hwframe+0x76/0x7e
Fixes: c0cfa2d8a788 ("vsock: add multi-transports support")
Reviewed-by: Stefano Garzarella
Signed-off-by: Michal Luczaj
---
net/vmw_vsock/af_vsock.c | 8 ++--
1 file changed, 6 insertions(+), 2 deletions(-)
diff --git a/net/vmw_vsock/af_vsock.c b/net
)
Reviewed-by: Stefano Garzarella
Reviewed-by: Luigi Leonardi
Signed-off-by: Michal Luczaj
---
net/vmw_vsock/af_vsock.c | 5 +
1 file changed, 5 insertions(+)
diff --git a/net/vmw_vsock/af_vsock.c b/net/vmw_vsock/af_vsock.c
index
cfe18bc8fdbe7ced073c6b3644d635
Distill timeout-guarded vsock_connect_fd(). Adapt callers.
Suggested-by: Stefano Garzarella
Reviewed-by: Stefano Garzarella
Signed-off-by: Michal Luczaj
---
tools/testing/vsock/util.c | 45 +
tools/testing/vsock/util.h | 1 +
2 files changed, 18
Deliberately fail a connect() attempt; expect error. Then verify that
subsequent attempt (using the same socket) can still succeed, rather than
fail outright.
Reviewed-by: Stefano Garzarella
Reviewed-by: Luigi Leonardi
Signed-off-by: Michal Luczaj
---
tools/testing/vsock/vsock_test.c | 47
Fail the autobind, then trigger a transport reassign. Socket might get
unbound from unbound_sockets, which then leads to a reference count
underflow.
Reviewed-by: Stefano Garzarella
Signed-off-by: Michal Luczaj
---
tools/testing/vsock/vsock_test.c | 58
Add a helper for socket()+bind(). Adapt callers.
Reviewed-by: Stefano Garzarella
Reviewed-by: Luigi Leonardi
Signed-off-by: Michal Luczaj
---
tools/testing/vsock/util.c | 57 +---
tools/testing/vsock/util.h | 1 +
tools/testing/vsock
opportunity to comment vsock_bind() (patch 3/6), and I've
kept your Reviewed-by. Is this okay?
Signed-off-by: Michal Luczaj
---
Changes in v3:
- Rebase
- Comment vsock_bind() (Luigi)
- Collect Reviewed-by (Stefano, Luigi)
- Link to v2:
https://lore.kernel.org/r/20250121-vsock-transport-vs-autobind
On 12/20/24 11:49, Stefano Garzarella wrote:
> ...
> Note that non-NULL -> NULL should only occur before a connection is
> established, so before any data is passed. Is this a problem for BPF?
Please take a look at vsock_bpf_update_proto(). The condition is to have a
transport assigned. BPF assum
On 12/19/24 16:12, Stefano Garzarella wrote:
> On Thu, 19 Dec 2024 at 16:05, Michal Luczaj wrote:
>>
>> On 12/19/24 15:48, Stefano Garzarella wrote:
>>> On Thu, 19 Dec 2024 at 15:36, Michal Luczaj wrote:
>>>>
>>>> On 12/19/24 09:19, Stefano Ga
On 12/19/24 15:48, Stefano Garzarella wrote:
> On Thu, 19 Dec 2024 at 15:36, Michal Luczaj wrote:
>>
>> On 12/19/24 09:19, Stefano Garzarella wrote:
>>> ...
>>> I think the best thing though is to better understand how to handle
>>> deassign, rather t
On 12/19/24 09:19, Stefano Garzarella wrote:
> ...
> I think the best thing though is to better understand how to handle
> deassign, rather than checking everywhere that it's not null, also
> because in some cases (like the one in virtio-vsock), it's also
> important that the transport is the sa
On 12/18/24 16:51, Hyunwoo Kim wrote:
> On Wed, Dec 18, 2024 at 04:31:03PM +0100, Stefano Garzarella wrote:
>> On Wed, Dec 18, 2024 at 03:40:40PM +0100, Stefano Garzarella wrote:
>>> On Wed, Dec 18, 2024 at 09:19:08AM -0500, Hyunwoo Kim wrote:
At least for vsock_loopback.c, this change doesn’t
On 11/19/24 11:31, Stefano Garzarella wrote:
> Hi Michal,
>
> On Thu, Nov 07, 2024 at 09:46:11PM +0100, Michal Luczaj wrote:
>> Short series fixing some memory leaks that I've stumbled upon while
>> toying
>> with the selftests.
