Signed-off-by: Alexander Potapenko <glider@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: Willy Tarreau <w@1wt.eu>
When we need to move xattrs into external xattr block, we call
ext4_xattr_block_set() from ext4_expand_extra_isize_ea(). That may end
up calling ext4_mark_inode_dirty() again which will recurse back into
the inode expansion code leading to deadlocks.
Protect from recursion using EXT4_STATE_NO_EXPAND inode flag and move
its management into ext4_expand_extra_isize_ea() since its manipulation
is safe there (due to xattr_sem) from possible races with
ext4_xattr_set_handle() which plays with it as well.
CC: stable@vger.kernel.org # 4.4.x Signed-off-by: Jan Kara <jack@suse.cz> Signed-off-by: Theodore Ts'o <tytso@mit.edu> Signed-off-by: Willy Tarreau <w@1wt.eu>
In the ext4 implementations of SEEK_HOLE and SEEK_DATA, make sure we
return -ENXIO for negative offsets instead of banging around inside
the extent code and returning -EFSCORRUPTED.
Reported-by: Mateusz S <muttdini@gmail.com> Signed-off-by: Darrick J. Wong <darrick.wong@oracle.com> Signed-off-by: Theodore Ts'o <tytso@mit.edu> Cc: stable@vger.kernel.org # 4.6 Signed-off-by: Willy Tarreau <w@1wt.eu>
If one cpu is doing nf_ct_extend_unregister while another cpu is doing
__nf_ct_ext_add_length, then we may hit BUG_ON(t == NULL). Moreover,
there's no synchronize_rcu invocation after set nf_ct_ext_types[id] to
NULL, so it's possible that we may access invalid pointer.
But actually, most of the ct extends are built-in, so the problem listed
above will not happen. However, there are two exceptions: NF_CT_EXT_NAT
and NF_CT_EXT_SYNPROXY.
For _EXT_NAT, the panic will not happen, since adding the nat extend and
unregistering the nat extend are located in the same file(nf_nat_core.c),
this means that after the nat module is removed, we cannot add the nat
extend too.
For _EXT_SYNPROXY, synproxy extend may be added by init_conntrack, while
synproxy extend unregister will be done by synproxy_core_exit. So after
nf_synproxy_core.ko is removed, we may still try to add the synproxy
extend, then kernel panic may happen.
I know it's very hard to reproduce this issue, but I can play a tricky
game to make it happen very easily :)
Step 1. Enable SYNPROXY for tcp dport 1234 at FORWARD hook:
# iptables -I FORWARD -p tcp --dport 1234 -j SYNPROXY
Step 2. Queue the syn packet to the userspace at raw table OUTPUT hook.
Also note, in the userspace we only add a 20s' delay, then
reinject the syn packet to the kernel:
# iptables -t raw -I OUTPUT -p tcp --syn -j NFQUEUE --queue-num 1
Step 3. Using "nc 2.2.2.2 1234" to connect the server.
Step 4. Now remove the nf_synproxy_core.ko quickly:
# iptables -F FORWARD
# rmmod ipt_SYNPROXY
# rmmod nf_synproxy_core
Step 5. After 20s' delay, the syn packet is reinjected to the kernel.
Now you will see the panic like this:
kernel BUG at net/netfilter/nf_conntrack_extend.c:91!
Call Trace:
? __nf_ct_ext_add_length+0x53/0x3c0 [nf_conntrack]
init_conntrack+0x12b/0x600 [nf_conntrack]
nf_conntrack_in+0x4cc/0x580 [nf_conntrack]
ipv4_conntrack_local+0x48/0x50 [nf_conntrack_ipv4]
nf_reinject+0x104/0x270
nfqnl_recv_verdict+0x3e1/0x5f9 [nfnetlink_queue]
? nfqnl_recv_verdict+0x5/0x5f9 [nfnetlink_queue]
? nla_parse+0xa0/0x100
nfnetlink_rcv_msg+0x175/0x6a9 [nfnetlink]
[...]
One possible solution is to make NF_CT_EXT_SYNPROXY extend built-in, i.e.
introduce nf_conntrack_synproxy.c and only do ct extend register and
unregister in it, similar to nf_conntrack_timeout.c.
But having such a obscure restriction of nf_ct_extend_unregister is not a
good idea, so we should invoke synchronize_rcu after set nf_ct_ext_types
to NULL, and check the NULL pointer when do __nf_ct_ext_add_length. Then
it will be easier if we add new ct extend in the future.
Last, we use kfree_rcu to free nf_ct_ext, so rcu_barrier() is unnecessary
anymore, remove it too.
core_kernel_text is used by MIPS in its function graph trace processing,
so having this method traced leads to an infinite set of recursive calls
such as:
Similar to the fix provided by Dominik Heidler in commit 9b3dc0a17d73 ("l2tp: cast l2tp traffic counter to unsigned")
we need to take care of 32bit kernels in dev_get_stats().
When using atomic_long_read(), we add a 'long' to u64 and
might misinterpret high order bit, unless we cast to unsigned.
Fixes: caf586e5f23ce ("net: add a core netdev->rx_dropped counter") Fixes: 015f0688f57ca ("net: net: add a core netdev->tx_dropped counter") Fixes: 6e7333d315a76 ("net: add rx_nohandler stat counter") Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Jarod Wilson <jarod@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Willy Tarreau <w@1wt.eu>
Currently, SEEK_HOLE implementation in ext4 may both return that there's
a hole at some offset although that offset already has data and skip
some holes during a search for the next hole. The first problem is
demostrated by:
xfs_io -c "falloc 0 256k" -c "pwrite 0 56k" -c "seek -h 0" file
wrote 57344/57344 bytes at offset 0
56 KiB, 14 ops; 0.0000 sec (2.054 GiB/sec and 538461.5385 ops/sec)
Whence Result
HOLE 0
Where we can see that SEEK_HOLE wrongly returned offset 0 as containing
a hole although we have written data there. The second problem can be
demonstrated by:
wrote 57344/57344 bytes at offset 0
56 KiB, 14 ops; 0.0000 sec (1.978 GiB/sec and 518518.5185 ops/sec)
wrote 8192/8192 bytes at offset 131072
8 KiB, 2 ops; 0.0000 sec (2 GiB/sec and 500000.0000 ops/sec)
Whence Result
HOLE 139264
Where we can see that hole at offsets 56k..128k has been ignored by the
SEEK_HOLE call.
The underlying problem is in the ext4_find_unwritten_pgoff() which is
just buggy. In some cases it fails to update returned offset when it
finds a hole (when no pages are found or when the first found page has
higher index than expected), in some cases conditions for detecting hole
are just missing (we fail to detect a situation where indices of
returned pages are not contiguous).
Fix ext4_find_unwritten_pgoff() to properly detect non-contiguous page
indices and also handle all cases where we got less pages then expected
in one place and handle it properly there.
Fixes: c8c0df241cc2719b1262e627f999638411934f60 CC: Zheng Liu <wenqing.lz@taobao.com> Signed-off-by: Jan Kara <jack@suse.cz> Signed-off-by: Theodore Ts'o <tytso@mit.edu> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: Willy Tarreau <w@1wt.eu>
When using get_options() it's possible to specify a range of numbers,
like 1-100500. The problem is that it doesn't track array size while
calling internally to get_range() which iterates over the range and
fills the memory with numbers.
Link: http://lkml.kernel.org/r/2613C75C-B04D-4BFF-82A6-12F97BA0F620@gmail.com Signed-off-by: Ilya V. Matveychikov <matvejchikov@gmail.com> Cc: Jonathan Corbet <corbet@lwn.net> Cc: <stable@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Willy Tarreau <w@1wt.eu>
Commit 57c67df(md/raid10: submit IO from originating thread instead of
md thread) submits bio directly for normal disks but not for replacement
disks. There is no point we shouldn't do this for replacement disks.
Otherwise, another CPU may access the invalid pointer. For example:
CPU0 CPU1
- rcu_read_lock();
- pfunc = _hook_;
_hook_ = NULL; -
mod unload -
- pfunc(); // invalid, panic
- rcu_read_unlock();
So we must call synchronize_rcu() to wait the rcu reader to finish.
Also note, in nf_nat_snmp_basic_fini, synchronize_rcu() will be invoked
by later nf_conntrack_helper_unregister, but I'm inclined to add a
explicit synchronize_rcu after set the nf_nat_snmp_hook to NULL. Depend
on such obscure assumptions is not a good idea.
Last, in nfnetlink_cttimeout, we use kfree_rcu to free the time object,
so in cttimeout_exit, invoking rcu_barrier() is not necessary at all,
remove it too.
digsig_verify() requests a user key, then accesses its payload.
However, a revoked key has a NULL payload, and we failed to check for
this. request_key() *does* skip revoked keys, but there is still a
window where the key can be revoked before we acquire its semaphore.
Fix it by checking for a NULL payload, treating it like a key which was
already revoked at the time it was requested.
Fixes: 051dbb918c7f ("crypto: digital signature verification support") Reviewed-by: James Morris <james.l.morris@oracle.com> Cc: <stable@vger.kernel.org> [v3.3+] Cc: Dmitry Kasatkin <dmitry.kasatkin@intel.com> Signed-off-by: Eric Biggers <ebiggers@google.com> Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Willy Tarreau <w@1wt.eu>
bitmap_resize() does not work for file-backed bitmaps.
The buffer_heads are allocated and initialized when
the bitmap is read from the file, but resize doesn't
read from the file, it loads from the internal bitmap.
When it comes time to write the new bitmap, the bh is
non-existent and we crash.
The common case when growing an array involves making the array larger,
and that normally means making the bitmap larger. Doing
that inside the kernel is possible, but would need more code.
It is probably easier to require people who use file-backed
bitmaps to remove them and re-add after a reshape.
So this patch disables the resizing of arrays which have
file-backed bitmaps. This is better than crashing.
