- 08 Feb, 2024 4 commits
-
-
Jakub Kicinski authored
Heiner Kallweit says: ==================== net: phy: realtek: complete 5Gbps support and replace private constants Realtek maps standard C45 registers to vendor-specific registers which can be accessed via C22 w/o MMD. For an unknown reason C22 MMD access to C45 registers isn't supported for integrated PHY's. However the vendor-specific registers preserve the format of the C45 registers, so we can use standard constants. First two patches are cherry-picked from a series posted by Marek some time ago. RTL8126 supports 5Gbps, therefore add the missing 5Gbps support to rtl822x_config_aneg(). ==================== Link: https://lore.kernel.org/r/31a83fd9-90ce-402a-84c7-d5c20540b730@gmail.comSigned-off-by:
Jakub Kicinski <kuba@kernel.org>
-
Heiner Kallweit authored
RTL8126 as an evolution of RTL8125 supports 5Gbps. rtl822x_config_aneg() is used by the PHY driver for the integrated PHY, therefore add 5Gbps support to it. Signed-off-by:
Heiner Kallweit <hkallweit1@gmail.com> Link: https://lore.kernel.org/r/5644ab50-e3e9-477c-96db-05cd5bdc2563@gmail.comSigned-off-by:
-
Marek Behún authored
Drop the ad-hoc MDIO constants used in the driver and use generic constants instead. Signed-off-by:
Marek Behún <kabel@kernel.org> Signed-off-by:
Andrew Lunn <andrew@lunn.ch> Link: https://lore.kernel.org/r/732a70d6-4191-4aae-8862-3716b062aa9e@gmail.comSigned-off-by:
-
Marek Behún authored
Add constants indicating 2.5g and 5g ability in the MMD PMA speed register. Signed-off-by:
-
- 07 Feb, 2024 26 commits
-
-
Jiri Pirko authored
Similar to what is done in dpll_device_unregister(), add assertion to __dpll_pin_unregister() to make sure driver does not try to unregister non-registered pin. Signed-off-by:
Jiri Pirko <jiri@nvidia.com> Reviewed-by:
Vadim Fedorenko <vadim.fedorenko@linux.dev> Reviewed-by:
Arkadiusz Kubalewski <arkadiusz.kubalewski@intel.com> Link: https://lore.kernel.org/r/20240206074853.345744-1-jiri@resnulli.usSigned-off-by:
-
Markus Elfring authored
A wrapper function is available since the commit 890cc39a ("drivers: provide devm_platform_get_and_ioremap_resource()"). Thus reuse existing functionality instead of keeping duplicate source code. This issue was detected by using the Coccinelle software. Signed-off-by:
Markus Elfring <elfring@users.sourceforge.net> Tested-by:
Gerhard Engleder <gerhard@engleder-embedded.com> Link: https://lore.kernel.org/r/29e9dc0f-5597-4fee-be5c-25a5ab4fe2dc@web.deSigned-off-by:
-
Amit Cohen authored
init_dummy_netdev() always returns zero and all the callers do not check the returned value. Set the function to not return value, as it is not really used today. Signed-off-by:
Amit Cohen <amcohen@nvidia.com> Reviewed-by:
Ido Schimmel <idosch@nvidia.com> Reviewed-by:
Simon Horman <horms@kernel.org> Link: https://lore.kernel.org/r/20240205103022.440946-1-amcohen@nvidia.comSigned-off-by:
-
Jakub Kicinski authored
Russell King says: ==================== net: eee network driver cleanups Since commit d1420bb9 ("net: phy: improve generic EEE ethtool functions") changed phylib to set eee->eee_active and eee->eee_enabled, overriding anything that drivers have set these to prior to calling phy_ethtool_get_eee(). Therefore, drivers setting these members becomes redundant, since phylib overwrites the values they set. This series finishes off Heiner's work in the referenced commit by removing these redundant writes in various drivers and any associated code or structure members that become unnecessary. ==================== Link: https://lore.kernel.org/r/Zb9/O81fVAZw4ANr@shell.armlinux.org.ukSigned-off-by:
-
Russell King (Oracle) authored
b53_get_mac_eee() sets both eee_enabled and eee_active, and then returns zero. dsa_slave_get_eee(), which calls this function, will then continue to call phylink_ethtool_get_eee(), which will return -EOPNOTSUPP if there is no PHY present, otherwise calling phy_ethtool_get_eee() which in turn will call genphy_c45_ethtool_get_eee(). genphy_c45_ethtool_get_eee() will overwrite eee_enabled and eee_active with its own interpretation from the PHYs settings and negotiation result. Thus, when there is no PHY, dsa_slave_get_eee() will fail with -EOPNOTSUPP, meaning eee_enabled and eee_active will not be returned to userspace. When there is a PHY, eee_enabled and eee_active will be overwritten by phylib, making the setting of these members in b53_get_mac_eee() entirely unnecessary. Remove this code, thus simplifying b53_get_mac_eee(). Reviewed-by:
