commit 474c90156c upstream.
gcc-7 has an "optimization" pass that completely screws up, and
generates the code expansion for the (impossible) case of calling
ilog2() with a zero constant, even when the code gcc compiles does not
actually have a zero constant.
And we try to generate a compile-time error for anybody doing ilog2() on
a constant where that doesn't make sense (be it zero or negative). So
now gcc7 will fail the build due to our sanity checking, because it
created that constant-zero case that didn't actually exist in the source
code.
There's a whole long discussion on the kernel mailing about how to work
around this gcc bug. The gcc people themselevs have discussed their
"feature" in
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=72785
but it's all water under the bridge, because while it looked at one
point like it would be solved by the time gcc7 was released, that was
not to be.
So now we have to deal with this compiler braindamage.
And the only simple approach seems to be to just delete the code that
tries to warn about bad uses of ilog2().
So now "ilog2()" will just return 0 not just for the value 1, but for
any non-positive value too.
It's not like I can recall anybody having ever actually tried to use
this function on any invalid value, but maybe the sanity check just
meant that such code never made it out in public.
Reported-by: Laura Abbott <labbott@redhat.com>
Cc: John Stultz <john.stultz@linaro.org>,
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
[bwh: Backported to 3.16: There's only one log2.h file]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit 82612de1c9 upstream.
After commit f6cc9c054e, the following conf is broken (note that the
default loopback mtu is 65536, ie IP_MAX_MTU + 1):
$ ip tunnel add gre1 mode gre local 10.125.0.1 remote 10.125.0.2 dev lo
add tunnel "gre0" failed: Invalid argument
$ ip l a type dummy
$ ip l s dummy1 up
$ ip l s dummy1 mtu 65535
$ ip tunnel add gre1 mode gre local 10.125.0.1 remote 10.125.0.2 dev dummy1
add tunnel "gre0" failed: Invalid argument
dev_set_mtu() doesn't allow to set a mtu which is too large.
First, let's cap the mtu returned by ip_tunnel_bind_dev(). Second, remove
the magic value 0xFFF8 and use IP_MAX_MTU instead.
0xFFF8 seems to be there for ages, I don't know why this value was used.
With a recent kernel, it's also possible to set a mtu > IP_MAX_MTU:
$ ip l s dummy1 mtu 66000
After that patch, it's also possible to bind an ip tunnel on that kind of
interface.
CC: Petr Machata <petrm@mellanox.com>
CC: Ido Schimmel <idosch@mellanox.com>
Link: https://git.kernel.org/pub/scm/linux/kernel/git/davem/netdev-vger-cvs.git/commit/?id=e5afd356a411a
Fixes: f6cc9c054e ("ip_tunnel: Emit events for post-register MTU changes")
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Reviewed-by: Ido Schimmel <idosch@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
[bwh: Backported to 3.16:
- Drop change in ip_tunnel_create()
- Adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit 8005b09d99 upstream.
The current error handling code has an issue where it does:
if (priv->txchan)
cpdma_chan_destroy(priv->txchan);
The problem is that ->txchan is either valid or an error pointer (which
would lead to an Oops). I've changed it to use multiple error labels so
that the test can be removed.
Also there were some missing calls to netif_napi_del().
Fixes: 3ef0fdb234 ("net: davinci_emac: switch to new cpdma layer")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit b5133e7a98 upstream.
Commit 3ba9738134 ("net: ethernet: davinci_emac: add pm_runtime support")
added support for runtime PM, but it causes issues on omap3 related devices
that actually gate the clocks:
Unhandled fault: external abort on non-linefetch (0x1008)
...
[<c04160f0>] (emac_dev_getnetstats) from [<c04d6a3c>] (dev_get_stats+0x78/0xc8)
[<c04d6a3c>] (dev_get_stats) from [<c04e9ccc>] (rtnl_fill_ifinfo+0x3b8/0x938)
[<c04e9ccc>] (rtnl_fill_ifinfo) from [<c04eade4>] (rtmsg_ifinfo+0x68/0xd8)
[<c04eade4>] (rtmsg_ifinfo) from [<c04dd35c>] (register_netdevice+0x3a0/0x4ec)
[<c04dd35c>] (register_netdevice) from [<c04dd4bc>] (register_netdev+0x14/0x24)
[<c04dd4bc>] (register_netdev) from [<c041755c>] (davinci_emac_probe+0x408/0x5c8)
[<c041755c>] (davinci_emac_probe) from [<c0396d78>] (platform_drv_probe+0x48/0xa4)
Let's fix it by moving the pm_runtime_get() call earlier, and also add it to
the emac_dev_getnetstats(). Also note that we want to use pm_runtime_get_sync()
as we don't want to have deferred_resume happen. And let's also check the
return value for pm_runtime_get_sync() as noted by Felipe Balbi <balbi@ti.com>.
Cc: Brian Hutchinson <b.hutchman@gmail.com>
Acked-by: Mark A. Greer <mgreer@animalcreek.com>
Reviewed-by: Felipe Balbi <balbi@ti.com>
Signed-off-by: Tony Lindgren <tony@atomide.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit 8a83c5d796 upstream.
It's not correct to return NULL when that is actually an error and
function returns errors in any other wrong case. In the same time,
the cpsw driver and davinci emac doesn't check error case while
creating channel and it can miss actual error. Also remove WARNs
replacing them on dev_err msgs.
