On x86_32 casting the unsigned int result of get_random_int() to
long may result in a negative value. On x86_32 the range of
mmap_rnd() therefore was -255 to 255. The 32bit mode on x86_64
used 0 to 255 as intended.
The bug was introduced by 675a081 ("x86: unify mmap_{32|64}.c")
in January 2008.
Signed-off-by: Ludwig Nussel <ludwig.nussel@suse.de> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: harvey.harrison@gmail.com Cc: "H. Peter Anvin" <hpa@zytor.com> Cc: Harvey Harrison <harvey.harrison@gmail.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Link: http://lkml.kernel.org/r/201111152246.pAFMklOB028527@wpaz5.hot.corp.google.com Signed-off-by: Ingo Molnar <mingo@elte.hu> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
Check against mistakenly passing in IPv6 addresses (which would result
in an INADDR_ANY bind) or similar incompatible sockaddrs.
Signed-off-by: Marcus Meissner <meissner@suse.de> Cc: Reinhard Max <max@suse.de> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
john stultz [Wed, 11 May 2011 23:10:28 +0000 (16:10 -0700)]
Fix time() inconsistencies caused by intermediate xtime_cache values being read
Currently with 2.6.32-longterm, its possible for time() to occasionally
return values one second earlier then the previous time() call.
This happens because update_xtime_cache() does:
xtime_cache = xtime;
timespec_add_ns(&xtime_cache, nsec);
Its possible that xtime is 1sec,999msecs, and nsecs is 1ms, resulting in
a xtime_cache that is 2sec,0ms.
get_seconds() (which is used by sys_time()) does not take the
xtime_lock, which is ok as the xtime.tv_sec value is a long and can be
atomically read safely.
The problem occurs the next call to update_xtime_cache() if xtime has
not increased:
/* This sets xtime_cache back to 1sec, 999msec */
xtime_cache = xtime;
/* get_seconds, calls here, and sees a 1second inconsistency */
timespec_add_ns(&xtime_cache, nsec);
In order to resolve this, we could add locking to get_seconds(), but it
needs to be lock free, as it is called from the machine check handler,
opening a possible deadlock.
So instead, this patch introduces an intermediate value for the
calculations, so that we only assign xtime_cache once with the correct
time, using ACCESS_ONCE to make sure the compiler doesn't optimize out
any intermediate values.
The xtime_cache manipulations were removed with 2.6.35, so that kernel
and later do not need this change.
In 2.6.33 and 2.6.34 the logarithmic accumulation should make it so
xtime is updated each tick, so it is unlikely that two updates to
xtime_cache could occur while the difference between xtime and
xtime_cache crosses the second boundary. However, the paranoid might
want to pull this into 2.6.33/34-longterm just to be sure.
Thanks to Stephen for helping finally narrow down the root cause and
many hours of help with testing and validation. Also thanks to Max,
Andi, Eric and Paul for review of earlier attempts and helping clarify
what is possible with regard to out of order execution.
Acked-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: John Stultz <johnstul@us.ibm.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
In 2.6.27, commit 393e52e33c6c2 (packet: deliver VLAN TCI to userspace)
added a small information leak.
Add padding field and make sure its zeroed before copy to user.
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> CC: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
there's a kernel bug related to reading the last allowed page on x86_64.
The _copy_to_user() and _copy_from_user() functions use the following
check for address limit:
if (buf + size >= limit)
fail();
while it should be more permissive:
if (buf + size > limit)
fail();
That's because the size represents the number of bytes being
read/write from/to buf address AND including the buf address.
So the copy function will actually never touch the limit
address even if "buf + size == limit".
Following program fails to use the last page as buffer
due to the wrong limit check:
The other place checking the addr limit is the access_ok() function,
which is working properly. There's just a misleading comment
for the __range_not_ok() macro - which this patch fixes as well.
The last page of the user-space address range is a guard page and
Brian Gerst observed that the guard page itself due to an erratum on K8 cpus
(#121 Sequential Execution Across Non-Canonical Boundary Causes Processor
Hang).
However, the test code is using the last valid page before the guard page.
The bug is that the last byte before the guard page can't be read
because of the off-by-one error. The guard page is left in place.
This bug would normally not show up because the last page is
part of the process stack and never accessed via syscalls.
blk_cleanup_queue() calls elevator_exit() and after this, we can't
touch the elevator without oopsing. __elv_next_request() must check
for this state because in the refcounted queue model, we can still
call it after blk_cleanup_queue() has been called.
This was reported as causing an oops attributable to scsi.
[WT: in 2.6.27, __elv_next_request() is in elevator.c]
Resulted from bonding driver registering packet handlers via dev_add_pack and
then trying to call pskb_may_pull. If another packet handler (like for AF_PACKET
sockets) gets called first, the delivered skb will have a user count > 1, which
causes pskb_may_pull to BUG halt when it does its skb_shared check. Fix this by
calling skb_share_check prior to the may_pull call sites in the bonding driver
to clone the skb when needed. Tested by myself and the reported successfully.
Signed-off-by: Neil Horman <nhorman@tuxdriver.com> CC: Andy Gospodarek <andy@greyhouse.net> CC: Jay Vosburgh <fubar@us.ibm.com> CC: "David S. Miller" <davem@davemloft.net> Signed-off-by: Jay Vosburgh <fubar@us.ibm.com> Signed-off-by: Andy Gospodarek <andy@greyhouse.net> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
It was recently brought to my attention that 802.3ad mode bonds would no
longer form when using some network hardware after a driver update.
After snooping around I realized that the particular hardware was using
page-based skbs and found that skb->data did not contain a valid LACPDU
as it was not stored there. That explained the inability to form an
802.3ad-based bond. For balance-alb mode bonds this was also an issue
as ARPs would not be properly processed.
This patch fixes the issue in my tests and should be applied to 2.6.36
and as far back as anyone cares to add it to stable.
Thanks to Alexander Duyck <alexander.h.duyck@intel.com> and Jesse
Brandeburg <jesse.brandeburg@intel.com> for the suggestions on this one.