Florian Fainelli <florian.fainelli@broadcom.com> Signed-off-by:
Russell King (Oracle) <rmk+kernel@armlinux.org.uk> Reviewed-by:
Vladimir Oltean <olteanv@gmail.com> Link: https://lore.kernel.org/r/E1rWbNI-002cCz-4x@rmk-PC.armlinux.org.ukSigned-off-by:
-
Russell King (Oracle) authored
bcmasp_get_eee() sets edata->eee_active and edata->eee_enabled from its own copy, and then calls phy_ethtool_get_eee() which in turn will call genphy_c45_ethtool_get_eee(). genphy_c45_ethtool_get_eee() will overwrite eee_enabled and eee_active with its own interpretation from the PHYs settings and negotiation result. Therefore, setting these members in bcmasp_get_eee() is redundant, and can be removed. This also makes intf->eee.eee_active unnecessary, so remove this and use a local variable where appropriate. Reviewed-by:
-
Russell King (Oracle) authored
bcmgenet_get_eee() sets edata->eee_active and edata->eee_enabled from its own copy, and then calls phy_ethtool_get_eee() which in turn will call genphy_c45_ethtool_get_eee(). genphy_c45_ethtool_get_eee() will overwrite eee_enabled and eee_active with its own interpretation from the PHYs settings and negotiation result. Therefore, setting these members in bcmgenet_get_eee() is redundant, and can be removed. This also makes priv->eee.eee_active unnecessary, so remove this and use a local variable where appropriate. Reviewed-by:
-
Russell King (Oracle) authored
fec_enet_get_eee() sets edata->eee_active and edata->eee_enabled from its own copy, and then calls phy_ethtool_get_eee() which in turn will call genphy_c45_ethtool_get_eee(). genphy_c45_ethtool_get_eee() will overwrite eee_enabled and eee_active with its own interpretation from the PHYs settings and negotiation result. Therefore, setting these members in fec_enet_get_eee() is redundant. Remove this, and remove the setting of fep->eee.eee_active member which becomes a write-only variable. Signed-off-by:
-
Russell King (Oracle) authored
sxgbe_get_eee() sets edata->eee_active and edata->eee_enabled from its own copy, and then calls phy_ethtool_get_eee() which in turn will call genphy_c45_ethtool_get_eee(). genphy_c45_ethtool_get_eee() will overwrite eee_enabled and eee_active with its own interpretation from the PHYs settings and negotiation result. Therefore, setting these members in sxgbe_get_eee() is redundant. Remove this, and remove the priv->eee_active member which then becomes a write-only variable. Reviewed-by:
-
Russell King (Oracle) authored
stmmac_ethtool_op_get_eee() sets both eee_enabled and eee_active, and then goes on to call phylink_ethtool_get_eee(). phylink_ethtool_get_eee() will return -EOPNOTSUPP if there is no PHY present, otherwise calling phy_ethtool_get_eee() which in turn will call genphy_c45_ethtool_get_eee(). genphy_c45_ethtool_get_eee() will overwrite eee_enabled and eee_active with its own interpretation from the PHYs settings and negotiation result. Thus, when there is no PHY, stmmac_ethtool_op_get_eee() will fail with -EOPNOTSUPP, meaning eee_enabled and eee_active will not be returned to userspace. When there is a PHY, eee_enabled and eee_active will be overwritten by phylib, making the setting of these members in stmmac_ethtool_op_get_eee() entirely unnecessary. Remove this code, thus simplifying stmmac_ethtool_op_get_eee(). Reviewed-by:
Serge Semin <fancer.lancer@gmail.com> Link: https://lore.kernel.org/r/E1rWbMs-002cCV-EE@rmk-PC.armlinux.org.ukSigned-off-by:
-
David S. Miller authored
