The test covered both nmi and tracepoint perf events.
$ ./test_send_signal_user
test_send_signal (tracepoint): OK
test_send_signal (perf_event): OK
Signed-off-by: Yonghong Song
---
tools/testing/selftests/bpf/Makefile | 5 +-
tools/testing/selftests/bpf/bpf_helpers.h | 2 +
Currently, bpf program can already collect stack traces
when certain events happens (e.g., cache miss counter or
cpu clock counter overflows). These stack traces can be
used for performance analysis. For jitted programs, e.g.,
hhvm (jited php), it is very hard to get the true stack
trace in the bpf
sync bpf uapi header bpf.h to tools directory.
Signed-off-by: Yonghong Song
---
tools/include/uapi/linux/bpf.h | 15 ++-
1 file changed, 14 insertions(+), 1 deletion(-)
diff --git a/tools/include/uapi/linux/bpf.h b/tools/include/uapi/linux/bpf.h
index 72336bac7573..e3e824848335 1006
This patch tries to solve the following specific use case.
Currently, bpf program can already collect stack traces
when certain events happens (e.g., cache miss counter or
cpu clock counter overflows). These stack traces can be
used for performance analysis. For jitted programs, e.g.,
hhvm (jited
--
Greetings
I wonder why you continue neglecting my emails. Please, acknowledge the
receipt of this message in reference to the subject above as I intend
to send to you the details of the project.
Sometimes, try to check your spam box because most of these
correspondences
fall out sometimes
From: "Matthew Wilcox (Oracle)"
One step closer to turning the skb_frag_t into a bio_vec.
Signed-off-by: Matthew Wilcox (Oracle)
---
include/linux/skbuff.h | 12 +---
net/core/skbuff.c | 2 +-
2 files changed, 6 insertions(+), 8 deletions(-)
diff --git a/include/linux/skbuff.h b
From: "Matthew Wilcox (Oracle)"
Add the dependency now, even though we're not using the bio_vec yet.
Signed-off-by: Matthew Wilcox (Oracle)
---
include/linux/skbuff.h | 1 +
1 file changed, 1 insertion(+)
diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index 9c6193a57241..bc416e5
From: "Matthew Wilcox (Oracle)"
To increase commonality between block and net, we are going to replace
the skb_frag_t with the bio_vec. This patch increases the size of
skb_frag_t on 32-bit machines from 8 bytes to 12 bytes. The size is
unchanged on 64-bit machines.
Signed-off-by: Matthew Wilc
From: "Matthew Wilcox (Oracle)"
Match the layout of bio_vec.
Signed-off-by: Matthew Wilcox (Oracle)
---
include/linux/skbuff.h | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/include/linux/skbuff.h b/include/linux/skbuff.h
index 23f05c64aa31..9c6193a57241 100644
--- a/inclu
From: "Matthew Wilcox (Oracle)"
It turns out there's a lot of accessors for the skb_frag, which would
make this conversion really easy if some drivers didn't bypass them.
This is what I've done so far; my laptop's not really beefy enough to
cope with changing skbuff.h too often ;-)
This would be
From: "Matthew Wilcox (Oracle)"
In preparation for renaming skb->page, use the fine accessors which
already exist.
Signed-off-by: Matthew Wilcox (Oracle)
---
drivers/hsi/clients/ssi_protocol.c | 3 ++-
drivers/net/ethernet/cavium/liquidio/lio_main.c| 2 +-
drivers/net/ether
From: Eric Dumazet
Date: Sun, 28 Apr 2019 12:22:25 -0700
> We had many syzbot reports that seem to be caused by use-after-free
> of struct fib6_info.
>
> ip6_dst_destroy(), fib6_drop_pcpu_from() and rt6_remove_exception()
> are writers vs rt->from, and use non consistent synchronization among
>
On Tue, Apr 30, 2019 at 5:46 AM Björn Töpel wrote:
>
> From: Björn Töpel
>
> When unmapping the AF_XDP memory regions used for the rings, an
> invalid address was passed to the munmap() calls. Instead of passing
> the beginning of the memory region, the descriptor region was passed
> to munmap.