>
> Are these tests alread
On 11/7/24 11:22, Stefano Garzarella wrote:
> On Wed, Nov 06, 2024 at 06:51:21PM +0100, Michal Luczaj wrote:
>> Macro vsock_connected_sockets_vsk() has been unused since its introduction.
>> Instead of removing it, utilise it in vsock_insert_connected() where it's
>>
Short series fixing some memory leaks that I've stumbled upon while toying
with the selftests.
Signed-off-by: Michal Luczaj
---
Changes in v2:
- Remove the refactoring patch from the series [Stefano]
- PATCH 2: Drop "virtio" from the commit title [Stefano]
- Collect Reviewed-by [
On 11/7/24 11:17, Stefano Garzarella wrote:
> On Wed, Nov 06, 2024 at 06:51:19PM +0100, Michal Luczaj wrote:
>> Kernel queues MSG_ZEROCOPY completion notifications on the error queue.
>> Where they remain, until explicitly recv()ed. To prevent memory leaks,
>> clean up the que
: Stefano Garzarella
Signed-off-by: Michal Luczaj
---
net/vmw_vsock/virtio_transport_common.c | 8
1 file changed, 8 insertions(+)
diff --git a/net/vmw_vsock/virtio_transport_common.c
b/net/vmw_vsock/virtio_transport_common.c
index
ccbd2bc0d2109aea4f19
Add a missing kfree_skb() to prevent memory leaks.
Fixes: 581512a6dc93 ("vsock/virtio: MSG_ZEROCOPY flag support")
Reviewed-by: Stefano Garzarella
Signed-off-by: Michal Luczaj
---
net/vmw_vsock/virtio_transport_common.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/net
+0x47/0x80
[] do_syscall_64+0x93/0x180
[] entry_SYSCALL_64_after_hwframe+0x76/0x7e
Fixes: 581512a6dc93 ("vsock/virtio: MSG_ZEROCOPY flag support")
Signed-off-by: Michal Luczaj
---
net/vmw_vsock/af_vsock.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/net/vmw_vsock/af_vsock.c b/net/vmw_vs
+0x47/0x80
[] do_syscall_64+0x93/0x180
[] entry_SYSCALL_64_after_hwframe+0x76/0x7e
Fixes: 581512a6dc93 ("vsock/virtio: MSG_ZEROCOPY flag support")
Signed-off-by: Michal Luczaj
---
net/vmw_vsock/af_vsock.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/net/vmw_vsock/af_vsock.c b/net/vmw_vs
Macro vsock_connected_sockets_vsk() has been unused since its introduction.
Instead of removing it, utilise it in vsock_insert_connected() where it's
been open-coded.
No functional change intended.
Fixes: d021c344051a ("VSOCK: Introduce VM Sockets")
Signed-off-by: Michal L
Add a missing kfree_skb() to prevent memory leaks.
Fixes: 581512a6dc93 ("vsock/virtio: MSG_ZEROCOPY flag support")
Signed-off-by: Michal Luczaj
---
net/vmw_vsock/virtio_transport_common.c | 1 +
1 file changed, 1 insertion(+)
diff --git a/net/vmw_vsock/virtio_transport_commo
x4dc/0x9a0
[] vsock_loopback_work+0xfd/0x140
[] process_one_work+0x20c/0x570
[] worker_thread+0x1bf/0x3a0
[] kthread+0xdd/0x110
[] ret_from_fork+0x2d/0x50
[] ret_from_fork_asm+0x1a/0x30
Fixes: 3fe356d58efa ("vsock/virtio: discard packets only when socket is really
closed")
Signed-o
Short series fixing some memory leaks that I've stumbled upon while toying
with the selftests.