Reported-by: Zhilong Liu <zlliu@suse.com> Fixes: d60b479d177a ("md/bitmap: add bitmap_resize function to allow bitmap resizing.") Cc: stable@vger.kernel.org (v3.5+). Signed-off-by: NeilBrown <neilb@suse.com> Signed-off-by: Shaohua Li <shli@fb.com> Signed-off-by: Willy Tarreau <w@1wt.eu>
A key of type "encrypted" references a "master key" which is used to
encrypt and decrypt the encrypted key's payload. However, when we
accessed the master key's payload, we failed to handle the case where
the master key has been revoked, which sets the payload pointer to NULL.
Note that request_key() *does* skip revoked keys, but there is still a
window where the key can be revoked before we acquire its semaphore.
Fix it by checking for a NULL payload, treating it like a key which was
already revoked at the time it was requested.
This was an issue for master keys of type "user" only. Master keys can
also be of type "trusted", but those cannot be revoked.
Fixes: 7e70cb497850 ("keys: add new key-type encrypted") Reviewed-by: James Morris <james.l.morris@oracle.com> Cc: <stable@vger.kernel.org> [v2.6.38+] Cc: Mimi Zohar <zohar@linux.vnet.ibm.com> Cc: David Safford <safford@us.ibm.com> Signed-off-by: Eric Biggers <ebiggers@google.com> Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Willy Tarreau <w@1wt.eu>
This is problematic because these "fake" keyrings won't have the right
permissions. In particular, the user who created them first will own
them and will have full access to them via the possessor permissions,
which can be used to compromise the security of a user's keys:
Fix it by marking user and user session keyrings with a flag
KEY_FLAG_UID_KEYRING. Then, when searching for a user or user session
keyring by name, skip all keyrings that don't have the flag set.
Fixes: 69664cf16af4 ("keys: don't generate user and user session keyrings unless they're accessed") Cc: <stable@vger.kernel.org> [v2.6.26+] Signed-off-by: Eric Biggers <ebiggers@google.com> Signed-off-by: David Howells <dhowells@redhat.com>
[wt: adjust context]
The inline asm retry check in the MIPS_ATOMIC_SET operation of the
sysmips system call has been backwards since commit f1e39a4a616c ("MIPS:
Rewrite sysmips(MIPS_ATOMIC_SET, ...) in C with inline assembler")
merged in v2.6.32, resulting in the non R10000_LLSC_WAR case retrying
until the operation was inatomic, before returning the new value that
was probably just written multiple times instead of the old value.
Invert the branch condition to fix that particular issue.
Fixes: f1e39a4a616c ("MIPS: Rewrite sysmips(MIPS_ATOMIC_SET, ...) in C with inline assembler") Signed-off-by: James Hogan <james.hogan@imgtec.com> Cc: linux-mips@linux-mips.org Cc: stable@vger.kernel.org
Patchwork: https://patchwork.linux-mips.org/patch/16148/ Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Willy Tarreau <w@1wt.eu>
When the file /proc/fs/fscache/objects (available with
CONFIG_FSCACHE_OBJECT_LIST=y) is opened, we request a user key with
description "fscache:objlist", then access its payload. However, a
revoked key has a NULL payload, and we failed to check for this.
request_key() *does* skip revoked keys, but there is still a window
where the key can be revoked before we access its payload.
Fix it by checking for a NULL payload, treating it like a key which was
already revoked at the time it was requested.
Fixes: 4fbf4291aa15 ("FS-Cache: Allow the current state of all objects to be dumped") Reviewed-by: James Morris <james.l.morris@oracle.com> Cc: <stable@vger.kernel.org> [v2.6.32+] Signed-off-by: Eric Biggers <ebiggers@google.com> Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Willy Tarreau <w@1wt.eu>
Currently, when passed a key that already exists, add_key() will call the
key's ->update() method if such exists. But this is heavily broken in the
case where the key is uninstantiated because it doesn't call
__key_instantiate_and_link(). Consequently, it doesn't do most of the
things that are supposed to happen when the key is instantiated, such as
setting the instantiation state, clearing KEY_FLAG_USER_CONSTRUCT and
awakening tasks waiting on it, and incrementing key->user->nikeys.
It also never takes key_construction_mutex, which means that
->instantiate() can run concurrently with ->update() on the same key. In
the case of the "user" and "logon" key types this causes a memory leak, at
best. Maybe even worse, the ->update() methods of the "encrypted" and
"trusted" key types actually just dereference a NULL pointer when passed an
uninstantiated key.
Change key_create_or_update() to wait interruptibly for the key to finish
construction before continuing.
This patch only affects *uninstantiated* keys. For now we still allow a
negatively instantiated key to be updated (thereby positively
instantiating it), although that's broken too (the next patch fixes it)
and I'm not sure that anyone actually uses that functionality either.
Here is a simple reproducer for the bug using the "encrypted" key type
(requires CONFIG_ENCRYPTED_KEYS=y), though as noted above the bug
pertained to more than just the "encrypted" key type:
Make sure to reset the USB-console port pointer when console setup fails
in order to avoid having the struct usb_serial be prematurely freed by
the console code when the device is later disconnected.
In the code added to function submit_page_section by commit b1058b981,
sdio->bio can currently be NULL when calling dio_bio_submit. This then
leads to a NULL pointer access in dio_bio_submit, so check for a NULL
bio in submit_page_section before trying to submit it instead.
Fixes xfstest generic/250 on gfs2.
Cc: stable@vger.kernel.org # v3.10+ Signed-off-by: Andreas Gruenbacher <agruenba@redhat.com> Reviewed-by: Jan Kara <jack@suse.cz> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Willy Tarreau <w@1wt.eu>
This patch fixes an issue that the usbhsf_fifo_clear() is possible
to cause 10 msec delay if the pipe is RX direction and empty because
the FRDY bit will never be set to 1 in such case.
Fixes: e8d548d54968 ("usb: renesas_usbhs: fifo became independent from pipe.") Cc: <stable@vger.kernel.org> # v3.1+ Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com> Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com> Signed-off-by: Willy Tarreau <w@1wt.eu>
This patch fixes an issue that the driver sets the BCLR bit of
{C,Dn}FIFOCTR register to 1 even when it's non-DCP pipe and
the FRDY bit of {C,Dn}FIFOCTR register is set to 1.
Fixes: e8d548d54968 ("usb: renesas_usbhs: fifo became independent from pipe.") Cc: <stable@vger.kernel.org> # v3.1+ Signed-off-by: Yoshihiro Shimoda <yoshihiro.shimoda.uh@renesas.com> Signed-off-by: Felipe Balbi <felipe.balbi@linux.intel.com> Signed-off-by: Willy Tarreau <w@1wt.eu>
At the default trace level, we only trace unsuccessful events including
FSF responses.
zfcp_dbf_hba_fsf_response() only used protocol status and FSF status to
decide on an unsuccessful response. However, this is only one of multiple
possible sources determining a failed struct zfcp_fsf_req.
An FSF request can also "fail" if its response runs into an ERP timeout
or if it gets dismissed because a higher level recovery was triggered
[trace tags "erscf_1" or "erscf_2" in zfcp_erp_strategy_check_fsfreq()].
FSF requests with ERP timeout are:
FSF_QTCB_EXCHANGE_CONFIG_DATA, FSF_QTCB_EXCHANGE_PORT_DATA,
FSF_QTCB_OPEN_PORT_WITH_DID or FSF_QTCB_CLOSE_PORT or
FSF_QTCB_CLOSE_PHYSICAL_PORT for target ports,
FSF_QTCB_OPEN_LUN, FSF_QTCB_CLOSE_LUN.
One example is slow queue processing which can cause follow-on errors,
e.g. FSF_PORT_ALREADY_OPEN after FSF_QTCB_OPEN_PORT_WITH_DID timed out.
In order to see the root cause, we need to see late responses even if the
channel presented them successfully with FSF_PROT_GOOD and FSF_GOOD.
Example trace records formatted with zfcpdbf from the s390-tools package:
Timestamp : ...
Area : REC
Subarea : 00
Level : 1
Exception : -
CPU ID : ..
Caller : ...
Record ID : 1
Tag : fcegpf1
LUN : 0xffffffffffffffff
WWPN : 0x<WWPN>
D_ID : 0x00<D_ID>
Adapter status : 0x5400050b
Port status : 0x41200000
LUN status : 0x00000000
Ready count : 0x00000001
Running count : 0x...
ERP want : 0x02 ZFCP_ERP_ACTION_REOPEN_PORT
ERP need : 0x02 ZFCP_ERP_ACTION_REOPEN_PORT
|
Timestamp : ... 30 seconds later
Area : REC
Subarea : 00
Level : 1
Exception : -
CPU ID : ..
Caller : ...
Record ID : 2
Tag : erscf_2
LUN : 0xffffffffffffffff
WWPN : 0x<WWPN>
D_ID : 0x00<D_ID>
Adapter status : 0x5400050b
Port status : 0x41200000
LUN status : 0x00000000
Request ID : 0x<request_ID>
ERP status : 0x10000000 ZFCP_STATUS_ERP_TIMEDOUT
ERP step : 0x0800 ZFCP_ERP_STEP_PORT_OPENING
ERP action : 0x02 ZFCP_ERP_ACTION_REOPEN_PORT
ERP count : 0x00
|
Timestamp : ... later than previous record
Area : HBA
Subarea : 00
Level : 5 > default level => 3 <= default level
Exception : -
CPU ID : 00
Caller : ...
Record ID : 1
Tag : fs_qtcb => fs_rerr
Request ID : 0x<request_ID>
Request status : 0x00001010 ZFCP_STATUS_FSFREQ_DISMISSED
| ZFCP_STATUS_FSFREQ_CLEANUP
FSF cmnd : 0x00000005
FSF sequence no: 0x...
FSF issued : ... > 30 seconds ago
FSF stat : 0x00000000 FSF_GOOD
FSF stat qual : 00000000000000000000000000000000
Prot stat : 0x00000001 FSF_PROT_GOOD
Prot stat qual : 00000000000000000000000000000000
Port handle : 0x...
LUN handle : 0x00000000
QTCB log length: ...
QTCB log info : ...
In case of problems detecting that new responses are waiting on the input
queue, we sooner or later trigger adapter recovery due to an FSF request
timeout (trace tag "fsrth_1").
