Skip to content

L
linux
Switch branch/tag
  1. 04 Oct, 2019 17 commits
  3. 03 Oct, 2019 10 commits
  5. 02 Oct, 2019 13 commits
    • Sudhakar Dindukurti's avatar
      net/rds: Log vendor error if send/recv Work requests fail · fab401e1
      Sudhakar Dindukurti authored 
      Log vendor error if work requests fail. Vendor error provides
more information that is used for debugging the issue.
Signed-off-by: default avatarSudhakar Dindukurti <sudhakar.dindukurti@oracle.com>
Acked-by: default avatarSantosh Shilimkar <santosh.shilimkar@oracle.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      fab401e1
    • Matteo Croce's avatar
      mvpp2: remove misleading comment · 80f60a91
      Matteo Croce authored 
      Recycling in mvpp2 has gone long time ago, but two comment still refers
to it. Remove those two misleading comments as they generate confusion.

Fixes: 7ef7e1d9 ("net: mvpp2: drop useless fields in mvpp2_bm_pool and related code")
Signed-off-by: default avatarMatteo Croce <mcroce@redhat.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      80f60a91
    • David S. Miller's avatar
      Merge branch 'CAIF-Kconfig-fixes' · bc389189
      David S. Miller authored 
      Randy Dunlap says:

====================
CAIF Kconfig fixes

This series of patches cleans up the CAIF Kconfig menus in
net/caif/Kconfig and drivers/net/caif/Kconfig and also puts the
CAIF Transport drivers into their own sub-menu.
====================
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      bc389189
    • rd.dunlab@gmail.com's avatar
      Minor fixes to the CAIF Transport drivers Kconfig file · 0f04f8ea
      rd.dunlab@gmail.com authored 
      Minor fixes to the CAIF Transport drivers Kconfig file:

- end sentence with period
- capitalize CAIF acronym
Signed-off-by: default avatarRandy Dunlap <rdunlap@infradead.org>
Cc: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      0f04f8ea
    • rd.dunlab@gmail.com's avatar
      Isolate CAIF transport drivers into their own menu · 21d54976
      rd.dunlab@gmail.com authored 
      Isolate CAIF transport drivers into their own menu.

This cleans up the main Network device support menu,
makes it easier to find the CAIF drivers, and makes it
easier to enable/disable them as a group.
Signed-off-by: default avatarRandy Dunlap <rdunlap@infradead.org>
Cc: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      21d54976
    • rd.dunlab@gmail.com's avatar
      Clean up the net/caif/Kconfig menu · 0903102f
      rd.dunlab@gmail.com authored 
      Clean up the net/caif/Kconfig menu:
- remove extraneous space
- minor language tweaks
- fix punctuation
Signed-off-by: default avatarRandy Dunlap <rdunlap@infradead.org>
Cc: Randy Dunlap <rdunlap@infradead.org>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      0903102f
    • Eric Dumazet's avatar
      net_sched: remove need_resched() from qdisc_run() · b60fa1c5
      Eric Dumazet authored 
      The introduction of this schedule point was done in commit
2ba2506c ("[NET]: Add preemption point in qdisc_run")
at a time the loop was not bounded.

Then later in commit d5b8aa1d ("net_sched: fix dequeuer fairness")
we added a limit on the number of packets.

Now is the time to remove the schedule point, since the default
limit of 64 packets matches the number of packets a typical NAPI
poll can process in a row.

This solves a latency problem for most TCP receivers under moderate load :

1) host receives a packet.
   NET_RX_SOFTIRQ is raised by NIC hard IRQ handler

2) __do_softirq() does its first loop, handling NET_RX_SOFTIRQ
   and calling the driver napi->loop() function

3) TCP stores the skb in socket receive queue:

4) TCP calls sk->sk_data_ready() and wakeups a user thread
   waiting for EPOLLIN (as a result, need_resched() might now be true)

5) TCP cooks an ACK and sends it.

6) qdisc_run() processes one packet from qdisc, and sees need_resched(),
   this raises NET_TX_SOFTIRQ (even if there are no more packets in
   the qdisc)

Then we go back to the __do_softirq() in 2), and we see that new
softirqs were raised. Since need_resched() is true, we end up waking
ksoftirqd in this path :

    if (pending) {
            if (time_before(jiffies, end) && !need_resched() &&
                --max_restart)
                    goto restart;

            wakeup_softirqd();
    }

So we have many wakeups of ksoftirqd kernel threads,
and more calls to qdisc_run() with associated lock overhead.