>
On Tue, Apr 30, 2019 at 5:46 AM Björn Töpel wrote:
>
> From: Björn Töpel
>
> The bpf_map_update_elem() function, when used on an XSKMAP, will fail
> if not a valid AF_XDP socket is passed as value. Therefore, this is
> function cannot be used to clear the XSKMAP. Instead, the
> bpf_map_delete_ele
From: Heiner Kallweit
Date: Sun, 28 Apr 2019 11:09:36 +0200
> This series aims at improving and simplifying the eri functions.
> No functional change intended.
Series applied, thanks.
The cited commit broke the offsets of hca cap struct, fix it.
While at it, cleanup a white space introduced by the same commit.
Fixes: b169e64a2444 ("net/mlx5: Geneve, Add flow table capabilities for Geneve
decap with TLV options")
Reported-by: Qian Cai
Cc: Yevgeny Kliteynik
Signed-off-by: Saee
From: Andrew Lunn
Date: Sun, 28 Apr 2019 02:56:20 +0200
> This patchset builds upon the previous patches to mv88e6060. It adds
> support for probing the switch as an MDIO device and then removes the
> legacy probe method. Since this is the last device supporting legacy
> probe, this allows legacy
From: Vladimir Oltean
Date: Sun, 28 Apr 2019 21:45:41 +0300
> In preparation of submitting the NXP SJA1105 driver, the Broadcom b53
> and Mediatek mt7530 drivers have been found to apply some VLAN
> workarounds that are needed in the new driver as well.
>
> Therefore this patchset is mostly simp
If we try to call strp_done on a parser that has never been
initialized, because the sockmap user is only using TX side for
example we get the following error.
[ 883.422081] WARNING: CPU: 1 PID: 208 at kernel/workqueue.c:3030
__flush_work+0x1ca/0x1e0
...
[ 883.422095] Workqueue: events s
It is possible (via shutdown()) for TCP socks to go through TCP_CLOSE
state via tcp_disconnect() without calling into close callback. This
would allow a kTLS enabled socket to exist outside of ESTABLISHED
state which is not supported.
Solve this the same way we solved the sock{map|hash} case by ad
In tcp bpf remove we free the cork list and purge the ingress msg
list. However we do this before the ref count reaches zero so it
could be possible some other access is in progress. In this case
(tcp close and/or tcp_unhash) we happen to also hold the sock
lock so no path exists but lets fix it ot
When converting a skb to msg->sg we forget to set the size after the
latest ktls/tls code conversion. This patch can be reached by doing
a redir into ingress path from BPF skb sock recv hook. Then trying to
read the size fails.
Fix this by setting the size.
Fixes: 604326b41a6fb ("bpf, sockmap: co
Series of fixes for sockmap and ktls, see patches for descriptions.
v2: fix build issue for CONFIG_TLS_DEVICE and fixup couple comments
from Jakub
v3: fix issue where release could call unhash resulting in a use after
free. Now we detach the ulp pointer before calling into destroy
or
Multi-line output in ss makes it difficult to search for things with
grep. This new option will make it easier to find sockets matching
certain criteria with simple grep commands.
Example without option:
$ ss -emoitn
State Recv-Q Send-Q Local Address:Port Peer Address:Port
ESTAB
On 2019-04-30 10:12, Uwe Kleine-König wrote:
> [Adding the mvebu guys and netdev to Cc]
>
> Hello,
>
> On Thu, Apr 25, 2019 at 09:17:32PM +0200, Aurelien Jarno wrote:
> > On 2019-04-25 14:50, Aurelien Jarno wrote:
> > > On 2019-04-23 22:16, Aurelien Jarno wrote:
> > > > Source: linux
> > > > Vers
Allow an interrupt number to be passed in the platform data. The
driver will then use it if not zero, otherwise it will poll for
interrupts.