The last patch is a refactoring.
Signed-off-by: Michal Luczaj
---
Michal Luczaj (4):
virtio/vsock: Fix accept_queue memory leak
virtio/vsock: Fix sk_error_queue memory leak
v
or the XDP Socket
> Zerocopy.
>
>
After taking a look at this series I haven't found adding
NETDEV_XDP_ACT_XSK_ZEROCOPY flag to netdev->xdp_features.
Is it intentional or just an oversight?
Thanks,
Michal
On Tue, Apr 20, 2021 at 01:55:03AM +0200, Michal Kubecek wrote:
> On Mon, Apr 19, 2021 at 05:29:46PM +0200, Michal Kubecek wrote:
> >
> > As pointed out in the discussion on v3, this patch may result in
> > significantly higher CPU consumption with multiple threads competi
On Tue 20-04-21 15:57:08, Michal Hocko wrote:
[...]
> Usual memory consumption is usually something like LRU pages + Slab
> memory + kernel stack + vmalloc used + pcp.
>
> > But I know that KernelStack is allocated through vmalloc these days,
> > and I don't know wh
Similarly, is Mlocked a subset of Unevictable?
>
> There is some attempt at explaining how these numbers fit together, but
> it's outdated, and doesn't include Mlocked, Unevictable or KernelStack
Agreed there is a lot of tribal knowledge or even misconceptions flying
around and it will take much more work to put everything into shape.
This is only one tiny step forward.
--
Michal Hocko
SUSE Labs
usage.
> >
> > Signed-off-by: Mike Rapoport
>
> Ooops, forgot to add Michal's Ack, sorry.
Let's make it more explicit
Acked-by: Michal Hocko
Thanks!
--
Michal Hocko
SUSE Labs
On Mon, Apr 19, 2021 at 05:29:46PM +0200, Michal Kubecek wrote:
>
> As pointed out in the discussion on v3, this patch may result in
> significantly higher CPU consumption with multiple threads competing on
> a saturated outgoing device. I missed this submission so that I haven't
the
> performance data.
> V2: Avoid the overhead of fixing the data race as much as
> possible.
As pointed out in the discussion on v3, this patch may result in
significantly higher CPU consumption with multiple threads competing on
a saturated outgoing device. I missed this submission so that I haven't
checked it yet but given the description of v3->v4 changes above, it's
quite likely that it suffers from the same problem.
Michal
On Mon, Apr 19, 2021 at 08:21:38PM +0800, Yunsheng Lin wrote:
> On 2021/4/19 10:04, Yunsheng Lin wrote:
> > On 2021/4/19 6:59, Michal Kubecek wrote:
> >> I tried this patch o top of 5.12-rc7 with real devices. I used two
> >> machines with 10Gb/s Intel ixgbe NICs, sender
> > helps if all of enqueue, first spin_trylock() failure and setting the
> > __QDISC_STATE_NEED_RESCHEDULE flag happen between us finding the
> > skb_array empty and checking the flag. If enqueue happens before we
> > check the array (and flag is set before the check), the retry is
> > useless. If the flag is set after we check it, we don't catch it (no
> > matter if the enqueue happened before or after we found skb_array
> > empty).
>
> In odrder to avoid doing the "set_bit(__QDISC_STATE_MISSED, &qdisc->state)"
> as much as possible, the __QDISC_STATE_NEED_RESCHEDULE need to be set as
> as much as possible, so only clear __QDISC_STATE_NEED_RESCHEDULE when the
> queue is empty.
This is what I'm worried about. We are trying to address a race
condition which is extremely rare so we should avoid adding too much
overhead to the normal use.
> And it has about 5% performance improvement.
OK then. It thought that it would do an unnecessary dequeue retry much
more often than prevent an unnecessary __netif_schedule() but I did not
run any special benchmark to check.