Signed-off-by: Ivan Khoronzhuk <ivan.khoronzhuk@linaro.org>
Reviewed-by: Grygorii Strashko <grygorii.strashko@ti.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
[bwh: Backported to 3.16:
- Channel pointers are stored in different fields
- Adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit efe3de79e0 upstream.
Call trace:
[<ffffff9203a8d7a8>] dump_backtrace+0x0/0x428
[<ffffff9203a8dbf8>] show_stack+0x28/0x38
[<ffffff920409bfb8>] dump_stack+0xd4/0x124
[<ffffff9203d187e8>] print_address_description+0x68/0x258
[<ffffff9203d18c00>] kasan_report.part.2+0x228/0x2f0
[<ffffff9203d1927c>] kasan_report+0x5c/0x70
[<ffffff9203d1776c>] check_memory_region+0x12c/0x1c0
[<ffffff9203d17cdc>] memcpy+0x34/0x68
[<ffffff9203d75348>] xattr_getsecurity+0xe0/0x160
[<ffffff9203d75490>] vfs_getxattr+0xc8/0x120
[<ffffff9203d75d68>] getxattr+0x100/0x2c8
[<ffffff9203d76fb4>] SyS_fgetxattr+0x64/0xa0
[<ffffff9203a83f70>] el0_svc_naked+0x24/0x28
If user get root access and calls security.selinux setxattr() with an
embedded NUL on a file and then if some process performs a getxattr()
on that file with a length greater than the actual length of the string,
it would result in a panic.
To fix this, add the actual length of the string to the security context
instead of the length passed by the userspace process.
Signed-off-by: Sachin Grover <sgrover@codeaurora.org>
Signed-off-by: Paul Moore <paul@paul-moore.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit 86b389ff22 upstream.
If a instance has an event trigger enabled when it is freed, it could cause
an access of free memory. Here's the case that crashes:
# cd /sys/kernel/tracing
# mkdir instances/foo
# echo snapshot > instances/foo/events/initcall/initcall_start/trigger
# rmdir instances/foo
Would produce:
general protection fault: 0000 [#1] PREEMPT SMP PTI
Modules linked in: tun bridge ...
CPU: 5 PID: 6203 Comm: rmdir Tainted: G W 4.17.0-rc4-test+ #933
Hardware name: Hewlett-Packard HP Compaq Pro 6300 SFF/339A, BIOS K01 v03.03 07/14/2016
RIP: 0010:clear_event_triggers+0x3b/0x70
RSP: 0018:ffffc90003783de0 EFLAGS: 00010286
RAX: 0000000000000000 RBX: 6b6b6b6b6b6b6b2b RCX: 0000000000000000
RDX: 0000000000000000 RSI: 0000000000000000 RDI: ffff8800c7130ba0
RBP: ffffc90003783e00 R08: ffff8801131993f8 R09: 0000000100230016
R10: ffffc90003783d80 R11: 0000000000000000 R12: ffff8800c7130ba0
R13: ffff8800c7130bd8 R14: ffff8800cc093768 R15: 00000000ffffff9c
FS: 00007f6f4aa86700(0000) GS:ffff88011eb40000(0000) knlGS:0000000000000000
CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 00007f6f4a5aed60 CR3: 00000000cd552001 CR4: 00000000001606e0
Call Trace:
event_trace_del_tracer+0x2a/0xc5
instance_rmdir+0x15c/0x200
tracefs_syscall_rmdir+0x52/0x90
vfs_rmdir+0xdb/0x160
do_rmdir+0x16d/0x1c0
__x64_sys_rmdir+0x17/0x20
do_syscall_64+0x55/0x1a0
entry_SYSCALL_64_after_hwframe+0x49/0xbe
This was due to the call the clears out the triggers when an instance is
being deleted not removing the trigger from the link list.
Fixes: 85f2b08268 ("tracing: Add basic event trigger framework")
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit 23d6aef74d upstream.
`resource' can be controlled by user-space, hence leading to a potential
exploitation of the Spectre variant 1 vulnerability.
This issue was detected with the help of Smatch:
kernel/sys.c:1474 __do_compat_sys_old_getrlimit() warn: potential spectre issue 'get_current()->signal->rlim' (local cap)
kernel/sys.c:1455 __do_sys_old_getrlimit() warn: potential spectre issue 'get_current()->signal->rlim' (local cap)
Fix this by sanitizing *resource* before using it to index
current->signal->rlim
Notice that given that speculation windows are large, the policy is to
kill the speculation on the first load and not worry if it can be
completed with a dependent load/store [1].
[1] https://marc.info/?l=linux-kernel&m=152449131114778&w=2
Link: http://lkml.kernel.org/r/20180515030038.GA11822@embeddedor.com
Signed-off-by: Gustavo A. R. Silva <gustavo@embeddedor.com>
Reviewed-by: Andrew Morton <akpm@linux-foundation.org>
Cc: Alexei Starovoitov <ast@kernel.org>
Cc: Dan Williams <dan.j.williams@intel.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
[bwh: Backported to 3.16:
- Drop changes to compat implementation, which is a wrapper for the
regular implementation here
- Adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit 8f89c007b6 upstream.
shmat()'s SHM_REMAP option forbids passing a nil address for; this is in
fact the very first thing we check for. Andrea reported that for
SHM_RND|SHM_REMAP cases we can end up bypassing the initial addr check,
but we need to check again if the address was rounded down to nil. As
of this patch, such cases will return -EINVAL.
