otherwise ECT(1) bit will get interpreted as RTO_ONLINK
and routing will fail with XfrmOutBundleGenError.
Signed-off-by: Ulrich Weber <uweber@astaro.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
When there are VLANs on a VETH device, the packets being transmitted
through the VETH device may be 4 bytes bigger than MTU. A check
in dev_forward_skb did not take this into account and so dropped
these packets.
This patch is needed at least as far back as 2.6.34.7 and should
be considered for -stable.
Signed-off-by: Ben Greear <greearb@candelatech.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
As tunnel devices are going to be lockless, we need to make sure a
misconfigured machine wont enter an infinite loop.
Add a percpu variable, and limit to three the number of stacked xmits.
Reported-by: Jesse Gross <jesse@nicira.com> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
TIPC needs to have its endianess issues fixed. Unfortunately, the format of a
subscriber message is passed in directly from user space, so requiring this
message to be in network byte order breaks user space ABI. Revert this change
until such time as we can determine how to do this in a backwards compatible
manner.
Signed-off-by: Neil Horman <nhorman@tuxdriver.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Backout the tipc changes to the flags int he subscription message. These
changees, while reasonable on the surface, interefere with user space ABI
compatibility which is a no-no. This was part of the changes to fix the
endianess issues in the TIPC protocol, which would be really nice to do but we
need to do so in a way that is backwards compatible with user space.
Signed-off-by: Neil Horman <nhorman@tuxdriver.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
This patch adds a syn_set parameter to the retransmits_timed_out()
routine and updates its callers. If not set, TCP_RTO_MIN is taken
as the calculation basis as before. If set, TCP_TIMEOUT_INIT is
used instead, so that sysctl_syn_retries represents the actual
amount of SYN retransmissions in case no SYNACKs are received when
establishing a new connection.
Signed-off-by: Damian Lukowski <damian@tvk.rwth-aachen.de> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
If a RST comes in immediately after checking sk->sk_err, tcp_poll will
return POLLIN but not POLLOUT. Fix this by checking sk->sk_err at the end
of tcp_poll. Additionally, ensure the correct order of operations on SMP
machines with memory barriers.
Signed-off-by: Tom Marshall <tdm.code@gmail.com> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
holt@sgi.com [Wed, 20 Oct 2010 02:03:37 +0000 (02:03 +0000)]
Limit sysctl_tcp_mem and sysctl_udp_mem initializers to prevent integer overflows.
On a 16TB x86_64 machine, sysctl_tcp_mem[2], sysctl_udp_mem[2], and
sysctl_sctp_mem[2] can integer overflow. Set limit such that they are
maximized without overflowing.
This patch fixes the condition (3rd arg) passed to sk_wait_event() in
sk_stream_wait_memory(). The incorrect check in sk_stream_wait_memory()
causes the following soft lockup in tcp_sendmsg() when the global tcp
memory pool has exhausted.
What is happening is, that the sk_wait_event() condition passed from
sk_stream_wait_memory() evaluates to true for the case of tcp global memory
exhaustion. This is because both sk_stream_memory_free() and vm_wait are true
which causes sk_wait_event() to *not* call schedule_timeout().
Hence sk_stream_wait_memory() returns immediately to the caller w/o sleeping.
This causes the caller to again try allocation, which again fails and again
calls sk_stream_wait_memory(), and so on.
Signed-off-by: Nagendra Singh Tomar <tomer_iisc@yahoo.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Just use explicit casts, since we really can't change the
types of structures exported to userspace which have been
around for 15 years or so.
Reported-by: Dan Rosenberg <dan.j.rosenberg@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
This code has been broken forever, but in several different and
creative ways.
So far as I can work out, the R6040 MAC filter has 4 exact-match
entries, the first of which the driver uses for its assigned unicast
address, plus a 64-entry hash-based filter for multicast addresses
(maybe unicast as well?).
The original version of this code would write the first 4 multicast
addresses as exact-match entries from offset 1 (bug #1: there is no
entry 4 so this could write to some PHY registers). It would fill the
remainder of the exact-match entries with the broadcast address (bug #2:
this would overwrite the last used entry). If more than 4 multicast
addresses were configured, it would set up the hash table, write some
random crap to the MAC control register (bug #3) and finally walk off
the end of the list when filling the exact-match entries (bug #4).
All of this seems to be pointless, since it sets the promiscuous bit
when the interface is made promiscuous or if >4 multicast addresses
are enabled, and never clears it (bug #5, masking bug #2).
The recent(ish) changes to the multicast list fixed bug #4, but
completely removed the limit on iteration over the exact-match entries
(bug #6).
Bug #4 was reported as
<https://bugzilla.kernel.org/show_bug.cgi?id=15355> and more recently
as <http://bugs.debian.org/600155>. Florian Fainelli attempted to fix
these in commit 3bcf8229a8c49769e48d3e0bd1e20d8e003f8106, but that
actually dealt with bugs #1-3, bug #4 having been fixed in mainline at
that point.
That commit fixes the most important current bug #6.