Michal
} else {
> WRITE_ONCE(qdisc->empty, true);
> }i
Does the retry really provide significant improvement? IIUC it only
helps if all of enqueue, first spin_trylock() failure and setting the
__QDISC_STATE_NEED_RESCHEDULE flag happen between us finding the
skb_array empty and checking the flag. If enqueue happens before we
check the array (and flag is set before the check), the retry is
useless. If the flag is set after we check it, we don't catch it (no
matter if the enqueue happened before or after we found skb_array
empty).
Michal
est is already implemented at UAPI level so all
we need is to make it available to user.
The syntax will be a bit tricky as some string sets are global and some
per device. Out of
ethtool --show-strings [devname]
ethtool --show-strings [devname] set
ethtool --show-strings [dev ]
the third seems nicest but also least consistent with the rest of
ethtool command line syntax. So probably one of the first two.
Michal
On 13. 04. 21 14:44, Andrew Lunn wrote:
On Tue, Apr 13, 2021 at 09:09:37AM +0200, Michal Vokáč wrote:
On 12. 04. 21 16:14, Andrew Lunn wrote:
The FEC does not have PHY and is connected to the switch at RGMII level.
Adding the fixed-link { speed = <1000>; full-duplex; }; subnode to FEC
do
ack to back PHYs?
If you are doing it RGMII level, the FEC also needs a fixed-link.
The FEC does not have PHY and is connected to the switch at RGMII level.
Adding the fixed-link { speed = <1000>; full-duplex; }; subnode to FEC
does not help.
Michal
ybe this change revealed that our DT
configuration [1] is not correct? Or after this change the driver does not
handle the rgmii-id configuration properly? Or something else..
Any ideas how to further debug the problem?
Thank you in advance,
Michal
[1]
https://elixir.bootlin.com/linux/v5.12-rc7/
On 3/29/21 10:52 PM, Ido Schimmel wrote:
ip_route_me_harder() does not set source / destination port in the
flow key, so it explains why fib rules that use them are not hit after
mangling the packet. These keys were added in 4.17, but I
don't think this use case every worked. You have a differen
.tx_lpi_enabled, .tx_lpi_timer value is interpreted as a number, not
a logical value (those two are also __u32). But I don't think it's
necessary to resubmit.
Reviewed-by: Michal Kubecek
> Fixes: fd77be7bd43c ("ethtool: set EEE settings with EEE_SET request")
> Cc: # 5.10.x
&
annot
make the "queue is empty" check and leaving the critical section atomic,
it would make sense to reimplement it in a way which would allow us
making it atomic.
Michal
CPU right after a data
transfer. Perhaps you could cheat a bit and insert a random delay
between the empty queue check and releasing q->seqlock to make it more
likely to happen.
Other than that, it's rather just "run this complex software in a xen VM
and wait".
Michal
cannot be any callers (it would cause a link error).
Such a guarantee would be nicer, I agree. I tried a bit but anandoned it
when I saw macros proliferate (which I found less readable than your
current variant). But YMMV.
Michal
signature.asc
Description: Digital signature
On 3/29/21 10:52 PM, Ido Schimmel wrote:
On Sun, Mar 28, 2021 at 04:05:29PM +0200, Michal Soltys wrote:
Hi,
I'm not sure how it behaved in earlier kernels (can check later), but it is
/ looks bugged in at least recent 5.x+ ones (tests were done with 5.11.8 and
5.10.25).
Consider foll
rocessor ugliness.
Reviewed-by: Michal Koutný
signature.asc
Description: Digital signature
On 3/28/21 4:05 PM, Michal Soltys wrote:
Hi,
iptables -t mangle -A OUTPUT -j MARK --set-mark 1
After some extra checks, the exact same issue happens if we mangle tos
instead of mark, so ...
iptables -t mangle -A OUTPUT -j TOS --set-tos 0x02
... will cause same routing issues with
remap_resource(&pdev->dev, ethres);
> if (IS_ERR(lp->regs)) {
> - dev_err(&pdev->dev, "could not map Axi Ethernet regs.\n");
> ret = PTR_ERR(lp->regs);
> goto cleanup_clk;
> }
>
I have already reviewed v1 and I can't see any v2 description what has
changed. Why am I getting this v2? Where is my tag?
