Add the missing unlock before return from function sk_buff()
in the error handling case.
Fixes: f3097be21bf1 ("net: dsa: sja1105: Add a state machine for RX
timestamping")
Signed-off-by: Wei Yongjun
---
net/dsa/tag_sja1105.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/net/dsa/tag_sja110
[sorry, resend again as previous one has come text messed out due to
networking issues]
On Tue, Jul 16, 2019 at 10:08 PM Y Song wrote:
>
> On Tue, Jul 16, 2019 at 4:59 AM Ilya Leoshkevich wrote:
> >
> > test_pkt_md_access is failing on s390, since the associated eBPF prog
> > returns TC_ACT_SHOT
On Tue, Jul 16, 2019 at 4:59 AM Ilya Leoshkevich wrote:
>
> test_pkt_md_access is failing on s390, since the associated eBPF prog
> returns TC_ACT_SHOT, which in turn happens because loading a part of a
> struct __sk_buff field produces an incorrect result.
>
> The problem is that when verifier em
I think there is a problem if dev_watchdog() is triggered before
netif_carrier_off(). dev_watchdog() might call ->ndo_tx_timeout(), i.e.
be_tx_timeout(), if txq timeout happens. Thus be_tx_timeout() could still be
able to access the memory which is being freed by be_update_queues().
Thanks,
Fi
On Mon, 15 Jul 2019 13:49:01 -0700, John Fastabend wrote:
> Resolve a series of splats discovered by syzbot and an unhash
> TLS issue noted by Eric Dumazet.
I spent most of today poking at this set, and I'll continue tomorrow.
I'm not capitulating yet, but if I can't get it to work for tls_device
Thank you for your feedback!
On Tue, Jul 16, 2019 at 7:26 PM Alexei Starovoitov
wrote:
>
> On Tue, Jul 16, 2019 at 09:59:26AM +0200, Eric Dumazet wrote:
> >
> >
> > On 7/16/19 2:26 AM, Petar Penkov wrote:
> > > From: Petar Penkov
> > >
> > > This helper function allows BPF programs to try to gen
On Tue, Jul 16, 2019 at 09:59:26AM +0200, Eric Dumazet wrote:
>
>
> On 7/16/19 2:26 AM, Petar Penkov wrote:
> > From: Petar Penkov
> >
> > This helper function allows BPF programs to try to generate SYN
> > cookies, given a reference to a listener socket. The function works
> > from XDP and wit
On Tue, Jul 16, 2019 at 10:42 AM Andrii Nakryiko
wrote:
>
> On Tue, Jul 16, 2019 at 5:59 AM Ilya Leoshkevich wrote:
> >
> > perf_buffer test fails for exactly the same reason test_attach_probe
> > used to fail: different nanosleep syscall kprobe name.
> >
> > Reuse the test_attach_probe fix.
> >
On Tue, Jul 16, 2019 at 5:22 PM Andrii Nakryiko
wrote:
> > Makefile becomes hairier by the day, thx for cleaning it up a bit :-)
> >
> > > But I also don't think we need to worry about creating them, because
> > > there is always at least one test (otherwise those tests are useless
> > > anyways)
On Tue, Jul 16, 2019 at 3:12 PM Jakub Kicinski
wrote:
>
> On Tue, 16 Jul 2019 09:17:03 -0700, Alexei Starovoitov wrote:
> > I don't think we have a test for such 'dead prog only due to verifier walk'
> > situation. I wonder what happens :)
>
> FWIW we do have verifier and BTF self tests for dead c
On 7/16/19 3:29 PM, Gerd Rausch wrote:
Since "rds_ib_free_frmr" and "rds_ib_free_frmr_list" simply put
the FRMR memory segments on the "drop_list" or "free_list",
and it is the job of "rds_ib_flush_mr_pool" to reap those entries
by ultimately issuing a "IB_WR_LOCAL_INV" work-request,
we need to t
On 7/16/19 3:29 PM, Gerd Rausch wrote:
Otherwise, if an IB connection is torn down before "rds_ib_setup_qp"
is called, the value of "ic->i_fastreg_wrs" is still at zero
(as it wasn't initialized by "rds_ib_setup_qp").