Link: http://lkml.kernel.org/r/20180503204934.kk63josdu6u53fbd@linux-n805
Signed-off-by: Davidlohr Bueso <dbueso@suse.de>
Reported-by: Andrea Arcangeli <aarcange@redhat.com>
Cc: Joe Lawrence <joe.lawrence@redhat.com>
Cc: Manfred Spraul <manfred@colorfullife.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit a73ab244f0 upstream.
Patch series "ipc/shm: shmat() fixes around nil-page".
These patches fix two issues reported[1] a while back by Joe and Andrea
around how shmat(2) behaves with nil-page.
The first reverts a commit that it was incorrectly thought that mapping
nil-page (address=0) was a no no with MAP_FIXED. This is not the case,
with the exception of SHM_REMAP; which is address in the second patch.
I chose two patches because it is easier to backport and it explicitly
reverts bogus behaviour. Both patches ought to be in -stable and ltp
testcases need updated (the added testcase around the cve can be
modified to just test for SHM_RND|SHM_REMAP).
[1] lkml.kernel.org/r/20180430172152.nfa564pvgpk3ut7p@linux-n805
This patch (of 2):
Commit 95e91b831f ("ipc/shm: Fix shmat mmap nil-page protection")
worked on the idea that we should not be mapping as root addr=0 and
MAP_FIXED. However, it was reported that this scenario is in fact
valid, thus making the patch both bogus and breaks userspace as well.
For example X11's libint10.so relies on shmat(1, SHM_RND) for lowmem
initialization[1].
[1] https://cgit.freedesktop.org/xorg/xserver/tree/hw/xfree86/os-support/linux/int10/linux.c#n347
Link: http://lkml.kernel.org/r/20180503203243.15045-2-dave@stgolabs.net
Fixes: 95e91b831f ("ipc/shm: Fix shmat mmap nil-page protection")
Signed-off-by: Davidlohr Bueso <dbueso@suse.de>
Reported-by: Joe Lawrence <joe.lawrence@redhat.com>
Reported-by: Andrea Arcangeli <aarcange@redhat.com>
Cc: Manfred Spraul <manfred@colorfullife.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit 322eaa06d5 upstream.
In commit 624dbf55a3 ("driver/net: enic: Try DMA 64 first, then
failover to DMA") DMA mask was changed from 40 bits to 64 bits.
Hardware actually supports only 47 bits.
Fixes: 624dbf55a3 ("driver/net: enic: Try DMA 64 first, then failover to DMA")
Signed-off-by: Govindarajulu Varadarajan <gvaradar@cisco.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit af8d3c7c00 upstream.
The PPPIOCDETACH ioctl effectively tries to "close" the given ppp file
before f_count has reached 0, which is fundamentally a bad idea. It
does check 'f_count < 2', which excludes concurrent operations on the
file since they would only be possible with a shared fd table, in which
case each fdget() would take a file reference. However, it fails to
account for the fact that even with 'f_count == 1' the file can still be
linked into epoll instances. As reported by syzbot, this can trivially
be used to cause a use-after-free.
Yet, the only known user of PPPIOCDETACH is pppd versions older than
ppp-2.4.2, which was released almost 15 years ago (November 2003).
Also, PPPIOCDETACH apparently stopped working reliably at around the
same time, when the f_count check was added to the kernel, e.g. see
https://lkml.org/lkml/2002/12/31/83. Also, the current 'f_count < 2'
check makes PPPIOCDETACH only work in single-threaded applications; it
always fails if called from a multithreaded application.
All pppd versions released in the last 15 years just close() the file
descriptor instead.
Therefore, instead of hacking around this bug by exporting epoll
internals to modules, and probably missing other related bugs, just
remove the PPPIOCDETACH ioctl and see if anyone actually notices. Leave
a stub in place that prints a one-time warning and returns EINVAL.
Reported-by: syzbot+16363c99d4134717c05b@syzkaller.appspotmail.com
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Eric Biggers <ebiggers@google.com>
Acked-by: Paul Mackerras <paulus@ozlabs.org>
Reviewed-by: Guillaume Nault <g.nault@alphalink.fr>
Tested-by: Guillaume Nault <g.nault@alphalink.fr>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit 0171c41835 upstream.
ppp_dev_uninit(), which is the .ndo_uninit() handler of PPP devices,
needs to lock pn->all_ppp_mutex. Therefore we mustn't call
register_netdevice() with pn->all_ppp_mutex already locked, or we'd
deadlock in case register_netdevice() fails and calls .ndo_uninit().
Fortunately, we can unlock pn->all_ppp_mutex before calling
register_netdevice(). This lock protects pn->units_idr, which isn't
used in the device registration process.
However, keeping pn->all_ppp_mutex locked during device registration
did ensure that no device in transient state would be published in
pn->units_idr. In practice, unlocking it before calling
register_netdevice() doesn't change this property: ppp_unit_register()
is called with 'ppp_mutex' locked and all searches done in
pn->units_idr hold this lock too.
Fixes: 8cb775bc0a ("ppp: fix device unregistration upon netns deletion")
Reported-and-tested-by: syzbot+367889b9c9e279219175@syzkaller.appspotmail.com
Signed-off-by: Guillaume Nault <g.nault@alphalink.fr>
Signed-off-by: David S. Miller <davem@davemloft.net>
[bwh: Backported to 3.16: adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
register_netdevice() will call the device's ndo_uninit operation if
registration fails after it calls the ndo_init operation. However
ppp_dev_uninit() uses ppp->ppp_net which is currently not set until
after register_netdevice() returns.