Signed-off-by: Ben Hutchings <ben@decadent.org.uk> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: Maciej Żenczykowski <maze@google.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
After running this bonding setup script
modprobe bonding miimon=100 mode=0 max_bonds=1
ifconfig bond0 10.1.1.1/16
ifenslave bond0 eth1
ifenslave bond0 eth3
on s390 with qeth-driven slaves, modprobe -r fails with this message
unregister_netdevice: waiting for bond0 to become free. Usage count = 1
due to twice detection of duplicate address.
Problem is caused by a missing decrease of ifp->refcnt in addrconf_dad_failure.
An extra call of in6_ifa_put(ifp) solves it.
Problem has been introduced with commit f2344a131bccdbfc5338e17fa71a807dee7944fa.
Signed-off-by: Ursula Braun <ursula.braun@de.ibm.com> Cc: David S. Miller <davem@davemloft.net> Cc: Herbert Xu <herbert@gondor.apana.org.au> Acked-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Special care should be taken when slow path is hit in ip_fragment() :
When walking through frags, we transfert truesize ownership from skb to
frags. Then if we hit a slow_path condition, we must undo this or risk
uncharging frags->truesize twice, and in the end, having negative socket
sk_wmem_alloc counter, or even freeing socket sooner than expected.
Many thanks to Nick Bowler, who provided a very clean bug report and
test program.
Thanks to Jarek for reviewing my first patch and providing a V2
While Nick bisection pointed to commit 2b85a34e911 (net: No more
expensive sock_hold()/sock_put() on each tx), underlying bug is older
(2.6.12-rc5)
A side effect is to extend work done in commit b2722b1c3a893e
(ip_fragment: also adjust skb->truesize for packets not owned by a
socket) to ipv6 as well.
Reported-and-bisected-by: Nick Bowler <nbowler@elliptictech.com> Tested-by: Nick Bowler <nbowler@elliptictech.com> Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com> CC: Jarek Poplawski <jarkao2@gmail.com> CC: Patrick McHardy <kaber@trash.net> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
The rx_recycle queue is global per device but can be accesed by many
napi handlers at the same time, so it needs full skb_queue primitives
(with locking). Otherwise, various crashes caused by broken skbs are
possible.
This patch resolves, at least partly, bugzilla bug 19692. (Because of
some doubts that there could be still something around which is hard
to reproduce my proposal is to leave this bug opened for a month.)
Reported-by: emin ak <eminak71@gmail.com> Tested-by: emin ak <eminak71@gmail.com> Signed-off-by: Jarek Poplawski <jarkao2@gmail.com> CC: Andy Fleming <afleming@freescale.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: Dan Carpenter <error27@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Several other ethtool functions leave heap uncleared (potentially) by
drivers. Some interfaces appear safe (eeprom, etc), in that the sizes
are well controlled. In some situations (e.g. unchecked error conditions),
the heap will remain unchanged in areas before copying back to userspace.
Note that these are less of an issue since these all require CAP_NET_ADMIN.
Cc: stable@kernel.org Signed-off-by: Kees Cook <kees.cook@canonical.com> Acked-by: Ben Hutchings <bhutchings@solarflare.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
During the delete, the SCSI device is in moved to SDEV_CANCEL. When
the FC transport class later calls scsi_target_unblock, this has no
effect, since scsi_internal_device_unblock ignores SCSI devics in this
state.
Fix by rejecting offline and cancel in the state transition.
Signed-off-by: Christof Schmitt <christof.schmitt@de.ibm.com>
[jejb: Original patch by Christof Schmitt, modified by Mike Christie] Signed-off-by: James Bottomley <James.Bottomley@suse.de> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
The function __scsi_remove_target iterates through the SCSI devices on
the host, but it drops the host_lock before calling
scsi_remove_device. When the SCSI device is deleted from another
thread, the pointer to the SCSI device in scsi_remove_device can
become invalid. Fix this by getting a reference to the SCSI device
before dropping the host_lock to keep the SCSI device alive for the
call to scsi_remove_device.
gdth_ioctl_alloc() takes the size variable as an int.
copy_from_user() takes the size variable as an unsigned long.
gen.data_len and gen.sense_len are unsigned longs.
On x86_64 longs are 64 bit and ints are 32 bit.
We could pass in a very large number and the allocation would truncate
the size to 32 bits and allocate a small buffer. Then when we do the
copy_from_user(), it would result in a memory corruption.
Signed-off-by: Dan Carpenter <error27@gmail.com> Signed-off-by: James Bottomley <James.Bottomley@suse.de> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Some cards (like mvsas) have issue troubles if non-NCQ commands are
mixed with NCQ ones. Fix this by using the libata default NCQ check
routine which waits until all NCQ commands are complete before issuing
a non-NCQ one. The impact to cards (like aic94xx) which don't need
this logic should be minimal
Signed-off-by: James Bottomley <James.Bottomley@suse.de> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Following a site power outage which re-enabled all the ports on my FC
switches, my system subsequently booted with far too many luns! I had
let it run hoping it would make multi-user. It didn't. :( It hung solid
after exhausting the last sd device, sdzzz, and attempting to create sdaaaa
and beyond. I was unable to get a dump.
Discovered using a 2.6.32.13 based system.
correct this by detecting when the last index is utilized and failing
the sd probe of the device. Patch applies to scsi-misc-2.6.