Consequently "rds_ib_conn_path_shutdown" will spin forever,
waiting for it t
On 7/16/19 3:29 PM, Gerd Rausch wrote:
Fix a bug where fr_state first goes to FRMR_IS_STALE, because of a failure
of operation IB_WR_LOCAL_INV, but then gets set back to "FRMR_IS_FREE"
uncoditionally, even though the operation failed.
Signed-off-by: Gerd Rausch
---
Acked-by: Santosh Shilimk
On 7/16/19 3:29 PM, Gerd Rausch wrote:
In order to:
1) avoid a silly bouncing between "clean_list" and "drop_list"
triggered by function "rds_ib_reg_frmr" as it is releases frmr
regions whose state is not "FRMR_IS_FREE" right away.
2) prevent an invalid access error in a race from a pend
On 7/16/19 3:29 PM, Gerd Rausch wrote:
Make function "rds_ib_try_reuse_ibmr" return NULL in case
memory region could not be allocated, since callers
simply check if the return value is not NULL.
Signed-off-by: Gerd Rausch
---
Acked-by: Santosh Shilimkar
Thank you for the reviews!
On Tue, Jul 16, 2019 at 4:56 AM Lorenz Bauer wrote:
>
> On Tue, 16 Jul 2019 at 08:59, Eric Dumazet wrote:
> > > + return -EINVAL;
> > > +
> > > + if (sk->sk_protocol != IPPROTO_TCP || sk->sk_state != TCP_LISTEN)
> > > + return -EINVAL;
> > >
On 7/16/19 3:28 PM, Gerd Rausch wrote:
Waiting for activity on the "clean_list" to quiesce is no substitute
for proper locking.
We can have multiple threads competing for "llist_del_first"
via "rds_ib_reuse_mr", and a single thread competing
for "llist_del_all" and "llist_del_first" via "rds_ib_
On 7/16/19 3:28 PM, Gerd Rausch wrote:
In the context of FRMR (ib_frmr.c):
Memory regions make it onto the "clean_list" via "rds_ib_flush_mr_pool",
after the memory region has been posted for invalidation via
"rds_ib_post_inv".
At that point in time, "fr_state" may still be in state "FRMR_IS_IN
On Tue, Jul 16, 2019 at 4:35 AM Lorenz Bauer wrote:
>
> On Tue, 16 Jul 2019 at 01:27, Petar Penkov wrote:
> >
> > From: Petar Penkov
> >
> > This patch allows generation of a SYN cookie before an SKB has been
> > allocated, as is the case at XDP.
> >
> > Signed-off-by: Petar Penkov
> > ---
> >
On Tue, Jul 16, 2019 at 5:15 PM Stanislav Fomichev wrote:
>
> On 07/16, Andrii Nakryiko wrote:
> > On Tue, Jul 16, 2019 at 3:57 PM Stanislav Fomichev wrote:
> > >
> > > On 07/16, Andrii Nakryiko wrote:
> > > > On Tue, Jul 16, 2019 at 12:55 PM Stanislav Fomichev
> > > > wrote:
> > > > >
> > > >
On 07/16, Andrii Nakryiko wrote:
> On Tue, Jul 16, 2019 at 3:57 PM Stanislav Fomichev wrote:
> >
> > On 07/16, Andrii Nakryiko wrote:
> > > On Tue, Jul 16, 2019 at 12:55 PM Stanislav Fomichev
> > > wrote:
> > > >
> > > > On 07/16, Andrii Nakryiko wrote:
> > > > > e46fc22e60a4 ("selftests/bpf: ma
On Tue, Jul 16, 2019 at 10:49 PM Marek Majkowski wrote:
>
> Morning,
>
> My poor man's fuzzer found something interesting in SCTP. It seems
> like creating large number of SCTP sockets + some magic dance, upsets
> a memory subsystem related to SCTP. The sequence:
>
> - create SCTP socket
> - cal
On Tue, Jul 16, 2019 at 3:57 PM Stanislav Fomichev wrote:
>
> On 07/16, Andrii Nakryiko wrote:
> > On Tue, Jul 16, 2019 at 12:55 PM Stanislav Fomichev
> > wrote:
> > >
> > > On 07/16, Andrii Nakryiko wrote:
> > > > e46fc22e60a4 ("selftests/bpf: make directory prerequisites order-only")
> > > > e
On Tue, Jul 16, 2019 at 11:47:40PM +0200, Marek Majkowski wrote:
> Morning,
>
> My poor man's fuzzer found something interesting in SCTP. It seems
> like creating large number of SCTP sockets + some magic dance, upsets
> a memory subsystem related to SCTP. The sequence:
>
> - create SCTP socket
On 07/16, Andrii Nakryiko wrote:
> On Tue, Jul 16, 2019 at 12:55 PM Stanislav Fomichev wrote:
> >
> > On 07/16, Andrii Nakryiko wrote:
> > > e46fc22e60a4 ("selftests/bpf: make directory prerequisites order-only")
> > > exposed existing problem in Makefile for test_verifier and test_maps
> > > tes
Fix a bug where fr_state first goes to FRMR_IS_STALE, because of a failure
of operation IB_WR_LOCAL_INV, but then gets set back to "FRMR_IS_FREE"
uncoditionally, even though the operation failed.
