In commit 85fe402 (fs: do not assign default i_ino in new_inode), the
initialisation of i_ino was removed from new_inode() and pushed down
into the callers. However spufs_new_inode() was not updated.
This exhibits as no files appearing in /spu, because all our dirents
have a zero inode, which readdir() seems to dislike.
Signed-off-by: Michael Ellerman <michael@ellerman.id.au> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In __after_prom_start we copy the kernel down to zero in two calls to
copy_and_flush. After the first call (copy from 0 to copy_to_here:)
we jump to the newly copied code soon after.
Unfortunately there's no isync between the copy of this code and the
jump to it. Hence it's possible that stale instructions could still be
in the icache or pipeline before we branch to it.
We've seen this on real machines and it's results in no console output
after:
calling quiesce...
returning from prom_init
The below adds an isync to ensure that the copy and flushing has
completed before any branching to the new instructions occurs.
Signed-off-by: Michael Neuling <mikey@neuling.org> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit 4a94445c9a5c (net: Use ip_route_input_noref() in input path)
added a bug in IP defragmentation handling, as non refcounted
dst could escape an RCU protected section.
Commit 64f3b9e203bd068 (net: ip_expire() must revalidate route) fixed
the case of timeouts, but not the general problem.
Tom Parkin noticed crashes in UDP stack and provided a patch,
but further analysis permitted us to pinpoint the root cause.
Before queueing a packet into a frag list, we must drop its dst,
as this dst has limited lifetime (RCU protected)
When/if a packet is finally reassembled, we use the dst of the very
last skb, still protected by RCU and valid, as the dst of the
reassembled packet.
Use same logic in IPv6, as there is no need to hold dst references.
Reported-by: Tom Parkin <tparkin@katalix.com> Tested-by: Tom Parkin <tparkin@katalix.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit 257b5358b32f ("scm: Capture the full credentials of the scm
sender") changed the credentials passing code to pass in the effective
uid/gid instead of the real uid/gid.
Obviously this doesn't matter most of the time (since normally they are
the same), but it results in differences for suid binaries when the wrong
uid/gid ends up being used.
This just undoes that (presumably unintentional) part of the commit.
Reported-by: Andy Lutomirski <luto@amacapital.net> Cc: Eric W. Biederman <ebiederm@xmission.com> Cc: Serge E. Hallyn <serge@hallyn.com> Cc: David S. Miller <davem@davemloft.net> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Acked-by: "Eric W. Biederman" <ebiederm@xmission.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If one does do something unfortunate and allow a
bad offload bug into the kernel, this the
skb_warn_bad_offload can effectively live-lock the
system, filling the logs with the same error over
and over.
Add rate limitation to this so that box remains otherwise
functional in this case.
Signed-off-by: Ben Greear <greearb@candelatech.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
commit bd090dfc634d (tcp: tcp_replace_ts_recent() should not be called
from tcp_validate_incoming()) introduced a TS ecr bug in slow path
processing.
1 A > B P. 1:10001(10000) ack 1 <nop,nop,TS val 1001 ecr 200>
2 B < A . 1:1(0) ack 1 win 257 <sack 9001:10001,TS val 300 ecr 1001>
3 A > B . 1:1001(1000) ack 1 win 227 <nop,nop,TS val 1002 ecr 200>
4 A > B . 1001:2001(1000) ack 1 win 227 <nop,nop,TS val 1002 ecr 200>
(ecr 200 should be ecr 300 in packets 3 & 4)
Problem is tcp_ack() can trigger send of new packets (retransmits),
reflecting the prior TSval, instead of the TSval contained in the
currently processed incoming packet.
Fix this by calling tcp_replace_ts_recent() from tcp_ack() after the
checks, but before the actions.
Reported-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Neal Cardwell <ncardwell@google.com> Acked-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
For sensitive data like keying material, it is common practice to zero
out keys before returning the memory back to the allocator. Thus, use
kzfree instead of kfree.
Signed-off-by: Daniel Borkmann <dborkman@redhat.com> Acked-by: Neil Horman <nhorman@tuxdriver.com> Acked-by: Vlad Yasevich <vyasevich@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If a TCP retransmission gets partially ACKed and collapsed multiple
times it is possible for the headroom to grow beyond 64K which will
overflow the 16bit skb->csum_start which is based on the start of
the headroom. It has been observed rarely in the wild with IPoIB due
to the 64K MTU.
Verify if the acking and collapsing resulted in a headroom exceeding
what csum_start can cover and reallocate the headroom if so.
A big thank you to Jim Foraker <foraker1@llnl.gov> and the team at
LLNL for helping out with the investigation and testing.
Reported-by: Jim Foraker <foraker1@llnl.gov> Signed-off-by: Thomas Graf <tgraf@suug.ch> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
There is a bug in cookie_v4_check (net/ipv4/syncookies.c):
flowi4_init_output(&fl4, 0, sk->sk_mark, RT_CONN_FLAGS(sk),
RT_SCOPE_UNIVERSE, IPPROTO_TCP,
inet_sk_flowi_flags(sk),
(opt && opt->srr) ? opt->faddr : ireq->rmt_addr,
ireq->loc_addr, th->source, th->dest);
Here we do not respect sk->sk_bound_dev_if, therefore wrong dst_entry may be
taken. This dst_entry is used by new socket (get_cookie_sock ->
tcp_v4_syn_recv_sock), so its packets may take the wrong path.
Signed-off-by: Dmitry Popov <dp@highloadlab.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Signed-off-by: Michael Riesch <michael.riesch@omicron.at> Cc: Jiri Benc <jbenc@redhat.com> Cc: "Theodore Ts'o" <tytso@mit.edu> Acked-by: Mark Rustad <mark.d.rustad@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit 130549fe ("netfilter: reset nf_trace in nf_reset") added code
to reset nf_trace in nf_reset(). This is wrong and unnecessary.
nf_reset() is used in the following cases:
- when passing packets up the the socket layer, at which point we want to
release all netfilter references that might keep modules pinned while
the packet is queued. nf_trace doesn't matter anymore at this point.
- when encapsulating or decapsulating IPsec packets. We want to continue
tracing these packets after IPsec processing.
- when passing packets through virtual network devices. Only devices on
that encapsulate in IPv4/v6 matter since otherwise nf_trace is not
used anymore. Its not entirely clear whether those packets should
be traced after that, however we've always done that.
- when passing packets through virtual network devices that make the
packet cross network namespace boundaries. This is the only cases
where we clearly want to reset nf_trace and is also what the
original patch intended to fix.
Add a new function nf_reset_trace() and use it in dev_forward_skb() to
fix this properly.
Signed-off-by: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
It was reported that the following LSB test case failed
https://lsbbugs.linuxfoundation.org/attachment.cgi?id=2144 because we
were not coallescing unix stream messages when the application was
expecting us to.
The problem was that the first send was before the socket was accepted
and thus sock->sk_socket was NULL in maybe_add_creds, and the second
send after the socket was accepted had a non-NULL value for sk->socket
and thus we could tell the credentials were not needed so we did not
bother.
The unnecessary credentials on the first message cause
unix_stream_recvmsg to start verifying that all messages had the same
credentials before coallescing and then the coallescing failed because
the second message had no credentials.
Ignoring credentials when we don't care in unix_stream_recvmsg fixes a
long standing pessimization which would fail to coallesce messages when
reading from a unix stream socket if the senders were different even if
we did not care about their credentials.
I have tested this and verified that the in the LSB test case mentioned
above that the messages do coallesce now, while the were failing to
coallesce without this change.