Signed-off-by: Andy Gospodarek <andy@greyhouse.net> CC: Alexander Duyck <alexander.h.duyck@intel.com> CC: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Jay Vosburgh <fubar@us.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
sym53c8xx_slave_destroy unconditionally assumes that sym53c8xx_slave_alloc has
succesesfully allocated a sym_lcb. This can lead to a NULL pointer dereference
(exposed by commit 4e6c82b).
expkey_parse() oopses when handling a 0 length export. This is easily
triggerable from usermode by writing 0 bytes into
'/proc/[proc id]/net/rpc/nfsd.fh/channel'.
The pool_to and to_pool fields of the global svc_pool_map are freed on
shutdown, but are initialized in nfsd startup only in the
SVC_POOL_PERCPU and SVC_POOL_PERNODE cases.
They *are* initialized to zero on kernel startup. So as long as you use
only SVC_POOL_GLOBAL (the default), this will never be a problem.
You're also OK if you only ever use SVC_POOL_PERCPU or SVC_POOL_PERNODE.
However, the following sequence events leads to a double-free:
1. set SVC_POOL_PERCPU or SVC_POOL_PERNODE
2. start nfsd: both fields are initialized.
3. shutdown nfsd: both fields are freed.
4. set SVC_POOL_GLOBAL
5. start nfsd: the fields are left untouched.
6. shutdown nfsd: now we try to free them again.
Step 4 is actually unnecessary, since (for some bizarre reason), nfsd
automatically resets the pool mode to SVC_POOL_GLOBAL on shutdown.
Signed-off-by: J. Bruce Fields <bfields@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
Echo vendor and product number of a non usb-storage device to
usb-storage driver's new_id, then plug in the device to host and you
will find following oops msg, the root cause is usb_stor_probe1()
refers invalid id entry if giving a dynamic id, so just disable the
feature.
When using a >8bpp framebuffer, offb advertises truecolor, not directcolor,
and doesn't touch the color map even if it has a corresponding access method
for the real hardware.
Thus it needs to set the pseudo-palette with all 3 components of the color,
like other truecolor framebuffers, not with copies of the color index like
a directcolor framebuffer would do.
This went unnoticed for a long time because it's pretty hard to get offb
to kick in with anything but 8bpp (old BootX under MacOS will do that and
qemu does it).
If the pte mapping in generic_perform_write() is unmapped between
iov_iter_fault_in_readable() and iov_iter_copy_from_user_atomic(), the
"copied" parameter to ->end_write can be zero. ext4 couldn't cope with
it with delayed allocations enabled. This skips the i_disksize
enlargement logic if copied is zero and no new data was appeneded to
the inode.
gdb> bt
#0 0xffffffff811afe80 in ext4_da_should_update_i_disksize (file=0xffff88003f606a80, mapping=0xffff88001d3824e0, pos=0x1\
08000, len=0x1000, copied=0x0, page=0xffffea0000d792e8, fsdata=0x0) at fs/ext4/inode.c:2467
#1 ext4_da_write_end (file=0xffff88003f606a80, mapping=0xffff88001d3824e0, pos=0x108000, len=0x1000, copied=0x0, page=0\
xffffea0000d792e8, fsdata=0x0) at fs/ext4/inode.c:2512
#2 0xffffffff810d97f1 in generic_perform_write (iocb=<value optimized out>, iov=<value optimized out>, nr_segs=<value o\
ptimized out>, pos=0x108000, ppos=0xffff88001e26be40, count=<value optimized out>, written=0x0) at mm/filemap.c:2440
#3 generic_file_buffered_write (iocb=<value optimized out>, iov=<value optimized out>, nr_segs=<value optimized out>, p\
os=0x108000, ppos=0xffff88001e26be40, count=<value optimized out>, written=0x0) at mm/filemap.c:2482
#4 0xffffffff810db5d1 in __generic_file_aio_write (iocb=0xffff88001e26bde8, iov=0xffff88001e26bec8, nr_segs=0x1, ppos=0\
xffff88001e26be40) at mm/filemap.c:2600
#5 0xffffffff810db853 in generic_file_aio_write (iocb=0xffff88001e26bde8, iov=0xffff88001e26bec8, nr_segs=<value optimi\
zed out>, pos=<value optimized out>) at mm/filemap.c:2632
#6 0xffffffff811a71aa in ext4_file_write (iocb=0xffff88001e26bde8, iov=0xffff88001e26bec8, nr_segs=0x1, pos=0x108000) a\
t fs/ext4/file.c:136
#7 0xffffffff811375aa in do_sync_write (filp=0xffff88003f606a80, buf=<value optimized out>, len=<value optimized out>, \
ppos=0xffff88001e26bf48) at fs/read_write.c:406
#8 0xffffffff81137e56 in vfs_write (file=0xffff88003f606a80, buf=0x1ec2960 <Address 0x1ec2960 out of bounds>, count=0x4\
000, pos=0xffff88001e26bf48) at fs/read_write.c:435
#9 0xffffffff8113816c in sys_write (fd=<value optimized out>, buf=0x1ec2960 <Address 0x1ec2960 out of bounds>, count=0x\
4000) at fs/read_write.c:487
#10 <signal handler called>
#11 0x00007f120077a390 in __brk_reservation_fn_dmi_alloc__ ()
#12 0x0000000000000000 in ?? ()
gdb> print offset
$22 = 0xffffffffffffffff
gdb> print idx
$23 = 0xffffffff
gdb> print inode->i_blkbits
$24 = 0xc
gdb> up
#1 ext4_da_write_end (file=0xffff88003f606a80, mapping=0xffff88001d3824e0, pos=0x108000, len=0x1000, copied=0x0, page=0\
xffffea0000d792e8, fsdata=0x0) at fs/ext4/inode.c:2512
2512 if (ext4_da_should_update_i_disksize(page, end)) {
gdb> print start
$25 = 0x0
gdb> print end
$26 = 0xffffffffffffffff
gdb> print pos
$27 = 0x108000
gdb> print new_i_size
$28 = 0x108000
gdb> print ((struct ext4_inode_info *)((char *)inode-((int)(&((struct ext4_inode_info *)0)->vfs_inode))))->i_disksize
$29 = 0xd9000
gdb> down
2467 for (i = 0; i < idx; i++)
gdb> print i
$30 = 0xd44acbee
This is 100% reproducible with some autonuma development code tuned in
a very aggressive manner (not normal way even for knumad) which does
"exotic" changes to the ptes. It wouldn't normally trigger but I don't
see why it can't happen normally if the page is added to swap cache in
between the two faults leading to "copied" being zero (which then
hangs in ext4). So it should be fixed. Especially possible with lumpy
reclaim (albeit disabled if compaction is enabled) as that would
ignore the young bits in the ptes.
