When INIT/SIPI sequence is sent to VCPU which before that
was in use by OS, VMRUN might fail with:
KVM: entry failed, hardware error 0xffffffff
EAX=00000000 EBX=00000000 ECX=00000000 EDX=000006d3
ESI=00000000 EDI=00000000 EBP=00000000 ESP=00000000
EIP=00000000 EFL=00000002 [-------] CPL=0 II=0 A20=1 SMM=0 HLT=0
ES =0000 00000000 0000ffff 00009300
CS =9a00 0009a000 0000ffff 00009a00
[...]
CR0=60000010 CR2=b6f3e000 CR3=01942000 CR4=000007e0
[...]
EFER=0000000000000000
with corresponding SVM error:
KVM: FAILED VMRUN WITH VMCB:
[...]
cpl: 0 efer: 0000000000001000
cr0: 0000000080010010 cr2: 00007fd7fe85bf90
cr3: 0000000187d0c000 cr4: 0000000000000020
[...]
What happens is that VCPU state right after offlinig:
CR0: 0x80050033 EFER: 0xd01 CR4: 0x7e0
-> long mode with CR3 pointing to longmode page tables
and when VCPU gets INIT/SIPI following transition happens
CR0: 0 -> 0x60000010 EFER: 0x0 CR4: 0x7e0
-> paging disabled with stale CR3
However SVM under the hood puts VCPU in Paged Real Mode*
which effectively translates CR0 0x60000010 -> 80010010 after
svm_vcpu_reset()
-> init_vmcb()
-> kvm_set_cr0()
-> svm_set_cr0()
but from kvm_set_cr0() perspective CR0: 0 -> 0x60000010
only caching bits are changed and
commit d81135a57a
("KVM: x86: do not reset mmu if CR0.CD and CR0.NW are changed")'
regressed svm_vcpu_reset() which relied on MMU being reset.
As result VMRUN after svm_vcpu_reset() tries to run
VCPU in Paged Real Mode with stale MMU context (longmode page tables),
which causes some AMD CPUs** to bail out with VMEXIT_INVALID.
Fix issue by unconditionally resetting MMU context
at init_vmcb() time.
* AMD64 Architecture Programmer’s Manual,
Volume 2: System Programming, rev: 3.25
15.19 Paged Real Mode
** Opteron 1216
Signed-off-by: Igor Mammedov <imammedo@redhat.com>
Fixes: d81135a57a
Cc: stable@vger.kernel.org
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
Commit 2ee507c472 ("sched: Add function single_task_running to let a task
check if it is the only task running on a cpu") referenced the current
runqueue with the smp_processor_id. When CONFIG_DEBUG_PREEMPT is enabled,
that is only allowed if preemption is disabled or the currrent task is
bound to the local cpu (e.g. kernel worker).
With commit f781951299 ("kvm: add halt_poll_ns module parameter") KVM
calls single_task_running. If CONFIG_DEBUG_PREEMPT is enabled that
generates a lot of kernel messages.
To avoid adding preemption in that cases, as it would limit the usefulness,
we change single_task_running to access directly the cpu local runqueue.
Cc: Tim Chen <tim.c.chen@linux.intel.com>
Suggested-by: Peter Zijlstra <peterz@infradead.org>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: <stable@vger.kernel.org>
Fixes: 2ee507c472
Signed-off-by: Dominik Dingel <dingel@linux.vnet.ibm.com>
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
As discussed on linux-arch all architectures should wire up the separate
system calls that are hidden behind the socketcall multiplexer system call.
It's just a couple more system calls and gives us a very small performance
improvement.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
A couple of compat wrapper functions are simply trampolines to the real
system call. This happened because the compat wrapper defines will only
sign and zero extend system call parameters which are of different size
on s390/s390x (longs and pointers).
All other parameters will be correctly sign and zero extended by the
normal system call wrappers.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Add notrace to the compat wrapper define to disable tracing of compat
wrapper functions. These are supposed to be very small and more or less
just a trampoline to the real system call.
Also fix indentation.
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
This patch adds ratelimited version of the BT_ERR macro.