Signed-off-by: Michael Reed <mdr@sgi.com> Signed-off-by: James Bottomley <James.Bottomley@suse.de> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
A few devices (such as the RCA VR5220 voice recorder) are so
non-compliant with the USB spec that they have invalid maxpacket sizes
for endpoint 0. Nevertheless, as long as we can safely use them, we
may as well do so.
This patch (as1432) softens our acceptance criterion by allowing
high-speed devices to have ep0-maxpacket sizes other than 64. A
warning is printed in the system log when this happens, and the
existing error message is clarified.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Reported-by: James <bjlockie@lockie.ca> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
The bulk-read callback had two bugs:
a) The bulk-in packet's leading two zeros were returned (and the two last
bytes truncated)
b) The wrong URB was transmitted for the second (and later) read requests,
causing further reads to return the entire packet (including leading
zeros)
Fix regression introduced by commit 214916f2ec6701e1c9972f26c60b3dc37d3153c6 (USB: visor: reimplement using
generic framework) which broke initialisation of UX50/TH55 devices that
used re-mapped bulk-out endpoint addresses.
Reported-by: Robert Gadsdon <rgadsdon@bayarea.net> Tested-by: Robert Gadsdon <rgadsdon@bayarea.net> Signed-off-by: Johan Hovold <jhovold@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
This patch (as1430) fixes a bug in usbcore. When a device
configuration change occurs or a device is removed, the endpoints for
the old config should be completely disabled. However it turns out
they aren't; this is because usb_unbind_interface() calls
usb_enable_interface() or usb_set_interface() to put interfaces back
in altsetting 0, which re-enables the interfaces' endpoints.
As a result, when a device goes through a config change or is
unconfigured, the ep_in[] and ep_out[] arrays may be left holding old
pointers to usb_host_endpoint structures. If the device is
deauthorized these structures get freed, and the stale pointers cause
errors when the the device is eventually unplugged.
The solution is to disable the endpoints after unbinding the
interfaces instead of before. This isn't as large a change as it
sounds, since usb_unbind_interface() disables all the interface's
endpoints anyway before calling the driver's disconnect routine,
unless the driver claims to support "soft" unbind.
This fixes Bugzilla #19192. Thanks to "Tom" Lei Ming for diagnosing
the underlying cause of the problem.
Signed-off-by: Alan Stern <stern@rowland.harvard.edu> Tested-by: Carsten Sommer <carsten_sommer@ymail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
The WAGO 750-923 USB Service Cable is used for configuration and firmware
updates of several industrial automation products from WAGO Kontakttechnik GmbH.
Bus 004 Device 002: ID 1be3:07a6
Device Descriptor:
bLength 18
bDescriptorType 1
bcdUSB 1.10
bDeviceClass 0 (Defined at Interface level)
bDeviceSubClass 0
bDeviceProtocol 0
bMaxPacketSize0 64
idVendor 0x1be3
idProduct 0x07a6
bcdDevice 1.00
iManufacturer 1 Silicon Labs
iProduct 2 WAGO USB Service Cable
iSerial 3 1277796751
. . .
Signed-off-by: Anders Larsen <al@alarsen.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
The protocol code is set 00 in IAD and it's set to 01 in ACM control
interface descriptor in f_acm.c file. Due to this, windows is unable to
install the modem(ACM) driver based on class-subclass-protocol matching.
This patch corrects the protocol code in ACM IAD to the same as in
acm_control_interface_desc protocol code.
Since commit 461972d8a4c94bc44f11a13046041c78a7cf18dd (musb_core: don't call
musb_platform_exit() twice), unloading the driver module results in a WARNING
"kobject: '(null)' (c73de788): is not initialized, yet kobject_put() is being
called." (or even kernel oops) on e.g. DaVincis, though only in the OTG mode.
There exists dubious and unbalanced put_device() call in musb_free() which
takes place only in the OTG mode. As this commit caused musb_platform_exit()
to be called (and so unregister the NOP transceiver) before this put_device()
call, this function references already freed memory.
On the other hand, all the glue layers miss the otg_put_transceiver() call,
complementary to the otg_get_transceiver() call that they do. So, I think
the solution is to get rid of the strange put_device() call, and instead
call otg_put_transceiver() in the glue layers...
Signed-off-by: Sergei Shtylyov <sshtylyov@ru.mvista.com> Signed-off-by: Felipe Balbi <balbi@ti.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
This patch for FTDI USB serial driver ads new VID/PIDs used on various
devices manufactured by Papouch (http://www.papouch.com). These devices
have their own VID/PID, although they're using standard FTDI chip. In
ftdi_sio.c, I also made small cleanup to have declarations for all
Papouch devices together.
Signed-off-by: Daniel Suchy <danny@danysek.cz> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
The OpenDCC project is developing a new hardware. This patch adds its
PID to the list of known FTDI devices. The PID can be found at
http://www.opendcc.de/elektronik/usb/opendcc_usb.html
RTS and DTR should not be modified based on CRTSCTS when calling
set_termios.
Modem control lines are raised at port open by the tty layer and should stay
raised regardless of whether hardware flow control is enabled or not.
This is in conformance with the way serial ports work today and many
applications depend on this behaviour to be able to talk to hardware
implementing hardware flow control (without the applications actually using
it).
Hardware which expects different behaviour on these lines can always
use TIOCMSET/TIOCMBI[SC] after port open to change them.