Backport for stable kernel v2.6.32.y to v2.6.36.y.
Current oprofile's x86 callgraph support may trigger page faults
throwing the BUG_ON(in_nmi()) message below. This patch fixes this by
using the same nmi-safe copy-from-user code as in perf.
------------[ cut here ]------------
kernel BUG at .../arch/x86/kernel/traps.c:436!
invalid opcode: 0000 [#1] SMP
last sysfs file: /sys/devices/pci0000:00/0000:00:0a.0/0000:07:00.0/0000:08:04.0/net/eth0/broadcast
CPU 5
Modules linked in:
Clement Lecigne reports a filesystem which causes a kernel oops in
hfs_find_init() trying to dereference sb->ext_tree which is NULL.
This proves to be because the filesystem has a corrupted MDB extent
record, where the extents file does not fit into the first three extents
in the file record (the first blocks).
In hfs_get_block() when looking up the blocks for the extent file
(HFS_EXT_CNID), it fails the first blocks special case, and falls
through to the extent code (which ultimately calls hfs_find_init())
which is in the process of being initialised.
Hfs avoids this scenario by always having the extents b-tree fitting
into the first blocks (the extents B-tree can't have overflow extents).
The fix is to check at mount time that the B-tree fits into first
blocks, i.e. fail if HFS_I(inode)->alloc_blocks >=
HFS_I(inode)->first_blocks
Note, the existing commit 47f365eb57573 ("hfs: fix oops on mount with
corrupted btree extent records") becomes subsumed into this as a special
case, but only for the extents B-tree (HFS_EXT_CNID), it is perfectly
acceptable for the catalog B-Tree file to grow beyond three extents,
with the remaining extent descriptors in the extents overfow.
[WT: patch edited - 47f365eb57573 was missing from 2.6.27.x]
This fixes CVE-2011-2203
Reported-by: Clement LECIGNE <clement.lecigne@netasq.com> Signed-off-by: Phillip Lougher <plougher@redhat.com> Cc: Jeff Mahoney <jeffm@suse.com> Cc: Christoph Hellwig <hch@lst.de> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Cc: Moritz Mühlenhoff <jmm@inutil.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
Ok, this isn't optimal, since it means that 'iotop' needs admin
capabilities, and we may have to work on this some more. But at the
same time it is very much not acceptable to let anybody just read
anybody elses IO statistics quite at this level.
Use of the GENL_ADMIN_PERM suggested by Johannes Berg as an alternative
to checking the capabilities by hand.
I hit a J_ASSERT(blocknr != 0) failure in cleanup_journal_tail() when
mounting a fsfuzzed ext3 image. It turns out that the corrupted ext3
image has s_first = 0 in journal superblock, and the 0 is passed to
journal->j_head in journal_reset(), then to blocknr in
cleanup_journal_tail(), in the end the J_ASSERT failed.
So validate s_first after reading journal superblock from disk in
journal_get_superblock() to ensure s_first is valid.
If oprofile uses the nmi timer interrupt there is a crash while
unloading the module. The bug can be triggered with oprofile build as
module and kernel parameter nolapic set. This patch fixes this.
oprofile: using NMI timer interrupt.
BUG: unable to handle kernel NULL pointer dereference at 0000000000000008
IP: [<ffffffff8123c226>] unregister_syscore_ops+0x41/0x58
PGD 42dbca067 PUD 41da6a067 PMD 0
Oops: 0002 [#1] PREEMPT SMP
CPU 5
Modules linked in: oprofile(-) [last unloaded: oprofile]
When we tear down a device we try to flush all outstanding
commands in scsi_free_queue(). However the check in
scsi_request_fn() is imperfect as it only signals that
we _might start_ aborting commands, not that we've actually
aborted some.
So move the printk inside the scsi_kill_request function,
this will also give us a hint about which commands are aborted.
Signed-off-by: Hannes Reinecke <hare@suse.de> Signed-off-by: James Bottomley <JBottomley@Parallels.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
ftdi_set_termios() is constantly setting the baud rate, data bits and parity
unnecessarily on every call, . When called while characters are being
transmitted can cause the FTDI chip to corrupt the serial port bit stream
output by stalling the output half a bit during the output of a character.
Simple fix by skipping this setting if the baud rate/data bits/parity are
unchanged.
Make sure that SCSI device removal via scsi_remove_host() does finish
all pending SCSI commands. Currently that's not the case and hence
removal of a SCSI host during I/O can cause a deadlock. See also
"blkdev_issue_discard() hangs forever if underlying storage device is
removed" (http://bugzilla.kernel.org/show_bug.cgi?id=40472). See also
http://lkml.org/lkml/2011/8/27/6.
Signed-off-by: Bart Van Assche <bvanassche@acm.org> Signed-off-by: James Bottomley <JBottomley@Parallels.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
The reason is that it forgot to mark dirty when splitting two extents in
ext4_ext_convert_to_initialized(). Althrough ex has been updated in
memory, it is not dirtied both in ext4_ext_convert_to_initialized() and
ext4_ext_insert_extent(). The disk layout is corrupted. Then it will
meet with a BUG_ON() when writting at the start of that extent again.