Signed-off-by: Szymon Janc <ext.szymon.janc@tieto.com>
Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
Memory placement in sch_dsmark is silly : Better place mask/value
in the same cache line.
Also, we can embed small arrays in the first cache line and
remove a potential cache miss.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Tracking idle time in bictcp_cwnd_event() is imprecise, as epoch_start
is normally set at ACK processing time, not at send time.
Doing a proper fix would need to add an additional state variable,
and does not seem worth the trouble, given CUBIC bug has been there
forever before Jana noticed it.
Let's simply not set epoch_start in the future, otherwise
bictcp_update() could overflow and CUBIC would again
grow cwnd too fast.
This was detected thanks to a packetdrill test Neal wrote that was flaky
before applying this fix.
Fixes: 30927520db ("tcp_cubic: better follow cubic curve after idle period")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Cc: Jana Iyengar <jri@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Dan Carpenter reported off-by-one error of fjes at
http://www.mail-archive.com/netdev@vger.kernel.org/msg77520.html
Actually this is a bug.
ep_shm_info[epidx].{es_status, zone} should be update
inside for loop.
This patch fixes this bug.
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Taku Izumi <izumi.taku@jp.fujitsu.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
I got this automated message from <shahed.shaikh@qlogic.com> when submitting
a qlcnic patch:
> Shahed Shaikh is no longer with QLogic. If you require assistance please
> contact Ariel Elior Ariel.Elior@qlogic.com
There's no point in having a bouncing address in MAINTAINERS.
CC: Dept-GELinuxNICDev@qlogic.com
CC: Ariel Elior <Ariel.Elior@qlogic.com>
Signed-off-by: Jiri Benc <jbenc@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Jiri Benc says:
====================
vxlan fixes
This fixes various issues with vxlan related to IPv6.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
The callback for adding vxlan port can be called with the same port for
both IPv4 and IPv6. Do not disable the offloading when the same port for
both protocols is added and later one of them removed.
Signed-off-by: Jiri Benc <jbenc@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The callback for adding vxlan port can be called with the same port for both
IPv4 and IPv6. Do not disable the offloading if this occurs.
Signed-off-by: Jiri Benc <jbenc@redhat.com>
Acked-by: Sathya Perla <sathya.perla@avagotech.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The callback for adding vxlan port can be called with the same port for
both IPv4 and IPv6. Do not disable the offloading when the same port for
both protocols is added and later one of them removed.
Signed-off-by: Jiri Benc <jbenc@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When IPv6 address is set without IPv6 configured, the vxlan socket is mostly
treated as an IPv4 one but various lookus in fdb etc. still take the
AF_INET6 into account. This creates incosistencies with weird consequences.
Just reject IPv6 addresses in such case.
Signed-off-by: Jiri Benc <jbenc@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
vxlan_setup is called when allocating the net_device, i.e. way before
vxlan_newlink (or vxlan_dev_configure) is called. This means
vxlan->default_dst is actually unset in vxlan_setup and the condition that
sets needed_headroom always takes the else branch.
Set the needed_headrom at the point when we have the information about
the address family available.
Fixes: e4c7ed4153 ("vxlan: add ipv6 support")
Fixes: 2853af6a2e ("vxlan: use dev->needed_headroom instead of dev->hard_header_len")
CC: Cong Wang <cwang@twopensource.com>
Signed-off-by: Jiri Benc <jbenc@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add entry for arcnet to MAINTAINERS file and add myself as the
maintainer of the subsystem.
Signed-off-by: Michael Grzeschik <m.grzeschik@pengutronix.de>
Cc: davem@davemloft.net
Cc: joe@perches.com
Signed-off-by: David S. Miller <davem@davemloft.net>
For arcnet the bare minimum header only contains the 4 bytes to
specify source, dest and offset (1, 1 and 2 bytes respectively).
The corresponding struct is struct arc_hardware.
The struct archdr contains additionally a union of possible soft
headers. When doing $insertusecasehere packets might well
include short (or even no?) soft headers.
For this reason only use arc_hardware instead of archdr to
determine the hard_header_len for an arcnet device.