FSF requests with FSF request timeout are:
typically FSF_QTCB_ABORT_FCP_CMND; but theoretically also
FSF_QTCB_EXCHANGE_CONFIG_DATA or FSF_QTCB_EXCHANGE_PORT_DATA via sysfs,
FSF_QTCB_OPEN_PORT_WITH_DID or FSF_QTCB_CLOSE_PORT for WKA ports,
FSF_QTCB_FCP_CMND for task management function (LUN / target reset).
One or more pending requests can meanwhile have FSF_PROT_GOOD and FSF_GOOD
because the channel filled in the response via DMA into the request's QTCB.
In a theroretical case, inject code can create an erroneous FSF request
on purpose. If data router is enabled, it uses deferred error reporting.
A READ SCSI command can succeed with FSF_PROT_GOOD, FSF_GOOD, and
SAM_STAT_GOOD. But on writing the read data to host memory via DMA,
it can still fail, e.g. if an intentionally wrong scatter list does not
provide enough space. Rather than getting an unsuccessful response,
we get a QDIO activate check which in turn triggers adapter recovery.
One or more pending requests can meanwhile have FSF_PROT_GOOD and FSF_GOOD
because the channel filled in the response via DMA into the request's QTCB.
Example trace records formatted with zfcpdbf from the s390-tools package:
Timestamp : ...
Area : HBA
Subarea : 00
Level : 6 > default level => 3 <= default level
Exception : -
CPU ID : ..
Caller : ...
Record ID : 1
Tag : fs_norm => fs_rerr
Request ID : 0x<request_ID2>
Request status : 0x00001010 ZFCP_STATUS_FSFREQ_DISMISSED
| ZFCP_STATUS_FSFREQ_CLEANUP
FSF cmnd : 0x00000001
FSF sequence no: 0x...
FSF issued : ...
FSF stat : 0x00000000 FSF_GOOD
FSF stat qual : 00000000000000000000000000000000
Prot stat : 0x00000001 FSF_PROT_GOOD
Prot stat qual : ........ ........ 0000000000000000
Port handle : 0x...
LUN handle : 0x...
|
Timestamp : ...
Area : SCSI
Subarea : 00
Level : 3
Exception : -
CPU ID : ..
Caller : ...
Record ID : 1
Tag : rsl_err
Request ID : 0x<request_ID2>
SCSI ID : 0x...
SCSI LUN : 0x...
SCSI result : 0x000e0000 DID_TRANSPORT_DISRUPTED
SCSI retries : 0x00
SCSI allowed : 0x05
SCSI scribble : 0x<request_ID2>
SCSI opcode : 28... Read(10)
FCP rsp inf cod: 0x00
FCP rsp IU : 00000000000000000000000000000000
^^ SAM_STAT_GOOD 0000000000000000
Only with luck in both above cases, we could see a follow-on trace record
of an unsuccesful event following a successful but late FSF response with
FSF_PROT_GOOD and FSF_GOOD. Typically this was the case for I/O requests
resulting in a SCSI trace record "rsl_err" with DID_TRANSPORT_DISRUPTED
[On ZFCP_STATUS_FSFREQ_DISMISSED, zfcp_fsf_protstatus_eval() sets
ZFCP_STATUS_FSFREQ_ERROR seen by the request handler functions as failure].
However, the reason for this follow-on trace was invisible because the
corresponding HBA trace record was missing at the default trace level
(by default hidden records with tags "fs_norm", "fs_qtcb", or "fs_open").
On adapter recovery, after we had shut down the QDIO queues, we perform
unsuccessful pseudo completions with flag ZFCP_STATUS_FSFREQ_DISMISSED
for each pending FSF request in zfcp_fsf_req_dismiss_all().
In order to find the root cause, we need to see all pseudo responses even
if the channel presented them successfully with FSF_PROT_GOOD and FSF_GOOD.
Therefore, check zfcp_fsf_req.status for ZFCP_STATUS_FSFREQ_DISMISSED
or ZFCP_STATUS_FSFREQ_ERROR and trace with a new tag "fs_rerr".
It does not matter that there are numerous places which set
ZFCP_STATUS_FSFREQ_ERROR after the location where we trace an FSF response
early. These cases are based on protocol status != FSF_PROT_GOOD or
== FSF_PROT_FSF_STATUS_PRESENTED and are thus already traced by default
as trace tag "fs_perr" or "fs_ferr" respectively.
NB: The trace record with tag "fssrh_1" for status read buffers on dismiss
all remains. zfcp_fsf_req_complete() handles this and returns early.
All other FSF request types are handled separately and as described above.
Signed-off-by: Steffen Maier <maier@linux.vnet.ibm.com> Fixes: 8a36e4532ea1 ("[SCSI] zfcp: enhancement of zfcp debug features") Fixes: 2e261af84cdb ("[SCSI] zfcp: Only collect FSF/HBA debug data for matching trace levels") Cc: <stable@vger.kernel.org> #2.6.38+ Reviewed-by: Benjamin Block <bblock@linux.vnet.ibm.com> Signed-off-by: Benjamin Block <bblock@linux.vnet.ibm.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Willy Tarreau <w@1wt.eu>
If the FCP_RSP UI has optional parts (FCP_SNS_INFO or FCP_RSP_INFO) and
thus does not fit into the fsp_rsp field built into a SCSI trace record,
trace the full FCP_RSP UI with all optional parts as payload record
instead of just FCP_SNS_INFO as payload and
a 1 byte RSP_INFO_CODE part of FCP_RSP_INFO built into the SCSI record.
That way we would also get the full FCP_SNS_INFO in case a
target would ever send more than
min(SCSI_SENSE_BUFFERSIZE==96, ZFCP_DBF_PAY_MAX_REC==256)==96.
The mandatory part of FCP_RSP IU is only 24 bytes.
PAYload costs at least one full PAY record of 256 bytes anyway.
We cap to the hardware response size which is only FSF_FCP_RSP_SIZE==128.
So we can just put the whole FCP_RSP IU with any optional parts into
PAYload similarly as we do for SAN PAY since v4.9 commit aceeffbb59bb
("zfcp: trace full payload of all SAN records (req,resp,iels)").
This does not cause any additional trace records wasting memory.
Decoded trace records were confusing because they showed a hard-coded
sense data length of 96 even if the FCP_RSP_IU field FCP_SNS_LEN showed
actually less.
Since the same commit, we set pl_len for SAN traces to the full length of a
request/response even if we cap the corresponding trace.
In contrast, here for SCSI traces we set pl_len to the pre-computed
length of FCP_RSP IU considering SNS_LEN or RSP_LEN if valid.
Nonetheless we trace a hardcoded payload of length FSF_FCP_RSP_SIZE==128
if there were optional parts.
This makes it easier for the zfcpdbf tool to format only the relevant
part of the long FCP_RSP UI buffer. And any trailing information is still
available in the payload trace record just in case.
Rename the payload record tag from "fcp_sns" to "fcp_riu" to make the new
content explicit to zfcpdbf which can then pick a suitable field name such
as "FCP rsp IU all:" instead of "Sense info :"
Also, the same zfcpdbf can still be backwards compatible with "fcp_sns".
Old example trace record before this fix, formatted with the tool zfcpdbf
from s390-tools:
For problem determination we need to see that we were in scsi_eh
as well as whether and why we were successful or not.
The following commits introduced new early returns without adding
a trace record:
v2.6.35 commit a1dbfddd02d2
("[SCSI] zfcp: Pass return code from fc_block_scsi_eh to scsi eh")
on fc_block_scsi_eh() returning != 0 which is FAST_IO_FAIL,
v2.6.30 commit 63caf367e1c9
("[SCSI] zfcp: Improve reliability of SCSI eh handlers in zfcp")
on not having gotten an FSF request after the maximum number of retry
attempts and thus could not issue a TMF and has to return FAILED.
Signed-off-by: Steffen Maier <maier@linux.vnet.ibm.com> Fixes: a1dbfddd02d2 ("[SCSI] zfcp: Pass return code from fc_block_scsi_eh to scsi eh") Fixes: 63caf367e1c9 ("[SCSI] zfcp: Improve reliability of SCSI eh handlers in zfcp") Cc: <stable@vger.kernel.org> #2.6.38+ Reviewed-by: Benjamin Block <bblock@linux.vnet.ibm.com> Signed-off-by: Benjamin Block <bblock@linux.vnet.ibm.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Willy Tarreau <w@1wt.eu>
Up until now zfcp would just ignore the FCP_RESID_OVER flag in the FCP
response IU. When this flag is set, it is possible, in regards to the
FCP standard, that the storage-server processes the command normally, up
to the point where data is missing and simply ignores those.
In this case no CHECK CONDITION would be set, and because we ignored the
FCP_RESID_OVER flag we resulted in at least a data loss or even
-corruption as a follow-up error, depending on how the
applications/layers on top behave. To prevent this, we now set the
host-byte of the corresponding scsi_cmnd to DID_ERROR.
Other storage-behaviors, where the same condition results in a CHECK
CONDITION set in the answer, don't need to be changed as they are
handled in the mid-layer already.
Following is an example trace record decoded with zfcpdbf from the
s390-tools package. We forcefully injected a fc_dl which is one byte too
small:
Timestamp : ...
Area : SCSI
Subarea : 00
Level : 3
Exception : -
CPU ID : ..
Caller : 0x...
Record ID : 1
Tag : rsl_err
Request ID : 0x...
SCSI ID : 0x...
SCSI LUN : 0x...
SCSI result : 0x00070000
^^DID_ERROR
SCSI retries : 0x..
SCSI allowed : 0x..
SCSI scribble : 0x...
SCSI opcode : 2a000000000000000800000000000000
FCP rsp inf cod: 0x00
FCP rsp IU : 00000000000000000000040000000001
^^fr_flags==FCP_RESID_OVER
^^fr_status==SAM_STAT_GOOD
^^^^^^^^fr_resid 0000000000000000
As of now, we don't actively handle to possibility that a response IU
has both flags - FCP_RESID_OVER and FCP_RESID_UNDER - set at once.