Cc: "Theodore Ts'o" <tytso@mit.edu> Cc: Xiaoyun Mao <xiaoyun.maoxy@aliyun-inc.com> Cc: Yingbin Wang <yingbin.wangyb@aliyun-inc.com> Cc: Jia Wan <jia.wanj@aliyun-inc.com> Signed-off-by: Zheng Liu <wenqing.lz@taobao.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
tc_fill_qdisc() should not be called for builtin qdisc, or it
dereference a NULL pointer to get device ifindex.
Fix is to always use tc_qdisc_dump_ignore() before calling
tc_fill_qdisc().
Reported-by: Ben Pfaff <blp@nicira.com> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
A remote user can provide a small value for the command size field in
the command header of an l2cap configuration request, resulting in an
integer underflow when subtracting the size of the configuration request
header. This results in copying a very large amount of data via
memcpy() and destroying the kernel heap. Check for underflow.
Signed-off-by: Dan Rosenberg <drosenberg@vsecurity.com> Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
If the NLM daemon is killed on the NFS server, we can currently end up
hanging forever on an 'unlock' request, instead of aborting. Basically,
if the rpcbind request fails, or the server keeps returning garbage, we
really want to quit instead of retrying.
ubd_file_size() cannot use ubd_dev->cow.file because at this time
ubd_dev->cow.file is not initialized.
Therefore, ubd_file_size() will always report a wrong disk size when
COW files are used.
Reading from /dev/ubd* would crash the kernel.
We have to read the correct disk size from the COW file's backing
file.
The following patch is admittedly a band-aid, and does not solve the
root cause, but it still is a good candidate for hardening as a pointer
check before reference.
Signed-off-by: Mark Salyzyn <mark_salyzyn@us.xyratex.com> Tested-by: Jack Wang <jack_wang@usish.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
This patch changes the call of tpm_transmit by supplying the size of the
userspace buffer instead of TPM_BUFSIZE.
This got assigned CVE-2011-1161.
[The first hunk didn't make sense given one could expect
way less data than TPM_BUFSIZE, so added tpm_transmit boundary
check over bufsiz instead
The last parameter of tpm_transmit() reflects the amount
of data expected from the device, and not the buffer size
being supplied to it. It isn't ideal to parse it directly,
so we just set it to the maximum the input buffer can handle
and let the userspace API to do such job.]
The name_len variable in CIFSFindNext is a signed int that gets set to
the resume_name_len in the cifs_search_info. The resume_name_len however
is unsigned and for some infolevels is populated directly from a 32 bit
value sent by the server.
If the server sends a very large value for this, then that value could
look negative when converted to a signed int. That would make that
value pass the PATH_MAX check later in CIFSFindNext. The name_len would
then be used as a length value for a memcpy. It would then be treated
as unsigned again, and the memcpy scribbles over a ton of memory.
Fix this by making the name_len an unsigned value in CIFSFindNext.
Reported-by: Darren Lavender <dcl@hppine99.gbr.hp.com> Signed-off-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Steve French <sfrench@us.ibm.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
On a box with 8TB of RAM the MMU hashtable is 64GB in size. That
means we have 4G PTEs. pSeries_lpar_hptab_clear was using a signed
int to store the index which will overflow at 2G.
Signed-off-by: Anton Blanchard <anton@samba.org> Acked-by: Michael Neuling <mikey@neuling.org> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
ie "IBM,Logh". OF got corrupted with a device tree string.
Looking at make_room and alloc_up, we claim the first chunk (1 MB)
but we never claim any more. mem_end is always set to alloc_top
which is the top of our available address space, guaranteeing we will
never call alloc_up and claim more memory.
Also alloc_up wasn't setting alloc_bottom to the bottom of the
available address space.
This doesn't help the box to boot, but we at least fail with
an obvious error. We could relocate the device tree in a future
patch.
Signed-off-by: Anton Blanchard <anton@samba.org> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
After commit 3262c816a3d7fb1eaabce633caa317887ed549ae "[PATCH] knfsd:
split svc_serv into pools", svc_delete_xprt (then svc_delete_socket) no
longer removed its xpt_ready (then sk_ready) field from whatever list it
was on, noting that there was no point since the whole list was about to
be destroyed anyway.
That was mostly true, but forgot that a few svc_xprt_enqueue()'s might
still be hanging around playing with the about-to-be-destroyed list, and
could get themselves into trouble writing to freed memory if we left
this xprt on the list after freeing it.
(This is actually functionally identical to a patch made first by Ben
Greear, but with more comments.)
Cc: gnb@fmeh.org Reported-by: Ben Greear <greearb@candelatech.com> Tested-by: Ben Greear <greearb@candelatech.com> Signed-off-by: J. Bruce Fields <bfields@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
Block allocation is called from two places: ext3_get_blocks_handle() and
ext3_xattr_block_set(). These two callers are not necessarily synchronized
because xattr code holds only xattr_sem and i_mutex, and
ext3_get_blocks_handle() may hold only truncate_mutex when called from
writepage() path. Block reservation code does not expect two concurrent
allocations to happen to the same inode and thus assertions can be triggered
or reservation structure corruption can occur.
Fix the problem by taking truncate_mutex in xattr code to serialize
allocations.
CC: Sage Weil <sage@newdream.net> Reported-by: Fyodor Ustinov <ufm@ufm.su> Signed-off-by: Jan Kara <jack@suse.cz> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
The existing code it pretty ugly. How about we clean it up even more
like this?
From: Anton Blanchard <anton@samba.org>
We check for timeout expiry in the outer loop, but we also need to
check it in the inner loop or we can lock up forever waiting for a
CPU to hit real mode.
Signed-off-by: Anton Blanchard <anton@samba.org> Signed-off-by: Michael Neuling <mikey@neuling.org> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
If expander discovery fails (sas_discover_expander()), remove the
expander from the port device list (sas_ex_discover_expander()),
before freeing it. Else the list is corrupted and, e.g., when we
attempt to send SMP commands to other devices, the kernel oopses.