Reported-by: Karel Srot <ksrot@redhat.com> Reported-by: Ding Tianhong <dingtianhong@huawei.com> 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@linuxfoundation.org>
While enslaving a new device and after IFF_BONDING flag is set, in case
of failure it is not stripped from the device's priv_flags while
cleaning up, which could lead to other problems.
Cleaning at err_close because the flag is set after dev_open().
v2: no change
Signed-off-by: Nikolay Aleksandrov <nikolay@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
While the bonding module is unloading, it is considered that after
rtnl_link_unregister all bond devices are destroyed but since no
synchronization mechanism exists, a new bond device can be created
via bonding_masters before unregister_pernet_subsys which would
lead to multiple problems (e.g. NULL pointer dereference, wrong RIP,
list corruption).
This patch fixes the issue by removing any bond devices left in the
netns after bonding_masters is removed from sysfs.
Signed-off-by: Nikolay Aleksandrov <nikolay@redhat.com> Acked-by: Veaceslav Falico <vfalico@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
A few drivers use dev_uc_sync/unsync to synchronize the
address lists from master down to slave/lower devices. In
some cases (bond/team) a single address list is synched down
to multiple devices. At the time of unsync, we have a leak
in these lower devices, because "synced" is treated as a
boolean and the address will not be unsynced for anything after
the first device/call.
Treat "synced" as a count (same as refcount) and allow all
unsync calls to work.
Signed-off-by: Vlad Yasevich <vyasevic@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
IPv6 Routing table becomes broken once we do ifdown, ifup of the loopback(lo)
interface. After down-up, routes of other interface's IPv6 addresses through
'lo' are lost.
IPv6 addresses assigned to all interfaces are routed through 'lo' for internal
communication. Once 'lo' is down, those routing entries are removed from routing
table. But those removed entries are not being re-created properly when 'lo' is
brought up. So IPv6 addresses of other interfaces becomes unreachable from the
same machine. Also this breaks communication with other machines because of
NDISC packet processing failure.
This patch fixes this issue by reading all interface's IPv6 addresses and adding
them to IPv6 routing table while bringing up 'lo'.
==Testing==
Before applying the patch:
$ route -A inet6
Kernel IPv6 routing table
Destination Next Hop Flag Met Ref Use If
2000::20/128 :: U 256 0 0 eth0
fe80::/64 :: U 256 0 0 eth0
::/0 :: !n -1 1 1 lo
::1/128 :: Un 0 1 0 lo
2000::20/128 :: Un 0 1 0 lo
fe80::xxxx:xxxx:xxxx:xxxx/128 :: Un 0 1 0 lo
ff00::/8 :: U 256 0 0 eth0
::/0 :: !n -1 1 1 lo
$ sudo ifdown lo
$ sudo ifup lo
$ route -A inet6
Kernel IPv6 routing table
Destination Next Hop Flag Met Ref Use If
2000::20/128 :: U 256 0 0 eth0
fe80::/64 :: U 256 0 0 eth0
::/0 :: !n -1 1 1 lo
::1/128 :: Un 0 1 0 lo
ff00::/8 :: U 256 0 0 eth0
::/0 :: !n -1 1 1 lo
$
After applying the patch:
$ route -A inet6
Kernel IPv6 routing
table
Destination Next Hop Flag Met Ref Use If
2000::20/128 :: U 256 0 0 eth0
fe80::/64 :: U 256 0 0 eth0
::/0 :: !n -1 1 1 lo
::1/128 :: Un 0 1 0 lo
2000::20/128 :: Un 0 1 0 lo
fe80::xxxx:xxxx:xxxx:xxxx/128 :: Un 0 1 0 lo
ff00::/8 :: U 256 0 0 eth0
::/0 :: !n -1 1 1 lo
$ sudo ifdown lo
$ sudo ifup lo
$ route -A inet6
Kernel IPv6 routing table
Destination Next Hop Flag Met Ref Use If
2000::20/128 :: U 256 0 0 eth0
fe80::/64 :: U 256 0 0 eth0
::/0 :: !n -1 1 1 lo
::1/128 :: Un 0 1 0 lo
2000::20/128 :: Un 0 1 0 lo
fe80::xxxx:xxxx:xxxx:xxxx/128 :: Un 0 1 0 lo
ff00::/8 :: U 256 0 0 eth0
::/0 :: !n -1 1 1 lo
$
Signed-off-by: Balakumaran Kannan <Balakumaran.Kannan@ap.sony.com> Signed-off-by: Maruthi Thotad <Maruthi.Thotad@ap.sony.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
currently cbq works incorrectly for limits > 10% real link bandwidth,
and practically does not work for limits > 50% real link bandwidth.
Below are results of experiments taken on 1 Gbit link
In shaper | Actual Result
-----------+---------------
100M | 108 Mbps
200M | 244 Mbps
300M | 412 Mbps
500M | 893 Mbps
This happen because of q->now changes incorrectly in cbq_dequeue():
when it is called before real end of packet transmitting,
L2T is greater than real time delay, q_now gets an extra boost
but never compensate it.
To fix this problem we prevent change of q->now until its synchronization
with real time.
Signed-off-by: Vasily Averin <vvs@openvz.org> Reviewed-by: Alexey Kuznetsov <kuznet@ms2.inr.ac.ru> Acked-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
sizeof() when applied to a pointer typed expression gives the size of the
pointer, not that of the pointed data.
Introduced by commit 3ce5ef(netrom: fix info leak via msg_name in nr_recvmsg)
Signed-off-by: Wei Yongjun <yongjun_wei@trendmicro.com.cn> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The code in set_orig_addr() does not initialize all of the members of
struct sockaddr_tipc when filling the sockaddr info -- namely the union
is only partly filled. This will make recv_msg() and recv_stream() --
the only users of this function -- leak kernel stack memory as the
msg_name member is a local variable in net/socket.c.
Additionally to that both recv_msg() and recv_stream() fail to update
the msg_namelen member to 0 while otherwise returning with 0, i.e.
"success". This is the case for, e.g., non-blocking sockets. This will
lead to a 128 byte kernel stack leak in net/socket.c.
Fix the first issue by initializing the memory of the union with
memset(0). Fix the second one by setting msg_namelen to 0 early as it
will be updated later if we're going to fill the msg_name member.
Signed-off-by: Mathias Krause <minipli@googlemail.com> Cc: Jon Maloy <jon.maloy@ericsson.com> Cc: Allan Stephens <allan.stephens@windriver.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The code in rose_recvmsg() does not initialize all of the members of
struct sockaddr_rose/full_sockaddr_rose when filling the sockaddr info.
Nor does it initialize the padding bytes of the structure inserted by
the compiler for alignment. This will lead to leaking uninitialized
kernel stack bytes in net/socket.c.
Fix the issue by initializing the memory used for sockaddr info with
memset(0).
Signed-off-by: Mathias Krause <minipli@googlemail.com> Cc: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The code in llcp_sock_recvmsg() does not initialize all the members of
struct sockaddr_nfc_llcp when filling the sockaddr info. Nor does it
initialize the padding bytes of the structure inserted by the compiler
for alignment.
Also, if the socket is in state LLCP_CLOSED or is shutting down during
receive the msg_namelen member is not updated to 0 while otherwise
returning with 0, i.e. "success". The msg_namelen update is also
missing for stream and seqpacket sockets which don't fill the sockaddr
info.
Both issues lead to the fact that the code will leak uninitialized
kernel stack bytes in net/socket.c.