Signed-off-by: Gerd Rausch
---
net/rds/ib_frmr.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletio
Since "rds_ib_free_frmr" and "rds_ib_free_frmr_list" simply put
the FRMR memory segments on the "drop_list" or "free_list",
and it is the job of "rds_ib_flush_mr_pool" to reap those entries
by ultimately issuing a "IB_WR_LOCAL_INV" work-request,
we need to trigger and then wait for all those memory
Otherwise, if an IB connection is torn down before "rds_ib_setup_qp"
is called, the value of "ic->i_fastreg_wrs" is still at zero
(as it wasn't initialized by "rds_ib_setup_qp").
Consequently "rds_ib_conn_path_shutdown" will spin forever,
waiting for it to go back to "RDS_IB_DEFAULT_FR_WR",
which o
In the context of FRMR (ib_frmr.c):
Memory regions make it onto the "clean_list" via "rds_ib_flush_mr_pool",
after the memory region has been posted for invalidation via
"rds_ib_post_inv".
At that point in time, "fr_state" may still be in state "FRMR_IS_INUSE",
since the only place where "fr_stat
Make function "rds_ib_try_reuse_ibmr" return NULL in case
memory region could not be allocated, since callers
simply check if the return value is not NULL.
Signed-off-by: Gerd Rausch
---
net/rds/ib_rdma.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/net/rds/ib_rdma.c b
In order to:
1) avoid a silly bouncing between "clean_list" and "drop_list"
triggered by function "rds_ib_reg_frmr" as it is releases frmr
regions whose state is not "FRMR_IS_FREE" right away.
2) prevent an invalid access error in a race from a pending
"IB_WR_LOCAL_INV" operation with a t
A number of net/rds fixes necessary to make "rds_rdma.ko"
pass some basic Oracle internal tests.
Gerd Rausch (7):
net/rds: Give fr_state a chance to transition to FRMR_IS_FREE
net/rds: Get rid of "wait_clean_list_grace" and add locking
net/rds: Wait for the FRMR_IS_FREE (or FRMR_IS_STALE) tr
Waiting for activity on the "clean_list" to quiesce is no substitute
for proper locking.
We can have multiple threads competing for "llist_del_first"
via "rds_ib_reuse_mr", and a single thread competing
for "llist_del_all" and "llist_del_first" via "rds_ib_flush_mr_pool".
Since "llist_del_first"
The following scenario was encountered during testing of logical
partition mobility on pseries partitions with bonded ibmvnic
adapters in LACP mode.
1. Driver receives a signal that the device has been
swapped, and it needs to reset to initialize the new
device.
2. Driver reports loss of ca
On Tue, 16 Jul 2019 09:17:03 -0700, Alexei Starovoitov wrote:
> I don't think we have a test for such 'dead prog only due to verifier walk'
> situation. I wonder what happens :)
FWIW we do have verifier and BTF self tests for dead code removal
of entire subprogs! :)
Morning,
My poor man's fuzzer found something interesting in SCTP. It seems
like creating large number of SCTP sockets + some magic dance, upsets
a memory subsystem related to SCTP. The sequence:
- create SCTP socket
- call setsockopts (SCTP_EVENTS)
- call bind(::1, port)
- call sendmsg(long
On Tue, Jul 16, 2019 at 1:58 PM Cong Wang wrote:
>
> Add a test case to simulate the loopback packet case fixed
> in the previous patch.