Reported-by: Luke M. Hopkins <lmhopkin@us.ibm.com> Reviewed-by: Steffen Maier <maier@linux.vnet.ibm.com> Fixes: 553448f6c483 ("[SCSI] zfcp: Message cleanup") Fixes: ea127f975424 ("[PATCH] s390 (7/7): zfcp host adapter.") (tglx/history.git) Cc: <stable@vger.kernel.org> #2.6.33+ Signed-off-by: Benjamin Block <bblock@linux.vnet.ibm.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Willy Tarreau <w@1wt.eu>
Since commit db007fc5e20c ("[SCSI] Command protection operation"),
scsi_eh_prep_cmnd() saves scmd->prot_op and temporarily resets it to
SCSI_PROT_NORMAL.
Other FCP LLDDs such as qla2xxx and lpfc shield their queuecommand()
to only access any of scsi_prot_sg...() if
(scsi_get_prot_op(cmd) != SCSI_PROT_NORMAL).
Do the same thing for zfcp, which introduced DIX support with
commit ef3eb71d8ba4 ("[SCSI] zfcp: Introduce experimental support for
DIF/DIX").
Otherwise, TUR SCSI commands as part of scsi_eh likely fail in zfcp,
because the regular SCSI command with DIX protection data, that scsi_eh
re-uses in scsi_send_eh_cmnd(), of course still has
(scsi_prot_sg_count() != 0) and so zfcp sends down bogus requests to the
FCP channel hardware.
This causes scsi_eh_test_devices() to have (finish_cmds == 0)
[not SCSI device is online or not scsi_eh_tur() failed]
so regular SCSI commands, that caused / were affected by scsi_eh,
are moved to work_q and scsi_eh_test_devices() itself returns false.
In turn, it unnecessarily escalates in our case in scsi_eh_ready_devs()
beyond host reset to finally scsi_eh_offline_sdevs()
which sets affected SCSI devices offline with the following kernel message:
"kernel: sd H:0:T:L: Device offlined - not ready after error recovery"
Signed-off-by: Steffen Maier <maier@linux.vnet.ibm.com> Fixes: ef3eb71d8ba4 ("[SCSI] zfcp: Introduce experimental support for DIF/DIX") Cc: <stable@vger.kernel.org> #2.6.36+ Reviewed-by: Benjamin Block <bblock@linux.vnet.ibm.com> Signed-off-by: Benjamin Block <bblock@linux.vnet.ibm.com> Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: Willy Tarreau <w@1wt.eu>
Before the patch, the flock flag could remain uninitialized for the
lifespan of the fuse_file allocation. Unless set to true in
fuse_file_flock(), it would remain in an indeterminate state until read in
an if statement in fuse_release_common(). This could consequently lead to
taking an unexpected branch in the code.
The bug was discovered by a runtime instrumentation designed to detect use
of uninitialized memory in the kernel.
This patch closes a race between se_lun deletion during configfs
unlink in target_fabric_port_unlink() -> core_dev_del_lun()
-> core_tpg_remove_lun(), when transport_clear_lun_ref() blocks
waiting for percpu_ref RCU grace period to finish, but a new
NodeACL mappedlun is added before the RCU grace period has
completed.
This can happen in target_fabric_mappedlun_link() because it
only checks for se_lun->lun_se_dev, which is not cleared until
after transport_clear_lun_ref() percpu_ref RCU grace period
finishes.
This bug originally manifested as NULL pointer dereference
OOPsen in target_stat_scsi_att_intr_port_show_attr_dev() on
v4.1.y code, because it dereferences lun->lun_se_dev without
a explicit NULL pointer check.
In post v4.1 code with target-core RCU conversion, the code
in target_stat_scsi_att_intr_port_show_attr_dev() no longer
uses se_lun->lun_se_dev, but the same race still exists.
To address the bug, go ahead and set se_lun>lun_shutdown as
early as possible in core_tpg_remove_lun(), and ensure new
NodeACL mappedlun creation in target_fabric_mappedlun_link()
fails during se_lun shutdown.
Reported-by: James Shen <jcs@datera.io> Cc: James Shen <jcs@datera.io> Tested-by: James Shen <jcs@datera.io> Cc: stable@vger.kernel.org # 3.10+ Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org> Signed-off-by: Willy Tarreau <w@1wt.eu>
this patch makes sure VPFE_CMD_S_CCDC_RAW_PARAMS ioctl no longer works
for vpfe_capture driver with a minimal patch suitable for backporting.
- This ioctl was never in public api and was only defined in kernel header.
- The function set_params constantly mixes up pointers and phys_addr_t
numbers.
- This is part of a 'VPFE_CMD_S_CCDC_RAW_PARAMS' ioctl command that is
described as an 'experimental ioctl that will change in future kernels'.
- The code to allocate the table never gets called after we copy_from_user
the user input over the kernel settings, and then compare them
for inequality.
- We then go on to use an address provided by user space as both the
__user pointer for input and pass it through phys_to_virt to come up
with a kernel pointer to copy the data to. This looks like a trivially
exploitable root hole.
Due to these reasons we make sure this ioctl now returns -EINVAL and backport
this patch as far as possible.
Fixes: 5f15fbb68fd7 ("V4L/DVB (12251): v4l: dm644x ccdc module for vpfe capture driver") Signed-off-by: Lad, Prabhakar <prabhakar.csengg@gmail.com> Cc: <stable@vger.kernel.org> # for v3.7 and up Signed-off-by: Hans Verkuil <hans.verkuil@cisco.com> Signed-off-by: Mauro Carvalho Chehab <mchehab@s-opensource.com> Signed-off-by: Willy Tarreau <w@1wt.eu>
On a 32-bit platform, the value of n_blcoks_count may be wrong during
the file system is resized to size larger than 2^32 blocks. This may
caused the superblock being corrupted with zero blocks count.
Fixes: 1c6bd7173d66 Signed-off-by: Jerry Lee <jerrylee@qnap.com> Signed-off-by: Theodore Ts'o <tytso@mit.edu> Cc: stable@vger.kernel.org # 3.7+ Signed-off-by: Willy Tarreau <w@1wt.eu>
ext4_find_unwritten_pgoff() does not properly handle a situation when
starting index is in the middle of a page and blocksize < pagesize. The
following command shows the bug on filesystem with 1k blocksize:
The combination of WQ_UNBOUND and max_active == 1 used to imply
ordered execution. After NUMA affinity 4c16bd327c74 ("workqueue:
implement NUMA affinity for unbound workqueues"), this is no longer
true due to per-node worker pools.
While the right way to create an ordered workqueue is
alloc_ordered_workqueue(), the documentation has been misleading for a
long time and people do use WQ_UNBOUND and max_active == 1 for ordered
workqueues which can lead to subtle bugs which are very difficult to
trigger.
It's unlikely that we'd see noticeable performance impact by enforcing
ordering on WQ_UNBOUND / max_active == 1 workqueues. Let's
automatically set __WQ_ORDERED for those workqueues.
Signed-off-by: Tejun Heo <tj@kernel.org> Reported-by: Christoph Hellwig <hch@infradead.org> Reported-by: Alexei Potashnik <alexei@purestorage.com> Fixes: 4c16bd327c74 ("workqueue: implement NUMA affinity for unbound workqueues") Cc: stable@vger.kernel.org # v3.10+ Signed-off-by: Willy Tarreau <w@1wt.eu>
My static checker complains that "devno" can be negative, meaning that
we read before the start of the loop. I've looked at the code, and I
think the warning is right. This come from /proc so it's root only or
it would be quite a quite a serious bug. The call tree looks like this:
proc_scsi_write() <- gets id and channel from simple_strtoul()
-> scsi_add_single_device() <- calls shost->transportt->user_scan()
-> ata_scsi_user_scan()
-> ata_find_dev()
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: Tejun Heo <tj@kernel.org> Cc: stable@vger.kernel.org # all versions at this point Signed-off-by: Willy Tarreau <w@1wt.eu>
Terminate FPU emulation immediately whenever an ISA mode switch has been
observed. This is so that we do not interpret machine code in the wrong
mode, for example when a regular MIPS FPU instruction has been placed in
a delay slot of a jump that switches into the MIPS16 mode, as with the
following code (taken from a GCC test suite case):
Here `set_fast_math' is called from `40130e' (`40130f' with the ISA bit)
and emulation triggers for the CTC1 instruction. As it is in a jump
delay slot emulation continues from `401312' (`401313' with the ISA
bit). However we have no path to handle MIPS16 FPU code emulation,
because there are no MIPS16 FPU instructions. So the default emulation
path is taken, interpreting a 32-bit word fetched by `get_user' from
`401313' as a regular MIPS instruction, which is:
This makes the FPU emulator proceed with the supposed SDC1 instruction
and consequently makes the program considered here terminate with
SIGSEGV.
A similar although less severe issue exists with pure-microMIPS
processors in the case where similarly an FPU instruction is emulated in
a delay slot of a register jump that (incorrectly) switches into the
regular MIPS mode. A subsequent instruction fetch from the jump's
target is supposed to cause an Address Error exception, however instead
we proceed with regular MIPS FPU emulation.
For simplicity then, always terminate the emulation loop whenever a mode
change is detected, denoted by an ISA mode bit flip. As from commit 377cb1b6c16a ("MIPS: Disable MIPS16/microMIPS crap for platforms not
supporting these ASEs.") the result of `get_isa16_mode' can be hardcoded
to 0, so we need to examine the ISA mode bit by hand.
This complements commit 102cedc32a6e ("MIPS: microMIPS: Floating point
support.") which added JALX decoding to FPU emulation.
Fixes: 102cedc32a6e ("MIPS: microMIPS: Floating point support.") Signed-off-by: Maciej W. Rozycki <macro@imgtec.com> Cc: James Hogan <james.hogan@imgtec.com> Cc: linux-mips@linux-mips.org Cc: stable@vger.kernel.org # 3.9+
Patchwork: https://patchwork.linux-mips.org/patch/16393/ Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Willy Tarreau <w@1wt.eu>
Fix a regression introduced with commit fb6883e5809c ("MIPS: microMIPS:
Support handling of delay slots.") and defer to `__compute_return_epc'
if the ISA bit is set in EPC with non-MIPS16, non-microMIPS hardware,
which will then arrange for a SIGBUS due to an unaligned instruction
reference. Returning EPC here is never correct as the API defines this
function's result to be either a negative error code on failure or one
of 0 and BRANCH_LIKELY_TAKEN on success.