Signed-off-by: Andrew Lunn
---
drivers/net/dsa/mv88e6xxx/chip.c| 13 +
include/linux/platform_data/mv88e6xxx.h | 1 +
2 files changed,
If the NO_CPU strap is set, the switch starts in 'dumb hub' mode, with
all ports enable. Ports which are then actively used are reconfigured
as required when the driver starts. However unused ports are left
alone. Change this to disable them, and turn off any SERDES
interface. This could save some
Save some power by disabling ports. The first patch fully disables a
port when it is runtime disabled. The second disables any ports which
are not used at all.
Depending on configuration strapping, this can lower the temperature
of an idle switch a few degrees.
Andrew Lunn (2):
net: dsa: mv88e6
When requested to disable a port, set the port STP state to disabled.
This fully disables the port and should save some power.
Signed-off-by: Andrew Lunn
---
drivers/net/dsa/mv88e6xxx/chip.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/drivers/net/dsa/mv88e6xxx/chip.c b/drivers/net/dsa
From: Bodong Wang
Some functions issue vport commands and access vport array using
vport_index/vport_num interchangeably which is OK for VFs vports.
However, this creates potential bug if those vports are not VFs
(E.g, uplink, sf) where their vport_index don't equal to vport_num.
Prepare code to
From: Bodong Wang
mlx5_query_nic_vport_vlans() is not used anymore. Hence remove it.
This patch doesn't change any functionality.
Signed-off-by: Bodong Wang
Reviewed-by: Parav Pandit
Signed-off-by: Saeed Mahameed
---
.../net/ethernet/mellanox/mlx5/core/vport.c | 61 ---
inc
From: Parav Pandit
Several functions need to access mlx5_vport and vport_num.
When these functions are called, caller already has mlx5_vport*
available.
Hence pass such mlx5_vport pointer.
This is preparation patch to add error checks to
mlx5_eswitch_get_vport() and to return error status.
By do
From: Bodong Wang
The check of legal vport is to ensure the vport number falls between
0 and total number of vports. Along with the introduction of uplink
rep, enabled vports are not consecutive any more.
Therefore, rely on the eswitch vport getter function to check if it's
a valid vport.
As the
From: Erez Alfasi
Added max EEPROM length defines for ethtool usage:
#define ETH_MODULE_SFF_8636_MAX_LEN 640
#define ETH_MODULE_SFF_8436_MAX_LEN 640
These definitions are exists in ethtool and used to
determine the EEPROM data length when reading high
pages as well.
For example, SFF-8
From: Bodong Wang
When the state of rep was introduced, it was also designed to prevent
duplicate unloading of the same rep. Considering the following two
flows when an eswitch manager is at switchdev mode with n VF reps loaded.
+--+---
From: Parav Pandit
Currently mlx5_esw_for_each_vf_vport iterates over mlx5_vport entries in
eswitch.c
Same macro in eswitch_offloads.c iterates over vport number in
eswitch_offloads.c
Instead of duplicate macro names, to avoid confusion and to reuse the
same macro in both files, move it to eswit
From: Masahiro Yamada
CFLAGS_tracepoint.o specifies CFLAGS for compiling tracepoint.c but
it does not exist under drivers/net/ethernet/mellanox/mlx5/core/.
CFLAGS_tracepoint.o is unused.
Signed-off-by: Masahiro Yamada
Signed-off-by: Saeed Mahameed
---
drivers/net/ethernet/mellanox/mlx5/core/
From: Erez Alfasi
Add the support to read additional EEPROM information from high pages.
Information for modules such as SFF-8436 and SFF-8636:
1) Application select table
2) User writable EEPROM
3) Thresholds and alarms
Signed-off-by: Erez Alfasi
Signed-off-by: Saeed Mahameed
---
.../ethe
From: Eli Britstein
Current ConnectX HW is unable to perform VLAN pop in TX path and VLAN
push on RX path. As a workaround, untagged packets are tagged with
VID 0x000 allowing pop/push actions to be exchanged with VLAN rewrite
actions.
