NEG_OFFSET to allow negative file offsets.
Signed-off-by: Wu Fengguang <fengguang.wu@intel.com>
Signed-off-by: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
Cc: Al Viro <viro@ZenIV.linux.org.uk>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---
drivers/char/mem.c | 4 ++++
fs/proc/base.c | 2 ++
fs/read_write.c | 21 +++++++++++++++++++--
include/linux/fs.h | 3 +++
4 files changed, 28 insertions(+), 2 deletions(-)
diff -puN drivers/char/mem.c~vfs-introduce-fmode_neg_offset-for-allowing-negative-f_pos drivers/char/mem.c
--- a/drivers/char/mem.c~vfs-introduce-fmode_neg_offset-for-allowing-negative-f_pos
+++ a/drivers/char/mem.c
@@ -873,6 +873,10 @@ static int memory_open(struct inode *ino
if (dev->dev_info)
filp->f_mapping->backing_dev_info = dev->dev_info;
+ /* Is /dev/mem or /dev/kmem ? */
+ if (dev->dev_info == &directly_mappable_cdev_bdi)
+ filp->f_mode |= FMODE_NEG_OFFSET;
+
if (dev->fops->open)
return dev->fops->open(inode, filp);
diff -puN fs/proc/base.c~vfs-introduce-fmode_neg_offset-for-allowing-negative-f_pos fs/proc/base.c
--- a/fs/proc/base.c~vfs-introduce-fmode_neg_offset-for-allowing-negative-f_pos
+++ a/fs/proc/base.c
@@ -773,6 +773,8 @@ static const struct file_operations proc
static int mem_open(struct inode* inode, struct file* file)
{
file->private_data = (void*)((long)current->self_exec_id);
+ /* OK to pass negative loff_t, we can catch out-of-range */
+ file->f_mode |= FMODE_NEG_OFFSET;
return 0;
}
diff -puN fs/read_write.c~vfs-introduce-fmode_neg_offset-for-allowing-negative-f_pos fs/read_write.c
--- a/fs/read_write.c~vfs-introduce-fmode_neg_offset-for-allowing-negative-f_pos
+++ a/fs/read_write.c
@@ -222,6 +222,20 @@ bad:
}
#endif
+static int
+__negative_fpos_check(struct file *file, loff_t pos, size_t count)
+{
+ /*
+ * pos or pos+count is negative here, check overflow.
+ * too big "count" will be caught in rw_verify_area().
+ */
+ if ((pos < 0) && (pos + count < pos))
+ return -EOVERFLOW;
+ if (file->f_mode & FMODE_NEG_OFFSET)
+ return 0;
+ return -EINVAL;
+}
+
/*
* rw_verify_area doesn't like huge counts. We limit
* them to something that fits in "int" so that others
@@ -239,8 +253,11 @@ int rw_verify_area(int read_write, struc
if (unlikely((ssize_t) count < 0))
return retval;
pos = *ppos;
- if (unlikely((pos < 0) || (loff_t) (pos + count) < 0))
- return retval;
+ if (unlikely((pos < 0) || (loff_t) (pos + count) < 0)) {
+ retval = __negative_fpos_check(file, pos, count);
+ if (retval)
+ return retval;
+ }
if (unlikely(inode->i_flock && mandatory_lock(inode))) {
retval = locks_mandatory_area(
diff -puN include/linux/fs.h~vfs-introduce-fmode_neg_offset-for-allowing-negative-f_pos include/linux/fs.h
--- a/include/linux/fs.h~vfs-introduce-fmode_neg_offset-for-allowing-negative-f_pos
+++ a/include/linux/fs.h
@@ -90,6 +90,9 @@ struct inodes_stat_t {
/* Expect random access pattern */
#define FMODE_RANDOM ((__force fmode_t)0x1000)
+/* File is huge (eg. /dev/kmem): treat loff_t as unsigned */
+#define FMODE_NEG_OFFSET ((__force fmode_t)0x2000)
+
/* File was opened by fanotify and shouldn't generate fanotify events */
/* File was opened by fanotify and shouldn't generate fanotify events */
#define FMODE_NONOTIFY ((__force fmode_t)0x1000000)
_
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [patch 3/5] vfs: introduce FMODE_NEG_OFFSET for allowing negative f_pos
2010-07-21 2:54 ` Nick Piggin
@ 2010-07-21 2:53 ` KAMEZAWA Hiroyuki
2010-07-21 3:10 ` Wu Fengguang
2010-07-21 4:03 ` Andrew Morton
0 siblings, 2 replies; 6+ messages in thread
From: KAMEZAWA Hiroyuki @ 2010-07-21 2:53 UTC (permalink / raw)
To: Nick Piggin; +Cc: akpm, viro, linux-fsdevel, fengguang.wu, heiko.carstens
On Wed, 21 Jul 2010 12:54:37 +1000
Nick Piggin <npiggin@suse.de> wrote:
> On Tue, Jul 20, 2010 at 03:29:48PM -0700, Andrew Morton wrote:
> > From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
> >
> > Now, rw_verify_area() checsk f_pos is negative or not. And if negative,
> > returns -EINVAL.
> >
> > But, some special files as /dev/(k)mem and /proc/<pid>/mem etc.. has
> > negative offsets. And we can't do any access via read/write to the
> > file(device).
> >
> > So introduce FMODE_NEG_OFFSET to allow negative file offsets.
>
> Minor nitpick but I don't understand why this is called NEG_OFFSET. It's
> a large positive offset into the file so FMODE_UNSIGNED_OFFSET seems
> like it would be better.