Fix the first issue by initializing the memory used for sockaddr info
with memset(0). Fix the second one by setting msg_namelen to 0 early.
It will be updated later if we're going to fill the msg_name member.
Signed-off-by: Mathias Krause <minipli@googlemail.com> Cc: Lauro Ramos Venancio <lauro.venancio@openbossa.org> Cc: Aloisio Almeida Jr <aloisio.almeida@openbossa.org> Cc: Samuel Ortiz <sameo@linux.intel.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In case msg_name is set the sockaddr info gets filled out, as
requested, but the code fails to initialize the padding bytes of
struct sockaddr_ax25 inserted by the compiler for alignment. Also
the sax25_ndigis member does not get assigned, leaking four more
bytes.
Both issues lead to the fact that the code will leak uninitialized
kernel stack bytes in net/socket.c.
Fix both issues by initializing the memory with memset(0).
Signed-off-by: Mathias Krause <minipli@googlemail.com> Cc: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
For stream sockets the code misses to update the msg_namelen member
to 0 and therefore makes net/socket.c leak the local, uninitialized
sockaddr_storage variable to userland -- 128 bytes of kernel stack
memory. The msg_namelen update is also missing for datagram sockets
in case the socket is shutting down during receive.
Fix both issues by setting msg_namelen to 0 early. It will be
updated later if we're going to fill the msg_name member.
Signed-off-by: Mathias Krause <minipli@googlemail.com> Cc: Arnaldo Carvalho de Melo <acme@ghostprotocols.net> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The current code does not fill the msg_name member in case it is set.
It also does not set the msg_namelen member to 0 and therefore makes
net/socket.c leak the local, uninitialized sockaddr_storage variable
to userland -- 128 bytes of kernel stack memory.
Fix that by simply setting msg_namelen to 0 as obviously nobody cared
about iucv_sock_recvmsg() not filling the msg_name in case it was set.
Signed-off-by: Mathias Krause <minipli@googlemail.com> Cc: Ursula Braun <ursula.braun@de.ibm.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The current code does not fill the msg_name member in case it is set.
It also does not set the msg_namelen member to 0 and therefore makes
net/socket.c leak the local, uninitialized sockaddr_storage variable
to userland -- 128 bytes of kernel stack memory.
Fix that by simply setting msg_namelen to 0 as obviously nobody cared
about irda_recvmsg_dgram() not filling the msg_name in case it was
set.
Signed-off-by: Mathias Krause <minipli@googlemail.com> Cc: Samuel Ortiz <samuel@sortiz.org> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The current code does not fill the msg_name member in case it is set.
It also does not set the msg_namelen member to 0 and therefore makes
net/socket.c leak the local, uninitialized sockaddr_storage variable
to userland -- 128 bytes of kernel stack memory.
Fix that by simply setting msg_namelen to 0 as obviously nobody cared
about caif_seqpkt_recvmsg() not filling the msg_name in case it was
set.
Signed-off-by: Mathias Krause <minipli@googlemail.com> Cc: Sjur Braendeland <sjur.brandeland@stericsson.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If RFCOMM_DEFER_SETUP is set in the flags, rfcomm_sock_recvmsg() returns
early with 0 without updating the possibly set msg_namelen member. This,
in turn, leads to a 128 byte kernel stack leak in net/socket.c.
Fix this by updating msg_namelen in this case. For all other cases it
will be handled in bt_sock_stream_recvmsg().
Signed-off-by: Mathias Krause <minipli@googlemail.com> Cc: Marcel Holtmann <marcel@holtmann.org> Cc: Gustavo Padovan <gustavo@padovan.org> Cc: Johan Hedberg <johan.hedberg@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In case the socket is already shutting down, bt_sock_recvmsg() returns
with 0 without updating msg_namelen leading to net/socket.c leaking the
local, uninitialized sockaddr_storage variable to userland -- 128 bytes
of kernel stack memory.
Fix this by moving the msg_namelen assignment in front of the shutdown
test.
Signed-off-by: Mathias Krause <minipli@googlemail.com> Cc: Marcel Holtmann <marcel@holtmann.org> Cc: Gustavo Padovan <gustavo@padovan.org> Cc: Johan Hedberg <johan.hedberg@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
When msg_namelen is non-zero the sockaddr info gets filled out, as
requested, but the code fails to initialize the padding bytes of struct
sockaddr_ax25 inserted by the compiler for alignment. Additionally the
msg_namelen value is updated to sizeof(struct full_sockaddr_ax25) but is
not always filled up to this size.
Both issues lead to the fact that the code will leak uninitialized
kernel stack bytes in net/socket.c.
Fix both issues by initializing the memory with memset(0).
Signed-off-by: Mathias Krause <minipli@googlemail.com> Cc: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The current code does not fill the msg_name member in case it is set.
It also does not set the msg_namelen member to 0 and therefore makes
net/socket.c leak the local, uninitialized sockaddr_storage variable
to userland -- 128 bytes of kernel stack memory.
Fix that by simply setting msg_namelen to 0 as obviously nobody cared
about vcc_recvmsg() not filling the msg_name in case it was set.
Signed-off-by: Mathias Krause <minipli@googlemail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
As reported by Dave Kleikamp, when we emit cross calls to do batched
TLB flush processing we have a race because we do not synchronize on
the sibling cpus completing the cross call.
So meanwhile the TLB batch can be reset (tb->tlb_nr set to zero, etc.)
and either flushes are missed or flushes will flush the wrong
addresses.
Fix this by using generic infrastructure to synchonize on the
completion of the cross call.
This first required getting the flush_tlb_pending() call out from
switch_to() which operates with locks held and interrupts disabled.
The problem is that smp_call_function_many() cannot be invoked with
IRQs disabled and this is explicitly checked for with WARN_ON_ONCE().
We get the batch processing outside of locked IRQ disabled sections by
using some ideas from the powerpc port. Namely, we only batch inside
of arch_{enter,leave}_lazy_mmu_mode() calls. If we're not in such a
region, we flush TLBs synchronously.
1) Get rid of xcall_flush_tlb_pending and per-cpu type
implementations.
a) Add 'active' member to struct tlb_batch
b) Define __HAVE_ARCH_ENTER_LAZY_MMU_MODE
c) Set 'active' in arch_enter_lazy_mmu_mode()
d) Run batch and clear 'active' in arch_leave_lazy_mmu_mode()
e) Check 'active' in tlb_batch_add_one() and do a synchronous
flush if it's clear.
4) Add infrastructure for synchronous TLB page flushes.
a) Implement __flush_tlb_page and per-cpu variants, patch
as needed.
b) Likewise for xcall_flush_tlb_page.
c) Implement smp_flush_tlb_page() to invoke the cross-call.
d) Wire up global_flush_tlb_page() to the right routine based
upon CONFIG_SMP
5) It turns out that singleton batches are very common, 2 out of every
3 batch flushes have only a single entry in them.
The batch flush waiting is very expensive, both because of the poll
on sibling cpu completeion, as well as because passing the tlb batch
pointer to the sibling cpus invokes a shared memory dereference.
Therefore, in flush_tlb_pending(), if there is only one entry in
the batch perform a completely asynchronous global_flush_tlb_page()
instead.
Reported-by: Dave Kleikamp <dave.kleikamp@oracle.com> Signed-off-by: David S. Miller <davem@davemloft.net> Acked-by: Dave Kleikamp <dave.kleikamp@oracle.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In commit b0de59b5733d ("TTY: do not update atime/mtime on read/write")
we removed timestamps from tty inodes to fix a security issue and waited
if something breaks. Well, 'w', the utility to find out logged users
and their inactivity time broke. It shows that users are inactive since
the time they logged in.