Well, I actually have to create a dummy1 to simulate it, as dummy0
is the gateway interface assigned to an IP address.
I will send v3.
Thanks.
On Tue, Jul 16, 2019 at 12:55 PM Stanislav Fomichev wrote:
>
> On 07/16, Andrii Nakryiko wrote:
> > e46fc22e60a4 ("selftests/bpf: make directory prerequisites order-only")
> > exposed existing problem in Makefile for test_verifier and test_maps tests:
> > their dependency on auto-generated header
From: Vedang Patel
Date: Tue, 16 Jul 2019 12:52:18 -0700
> During the review of the iproute2 patches for txtime-assist mode, it was
> pointed out that it does not make sense for the txtime-delay parameter to
> be negative. So, change the type of the parameter from s32 to u32.
>
> Fixes: 4cfd5779
Add a test case to simulate the loopback packet case fixed
in the previous patch.
This test gets passed after the fix:
IPv4 rp_filter tests
TEST: rp_filter passes local packets[ OK ]
TEST: rp_filter passes loopback packets [ OK ]
In a rare case where we redirect local packets from veth to lo,
these packets fail to pass the source validation when rp_filter
is turned on, as the tracing shows:
<...>-311708 [040] ..s1 7951180.957825: fib_table_lookup: table 254 oif 0 iif
1 src 10.53.180.130 dst 10.53.180.130 tos 0 scope 0 f
For qdisc's that support TC filters and set TCQ_F_CAN_BYPASS,
notably fq_codel, it makes no sense to let packets bypass the TC
filters we setup in any scenario, otherwise our packets steering
policy could not be enforced.
This can be reproduced easily with the following script:
ip li add dev dum
I request that this patch should also be considered for the net tree since it
fixes the data type of of the txtime_delay parameter and should go in with the
iproute2 patches which implement support for txtime-assist mode.
Thanks,
Vedang Patel
> On Jul 16, 2019, at 12:52 PM, Patel, Vedang wrot
On 07/16, Andrii Nakryiko wrote:
> e46fc22e60a4 ("selftests/bpf: make directory prerequisites order-only")
> exposed existing problem in Makefile for test_verifier and test_maps tests:
> their dependency on auto-generated header file with a list of all tests wasn't
> recorded explicitly. This patch
From: Vinicius Costa Gomes
This allows a new parameter, flags, to be passed to taprio. Currently, it
only supports enabling the txtime-assist mode. But, we plan to add
different modes for taprio (e.g. hardware offloading) and this parameter
will be useful in enabling those modes.
Signed-off-by:
Add documentation for the latest options, flags and txtime-delay, to the
taprio manpage.
This also adds an example to run tc in txtime offload mode.
Signed-off-by: Vedang Patel
---
man/man8/tc-taprio.8 | 40
1 file changed, 40 insertions(+)
diff --git a
ETF Qdisc currently checks for a socket with SO_TXTIME socket option. If
either is not present, the packet is dropped. In the future commits, we
want other Qdiscs to add packet with launchtime to the ETF Qdisc. Also,
there are some packets (e.g. ICMP packets) which may not have a socket
associated
The type for txtime-delay parameter will change from s32 to u32. So,
make the corresponding change in the ABI file as well.
Signed-off-by: Vedang Patel
---
include/uapi/linux/pkt_sched.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/include/uapi/linux/pkt_sched.h b/include/
Document the newly added option (skip-skb-check) on the etf man-page.
Signed-off-by: Vedang Patel
---
man/man8/tc-etf.8 | 10 ++
1 file changed, 10 insertions(+)
diff --git a/man/man8/tc-etf.8 b/man/man8/tc-etf.8
index 30a12de7d2c7..2e01a591dbaa 100644
--- a/man/man8/tc-etf.8
+++ b/man/
During the review of the iproute2 patches for txtime-assist mode, it was
pointed out that it does not make sense for the txtime-delay parameter to
be negative. So, change the type of the parameter from s32 to u32.
Fixes: 4cfd5779bd6e ("taprio: Add support for txtime-assist mode")
Reported-by: Step
This adds support for setting the txtime_delay parameter which is useful
for the txtime offload mode of taprio.