>
Ah, agreed.
I'll prepare a macro-renaming patch later (if this patch goes now.)
Thank you for review.
-Kame
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [patch 3/5] vfs: introduce FMODE_NEG_OFFSET for allowing negative f_pos
2010-07-20 22:29 [patch 3/5] vfs: introduce FMODE_NEG_OFFSET for allowing negative f_pos akpm
@ 2010-07-21 2:54 ` Nick Piggin
2010-07-21 2:53 ` KAMEZAWA Hiroyuki
0 siblings, 1 reply; 6+ messages in thread
From: Nick Piggin @ 2010-07-21 2:54 UTC (permalink / raw)
To: akpm; +Cc: viro, linux-fsdevel, kamezawa.hiroyu, fengguang.wu,
heiko.carstens
On Tue, Jul 20, 2010 at 03:29:48PM -0700, Andrew Morton wrote:
> From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
>
> Now, rw_verify_area() checsk f_pos is negative or not. And if negative,
> returns -EINVAL.
>
> But, some special files as /dev/(k)mem and /proc/<pid>/mem etc.. has
> negative offsets. And we can't do any access via read/write to the
> file(device).
>
> So introduce FMODE_NEG_OFFSET to allow negative file offsets.
Minor nitpick but I don't understand why this is called NEG_OFFSET. It's
a large positive offset into the file so FMODE_UNSIGNED_OFFSET seems
like it would be better.
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [patch 3/5] vfs: introduce FMODE_NEG_OFFSET for allowing negative f_pos
2010-07-21 2:53 ` KAMEZAWA Hiroyuki
@ 2010-07-21 3:10 ` Wu Fengguang
2010-07-21 4:03 ` Andrew Morton
1 sibling, 0 replies; 6+ messages in thread
From: Wu Fengguang @ 2010-07-21 3:10 UTC (permalink / raw)
To: KAMEZAWA Hiroyuki
Cc: Nick Piggin, akpm@linux-foundation.org, viro@zeniv.linux.org.uk,
linux-fsdevel@vger.kernel.org, heiko.carstens@de.ibm.com
On Wed, Jul 21, 2010 at 10:53:59AM +0800, KAMEZAWA Hiroyuki wrote:
> On Wed, 21 Jul 2010 12:54:37 +1000
> Nick Piggin <npiggin@suse.de> wrote:
>
> > On Tue, Jul 20, 2010 at 03:29:48PM -0700, Andrew Morton wrote:
> > > From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
> > >
> > > Now, rw_verify_area() checsk f_pos is negative or not. And if negative,
> > > returns -EINVAL.
> > >
> > > But, some special files as /dev/(k)mem and /proc/<pid>/mem etc.. has
> > > negative offsets. And we can't do any access via read/write to the
> > > file(device).
> > >
> > > So introduce FMODE_NEG_OFFSET to allow negative file offsets.
> >
> > Minor nitpick but I don't understand why this is called NEG_OFFSET. It's
> > a large positive offset into the file so FMODE_UNSIGNED_OFFSET seems
> > like it would be better.
> >
> Ah, agreed.
Agreed too.
> I'll prepare a macro-renaming patch later (if this patch goes now.)
> Thank you for review.
Thanks,
Fengguang
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: [patch 3/5] vfs: introduce FMODE_NEG_OFFSET for allowing negative f_pos
2010-07-21 2:53 ` KAMEZAWA Hiroyuki
2010-07-21 3:10 ` Wu Fengguang
@ 2010-07-21 4:03 ` Andrew Morton
1 sibling, 0 replies; 6+ messages in thread
From: Andrew Morton @ 2010-07-21 4:03 UTC (permalink / raw)
To: KAMEZAWA Hiroyuki
Cc: Nick Piggin, viro, linux-fsdevel, fengguang.wu, heiko.carstens
On Wed, 21 Jul 2010 11:53:59 +0900 KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com> wrote:
> On Wed, 21 Jul 2010 12:54:37 +1000
> Nick Piggin <npiggin@suse.de> wrote:
>
> > On Tue, Jul 20, 2010 at 03:29:48PM -0700, Andrew Morton wrote:
> > > From: KAMEZAWA Hiroyuki <kamezawa.hiroyu@jp.fujitsu.com>
> > >
> > > Now, rw_verify_area() checsk f_pos is negative or not. And if negative,
> > > returns -EINVAL.
> > >
> > > But, some special files as /dev/(k)mem and /proc/<pid>/mem etc.. has
> > > negative offsets. And we can't do any access via read/write to the
> > > file(device).
> > >
> > > So introduce FMODE_NEG_OFFSET to allow negative file offsets.
> >
> > Minor nitpick but I don't understand why this is called NEG_OFFSET. It's
> > a large positive offset into the file so FMODE_UNSIGNED_OFFSET seems
> > like it would be better.
> >
> Ah, agreed.
>
> I'll prepare a macro-renaming patch later (if this patch goes now.)
It won't. I hand-edited the diff...
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2010-07-21 4:04 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2010-07-20 22:29 [patch 3/5] vfs: introduce FMODE_NEG_OFFSET for allowing negative f_pos akpm
2010-07-21 2:54 ` Nick Piggin
2010-07-21 2:53 ` KAMEZAWA Hiroyuki
2010-07-21 3:10 ` Wu Fengguang
2010-07-21 4:03 ` Andrew Morton
-- strict thread matches above, loose matches on Subject: below --
2010-05-24 19:24 akpm
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).