Reported-by: Daniel Mack <daniel@caiaq.de> Reported-by: Dave Mielke <dave@mielke.cc> Signed-off-by: Johan Hovold <jhovold@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
This patch fixes the vendor and product ID the gadget uses
by replacing the temporary IDs that were used during
development (which should never get into mainline) with
proper IDs.
Signed-off-by: Michal Nazarewicz <m.nazarewicz@samsung.com> Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
This patch fixes the vendor and product ID the gadget uses
by replacing the temporary IDs that were used during
development (which should never get into mainline) with
proper IDs.
Signed-off-by: Michal Nazarewicz <m.nazarewicz@samsung.com> Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
The composite gadget will OOPS if the host sends a control request
targetted to an interface of an un-configured composite device. This patch
prevents this.
The OOPS was observed during WHQL USB CV tests. With this patch, the device
STALLs as per requirement.
Failing test case: From host do the following. I used libusb-1.0
1) Set configuration to zero.
libusb_control_transfer(device_handle,
0, /* standard OUT */
0x9, /* setConfiguration */
0, 0, NULL, 0, 0);
2) Query current configuratioan.
libusb_control_transfer(device_handle,
0x80, /* standard IN*/
0x8, /* getConfiguration */
0, 0, data, 1, 0);
3) Send the non-standard ctrl transfer targetted to interface
libusb_control_transfer(device_handle,
0x81, /* standard IN to interface*/
0x6, /* getDescriptor */
0x2300, 0, data, 0x12, 0);
Signed-off-by: Roger Quadros <roger.quadros@nokia.com> Cc: David Brownell <dbrownell@users.sourceforge.net> Cc: Michal Nazarewicz <m.nazarewicz@samsung.com> Cc: Robert Lukassen <robert.lukassen@tomtom.com> Cc: Kyungmin Park <kyungmin.park@samsung.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
ret = 0
... when != ret = e1
*x = \(kmalloc\|kcalloc\|kzalloc\)(...)
... when != ret = e2
if (x == NULL) { ... when != ret = e3
return ret;
}
// </smpl>
Signed-off-by: Julia Lawall <julia@diku.dk> Acked-by: Christian Lamparter <chunkeey@googlemail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Signed-off-by: Christian Lamparter <chunkeey@googlemail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
The ISL3887 chip needs a USB reset, whenever the
usb-frontend module "p54usb" is reloaded.
This patch fixes an off-by-one bug, if the user
is running a kernel without the CONFIG_PM option
set and for some reason (e.g.: compat-wireless)
wants to switch between different p54usb modules.
Signed-off-by: Christian Lamparter <chunkeey@googlemail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
This driver uses non-standard semantics for the read urb which makes the
generic close function fail to kill it (the read urb is actually an
interrupt urb and therefore bulk_in size is zero).
Reported-by: Eric Shattow "Eprecocious" <lucent@gmail.com> Tested-by: Eric Shattow "Eprecocious" <lucent@gmail.com> Signed-off-by: Johan Hovold <jhovold@gmail.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
This patch implements restoring of the correct rip, rsp, and
rax after the svm emulation in KVM injected a selective_cr0
write intercept into the guest hypervisor. The problem was
that the vmexit is emulated in the instruction emulation
which later commits the registers right after the write-cr0
instruction. So the l1 guest will continue to run with the
l2 rip, rsp and rax resulting in unpredictable behavior.
This patch is not the final word, it is just an easy patch
to fix the issue. The real fix will be done when the
instruction emulator is made aware of nested virtualization.
Until this is done this patch fixes the issue and provides
an easy way to fix this in -stable too.
This patch fixes a bug in KVM where it _always_ reports the
support of the SVM feature to userspace. But KVM only
supports SVM on AMD hardware and only when it is enabled in
the kernel module. This patch fixes the wrong reporting.
Impact: fix kernel bug such as:
BUG: scheduling while atomic: dosemu.bin/19680/0x00000004
See also Ubuntu bug 455067 at
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/455067
Commits 4915a35e35a037254550a2ba9f367a812bc37d40
("Use preempt_conditional_sti/cli in do_int3, like on x86_64.")
and 3d2a71a596bd9c761c8487a2178e95f8a61da083
("x86, traps: converge do_debug handlers")
started disabling preemption in int1 and int3 handlers on i386.
The problem with vm86 is that the call to handle_vm86_trap() may jump
straight to entry_32.S and never returns so preempt is never enabled
again, and there is an imbalance in the preempt count.
Commit be716615fe596ee117292dc615e95f707fb67fd1 ("x86, vm86:
fix preemption bug"), which was later (accidentally?) reverted by commit 08d68323d1f0c34452e614263b212ca556dae47f ("hw-breakpoints: modifying
generic debug exception to use thread-specific debug registers")
fixed the problem for debug exceptions but not for breakpoints.
There are three solutions to this problem.
1. Reenable preemption before calling handle_vm86_trap(). This
was the approach that was later reverted.
2. Do not disable preemption for i386 in breakpoint and debug handlers.
This was the situation before October 2008. As far as I understand
preemption only needs to be disabled on x86_64 because a seperate stack is
used, but it's nice to have things work the same way on
i386 and x86_64.