Signed-off-by: Michael Grzeschik <m.grzeschik@pengutronix.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Johan Hedberg says:
====================
pull request: bluetooth 2015-09-17
Here's one important patch for the 4.3-rc series that fixes an issue
with Bluetooth LE encryption failing because of a too early check for
the SMP context.
Please let me know if there are any issues pulling. Thanks.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Florian Fainelli says:
====================
net: bcmgenet: Interrupt coalescing
This patch series adds support for interrupt coalescing for GENET
adapters.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Add support for the ethtool rx-frames coalescing parameter which allows
defining the number of RX interrupts per frames received. The RDMA
engine supports a configurable timeout with a resolution of
approximately 8.192 us.
We can no longer enable the BDONE/PDONE interrupts as those would
fire for each packet/buffer received, which would defeat the MBDONE
interrupt purpose. The MBDONE interrupt is guaranteed to correspond to a
PDONE/BDONE interrupt when the threshold is set to 1.
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Configuring the ethtool tx-frames property, which translates into N
packets before a TX interrupt is the simplest configuration scheme
because it requires no locking neither at the softare nor hardware
level, and is completely indepedent from the link speed. Since ethtool
does not allow per-tx queue coalescing parameters, we apply the same
setting to any transmit queue.
We can no longer enable the BDONE/PDONE interrupts as those would fire
for each packet/buffer received, which would defeat the MBDONE interrupt
purpose. The MBDONE interrupt is guaranteed to correspond to a
PDONE/BDONE interrupt when the threshold is set to 1, but offers
interrupt coalescing when the value is > 1.
Since the HW is configured to generate an interrupt when the ring
becomes emtpy, we have to deny any timeout/timer settings coming from
user-space to indicate we can only generate an interrupt very <N>
packets.
While we are at it, fix the DMA_INTR_THRESHOLD_MASK value which was off
by one bit (0xff vs. 0x1ff).
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Remove not defined MAC_CR_GMII_EN_ bit from MAC_CR.
Signed-off-by: Woojung Huh <woojung.huh@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Create lan78xx_get_mdix_status() and lan78xx_set_mdix_status() for MDIX control.
Signed-off-by: Woojung Huh <woojung.huh@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Remove phy defines in lan78xx.h and use defines in include/linux/microchipphy.h.
Signed-off-by: Woojung Huh <woojung.huh@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Update to use phylib instead of mii_if_info.
Signed-off-by: Woojung Huh <woojung.huh@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add PHYLIB and MICROCHIP_PHY as default configuration for lan78xx.
Signed-off-by: Woojung Huh <woojung.huh@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Check device ready bit (PMT_CTL_READY_) after reset the PHY.
Device may not be ready even if PHY_RST_ is cleared depends on configuration.
Signed-off-by: Woojung Huh <woojung.huh@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Fix an issue introduced by the previous major toshiba rework. Add a quirk.
Workaround a few platform specific firmware items. One cleanup to wmi I
inadvertently dropped from a previous pull request.
hp-wmi
- limit hotkey enable
toshiba_acpi:
- Fix hotkeys registration on some toshiba models
- Fix USB Sleep and Music always disabled
wmi:
- Remove private %pUL implementation
asus-nb-wmi:
- Add wapf=4 quirk for X456UA/X456UF
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1
iQEcBAABAgAGBQJV+2i5AAoJEKbMaAwKp364AmwIAKZd3N2lsI15wqx8fYKrvBTR
tdov/1v4N0T79BAngZNF7dIrSCzRGtjCK7w1zoR6B8vkrpJg2vgXS89GA9p0gkdq
jOQGuDswBDDD9KWDhlR0SjXE+P8T35gQMeeHNC8mpAfn9S9IGKEkLW4UsPF8tDxx
TmIv2b2kH8LN2SdmgkK1aNrDvm/Z7Z+2VH1a87NlNKQtJJFjxx5CVKs09UrzGOVG
ETOZ5+vq+7LR/ajzquEyPIZIFncz6adkxVKvFs9FuTFYza6VNBH+ED0QzijCY4mn
nAD7cIEa5b9/CSDgwaueBbCzLAkl0csPwjJiA4JaBWgVSn5ctZUUi0b+dApIU88=
=DQTr
-----END PGP SIGNATURE-----
Merge tag 'platform-drivers-x86-v4.3-2' of git://git.infradead.org/users/dvhart/linux-platform-drivers-x86
Pull x86 platform driver fixes from Darren Hart:
"Fix an issue introduced by the previous major toshiba rework. Add a
quirk. Workaround a few platform specific firmware items. One
cleanup to wmi I inadvertently dropped from a previous pull request.