From: Andrew Lunn <andrew@lunn.ch> To: Heiner Kallweit <hkallweit1@gmail.com>, Russell King <linux@armlinux.org.uk>, "David S. Miller" <davem@davemloft.net>, Eric Dumazet <edumazet@google.com>, Jakub Kicinski <kuba@kernel.org>, Paolo Abeni <pabeni@redhat.com>, Florian Fainelli <f.fainelli@gmail.com>, Vladimir Oltean <olteanv@gmail.com> Cc: netdev@vger.kernel.org, Tim Menninger <tmenninger@purestorage.com>, Andrew Lunn <andrew@lunn.ch> ==================== net: Unify C22 and C45 error handling during bus enumeration When enumerating an MDIO bus, an MDIO bus driver can return -ENODEV to a C22 read transaction to indicate there is no device at that address on the bus. Enumeration will then continue with the next address on the bus. Modify C45 enumeration so that it also accepts -ENODEV and moves to the next address on the bus, rather than consider -ENODEV as a fatal error. Convert the mv88e6xxx driver to return -ENODEV rather than 0xffff on read for families which do not support C45 bus transactions. This is more efficient, since enumeration will scan multiple devices at one address when 0xffff is returned, where as -EONDEV immediately jumps to the next address on the bus. ==================== Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Andrew Lunn authored
MDIO bus drivers can return -ENODEV when they know the bus does not have a device at the given address, e.g. because of hardware limitation. One such limitation is that the bus does not support C45 at all. This is more efficient than returning 0xffff, since it immediately stops the probing on the given address, where as further reads can be made when 0xffff is returned. Signed-off-by:
-
Andrew Lunn authored
When scanning the MDIO bus for C22 devices, the driver returning -ENODEV is not considered fatal, it just indicates the MDIO bus master knows there is no device at that address, maybe because of hardware limitation. Make the C45 scan code act on -ENODEV the same way, to make C22 and C45 more uniform. It is expected all reads for a given address will return -ENODEV, so within get_phy_c45_ids() only the first place a read occurs has been changed. Reviewed-by:
-
Ricardo B. Marliere authored
Now that the driver core can properly handle constant struct bus_type, move the nsim_bus variable to be a constant structure as well, placing it into read-only memory which can not be modified at runtime. Cc: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Suggested-by:
Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by:
Ricardo B. Marliere <ricardo@marliere.net> Reviewed-by:
-
Masahiro Yamada authored
The 'bool' is already specified for these options. The second 'bool' under the help message is redundant. While I am here, I moved 'default y' above, as it is common to place the help text last. Signed-off-by:
Masahiro Yamada <masahiroy@kernel.org> Reviewed-by:
Randy Dunlap <rdunlap@infradead.org> Reviewed-by:
-
Masahiro Yamada authored
Both drivers/net/Kconfig and drivers/net/ethernet/Kconfig contain the same config entry: config SUNGEM_PHY tristate Commit f860b052 ("drivers/net: Kconfig and Makefile cleanup") moved SUNGEM_PHY from drivers/net/Kconfig to drivers/net/ethernet/Kconfig. Shortly after it was applied, commit 19e2f6fe ("net: Fix sungem_phy sharing.") added the second one to drivers/net/Kconfig. I kept the one in drivers/net/Kconfig because this CONFIG option controls the compilation of drivers/net/sungem_phy.c. Signed-off-by: -
George Guo authored
Simplify the code from macro NETLBL_CATMAP_MAPTYPE to u64, and fix warning "Macros with complex values should be enclosed in parentheses" on "#define NETLBL_CATMAP_BIT (NETLBL_CATMAP_MAPTYPE)0x01", which is modified to "#define NETLBL_CATMAP_BIT ((u64)0x01)". Signed-off-by:
George Guo <guodongtai@kylinos.cn> Acked-by:
Paul Moore <paul@paul-moore.com> Signed-off-by:
-
Heiner Kallweit authored
Convert EEE handling to use linkmode bitmaps. This prepares for removing the legacy bitmaps from struct ethtool_keee. No functional change intended. Note: The change to mii_eee_cap1_mod_linkmode_t(tp->eee.advertised, val) in tg3_phy_autoneg_cfg() isn't completely obvious, but it doesn't change the current functionality. Signed-off-by:
-
Jakub Kicinski authored
Heiner Kallweit says: ==================== net: phy: add and use helper phy_advertise_eee_all Per default phylib preserves the EEE advertising at the time of phy probing. The EEE advertising can be changed from user space, in addition this helper allows to set the EEE advertising to all supported modes from drivers in kernel space. ==================== Link: https://lore.kernel.org/r/0d886510-b2b7-43f2-b8a6-fb770d97266d@gmail.comSigned-off-by:
-
Heiner Kallweit authored