Use the ingress ACL table, preceding the FDB, to push VLAN 0x
From: Vlad Buslov
With unlocked TC it is possible to have spurious deletes and inserts of
same filter. TC layer needs drivers to always return error when flow
insertion failed in order to correctly calculate "in_hw_count" for each
filter. Fix mlx5e_configure_flower() to return -EEXIST when TC tri
From: Maxim Mikityanskiy
The same code that returns XDP frames and releases pages is used both in
mlx5e_poll_xdpsq_cq and mlx5e_free_xdpsq_descs. Create a function that
cleans up an MPWQE.
Signed-off-by: Maxim Mikityanskiy
Signed-off-by: Saeed Mahameed
---
.../net/ethernet/mellanox/mlx5/core/
From: Eli Britstein
Current ConnectX HW is unable to perform VLAN pop in TX path and VLAN
push on RX path. To workaround that limitation untagged packets are
tagged with VLAN ID 0x000 (priority tag) and pop/push actions are
replaced by VLAN re-write actions (which are supported by the HW).
Replac
From: Tariq Toukan
Many TIR context settings are common to different TIR types,
take them into a common function.
Signed-off-by: Tariq Toukan
Reviewed-by: Aya Levin
Signed-off-by: Saeed Mahameed
---
.../net/ethernet/mellanox/mlx5/core/en_main.c | 49 ---
1 file changed, 21 in
From: Tariq Toukan
Hardware requires that all TIRs that steer traffic to the same RQ
should share identical tunneled_offload_en value.
For that, the tunneled_offload_en bit should be set/unset (according to
the HW capability) for all TIRs', not only the ones dedicated for
tunneled (inner) traffic
Hi Dave,
This series provides misc updates to mlx5 driver.
There is one patch of this series that is touching outside mlx5 driver:
ethtool.h: Add SFF-8436 and SFF-8636 max EEPROM length definitions
Added max EEPROM length defines for ethtool usage:
#define ETH_MODULE_SFF_8636_MAX_LEN 640
On 4/30/19 12:41 PM, David Ahern wrote:
On 4/30/19 12:55 PM, Josh Hunt wrote:
Actually, David can you clarify what you meant by "use 'oneline' as the
long option without the '-'."?
for your patch:
1,$s/one-line/oneline/
ip has -oneline which is most likely used as 'ip -o'. having ss with
--on
On 4/30/19 12:55 PM, Josh Hunt wrote:
> Actually, David can you clarify what you meant by "use 'oneline' as the
> long option without the '-'."?
for your patch:
1,$s/one-line/oneline/
ip has -oneline which is most likely used as 'ip -o'. having ss with
--one-line vs --oneline is at least consiste
On Tue, 30 Apr 2019 08:31:53 -0700, Y Song wrote:
> On Tue, Apr 30, 2019 at 2:34 AM Quentin Monnet
> wrote:
> >
> > Hi Yonghong,
> >
> > 2019-04-29 16:32 UTC-0700 ~ Y Song
> > > On Mon, Apr 29, 2019 at 2:53 AM Quentin Monnet
> > > wrote:
> > >>
> > >> libbpf has three levels of priority for
On 4/30/19 11:31 AM, Josh Hunt wrote:
On 4/30/19 11:30 AM, David Ahern wrote:
On 4/25/19 3:21 PM, Josh Hunt wrote:
@@ -4877,6 +4903,7 @@ static void _usage(FILE *dest)
"\n"
" -K, --kill forcibly close sockets, display what was
closed\n"
" -H, --no-header Suppress header
Although devlink health report does a nice job on reporting TX
timeout and other NIC errors, unfortunately it requires drivers
to support it but currently only mlx5 has implemented it.
Before other drivers could catch up, it is useful to have a
generic tracepoint to monitor this kind of TX timeout.