Details:
hp-wmi:
- limit hotkey enable
toshiba_acpi:
- Fix hotkeys registration on some toshiba models
- Fix USB Sleep and Music always disabled
wmi:
- Remove private %pUL implementation
asus-nb-wmi:
- Add wapf=4 quirk for X456UA/X456UF"
* tag 'platform-drivers-x86-v4.3-2' of git://git.infradead.org/users/dvhart/linux-platform-drivers-x86:
hp-wmi: limit hotkey enable
toshiba_acpi: Fix hotkeys registration on some toshiba models
toshiba_acpi: Fix USB Sleep and Music always disabled
wmi: Remove private %pUL implementation
asus-nb-wmi: Add wapf=4 quirk for X456UA/X456UF
David Woodhouse reports skb_under_panic when we try to push ethernet
header to fragmented ipv6 skbs:
skbuff: skb_under_panic: text:c1277f1e len:1294 put:14 head:dec98000
data:dec97ffc tail:0xdec9850a end:0xdec98f40 dev:br-lan
[..]
ip6_finish_output2+0x196/0x4da
David further debugged this:
[..] offending fragments were arriving here with skb_headroom(skb)==10.
Which is reasonable, being the Solos ADSL card's header of 8 bytes
followed by 2 bytes of PPP frame type.
The problem is that if netfilter ipv6 defragmentation is used, skb_cow()
in ip6_forward will only see reassembled skb.
Therefore, headroom is overestimated by 8 bytes (we pulled fragment
header) and we don't check the skbs in the frag_list either.
We can't do these checks in netfilter defrag since outdev isn't known yet.
Furthermore, existing tests in ip6_fragment did not consider the fragment
or ipv6 header size when checking headroom of the fraglist skbs.
While at it, also fix a skb leak on memory allocation -- ip6_fragment
must consume the skb.
I tested this e1000 driver hacked to not allocate additional headroom
(we end up in slowpath, since LL_RESERVED_SPACE is 16).
If 2 bytes of headroom are allocated, fastpath is taken (14 byte
ethernet header was pulled, so 16 byte headroom available in all
fragments).
Reported-by: David Woodhouse <dwmw2@infradead.org>
Diagnosed-by: David Woodhouse <dwmw2@infradead.org>
Signed-off-by: Florian Westphal <fw@strlen.de>
Tested-by: David Woodhouse <David.Woodhouse@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
A comment in include/linux/skbuff.h says that:
* Various parts of the networking layer expect at least 32 bytes of
* headroom, you should not reduce this.
This was demonstrated by a panic when handling fragmented IPv6 packets:
http://marc.info/?l=linux-netdev&m=144236093519172&w=2
It's not entirely clear if that comment is still valid — and if it is,
perhaps netif_rx() ought to be enforcing it with a warning.
But either way, it is rather stupid from a performance point of view
for us to be receiving packets into a buffer which doesn't have enough
room to prepend an Ethernet header — it means that *every* incoming
packet is going to be need to be reallocated. So let's fix that.
Signed-off-by: David Woodhouse <David.Woodhouse@intel.com>
Acked-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Drivers needs to export the OF id table and this be built into
the module or udev won't have the necessary information to autoload
the driver module when the device is registered via OF.