Signed-off-by: Vedang Patel
---
tc/q_taprio.c | 24 +++-
1 file changed, 23 insertions(+), 1 deletion(-)
diff --git a/tc/q_taprio.c b/tc/q_taprio.c
index 1db2aba6e
From: Benjamin Poirier
Date: Tue, 16 Jul 2019 17:16:55 +0900
> While changing the number of interrupt channels, be2net stops adapter
> operation (including netif_tx_disable()) but it doesn't signal that it
> cannot transmit. This may lead dev_watchdog() to falsely trigger during
> that time.
>
>
On Tue, Jul 16, 2019 at 10:49 AM Alexei Starovoitov
wrote:
>
> On Mon, Jul 15, 2019 at 3:22 PM Daniel Borkmann wrote:
> >
> > On 7/12/19 3:56 PM, Ilya Leoshkevich wrote:
> > > When directories are used as prerequisites in Makefiles, they can cause
> > > a lot of unnecessary rebuilds, because a di
Alexei Starovoitov writes:
> On Tue, Jul 16, 2019 at 09:50:25AM +0100, Jiong Wang wrote:
>>
>> Let me digest a little bit and do some coding, then I will come back. Some
>> issues can only shown up during in-depth coding. I kind of feel handling
>> aux reference in verifier layer is the part th
It's easier to follow the logic if it's structured the same.
There is just slight difference between test_progs/test_maps and
test_verifier. test_verifier's verifier/*.c files are not really compilable
C files (they are more of include headers), so they can't be specified as
explicit dependencies o
e46fc22e60a4 ("selftests/bpf: make directory prerequisites order-only")
exposed existing problem in Makefile for test_verifier and test_maps tests:
their dependency on auto-generated header file with a list of all tests wasn't
recorded explicitly. This patch fixes these issues.
Fixes: 51a0e301a563
On 2019-07-15 16:38, Paul Moore wrote:
> On Mon, Jul 8, 2019 at 1:51 PM Richard Guy Briggs wrote:
> > On 2019-05-29 11:29, Paul Moore wrote:
>
> ...
>
> > > The idea is that only container orchestrators should be able to
> > > set/modify the audit container ID, and since setting the audit
> > >
From: Nishka Dasgupta
Date: Tue, 16 Jul 2019 11:18:43 +0530
> Each iteration of for_each_available_child_of_node puts the previous
> node, but in the case of a return or break from the middle of the loop,
> there is no put, thus causing a memory leak.
What an incredible terribly designed loop ma
On Tue, 9 Jul 2019 15:16:49 +0200
Andrea Claudi wrote:
> Commit ba126dcad20e6 ("ip6tunnel: fix 'ip -6 {show|change} dev
> ' cmds") breaks IPv6 tunnel creation when dev parameter
> is used.
>
> This series revert the original commit, which mistakenly use
> dev for tunnel name, while addressing a
On Tue, Jul 16, 2019 at 6:53 PM Nathan Chancellor
wrote:
>
> On Tue, Jul 16, 2019 at 11:43:05AM -0400, Qian Cai wrote:
> > The commit 6413139dfc64 ("skbuff: increase verbosity when dumping skb
> > data") introduced a few compilation warnings.
> >
> > net/core/skbuff.c:766:32: warning: format speci
On Wed, 10 Jul 2019 10:24:49 +0300
Leon Romanovsky wrote:
> From: Mark Zhang
>
> This patch presents link, id, task name, lqpn, as well as all sub
> counters of a QP counter.
> A QP counter is a dynamically allocated statistic counter that is
> bound with one or more QPs. It has several sub-cou
On Sat, Jul 06, 2019 at 07:55:21PM +0300, Alexey Dobriyan wrote:
> From: "Hallsmark, Per"
>
> proc_net_mkdir() should be used to create stuff under /proc/net,
> so that dentry revalidation kicks in.