On 4/30/19 11:30 AM, David Ahern wrote:
On 4/25/19 3:21 PM, Josh Hunt wrote:
@@ -4877,6 +4903,7 @@ static void _usage(FILE *dest)
"\n"
" -K, --kill forcibly close sockets, display what was closed\n"
" -H, --no-header Suppress header line\n"
+" -O, --one-line socket'
On 4/25/19 3:21 PM, Josh Hunt wrote:
> @@ -4877,6 +4903,7 @@ static void _usage(FILE *dest)
> "\n"
> " -K, --kill forcibly close sockets, display what was closed\n"
> " -H, --no-header Suppress header line\n"
> +" -O, --one-line socket's data printed on a single line\n"
>
On Tue, Apr 30, 2019 at 10:45 AM Martin KaFai Lau wrote:
>
> It is a followup after the fix in
> commit 9c69a1320515 ("route: Avoid crash from dereferencing NULL rt->from")
>
> rt6_do_redirect():
> 1. NULL checking is needed on rt->from because a parallel
>fib6_info delete could happen that se
On 2019-04-30 20:11, Maxim Mikityanskiy wrote:
I'm going to respin this series, adding the mlx5 patches and addressing
the other comments. If you feel like there are more things to discuss
regarding this patch, please move on to the v2.
Very cool, thanks for working on this Maxim!
It's Labor D
On 4/30/19 2:42 AM, Ido Schimmel wrote:
> From: Ido Schimmel
>
> With current number of spaces the output is mangled if the shared buffer
> is congested.
>
...
>
> v2:
> * Increase number of spaces to make the change more future-proof
>
> Signed-off-by: Ido Schimmel
> Reported-by: Alex Kush
The PCI API for DMA is deprecated, and PCI_DMA_TODEVICE is just defined
to DMA_TO_DEVICE for backward compatibility. Just use DMA_TO_DEVICE.
Signed-off-by: Maxim Mikityanskiy
Reviewed-by: Tariq Toukan
Acked-by: Saeed Mahameed
---
drivers/net/ethernet/mellanox/mlx5/core/en/xdp.c | 2 +-
1 file
Put the XDP SQ that is used for XDP_TX into the channel. It used to be a
part of the RQ, but with introduction of AF_XDP there will be one more
RQ that could share the same XDP SQ. This patch is a preparation for
that change.
Separate XDP_TX statistics per RQ were implemented in one of the previou
Prepare to creation of the XSK RQ, which will require posting UMRs, too.
The same ICO SQ will be used for both RQs and also to trigger interrupts
by posting NOPs. UMR WQEs can't be reused any more. Optimization
introduced in commit ab966d7e4ff98 ("net/mlx5e: RX, Recycle buffer of
UMR WQEs") is reve
structs mlx5e_{rq,sq,cq,channel}_param are going to be used in the
upcoming XSK RX and TX patches. Move them to a header file to make
them accessible from other C files.
Signed-off-by: Maxim Mikityanskiy
Reviewed-by: Tariq Toukan
Acked-by: Saeed Mahameed
---
.../ethernet/mellanox/mlx5/core/en/
Use the existing mlx5e_get_linear_rq_headroom function to calculate the
headroom for mlx5e_xdp_max_mtu. This function takes the XSK headroom
into consideration, which will be used in the following patches.
Signed-off-by: Maxim Mikityanskiy
Reviewed-by: Tariq Toukan
Acked-by: Saeed Mahameed
---
Create new functions mlx5e_{open,close}_queues to encapsulate opening
and closing RQs and SQs, and call the new functions from
mlx5e_{open,close}_channel. It simplifies the existing functions a bit
and prepares them for the upcoming AF_XDP changes.
Signed-off-by: Maxim Mikityanskiy
Reviewed-by: T
Additional conditions introduced:
- XSK implies XDP.
- Headroom includes the XSK headroom if it exists.
- No space is reserved for struct shared_skb_info in XSK mode.
- Fragment size smaller than the XSK chunk size is not allowed.