3. Let handle_vm86_trap() return instead of jumping to assembly code.
By setting a flag in _TIF_WORK_MASK, either TIF_IRET or TIF_NOTIFY_RESUME,
the code in entry_32.S is instructed to return to 32 bit mode from
V86 mode. The logic in entry_32.S was already present to handle signals.
(I chose TIF_IRET because it's slightly more efficient in
do_notify_resume() in signal.c, but in fact TIF_IRET can probably be
replaced by TIF_NOTIFY_RESUME everywhere.)
I'm submitting approach 3, because I believe it is the most elegant
and prevents future confusion. Still, an obvious
preempt_conditional_cli(regs); is necessary in traps.c to correct the
bug.
[ hpa: This is technically a regression, but because:
1. the regression is so old,
2. the patch seems relatively high risk, justifying more testing, and
3. we're late in the 2.6.36-rc cycle,
I'm queuing it up for the 2.6.37 merge window. It might, however,
justify as a -stable backport at a latter time, hence Cc: stable. ]
Signed-off-by: Bart Oldeman <bartoldeman@users.sourceforge.net>
LKML-Reference: <alpine.DEB.2.00.1009231312330.4732@localhost.localdomain> Cc: Frederic Weisbecker <fweisbec@gmail.com> Cc: K.Prasad <prasad@linux.vnet.ibm.com> Cc: Alan Stern <stern@rowland.harvard.edu> Cc: Alexander van Heukelum <heukelum@fastmail.fm> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
The copy of /proc/vmcore to a user buffer proceeds much faster
if the kernel addresses memory as cached.
With this patch we have seen an increase in transfer rate from
less than 15MB/s to 80-460MB/s, depending on size of the
transfer. This makes a big difference in time needed to save a
system dump.
Currently the redirection hint in the interrupt-remapping table entry
is set to 0, which means the remapped interrupt is directed to the
processors listed in the destination. So in logical flat mode
in the presence of intr-remapping, this results in a single
interrupt multi-casted to multiple cpu's as specified by the destination
bit mask. But what we really want is to send that interrupt to one of the cpus
based on the lowest priority delivery mode.
Set the redirection hint in the IRTE to '1' to indicate that we want
the remapped interrupt to be directed to only one of the processors
listed in the destination.
This fixes the issue of same interrupt getting delivered to multiple cpu's
in the logical flat mode in the presence of interrupt-remapping. While
there is no functional issue observed with this behavior, this will
impact performance of such configurations (<=8 cpu's using logical flat
mode in the presence of interrupt-remapping)
Signed-off-by: Suresh Siddha <suresh.b.siddha@intel.com>
LKML-Reference: <20100827181049.013051492@sbsiddha-MOBL3.sc.intel.com> Cc: Weidong Han <weidong.han@intel.com> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Instead of adapting the CPU family check in amd_special_default_mtrr()
for each new CPU family assume that all new AMD CPUs support the
necessary bits in SYS_CFG MSR.
Tom2Enabled is architectural (defined in APM Vol.2).
Tom2ForceMemTypeWB is defined in all BKDGs starting with K8 NPT.
In pre K8-NPT BKDG this bit is reserved (read as zero).
W/o this adaption Linux would unnecessarily complain about bad MTRR
settings on every new AMD CPU family, e.g.
[ 0.000000] WARNING: BIOS bug: CPU MTRRs don't cover all of memory, losing 4863MB of RAM.
Signed-off-by: Andreas Herrmann <andreas.herrmann3@amd.com>
LKML-Reference: <20100930123235.GB20545@loge.amd.com> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
x86 smp_ops now has a new op, stop_other_cpus which takes a parameter
"wait" this allows the caller to specify if it wants to stop until all
the cpus have processed the stop IPI. This is required specifically
for the kexec case where we should wait for all the cpus to be stopped
before starting the new kernel. We now wait for the cpus to stop in
all cases except for panic/kdump where we expect things to be broken
and we are doing our best to make things work anyway.
Signed-off-by: Alok N Kataria <akataria@vmware.com>
LKML-Reference: <1286833028.1372.20.camel@ank32.eng.vmware.com> Cc: Eric W. Biederman <ebiederm@xmission.com> Cc: Jeremy Fitzhardinge <jeremy@xensource.com> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
The AMD SSE5 feature set as-it has been replaced by some extensions
to the AVX instruction set. Thus the bit formerly advertised as SSE5
is re-used for one of these extensions (XOP).
Although this changes the /proc/cpuinfo output, it is not user visible, as
there are no CPUs (yet) having this feature.
To avoid confusion this should be added to the stable series, too.
Signed-off-by: Andre Przywara <andre.przywara@amd.com>
LKML-Reference: <1283778860-26843-2-git-send-email-andre.przywara@amd.com> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
During the reading of /proc/vmcore the kernel is doing
ioremap()/iounmap() repeatedly. And the buildup of un-flushed
vm_area_struct's is causing a great deal of overhead. (rb_next()
is chewing up most of that time).
This solution is to provide function set_iounmap_nonlazy(). It
causes a subsequent call to iounmap() to immediately purge the
vma area (with try_purge_vmap_area_lazy()).
With this patch we have seen the time for writing a 250MB
compressed dump drop from 71 seconds to 44 seconds.