>
> See
>
> commit 1fde6f21d90f8ba5da3cb9c54ca991ed72696c43
> proc: fix /proc/net/*
On Tue, Jul 16, 2019 at 1:50 AM Jiong Wang wrote:
>
>
> Andrii Nakryiko writes:
>
> > On Mon, Jul 15, 2019 at 2:21 AM Jiong Wang wrote:
> >>
> >>
> >> Andrii Nakryiko writes:
> >>
> >> > On Thu, Jul 11, 2019 at 4:22 AM Jiong Wang
> >> > wrote:
> >> >>
> >> >>
> >> >> Andrii Nakryiko writes:
> >
On Mon, Jul 15, 2019 at 3:22 PM Daniel Borkmann wrote:
>
> On 7/12/19 3:56 PM, Ilya Leoshkevich wrote:
> > When directories are used as prerequisites in Makefiles, they can cause
> > a lot of unnecessary rebuilds, because a directory is considered changed
> > whenever a file in this directory is a
On Tue, Jul 16, 2019 at 5:59 AM Ilya Leoshkevich wrote:
>
> perf_buffer test fails for exactly the same reason test_attach_probe
> used to fail: different nanosleep syscall kprobe name.
>
> Reuse the test_attach_probe fix.
>
> Fixes: ee5cf82ce04a ("selftests/bpf: test perf buffer API")
> Signed-of
On Tue, Jul 16, 2019 at 10:31 AM Magnus Karlsson
wrote:
>
> On Tue, Jul 16, 2019 at 5:59 AM Andrii Nakryiko wrote:
> >
> > Similar issue was fixed in cdfc7f888c2a ("libbpf: fix GCC8 warning for
> > strncpy") already. This one was missed. Fixing now.
>
> Thanks Andrii.
>
> Acked-by: Magnus Karlsso
On 07/16, Martin Lau wrote:
> On Tue, Jul 09, 2019 at 09:33:21AM -0700, Stanislav Fomichev wrote:
> > On 06/11, Martin KaFai Lau wrote:
> > > The cloned sk should not carry its parent-listener's sk_bpf_storage.
> > > This patch fixes it by setting it back to NULL.
> > Have you thought about some ki
On Tue, Jul 16, 2019 at 5:59 AM Andrii Nakryiko wrote:
>
> Similar issue was fixed in cdfc7f888c2a ("libbpf: fix GCC8 warning for
> strncpy") already. This one was missed. Fixing now.
Thanks Andrii.
Acked-by: Magnus Karlsson
> Cc: Magnus Karlsson
> Signed-off-by: Andrii Nakryiko
> ---
> too
On 7/16/19 7:56 AM, Paul Donohue wrote:
> After establishing an IPsec tunnel to carry the L2TP traffic, the first L2TP
> packet through the IPsec tunnel permanently breaks the associated L2TP
> tunnel. Tearing down the IPsec tunnel does not restore functionality of the
> L2TP tunnel - I have to
On Tue, Jul 16, 2019 at 07:57:04AM -0700, Andrii Nakryiko wrote:
> On Tue, Jul 16, 2019 at 3:56 AM Ilya Leoshkevich wrote:
> >
> > Some setups (e.g. virtual machines) might run with hardware perf events
> > disabled. If this is the case, skip the test_send_signal_nmi test.
> >
> > Add a separate t
On Tue, Jul 16, 2019 at 12:53:53PM +0200, Ilya Leoshkevich wrote:
> BPF_LDX_MEM is used to load the least significant byte of the retrieved
> test_val.index, however, on big-endian machines it ends up retrieving
> the most significant byte.
>
> Change the test to load the whole int in order to mak
On Tue, Jul 16, 2019 at 09:50:25AM +0100, Jiong Wang wrote:
>
> Let me digest a little bit and do some coding, then I will come back. Some
> issues can only shown up during in-depth coding. I kind of feel handling
> aux reference in verifier layer is the part that will still introduce some
> un-cl
2019-07-12, 16:29:48 +0100, Edward Cree wrote:
> On 10/07/2019 23:47, Sabrina Dubroca wrote:
> > 2019-07-10, 16:07:43 +0100, Edward Cree wrote:
> >> On 10/07/2019 14:52, Sabrina Dubroca wrote:
> >>> -static int __netif_receive_skb_core(struct sk_buff *skb, bool pfmemalloc,
> >>> +static int __netif
On 7/11/2019 11:15 PM, Prout, Andrew - LLSC - MITLL wrote:
> I in no way intended to imply that I had confirmed the small SO_SNDBUF was a
> prerequisite to our problem. With my synthetic test, I was looking for a
> concise reproducer and trying things to > stress the system.
