Skip to content

L
linux
Switch branch/tag
  1. 29 Jul, 2019 8 commits
  3. 27 Jul, 2019 16 commits
  5. 26 Jul, 2019 3 commits
  7. 25 Jul, 2019 8 commits
    • David S. Miller's avatar
      Merge branch 'tipc-link-changeover-issues' · b591c6f6
      David S. Miller authored 
      Tuong Lien says:

====================
tipc: link changeover issues

This patch series is to resolve some issues found with the current link
changeover mechanism, it also includes an optimization for the link
synching.
====================
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      b591c6f6
    • Tuong Lien's avatar
      tipc: fix changeover issues due to large packet · 2320bcda
      Tuong Lien authored 
      In conjunction with changing the interfaces' MTU (e.g. especially in
the case of a bonding) where the TIPC links are brought up and down
in a short time, a couple of issues were detected with the current link
changeover mechanism:

1) When one link is up but immediately forced down again, the failover
procedure will be carried out in order to failover all the messages in
the link's transmq queue onto the other working link. The link and node
state is also set to FAILINGOVER as part of the process. The message
will be transmited in form of a FAILOVER_MSG, so its size is plus of 40
bytes (= the message header size). There is no problem if the original
message size is not larger than the link's MTU - 40, and indeed this is
the max size of a normal payload messages. However, in the situation
above, because the link has just been up, the messages in the link's
transmq are almost SYNCH_MSGs which had been generated by the link
synching procedure, then their size might reach the max value already!
When the FAILOVER_MSG is built on the top of such a SYNCH_MSG, its size
will exceed the link's MTU. As a result, the messages are dropped
silently and the failover procedure will never end up, the link will
not be able to exit the FAILINGOVER state, so cannot be re-established.

2) The same scenario above can happen more easily in case the MTU of
the links is set differently or when changing. In that case, as long as
a large message in the failure link's transmq queue was built and
fragmented with its link's MTU > the other link's one, the issue will
happen (there is no need of a link synching in advance).

3) The link synching procedure also faces with the same issue but since
the link synching is only started upon receipt of a SYNCH_MSG, dropping
the message will not result in a state deadlock, but it is not expected
as design.

The 1) & 3) issues are resolved by the last commit that only a dummy
SYNCH_MSG (i.e. without data) is generated at the link synching, so the
size of a FAILOVER_MSG if any then will never exceed the link's MTU.

For the 2) issue, the only solution is trying to fragment the messages
in the failure link's transmq queue according to the working link's MTU
so they can be failovered then. A new function is made to accomplish
this, it will still be a TUNNEL PROTOCOL/FAILOVER MSG but if the
original message size is too large, it will be fragmented & reassembled
at the receiving side.
Acked-by: default avatarYing Xue <ying.xue@windriver.com>
Acked-by: default avatarJon Maloy <jon.maloy@ericsson.com>
Signed-off-by: default avatarTuong Lien <tuong.t.lien@dektech.com.au>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      2320bcda
    • Tuong Lien's avatar
      tipc: optimize link synching mechanism · 4929a932
      Tuong Lien authored 
      This commit along with the next one are to resolve the issues with the
link changeover mechanism. See that commit for details.

Basically, for the link synching, from now on, we will send only one
single ("dummy") SYNCH message to peer. The SYNCH message does not
contain any data, just a header conveying the synch point to the peer.

A new node capability flag ("TIPC_TUNNEL_ENHANCED") is introduced for
backward compatible!
Acked-by: default avatarYing Xue <ying.xue@windriver.com>
Acked-by: default avatarJon Maloy <jon.maloy@ericsson.com>
Suggested-by: default avatarJon Maloy <jon.maloy@ericsson.com>
Signed-off-by: default avatarTuong Lien <tuong.t.lien@dektech.com.au>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      4929a932
    • Ding Xiang's avatar
      ptp: ptp_dte: remove redundant dev_err message · 37f7c66f
      Ding Xiang authored 
      devm_ioremap_resource already contains error message, so remove
the redundant dev_err message
Signed-off-by: default avatarDing Xiang <dingxiang@cmss.chinamobile.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      37f7c66f
    • David S. Miller's avatar
      Merge branch 'mlxsw-Two-small-updates' · f2ad83af
      David S. Miller authored 
      Ido Schimmel says:

====================
mlxsw: Two small updates

Patch #1, from Amit, exposes the size of the key-value database (KVD)
where different entries (e.g., routes, neighbours) are stored in the
device. This allows users to understand how many entries can be
offloaded and is also useful for writing scale tests.

Patch #2 increases the number of IPv6 nexthop groups mlxsw can offload.
The problem and solution are explained in detail in the commit message.
====================
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      f2ad83af
    • Ido Schimmel's avatar
      mlxsw: spectrum_router: Increase scale of IPv6 nexthop groups · fc25996e
      Ido Schimmel authored 
      Unlike IPv4, the kernel does not consolidate IPv6 nexthop groups. To
avoid exhausting the device's adjacency table - where nexthops are
stored - the driver does this consolidation instead.

Each nexthop group is hashed by XOR-ing the interface indexes of all the
member nexthop devices. However, the ifindex itself is not hashed, which
can result in identical keys used for different groups and finally an
-EBUSY error from rhashtable due to too long objects list.