ATA devices don't send D2H Reg FIS after an successful ATA PIO data-in
command. The host is supposed to take the TF and E_Status of the
preceding PIO Setup FIS. Update ahci_qc_fill_rtf() such that it takes
TF + E_Status from PIO Setup FIS after a successful ATA PIO data-in
command.
Without this patch, result_tf for such a command is filled with the
content of the previous D2H Reg FIS which belongs to a previous
command, which can make the command incorrectly seen as failed.
* Patch updated to grab the whole TF + E_Status from PIO Setup FIS
instead of just E_Status as suggested by Robert Hancock.
Signed-off-by: Tejun Heo <tj@kernel.org> Reported-by: Mark Lord <kernel@teksavvy.com> Cc: Robert Hancock <hancockrwd@gmail.com> Signed-off-by: Jeff Garzik <jgarzik@redhat.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
cifs_new_fileinfo() does not use the 'oplock' value from the callers. Instead,
it sets it to REQ_OPLOCK which seems wrong. We should be using the oplock value
obtained from the Server to set the inode's clientCanCacheAll or
clientCanCacheRead flags. Fix this by passing oplock from the callers to
cifs_new_fileinfo().
This change dates back to commit a6ce4932 (2.6.30-rc3). So, all the affected
versions will need this fix. Please Cc stable once reviewed and accepted.
Reviewed-by: Jeff Layton <jlayton@redhat.com> Signed-off-by: Suresh Jayaraman <sjayaraman@suse.de> Signed-off-by: Steve French <sfrench@us.ibm.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
futex_wait() is leaking key references due to futex_wait_setup()
acquiring an additional reference via the queue_lock() routine. The
nested key ref-counting has been masking bugs and complicating code
analysis. queue_lock() is only called with a previously ref-counted
key, so remove the additional ref-counting from the queue_(un)lock()
functions.
Also futex_wait_requeue_pi() drops one key reference too many in
unqueue_me_pi(). Remove the key reference handling from
unqueue_me_pi(). This was paired with a queue_lock() in
futex_lock_pi(), so the count remains unchanged.
Document remaining nested key ref-counting sites.
Signed-off-by: Darren Hart <dvhart@linux.intel.com> Reported-and-tested-by: Matthieu Fertré<matthieu.fertre@kerlabs.com> Reported-by: Louis Rilling<louis.rilling@kerlabs.com> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Eric Dumazet <eric.dumazet@gmail.com> Cc: John Kacur <jkacur@redhat.com> Cc: Rusty Russell <rusty@rustcorp.com.au>
LKML-Reference: <4CBB17A8.70401@linux.intel.com> Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Load weights are for the CFS, they do not belong in the RT task. This makes all
RT scheduling classes leave the CFS weights alone.
This fixes a real bug as well: I noticed the following phonomena: a process
elevated to SCHED_RR forks with SCHED_RESET_ON_FORK set, and the child is
indeed SCHED_OTHER, and the niceval is indeed reset to 0. However the weight
inserted by set_load_weight() remains at 0, giving the task insignificat
priority.
With this fix, the weight is reset to what the task had before being elevated
to SCHED_RR/SCHED_FIFO.
The logic to distinguish marked instruction events from ordinary events
on PPC970 and derivatives was flawed. The result is that instruction
sampling didn't get enabled in the PMU for some marked instruction
events, so they would never trigger. This fixes it by adding the
appropriate break statements in the switch statement.
Reported-by: David Binderman <dcb314@hotmail.com> Signed-off-by: Paul Mackerras <paulus@samba.org> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
You can only call update_context_time() when the context
is active, i.e., the thread it is attached to is still running.
However, perf_event_read() can be called even when the context
is inactive, e.g., user read() the counters. The call to
update_context_time() must be conditioned on the status of
the context, otherwise, bogus time_enabled, time_running may
be returned. Here is an example on AMD64. The task program
is an example from libpfm4. The -p prints deltas every 1s.
- Event thread wakes up and passes the event to
event_handler() to process.
- It processes and clears the USBIP_EH_SHUTDOWN
flag then returns.
- The outer event loop (event_handler_loop()) calls
wait_event_interruptible().
The processing of the second flag which is part of
VDEV_EVENT_DOWN (USBIP_EH_RESET) did not happen yet.
It is delayed until the next event.
This means the ->reset callback may not happen for
a long time (if ever), leaving the usbip port in a
weird state which prevents its reuse.
This patch changes the handler to process all flags
before waiting for another wakeup.
I have verified this change to fix a problem which
prevented reattach of a usbip device. It also helps
for socket errors which missed the RESET as well.
The delayed event processing also affects the stub
side of usbip and the error handling there.
Signed-off-by: Max Vozeler <mvz@vozeler.com> Reported-by: Marco Lancione <marco@optikam.com> Tested-by: Luc Jalbert <ljalbert@optikam.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
An execve with a very large total of argument/environment strings
can take a really long time in the execve system call. It runs
uninterruptibly to count and copy all the strings. This change
makes it abort the exec quickly if sent a SIGKILL.
Note that this is the conservative change, to interrupt only for
SIGKILL, by using fatal_signal_pending(). It would be perfectly
correct semantics to let any signal interrupt the string-copying in
execve, i.e. use signal_pending() instead of fatal_signal_pending().