Signed-off-by: Luben Tuikov <ltuikov@yahoo.com> Reviewed-by: Jack Wang <jack_wang@usish.com> Signed-off-by: James Bottomley <JBottomley@Parallels.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
Rebooting on the Dell E5420 often hangs with the keyboard or ACPI
methods, but is reliable via the PCI method.
[ hpa: this was deferred because we believed for a long time that the
recent reshuffling of the boot priorities in commit 660e34cebf0a11d54f2d5dd8838607452355f321 fixed this platform.
Unfortunately that turned out to be incorrect. ]
Because struct rpcbind_args *map was declared static, if two
threads entered this method at the same time, the values
assigned to map could be sent two two differen tasks.
This could cause all sorts of problems, include use-after-free
and double-free of memory.
Fix this by removing the static declaration so that the map
pointer is on the stack.
Since we are modifying this RCU pointer, we need to hold
the lock protecting it around it.
This fixes a potential reuse and double free of a cfq
io_context structure. The bug has been in CFQ for a long
time, it hit very few people but those it did hit seemed
to see it a lot.
Check pers->hot_remove_disk instead of pers->hot_add_disk in slot_store()
during disk removal. The linear personality only has ->hot_add_disk and
no ->hot_remove_disk, so that removing disk in the array resulted to
following kernel bug:
The driver contains several loops counting on an u16 value
where the exit-condition is checked against variables that
can have values up to 0xffff. In this case the loops will
never exit. This patch fixed 3 such loops.
Move the smp_rmb after cpu_relax loop in read_seqlock and add
ACCESS_ONCE to make sure the test and return are consistent.
A multi-threaded core in the lab didn't like the update
from 2.6.35 to 2.6.36, to the point it would hang during
boot when multiple threads were active. Bisection showed af5ab277ded04bd9bc6b048c5a2f0e7d70ef0867 (clockevents:
Remove the per cpu tick skew) as the culprit and it is
supported with stack traces showing xtime_lock waits including
tick_do_update_jiffies64 and/or update_vsyscall.
Experimentation showed the combination of cpu_relax and smp_rmb
was significantly slowing the progress of other threads sharing
the core, and this patch is effective in avoiding the hang.
A theory is the rmb is affecting the whole core while the
cpu_relax is causing a resource rebalance flush, together they
cause an interfernce cadance that is unbroken when the seqlock
reader has interrupts disabled.
At first I was confused why the refactor in 3c22cd5709e8143444a6d08682a87f4c57902df3 (kernel: optimise
seqlock) didn't affect this patch application, but after some
study that affected seqcount not seqlock. The new seqcount was
not factored back into the seqlock. I defer that the future.
While the removal of the timer interrupt offset created
contention for the xtime lock while a cpu does the
additonal work to update the system clock, the seqlock
implementation with the tight rmb spin loop goes back much
further, and is just waiting for the right trigger.
Signed-off-by: Milton Miller <miltonm@bga.com> Cc: <linuxppc-dev@lists.ozlabs.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Andi Kleen <andi@firstfloor.org> Cc: Nick Piggin <npiggin@kernel.dk> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org> Cc: Anton Blanchard <anton@samba.org> Cc: Paul McKenney <paulmck@linux.vnet.ibm.com> Acked-by: Eric Dumazet <eric.dumazet@gmail.com> Link: http://lkml.kernel.org/r/%3Cseqlock-rmb%40mdm.bga.com%3E Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
As Ben Hutchings discovered [1], the patch for CVE-2011-1017 (buffer
overflow in ldm_frag_add) is not sufficient. The original patch in
commit c340b1d64000 ("fs/partitions/ldm.c: fix oops caused by corrupted
partition table") does not consider that, for subsequent fragments,
previously allocated memory is used.
[1] http://lkml.org/lkml/2011/5/6/407
Reported-by: Ben Hutchings <ben@decadent.org.uk> Signed-off-by: Timo Warns <warns@pre-sense.de> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
When make_indexed_dir() fails (e.g. because of ENOSPC) after it has allocated
block for index tree root, we did not properly mark all changed buffers dirty.
This lead to only some of these buffers being written out and thus effectively
corrupting the directory.
Fix the issue by marking all changed data dirty even in the error failure case.
Signed-off-by: Jan Kara <jack@suse.cz> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
Commit 0837e3242c73566fc1c0196b4ec61779c25ffc93 fixes a situation on POWER7
where events can roll back if a specualtive event doesn't actually complete.
This can raise a performance monitor exception. We need to catch this to ensure
that we reset the PMC. In all cases the PMC will be less than 256 cycles from
overflow.
This patch lifts Anton's fix for the problem in perf and applies it to oprofile
as well.
Signed-off-by: Eric B Munson <emunson@mgebm.net> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
We occasionally see list corruption using libertas.
While we haven't been able to diagnose this precisely, we have spotted
a possible cause: cmdpendingq is generally modified with driver_lock
held. However, there are a couple of points where this is not the case.
Fix up those operations to execute under the lock, it seems like
the correct thing to do and will hopefully improve the situation.
Signed-off-by: Paul Fox <pgf@laptop.org> Signed-off-by: Daniel Drake <dsd@laptop.org> Acked-by: Dan Williams <dcbw@redhat.com> Signed-off-by: John W. Linville <linville@tuxdriver.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
MUSB is a non-standard host implementation which
can handle all speeds with the same core. We need
to set has_tt flag after commit d199c96d41d80a567493e12b8e96ea056a1350c1 (USB: prevent
buggy hubs from crashing the USB stack) in order for
MUSB HCD to continue working.
Signed-off-by: Felipe Balbi <balbi@ti.com> Cc: Alan Stern <stern@rowland.harvard.edu> Tested-by: Michael Jones <michael.jones@matrix-vision.de> Tested-by: Alexander Holler <holler@ahsoftware.de> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
ata_pio_sectors() expects buffer for each sector to be contained in a
single page; otherwise, it ends up overrunning the first page. This
is achieved by setting queue DMA alignment. If sector_size is smaller
than PAGE_SIZE and all buffers are sector_size aligned, buffer for
each sector is always contained in a single page.