I've looked into th
On Tue, Jul 16, 2019 at 3:56 AM Ilya Leoshkevich wrote:
>
> Some setups (e.g. virtual machines) might run with hardware perf events
> disabled. If this is the case, skip the test_send_signal_nmi test.
>
> Add a separate test involving a software perf event. This allows testing
> the perf event pat
On Mon, Jul 15, 2019 at 12:55:48PM -0600, David Ahern wrote:
> As an FYI, gmail thinks your emails are spam.
Ugh. Thanks for letting me know. I'll look into it.
> On 7/15/19 10:18 AM, Paul Donohue wrote:
> > Reverting commit 93531c6743157d7e8c5792f8ed1a57641149d62c (identified by
> > bisection)
perf_buffer test fails for exactly the same reason test_attach_probe
used to fail: different nanosleep syscall kprobe name.
Reuse the test_attach_probe fix.
Fixes: ee5cf82ce04a ("selftests/bpf: test perf buffer API")
Signed-off-by: Ilya Leoshkevich
---
.../testing/selftests/bpf/prog_tests/attac
On Tue, 16 Jul 2019 at 03:15, Pravin Shelar wrote:
>
Hi Pravin,
> On Fri, Jul 5, 2019 at 9:08 AM Taehee Yoo wrote:
> >
> > When a vport is deleted, the maximum headroom size would be changed.
> > If the vport which has the largest headroom is deleted,
> > the new max_headroom would be set.
> >
On Tue, Jul 16, 2019 at 11:19:26AM +0300, Gal Pressman wrote:
> On 10/07/2019 10:24, Leon Romanovsky wrote:
> > +.SH "EXAMPLES"
> > +.PP
> > +rdma statistic show
> > +.RS 4
> > +Shows the state of the default counter of all RDMA devices on the system.
> > +.RE
> > +.PP
> > +rdma statistic show link
test_pkt_md_access is failing on s390, since the associated eBPF prog
returns TC_ACT_SHOT, which in turn happens because loading a part of a
struct __sk_buff field produces an incorrect result.
The problem is that when verifier emits the code to replace partial load
of a field with a full load, a
On Tue, 16 Jul 2019 at 08:59, Eric Dumazet wrote:
> > + return -EINVAL;
> > +
> > + if (sk->sk_protocol != IPPROTO_TCP || sk->sk_state != TCP_LISTEN)
> > + return -EINVAL;
> > +
> > + if (!sock_net(sk)->ipv4.sysctl_tcp_syncookies)
> > + return -EINVAL;
>
On Tue, 16 Jul 2019 at 01:27, Petar Penkov wrote:
>
> From: Petar Penkov
>
> This helper function allows BPF programs to try to generate SYN
> cookies, given a reference to a listener socket. The function works
> from XDP and with an skb context since bpf_skc_lookup_tcp can lookup a
> socket in b
On Tue, 16 Jul 2019 at 01:27, Petar Penkov wrote:
>
> From: Petar Penkov
>
> This patch allows generation of a SYN cookie before an SKB has been
> allocated, as is the case at XDP.
>
> Signed-off-by: Petar Penkov
> ---
> include/net/tcp.h| 11 ++
> net/ipv4/tcp_input.c | 79
This patch adds a 'flags' field to the umem_config and umem_reg structs.
This will allow for more options to be added for configuring umems.
The first use for the flags field is to add a flag for unaligned chunks
mode. These flags can either be user-provided or filled with a default.
Signed-off-b
This patch adds support for the unaligned chunks mode. The addition of the
unaligned chunks option will allow users to run the application with more
relaxed chunk placement in the XDP umem.
Unaligned chunks mode can be used with the '-u' or '--unaligned' command
line options.
Signed-off-by: Kevin
With the addition of the unaligned chunks option, we need to make sure we
handle the offsets accordingly based on the mode we are currently running
in. This patch modifies the driver to appropriately mask the address for
each case.