Improve the situation by hashing the ifindex itself.
Signed-off-by: default avatarIdo Schimmel <idosch@mellanox.com>
Acked-by: default avatarJiri Pirko <jiri@mellanox.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      fc25996e
    • Amit Cohen's avatar
      mlxsw: spectrum: Expose KVD size for Spectrum-2 · b06689cc
      Amit Cohen authored 
      Unlike Spectrum-1, the KVD (Key-value database) of Spectrum-2 is not
partitioned, so only expose the entire KVD size. This enables users to
query the total size of the KVD.
Signed-off-by: default avatarAmit Cohen <amitc@mellanox.com>
Acked-by: default avatarJiri Pirko <jiri@mellanox.com>
Signed-off-by: default avatarIdo Schimmel <idosch@mellanox.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      b06689cc
    • Wolfram Sang's avatar
      net: sfc: falcon: convert to i2c_new_dummy_device · c93496e9
      Wolfram Sang authored 
      Move from i2c_new_dummy() to i2c_new_dummy_device(). So, we now get an
ERRPTR which we use in error handling.
Signed-off-by: default avatarWolfram Sang <wsa+renesas@sang-engineering.com>
Acked-by: default avatarEdward Cree <ecree@solarflare.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      c93496e9
  9. 24 Jul, 2019 5 commits
    • Arnd Bergmann's avatar
      mlx4: avoid large stack usage in mlx4_init_hca() · 9eed21c0
      Arnd Bergmann authored 
      The mlx4_dev_cap and mlx4_init_hca_param are really too large
to be put on the kernel stack, as shown by this clang warning:

drivers/net/ethernet/mellanox/mlx4/main.c:3304:12: error: stack frame size of 1088 bytes in function 'mlx4_load_one' [-Werror,-Wframe-larger-than=]

With gcc, the problem is the same, but it does not warn because
it does not inline this function, and therefore stays just below
the warning limit, while clang is just above it.

Use kzalloc for dynamic allocation instead of putting them
on stack. This gets the combined stack frame down to 424 bytes.
Signed-off-by: default avatarArnd Bergmann <arnd@arndb.de>
Reviewed-by: default avatarSaeed Mahameed <saeedm@mellanox.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      9eed21c0
    • Arnd Bergmann's avatar
      qed: reduce maximum stack frame size · 7c116e02
      Arnd Bergmann authored 
      clang warns about an overly large stack frame in one function
when it decides to inline all __qed_get_vport_*() functions into
__qed_get_vport_stats():

drivers/net/ethernet/qlogic/qed/qed_l2.c:1889:13: error: stack frame size of 1128 bytes in function '_qed_get_vport_stats' [-Werror,-Wframe-larger-than=]

Use a noinline_for_stack annotation to prevent clang from inlining
these, which keeps the maximum stack usage at around half of that
in the worst case, similar to what we get with gcc.

Fixes: 86622ee7 ("qed: Move statistics to L2 code")
Signed-off-by: default avatarArnd Bergmann <arnd@arndb.de>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      7c116e02
    • Heiner Kallweit's avatar
      r8169: improve rtl_set_rx_mode · 81cd17a4
      Heiner Kallweit authored 
      This patch improves and simplifies rtl_set_rx_mode a little.
No functional change intended.
Signed-off-by: default avatarHeiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      81cd17a4
    • David S. Miller's avatar
      Merge branch '1GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/next-queue · ce599b1a
      David S. Miller authored 
      Jeff Kirsher says:

====================
1GbE Intel Wired LAN Driver Updates 2019-07-24

This series contains updates to igc and e1000e client drivers only.

Sasha provides a couple of cleanups to remove code that is not needed
and reduce structure sizes.  Updated the MAC reset flow to use the
device reset flow instead of a port reset flow.  Added addition device
id's that will be supported.

Kai-Heng Feng provides a workaround for a possible stalled packet issue
in our ICH devices due to a clock recovery from the PCH being too slow.

v2: removed the last patch in the series that supposedly fixed a MAC/PHY
    de-sync potential issue while waiting for additional information from
    hardware engineers.
====================
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      ce599b1a
    • Qian Cai's avatar
      net/ixgbevf: fix a compilation error of skb_frag_t · d601be97
      Qian Cai authored 
      The linux-next commit "net: Rename skb_frag_t size to bv_len" [1]
introduced a compilation error on powerpc as it forgot to deal with the
renaming from "size" to "bv_len" for ixgbevf.

[1] https://lore.kernel.org/netdev/20190723030831.11879-1-willy@infradead.org/T/#md052f1c7de965ccd1bdcb6f92e1990a52298eac5

In file included from ./include/linux/cache.h:5,
                 from ./include/linux/printk.h:9,
                 from ./include/linux/kernel.h:15,
                 from ./include/linux/list.h:9,
                 from ./include/linux/module.h:9,
                 from
drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c:12:
drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c: In function
'ixgbevf_xmit_frame_ring':
drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c:4138:51: error:
'skb_frag_t' {aka 'struct bio_vec'} has no member named 'size'
   count += TXD_USE_COUNT(skb_shinfo(skb)->frags[f].size);
                                                   ^
./include/uapi/linux/kernel.h:13:40: note: in definition of macro
'__KERNEL_DIV_ROUND_UP'
 #define __KERNEL_DIV_ROUND_UP(n, d) (((n) + (d) - 1) / (d))
                                        ^
drivers/net/ethernet/intel/ixgbevf/ixgbevf_main.c:4138:12: note: in
expansion of macro 'TXD_USE_COUNT'
   count += TXD_USE_COUNT(skb_shinfo(skb)->frags[f].size);
Signed-off-by: default avatarQian Cai <cai@lca.pw>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      d601be97
GitLab Nexedi Edition | About GitLab | About Nexedi | 沪ICP备2021021310号-2 | 沪ICP备2021021310号-7