Fixes: fb6883e5809c ("MIPS: microMIPS: Support handling of delay slots.") Signed-off-by: Maciej W. Rozycki <macro@imgtec.com> Cc: James Hogan <james.hogan@imgtec.com> Cc: linux-mips@linux-mips.org Cc: stable@vger.kernel.org # 3.9+
Patchwork: https://patchwork.linux-mips.org/patch/16395/ Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Willy Tarreau <w@1wt.eu>
Complement commit fb6883e5809c ("MIPS: microMIPS: Support handling of
delay slots.") and actually decode the regular MIPS JALX major
instruction opcode, the handling of which has been added with the said
commit for EPC calculation in `__compute_return_epc_for_insn'.
Fixes: fb6883e5809c ("MIPS: microMIPS: Support handling of delay slots.") Signed-off-by: Maciej W. Rozycki <macro@imgtec.com> Cc: James Hogan <james.hogan@imgtec.com> Cc: linux-mips@linux-mips.org Cc: stable@vger.kernel.org # 3.9+
Patchwork: https://patchwork.linux-mips.org/patch/16394/ Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Willy Tarreau <w@1wt.eu>
This patch fixes an issue that some registers may be not initialized
after resume if the USBHSF_RUNTIME_PWCTRL is not set. Otherwise,
if a cable is not connected, the driver will not enable INTENB0.VBSE
after resume. And then, the driver cannot detect the VBUS.
The workaround for the CELL timebase bug does not correctly mark cr0 as
being clobbered. This means GCC doesn't know that the asm block changes cr0 and
might leave the result of an unrelated comparison in cr0 across the block, which
we then trash, leading to basically random behaviour.
Fixes: 859deea949c3 ("[POWERPC] Cell timebase bug workaround") Cc: stable@vger.kernel.org # v2.6.19+ Signed-off-by: Oliver O'Halloran <oohall@gmail.com>
[mpe: Tweak change log and flag for stable] Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Signed-off-by: Willy Tarreau <w@1wt.eu>
From POWER4 onwards, mfocrf() only places the specified CR field into
the destination GPR, and the rest of it is set to 0. The PowerPC AS
from version 3.0 now requires this behaviour.
The emulation code currently puts the entire CR into the destination GPR.
Fix it.
Fixes: 6888199f7fe5 ("[POWERPC] Emulate more instructions in software") Cc: stable@vger.kernel.org # v2.6.22+ Signed-off-by: Anton Blanchard <anton@samba.org> Acked-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Signed-off-by: Willy Tarreau <w@1wt.eu>
Although it's not documented anywhere, there is an expectation that
atomic64_inc_not_zero() returns a result which fits in an int. This is
the behaviour implemented on all arches except powerpc.
This has caused at least one bug in practice, in the percpu-refcount
code, where the long result from our atomic64_inc_not_zero() was
truncated to an int leading to lost references and stuck systems. That
was worked around in that code in commit 966d2b04e070 ("percpu-refcount:
fix reference leak during percpu-atomic transition").
To the best of my grepping abilities there are no other callers
in-tree which truncate the value, but we should fix it anyway. Because
the breakage is subtle and potentially very harmful I'm also tagging
it for stable.
Code generation is largely unaffected because in most cases the
callers are just using the result for a test anyway. In particular the
case of fget() that was mentioned in commit a6cf7ed5119f
("powerpc/atomic: Implement atomic*_inc_not_zero") generates exactly
the same code.
An updated patch that also handles the additional key length requirements
for the AEAD algorithms.
The max keysize is not 96. For SHA384/512 it's 128, and for the AEAD
algorithms it's longer still. Extend the max keysize for the
AEAD size for AES256 + HMAC(SHA512).
Cc: <stable@vger.kernel.org> # 3.6+ Fixes: 357fb60502ede ("crypto: talitos - add sha224, sha384 and sha512 to existing AEAD algorithms") Signed-off-by: Martin Hicks <mort@bork.org> Acked-by: Horia Geantă <horia.geanta@nxp.com> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: Willy Tarreau <w@1wt.eu>
Actually, at the moment this is not an issue, as every in-tree arch does
the same manual checks for VERIFY_READ vs VERIFY_WRITE, relying on the MMU
to tell them apart, but this wasn't the case in the past and may happen
again on some odd arch in the future.
If anyone cares about 3.7 and earlier, this is a security hole (untested)
on real 80386 CPUs.
The lower level nl80211 code in cfg80211 ensures that "len" is between
25 and NL80211_ATTR_FRAME (2304). We subtract DOT11_MGMT_HDR_LEN (24) from
"len" so thats's max of 2280. However, the action_frame->data[] buffer is
only BRCMF_FIL_ACTION_FRAME_SIZE (1800) bytes long so this memcpy() can
overflow.
There is a clean-up bug in the core comedi module initialization
functions, `comedi_init()`. If the `comedi_num_legacy_minors` module
parameter is non-zero (and valid), it creates that many "legacy" devices
and registers them in SysFS. A failure causes the function to clean up
and return an error. Unfortunately, it fails to destroy the "comedi"
class that was created earlier. Fix it by adding a call to
`class_destroy(comedi_class)` at the appropriate place in the clean-up
sequence.
This fixes a crash when function_graph and jprobes are used together.
This is essentially commit 237d28db036e ("ftrace/jprobes/x86: Fix
conflict between jprobes and function graph tracing"), but for powerpc.
Jprobes breaks function_graph tracing since the jprobe hook needs to use
jprobe_return(), which never returns back to the hook, but instead to
the original jprobe'd function. The solution is to momentarily pause
function_graph tracing before invoking the jprobe hook and re-enable it
when returning back to the original jprobe'd function.
Fixes: 6794c78243bf ("powerpc64: port of the function graph tracer") Cc: stable@vger.kernel.org # v2.6.30+ Signed-off-by: Naveen N. Rao <naveen.n.rao@linux.vnet.ibm.com> Acked-by: Masami Hiramatsu <mhiramat@kernel.org> Acked-by: Steven Rostedt (VMware) <rostedt@goodmis.org> Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Signed-off-by: Willy Tarreau <w@1wt.eu>
Commit 27ed3cd2ebf4 (cpufreq: conservative: Fix the logic in frequency
decrease checking) removed the 10 point substraction when comparing the
load against down_threshold but did not remove the related limit for the
down_threshold value. As a result, down_threshold lower than 11 is not
allowed even though values from 1 to 10 do work correctly too. The
comment ("cannot be lower than 11 otherwise freq will not fall") also
does not apply after removing the substraction.
For this reason, allow down_threshold to take any value from 1 to 99
and fix the related comment.
Fixes: 27ed3cd2ebf4 (cpufreq: conservative: Fix the logic in frequency decrease checking) Signed-off-by: Tomasz Wilczyński <twilczynski@naver.com> Acked-by: Viresh Kumar <viresh.kumar@linaro.org> Cc: 3.10+ <stable@vger.kernel.org> # 3.10+ Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by: Willy Tarreau <w@1wt.eu>
sys_add_key() and the KEYCTL_UPDATE operation of sys_keyctl() allowed a
NULL payload with nonzero length to be passed to the key type's
->preparse(), ->instantiate(), and/or ->update() methods. Various key
types including asymmetric, cifs.idmap, cifs.spnego, and pkcs7_test did
not handle this case, allowing an unprivileged user to trivially cause a
NULL pointer dereference (kernel oops) if one of these key types was
present. Fix it by doing the copy_from_user() when 'plen' is nonzero
rather than when '_payload' is non-NULL, causing the syscall to fail
with EFAULT as expected when an invalid buffer is specified.
Cc: stable@vger.kernel.org # 2.6.10+ Signed-off-by: Eric Biggers <ebiggers@google.com> Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: James Morris <james.l.morris@oracle.com> Signed-off-by: Willy Tarreau <w@1wt.eu>
Make sure to deregister the SPI driver before releasing the tty driver
to avoid use-after-free in the SPI remove callback where the tty
devices are deregistered.
Fixes: 72d4724ea54c ("serial: ifx6x60: Add modem power off function in the platform reboot process") Cc: stable <stable@vger.kernel.org> # 3.8 Cc: Jun Chen <jun.d.chen@intel.com> Signed-off-by: Johan Hovold <johan@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: Willy Tarreau <w@1wt.eu>
Fix commit e50c0a8fa60d ("Support the MIPS32 / MIPS64 DSP ASE.") and
send SIGILL rather than SIGBUS whenever an unimplemented BPOSGE32 DSP
ASE instruction has been encountered in `__compute_return_epc_for_insn'
as our Reserved Instruction exception handler would in response to an
attempt to actually execute the instruction. Sending SIGBUS only makes
sense for the unaligned PC case, since moved to `__compute_return_epc'.
Adjust function documentation accordingly, correct formatting and use
`pr_info' rather than `printk' as the other exit path already does.
Fixes: e50c0a8fa60d ("Support the MIPS32 / MIPS64 DSP ASE.") Signed-off-by: Maciej W. Rozycki <macro@imgtec.com> Cc: James Hogan <james.hogan@imgtec.com> Cc: linux-mips@linux-mips.org Cc: stable@vger.kernel.org # 2.6.14+
Patchwork: https://patchwork.linux-mips.org/patch/16396/ Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Willy Tarreau <w@1wt.eu>
There is a potential race window opened at creating and deleting a
port via ioctl, as spotted by fuzzing. snd_seq_create_port() creates
a port object and returns its pointer, but it doesn't take the
refcount, thus it can be deleted immediately by another thread.