To revert to the old behaviour while still preventing attackers to
guess the password length, we update the timestamps in one-minute
intervals by this patch.
On http://vladz.devzero.fr/013_ptmx-timing.php, we can see how to find
out length of a password using timestamps of /dev/ptmx. It is
documented in "Timing Analysis of Keystrokes and Timing Attacks on
SSH". To avoid that problem, do not update time when reading
from/writing to a TTY.
I am afraid of regressions as this is a behavior we have since 0.97
and apps may expect the time to be current, e.g. for monitoring
whether there was a change on the TTY. Now, there is no change. So
this would better have a lot of testing before it goes upstream.
While trying to track down a tree log replay bug I noticed that fsck was always
complaining about nbytes not being right for our fsynced file. That is because
the new fsync stuff doesn't wait for ordered extents to complete, so the inodes
nbytes are not necessarily updated properly when we log it. So to fix this we
need to set nbytes to whatever it is on the inode that is on disk, so when we
replay the extents we can just add the bytes that are being added as we replay
the extent. This makes it work for the case that we have the wrong nbytes or
the case that we logged everything and nbytes is actually correct. With this
I'm no longer getting nbytes errors out of btrfsck.
Signed-off-by: Josef Bacik <jbacik@fusionio.com> Signed-off-by: Chris Mason <chris.mason@fusionio.com> Signed-off-by: Lingzhu Xiang <lxiang@redhat.com> Reviewed-by: CAI Qian <caiqian@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This is my example conversion of a few existing mmap users. The mtdchar
case is actually disabled right now (and stays disabled), but I did it
because it showed up on my "git grep", and I was familiar with the code
due to fixing an overflow problem in the code in commit 9c603e53d380
("mtdchar: fix offset overflow detection").
This is my example conversion of a few existing mmap users. The HPET
case is simple, widely available, and easy to test (Clemens Ladisch sent
a trivial test-program for it).
This is my example conversion of a few existing mmap users. The
fb_mmap() case is a good example because it is a bit more complicated
than some: fb_mmap() mmaps one of two different memory areas depending
on the page offset of the mmap (but happily there is never any mixing of
the two, so the helper function still works).
Various drivers end up replicating the code to mmap() their memory
buffers into user space, and our core memory remapping function may be
very flexible but it is unnecessarily complicated for the common cases
to use.
Our internal VM uses pfn's ("page frame numbers") which simplifies
things for the VM, and allows us to pass physical addresses around in a
denser and more efficient format than passing a "phys_addr_t" around,
and having to shift it up and down by the page size. But it just means
that drivers end up doing that shifting instead at the interface level.
It also means that drivers end up mucking around with internal VM things
like the vma details (vm_pgoff, vm_start/end) way more than they really
need to.
So this just exports a function to map a certain physical memory range
into user space (using a phys_addr_t based interface that is much more
natural for a driver) and hides all the complexity from the driver.
Some drivers will still end up tweaking the vm_page_prot details for
things like prefetching or cacheability etc, but that's actually
relevant to the driver, rather than caring about what the page offset of
the mapping is into the particular IO memory region.
Okay so Alan's patch handled the case where there was no registered fbcon,
however the other path entered in set_con2fb_map pit.
In there we called fbcon_takeover, but we also took the console lock in a couple
of places. So push the console lock out to the callers of set_con2fb_map,
this means fbmem and switcheroo needed to take the lock around the fb notifier
entry points that lead to this.
This should fix the efifb regression seen by Maarten.
Tested-by: Maarten Lankhorst <maarten.lankhorst@canonical.com> Tested-by: Lu Hua <huax.lu@intel.com> Signed-off-by: Dave Airlie <airlied@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The valid mask for both offcore_response_0 and
offcore_response_1 was wrong for SNB/SNB-EP,
IVB/IVB-EP. It was possible to write to
reserved bit and cause a GP fault crashing
the kernel.
This patch fixes the problem by correctly marking the
reserved bits in the valid mask for all the processors
mentioned above.
A distinction between desktop and server parts is introduced
because bits 24-30 are only available on the server parts.
This version of the patch is just a rebase to perf/urgent tree
and should apply to older kernels as well.
Trinity discovered that we fail to check all 64 bits of
attr.config passed by user space, resulting to out-of-bounds
access of the perf_swevent_enabled array in
sw_perf_event_destroy().
Introduced in commit b0a873ebb ("perf: Register PMU
implementations").
Signed-off-by: Tommi Rantala <tt.rantala@gmail.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: davej@redhat.com Cc: Paul Mackerras <paulus@samba.org> Cc: Arnaldo Carvalho de Melo <acme@ghostprotocols.net> Link: http://lkml.kernel.org/r/1365882554-30259-1-git-send-email-tt.rantala@gmail.com Signed-off-by: Ingo Molnar <mingo@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The current code does not set the msg_namelen member to 0 and therefore
makes net/socket.c leak the local sockaddr_storage variable to userland
-- 128 bytes of kernel stack memory. Fix that.
And make use of it in b43. This fixes a regression introduced with 49d55cef5b1925a5c1efb6aaddaa40fc7c693335
b43: N-PHY: implement spurious tone avoidance
This commit made BCM4322 use only MCS 0 on channel 13, which of course
resulted in performance drop (down to 0.7Mb/s).
Reported-by: Stefan Brüns <stefan.bruens@rwth-aachen.de> Signed-off-by: Rafał Miłecki <zajec5@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The hardware parsing of Control Wrapper Frames needs to be disabled, as
it has been causing spurious decryption error reports. The initvals for
other chips have been updated to disable it, but AR9580 was left out for
some reason.
Signed-off-by: Felix Fietkau <nbd@openwrt.org> Signed-off-by: John W. Linville <linville@tuxdriver.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Since the firmware has been open sourced, the minor version has been
bumped to 1.4 and the API/ABI will stay compatible across further 1.x
releases.
Signed-off-by: Felix Fietkau <nbd@openwrt.org> Signed-off-by: John W. Linville <linville@tuxdriver.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Events may be created with attr->disabled == 1 and attr->enable_on_exec
== 1, which confuses the group validation code because events with the
PERF_EVENT_STATE_OFF are not considered candidates for scheduling, which
may lead to failure at group scheduling time.
This patch fixes the validation check for ARM, so that events in the
OFF state are still considered when enable_on_exec is true.
Reported-by: Sudeep KarkadaNagesha <Sudeep.KarkadaNagesha@arm.com> Cc: Peter Zijlstra <a.p.zijlstra@chello.nl> Cc: Arnaldo Carvalho de Melo <acme@ghostprotocols.net> Cc: Jiri Olsa <jolsa@redhat.com> Signed-off-by: Will Deacon <will.deacon@arm.com> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
On Feroceon the L2 cache becomes non-coherent with the CPU
when the L1 caches are disabled. Thus the L2 needs to be invalidated
after both L1 caches are disabled.
On kexec before the starting the code for relocation the kernel,
the L1 caches are disabled in cpu_froc_fin (cpu_v7_proc_fin for Feroceon),
but after L2 cache is never invalidated, because inv_all is not set
in cache-feroceon-l2.c.
So kernel relocation and decompression may has (and usually has) errors.
Setting the function enables L2 invalidation and fixes the issue.