Thanks,
Michal
Hi,
I'm not sure how it behaved in earlier kernels (can check later), but it
is / looks bugged in at least recent 5.x+ ones (tests were done with
5.11.8 and 5.10.25).
Consider following setup:
# ip -o ad sh
1: loinet 127.0.0.1/8 scope host lo
2: right1inet 10.0.10.2/24 scope global
3
On Thu 25-03-21 12:47:04, Johannes Weiner wrote:
> On Thu, Mar 25, 2021 at 10:02:28AM +0100, Michal Hocko wrote:
> > On Wed 24-03-21 15:49:15, Arjun Roy wrote:
> > > On Wed, Mar 24, 2021 at 2:24 PM Johannes Weiner
> > > wrote:
> > > >
> > > &g
On Wed 24-03-21 15:49:15, Arjun Roy wrote:
> On Wed, Mar 24, 2021 at 2:24 PM Johannes Weiner wrote:
> >
> > On Wed, Mar 24, 2021 at 10:12:46AM +0100, Michal Hocko wrote:
> > > On Tue 23-03-21 11:47:54, Arjun Roy wrote:
> > > > On Tue, Mar 23,
had
> > reasons against, and it isn't always the case that MEMCG support is
> > enabled.
> > 3) Use compound_dtor - but I think this would have problems for the
> > prior reasons.
>
> I don't think Michal is suggesting to use PageCompound() or
> PageHead().
On Tue 23-03-21 11:47:54, Arjun Roy wrote:
> On Tue, Mar 23, 2021 at 7:34 AM Michal Hocko wrote:
> >
> > On Wed 17-03-21 18:12:55, Johannes Weiner wrote:
> > [...]
> > > Here is an idea of how it could work:
> > >
> > > struct pag
always valid when a page is freed
this would be really a nice and useful abstraction because you wouldn't
have to care about the specific type of page.
But maybe I am just overlooking the real complexity there.
--
Michal Hocko
SUSE Labs
EE`` ``ETHTOOL_MSG_EEE_GET``
> >``ETHTOOL_SEEE````ETHTOOL_MSG_EEE_SET``
> >``ETHTOOL_GRSSH`` n/a
>
> We should check with Michal about this. It is not a direct replacement
> of the old IOCTL API, it is new API. He ma
itionally avoiding the warning.
Fixes: 635e442f4a48 ("ibmvnic: merge ibmvnic_reset_init and ibmvnic_init")
Signed-off-by: Michal Suchanek
---
drivers/net/ethernet/ibm/ibmvnic.c | 8 +++-
1 file changed, 3 insertions(+), 5 deletions(-)
diff --git a/drivers/net/ethernet/ibm/ibmvn
On Tue, Feb 23, 2021 at 05:02:06PM -0800, Jakub Kicinski wrote:
> On Wed, 24 Feb 2021 01:32:51 +0100 Michal Kubecek wrote:
> > On Tue, Feb 23, 2021 at 02:24:40PM +0100, Simon Horman wrote:
> > > From: Yinjun Zhang
> > >
> > > The command "ethtool -L
]
> are NULL in this case when recent ethtool is used.
>
> Tested using ethtool v5.10.
>
> Fixes: 7be92514b99c ("ethtool: check if there is at least one channel for
> TX/RX in the core")
> Signed-off-by: Yinjun Zhang
> Signed-off-by: Simon Horman
> Signed-of
or
> > TX/RX in the core")
> > Signed-off-by: Yinjun Zhang
> > Signed-off-by: Simon Horman
> > Signed-off-by: Louis Peens
>
> Please make sure you CC Michal on ethtool patches.
>
> > diff --git a/net/ethtool/channels.c b/net/ethtool/channels.c
> >
eported-by: Bruce LIU
Signed-off-by: Michal Kubecek
---
ethtool.c | 12
1 file changed, 12 insertions(+)
diff --git a/ethtool.c b/ethtool.c
index fb90e9e456b9..15e9d34831b3 100644
--- a/ethtool.c
+++ b/ethtool.c
@@ -72,6 +72,16 @@ static void exit_bad_args(void)
exit(1);
}
+s
ion for lanes.