Meanwhile, snd_seq_ioctl_create_port() still calls the function
snd_seq_system_client_ev_port_start() with the created port object
that is being deleted, and this triggers use-after-free like:
We may fix this in a few different ways, and in this patch, it's fixed
simply by taking the refcount properly at snd_seq_create_port() and
letting the caller unref the object after use. Also, there is another
potential use-after-free by sprintf() call in snd_seq_create_port(),
and this is moved inside the lock.
nl80211_set_rekey_data() does not check if the required attributes
NL80211_REKEY_DATA_{REPLAY_CTR,KEK,KCK} are present when processing
NL80211_CMD_SET_REKEY_OFFLOAD request. This request can be issued by
users with CAP_NET_ADMIN privilege and may result in NULL dereference
and a system crash. Add a check for the required attributes presence.
This patch is based on the patch by bo Zhang.
This fixes CVE-2017-12153.
References: https://bugzilla.redhat.com/show_bug.cgi?id=1491046 Fixes: e5497d766ad ("cfg80211/nl80211: support GTK rekey offload") Cc: <stable@vger.kernel.org> # v3.1-rc1 Reported-by: bo Zhang <zhangbo5891001@gmail.com> Signed-off-by: Vladis Dronov <vdronov@redhat.com> Signed-off-by: Johannes Berg <johannes.berg@intel.com> Signed-off-by: Willy Tarreau <w@1wt.eu>
The 'dir' parameter in xfrm_migrate() is a user-controlled byte which is used
as an array index. This can lead to an out-of-bound access, kernel lockup and
DoS. Add a check for the 'dir' value.
When a kerberos 5 ticket is being decoded so that it can be loaded into an
rxrpc-type key, there are several places in which the length of a
variable-length field is checked to make sure that it's not going to
overrun the available data - but the data is padded to the nearest
four-byte boundary and the code doesn't check for this extra. This could
lead to the size-remaining variable wrapping and the data pointer going
over the end of the buffer.
Fix this by making the various variable-length data checks use the padded
length.
Reported-by: 石磊 <shilei-c@360.cn> Signed-off-by: David Howells <dhowells@redhat.com> Reviewed-by: Marc Dionne <marc.c.dionne@auristor.com> Reviewed-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: Willy Tarreau <w@1wt.eu>
When limiting the argv/envp strings during exec to 1/4 of the stack limit,
the storage of the pointers to the strings was not included. This means
that an exec with huge numbers of tiny strings could eat 1/4 of the stack
limit in strings and then additional space would be later used by the
pointers to the strings.
For example, on 32-bit with a 8MB stack rlimit, an exec with 1677721
single-byte strings would consume less than 2MB of stack, the max (8MB /
4) amount allowed, but the pointers to the strings would consume the
remaining additional stack space (1677721 * 4 == 6710884).
The result (1677721 + 6710884 == 8388605) would exhaust stack space
entirely. Controlling this stack exhaustion could result in
pathological behavior in setuid binaries (CVE-2017-1000365).
[akpm@linux-foundation.org: additional commenting from Kees] Fixes: b6a2fea39318 ("mm: variable length argument support") Link: http://lkml.kernel.org/r/20170622001720.GA32173@beast Signed-off-by: Kees Cook <keescook@chromium.org> Acked-by: Rik van Riel <riel@redhat.com> Acked-by: Michal Hocko <mhocko@suse.com> Cc: Alexander Viro <viro@zeniv.linux.org.uk> Cc: Qualys Security Advisory <qsa@qualys.com> Cc: <stable@vger.kernel.org> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Willy Tarreau <w@1wt.eu>
The DREQE bit of the DnFIFOSEL should be set to 1 after the DE bit of
USB-DMAC on R-Car SoCs is set to 1 after the USB-DMAC received a
zero-length packet. Otherwise, a transfer completion interruption
of USB-DMAC doesn't happen. Even if the driver changes the sequence,
normal operations (transmit/receive without zero-length packet) will
not cause any side-effects. So, this patch fixes the sequence anyway.
Some gadget drivers will call usb_ep_queue() more than once before
the first queue doesn't finish. However, this driver didn't handle
it correctly. So, this patch fixes the behavior of some
usbhs_pkt_handle using the "running" flag. Otherwise, the oops below
happens if we use g_ncm driver and when the "iperf -u -c host -b 200M"
is running.
Al Viro [Sat, 30 Sep 2017 04:28:05 +0000 (05:28 +0100)]
leak in O_DIRECT readv past the EOF
In all versions from 2.5.62 to 3.15, on each iteration through the loop
by iovec array in do_blockdev_direct_IO() we used to do this:
sdio.head = 0;
sdio.tail = 0;
...
retval = do_direct_IO(dio, &sdio, &map_bh);
if (retval) {
dio_cleanup(dio, &sdio);
break;
}
with another dio_cleanup() done after the loop, catching the situation when
retval had been 0. Consider the situation when e.g. the 3rd iovec in 4-iovec
array passed to readv() has crossed the EOF. do_direct_IO() returns 0 and
buggers off *without* exhausting the page array. The loop proceeds to the
next iovec without calling dio_cleanup() and resets sdio.head and sdio.tail.
That reset of sdio.{head,tail} has prevented the eventual dio_cleanup() from
seeing anything and the page reference end up leaking.
Commit 7b2c99d15559 (new helper: iov_iter_get_pages()) in 3.16 had eliminated
the loop by iovec array, along with sdio.head and sdio.tail resets. Backporting
that is too much work - the minimal fix is simply to make sure that the only case
when do_direct_IO() buggers off early without returning non-zero will not skip
dio_cleanup().
The fix applies to all versions from 2.5.62 to 3.15.
Linus suggested we try to remove some of the low-hanging fruit related
to kernel address exposure in dmesg. The only leaks I see on my local
system are:
"I suspect we should just remove [the addresses in the 'Freeing'
messages]. I'm sure they are useful in theory, but I suspect they
were more useful back when the whole "free init memory" was
originally done.
These days, if we have a use-after-free, I suspect the init-mem
situation is the easiest situation by far. Compared to all the dynamic
allocations which are much more likely to show it anyway. So having
debug output for that case is likely not all that productive."
With this patch the freeing messages now look like this:
Fix a TCP loss recovery performance bug raised recently on the netdev
list, in two threads:
(i) July 26, 2017: netdev thread "TCP fast retransmit issues"
(ii) July 26, 2017: netdev thread:
"[PATCH V2 net-next] TLP: Don't reschedule PTO when there's one
outstanding TLP retransmission"
The basic problem is that incoming TCP packets that did not indicate
forward progress could cause the xmit timer (TLP or RTO) to be rearmed
and pushed back in time. In certain corner cases this could result in
the following problems noted in these threads:
- Repeated ACKs coming in with bogus SACKs corrupted by middleboxes
could cause TCP to repeatedly schedule TLPs forever. We kept
sending TLPs after every ~200ms, which elicited bogus SACKs, which
caused more TLPs, ad infinitum; we never fired an RTO to fill in
the holes.
- Incoming data segments could, in some cases, cause us to reschedule
our RTO or TLP timer further out in time, for no good reason. This
could cause repeated inbound data to result in stalls in outbound
data, in the presence of packet loss.
This commit fixes these bugs by changing the TLP and RTO ACK
processing to:
(a) Only reschedule the xmit timer once per ACK.
(b) Only reschedule the xmit timer if tcp_clean_rtx_queue() deems the
ACK indicates sufficient forward progress (a packet was
cumulatively ACKed, or we got a SACK for a packet that was sent
before the most recent retransmit of the write queue head).
This brings us back into closer compliance with the RFCs, since, as
the comment for tcp_rearm_rto() notes, we should only restart the RTO
timer after forward progress on the connection. Previously we were
restarting the xmit timer even in these cases where there was no
forward progress.
As a side benefit, this commit simplifies and speeds up the TCP timer
arming logic. We had been calling inet_csk_reset_xmit_timer() three
times on normal ACKs that cumulatively acknowledged some data:
1) Once near the top of tcp_ack() to switch from TLP timer to RTO:
if (icsk->icsk_pending == ICSK_TIME_LOSS_PROBE)
tcp_rearm_rto(sk);
2) Once in tcp_clean_rtx_queue(), to update the RTO:
if (flag & FLAG_ACKED) {
tcp_rearm_rto(sk);
3) Once in tcp_ack() after tcp_fastretrans_alert() to switch from RTO
to TLP:
if (icsk->icsk_pending == ICSK_TIME_RETRANS)
tcp_schedule_loss_probe(sk);
This commit, by only rescheduling the xmit timer once per ACK,
simplifies the code and reduces CPU overhead.
This commit was tested in an A/B test with Google web server
traffic. SNMP stats and request latency metrics were within noise
levels, substantiating that for normal web traffic patterns this is a
rare issue. This commit was also tested with packetdrill tests to
verify that it fixes the timer behavior in the corner cases discussed
in the netdev threads mentioned above.
This patch is a bug fix patch intended to be queued for -stable
relases.
[This version of the commit was compiled and briefly tested
based on top of v3.10.107.]
Have tcp_schedule_loss_probe() base the TLP scheduling decision based
on when the RTO *should* fire. This is to enable the upcoming xmit
timer fix in this series, where tcp_schedule_loss_probe() cannot
assume that the last timer installed was an RTO timer (because we are
no longer doing the "rearm RTO, rearm RTO, rearm TLP" dance on every
ACK). So tcp_schedule_loss_probe() must independently figure out when
an RTO would want to fire.
In the new TLP implementation following in this series, we cannot
assume that icsk_timeout was set based on an RTO; after processing a
cumulative ACK the icsk_timeout we see can be from a previous TLP or
RTO. So we need to independently recalculate the RTO time (instead of
reading it out of icsk_timeout). Removing this dependency on the
nature of icsk_timeout makes things a little easier to reason about
anyway.
Note that the old and new code should be equivalent, since they are
both saying: "if the RTO is in the future, but at an earlier time than
the normal TLP time, then set the TLP timer to fire when the RTO would
have fired".
[This version of the commit was compiled and briefly tested
based on top of v3.10.107.]
Pure refactor. This helper will be required in the xmit timer fix
later in the patch series. (Because the TLP logic will want to make
this calculation.)
[This version of the commit was compiled and briefly tested
based on top of v3.10.107.]