Signed-off-by: Illia Ragozin <illia.ragozin@grapecom.com> Acked-by: Jason Cooper <jason@lakedaemon.net> Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
try_to_wake_up_local() should only be invoked to wake up another
task in the same runqueue and BUG_ON()s are used to enforce the
rule. Missing try_to_wake_up_local() can stall workqueue
execution but such stalls are likely to be finite either by
another work item being queued or the one blocked getting
unblocked. There's no reason to trigger BUG while holding rq
lock crashing the whole system.
Convert BUG_ON()s in try_to_wake_up_local() to WARN_ON_ONCE()s.
This patch adds support for kvm_gfn_to_hva_cache_init functions for
reads and writes that will cross a page. If the range falls within
the same memslot, then this will be a fast operation. If the range
is split between two memslots, then the slower kvm_read_guest and
kvm_write_guest are used.
Tested: Test against kvm_clock unit tests.
Signed-off-by: Andrew Honig <ahonig@google.com> Signed-off-by: Gleb Natapov <gleb@redhat.com> Cc: Ben Hutchings <ben@decadent.org.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If the guest specifies a IOAPIC_REG_SELECT with an invalid value and follows
that with a read of the IOAPIC_REG_WINDOW KVM does not properly validate
that request. ioapic_read_indirect contains an
ASSERT(redir_index < IOAPIC_NUM_PINS), but the ASSERT has no effect in
non-debug builds. In recent kernels this allows a guest to cause a kernel
oops by reading invalid memory. In older kernels (pre-3.3) this allows a
guest to read from large ranges of host memory.
Tested: tested against apic unit tests.
Signed-off-by: Andrew Honig <ahonig@google.com> Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com> Cc: Ben Hutchings <ben@decadent.org.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
There is a potential use after free issue with the handling of
MSR_KVM_SYSTEM_TIME. If the guest specifies a GPA in a movable or removable
memory such as frame buffers then KVM might continue to write to that
address even after it's removed via KVM_SET_USER_MEMORY_REGION. KVM pins
the page in memory so it's unlikely to cause an issue, but if the user
space component re-purposes the memory previously used for the guest, then
the guest will be able to corrupt that memory.
Tested: Tested against kvmclock unit test
Signed-off-by: Andrew Honig <ahonig@google.com> Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com> Cc: Ben Hutchings <ben@decadent.org.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If the guest sets the GPA of the time_page so that the request to update the
time straddles a page then KVM will write onto an incorrect page. The
write is done byusing kmap atomic to get a pointer to the page for the time
structure and then performing a memcpy to that page starting at an offset
that the guest controls. Well behaved guests always provide a 32-byte aligned
address, however a malicious guest could use this to corrupt host kernel
memory.
Tested: Tested against kvmclock unit test.
Signed-off-by: Andrew Honig <ahonig@google.com> Signed-off-by: Marcelo Tosatti <mtosatti@redhat.com> Cc: Ben Hutchings <ben@decadent.org.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
With applying the previous patch "hugetlbfs: stop setting VM_DONTDUMP in
initializing vma(VM_HUGETLB)" to reenable hugepage coredump, if a memory
error happens on a hugepage and the affected processes try to access the
error hugepage, we hit VM_BUG_ON(atomic_read(&page->_count) <= 0) in
get_page().
The reason for this bug is that coredump-related code doesn't recognise
"hugepage hwpoison entry" with which a pmd entry is replaced when a memory
error occurs on a hugepage.
In other words, physical address information is stored in different bit
layout between hugepage hwpoison entry and pmd entry, so
follow_hugetlb_page() which is called in get_dump_page() returns a wrong
page from a given address.
With this patch, we can call hugetlb_fault() and take proper actions (we
wait for migration entries, fail with VM_FAULT_HWPOISON_LARGE for
hwpoisoned entries,) and as the result we can dump all hugepages except
for hwpoisoned ones.
Signed-off-by: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com> Cc: Rik van Riel <riel@redhat.com> Acked-by: Michal Hocko <mhocko@suse.cz> Cc: HATAYAMA Daisuke <d.hatayama@jp.fujitsu.com> Acked-by: KOSAKI Motohiro <kosaki.motohiro@jp.fujitsu.com> Acked-by: David Rientjes <rientjes@google.com> 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@linuxfoundation.org>
To get correct endianes on little endian cpus (like arm) while reading device
tree properties, this patch replaces of_get_property() with
of_property_read_u32(). While there use of_property_read_bool() for the
handling of the boolean "nxp,no-comparator-bypass" property.
Signed-off-by: Christoph Fritz <chf.fritz@googlemail.com> Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The current code makes the assumption that a cpu_base lock won't be
held if the CPU corresponding to that cpu_base is offline, which isn't
always true.
If a hrtimer is not queued, then it will not be migrated by
migrate_hrtimers() when a CPU is offlined. Therefore, the hrtimer's
cpu_base may still point to a CPU which has subsequently gone offline
if the timer wasn't enqueued at the time the CPU went down.
Normally this wouldn't be a problem, but a cpu_base's lock is blindly
reinitialized each time a CPU is brought up. If a CPU is brought
online during the period that another thread is performing a hrtimer
operation on a stale hrtimer, then the lock will be reinitialized
under its feet, and a SPIN_BUG() like the following will be observed:
<0>[ 28.082085] BUG: spinlock already unlocked on CPU#0, swapper/0/0
<0>[ 28.087078] lock: 0xc4780b40, value 0x0 .magic: dead4ead, .owner: <none>/-1, .owner_cpu: -1
<4>[ 42.451150] [<c0014398>] (unwind_backtrace+0x0/0x120) from [<c0269220>] (do_raw_spin_unlock+0x44/0xdc)
<4>[ 42.460430] [<c0269220>] (do_raw_spin_unlock+0x44/0xdc) from [<c071b5bc>] (_raw_spin_unlock+0x8/0x30)
<4>[ 42.469632] [<c071b5bc>] (_raw_spin_unlock+0x8/0x30) from [<c00a9ce0>] (__hrtimer_start_range_ns+0x1e4/0x4f8)
<4>[ 42.479521] [<c00a9ce0>] (__hrtimer_start_range_ns+0x1e4/0x4f8) from [<c00aa014>] (hrtimer_start+0x20/0x28)
<4>[ 42.489247] [<c00aa014>] (hrtimer_start+0x20/0x28) from [<c00e6190>] (rcu_idle_enter_common+0x1ac/0x320)
<4>[ 42.498709] [<c00e6190>] (rcu_idle_enter_common+0x1ac/0x320) from [<c00e6440>] (rcu_idle_enter+0xa0/0xb8)
<4>[ 42.508259] [<c00e6440>] (rcu_idle_enter+0xa0/0xb8) from [<c000f268>] (cpu_idle+0x24/0xf0)
<4>[ 42.516503] [<c000f268>] (cpu_idle+0x24/0xf0) from [<c06ed3c0>] (rest_init+0x88/0xa0)
<4>[ 42.524319] [<c06ed3c0>] (rest_init+0x88/0xa0) from [<c0c00978>] (start_kernel+0x3d0/0x434)
As an example, this particular crash occurred when hrtimer_start() was
executed on CPU #0. The code locked the hrtimer's current cpu_base
corresponding to CPU #1. CPU #0 then tried to switch the hrtimer's
cpu_base to an optimal CPU which was online. In this case, it selected
the cpu_base corresponding to CPU #3.