Note that another way to solve the issue would be to change TCP
to first send the ACK packet, then signal the EPOLLIN,
but this changes P99 latencies, as sending the ACK packet
can add a long delay.
Signed-off-by: default avatarEric Dumazet <edumazet@google.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      b60fa1c5
    • Vladimir Oltean's avatar
      net: dsa: Remove unused __DSA_SKB_CB macro · 37048e94
      Vladimir Oltean authored 
      The struct __dsa_skb_cb is supposed to span the entire 48-byte skb
control block, while the struct dsa_skb_cb only the portion of it which
is used by the DSA core (the rest is available as private data to
drivers).

The DSA_SKB_CB and __DSA_SKB_CB helpers are supposed to help retrieve
this pointer based on a skb, but it turns out there is nobody directly
interested in the struct __dsa_skb_cb in the kernel. So remove it.
Signed-off-by: default avatarVladimir Oltean <olteanv@gmail.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      37048e94
    • David S. Miller's avatar
      Merge branch 'sja1105-cleanups' · b74d402e
      David S. Miller authored 
      Vladimir Oltean says:

====================
SJA1105 DSA coding style cleanup

This series provides some mechanical cleanup patches related to function
names and prototypes.
====================
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      b74d402e
    • Vladimir Oltean's avatar
      net: dsa: sja1105: Rename sja1105_spi_send_packed_buf to sja1105_xfer_buf · 1bd44870
      Vladimir Oltean authored 
      The most commonly called function in the driver is long due for a
rename. The "packed" word is redundant (it doesn't make sense to
transfer an unpacked structure, since that is in CPU endianness yadda
yadda), and the "spi" word is also redundant since argument 2 of the
function is SPI_READ or SPI_WRITE.

As for the sja1105_spi_send_long_packed_buf function, it is only being
used from sja1105_spi.c, so remove its global prototype.
Signed-off-by: default avatarVladimir Oltean <olteanv@gmail.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      1bd44870
    • Vladimir Oltean's avatar
      net: dsa: sja1105: Replace sja1105_spi_send_int with sja1105_xfer_{u32, u64} · dff79620
      Vladimir Oltean authored 
      Having a function that takes a variable number of unpacked bytes which
it generically calls an "int" is confusing and makes auditing patches
next to impossible.

We only use spi_send_int with the int sizes of 32 and 64 bits. So just
make the spi_send_int function less generic and replace it with the
appropriate two explicit functions, which can now type-check the int
pointer type.

Note that there is still a small weirdness in the u32 function, which
has to convert it to a u64 temporary. This is because of how the packing
API works at the moment, but the weirdness is at least hidden from
callers of sja1105_xfer_u32 now.
Suggested-by: default avatarDavid S. Miller <davem@davemloft.net>
Signed-off-by: default avatarVladimir Oltean <olteanv@gmail.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      dff79620
    • Vladimir Oltean's avatar
      net: dsa: sja1105: Don't use "inline" function declarations in C files · 09c1b412
      Vladimir Oltean authored 
      Let the compiler decide.
Signed-off-by: default avatarVladimir Oltean <olteanv@gmail.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      09c1b412
    • David S. Miller's avatar
      Merge branch 'SMB-rootfs' · 5cf37738
      David S. Miller authored 
      Paulo Alcantara says:

====================
Experimental SMB rootfs support

This patch series enables Linux to mount root file systems over the
network by utilizing SMB protocol.

Upstream commit 8eecd1c2 ("cifs: Add support for root file
systems") introduced a new CONFIG_CIFS_ROOT option, a virtual device
(Root_CIFS) and a kernel cmdline parameter "cifsroot=" which tells the
kernel to actually mount the root filesystem over a SMB share.

The feature relies on ipconfig to set up the network prior to mounting
the rootfs, so when it is set along with "cifsroot=" parameter:

    (1) cifs_root_setup() parses all necessary data out of "cifsroot="
    parameter for the init process know how to mount the SMB rootfs
    (e.g. SMB server address, mount options).

    (2) If DHCP failed for some reason in ipconfig, we keep retrying
    forever as we have nowhere to go for NFS or SMB root
    filesystems (see PATCH 2/2). Otherwise go to (3).

    (3) mount_cifs_root() is then called by mount_root() (ROOT_DEV ==
    Root_CIFS), retrieves early parsed data from (1), then attempt to
    mount SMB rootfs by CIFSROOT_RETRY_MAX times at most (see PATCH
    1/2).

    (4) If all attempts failed, fall back to floppy drive, otherwise
    continue the boot process with rootfs mounted over a SMB share.

My idea was to keep the same behavior of nfsroot - as it seems to work
for most users so far.

For more information on how this feature works, see
Documentation/filesystems/cifs/cifsroot.txt.
====================
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      5cf37738
GitLab Nexedi Edition | About GitLab | About Nexedi | 沪ICP备2021021310号-2 | 沪ICP备2021021310号-7