We'll save that change for later, since it could have user-visible
consequences, such as having a timer set too quickly make it so that
an execve can never complete, though it always happened to work before.
This adds a preemption point during the copying of the argument and
environment strings for execve, in copy_strings(). There is already
a preemption point in the count() loop, so this doesn't add any new
points in the abstract sense.
When the total argument+environment strings are very large, the time
spent copying them can be much more than a normal user time slice.
So this change improves the interactivity of the rest of the system
when one process is doing an execve with very large arguments.
The CONFIG_STACK_GROWSDOWN variant of setup_arg_pages() does not
check the size of the argument/environment area on the stack.
When it is unworkably large, shift_arg_pages() hits its BUG_ON.
This is exploitable with a very large RLIMIT_STACK limit, to
create a crash pretty easily.
Check that the initial stack is not too large to make it possible
to map in any executable. We're not checking that the actual
executable (or intepreter, for binfmt_elf) will fit. So those
mappings might clobber part of the initial stack mapping. But
that is just userland lossage that userland made happen, not a
kernel problem.
rc2 kernel crashes when booting second cpu on this CONFIG_VMSPLIT_2G_OPT
laptop: whereas cloning from kernel to low mappings pgd range does need
to limit by both KERNEL_PGD_PTRS and KERNEL_PGD_BOUNDARY, cloning kernel
pgd range itself must not be limited by the smaller KERNEL_PGD_BOUNDARY.
Signed-off-by: Hugh Dickins <hughd@google.com>
LKML-Reference: <alpine.LSU.2.00.1008242235120.2515@sister.anvils> Signed-off-by: H. Peter Anvin <hpa@zytor.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
This patch fixes machine crashes which occur when heavily exercising the
CPU hotplug codepaths on a 32-bit kernel. These crashes are caused by
AMD Erratum 383 and result in a fatal machine check exception. Here's
the scenario:
1. On 32-bit, the swapper_pg_dir page table is used as the initial page
table for booting a secondary CPU.
2. To make this work, swapper_pg_dir needs a direct mapping of physical
memory in it (the low mappings). By adding those low, large page (2M)
mappings (PAE kernel), we create the necessary conditions for Erratum
383 to occur.
3. Other CPUs which do not participate in the off- and onlining game may
use swapper_pg_dir while the low mappings are present (when leave_mm is
called). For all steps below, the CPU referred to is a CPU that is using
swapper_pg_dir, and not the CPU which is being onlined.
4. The presence of the low mappings in swapper_pg_dir can result
in TLB entries for addresses below __PAGE_OFFSET to be established
speculatively. These TLB entries are marked global and large.
5. When the CPU with such TLB entry switches to another page table, this
TLB entry remains because it is global.
6. The process then generates an access to an address covered by the
above TLB entry but there is a permission mismatch - the TLB entry
covers a large global page not accessible to userspace.
7. Due to this permission mismatch a new 4kb, user TLB entry gets
established. Further, Erratum 383 provides for a small window of time
where both TLB entries are present. This results in an uncorrectable
machine check exception signalling a TLB multimatch which panics the
machine.
There are two ways to fix this issue:
1. Always do a global TLB flush when a new cr3 is loaded and the
old page table was swapper_pg_dir. I consider this a hack hard
to understand and with performance implications
2. Do not use swapper_pg_dir to boot secondary CPUs like 64-bit
does.
This patch implements solution 2. It introduces a trampoline_pg_dir
which has the same layout as swapper_pg_dir with low_mappings. This page
table is used as the initial page table of the booting CPU. Later in the
bringup process, it switches to swapper_pg_dir and does a global TLB
flush. This fixes the crashes in our test cases.
-v2: switch to swapper_pg_dir right after entering start_secondary() so
that we are able to access percpu data which might not be mapped in the
trampoline page table.
Signed-off-by: Joerg Roedel <joerg.roedel@amd.com>
LKML-Reference: <20100816123833.GB28147@aftab> Signed-off-by: Borislav Petkov <borislav.petkov@amd.com> Signed-off-by: H. Peter Anvin <hpa@zytor.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Sandybridge GTT has new cache control bits in PTE, which controls
graphics page cache in LLC or LLC/MLC, so we need to extend the mask
function to respect the new bits.
And set cache control to always LLC only by default on Gen6.
Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com> Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Some extra CPU features such as ARAT is needed in early boot so
that x86_init function pointers can be set up properly.
http://lkml.org/lkml/2010/5/18/519
At start_kernel() level, this patch moves init_scattered_cpuid_features()
from check_bugs() to setup_arch() -> early_cpu_init() which is earlier than
platform specific x86_init layer setup. Suggested by HPA.