Before it could proceed, CPU #1 came online and reinitialized the
spinlock corresponding to its cpu_base. Thus now CPU #0 held a lock
which was reinitialized. When CPU #0 finally ended up unlocking the
old cpu_base corresponding to CPU #1 so that it could switch to CPU
#3, we hit this SPIN_BUG() above while in switch_hrtimer_base().
CPU #0 CPU #1
---- ----
... <offline>
hrtimer_start()
lock_hrtimer_base(base #1)
... init_hrtimers_cpu()
switch_hrtimer_base() ...
... raw_spin_lock_init(&cpu_base->lock)
raw_spin_unlock(&cpu_base->lock) ...
<spin_bug>
This code was broken because it assumed that all MTD devices were map-based.
Disable it for now, until it can be fixed properly for the next merge window.
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
It would cause no link after suspending or shutdowning when the
nic changes the speed to 10M and connects to a link partner which
forces the speed to 100M.
Check the link partner ability to determine which speed to set.
Signed-off-by: Hayes Wang <hayeswang@realtek.com> Acked-by: Francois Romieu <romieu@fr.zoreil.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Anatol Pomozov identified a race condition that hits module unloading
and re-loading. To quote Anatol:
"This is a race codition that exists between kset_find_obj() and
kobject_put(). kset_find_obj() might return kobject that has refcount
equal to 0 if this kobject is freeing by kobject_put() in other
thread.
Here is timeline for the crash in case if kset_find_obj() searches for
an object tht nobody holds and other thread is doing kobject_put() on
the same kobject:
THREAD A (calls kset_find_obj()) THREAD B (calls kobject_put())
splin_lock()
atomic_dec_return(kobj->kref), counter gets zero here
... starts kobject cleanup ....
spin_lock() // WAIT thread A in kobj_kset_leave()
iterate over kset->list
atomic_inc(kobj->kref) (counter becomes 1)
spin_unlock()
spin_lock() // taken
// it does not know that thread A increased counter so it
remove obj from list
spin_unlock()
vfree(module) // frees module object with containing kobj
// kobj points to freed memory area!!
kobject_put(kobj) // OOPS!!!!
The race above happens because module.c tries to use kset_find_obj()
when somebody unloads module. The module.c code was introduced in
commit 6494a93d55fa"
Anatol supplied a patch specific for module.c that worked around the
problem by simply not using kset_find_obj() at all, but rather than make
a local band-aid, this just fixes kset_find_obj() to be thread-safe
using the proper model of refusing the get a new reference if the
refcount has already dropped to zero.
See examples of this proper refcount handling not only in the kref
documentation, but in various other equivalent uses of this pattern by
grepping for atomic_inc_not_zero().
[ Side note: the module race does indicate that module loading and
unloading is not properly serialized wrt sysfs information using the
module mutex. That may require further thought, but this is the
correct fix at the kobject layer regardless. ]
Sasha Levin has been running trinity in a KVM tools guest, and was able
to trigger the BUG_ON() at arch/x86/mm/pat.c:279 (verifying the range of
the memory type). The call trace showed that it was mtdchar_mmap() that
created an invalid remap_pfn_range().
The problem is that mtdchar_mmap() does various really odd and subtle
things with the vma page offset etc, and uses the wrong types (and the
wrong overflow) detection for it.
For example, the page offset may well be 32-bit on a 32-bit
architecture, but after shifting it up by PAGE_SHIFT, we need to use a
potentially 64-bit resource_size_t to correctly hold the full value.
Also, we need to check that the vma length plus offset doesn't overflow
before we check that it is smaller than the length of the mtdmap region.
This fixes things up and tries to make the code a bit easier to read.
Invoking arch_flush_lazy_mmu_mode() results in calls to
preempt_enable()/disable() which may have performance impact.
Since lazy MMU is not used on bare metal we can patch away
arch_flush_lazy_mmu_mode() so that it is never called in such
environment.
[ hpa: the previous patch "Fix vmalloc_fault oops during lazy MMU
updates" may cause a minor performance regression on
bare metal. This patch resolves that performance regression. It is
somewhat unclear to me if this is a good -stable candidate. ]
Signed-off-by: Boris Ostrovsky <boris.ostrovsky@oracle.com> Link: http://lkml.kernel.org/r/1364045796-10720-2-git-send-email-konrad.wilk@oracle.com Tested-by: Josh Boyer <jwboyer@redhat.com> Tested-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> Acked-by: Borislav Petkov <bp@suse.de> Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In paravirtualized x86_64 kernels, vmalloc_fault may cause an oops
when lazy MMU updates are enabled, because set_pgd effects are being
deferred.
One instance of this problem is during process mm cleanup with memory
cgroups enabled. The chain of events is as follows:
- zap_pte_range enables lazy MMU updates
- zap_pte_range eventually calls mem_cgroup_charge_statistics,
which accesses the vmalloc'd mem_cgroup per-cpu stat area
- vmalloc_fault is triggered which tries to sync the corresponding
PGD entry with set_pgd, but the update is deferred
- vmalloc_fault oopses due to a mismatch in the PUD entries
The sched_clock_remote() implementation has the following inatomicity
problem on 32bit systems when accessing the remote scd->clock, which
is a 64bit value.
While the update of scd->clock is using an atomic64 mechanism, the
readout on the remote cpu is not, which can cause completely bogus
readouts.
It is a quite rare problem, because it requires the update to hit the
narrow race window between the low/high readout and the update must go
across the 32bit boundary.
The resulting misbehaviour is, that CPU1 will see the sched_clock on
CPU1 ~4 seconds ahead of it's own and update CPU1s sched_clock value
to this bogus timestamp. This stays that way due to the clamping
implementation for about 4 seconds until the synchronization with
CLOCK_MONOTONIC undoes the problem.
The issue is hard to observe, because it might only result in a less
accurate SCHED_OTHER timeslicing behaviour. To create observable
damage on realtime scheduling classes, it is necessary that the bogus
update of CPU1 sched_clock happens in the context of an realtime
thread, which then gets charged 4 seconds of RT runtime, which results
in the RT throttler mechanism to trigger and prevent scheduling of RT
tasks for a little less than 4 seconds. So this is quite unlikely as
well.
The issue was quite hard to decode as the reproduction time is between
2 days and 3 weeks and intrusive tracing makes it less likely, but the
following trace recorded with trace_clock=global, which uses
sched_clock_local(), gave the final hint:
What happens is that CPU0 goes idle and invokes
sched_clock_idle_sleep_event() which invokes sched_clock_local() and
CPU1 runs a remote wakeup for CPU0 at the same time, which invokes
sched_remote_clock(). The time jump gets propagated to CPU0 via
sched_remote_clock() and stays stale on both cores for ~4 seconds.
There are only two other possibilities, which could cause a stale
sched clock:
1) ktime_get() which reads out CLOCK_MONOTONIC returns a sporadic
wrong value.
2) sched_clock() which reads the TSC returns a sporadic wrong value.
#1 can be excluded because sched_clock would continue to increase for
one jiffy and then go stale.
#2 can be excluded because it would not make the clock jump
forward. It would just result in a stale sched_clock for one jiffy.
After quite some brain twisting and finding the same pattern on other
traces, sched_clock_remote() remained the only place which could cause
such a problem and as explained above it's indeed racy on 32bit
systems.
So while on 64bit systems the readout is atomic, we need to verify the
remote readout on 32bit machines. We need to protect the local->clock
readout in sched_clock_remote() on 32bit as well because an NMI could
hit between the low and the high readout, call sched_clock_local() and
modify local->clock.
