- 12 Feb, 2019 31 commits
-
-
David S. Miller authored
Christoph Hellwig says: ==================== net: don't pass a NULL struct device to DMA API functions v2 We still have a few drivers which pass a NULL struct device pointer to DMA API functions, which generally is a bad idea as the API implementations rely on the device not only for ops selection, but also the dma mask and various other attributes. This series contains all easy conversions to pass a struct device, besides that there also is some arch code that needs separate handling, a driver that should not use the DMA API at all, and one that is a complete basket case to be deal with separately. Changes since v1: - fix an inverted ifdef in CAIF - update the smc911x changelog - split the series, this only contains the networking patches ==================== Signed-off-by:David S. Miller <davem@davemloft.net>
-
Christoph Hellwig authored
The DMA API generally relies on a struct device to work properly, and only barely works without one for legacy reasons. Pass the easily available struct device from the platform_device to remedy this. Note that smc911x apparently is a PIO chip with an external DMA handshake, and we probably use the wrong device here. But at least it matches the mapping side, which apparently works or at least worked in the not too distant past. Signed-off-by:
Christoph Hellwig <hch@lst.de> Signed-off-by:
-
Christoph Hellwig authored
The DMA API generally relies on a struct device to work properly, and only barely works without one for legacy reasons. Pass the easily available struct device from the platform_device to remedy this. Also use GFP_KERNEL instead of GFP_ATOMIC as the gfp_t for the memory allocation, as we aren't in interrupt context or under a lock. Signed-off-by:
-
Christoph Hellwig authored
The DMA API generally relies on a struct device to work properly, and only barely works without one for legacy reasons. Pass the easily available struct device from the platform_device to remedy this. Signed-off-by:
-
Christoph Hellwig authored
The DMA API generally relies on a struct device to work properly, and only barely works without one for legacy reasons. Pass the easily available struct device from the platform_device to remedy this. Note that this driver seems to entirely lack dma_map_single error handling, but that is left for another time. Signed-off-by:
-
Christoph Hellwig authored
The DMA API generally relies on a struct device to work properly, and only barely works without one for legacy reasons. Pass the easily available struct device from the platform_device to remedy this. Note this driver seems to lack dma_unmap_* calls entirely, but fixing that is left for another time. Signed-off-by:
-
Christoph Hellwig authored
The DMA API generally relies on a struct device to work properly, and only barely works without one for legacy reasons. Pass the easily available struct device from the platform_device to remedy this. Signed-off-by:
Nicolas Ferre <nicolas.ferre@microchip.com> Signed-off-by:
-
Christoph Hellwig authored
The DMA API generally relies on a struct device to work properly, and only barely works without one for legacy reasons. Pass the easily available struct device from the platform_device to remedy this. Signed-off-by:
-
Christoph Hellwig authored
The DMA API generally relies on a struct device to work properly, and only barely works without one for legacy reasons. Pass the easily available struct device from the platform_device to remedy this. Also use the proper Kconfig symbol to check for DMA API availability. Signed-off-by:
-
David S. Miller authored
Ido Schimmel says: ==================== mlxsw: Several updates Patches #1-#3 contain misc updates for the mlxsw driver, one of which is a fix following recent introduction of flow_rule infrastructure. Patch #4 avoids double sourcing of lib.sh in forwarding selftests. ==================== Signed-off-by: -
Jiri Pirko authored
Don't source lib.sh 2 times and make the script work with ifnames passed on the command line. Signed-off-by:
Jiri Pirko <jiri@mellanox.com> Signed-off-by:
Ido Schimmel <idosch@mellanox.com> Signed-off-by:
-
Ido Schimmel authored
The driver does not support VLAN push and pop, but only VLAN modify. Fixes: 73867881 ("drivers: net: use flow action infrastructure") Signed-off-by:
-
Ido Schimmel authored
In case the register access failed an error would be logged anyway, so we can drop the warning. Signed-off-by:
-