This wasn't applied to ATAPI devices but IDENTIFY_PACKET is executed
as ATA_PROT_PIO and thus uses ata_pio_sectors(). Newer versions of
udev issue IDENTIFY_PACKET with unaligned buffer triggering the
problem and causing oops.
This patch fixes the problem by setting sdev->sector_size to
ATA_SECT_SIZE on ATATPI devices and always setting DMA alignment to
sector_size. While at it, add a warning for the unlikely but still
possible scenario where sector_size is larger than PAGE_SIZE, in which
case the alignment wouldn't be enough.
Signed-off-by: Tejun Heo <tj@kernel.org> Reported-by: John Stanley <jpsinthemix@verizon.net> Tested-by: John Stanley <jpsinthemix@verizon.net> Signed-off-by: Jeff Garzik <jgarzik@redhat.com> Signed-off-by: Jonathan Liu <net147@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
When writing a disc on certain lite-on dvd-writers (also rebadged
as optiarc/LG/...) connected to a vt6420, the ATAPI CDB ends
up in the datastream and on the disc, causing silent corruption.
Delaying between sending the CDB and starting DMA seems to
prevent this.
I do not know if there are burners that do not suffer from
this, but the patch should be safe for those as well.
There are many reports of this issue, but AFAICT no solution was
found before. For example:
http://lkml.indiana.edu/hypermail/linux/kernel/0802.3/0561.html
Signed-off-by: Bart Hartgers <bart.hartgers@gmail.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com>
[bwh: Remove version bump for 2.6.32] Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
The kernel automatically evaluates partition tables of storage devices.
The code for evaluating OSF partitions contains a bug that leaks data
from kernel heap memory to userspace for certain corrupted OSF
partitions.
In more detail:
for (i = 0 ; i < le16_to_cpu(label->d_npartitions); i++, partition++) {
iterates from 0 to d_npartitions - 1, where d_npartitions is read from
the partition table without validation and partition is a pointer to an
array of at most 8 d_partitions.
Add the proper and obvious validation.
Signed-off-by: Timo Warns <warns@pre-sense.de> Cc: stable@kernel.org
[ Changed the patch trivially to not repeat the whole le16_to_cpu()
thing, and to use an explicit constant for the magic value '8' ] Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
When reusing a TCP connection, ensure that it's aborted if a previous
shutdown attempt has been made on that connection so that the RPC over
TCP recovery mechanism succeeds.
The kernel automatically evaluates partition tables of storage devices.
The code for evaluating LDM partitions (in fs/partitions/ldm.c) contains
a bug that causes a kernel oops on certain corrupted LDM partitions.
A kernel subsystem seems to crash, because, after the oops, the kernel no
longer recognizes newly connected storage devices.
The patch validates the value of vblk_size.
[akpm@linux-foundation.org: coding-style fixes] Signed-off-by: Timo Warns <warns@pre-sense.de> Cc: Eugene Teo <eugeneteo@kernel.sg> Cc: Harvey Harrison <harvey.harrison@gmail.com> Cc: Richard Russon <rich@flatcap.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
An open on a NFS4 share using the O_CREAT flag on an existing file for
which we have permissions to open but contained in a directory with no
write permissions will fail with EACCES.
A tcpdump shows that the client had set the open mode to UNCHECKED which
indicates that the file should be created if it doesn't exist and
encountering an existing flag is not an error. Since in this case the
file exists and can be opened by the user, the NFS server is wrong in
attempting to check create permissions on the parent directory.
The patch adds a conditional statement to check for create permissions
only if the file doesn't exist.
Signed-off-by: Sachin S. Prabhu <sprabhu@redhat.com> Signed-off-by: J. Bruce Fields <bfields@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
When CONFIG_OABI_COMPAT is set, the wrapper for semtimedop does not
bound the nsops argument. A sufficiently large value will cause an
integer overflow in allocation size, followed by copying too much data
into the allocated buffer. Fix this by restricting nsops to SEMOPM.
Untested.
Signed-off-by: Dan Rosenberg <drosenberg@vsecurity.com> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
This fixes the following oops discovered by Dan Aloni:
> Anyway, the following is the output of the Oops that I got on the
> Ubuntu kernel on which I first detected the problem
> (2.6.37-12-generic). The Oops that followed will be more useful, I
> guess.
The bug was that unix domain sockets use a pseduo packet for
connecting and accept uses that psudo packet to get the socket.
In the buggy seqpacket case we were allowing unconnected
sockets to call recvmsg and try to receive the pseudo packet.
That is always wrong and as of commit 7361c36c5 the pseudo
packet had become enough different from a normal packet
that the kernel started oopsing.
Do for seqpacket_recv what was done for seqpacket_send in 2.5
and only allow it on connected seqpacket sockets.
Tested-by: Dan Aloni <dan@aloni.org> Signed-off-by: Eric W. Biederman <ebiederm@xmission.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
It seems that under certain circumstances the sdhci_tasklet_finish()
call can be entered with mrq set to NULL, causing the system to crash
with a NULL pointer de-reference.
Seen on S3C6410 system. Based on a patch by Dimitris Papastamos.
It seems that under certain circumstances that the sdhci_tasklet_finish()
call can be entered with mrq->cmd set to NULL, causing the system to crash
with a NULL pointer de-reference.
Unable to handle kernel NULL pointer dereference at virtual address 00000000
PC is at sdhci_tasklet_finish+0x34/0xe8
LR is at sdhci_tasklet_finish+0x24/0xe8
Seen on S3C6410 system.
Signed-off-by: Ben Dooks <ben-linux@fluff.org> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Signed-off-by: Chris Ball <cjb@laptop.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
If pci_ioremap_bar() fails during probe, we "goto release;" and free the
host, but then we return 0 -- which tells sdhci_pci_probe() that the probe
succeeded. Since we think the probe succeeded, when we unload sdhci we'll
go to sdhci_pci_remove_slot() and it will try to dereference slot->host,
which is now NULL because we freed it in the error path earlier.