Signed-off-by: Jacob Pan <jacob.jun.pan@linux.intel.com>
LKML-Reference: <1274295685-6774-2-git-send-email-jacob.jun.pan@linux.intel.com> Acked-by: Thomas Gleixner <tglx@linutronix.de> Signed-off-by: H. Peter Anvin <hpa@linux.intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Commit 2a6b69765ad794389f2fc3e14a0afa1a995221c2 (ACPI: Store NVS
state even when entering suspend to RAM) changed the ACPI suspend
to RAM code so that the NVS memory area is always unconditionally
saved during suspend and restored during resume, since some systems
evidently need that for the suspend-resume to work on them. However,
it turned out that this change broke suspend-resume on a few systems,
so commit 72ad5d77fb981963edae15eee8196c80238f5ed0 (ACPI / Sleep:
Allow the NVS saving to be skipped during suspend to RAM) introduced
the acpi_sleep=nonvs command line switch to allow their users to
work around this issue. To keep track of the systems that require
this workaround and to make the life of their users slightly easier
blacklist them in acpisleep_dmi_table[].
https://bugzilla.kernel.org/show_bug.cgi?id=16396
Signed-off-by: Rafael J. Wysocki <rjw@sisk.pl> Signed-off-by: Len Brown <len.brown@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Section 4.7.3.1.1 (PM1 Status Registers) of version 4.0 of
the ACPI spec concerning PCIEXP_WAKE_STS points out in
in the final note field in table 4-11 that if this bit is
set to 1 and the system is put into a sleeping state then
the system will not automatically wake.
This bit gets set by hardware to indicate that the system
woke up due to a PCI Express wakeup event, so clear it during
acpi_hw_clear_acpi_status() calls to enable subsequent
resumes to work.
BugLink: http://bugs.launchpad.net/bugs/613381 Signed-off-by: Colin Ian King <colin.king@canonical.com> Signed-off-by: Len Brown <len.brown@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Lenovo IdeaPad Y430 has an additional subwoofer connected at pin 0x1b,
which isn't muted when headphone is plugged in. This adds additional
support to the extra subwoofer via new ideapad model.
Signed-off-by: Herton Ronaldo Krzesinski <herton@mandriva.com.br> Signed-off-by: Takashi Iwai <tiwai@suse.de> Cc: David Henningsson <david.henningsson@canonical.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
For carrier detection to work properly when binding the driver with a cable
unplugged, netif_carrier_off() should be called after register_netdev(),
not before.
Signed-off-by: Paul Fertser <fercerpav@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Unfortunately, this stack pointer can't be used when translation is off
on PHYP as this stack pointer might be outside the RMO. This results in
the following on all non zero cpus:
cpu 0x1: Vector: 300 (Data Access) at [c00000001639fd10]
pc: 000000000001c50c
lr: 000000000000821c
sp: c00000001639ff90
msr: 8000000000001000
dar: c00000001639ffa0
dsisr: 42000000
current = 0xc000000016393540
paca = 0xc000000006e00200
pid = 0, comm = swapper
The original patch was only tested on bare metal system, so it never
caught this problem.
This changes __secondary_start so that we calculate the new stack
pointer but only start using it after we've called early_setup_secondary.
With this patch, the above problem goes away.
Signed-off-by: Michael Neuling <mikey@neuling.org> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
As early setup calls down to slb_initialize(), we must have kstack
initialised before checking "should we add a bolted SLB entry for our kstack?"
Failing to do so means stack access requires an SLB miss exception to refill
an entry dynamically, if the stack isn't accessible via SLB(0) (kernel text
& static data). It's not always allowable to take such a miss, and
intermittent crashes will result.
Primary CPUs don't have this issue; an SLB entry is not bolted for their
stack anyway (as that lives within SLB(0)). This patch therefore only
affects the init of secondaries.
Signed-off-by: Matt Evans <matt@ozlabs.org> Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
After
| commit d8191fa4a33fdc817277da4f2b7f771ff605a41c
| Author: Alex Chiang <achiang@hp.com>
| Date: Mon Feb 22 12:11:39 2010 -0700
|
| ACPI: processor: driver doesn't need to evaluate _PDC
|
| Now that the early _PDC evaluation path knows how to correctly
| evaluate _PDC on only physically present processors, there's no
| need for the processor driver to evaluate it later when it loads.
|
| To cover the hotplug case, push _PDC evaluation down into the
| hotplug paths.
only cpu with Processor Statement get processed with _PDC
If bios is using Device object instead of Processor statement.
SSDTs for Pstate/Cstate/Tstate can not be loaded dynamically.
Need to try to scan ACPI0007 in addition to Processor.
That commit is between 2.6.34-rc1 and 2.6.34-rc2, so stable tree for 2.6.34+
need this patch.
Signed-off-by: Yinghai Lu <yinghai@kernel.org> Reviewed-by: Bjorn Helgaas <bjorn.helgaas@hp.com> Signed-off-by: Len Brown <len.brown@intel.com> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
If the host is slow in reading data or doesn't read data at all,
blocking write calls not only blocked the program that called write()
but the entire guest itself.
To overcome this, let's not block till the host signals it has given
back the virtio ring element we passed it. Instead, send the buffer to
the host and return to userspace. This operation then becomes similar
to how non-blocking writes work, so let's use the existing code for this
path as well.
This code change also ensures blocking write calls do get blocked if
there's not enough room in the virtio ring as well as they don't return
-EAGAIN to userspace.
Signed-off-by: Amit Shah <amit.shah@redhat.com> Acked-by: Hans de Goede <hdegoede@redhat.com> Signed-off-by: Rusty Russell <rusty@rustcorp.com.au> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
- Don't return failure if the user asked for page 0
- The end of buffer check failed to account for the page header size
and consequently didn't work
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com> Signed-off-by: James Bottomley <James.Bottomley@suse.de> Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
projects / karo-tx-linux.git / log