Thanks to Siegfried Wulsch for bearing with my debug requests and
going through the tedious tasks of running a bunch of reproducer
systems to generate the debug information which let me decode the
issue.
Reported-by: Siegfried Wulsch <Siegfried.Wulsch@rovema.de> Acked-by: Peter Zijlstra <peterz@infradead.org> Cc: Steven Rostedt <rostedt@goodmis.org> Link: http://lkml.kernel.org/r/alpine.LFD.2.02.1304051544160.21884@ionos Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This function is intended to simplify locking around refcounting for
objects that can be looked up from a lookup structure, and which are
removed from that lookup structure in the object destructor.
Operations on such objects require at least a read lock around
lookup + kref_get, and a write lock around kref_put + remove from lookup
structure. Furthermore, RCU implementations become extremely tricky.
With a lookup followed by a kref_get_unless_zero *with return value check*
locking in the kref_put path can be deferred to the actual removal from
the lookup structure and RCU lookups become trivial.
v2: Formatting fixes.
v3: Invert the return value.
Signed-off-by: Thomas Hellstrom <thellstrom@vmware.com> Signed-off-by: Dave Airlie <airlied@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Revert commit 62a3ddef6181 ("vfs: fix spinning prevention in prune_icache_sb").
This commit doesn't look right: since we are looking at the tail of the
list (sb->s_inode_lru.prev) if we want to skip an inode, we should put
it back at the head of the list instead of the tail, otherwise we will
keep spinning on it.
Discovered when investigating why prune_icache_sb came top in perf
reports of a swapping load.
This patch fixes a bug where a handful of informational / control CDBs
that should be allowed during ALUA access state Standby/Offline/Transition
where incorrectly returning CHECK_CONDITION + ASCQ_04H_ALUA_TG_PT_*.
This includes INQUIRY + REPORT_LUNS, which would end up preventing LUN
registration when LUN scanning occured during these ALUA access states.
Signed-off-by: Nicholas Bellinger <nab@linux-iscsi.org> Cc: Hannes Reinecke <hare@suse.de> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
If a result of the SMP discover function is PHY VACANT,
the content of discover response structure (dr) is not valid.
It sometimes happens that dr->attached_sas_addr can contain
even SAS address of other phy. In such case an invalid phy
is created, what causes NULL pointer dereference during
destruction of expander's phys.
So if a result of SMP function is PHY VACANT, the content of discover
response structure (dr) must not be copied to phy structure.
The current layout is to place the per-process tables at the end of the
GTT. However, this is currently using a hardcoded maximum size for the GTT
and not taking in account limitations imposed by the BIOS. Use the value
for the total number of entries allocated in the table as provided by
the configuration registers.
Reported-by: Matthew Garrett <mjg@redhat.com> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Cc: Daniel Vetter <daniel.vetter@ffwll.ch> Cc: Ben Widawsky <ben@bwidawsk.net> Cc: Matthew Garret <mjg@redhat.com> Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch> Cc: Jonathan Nieder <jrnieder@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
As commit 40dc166c (PM / Core: Introduce struct syscore_ops for core
subsystems PM) say, syscore_ops operations should be carried with one
CPU on-line and interrupts disabled. However, after commit f96972f2d
(kernel/sys.c: call disable_nonboot_cpus() in kernel_restart()),
syscore_shutdown() is called before disable_nonboot_cpus(), so break
the rules. We have a MIPS machine with a 8259A PIC, and there is an
external timer (HPET) linked at 8259A. Since 8259A has been shutdown
too early (by syscore_shutdown()), disable_nonboot_cpus() runs without
timer interrupt, so it hangs and reboot fails. This patch call
syscore_shutdown() a little later (after disable_nonboot_cpus()) to
avoid reboot failure, this is the same way as poweroff does.
For consistency, add disable_nonboot_cpus() to kernel_halt().
Signed-off-by: Huacai Chen <chenhc@lemote.com> Signed-off-by: Rafael J. Wysocki <rafael.j.wysocki@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The Charge Pump needs the DSP clock to work properly, without it the
bypass to HP/LINEOUT is not working properly. This requirement is not
mentioned in the datasheet but has been confirmed by Mark Brown from
Wolfson.
Signed-off-by: Alban Bedel <alban.bedel@avionic-design.de> Signed-off-by: Mark Brown <broonie@opensource.wolfsonmicro.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The usb_control_msg() function expects __u16 types and performs
the endianness conversions by itself.
However, in three places, a conversion is performed before it is
handed over to usb_control_msg(), which leads to a double conversion
(= no conversion):
* snd_usb_nativeinstruments_boot_quirk()
* snd_nativeinstruments_control_get()
* snd_nativeinstruments_control_put()
Caught by sparse:
sound/usb/mixer_quirks.c:512:38: warning: incorrect type in argument 6 (different base types)
sound/usb/mixer_quirks.c:512:38: expected unsigned short [unsigned] [usertype] index
sound/usb/mixer_quirks.c:512:38: got restricted __le16 [usertype] <noident>
sound/usb/mixer_quirks.c:543:35: warning: incorrect type in argument 5 (different base types)
sound/usb/mixer_quirks.c:543:35: expected unsigned short [unsigned] [usertype] value
sound/usb/mixer_quirks.c:543:35: got restricted __le16 [usertype] <noident>
sound/usb/mixer_quirks.c:543:56: warning: incorrect type in argument 6 (different base types)
sound/usb/mixer_quirks.c:543:56: expected unsigned short [unsigned] [usertype] index
sound/usb/mixer_quirks.c:543:56: got restricted __le16 [usertype] <noident>
sound/usb/quirks.c:502:35: warning: incorrect type in argument 5 (different base types)
sound/usb/quirks.c:502:35: expected unsigned short [unsigned] [usertype] value
sound/usb/quirks.c:502:35: got restricted __le16 [usertype] <noident>
BugLink: http://bugs.launchpad.net/bugs/1128840
It appears that when this register read fails it never recovers, so
I think there is no need to repeat the same error message ad infinitum.
Signed-off-by: Tim Gardner <tim.gardner@canonical.com> Cc: Ivo van Doorn <IvDoorn@gmail.com> Cc: Gertjan van Wingerde <gwingerde@gmail.com> Cc: Helmut Schaa <helmut.schaa@googlemail.com> Cc: "John W. Linville" <linville@tuxdriver.com> Cc: linux-wireless@vger.kernel.org Cc: users@rt2x00.serialmonkey.com Cc: netdev@vger.kernel.org Cc: stable@vger.kernel.org Signed-off-by: John W. Linville <linville@tuxdriver.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
find_vma() can be called by multiple threads with read lock
held on mm->mmap_sem and any of them can update mm->mmap_cache.
Prevent compiler from re-fetching mm->mmap_cache, because other
readers could update it in the meantime:
thread 1 thread 2
|
find_vma() | find_vma()
struct vm_area_struct *vma = NULL; |
vma = mm->mmap_cache; |
if (!(vma && vma->vm_end > addr |
&& vma->vm_start <= addr)) { |
| mm->mmap_cache = vma;
return vma; |
^^ compiler may optimize this |
local variable out and re-read |
mm->mmap_cache |
This issue can be reproduced with gcc-4.8.0-1 on s390x by running
mallocstress testcase from LTP, which triggers:
panic_lock is meant to ensure that panic processing takes place only on
one cpu; if any of the other cpus encounter a panic, they will spin
waiting to be shut down.
However, this causes a regression in this scenario:
1. Cpu 0 encounters a panic and acquires the panic_lock
and proceeds with the panic processing.