This was fixed upstream as part of commit 96d934c70d "ppp: add
rtnetlink device creation support".
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit 6151b8b37b upstream.
ppp_release() tries to ensure that netdevices are unregistered before
decrementing the unit refcount and running ppp_destroy_interface().
This is all fine as long as the the device is unregistered by
ppp_release(): the unregister_netdevice() call, followed by
rtnl_unlock(), guarantee that the unregistration process completes
before rtnl_unlock() returns.
However, the device may be unregistered by other means (like
ppp_nl_dellink()). If this happens right before ppp_release() calling
rtnl_lock(), then ppp_release() has to wait for the concurrent
unregistration code to release the lock.
But rtnl_unlock() releases the lock before completing the device
unregistration process. This allows ppp_release() to proceed and
eventually call ppp_destroy_interface() before the unregistration
process completes. Calling free_netdev() on this partially unregistered
device will BUG():
------------[ cut here ]------------
kernel BUG at net/core/dev.c:8141!
invalid opcode: 0000 [#1] SMP
CPU: 1 PID: 1557 Comm: pppd Not tainted 4.14.0-rc2+ #4
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.10.2-1.fc26 04/01/2014
Call Trace:
ppp_destroy_interface+0xd8/0xe0 [ppp_generic]
ppp_disconnect_channel+0xda/0x110 [ppp_generic]
ppp_unregister_channel+0x5e/0x110 [ppp_generic]
pppox_unbind_sock+0x23/0x30 [pppox]
pppoe_connect+0x130/0x440 [pppoe]
SYSC_connect+0x98/0x110
? do_fcntl+0x2c0/0x5d0
SyS_connect+0xe/0x10
entry_SYSCALL_64_fastpath+0x1a/0xa5
RIP: free_netdev+0x107/0x110 RSP: ffffc28a40573d88
---[ end trace ed294ff0cc40eeff ]---
We could set the ->needs_free_netdev flag on PPP devices and move the
ppp_destroy_interface() logic in the ->priv_destructor() callback. But
that'd be quite intrusive as we'd first need to unlink from the other
channels and units that depend on the device (the ones that used the
PPPIOCCONNECT and PPPIOCATTACH ioctls).
Instead, we can just let the netdevice hold a reference on its
ppp_file. This reference is dropped in ->priv_destructor(), at the very
end of the unregistration process, so that neither ppp_release() nor
ppp_disconnect_channel() can call ppp_destroy_interface() in the interim.
Reported-by: Beniamino Galvani <bgalvani@redhat.com>
Fixes: 8cb775bc0a ("ppp: fix device unregistration upon netns deletion")
Signed-off-by: Guillaume Nault <g.nault@alphalink.fr>
Signed-off-by: David S. Miller <davem@davemloft.net>
[bwh: Backported to 3.16: Set net_device::destructor instead of
priv_destructor, and call ppp_dev_priv_destructor() if register_netdevice()
fails after call ppp_dev_init().]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit 8cb775bc0a upstream.
PPP devices may get automatically unregistered when their network
namespace is getting removed. This happens if the ppp control plane
daemon (e.g. pppd) exits while it is the last user of this namespace.
This leads to several races:
* ppp_exit_net() may destroy the per namespace idr (pn->units_idr)
before all file descriptors were released. Successive ppp_release()
calls may then cleanup PPP devices with ppp_shutdown_interface() and
try to use the already destroyed idr.
* Automatic device unregistration may also happen before the
ppp_release() call for that device gets executed. Once called on
the file owning the device, ppp_release() will then clean it up and
try to unregister it a second time.
To fix these issues, operations defined in ppp_shutdown_interface() are
moved to the PPP device's ndo_uninit() callback. This allows PPP
devices to be properly cleaned up by unregister_netdev() and friends.
So checking for ppp->owner is now an accurate test to decide if a PPP
device should be unregistered.
Setting ppp->owner is done in ppp_create_interface(), before device
registration, in order to avoid unprotected modification of this field.
Finally, ppp_exit_net() now starts by unregistering all remaining PPP
devices to ensure that none will get unregistered after the call to
idr_destroy().
Signed-off-by: Guillaume Nault <g.nault@alphalink.fr>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit c4d2188206 upstream.
The CPUID bits of OSXSAVE (function=0x1) and OSPKE (func=0x7, leaf=0x0)
allows user apps to detect if OS has set CR4.OSXSAVE or CR4.PKE. KVM is
supposed to update these CPUID bits when CR4 is updated. Current KVM
code doesn't handle some special cases when updates come from emulator.
Here is one example:
Step 1: guest boots
Step 2: guest OS enables XSAVE ==> CR4.OSXSAVE=1 and CPUID.OSXSAVE=1
Step 3: guest hot reboot ==> QEMU reset CR4 to 0, but CPUID.OSXAVE==1
Step 4: guest os checks CPUID.OSXAVE, detects 1, then executes xgetbv
Step 4 above will cause an #UD and guest crash because guest OS hasn't
turned on OSXAVE yet. This patch solves the problem by comparing the the
old_cr4 with cr4. If the related bits have been changed,
kvm_update_cpuid() needs to be called.
Signed-off-by: Wei Huang <wei@redhat.com>
Reviewed-by: Bandan Das <bsd@redhat.com>
Signed-off-by: Radim Krčmář <rkrcmar@redhat.com>
[bwh: Backported to 3.16: PKE is not supported]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit 4544e403eb upstream.
This one should be using the default LPM policy for mobile chipsets so
add the PCI ID to the driver list of supported revices.
Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
[bwh: Backported to 3.16: Use board_ahci as we don't have board_ahci_mobile]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit c7e814628d upstream.
Use 64-bit accesses for 64-bit floating-point general registers with
PTRACE_PEEKUSR, removing the truncation of their upper halves in the
FR=1 mode, caused by commit bbd426f542 ("MIPS: Simplify FP context
access"), which inadvertently switched them to using 32-bit accesses.
The PTRACE_POKEUSR side is fine as it's never been broken and continues
using 64-bit accesses.
Fixes: bbd426f542 ("MIPS: Simplify FP context access")
Signed-off-by: Maciej W. Rozycki <macro@mips.com>
Cc: Ralf Baechle <ralf@linux-mips.org>
Cc: linux-mips@linux-mips.org
Patchwork: https://patchwork.linux-mips.org/patch/19334/
Signed-off-by: James Hogan <jhogan@kernel.org>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit 4faa99965e upstream.
If io_destroy() gets to cancelling everything that can be cancelled and
gets to kiocb_cancel() calling the function driver has left in ->ki_cancel,
it becomes vulnerable to a race with IO completion. At that point req
is already taken off the list and aio_complete() does *NOT* spin until
we (in free_ioctx_users()) releases ->ctx_lock. As the result, it proceeds
to kiocb_free(), freing req just it gets passed to ->ki_cancel().
Fix is simple - remove from the list after the call of kiocb_cancel(). All
instances of ->ki_cancel() already have to cope with the being called with
iocb still on list - that's what happens in io_cancel(2).
Fixes: 0460fef2a9 "aio: use cancellation list lazily"
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
[bwh: Backported to 3.16: adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit d546b67cda upstream.
spin_lock/unlock was used instead of spin_un/lock_irq
in a procedure used in process space, on a spinlock
which can be grabbed in an interrupt.
This caused the stack trace below to be displayed (on kernel
4.17.0-rc1 compiled with Lock Debugging enabled):
[ 154.661474] WARNING: SOFTIRQ-safe -> SOFTIRQ-unsafe lock order detected
[ 154.668909] 4.17.0-rc1-rdma_rc_mlx+ #3 Tainted: G I
[ 154.675856] -----------------------------------------------------
[ 154.682706] modprobe/10159 [HC0[0]:SC0[0]:HE0:SE1] is trying to acquire:
[ 154.690254] 00000000f3b0e495 (&(&qp_table->lock)->rlock){+.+.}, at: mlx4_qp_remove+0x20/0x50 [mlx4_core]
[ 154.700927]
and this task is already holding:
[ 154.707461] 0000000094373b5d (&(&cq->lock)->rlock/1){....}, at: destroy_qp_common+0x111/0x560 [mlx4_ib]
[ 154.718028] which would create a new lock dependency:
[ 154.723705] (&(&cq->lock)->rlock/1){....} -> (&(&qp_table->lock)->rlock){+.+.}
[ 154.731922]
but this new dependency connects a SOFTIRQ-irq-safe lock:
[ 154.740798] (&(&cq->lock)->rlock){..-.}
[ 154.740800]
... which became SOFTIRQ-irq-safe at:
[ 154.752163] _raw_spin_lock_irqsave+0x3e/0x50
[ 154.757163] mlx4_ib_poll_cq+0x36/0x900 [mlx4_ib]
[ 154.762554] ipoib_tx_poll+0x4a/0xf0 [ib_ipoib]
...
to a SOFTIRQ-irq-unsafe lock:
[ 154.815603] (&(&qp_table->lock)->rlock){+.+.}
[ 154.815604]
... which became SOFTIRQ-irq-unsafe at:
[ 154.827718] ...
[ 154.827720] _raw_spin_lock+0x35/0x50
[ 154.833912] mlx4_qp_lookup+0x1e/0x50 [mlx4_core]
[ 154.839302] mlx4_flow_attach+0x3f/0x3d0 [mlx4_core]
Since mlx4_qp_lookup() is called only in process space, we can
simply replace the spin_un/lock calls with spin_un/lock_irq calls.
Fixes: 6dc06c08be ("net/mlx4: Fix the check in attaching steering rules")
Signed-off-by: Jack Morgenstein <jackm@dev.mellanox.co.il>
Signed-off-by: Tariq Toukan <tariqt@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit baf10564fb upstream.
kill_ioctx() used to have an explicit RCU delay between removing the
reference from ->ioctx_table and percpu_ref_kill() dropping the refcount.
At some point that delay had been removed, on the theory that
percpu_ref_kill() itself contained an RCU delay. Unfortunately, that was
the wrong kind of RCU delay and it didn't care about rcu_read_lock() used
by lookup_ioctx(). As the result, we could get ctx freed right under
lookup_ioctx(). Tejun has fixed that in a6d7cff472 ("fs/aio: Add explicit
RCU grace period when freeing kioctx"); however, that fix is not enough.
Suppose io_destroy() from one thread races with e.g. io_setup() from another;
CPU1 removes the reference from current->mm->ioctx_table[...] just as CPU2
has picked it (under rcu_read_lock()). Then CPU1 proceeds to drop the
refcount, getting it to 0 and triggering a call of free_ioctx_users(),
which proceeds to drop the secondary refcount and once that reaches zero
calls free_ioctx_reqs(). That does
INIT_RCU_WORK(&ctx->free_rwork, free_ioctx);
queue_rcu_work(system_wq, &ctx->free_rwork);
and schedules freeing the whole thing after RCU delay.