Signed-off-by: Kevin Laatz
---
drivers/net/ethernet/intel/ixgbe/
This patch adds buffer recycling support for unaligned buffers. Since we
don't mask the addr to 2k at umem_reg in unaligned mode, we need to make
sure we give back the correct (original) addr to the fill queue. We achieve
this using the new descriptor format and associated masks. The new format
use
The addition of unaligned chunks mode, the documentation needs to be
updated to indicate that the incoming addr to the fill ring will only be
masked if the user application is run in the aligned chunk mode. This patch
also adds a line to explicitly indicate that the incoming addr will not be
masked
This patch modifies xdpsock to use mmap instead of posix_memalign. With
this change, we can use hugepages when running the application in unaligned
chunks mode. Using hugepages makes it more likely that we have physically
contiguous memory, which supports the unaligned chunk mode better.
Signed-of
Currently, addresses are chunk size aligned. This means, we are very
restricted in terms of where we can place chunk within the umem. For
example, if we have a chunk size of 2k, then our chunks can only be placed
at 0,2k,4k,6k,8k... and so on (ie. every 2k starting from 0).
This patch introduces t
With the addition of the unaligned chunks option, we need to make sure we
handle the offsets accordingly based on the mode we are currently running
in. This patch modifies the driver to appropriately mask the address for
each case.
Signed-off-by: Bruce Richardson
Signed-off-by: Kevin Laatz
---
Currently, the dma, addr and handle are modified when we reuse Rx buffers
in zero-copy mode. However, this is not required as the inputs to the
function are copies, not the original values themselves. As we use the
copies within the function, we can use the original 'old_bi' values
directly without
Currently, the dma, addr and handle are modified when we reuse Rx buffers
in zero-copy mode. However, this is not required as the inputs to the
function are copies, not the original values themselves. As we use the
copies within the function, we can use the original 'obi' values
directly without ha
This patch set adds the ability to use unaligned chunks in the XDP umem.
Currently, all chunk addresses passed to the umem are masked to be chunk
size aligned (default is 2k, max is PAGE_SIZE). This limits where we can
place chunks within the umem as well as limiting the packet sizes that are
supp
Some setups (e.g. virtual machines) might run with hardware perf events
disabled. If this is the case, skip the test_send_signal_nmi test.
Add a separate test involving a software perf event. This allows testing
the perf event path regardless of hardware perf event support.
Signed-off-by: Ilya Le
BPF_LDX_MEM is used to load the least significant byte of the retrieved
test_val.index, however, on big-endian machines it ends up retrieving
the most significant byte.
Change the test to load the whole int in order to make it
endianness-independent.
Signed-off-by: Ilya Leoshkevich
---
v1->v2:
On Tue, Jul 16, 2019 at 06:01:33AM -0400, Michael S. Tsirkin wrote:
> On Tue, Jul 16, 2019 at 11:40:24AM +0200, Stefano Garzarella wrote:
> > On Mon, Jul 15, 2019 at 01:50:28PM -0400, Michael S. Tsirkin wrote:
> > > On Mon, Jul 15, 2019 at 09:44:16AM +0200, Stefano Garzarella wrote:
> > > > On Fri,
On Tue, Jul 16, 2019 at 11:40:24AM +0200, Stefano Garzarella wrote:
> On Mon, Jul 15, 2019 at 01:50:28PM -0400, Michael S. Tsirkin wrote:
> > On Mon, Jul 15, 2019 at 09:44:16AM +0200, Stefano Garzarella wrote:
> > > On Fri, Jul 12, 2019 at 06:14:39PM +0800, Jason Wang wrote:
>
> [...]
>
> > > >
On Mon, Jul 15, 2019 at 01:50:28PM -0400, Michael S. Tsirkin wrote:
> On Mon, Jul 15, 2019 at 09:44:16AM +0200, Stefano Garzarella wrote:
> > On Fri, Jul 12, 2019 at 06:14:39PM +0800, Jason Wang wrote:
[...]
> > >
> > >
> > > I think it's just a branch, for ethernet, go for networking stack.
>
Andrii Nakryiko writes:
> On Mon, Jul 15, 2019 at 2:21 AM Jiong Wang wrote:
>>
>>
>> Andrii Nakryiko writes:
>>
>> > On Thu, Jul 11, 2019 at 4:22 AM Jiong Wang
>> > wrote:
>> >>
>> >>
>> >> Andrii Nakryiko writes:
>> >>
>> >> > On Thu, Jul 4, 2019 at 2:31 PM Jiong Wang
>> >> > wrote:
>> >>
1 - 100 of 104 matches
Mail list logo