Signed-off-by: Javier Martinez Canillas <javier@osg.samsung.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Merge misc fixes from ANdrew Morton:
"8 fixes"
* emailed patches from Andrew Morton <akpm@linux-foundation.org>:
revert "mm: make sure all file VMAs have ->vm_ops set"
MAINTAINERS: update LTP mailing list
userfaultfd: add missing mmput() in error path
lib/string_helpers.c: fix infinite loop in string_get_size()
alpha: lib: export __delay
alpha: io: define ioremap_uc
kasan: fix last shadow judgement in memory_is_poisoned_16()
zram: fix possible use after free in zcomp_create()
Revert commit 6dc296e7df "mm: make sure all file VMAs have ->vm_ops
set".
Will Deacon reports that it "causes some mmap regressions in LTP, which
appears to use a MAP_PRIVATE mmap of /dev/zero as a way to get anonymous
pages in some of its tests (specifically mmap10 [1])".
William Shuman reports Oracle crashes.
So revert the patch while we work out what to do.
Reported-by: William Shuman <wshuman3@gmail.com>
Reported-by: Will Deacon <will.deacon@arm.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Oleg Nesterov <oleg@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
This fixes a memleak if anon_inode_getfile() fails in userfaultfd().
Signed-off-by: Eric Biggers <ebiggers3@gmail.com>
Signed-off-by: Andrea Arcangeli <aarcange@redhat.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Some string_get_size() calls (e.g.:
string_get_size(1, 512, STRING_UNITS_10, ..., ...)
string_get_size(15, 64, STRING_UNITS_10, ..., ...)
) result in an infinite loop. The problem is that if size is equal to
divisor[units]/blk_size and is smaller than divisor[units] we'll end
up with size == 0 when we start doing sf_cap calculations:
For string_get_size(1, 512, STRING_UNITS_10, ..., ...) case:
...
remainder = do_div(size, divisor[units]); -> size is 0, remainder is 1
remainder *= blk_size; -> remainder is 512
...
size *= blk_size; -> size is still 0
size += remainder / divisor[units]; -> size is still 0
The caller causing the issue is sd_read_capacity(), the problem was
noticed on Hyper-V, such weird size was reported by host when scanning
collides with device removal. This is probably a separate issue worth
fixing, this patch is intended to prevent the library routine from
infinite looping.
Signed-off-by: Vitaly Kuznetsov <vkuznets@redhat.com>
Acked-by: James Bottomley <JBottomley@Odin.com>
Cc: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Cc: Rasmus Villemoes <linux@rasmusvillemoes.dk>
Cc: "K. Y. Srinivasan" <kys@microsoft.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
__delay was not exported as a result while building with allmodconfig we
were getting build error of undefined symbol. __delay is being used by:
drivers/net/phy/mdio-octeon.c
Signed-off-by: Sudip Mukherjee <sudip@vectorindia.org>
Cc: Richard Henderson <rth@twiddle.net>
Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: Matt Turner <mattst88@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
ioremap_uc was not defined and as a result while building with
allmodconfig were getting build error of: implicit declaration of
function 'ioremap_uc'.
Signed-off-by: Sudip Mukherjee <sudip@vectorindia.org>
Cc: Richard Henderson <rth@twiddle.net>
Cc: Ivan Kokshaysky <ink@jurassic.park.msu.ru>
Cc: Matt Turner <mattst88@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The shadow which correspond 16 bytes memory may span 2 or 3 bytes. If
the memory is aligned on 8, then the shadow takes only 2 bytes. So we
check "shadow_first_bytes" is enough, and need not to call
"memory_is_poisoned_1(addr + 15);". But the code "if
(likely(!last_byte))" is wrong judgement.
e.g. addr=0, so last_byte = 15 & KASAN_SHADOW_MASK = 7, then the code
will continue to call "memory_is_poisoned_1(addr + 15);"
Signed-off-by: Xishi Qiu <qiuxishi@huawei.com>
Acked-by: Andrey Ryabinin <aryabinin@virtuozzo.com>
Cc: Andrey Konovalov <adech.fo@gmail.com>
Cc: Rusty Russell <rusty@rustcorp.com.au>
Cc: Michal Marek <mmarek@suse.cz>
Cc: <zhongjiang@huawei.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
zcomp_create() verifies the success of zcomp_strm_{multi,single}_create()
through comp->stream, which can potentially be pointing to memory that
was freed if these functions returned an error.