The patch simply sets ret appropriately, so that sdhci_pci_probe() will
detect the failure immediately and bail out.
SCSI uses request_queue->queuedata == NULL as a signal that the queue
is dying. We set this state in the sdev release function. However,
this allows a small window where we release the last reference but
haven't quite got to this stage yet and so something will try to take
a reference in scsi_request_fn and oops. It's very rare, but we had a
report here, so we're pushing this as a bug fix
The actual fix is to set request_queue->queuedata to NULL in
scsi_remove_device() before we drop the reference. This causes
correct automatic rejects from scsi_request_fn as people who hold
additional references try to submit work and prevents anything from
getting a new reference to the sdev that way.
page_count is copied from userspace. agp_allocate_memory() tries to
check whether this number is too big, but doesn't take into account the
wrap case. Also agp_create_user_memory() doesn't check whether
alloc_size is calculated from num_agp_pages variable without overflow.
This may lead to allocation of too small buffer with following buffer
overflow.
Another problem in agp code is not addressed in the patch - kernel memory
exhaustion (AGPIOC_RESERVE and AGPIOC_ALLOCATE ioctls). It is not checked
whether requested pid is a pid of the caller (no check in agpioc_reserve_wrap()).
Each allocation is limited to 16KB, though, there is no per-process limit.
This might lead to OOM situation, which is not even solved in case of the
caller death by OOM killer - the memory is allocated for another (faked) process.
pg_start is copied from userspace on AGPIOC_BIND and AGPIOC_UNBIND ioctl
cmds of agp_ioctl() and passed to agpioc_bind_wrap(). As said in the
comment, (pg_start + mem->page_count) may wrap in case of AGPIOC_BIND,
and it is not checked at all in case of AGPIOC_UNBIND. As a result, user
with sufficient privileges (usually "video" group) may generate either
local DoS or privilege escalation.
When a DISCONTIGMEM memory range is brought online as a NUMA node, it
also needs to have its bet set in N_NORMAL_MEMORY. This is necessary for
generic kernel code that utilizes N_NORMAL_MEMORY as a subset of N_ONLINE
for memory savings.
These types of hacks can hopefully be removed once DISCONTIGMEM is either
removed or abstracted away from CONFIG_NUMA.
Fixes a panic in the slub code which only initializes structures for
N_NORMAL_MEMORY to save memory:
This patch fixes the following symptoms:
1. Unmount UBIFS cleanly.
2. Start mounting UBIFS R/W and have a power cut immediately
3. Start mounting UBIFS R/O, this succeeds
4. Try to re-mount UBIFS R/W - this fails immediately or later on,
because UBIFS will write the master node to the flash area
which has been written before.
The analysis of the problem:
1. UBIFS is unmounted cleanly, both copies of the master node are clean.
2. UBIFS is being mounter R/W, starts changing master node copy 1, and
a power cut happens. The copy N1 becomes corrupted.
3. UBIFS is being mounted R/O. It notices the copy N1 is corrupted and
reads copy N2. Copy N2 is clean.
4. Because of R/O mode, UBIFS cannot recover copy 1.
5. The mount code (ubifs_mount()) sees that the master node is clean,
so it decides that no recovery is needed.
6. We are re-mounting R/W. UBIFS believes no recovery is needed and
starts updating the master node, but copy N1 is still corrupted
and was not recovered!
Fix this problem by marking the master node as dirty every time we
recover it and we are in R/O mode. This forces further recovery and
the UBIFS cleans-up the corruptions and recovers the copy N1 when
re-mounting R/W later.
I may have an explanation for the LSI 1068 HBA hangs provoked by ATA
pass-through commands, in particular by smartctl.
First, my version of the symptoms. On an LSI SAS1068E B3 HBA running
01.29.00.00 firmware, with SATA disks, and with smartd running, I'm seeing
occasional task, bus, and host resets, some of which lead to hard faults of
the HBA requiring a reboot. Abusively looping the smartctl command,
# while true; do smartctl -a /dev/sdb > /dev/null; done
dramatically increases the frequency of these failures to nearly one per
minute. A high IO load through the HBA while looping smartctl seems to
improve the chance of a full scsi host reset or a non-recoverable hang.
I reduced what smartctl was doing down to a simple test case which
causes the hang with a single IO when pointed at the sd interface. See
the code at the bottom of this e-mail. It uses an SG_IO ioctl to issue
a single pass-through ATA identify device command. If the buffer
userspace gives for the read data has certain alignments, the task is
issued to the HBA but the HBA fails to respond. If run against the sg
interface, neither the test code nor smartctl causes a hang.
sd and sg handle the SG_IO ioctl slightly differently. Unless you
specifically set a flag to do direct IO, sg passes a buffer of its own,
which is page-aligned, to the block layer and later copies the result
into the userspace buffer regardless of its alignment. sd, on the other
hand, always does direct IO unless the userspace buffer fails an
alignment test at block/blk-map.c line 57, in which case a page-aligned
buffer is created and used for the transfer.
The alignment test currently checks for word-alignment, the default
setup by scsi_lib.c; therefore, userspace buffers of almost any
alignment are given directly to the HBA as DMA targets. The LSI 1068
hardware doesn't seem to like at least a couple of the alignments which
cross a page boundary (see the test code below). Curiously, many
page-boundary-crossing alignments do work just fine.
So, either the hardware has an bug handling certain alignments or the
hardware has a stricter alignment requirement than the driver is
advertising. If stricter alignment is required, then in no case should
misaligned buffers from userspace be allowed through without being
bounced or at least causing an error to be returned.
It seems the mptsas driver could use blk_queue_dma_alignment() to advertise
a stricter alignment requirement. If it does, sd does the right thing and
bounces misaligned buffers (see block/blk-map.c line 57). The following
patch to 2.6.34-rc5 makes my symptoms go away. I'm sure this is the wrong
place for this code, but it gets my idea across.