In the meanwhile CPU2 has gotten around to percpu_ref_get(), bumping the
refcount from 0 to 1 and returned the reference to io_setup().
Tejun's fix (that queue_rcu_work() in there) guarantees that ctx won't get
freed until after percpu_ref_get(). Sure, we'd increment the counter before
ctx can be freed. Now we are out of rcu_read_lock() and there's nothing to
stop freeing of the whole thing. Unfortunately, CPU2 assumes that since it
has grabbed the reference, ctx is *NOT* going away until it gets around to
dropping that reference.
The fix is obvious - use percpu_ref_tryget_live() and treat failure as miss.
It's not costlier than what we currently do in normal case, it's safe to
call since freeing *is* delayed and it closes the race window - either
lookup_ioctx() comes before percpu_ref_kill() (in which case ctx->users
won't reach 0 until the caller of lookup_ioctx() drops it) or lookup_ioctx()
fails, ctx->users is unaffected and caller of lookup_ioctx() doesn't see
the object in question at all.
Fixes: a6d7cff472 "fs/aio: Add explicit RCU grace period when freeing kioctx"
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit 5aa1437d2d upstream.
open file, unlink it, then use ioctl(2) to make it immutable or
append only. Now close it and watch the blocks *not* freed...
Immutable/append-only checks belong in ->setattr().
Note: the bug is old and backport to anything prior to 737f2e93b9
("ext2: convert to use the new truncate convention") will need
these checks lifted into ext2_setattr().
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
[bwh: Backported to 3.16: adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit 30da870ce4 upstream.
we unlock the directory hash too early - if we are looking at secondary
link and primary (in another directory) gets removed just as we unlock,
we could have the old primary moved in place of the secondary, leaving
us to look into freed entry (and leaving our dentry with ->d_fsdata
pointing to a freed entry).
Acked-by: David Sterba <dsterba@suse.com>
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit 136d769e0b upstream.
While whitelisting Micron M500DC drives, the tweaked blacklist entry
enabled queued TRIM from M500IT variants also. But these do not support
queued TRIM. And while using those SSDs with the latest kernel we have
seen errors and even the partition table getting corrupted.
Some part from the dmesg:
[ 6.727384] ata1.00: ATA-9: Micron_M500IT_MTFDDAK060MBD, MU01, max UDMA/133
[ 6.727390] ata1.00: 117231408 sectors, multi 16: LBA48 NCQ (depth 31/32), AA
[ 6.741026] ata1.00: supports DRM functions and may not be fully accessible
[ 6.759887] ata1.00: configured for UDMA/133
[ 6.762256] scsi 0:0:0:0: Direct-Access ATA Micron_M500IT_MT MU01 PQ: 0 ANSI: 5
and then for the error:
[ 120.860334] ata1.00: exception Emask 0x1 SAct 0x7ffc0007 SErr 0x0 action 0x6 frozen
[ 120.860338] ata1.00: irq_stat 0x40000008
[ 120.860342] ata1.00: failed command: SEND FPDMA QUEUED
[ 120.860351] ata1.00: cmd 64/01:00:00:00:00/00:00:00:00:00/a0 tag 0 ncq dma 512 out
res 40/00:00:00:00:00/00:00:00:00:00/00 Emask 0x5 (timeout)
[ 120.860353] ata1.00: status: { DRDY }
[ 120.860543] ata1: hard resetting link
[ 121.166128] ata1: SATA link up 3.0 Gbps (SStatus 123 SControl 300)
[ 121.166376] ata1.00: supports DRM functions and may not be fully accessible
[ 121.186238] ata1.00: supports DRM functions and may not be fully accessible
[ 121.204445] ata1.00: configured for UDMA/133
[ 121.204454] ata1.00: device reported invalid CHS sector 0
[ 121.204541] sd 0:0:0:0: [sda] tag#18 UNKNOWN(0x2003) Result: hostbyte=0x00 driverbyte=0x08
[ 121.204546] sd 0:0:0:0: [sda] tag#18 Sense Key : 0x5 [current]
[ 121.204550] sd 0:0:0:0: [sda] tag#18 ASC=0x21 ASCQ=0x4
[ 121.204555] sd 0:0:0:0: [sda] tag#18 CDB: opcode=0x93 93 08 00 00 00 00 00 04 28 80 00 00 00 30 00 00
[ 121.204559] print_req_error: I/O error, dev sda, sector 272512
After few reboots with these errors, and the SSD is corrupted.
After blacklisting it, the errors are not seen and the SSD does not get
corrupted any more.
Fixes: 243918be63 ("libata: Do not blacklist Micron M500DC")
Cc: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Sudip Mukherjee <sudipm.mukherjee@gmail.com>
Signed-off-by: Tejun Heo <tj@kernel.org>
[bwh: Backported to 3.16: Drop ATA_HORKAGE_ZERO_AFTER_TRIM flag]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit 0d73c3f8e7 upstream.
Since do_undefinstr() uses get_user to get the undefined
instruction, it can be called before kprobes processes
recursive check. This can cause an infinit recursive
exception.
Prohibit probing on get_user functions.
Fixes: 24ba613c9d ("ARM kprobes: core code")
Signed-off-by: Masami Hiramatsu <mhiramat@kernel.org>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
[bwh: Backported to 3.16: Drop changes to __get_user_{8,32_t_8,64t_{1,2,4}}]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit eb0146daef upstream.