A new auxiliary function mlx5e_get_linear_rq_headroom with the sup
Currently, struct mlx5e_xdp_info has some issues that have to be cleaned
up before the upcoming AF_XDP support makes things too complicated and
messy. This structure is used both when sending the packet and on
completion. Moreover, the cleanup procedure on completion depends on the
origin of the pa
When an XDP program returns XDP_TX, and the RQ is XSK-enabled, it
requires careful handling, because convert_to_xdp_frame creates a new
page and copies the data there, while our driver expects the xdp_frame
to point to the same memory as the xdp_buff. Handle this case
separately: map the page, and
Some drivers want to access the data transmitted in order to implement
acceleration features of the NICs. It is also useful in AF_XDP TX flow.
Change the xsk_umem_consume_tx API to return the whole xdp_desc, that
contains the data pointer, length and DMA address, instead of only the
latter two. Ad
The typical XDP memory scheme is one packet per page. Change the AF_XDP
frame size in libbpf to 4096, which is the page size on x86, to allow
libbpf to be used with the drivers with the packet-per-page scheme.
Add a command line option -f to xdpsock to allow to specify a custom
frame size.
Signed
Currently, the drivers that implement AF_XDP zero-copy support (e.g.,
i40e) switch the channel into a different mode when an XSK is opened. It
causes some issues that have to be taken into account. For example, RSS
needs to be reconfigured to skip the XSK-enabled channels, or the XDP
program should
Query XDP_OPTIONS in libbpf to determine if the zero-copy mode is active
or not.
Signed-off-by: Maxim Mikityanskiy
Reviewed-by: Tariq Toukan
Acked-by: Saeed Mahameed
---
tools/lib/bpf/xsk.c | 11 +++
1 file changed, 11 insertions(+)
diff --git a/tools/lib/bpf/xsk.c b/tools/lib/bpf/xsk
Make it possible for the application to determine whether the AF_XDP
socket is running in zero-copy mode. To achieve this, add a new
getsockopt option XDP_OPTIONS that returns flags. The only flag
supported for now is the zero-copy mode indicator.
Signed-off-by: Maxim Mikityanskiy
Reviewed-by: Ta
This series contains improvements to the AF_XDP kernel infrastructure
and AF_XDP support in mlx5e. The infrastructure improvements are
required for mlx5e, but also some of them benefit to all drivers, and
some can be useful for other drivers that want to implement AF_XDP.
The performance testing w
Add a function that checks whether the Fill Ring has the specified
amount of descriptors available. It will be useful for mlx5e that wants
to check in advance, whether it can allocate a bulk of RX descriptors,
to get the best performance.
Signed-off-by: Maxim Mikityanskiy
Reviewed-by: Tariq Touka
On 2019-04-26 23:26, Björn Töpel wrote:
> On 2019-04-26 13:42, Maxim Mikityanskiy wrote:
>> Currently, the drivers that implement AF_XDP zero-copy support (e.g.,
>> i40e) switch the channel into a different mode when an XSK is opened. It
>> causes some issues that have to be taken into account. For
On 2019-04-26 22:11, Jakub Kicinski wrote:
> On Fri, 26 Apr 2019 11:42:37 +, Maxim Mikityanskiy wrote:
>> Currently, the drivers that implement AF_XDP zero-copy support (e.g.,
>> i40e) switch the channel into a different mode when an XSK is opened. It
>> causes some issues that have to be taken
On 4/29/19 8:37 PM, Hangbin Liu wrote:
> An other issue is The IPv4 rule 'from iif' check test failed while IPv6
> passed. I haven't found out the reason yet.
>
> # ip -netns testns rule add from 192.51.100.3 iif dummy0 table 100
> # ip -netns testns route get 192.51.100.2 from 192.51.100.3 iif du
It is a followup after the fix in
commit 9c69a1320515 ("route: Avoid crash from dereferencing NULL rt->from")
rt6_do_redirect():
1. NULL checking is needed on rt->from because a parallel
fib6_info delete could happen that sets rt->from to NULL.
(e.g. rt6_remove_exception() and fib6_drop_pcpu
sch_plug can be used to perform functional qdisc unit tests
controlling explicitly the queuing behaviour from user-space.