Nir Dotan authored
The LAG port collecting (receive) function was mistakenly set when the port was registered as a LAG member, while it should be set only when the port collection state is set to true. Set LAG port to collecting when it is set to distributing, as described in the IEEE link aggregation standard coupled control mux machine state diagram. Signed-off-by:
Nir Dotan <nird@mellanox.com> Acked-by:
-
David S. Miller authored
Ursula Braun says: ==================== net/smc: patches 2019-02-12 here are patches for SMC: * patches 1 and 3 optimize SMC-R tx logic * patch 2 is a cleanup without functional change * patch 4 optimizes rx logic * patches 5 and 6 improve robustness in link group and IB event handling * patch 7 establishes Karsten Graul as another SMC maintainer ==================== Signed-off-by: -
Ursula Braun authored
Add Karsten as additional maintainer for Shared Memory Communications (SMC) Sockets. Acked-by:
Karsten Graul <kgraul@linux.ibm.com> Signed-off-by:
Ursula Braun <ubraun@linux.ibm.com> Signed-off-by:
-
Karsten Graul authored
For robustness protect of higher port numbers than expected to avoid setting bits behind our port_event_mask. In case of an DEVICE_FATAL event all ports must be checked. The IB_EVENT_GID_CHANGE event is provided in the global event handler, so handle it there. And handle a QP_FATAL event instead of an DEVICE_FATAL event in the qp handler. Signed-off-by:
-
Karsten Graul authored
Remove the shortcut that smc_lgr_free() would skip the check for existing connections when the link group is not in the link group list. Signed-off-by:
-
Karsten Graul authored
In smc_cdc_msg_recv_action() the received cdc message is evaluated. To reduce the number of messaged triggered by this evaluation the logic is streamlined. For the write_blocked condition we do not need to send a response immediately. The remaining conditions can be put together into one if clause. Signed-off-by:
-
Karsten Graul authored
When no free transfer buffers are available then a work to call smc_tx_work() is scheduled. Set the schedule delay to zero, because for the out-of-buffers condition the work can start immediately and will block in the called function smc_wr_tx_get_free_slot(), waiting for free buffers. Signed-off-by:
-
Karsten Graul authored
Move the call to smc_close_wake_tx_prepared() (which wakes up a possibly waiting close processing that might wait for 'all data sent') to smc_tx_sndbuf_nonempty() (which is the main function to send data). Signed-off-by:
-
Karsten Graul authored
When an updated rx_cursor_confirmed field was sent to the peer then reset the cons_curs_upd_req flag. And remove the duplicate reset and cursor update in smc_tx_consumer_update(). Signed-off-by:
-
Jakub Kicinski authored
Recent cls_flower offload rewrite added a double new line. Signed-off-by:
Jakub Kicinski <jakub.kicinski@netronome.com> Reviewed-by:
Quentin Monnet <quentin.monnet@netronome.com> Signed-off-by:
-
Eli Britstein authored
Implementation of macro "flow_action_for_each" introduced in commit e3ab786b ("flow_offload: add flow action infrastructure") and used in commit 73867881 ("drivers: net: use flow action infrastructure") iterated the first item twice and did not reach the last one. Fix it. Fixes: e3ab786b ("flow_offload: add flow action infrastructure") Fixes: 73867881 ("drivers: net: use flow action infrastructure") Signed-off-by:
Eli Britstein <elibr@mellanox.com> Reviewed-by:
Roi Dayan <roid@mellanox.com> Acked-by:
Pablo Neira Ayuso <pablo@netfilter.org> Signed-off-by:
-
Randy Dunlap authored
Fix markup warnings for one list by using correct list syntax. Fix markup warnings for another list by using blank lines before the list. Documentation/networking/device_drivers/freescale/dpaa2/dpio-driver.rst:30: WARNING: Unexpected indentation. Documentation/networking/device_drivers/freescale/dpaa2/dpio-driver.rst:143: WARNING: Unexpected indentation. Signed-off-by:
Randy Dunlap <rdunlap@infradead.org> Cc: Stuart Yoder <stuyoder@gmail.com> Cc: Laurentiu Tudor <laurentiu.tudor@nxp.com> Cc: Ioana Radulescu <ruxandra.radulescu@nxp.com> Cc: netdev@vger.kernel.org Cc: Madalin Bucur <madalin.bucur@nxp.com> Signed-off-by:
-
David S. Miller authored