> Patch #5: Add lanes to man page.
>
> v3:
> * Add a seperated patch, patch #1, for uapi headers and squash
> * the rest of it to patch #2.
Series applied, thank you.
Michal
>
> Danielle Ratson (5):
> update UAPI header copies
> netlink
er. The important point is to have
a consistent snapshot of all copied (and sanitized) uapi headers.
Michal
signature.asc
Description: PGP signature
tion/bash/ethtool | 4
> uapi/linux/ethtool_netlink.h | 1 +
> 6 files changed, 25 insertions(+)
Sorry for the delay, I was busy with other stuff last week and missed
that with kernel part accepted, I should take care of the userspace
counterpart.
The series looks good to me, except for tw
to your local git
repository with kernel tree and takes one argument identifying the
commit you want to import the uapi headers from (commit id, tag or
branch name can be used). In your case, net-next would be the most
likely choice.
Michal
> Notes:
> v2:
> * Update headers after
+20,7 @@
> struct link_mode_info {
> enum link_mode_classclass;
> u32 speed;
> + u32 lanes;
> u8 duplex;
> };
>
This structure member is not used anywhere in this patc
Use already prepared dev_err_probe() introduced by commit a787e5400a1c
("driver core: add device probe log helper").
It simplifies EPROBE_DEFER handling.
Also unify message format for similar error cases.
Signed-off-by: Michal Simek
---
drivers/net/can/xilinx_can.c | 10
> ; Andrew Lunn ; f.faine...@gmail.com;
> > Michal Kubecek ; mlxsw
> > ; Ido Schimmel
> > Subject: Re: [PATCH net-next v3 2/7] ethtool: Get link mode in use instead
> > of speed and duplex parameters
> >
> > For one thing, it's cleaner if the driver
On Wed, Jan 20, 2021 at 09:21:22PM +0900, Yuusuke Ashizuka wrote:
> Fixes: 558f7cc33daf ("netlink: add master/slave configuration support")
> Signed-off-by: Yuusuke Ashizuka
Applied, thank you.
Michal
> ---
> ethtool.c | 2 +-
> 1 file changed, 1 insertion(+), 1 dele
overflowing and inform userspace that
the list is incomplete. Not perfect but there is no perfect solution
which would not require userspace changes to work properly for devices
with "too many" VFs.
Michal
proximately 300 bytes, depending on alignment and optional
> > fields). Exceeding the limit causes IFLA_VFINFO_LIST's length to wrap
> > modulo 16 bits in the kernel's nla_nest_end().
>
> Let's add Michal to CC, my faulty memory tells me he was fighting with
>
On Wed, Dec 30, 2020 at 03:23:14PM +0800, YANG LI wrote:
> The error is due to dereference a null pointer in function
> reset_one_sub_crq_queue():
>
> if (!scrq) {
> netdev_dbg(adapter->netdev,
>"Invalid scrq reset. irq (%d) or msgs(%p).\n",
> scrq->irq, scrq->msg
t to write terminating null character
into a string which may be read only. There is also a proposed fix.
Michal
On Sun, Dec 27, 2020 at 12:34:09PM +0800, Bruce LIU wrote:
> Hi Michal Kubecek and Network dev team,
>
> Good day! Hope you are doing well.
> This is Bruce from China, and please allow me to cc Rudy from Cisco Systems
> in China team.
>
> We are facing a weird behavio
with at this
point please? If so, do not hesitate to let me know.
Thank you very much for reading this rather lengthy report and for any
reply. With wishing of all the best,
Michal Tarana
--
Mgr. Michal Tarana, PhD
Department of Theoretical Chemistry
J Heyrovský Institute of Physi
: leaked instances of struct nl_socket
* Fix: improve compatibility between netlink and ioctl (-s)
Enjoy,
Michal
signature.asc
Description: PGP signature
On Tue, Dec 15, 2020 at 10:08:10AM +0100, Ivan Vecera wrote:
> Fix two error paths in ethnl_set_channels() to avoid lock-up caused
> but unreleased RTNL.