Plug support lacks since its introduction in 2012. This change
introduces basic support, to control the tc status.
v1 -> v2:
- use the SPDX identifier
Signed-off-by: Paolo A
From: Eric Dumazet
Date: Tue, 30 Apr 2019 06:27:58 -0700
> Before taking a refcount on a rcu protected structure,
> we need to make sure the refcount is not zero.
>
> syzbot reported :
...
> Fixes: 54652eb12c1b ("l2tp: hold tunnel while looking up sessions in
> l2tp_netlink")
> Signed-off-by:
On 30 Apr 2019, at 5:45, Björn Töpel wrote:
> William found two bugs, when doing socket teardown within the same
> process.
>
> The first issue was an invalid munmap call, and the second one was an
> invalid XSKMAP cleanup. Both resulted in that the process kept
> references to the socket, whic
On Tue, Apr 30, 2019 at 2:34 AM Quentin Monnet
wrote:
>
> Hi Yonghong,
>
> 2019-04-29 16:32 UTC-0700 ~ Y Song
> > On Mon, Apr 29, 2019 at 2:53 AM Quentin Monnet
> > wrote:
> >>
> >> libbpf has three levels of priority for output: warn, info, debug. By
> >> default, debug output is not printed to
From: Dan Carpenter
Date: Tue, 30 Apr 2019 13:44:19 +0300
> The "fs->location" is a u32 that comes from the user in ethtool_set_rxnfc().
> We can't pass unclamped values to test_bit() or it results in an out of
> bounds access beyond the end of the bitmap.
>
> Fixes: 7318166cacad ("net: dsa: bcm
On Mon, 2019-04-29 at 10:16 -0700, Stephen Hemminger wrote:
> On Fri, 26 Apr 2019 10:47:52 +0200
> Paolo Abeni wrote:
> > The problem here is that the sch_plug qdisc does not implement the
> > dump() qdisc_op, so this callback has nothing to dump.
> >
> > Must I patch sch_plug first?
> >
> > Tha
On Tue, Apr 30, 2019 at 07:45:48AM -0700, David Ahern wrote:
> From: David Ahern
>
> While the cached routes, nh_pcpu_rth_output and nh_rth_input, are IPv4
> specific, a later patch wants to make them accessible for IPv6 nexthops
> with IPv4 routes using a fib6_nh. Move the cached routes from fib
From: David Ahern
Now that the cached routes are in fib_nh_common, pass it to
rt_cache_route and simplify its callers. For rt_set_nexthop,
the tclassid becomes the last user of fib_nh so move the
container_of under the #ifdef CONFIG_IP_ROUTE_CLASSID.
Signed-off-by: David Ahern
Reviewed-by: Ido
From: David Ahern
Similar to the cached routes, make IPv4 exceptions accessible when
using an IPv6 nexthop struct with IPv4 routes. Simplify the exception
functions by passing in fib_nh_common since that is all it needs,
and then cleanup the call sites that have extraneous fib_nh conversions.
As
From: David Ahern
This series moves IPv4 pcpu cached routes from fib_nh to fib_nh_common
to make the caches available for IPv6 nexthops (fib6_nh) with IPv4
routes. This allows a fib6_nh struct to be used with both IPv4 and
and IPv6 routes.
v4
- fixed memleak if encap_type is not set as noticed b
From: David Ahern
While the cached routes, nh_pcpu_rth_output and nh_rth_input, are IPv4
specific, a later patch wants to make them accessible for IPv6 nexthops
with IPv4 routes using a fib6_nh. Move the cached routes from fib_nh to
fib_nh_common and update references.
Initialization of the cach
On 4/30/19 12:40 AM, Ido Schimmel wrote:
> On Mon, Apr 29, 2019 at 09:16:17AM -0700, David Ahern wrote:
>> /* Release a nexthop info record */
>> @@ -491,9 +491,15 @@ int fib_nh_common_init(struct fib_nh_common *nhc,
>> struct nlattr *encap,
>> u16 encap_type, void *cfg, gfp_t
From: Johannes Berg
Date: Tue, 30 Apr 2019 08:58:10 +0200
> If you prefer to have the safeguard in net even if it shouldn't be
> needed now, let me know and I'll make a version that applies there, but
> note that will invariably cause conflicts with all the other changes in
> lib/nlattr.c.