Jakub Kicinski says: ==================== devlink: minor tweaks to reported device info This series contains two minor touch ups for devlink code. First || is corrected to && in the ethtool compat code. Next patch decreases the stack allocation size. On the nfp side after further discussions with the manufacturing team we decided to realign the serial number contents slightly and rename one of the other fields from "vendor" to "mfr", short for "manufacture". v2: - add patch 3 - move board maker as a generic attribute. ==================== Signed-off-by: -
Jakub Kicinski authored
The manufacturing team requests we include vendor and product in the serial number field, as the serial number itself is not unique across manufacturing facilities and products. Signed-off-by:
Dirk van der Merwe <dirk.vandermerwe@netronome.com> Acked-by:
-
Jakub Kicinski authored
Vendor may sound ambiguous, let's rename the fab string to "board.manufacture" (which was just added as a generic identifier). Signed-off-by:
-
Jakub Kicinski authored
At Jiri's suggestion add a generic "board.manufacture" version identifier. Signed-off-by:
-
Jakub Kicinski authored
Number of devlink attributes has grown over 128, causing the following warning: ../net/core/devlink.c: In function ‘devlink_nl_cmd_region_read_dumpit’: ../net/core/devlink.c:3740:1: warning: the frame size of 1064 bytes is larger than 1024 bytes [-Wframe-larger-than=] } ^ Since the number of attributes is only going to grow allocate the array dynamically. Signed-off-by:
-
Jakub Kicinski authored
We need the port to be both ethernet and have the rigth netdev, not one or the other. Fixes: ddb6e99e ("ethtool: add compat for devlink info") Signed-off-by:
-
- 11 Feb, 2019 2 commits
-
-
Parav Pandit authored
Add WARN_ON to make sure that all sub objects of a devlink device are cleanedup before freeing the devlink device. This helps to catch any driver bugs. Signed-off-by:
Parav Pandit <parav@mellanox.com> Signed-off-by:
-
Tristram Ha authored
The flag offload_fwd_mark is set as the switch can forward frames by itself. This can be considered a fix to a problem introduced in commit c2e86691 where the port membership are not set in sync. The flag offload_fwd_mark just needs to be set in tag_ksz.c to prevent the software bridge from forwarding duplicate multicast frames. Fixes: c2e86691 ("microchip: break KSZ9477 DSA driver into two files") Signed-off-by:
Tristram Ha <Tristram.Ha@microchip.com> Signed-off-by:
-
- 10 Feb, 2019 7 commits
-
-
David S. Miller authored
Heiner Kallweit says: ==================== net: phy: add and use register modifying helpers returning 1 on change Add and use register modifying helpers returning 1 on change. ==================== Signed-off-by: -
Heiner Kallweit authored
Use phy_modify_changed() to simplify the code. Signed-off-by:
Heiner Kallweit <hkallweit1@gmail.com> Reviewed-by:
Andrew Lunn <andrew@lunn.ch> Reviewed-by:
Florian Fainelli <f.fainelli@gmail.com> Signed-off-by:
-
Heiner Kallweit authored
When replacing mv3310_modify() with phy_modify_mmd() we missed that they behave differently, mv3310_modify() returns 1 on a changed register value whilst phy_modify_mmd() returns 0. Fix this by replacing phy_modify_mmd() with phy_modify_mmd_changed() where needed. Fixes: b52c018d ("net: phy: make use of new MMD accessors") Signed-off-by:
-
Heiner Kallweit authored
When modifying registers there are scenarios where we need to know whether the register content actually changed. This patch adds new helpers to not break users of the current ones, phy_modify() etc. Signed-off-by:
-
David S. Miller authored
Eli Cohen says: ==================== Change tc action identifiers to be more consistent This two patch series modifies TC actions identifiers to be more consistent and also puts them in one place so new identifiers numbers can be chosen more easily. ==================== Signed-off-by: -
Eli Cohen authored
Modify the kernel users of the TCA_ACT_* macros to use TCA_ID_*. For example, use TCA_ID_GACT instead of TCA_ACT_GACT. This will align with TCA_ID_POLICE and also differentiates these identifier, used in struct tc_action_ops type field, from other macros starting with TCA_ACT_. To make things clearer, we name the enum defining the TCA_ID_* identifiers and also change the "type" field of struct tc_action to id. Signed-off-by:
Eli Cohen <eli@mellanox.com> Acked-by:
-
Eli Cohen authored
Move all the TC identifiers to one place, to the same enum that defines the identifier of police action. This makes it easier choose numbers for new actions since they are now defined in one place. We preserve the original values for binary compatibility. New IDs should be added inside the enum. Signed-off-by:
-