Acked-by: "Desai Kashyap" <Kashyap.Desai@lsi.com> Signed-off-by: James Bottomley <James.Bottomley@suse.de>
tcp_read_sock() can have a eat skbs without immediately advancing copied_seq.
This can cause a panic in tcp_collapse() if it is called as a result
of the recv_actor dropping the socket lock.
A userspace program that splices data from a socket to either another
socket or to a file can trigger this bug.
Signed-off-by: Steven J. Magnani <steve@digidescorp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
This patch (as1458) fixes a problem affecting ultra-reliable systems:
When hardware failover of an EHCI controller occurs, the data
structures do not get released correctly. This is because the routine
responsible for removing unused QHs from the async schedule assumes
the controller is running properly (the frame counter is used in
determining how long the QH has been idle) -- but when a failover
causes the controller to be electronically disconnected from the PCI
bus, obviously it stops running.
The solution is simple: Allow scan_async() to remove a QH from the
async schedule if it has been idle for long enough _or_ if the
controller is stopped.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Reported-and-Tested-by: Dan Duval <dan.duval@stratus.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
next_pidmap() just quietly accepted whatever 'last' pid that was passed
in, which is not all that safe when one of the users is /proc.
Admittedly the proc code should do some sanity checking on the range
(and that will be the next commit), but that doesn't mean that the
helper functions should just do that pidmap pointer arithmetic without
checking the range of its arguments.
So clamp 'last' to PID_MAX_LIMIT. The fact that we then do "last+1"
doesn't really matter, the for-loop does check against the end of the
pidmap array properly (it's only the actual pointer arithmetic overflow
case we need to worry about, and going one bit beyond isn't going to
overflow).
[ Use PID_MAX_LIMIT rather than pid_max as per Eric Biederman ]
Reported-by: Tavis Ormandy <taviso@cmpxchg8b.com> Analyzed-by: Robert Święcki <robert@swiecki.net> Cc: Eric W. Biederman <ebiederm@xmission.com> Cc: Pavel Emelyanov <xemul@openvz.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
This patch fixes severe UBIFS bug: UBIFS oopses when we 'fsync()' an
file on R/O-mounter file-system. We (the UBIFS authors) incorrectly
thought that VFS would not propagate 'fsync()' down to the file-system
if it is read-only, but this is not the case.
It is easy to exploit this bug using the following simple perl script:
use strict;
use File::Sync qw(fsync sync);
die "File path is not specified" if not defined $ARGV[0];
my $path = $ARGV[0];
open FILE, "<", "$path" or die "Cannot open $path: $!";
fsync(\*FILE) or die "cannot fsync $path: $!";
close FILE or die "Cannot close $path: $!";
Thanks to Reuben Dowle <Reuben.Dowle@navico.com> for reporting about this
issue.
The pages attached to a ramfs inode's pagecache by truncation from nothing
- as done by SYSV SHM for example - may get discarded under memory
pressure.
The problem is that the pages are not marked dirty. Anything that creates
data in an MMU-based ramfs will cause the pages holding that data will
cause the set_page_dirty() aop to be called.
For the NOMMU-based mmap, set_page_dirty() may be called by write(), but
it won't be called by page-writing faults on writable mmaps, and it isn't
called by ramfs_nommu_expand_for_mapping() when a file is being truncated
from nothing to allocate a contiguous run.
The solution is to mark the pages dirty at the point of allocation by the
truncation code.
Signed-off-by: Enrik Berkhan <Enrik.Berkhan@ge.com> Signed-off-by: David Howells <dhowells@redhat.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Nick Piggin <nickpiggin@yahoo.com.au> Cc: Johannes Weiner <hannes@cmpxchg.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
'simple' would have required specifying current frame address
and return address location manually, but that's obviously not
the case (and not necessary) here.
Unfortunately, one of the callers of that function passes the
address of a smaller data type, cast to fit the type that
xfs_fs_geometry() requires. As a result, this can happen:
Kernel panic - not syncing: stack-protector: Kernel stack is corrupted
in: f87aca93
Pid: 262, comm: xfs_fsr Not tainted 2.6.38-rc6-493f3358cb2+ #1
Call Trace:
Note: this patch targets 2.6.37 and tries to be as simple as possible.
That is why it adds more copy-and-paste horror into fs/compat.c and
uglifies fs/exec.c, this will be cleanuped later.
compat_copy_strings() plays with bprm->vma/mm directly and thus has
two problems: it lacks the RLIMIT_STACK check and argv/envp memory
is not visible to oom killer.
Export acct_arg_size() and get_arg_page(), change compat_copy_strings()
to use get_arg_page(), change compat_do_execve() to do acct_arg_size(0)
as do_execve() does.
Add the fatal_signal_pending/cond_resched checks into compat_count() and
compat_copy_strings(), this matches the code in fs/exec.c and certainly
makes sense.
Brad Spengler published a local memory-allocation DoS that
evades the OOM-killer (though not the virtual memory RLIMIT):
http://www.grsecurity.net/~spender/64bit_dos.c
execve()->copy_strings() can allocate a lot of memory, but
this is not visible to oom-killer, nobody can see the nascent
bprm->mm and take it into account.
With this patch get_arg_page() increments current's MM_ANONPAGES
counter every time we allocate the new page for argv/envp. When
do_execve() succeds or fails, we change this counter back.
Technically this is not 100% correct, we can't know if the new
page is swapped out and turn MM_ANONPAGES into MM_SWAPENTS, but
I don't think this really matters and everything becomes correct
once exec changes ->mm or fails.
Since the socket address is just being used as a unique identifier, its
inode number is an alternative that does not leak potentially sensitive
information.
CC-ing stable because MITRE has assigned CVE-2010-4565 to the issue.
Signed-off-by: Dan Rosenberg <drosenberg@vsecurity.com> Acked-by: Oliver Hartkopp <socketcan@hartkopp.net> Signed-off-by: David S. Miller <davem@davemloft.net> Cc: Moritz Muehlenhoff <jmm@debian.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
projects / karo-tx-linux.git / log