While at it, replace a 'ERR_PTR(-ENOMEM)' by a more generic
'ERR_PTR(error)' as in the future zcomp_strm_{multi,siggle}_create()
could return other error codes. Function documentation updated
accordingly.
Fixes: beca3ec71f ("zram: add multi stream functionality")
Signed-off-by: Luis Henriques <luis.henriques@canonical.com>
Acked-by: Sergey Senozhatsky <sergey.senozhatsky@gmail.com>
Acked-by: Minchan Kim <minchan@kernel.org>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Alexei Starovoitov says:
====================
bpf: performance improvements
v1->v2: dropped redundant iff_up check in patch 2
At plumbers we discussed different options on how to get rid of skb_clone
from bpf_clone_redirect(), the patch 2 implements the best option.
Patch 1 adds 'integrated exts' to cls_bpf to improve performance by
combining simple actions into bpf classifier.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Existing bpf_clone_redirect() helper clones skb before redirecting
it to RX or TX of destination netdev.
Introduce bpf_redirect() helper that does that without cloning.
Benchmarked with two hosts using 10G ixgbe NICs.
One host is doing line rate pktgen.
Another host is configured as:
$ tc qdisc add dev $dev ingress
$ tc filter add dev $dev root pref 10 u32 match u32 0 0 flowid 1:2 \
action bpf run object-file tcbpf1_kern.o section clone_redirect_xmit drop
so it receives the packet on $dev and immediately xmits it on $dev + 1
The section 'clone_redirect_xmit' in tcbpf1_kern.o file has the program
that does bpf_clone_redirect() and performance is 2.0 Mpps
$ tc filter add dev $dev root pref 10 u32 match u32 0 0 flowid 1:2 \
action bpf run object-file tcbpf1_kern.o section redirect_xmit drop
which is using bpf_redirect() - 2.4 Mpps
and using cls_bpf with integrated actions as:
$ tc filter add dev $dev root pref 10 \
bpf run object-file tcbpf1_kern.o section redirect_xmit integ_act classid 1
performance is 2.5 Mpps
To summarize:
u32+act_bpf using clone_redirect - 2.0 Mpps
u32+act_bpf using redirect - 2.4 Mpps
cls_bpf using redirect - 2.5 Mpps
For comparison linux bridge in this setup is doing 2.1 Mpps
and ixgbe rx + drop in ip_rcv - 7.8 Mpps
Signed-off-by: Alexei Starovoitov <ast@plumgrid.com>
Acked-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: John Fastabend <john.r.fastabend@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Often cls_bpf classifier is used with single action drop attached.
Optimize this use case and let cls_bpf return both classid and action.
For backwards compatibility reasons enable this feature under
TCA_BPF_FLAG_ACT_DIRECT flag.
Then more interesting programs like the following are easier to write:
int cls_bpf_prog(struct __sk_buff *skb)
{
/* classify arp, ip, ipv6 into different traffic classes
* and drop all other packets
*/
switch (skb->protocol) {
case htons(ETH_P_ARP):
skb->tc_classid = 1;
break;
case htons(ETH_P_IP):
skb->tc_classid = 2;
break;
case htons(ETH_P_IPV6):
skb->tc_classid = 3;
break;
default:
return TC_ACT_SHOT;
}
return TC_ACT_OK;
}
Joint work with Daniel Borkmann.
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Signed-off-by: Alexei Starovoitov <ast@plumgrid.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When changing rss key, we do not want to overwrite user provided key
by the one provided by netdev_rss_key_fill(), which is the host random
key generated at boot time.
Fixes: 947cbb0ac2 ("net/mlx4_en: Support for configurable RSS hash function")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Eyal Perry <eyalpe@mellanox.com>
CC: Amir Vadai <amirv@mellanox.com>
Acked-by: Or Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The permanent protocol nodes are at the head of the list,
So only need check all these nodes.
No matter the new node is permanent or not,
insert the new node after the last permanent protocol node,
If the new node conflicts with existing permanent node,
return error.
Signed-off-by: Martin Zhang <martinbj2008@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