No, t
From: Steffen Klassert
Date: Tue, 30 Apr 2019 08:37:09 +0200
> 1) A lot of work to remove indirections from the xfrm code.
>From Florian Westphal.
>
> 2) Support ESP offload in combination with gso partial.
>From Boris Pismenny.
>
> 3) Remove some duplicated code from vti4.
>From Je
Before taking a refcount on a rcu protected structure,
we need to make sure the refcount is not zero.
syzbot reported :
refcount_t: increment on 0; use-after-free.
WARNING: CPU: 1 PID: 23533 at lib/refcount.c:156 refcount_inc_checked
lib/refcount.c:156 [inline]
WARNING: CPU: 1 PID: 23533 at lib/
Hi Daniel,
I'm running selftests/bpf/test_tag and it's taking roughly half an hour
to complete, is that expected?
I don't really grok what the test is doing TBH, but it does appear to be
doing it 5 times :)
for (i = 0; i < 5; i++) {
do_test(&tests, 2, -1, bpf_gen_imm_
From: Steffen Klassert
Date: Tue, 30 Apr 2019 07:30:18 +0200
...
> Please pull or let me know if there are problems.
Pulled, thanks Steffen.
From: Björn Töpel
The bpf_map_update_elem() function, when used on an XSKMAP, will fail
if not a valid AF_XDP socket is passed as value. Therefore, this is
function cannot be used to clear the XSKMAP. Instead, the
bpf_map_delete_elem() function should be used for that.
This patch also simplifies
From: Björn Töpel
When unmapping the AF_XDP memory regions used for the rings, an
invalid address was passed to the munmap() calls. Instead of passing
the beginning of the memory region, the descriptor region was passed
to munmap.
When the userspace application tried to tear down an AF_XDP socke
William found two bugs, when doing socket teardown within the same
process.
The first issue was an invalid munmap call, and the second one was an
invalid XSKMAP cleanup. Both resulted in that the process kept
references to the socket, which was not correctly cleaned up. When a
new socket was creat
On Fri, Apr 26, 2019 at 08:41:03AM -0700, Peter Oskolkov wrote:
> This is a backport of a 5.1rc patchset:
> https://patchwork.ozlabs.org/cover/1029418/
>
> Which was backported into 4.19:
> https://patchwork.ozlabs.org/cover/1081619/
>
> and into 4.14:
> https://patchwork.ozlabs.org/cover/1
On 4/30/2019 3:54 AM, Jakub Kicinski wrote:
> On Mon, 29 Apr 2019 17:17:39 +0300, Aya Levin wrote:
>> In order to offline translate the raw memory into a human readable
>> format, the user can use some out-of-kernel scripts which receives as an
>> input the following:
>> - Object raw memory
>> -
On 4/29/2019 9:32 PM, Saeed Mahameed wrote:
> On Mon, 2019-04-29 at 17:17 +0300, Aya Levin wrote:
>> TX reporter reports an error on two scenarios:
>> - TX timeout on a specific tx queue
>> - TX completion error on a specific send queue
>> Prior to this patch, no dump data was supported by the tx
The "fs->location" is a u32 that comes from the user in ethtool_set_rxnfc().
We can't pass unclamped values to test_bit() or it results in an out of
bounds access beyond the end of the bitmap.
Fixes: 7318166cacad ("net: dsa: bcm_sf2: Add support for ethtool::rxnfc")
Signed-off-by: Dan Carpenter
-
From: Josh Elsasser
> Sent: 29 April 2019 21:02
> On Apr 29, 2019, at 12:16 PM, Jeff Kirsher
> wrote:
>
> > From: Björn Töpel
> >
> > GCC will generate jump tables for switch-statements with more than 5
> > case statements. An entry into the jump table is an indirect call,
> > which means that
1 - 100 of 106 matches
Mail list logo