Prohibit kprobes on do_undefinstr because kprobes on
arm is implemented by undefined instruction. This means
if we probe do_undefinstr(), it can cause infinit
recursive exception.
Fixes: 24ba613c9d ("ARM kprobes: core code")
Signed-off-by: Masami Hiramatsu <mhiramat@kernel.org>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit e07e3c33b9 upstream.
In commit 639da5ee37 ("ARM: add an extra temp register to the low
level debugging addruart macro") an additional temporary register was
added to the addruart macro, but the decompressor code wasn't updated.
Fixes: 639da5ee37 ("ARM: add an extra temp register to the low level debugging addruart macro")
Signed-off-by: Łukasz Stelmach <l.stelmach@samsung.com>
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit 4855c92dbb upstream.
When run raidconfig from Dom0 we found that the Xen DMA heap is reduced,
but Dom Heap is increased by the same size. Tracing raidconfig we found
that the related ioctl() in megaraid_sas will call dma_alloc_coherent()
to apply memory. If the memory allocated by Dom0 is not in the DMA area,
it will exchange memory with Xen to meet the requiment. Later drivers
call dma_free_coherent() to free the memory, on xen_swiotlb_free_coherent()
the check condition (dev_addr + size - 1 <= dma_mask) is always false,
it prevents calling xen_destroy_contiguous_region() to return the memory
to the Xen DMA heap.
This issue introduced by commit 6810df88dc "xen-swiotlb: When doing
coherent alloc/dealloc check before swizzling the MFNs.".
Signed-off-by: Joe Jin <joe.jin@oracle.com>
Tested-by: John Sobecki <john.sobecki@oracle.com>
Reviewed-by: Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit 3ae1809725 upstream.
Commit f65e0d2998 ("ALSA: timer: Call notifier in the same spinlock")
combined the start/continue and stop/pause functions, and in doing so
changed the event code for the pause case to SNDRV_TIMER_EVENT_CONTINUE.
Change it back to SNDRV_TIMER_EVENT_PAUSE.
Fixes: f65e0d2998 ("ALSA: timer: Call notifier in the same spinlock")
Signed-off-by: Ben Hutchings <ben.hutchings@codethink.co.uk>
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit f65e0d2998 upstream.
snd_timer_notify1() is called outside the spinlock and it retakes the
lock after the unlock. This is rather racy, and it's safer to move
snd_timer_notify() call inside the main spinlock.
The patch also contains a slight refactoring / cleanup of the code.
Now all start/stop/continue/pause look more symmetric and a bit better
readable.
Signed-off-by: Takashi Iwai <tiwai@suse.de>
[bwh: Backported to 3.16:
- Fix up another use of "event" in _snd_timer_stop()
- Adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit faf37c44a1 upstream.
Clear the PCR (Processor Compatibility Register) on boot to ensure we
are not running in a compatibility mode.
We've seen this cause problems when a crash (and kdump) occurs while
running compat mode guests. The kdump kernel then runs with the PCR
set and causes problems. The symptom in the kdump kernel (also seen in
petitboot after fast-reboot) is early userspace programs taking
sigills on newer instructions (seen in libc).
Signed-off-by: Michael Neuling <mikey@neuling.org>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
[bwh: Backported to 3.16: Drop changes in __{setup,restore}_cpu_power9
and __restore_cpu_cpufeatures()]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit 113f99c335 upstream.
Device features may change during transmission. In particular with
corking, a device may toggle scatter-gather in between allocating
and writing to an skb.
Do not unconditionally assume that !NETIF_F_SG at write time implies
that the same held at alloc time and thus the skb has sufficient
tailroom.
This issue predates git history.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Reported-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Willem de Bruijn <willemb@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit c1d2a31397 upstream.
Similarly to opal_event_shutdown, opal_nvram_write can be called in
the crash path with irqs disabled. Special case the delay to avoid
sleeping in invalid context.
Fixes: 3b8070335f ("powerpc/powernv: Fix OPAL NVRAM driver OPAL_BUSY loops")
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
Signed-off-by: Michael Ellerman <mpe@ellerman.id.au>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit 30035e4575 upstream.
The strscpy() API is intended to be used instead of strlcpy(),
and instead of most uses of strncpy().
- Unlike strlcpy(), it doesn't read from memory beyond (src + size).
- Unlike strlcpy() or strncpy(), the API provides an easy way to check
for destination buffer overflow: an -E2BIG error return value.
- The provided implementation is robust in the face of the source
buffer being asynchronously changed during the copy, unlike the
current implementation of strlcpy().
- Unlike strncpy(), the destination buffer will be NUL-terminated
if the string in the source buffer is too long.
- Also unlike strncpy(), the destination buffer will not be updated
beyond the NUL termination, avoiding strncpy's behavior of zeroing
the entire tail end of the destination buffer. (A memset() after
the strscpy() can be used if this behavior is desired.)
- The implementation should be reasonably performant on all
platforms since it uses the asm/word-at-a-time.h API rather than
simple byte copy. Kernel-to-kernel string copy is not considered
to be performance critical in any case.
Signed-off-by: Chris Metcalf <cmetcalf@ezchip.com>
[bwh: Backported to 3.16: adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit 19c22f3a29 upstream.
arch/tile added word-at-a-time.h after the patch that added generic-y
entries; the generic-y entry is now stale.
arch/h8300 is newer than the generic-y patch for word-at-a-time.h,
and needs a generic-y entry.
arch/powerpc seems to have gotten a generic-y entry by mistake in
the first patch; this change removes it.