>
> Fixes: e19c591eafad ("ethtool: set device channel counts with CHANNELS_SET
> request")
> Cc: Michal Kub
tring sets with STRSET_GET request")
Reported-by: syzbot+96523fb438937cd01...@syzkaller.appspotmail.com
Signed-off-by: Michal Kubecek
---
net/ethtool/strset.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/net/ethtool/strset.c b/net/ethtool/strset.c
index 0baad0ce1832..c3
ny request
fails or if all fail. (Personally, I would prefer the latter.) And we
should probably modify nl_sset() to behave the same way as it currently
bails out on first failed request.
Michal
recognize), we also need to cap change_bits to nbits so
that we don't try to write past the prepared bitmaps.
Fixes: 88db6d1e4f62 ("ethtool: add ethnl_parse_bitset() helper")
Reported-by: syzbot+9d39fa49d4df294aa...@syzkaller.appspotmail.com
Signed-off-by: Michal Kubecek
---
net/e
Can we be sure that newer version of the standard cannot change this,
e.g. come with a finer granularity? Perhaps it would be safer to express
the size in bytes in the userspace API and translate to this internal
representation in common ethtool code.
Also, please don't forget to update Documentation/networking/ethtool-netlink.rst
Michal
ason for the failure.
>
> Signed-off-by: Baruch Siach
Applied (with minor whitespace formatting cleanup), thank you.
Michal
> ---
> v2: Limit message to likely errno values (Andrew Lunn)
> ---
> ethtool.c | 4
> 1 file changed, 4 insertions(+)
>
> dif
icit list types or default. The logic still makes
sense to me. Couldn't we address the issue by either setting features in
NETIF_F_UPPER_DISABLES) by default for a new vlan (and probably macvlan)
device? Or perhaps inheriting their values from the lower device.
Michal
unix_mknod() out of bindlock"), this
should revert it to the original value. As far as I'm aware, only LTP
guys noticed back then. I tried to raise the question in this list but
nobody seemed to care so I didn't pursue.
Technically, both errors are correct as both "address already in use"
and "socket already bound" conditions are met and the manual page
(neither linux nor posix) doesn't seem to specify which should take
precedence.
Michal
dn't go in right now in the
> development cycle anyway, all features for 5.10 had to be in linux-next
> before 5.9 was released.
>From the context, I believe Stephen meant merging into iproute2 5.10,
not kernel.
Michal Kubecek
On Wed, Nov 25, 2020 at 10:35:35AM +, Danielle Ratson wrote:
> > > What do you think of passing the link modes you have suggested as a
> > > bitmask, similar to "supported", that contains only one positive bit?
> > > Something like that:
>
&g
On Mon, Nov 23, 2020 at 09:47:53AM +, Danielle Ratson wrote:
>
>
> > -Original Message-
> > From: Michal Kubecek
> > Sent: Thursday, October 22, 2020 7:28 PM
> > To: Danielle Ratson
> > Cc: Jiri Pirko ; Andrew Lunn ; Jakub
> > Kicinski ;
On Fri, Nov 20, 2020 at 02:39:38PM +0100, Andrew Lunn wrote:
> On Fri, Nov 20, 2020 at 08:23:22AM +0100, Michal Kubecek wrote:
> > On Fri, Nov 20, 2020 at 10:59:59AM +0800, tanhuazhong wrote:
> > > On 2020/11/20 6:02, Michal Kubecek wrote:
> > > >
> > > >
On Fri, Nov 20, 2020 at 10:59:59AM +0800, tanhuazhong wrote:
> On 2020/11/20 6:02, Michal Kubecek wrote:
> >
> > We could use a similar approach as struct ethtool_link_ksettings, e.g.
> >
> > struct kernel_ethtool_coalesce {
> >
ct
kernel_ethtool_coalesce and ioctl code would be modified to only touch
the base (legacy?) part.