2. There is an interrupt on cpu 0 that also encounters
an error condition and invokes panic.
3. This second invocation fails to acquire the panic_lock
and enters the infinite while loop in panic_smp_self_stop.
Thus all panic processing is stopped, and the cpu is stuck for eternity
in the while(1) inside panic_smp_self_stop.
To address this, disable local interrupts with local_irq_disable before
acquiring the panic_lock. This will prevent interrupt handlers from
executing during the panic processing, thus avoiding this particular
problem.
Signed-off-by: Vikram Mulukutla <markivx@codeaurora.org> Reviewed-by: Stephen Boyd <sboyd@codeaurora.org> Cc: Michael Holzheu <holzheu@linux.vnet.ibm.com> Signed-off-by: Andrew Morton <akpm@linux-foundation.org> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Cc: Paul Gortmaker <paul.gortmaker@windriver.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The return code from the registration of the thermal class is used to
unallocate resources, but this failure isn't passed back to the caller of
thermal_init. Return this failure back to the caller.
This bug was introduced in changeset 4cb18728 which overwrote the return code
when the variable was re-used to catch the return code of the registration of
the genetlink thermal socket family.
Signed-off-by: Richard Guy Briggs <rbriggs@redhat.com> Signed-off-by: Zhang Rui <rui.zhang@intel.com> Cc: Jonghwan Choi <jhbird.choi@samsung.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
eboot.o and efi_stub_$(BITS).o didn't get added to "targets", and hence
their .cmd files don't get included by the build machinery, leading to
the files always getting rebuilt.
Rather than adding the two files individually, take the opportunity and
add $(VMLINUX_OBJS) to "targets" instead, thus allowing the assignment
at the top of the file to be shrunk quite a bit.
At the same time, remove a pointless flags override line - the variable
assigned to was misspelled anyway, and the options added are
meaningless for assembly sources.
[ hpa: the patch is not minimal, but I am taking it for -urgent anyway
since the excess impact of the patch seems to be small enough. ]
Signed-off-by: Jan Beulich <jbeulich@suse.com> Link: http://lkml.kernel.org/r/515C5D2502000078000CA6AD@nat28.tlf.novell.com Cc: Matthew Garrett <mjg@redhat.com> Cc: Matt Fleming <matt.fleming@intel.com> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
As found by gcc-4.8, the QUEUE_SYSFS_BIT_FNS macro creates functions
that use a value generated by queue_var_store independent of whether
that value was set or not.
block/blk-sysfs.c: In function 'queue_store_nonrot':
block/blk-sysfs.c:244:385: warning: 'val' may be used uninitialized in this function [-Wmaybe-uninitialized]
Unlike most other such warnings, this one is not a false positive,
writing any non-number string into the sysfs files indeed has
an undefined result, rather than returning an error.
rfc4543(gcm(*)) code for GMAC assumes that assoc scatterlist always contains
only one segment and only makes use of this first segment. However ipsec passes
assoc with three segments when using 'extended sequence number' thus in this
case rfc4543(gcm(*)) fails to function correctly. Patch fixes this issue.
Reported-by: Chaoxing Lin <Chaoxing.Lin@ultra-3eti.com> Tested-by: Chaoxing Lin <Chaoxing.Lin@ultra-3eti.com> Signed-off-by: Jussi Kivilinna <jussi.kivilinna@iki.fi> Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In UP and non-preempt respectively, the spinlocks and preemption
disable/enable points are stubbed out entirely, because there is no
regular code that can ever hit the kind of concurrency they are meant to
protect against.
However, while there is no regular code that can cause scheduling, we
_do_ end up having some exceptional (literally!) code that can do so,
and that we need to make sure does not ever get moved into the critical
region by the compiler.
In particular, get_user() and put_user() is generally implemented as
inline asm statements (even if the inline asm may then make a call
instruction to call out-of-line), and can obviously cause a page fault
and IO as a result. If that inline asm has been scheduled into the
middle of a preemption-safe (or spinlock-protected) code region, we
obviously lose.
Now, admittedly this is *very* unlikely to actually ever happen, and
we've not seen examples of actual bugs related to this. But partly
exactly because it's so hard to trigger and the resulting bug is so
subtle, we should be extra careful to get this right.
So make sure that even when preemption is disabled, and we don't have to
generate any actual *code* to explicitly tell the system that we are in
a preemption-disabled region, we need to at least tell the compiler not
to move things around the critical region.
This patch grew out of the same discussion that caused commits 79e5f05edcbf ("ARC: Add implicit compiler barrier to raw_local_irq*
functions") and 3e2e0d2c222b ("tile: comment assumption about
__insn_mtspr for <asm/irqflags.h>") to come about.
Note for stable: use discretion when/if applying this. As mentioned,
this bug may never have actually bitten anybody, and gcc may never have
done the required code motion for it to possibly ever trigger in
practice.
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Cc: Steven Rostedt <srostedt@redhat.com> Cc: Peter Zijlstra <peterz@infradead.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Even in failed case of pm_runtime_get_sync, the usage_count
is incremented. In order to keep the usage_count with correct
value and runtime power management to behave correctly, call
pm_runtime_put_noidle in such case.
In __hwspin_lock_request, module_put is also called before
return in pm_runtime_get_sync failed case.
Signed-off-by Liu Chuansheng <chuansheng.liu@intel.com> Signed-off-by: Li Fei <fei.li@intel.com>
[edit commit log] Signed-off-by: Ohad Ben-Cohen <ohad@wizery.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Commit fca460f95e928bae373daa8295877b6905bc62b8 simplified the x32
implementation by creating a syscall bitmask, equal to 0x40000000, that
could be applied to x32 syscalls such that the masked syscall number
would be the same as a x86_64 syscall. While that patch was a nice
way to simplify the code, it went a bit too far by adding the mask to
syscall_get_nr(); returning the masked syscall numbers can cause
confusion with callers that expect syscall numbers matching the x32
ABI, e.g. unmasked syscall numbers.
This patch fixes this by simply removing the mask from syscall_get_nr()
while preserving the other changes from the original commit. While
there are several syscall_get_nr() callers in the kernel, most simply
check that the syscall number is greater than zero, in this case this
patch will have no effect. Of those remaining callers, they appear
to be few, seccomp and ftrace, and from my testing of seccomp without
this patch the original commit definitely breaks things; the seccomp
filter does not correctly filter the syscalls due to the difference in
syscall numbers in the BPF filter and the value from syscall_get_nr().
Applying this patch restores the seccomp BPF filter functionality on
x32.
I've tested this patch with the seccomp BPF filters as well as ftrace
and everything looks reasonable to me; needless to say general usage
seemed fine as well.
Signed-off-by: Paul Moore <pmoore@redhat.com> Link: http://lkml.kernel.org/r/20130215172143.12549.10292.stgit@localhost Cc: Will Drewry <wad@chromium.org> Cc: H. Peter Anvin <hpa@zytor.com> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Some versions of pHyp will perform the adjunct partition test before the
ANDCOND test. The result of this is that H_RESOURCE can be returned and
cause the BUG_ON condition to occur. The HPTE is not removed. So add a
check for H_RESOURCE, it is ok if this HPTE is not removed as
pSeries_lpar_hpte_remove is looking for an HPTE to remove and not a
specific HPTE to remove. So it is ok to just move on to the next slot
and try again.
Signed-off-by: Michael Wolf <mjw@linux.vnet.ibm.com> Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
projects / karo-tx-linux.git / log