... rather than relying on ciptool(8) never passing it anything else. Give
it e.g. an AF_UNIX connected socket (from socketpair(2)) and it'll oops,
trying to evaluate &l2cap_pi(sock->sk)->chan->dst...
My recent change missed fact that UFO would perform a complete
UDP checksum before segmenting in frags.
In this case skb->ip_summed is set to CHECKSUM_NONE.
We need to add this valid case to skb_needs_check()
Fixes: b2504a5dbef3 ("net: reduce skb_warn_bad_offload() noise") Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Willem de Bruijn <willemb@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Willy Tarreau <w@1wt.eu>
Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: Dmitry Vyukov <dvyukov@google.com> Cc: Willem de Bruijn <willemb@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Willy Tarreau <w@1wt.eu>
We do not check if packet from real server is for NAT
connection before performing SNAT. This causes problems
for setups that use DR/TUN and allow local clients to
access the real server directly, for example:
- local client in director creates IPVS-DR/TUN connection
CIP->VIP and the request packets are routed to RIP.
Talks are finished but IPVS connection is not expired yet.
- second local client creates non-IPVS connection CIP->RIP
with same reply tuple RIP->CIP and when replies are received
on LOCAL_IN we wrongly assign them for the first client
connection because RIP->CIP matches the reply direction.
As result, IPVS SNATs replies for non-IPVS connections.
The problem is more visible to local UDP clients but in rare
cases it can happen also for TCP or remote clients when the
real server sends the reply traffic via the director.
So, better to be more precise for the reply traffic.
As replies are not expected for DR/TUN connections, better
to not touch them.
Fix expand_upwards() on architectures with an upward-growing stack (parisc,
metag and partly IA-64) to allow the stack to reliably grow exactly up to
the address space limit given by TASK_SIZE.
Trinity gets kernel BUG at mm/mmap.c:1963! in about 3 minutes of
mmap testing. That's the VM_BUG_ON(gap_end < gap_start) at the
end of unmapped_area_topdown(). Linus points out how MAP_FIXED
(which does not have to respect our stack guard gap intentions)
could result in gap_end below gap_start there. Fix that, and
the similar case in its alternative, unmapped_area().
Cc: stable@vger.kernel.org Fixes: 1be7107fbe18 ("mm: larger stack guard gap, between vmas") Reported-by: Dave Jones <davej@codemonkey.org.uk> Debugged-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Hugh Dickins <hughd@google.com> Acked-by: Michal Hocko <mhocko@suse.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Willy Tarreau <w@1wt.eu>
Stack guard page is a useful feature to reduce a risk of stack smashing
into a different mapping. We have been using a single page gap which
is sufficient to prevent having stack adjacent to a different mapping.
But this seems to be insufficient in the light of the stack usage in
userspace. E.g. glibc uses as large as 64kB alloca() in many commonly
used functions. Others use constructs liks gid_t buffer[NGROUPS_MAX]
which is 256kB or stack strings with MAX_ARG_STRLEN.
This will become especially dangerous for suid binaries and the default
no limit for the stack size limit because those applications can be
tricked to consume a large portion of the stack and a single glibc call
could jump over the guard page. These attacks are not theoretical,
unfortunatelly.
Make those attacks less probable by increasing the stack guard gap
to 1MB (on systems with 4k pages; but make it depend on the page size
because systems with larger base pages might cap stack allocations in
the PAGE_SIZE units) which should cover larger alloca() and VLA stack
allocations. It is obviously not a full fix because the problem is
somehow inherent, but it should reduce attack space a lot.
One could argue that the gap size should be configurable from userspace,
but that can be done later when somebody finds that the new 1MB is wrong
for some special case applications. For now, add a kernel command line
option (stack_guard_gap) to specify the stack gap size (in page units).
Implementation wise, first delete all the old code for stack guard page:
because although we could get away with accounting one extra page in a
stack vma, accounting a larger gap can break userspace - case in point,
a program run with "ulimit -S -v 20000" failed when the 1MB gap was
counted for RLIMIT_AS; similar problems could come with RLIMIT_MLOCK
and strict non-overcommit mode.
Instead of keeping gap inside the stack vma, maintain the stack guard
gap as a gap between vmas: using vm_start_gap() in place of vm_start
(or vm_end_gap() in place of vm_end if VM_GROWSUP) in just those few
places which need to respect the gap - mainly arch_get_unmapped_area(),
and and the vma tree's subtree_gap support for that.
Original-patch-by: Oleg Nesterov <oleg@redhat.com> Original-patch-by: Michal Hocko <mhocko@suse.com> Signed-off-by: Hugh Dickins <hughd@google.com>
[wt: backport to 4.11: adjust context]
[wt: backport to 4.9: adjust context ; kernel doc was not in admin-guide]
[wt: backport to 4.4: adjust context ; drop ppc hugetlb_radix changes]
[wt: backport to 3.18: adjust context ; no FOLL_POPULATE ;
s390 uses generic arch_get_unmapped_area()]
[wt: backport to 3.16: adjust context]
[wt: backport to 3.10: adjust context ; code logic in PARISC's
arch_get_unmapped_area() wasn't found ; code inserted into
expand_upwards() and expand_downwards() runs under anon_vma lock;
changes for gup.c:faultin_page go to memory.c:__get_user_pages();
included Hugh Dickins' fixes] Signed-off-by: Willy Tarreau <w@1wt.eu>
Currently on i386 and on X86_64 when emulating X86_32 in legacy mode, only
the stack and the executable are randomized but not other mmapped files
(libraries, vDSO, etc.). This patch enables randomization for the
libraries, vDSO and mmap requests on i386 and in X86_32 in legacy mode.
By default on i386 there are 8 bits for the randomization of the libraries,
vDSO and mmaps which only uses 1MB of VA.
This patch preserves the original randomness, using 1MB of VA out of 3GB or
4GB. We think that 1MB out of 3GB is not a big cost for having the ASLR.
The first obvious security benefit is that all objects are randomized (not
only the stack and the executable) in legacy mode which highly increases
the ASLR effectiveness, otherwise the attackers may use these
non-randomized areas. But also sensitive setuid/setgid applications are
more secure because currently, attackers can disable the randomization of
these applications by setting the ulimit stack to "unlimited". This is a
very old and widely known trick to disable the ASLR in i386 which has been
allowed for too long.
Another trick used to disable the ASLR was to set the ADDR_NO_RANDOMIZE
personality flag, but fortunately this doesn't work on setuid/setgid
applications because there is security checks which clear Security-relevant
flags.
This patch always randomizes the mmap_legacy_base address, removing the
possibility to disable the ASLR by setting the stack to "unlimited".
Signed-off-by: Hector Marco-Gisbert <hecmargi@upv.es> Acked-by: Ismael Ripoll Ripoll <iripoll@upv.es> Acked-by: Kees Cook <keescook@chromium.org> Acked-by: Arjan van de Ven <arjan@linux.intel.com> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: akpm@linux-foundation.org Cc: kees Cook <keescook@chromium.org> Link: http://lkml.kernel.org/r/1457639460-5242-1-git-send-email-hecmargi@upv.es Signed-off-by: Ingo Molnar <mingo@kernel.org> Signed-off-by: Ben Hutchings <ben@decadent.org.uk> Signed-off-by: Willy Tarreau <w@1wt.eu>
In preparation for splitting out ET_DYN ASLR, this refactors the use of
mmap_rnd() to be used similarly to arm, and extracts the checking of
PF_RANDOMIZE.
In situations where an skb is paged, the transport header pointer and
tail pointer can be the same because the skb contents are in frags.
This results in ioctl(SIOCINQ/FIONREAD) incorrectly returning a
length of 0 when the length to receive is actually greater than zero.
skb->len is already correctly set in ip6_input_finish() with
pskb_pull(), so use skb->len as it always returns the correct result
for both linear and paged data.
This RCU warning, however, is suppressed by lockdep_off() in printk().
lockdep_off() increments the ->lockdep_recursion counter and thus
disables RCU_LOCKDEP_WARN() and debug_lockdep_rcu_enabled(), which want
lockdep to be enabled "current->lockdep_recursion == 0".
Link: http://lkml.kernel.org/r/20170217015932.11898-1-sergey.senozhatsky@gmail.com Signed-off-by: Sergey Senozhatsky <sergey.senozhatsky@gmail.com> Reported-by: Tony Lindgren <tony@atomide.com> Tested-by: Tony Lindgren <tony@atomide.com> Acked-by: Paul E. McKenney <paulmck@linux.vnet.ibm.com> Acked-by: Steven Rostedt (VMware) <rostedt@goodmis.org> Cc: Petr Mladek <pmladek@suse.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Tony Lindgren <tony@atomide.com> Cc: Russell King <rmk@armlinux.org.uk> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
[wt: changes are in kernel/printk.c in 3.10] Signed-off-by: Willy Tarreau <w@1wt.eu>
When IFF_VNET_HDR is enabled, a virtio_net header must precede data.
Data length is verified to be greater than or equal to expected header
length tun->vnet_hdr_sz before copying.
Read this value once and cache locally, as it can be updated between
the test and use (TOCTOU).
[js] we have TUN_VNET_HDR in 3.12
Signed-off-by: Willem de Bruijn <willemb@google.com> Reported-by: Dmitry Vyukov <dvyukov@google.com> CC: Eric Dumazet <edumazet@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Jiri Slaby <jslaby@suse.cz>
[wt: s/READ_ONCE/ACCESS_ONCE] Signed-off-by: Willy Tarreau <w@1wt.eu>
When L2 exits to L0 due to "exception or NMI", software exceptions
(#BP and #OF) for which L1 has requested an intercept should be
handled by L1 rather than L0. Previously, only hardware exceptions
were forwarded to L1.
Signed-off-by: Jim Mattson <jmattson@google.com> Signed-off-by: Paolo Bonzini <pbonzini@redhat.com> Signed-off-by: Willy Tarreau <w@1wt.eu>
On x86-32, with CONFIG_FIRMWARE and multiple CPUs, if you enable function
graph tracing and then suspend to RAM, it will triple fault and reboot when
it resumes.