While already changing the ops arguments, we could also add extack
pointer, either as a separate argument or as struct member (I slightly
prefer the former).
BtW, please don't forget to update the message descriptions in
Documentation/networking/ethtool-netlink.rst
Michal
suffice for a read only bool attribute but if the
attribute is expected to be set, we need to distinguish three cases:
set to true, set to false and keep current value. That's why we use
NLA_U8 for read write bool attributes (0 false, anything else true and
attribute not present means leave untouched).
Michal
On Tue, Nov 17, 2020 at 04:20:14PM +0100, Antonio Cardace wrote:
> As pointed out by Michal Kubecek, getting the name
> with the previous approach was racy, it's better
> and easier to get the name of the device with this
> patch's approach.
>
> Essentialy the fu
!iwarp_info->mpa_bufs)
> + if (!iwarp_info->mpa_bufs) {
> + rc = -ENOMEM;
> goto err;
> + }
>
> INIT_LIST_HEAD(&iwarp_info->mpa_buf_pending_list);
> INIT_LIST_HEAD(&iwarp_info->mpa_buf_list);
> --
> 2.9.5
Thanks,
Acked-by: Michal Kalderon
On Sat, Nov 14, 2020 at 12:16:54AM +0100, Antonio Cardace wrote:
> Factor out some useful functions so that they can be reused
> by other ethtool-netdevsim scripts.
>
> Signed-off-by: Antonio Cardace
Reviewed-by: Michal Kubecek
Just one comment:
[...]
> +function
On Sat, Nov 14, 2020 at 12:16:53AM +0100, Antonio Cardace wrote:
> Add ethtool ring and coalesce settings support for testing.
>
> Signed-off-by: Antonio Cardace
Reviewed-by: Michal Kubecek
signature.asc
Description: PGP signature
On Sat, Nov 14, 2020 at 12:16:52AM +0100, Antonio Cardace wrote:
> This bitmask represents all existing coalesce parameters.
>
> Signed-off-by: Antonio Cardace
Reviewed-by: Michal Kubecek
> ---
> include/linux/ethtool.h | 1 +
> 1 file changed, 1 insertion(+)
>
> di
ns->ethtool.ring.rx_mini_max_pending = UINT32_MAX;
> + ns->ethtool.ring.tx_max_pending = UINT32_MAX;
> +}
This way an ETHTOOL_MSG_RINGS_SET request would never fail. It would be
more useful for selftests if the max values were more realistic and
ideally also configurable via debugfs.
Michal
goto cleanup;
}
if (json_output)
jsonw_null(json_wtr);
- return 0;
+ err = 0;
+
+cleanup:
+ close(progfd);
+ return err;
}
static int do_detach(int argc, char **argv)
LGTM, thanks!
Reviewed-By: Michal Rostecki
netdev features with FEATURES_SET request")
> Signed-off-by: Alexander Lobakin
Reviewed-by: Michal Kubecek
> ---
> net/ethtool/features.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/net/ethtool/features.c b/net/ethtool/features.c
> index 8e
.
Signed-off-by: Michal Kubecek
---
netlink/cable_test.c | 2 +-
netlink/channels.c | 2 +-
netlink/coalesce.c | 2 +-
netlink/eee.c| 2 +-
netlink/parser.c | 43 ---
netlink/parser.h | 3 ++-
netlink/pause.c | 2 +-
netlink
tch cleans up the parser interface; it allows nl_sset() to inspect
the composed message and append an attribute to it if needed.
Ido Schimmel (1):
ethtool: Improve compatibility between netlink and ioctl interfaces
Michal Kubecek (1):
netlink: do not send messages and process replies in
ransceiver: internal
MDI-X: on (auto)
Supports Wake-on: umbg
Wake-on: d
Current message level: 0x0007 (7)
drv probe link
Link detected: yes
Signed-off-by: Ido Schimmel
Signed-off-by: Michal Kubecek
---
net
1 - 100 of 1557 matches
Mail list logo