Signed-off-by: Chris Metcalf <cmetcalf@ezchip.com>
[bwh: Backported to 3.16:
- Drop change in arch/h8300, which doesn't exist here
- Drop change in arch/tile, which is still using the generic implementation]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit a6e2f029ae upstream.
Added the x86 implementation of word-at-a-time to the
generic version, which previously only supported big-endian.
Omitted the x86-specific load_unaligned_zeropad(), which in
any case is also not present for the existing BE-only
implementation of a word-at-a-time, and is only used under
CONFIG_DCACHE_WORD_ACCESS.
Added as a "generic-y" to the Kbuilds of all architectures
that didn't previously have it.
Signed-off-by: Chris Metcalf <cmetcalf@ezchip.com>
[bwh: Backported to 3.16:
- Drop change in arch/nios2
- Adjust context]
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit 378831e4da upstream.
Doing faccessat("/afs/some/directory", 0) triggers a BUG in the permissions
check code.
Fix this by just removing the BUG section. If no permissions are asked
for, just return okay if the file exists.
Also:
(1) Split up the directory check so that it has separate if-statements
rather than if-else-if (e.g. checking for MAY_EXEC shouldn't skip the
check for MAY_READ and MAY_WRITE).
(2) Check for MAY_CHDIR as MAY_EXEC.
Without the main fix, the following BUG may occur:
kernel BUG at fs/afs/security.c:386!
invalid opcode: 0000 [#1] SMP PTI
...
RIP: 0010:afs_permission+0x19d/0x1a0 [kafs]
...
Call Trace:
? inode_permission+0xbe/0x180
? do_faccessat+0xdc/0x270
? do_syscall_64+0x60/0x1f0
? entry_SYSCALL_64_after_hwframe+0x49/0xbe
Fixes: 00d3b7a453 ("[AFS]: Add security support.")
Reported-by: Jonathan Billings <jsbillings@jsbillings.org>
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit fd2498211a upstream.
The AFS_ACE_READ and AFS_ACE_WRITE permission bits should not
be used to make access decisions for the directory itself. They
are meant to control access for the objects contained in that
directory.
Reading a directory is allowed if the AFS_ACE_LOOKUP bit is set.
This would cause an incorrect access denied error for a directory
with AFS_ACE_LOOKUP but not AFS_ACE_READ.
The AFS_ACE_WRITE bit does not allow operations that modify the
directory. For a directory with AFS_ACE_WRITE but neither
AFS_ACE_INSERT nor AFS_ACE_DELETE, this would result in trying
operations that would ultimately be denied by the server.
Signed-off-by: Marc Dionne <marc.dionne@auristor.com>
Signed-off-by: David Howells <dhowells@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit 5596fe3449 upstream.
for_each_cpu() unintuitively reports CPU0 as set independent of the actual
cpumask content on UP kernels. This causes an unexpected PIT interrupt
storm on a UP kernel running in an SMP virtual machine on Hyper-V, and as
a result, the virtual machine can suffer from a strange random delay of 1~20
minutes during boot-up, and sometimes it can hang forever.
Protect if by checking whether the cpumask is empty before entering the
for_each_cpu() loop.
[ tglx: Use !IS_ENABLED(CONFIG_SMP) instead of #ifdeffery ]
Signed-off-by: Dexuan Cui <decui@microsoft.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: Josh Poulson <jopoulso@microsoft.com>
Cc: "Michael Kelley (EOSG)" <Michael.H.Kelley@microsoft.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Rakib Mullick <rakib.mullick@gmail.com>
Cc: Jork Loeser <Jork.Loeser@microsoft.com>
Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: KY Srinivasan <kys@microsoft.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Alexey Dobriyan <adobriyan@gmail.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Link: https://lkml.kernel.org/r/KL1P15301MB000678289FE55BA365B3279ABF990@KL1P15301MB0006.APCP153.PROD.OUTLOOK.COM
Link: https://lkml.kernel.org/r/KL1P15301MB0006FA63BC22BEB64902EAA0BF930@KL1P15301MB0006.APCP153.PROD.OUTLOOK.COM
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit bb7298a7e8 upstream.
VPIF capture driver expects card name to be set since it
uses it without checking for NULL. The commit which
introduced VPIF display and capture support added card
name only for display, not for capture.
Set it in platform data to probe driver successfully.
While at it, also fix the display card name to something more
appropriate.
Fixes: 85609c1ccd ("DaVinci: DM646x - platform changes for vpif capture and display drivers")
Signed-off-by: Sekhar Nori <nsekhar@ti.com>
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit 35cd67a0ca upstream.
Returning zero is wrong in this case.
Signed-off-by: Peter Rosin <peda@axentia.se>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
Fixes: 174a13aa86 ("i2c: Add viperboard i2c master driver")
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit 12d9bbc5a7 upstream.
Returning -1 (-EPERM) is not appropriate here, go with -EIO.
Signed-off-by: Peter Rosin <peda@axentia.se>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
Fixes: 1b144df1d7 ("i2c: New PMC MSP71xx TWI bus driver")
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
commit de9a8634f1 upstream.
Returning zero is wrong in this case.
Signed-off-by: Peter Rosin <peda@axentia.se>
Signed-off-by: Wolfram Sang <wsa@the-dreams.de>
Fixes: 1b144df1d7 ("i2c: New PMC MSP71xx TWI bus driver")
Signed-off-by: Ben Hutchings <ben@decadent.org.uk>