The first fault happens when booting a secondary CPU:
The early head_32.S code calls into load_ucode_ap(), which has an an
ftrace hook, so it calls prepare_ftrace_return(), which calls
ftrace_graph_is_dead(), which tries to access the global
'kill_ftrace_graph' variable with a virtual address, causing a fault
because the CPU is still in real mode.
The fix is to add a check in prepare_ftrace_return() to make sure it's
running in protected mode before continuing. The check makes sure the
stack pointer is a virtual kernel address. It's a bit of a hack, but
it's not very intrusive and it works well enough.
For reference, here are a few other (more difficult) ways this could
have potentially been fixed:
- Move startup_32_smp()'s call to load_ucode_ap() down to *after* paging
is enabled. (No idea what that would break.)
- Track down load_ucode_ap()'s entire callee tree and mark all the
functions 'notrace'. (Probably not realistic.)
- Pause graph tracing in ftrace_suspend_notifier_call() or bringup_cpu()
or __cpu_up(), and ensure that the pause facility can be queried from
real mode.
Reported-by: Paul Menzel <pmenzel@molgen.mpg.de> Signed-off-by: Josh Poimboeuf <jpoimboe@redhat.com> Tested-by: Paul Menzel <pmenzel@molgen.mpg.de> Reviewed-by: Steven Rostedt (VMware) <rostedt@goodmis.org> Cc: "Rafael J . Wysocki" <rjw@rjwysocki.net> Cc: linux-acpi@vger.kernel.org Cc: Borislav Petkov <bp@alien8.de> Cc: Len Brown <lenb@kernel.org> Link: http://lkml.kernel.org/r/5c1272269a580660703ed2eccf44308e790c7a98.1492123841.git.jpoimboe@redhat.com Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: Willy Tarreau <w@1wt.eu>
A client can append random data to the end of an NFSv2 or NFSv3 RPC call
without our complaining; we'll just stop parsing at the end of the
expected data and ignore the rest.
Encoded arguments and replies are stored together in an array of pages,
and if a call is too large it could leave inadequate space for the
reply. This is normally OK because NFS RPC's typically have either
short arguments and long replies (like READ) or long arguments and short
replies (like WRITE). But a client that sends an incorrectly long reply
can violate those assumptions. This was observed to cause crashes.
Also, several operations increment rq_next_page in the decode routine
before checking the argument size, which can leave rq_next_page pointing
well past the end of the page array, causing trouble later in
svc_free_pages.
So, following a suggestion from Neil Brown, add a central check to
enforce our expectation that no NFSv2/v3 call has both a large call and
a large reply.
As followup we may also want to rewrite the encoding routines to check
more carefully that they aren't running off the end of the page array.
We may also consider rejecting calls that have any extra garbage
appended. That would be safer, and within our rights by spec, but given
the age of our server and the NFS protocol, and the fact that we've
never enforced this before, we may need to balance that against the
possibility of breaking some oddball client.
Reported-by: Tuomas Haanpää <thaan@synopsys.com> Reported-by: Ari Kauppi <ari@synopsys.com> Reviewed-by: NeilBrown <neilb@suse.com> Signed-off-by: J. Bruce Fields <bfields@redhat.com> Signed-off-by: Willy Tarreau <w@1wt.eu>
On slow platforms with unreliable TSC, such as QEMU emulated machines,
it is possible for the kernel to request the next event in the past. In
that case, in the current implementation of xen_vcpuop_clockevent, we
simply return -ETIME. To be precise the Xen returns -ETIME and we pass
it on. However the result of this is a missed event, which simply causes
the kernel to hang.
Instead it is better to always ask the hypervisor for a timer event,
even if the timeout is in the past. That way there are no lost
interrupts and the kernel survives. To do that, remove the
VCPU_SSHOTTMR_future flag.
Two different threads with different rds sockets may be in
rds_recv_rcvbuf_delta() via receive path. If their ports
both map to the same word in the congestion map, then
using non-atomic ops to update it could cause the map to
be incorrect. Lets use atomics to avoid such an issue.
Full credit to Wengang <wen.gang.wang@oracle.com> for
finding the issue, analysing it and also pointing out
to offending code with spin lock based fix.
Reviewed-by: Leon Romanovsky <leon@leon.nu> Signed-off-by: Wengang Wang <wen.gang.wang@oracle.com> Signed-off-by: Santosh Shilimkar <santosh.shilimkar@oracle.com> Signed-off-by: David S. Miller <davem@davemloft.net> Cc: Julia Lawall <julia.lawall@lip6.fr> Signed-off-by: Jiri Slaby <jslaby@suse.cz> Signed-off-by: Willy Tarreau <w@1wt.eu>
As part of handling a crash on an SMP system, an IPI is send to
all other CPUs to save their current registers and stop. It was
using task_pt_regs(current) to get the registers, but that will
only be accurate if the CPU was interrupted running in userland.
Instead allow the architecture to pass in the registers (all
pass NULL now, but allow for the future) and then use get_irq_regs()
which should be accurate as we are in an interrupt. Fall back to
task_pt_regs(current) if nothing else is available.
Andrey Konovalov reported a BUG caused by the ip6mr code which is caused
because we call unregister_netdevice_many for a device that is already
being destroyed. In IPv4's ipmr that has been resolved by two commits
long time ago by introducing the "notify" parameter to the delete
function and avoiding the unregister when called from a notifier, so
let's do the same for ip6mr.
Now sctp doesn't check sock's state before listening on it. It could
even cause changing a sock with any state to become a listening sock
when doing sctp_listen.
This patch is to fix it by checking sock's state in sctp_listen, so
that it will listen on the sock with right state.
Reported-by: Andrey Konovalov <andreyknvl@google.com> Tested-by: Andrey Konovalov <andreyknvl@google.com> Signed-off-by: Xin Long <lucien.xin@gmail.com> Acked-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Willy Tarreau <w@1wt.eu>
Dmitry posted a nice reproducer of a bug triggering in neigh_probe()
when dereferencing a NULL neigh->ops->solicit method.
This can happen for arp_direct_ops/ndisc_direct_ops and similar,
which can be used for NUD_NOARP neighbours (created when dev->header_ops
is NULL). Admin can then force changing nud_state to some other state
that would fire neigh timer.
Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: Dmitry Vyukov <dvyukov@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Willy Tarreau <w@1wt.eu>
We get a bogus warning about a potential uninitialized variable
use in gfs2, because the compiler does not figure out that we
never use the leaf number if get_leaf_nr() returns an error:
fs/gfs2/dir.c: In function 'get_first_leaf':
fs/gfs2/dir.c:802:9: warning: 'leaf_no' may be used uninitialized in this function [-Wmaybe-uninitialized]
fs/gfs2/dir.c: In function 'dir_split_leaf':
fs/gfs2/dir.c:1021:8: warning: 'leaf_no' may be used uninitialized in this function [-Wmaybe-uninitialized]
Changing the 'if (!error)' to 'if (!IS_ERR_VALUE(error))' is
sufficient to let gcc understand that this is exactly the same
condition as in IS_ERR() so it can optimize the code path enough
to understand it.
The driver reads a value from hfa384x_from_bap(), which may fail,
and then assigns the value to a local variable. gcc detects that
in in the failure case, the 'rlen' variable now contains
uninitialized data:
In file included from ../drivers/net/wireless/intersil/hostap/hostap_pci.c:220:0:
drivers/net/wireless/intersil/hostap/hostap_hw.c: In function 'hfa384x_get_rid':
drivers/net/wireless/intersil/hostap/hostap_hw.c:842:5: warning: 'rec' may be used uninitialized in this function [-Wmaybe-uninitialized]
if (le16_to_cpu(rec.len) == 0) {
This restructures the function as suggested by Russell King, to
make it more readable and get more reliable error handling, by
handling each failure mode using a goto.
Signed-off-by: Arnd Bergmann <arnd@arndb.de> Signed-off-by: Kalle Valo <kvalo@codeaurora.org> Signed-off-by: Willy Tarreau <w@1wt.eu>
The nozomi wireless data driver has its own helper function to
transfer data from a FIFO, doing an extra byte swap on big-endian
architectures, presumably to bring the data back into byte-serial
order after readw() or readl() perform their implicit byteswap.
This helper function is used in the receive_data() function to
first read the length into a 32-bit variable, which causes
a compile-time warning:
drivers/tty/nozomi.c: In function 'receive_data':
drivers/tty/nozomi.c:857:9: warning: 'size' may be used uninitialized in this function [-Wmaybe-uninitialized]
The problem is that gcc is unsure whether the data was actually
read or not. We know that it is at this point, so we can replace
it with a single readl() to shut up that warning.
I am leaving the byteswap in there, to preserve the existing
behavior, even though this seems fishy: Reading the length of
the data into a cpu-endian variable should normally not use
a second byteswap on big-endian systems, unless the hardware
is aware of the CPU endianess.
There appears to be a lot more confusion about endianess in this
driver, so it probably has not worked on big-endian systems in
a long time, if ever, and I have no way to test it. It's well
possible that this driver has not been used by anyone in a while,
the last patch that looks like it was tested on the hardware is
from 2008.
When calculating po->tp_hdrlen + po->tp_reserve the result can overflow.
Fix by checking that tp_reserve <= INT_MAX on assign.
Signed-off-by: Andrey Konovalov <andreyknvl@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Willy Tarreau <w@1wt.eu>
When calculating rb->frames_per_block * req->tp_block_nr the result
can overflow.
Add a check that tp_block_size * tp_block_nr <= UINT_MAX.
Since frames_per_block <= tp_block_size, the expression would
never overflow.
Signed-off-by: Andrey Konovalov <andreyknvl@google.com> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Willy Tarreau <w@1wt.eu>
There is a bug in binutils 2.24 which causes miscompilation if we're
building little endian and using weak symbols (which the kernel does).
It is fixed in binutils commit 57fa7b8c7e59 "Correct elf_merge_st_other
arguments for weak symbols", which is in binutils 2.25 and has been
backported to the binutils 2.24 branch and has been picked up by most
distros it seems.
However if we're running stock 2.24 (no extra version) then the bug is
present, so check for that and bail.
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au> Signed-off-by: Willy Tarreau <w@1wt.eu>