Use new helper phy_advertise_eee_all() to simplify the code. Reviewed-by:
-
Heiner Kallweit authored
Per default phylib preserves the EEE advertising at the time of phy probing. The EEE advertising can be changed from user space, in addition this helper allows to set the EEE advertising to all supported modes from drivers in kernel space. Suggested-by:
-
Jakub Kicinski authored
Alessandro Marcolini says: ==================== Add support for encoding multi-attr to ynl This patchset add the support for encoding multi-attr attributes, making it possible to use ynl with qdisc which have this kind of attributes (e.g: taprio, ets). Patch 1 corrects two docstrings in nlspec.py Patch 2 adds the multi-attr attribute to taprio entry Patch 3 adds the support for encoding multi-attr Some examples of what is now possible with the ynl cli: - Add a taprio qdisc --do newqdisc --create --json '{ "family":1, "ifindex":4, "handle":65536, "parent":4294967295, "info":0, "kind":"taprio", "stab":{ "base": { "cell-log": 0, "size-log": 0, "cell-align": 0, "overhead": 31, "linklayer": 0, "mpu": 0, "mtu": 0, "tsize": 0 } }, "options":{ "priomap": { "num-tc": 3, "prio-tc-map": "01010101010101010101010101010101", "hw": 0, "count": "0100010002000000000000000000000000000000000000000000000000000000", "offset": "0100020003000000000000000000000000000000000000000000000000000000" }, "sched-clockid":11, "sched-entry-list": {"entry": [ {"index":0, "cmd":0, "gate-mask":1, "interval":300000}, {"index":1, "cmd":0, "gate-mask":2, "interval":300000}, {"index":2, "cmd":0, "gate-mask":4, "interval":400000} ] }, "sched-base-time":1528743495910289987, "flags": 1 } }' - Add an ets qdisc --create --json '{ "family":1, "ifindex":4, "handle":65536, "parent":4294967295, "kind":"ets", "options":{ "nbands":6, "nstrict":3, "quanta":{ "quanta-band": [3500, 3000, 2500] }, "priomap":{ "priomap-band":[0, 1, 1, 1, 2, 3, 4, 5] } } }' ==================== Link: https://lore.kernel.org/r/cover.1706962013.git.alessandromarcolini99@gmail.comSigned-off-by: -
Alessandro Marcolini authored
Multi-attr elements could not be encoded because of missing logic in the ynl code. Enable encoding of these attributes by checking if the attribute is a multi-attr and if the value to be processed is a list. This has been tested both with the taprio and ets qdisc which contain this kind of attributes. Signed-off-by:
Alessandro Marcolini <alessandromarcolini99@gmail.com> Reviewed-by:
Donald Hunter <donald.hunter@gmail.com> Reviewed-by:
-
Alessandro Marcolini authored
Add multi-attr attribute to tc-taprio-sched-entry to specify multiple entries. Signed-off-by:
-
Alessandro Marcolini authored
Correct typo in SpecAttr docstring. Changed SpecSubMessageFormat docstring. Signed-off-by:
-
Simon Horman authored
Address spelling errors flagged by codespell. This patch follows-up on an earlier patch by Colin Ian King, which addressed a spelling error in a user-visible log message [1]. This patch includes that change. [1] https://lore.kernel.org/netdev/20231209225135.4055334-1-colin.i.king@gmail.com/ This patch is intended to cover all files under drivers/net/ethernet/mellanox/mlx4 Signed-off-by:
Tariq Toukan <tariqt@nvidia.com> Reviewed-by:
-
- 06 Feb, 2024 10 commits
-
-
Paolo Abeni authored
Christophe says: ==================== nfc: hci: Save a few bytes of memory when registering a 'nfc_llc' engine nfc_llc_register() calls pass a string literal as the 'name' parameter. So kstrdup_const() can be used instead of kfree() to avoid a memory allocation in such cases. ==================== Link: https://lore.kernel.org/r/cover.1706946099.git.christophe.jaillet@wanadoo.frSigned-off-by:
Paolo Abeni <pabeni@redhat.com>
-
Christophe JAILLET authored
nfc_llc_register() calls pass a string literal as the 'name' parameter. So kstrdup_const() can be used instead of kfree() to avoid a memory allocation in such cases. Signed-off-by:
Christophe JAILLET <christophe.jaillet@wanadoo.fr> Signed-off-by:
-
Christophe JAILLET authored
Add a new helper to avoid code duplication between nfc_llc_exit() and nfc_llc_unregister(). Signed-off-by:
-
Xin Long authored
It is not appropriate for TIPC to use "diag" as its diag module name while the other protocols are using "$(protoname)_diag" like tcp_diag, udp_diag and sctp_diag etc. So this patch is to rename diag.ko to tipc_diag.ko in tipc's Makefile. Signed-off-by:
Xin Long <lucien.xin@gmail.com> Reviewed-by:
Tung Nguyen <tung.q.nguyen@dektech.com.au> Link: https://lore.kernel.org/r/d909edeef072da1810bd5869fdbbfe84411efdb2.1706904669.git.lucien.xin@gmail.comSigned-off-by:
-
Paolo Abeni authored
Brett Creeley says: ==================== pds_core: Various improvements/cleanups This series contains various improvements and cleanups for the pds_core driver. These patches were originally part of the following net-next series: https://lore.kernel.org/netdev/20240126174255.17052-1-brett.creeley@amd.com/ However, some of the patches from the above series were actually fixes, so they were pushed and accepted to net. That series can be found here: https://lore.kernel.org/netdev/20240129234035.69802-1-brett.creeley@amd.com/ Also, the Reviewed-by tags were left in place from the original net-next reviews as the patches didn't change. ==================== Link: https://lore.kernel.org/r/20240202195911.65338-1-brett.creeley@amd.comSigned-off-by:
-
Brett Creeley authored
The setup and teardown flows are somewhat hard to follow regarding pdsc_core_init()/pdsc_dev_init() and their corresponding teardown flows being in pdsc_teardown(). Improve the readability by adding new pdsc_core_uninit()/pdsc_dev_unint() functions that mirror their init counterparts. Also, move the notify and admin qcq allocations into pdsc_core_init(), so they can be freed in pdsc_core_uninit(). Signed-off-by:
Brett Creeley <brett.creeley@amd.com> Reviewed-by:
Shannon Nelson <shannon.nelson@amd.com> Signed-off-by:
-
Brett Creeley authored
Running xmastree.py against the driver found some RCT issues, so fix them. Also, if allocating pdsc->intr_info in pdsc_dev_init() fails the driver still tries to free pdsc->intr_info. Fix this by just returning -ENOMEM since there's nothing to free at this point of failure. Signed-off-by:
-
Brett Creeley authored
Unmasking the interrupt during the pdsc_adminq_isr is a bit early and could cause unnecessary interrupts. Instead always unmask after processing the adminq and notifyq in pdsc_work_thread()->pdsc_process_adminq(). Also, since we are always unmasking, there's no need for the local credits variable in pdsc_process_adminq(). Signed-off-by:
Przemek Kitszel <przemyslaw.kitszel@intel.com> Signed-off-by:
-
Brett Creeley authored
The notifyq rides on the adminq's interrupt, so there's no need to setup and/or access the notifyq's interrupt index or bound_intr. The driver sets the bound_intr using qcq->intx = -1 for the notifyq, but luckily nothing accesses that field for notifyq. Instead of expecting that remains the case, just clean up the notifyq's interrupt index and bound_intr fields. Signed-off-by:
-
Aahil Awatramani authored
Add support for the independent control state machine per IEEE 802.1AX-2008 5.4.15 in addition to the existing implementation of the coupled control state machine. Introduces two new states, AD_MUX_COLLECTING and AD_MUX_DISTRIBUTING in the LACP MUX state machine for separated handling of an initial Collecting state before the Collecting and Distributing state. This enables a port to be in a state where it can receive incoming packets while not still distributing. This is useful for reducing packet loss when a port begins distributing before its partner is able to collect. Added new functions such as bond_set_slave_tx_disabled_flags and bond_set_slave_rx_enabled_flags to precisely manage the port's collecting and distributing states. Previously, there was no dedicated method to disable TX while keeping RX enabled, which this patch addresses. Note that the regular flow process in the kernel's bonding driver remains unaffected by this patch. The extension requires explicit opt-in by the user (in order to ensure no disruptions for existing setups) via netlink support using the new bonding parameter coupled_control. The default value for coupled_control is set to 1 so as to preserve existing behaviour. Signed-off-by:
Aahil Awatramani <aahila@google.com> Reviewed-by:
Hangbin Liu <liuhangbin@gmail.com> Link: https://lore.kernel.org/r/20240202175858.1573852-1-aahila@google.comSigned-off-by:
-
