ethernet: fix PTP_1588_CLOCK dependencies
Subsystems (21)
PATCH v3 1/1 ethernet: fix PTP_1588_CLOCK dependencies
From: Arnd Bergmann <[email protected]>
The 'imply' keyword does not do what most people think it does, it only politely asks Kconfig to turn on another symbol, but does not prevent it from being disabled manually or built as a loadable module when the user is built-in. In the ICE driver, the latter now causes a link failure:
aarch64-linux-ld: drivers/net/ethernet/intel/ice/ice_main.o: in function `ice_eth_ioctl':
ice_main.c:(.text+0x13b0): undefined reference to `ice_ptp_get_ts_config'
ice_main.c:(.text+0x13b0): relocation truncated to fit: R_AARCH64_CALL26 against undefined symbol `ice_ptp_get_ts_config'
aarch64-linux-ld: ice_main.c:(.text+0x13bc): undefined reference to ice_ptp_set_ts_config' ice_main.c:(.text+0x13bc): relocation truncated to fit: R_AARCH64_CALL26 against undefined symbol ice_ptp_set_ts_config'
aarch64-linux-ld: drivers/net/ethernet/intel/ice/ice_main.o: in function `ice_prepare_for_reset':
ice_main.c:(.text+0x31fc): undefined reference to `ice_ptp_release'
ice_main.c:(.text+0x31fc): relocation truncated to fit: R_AARCH64_CALL26 against undefined symbol `ice_ptp_release'
aarch64-linux-ld: drivers/net/ethernet/intel/ice/ice_main.o: in function `ice_rebuild':
This is a recurring problem in many drivers, and we have discussed it several times befores, without reaching a consensus. I'm providing a link to the previous email thread for reference, which discusses some related problems.
To solve the dependency issue better than the 'imply' keyword, introduce a separate Kconfig symbol "CONFIG_PTP_1588_CLOCK_OPTIONAL" that any driver can depend on if it is able to use PTP support when available, but works fine without it. Whenever CONFIG_PTP_1588_CLOCK=m, those drivers are then prevented from being built-in, the same way as with a 'depends on PTP_1588_CLOCK || !PTP_1588_CLOCK' dependency that does the same trick, but that can be rather confusing when you first see it.
Since this should cover the dependencies correctly, the IS_REACHABLE() hack in the header is no longer needed now, and can be turned back into a normal IS_ENABLED() check. Any driver that gets the dependency wrong will now cause a link time failure rather than being unable to use PTP support when that is in a loadable module.
However, the two recently added ptp_get_vclocks_index() and ptp_convert_timestamp() interfaces are only called from builtin code with ethtool and socket timestamps, so keep the current behavior by stubbing those out completely when PTP is in a loadable module. This should be addressed properly in a follow-up.
As Richard suggested, we may want to actually turn PTP support into a 'bool' option later on, preventing it from being a loadable module altogether, which would be one way to solve the problem with the ethtool interface.
---
Changes in v3:
- rewrite to introduce a new PTP_1588_CLOCK_OPTIONAL symbol
- use it for all driver, not just Intel's
- change IS_REACHABLE() to IS_ENABLED() in the header
Changes in v2:
- include a missing patch hunk
- link to a previous discussion with Richard Cochran
--- drivers/net/dsa/mv88e6xxx/Kconfig | 1 + drivers/net/dsa/ocelot/Kconfig | 2 + drivers/net/ethernet/amd/Kconfig | 2 +- drivers/net/ethernet/broadcom/Kconfig | 6 +-- drivers/net/ethernet/cadence/Kconfig | 1 + drivers/net/ethernet/cavium/Kconfig | 4 +- drivers/net/ethernet/freescale/Kconfig | 2 +- drivers/net/ethernet/hisilicon/Kconfig | 2 +- drivers/net/ethernet/intel/Kconfig | 14 +++--- drivers/net/ethernet/mellanox/mlx4/Kconfig | 2 +- .../net/ethernet/mellanox/mlx5/core/Kconfig | 2 +- drivers/net/ethernet/mellanox/mlxsw/Kconfig | 2 +- drivers/net/ethernet/mscc/Kconfig | 2 + drivers/net/ethernet/oki-semi/pch_gbe/Kconfig | 1 + drivers/net/ethernet/pensando/Kconfig | 2 +- drivers/net/ethernet/qlogic/Kconfig | 2 +- drivers/net/ethernet/renesas/Kconfig | 2 +- drivers/net/ethernet/samsung/Kconfig | 2 +- drivers/net/ethernet/sfc/Kconfig | 2 +- drivers/net/ethernet/stmicro/stmmac/Kconfig | 2 +- drivers/net/phy/Kconfig | 2 + drivers/ptp/Kconfig | 15 +++++- drivers/ptp/ptp_vclock.c | 2 + include/linux/ptp_clock_kernel.h | 48 +++++++++++-------- 24 files changed, 77 insertions(+), 45 deletions(-)
drivers/net/dsa/mv88e6xxx/Kconfig +1 -0
@@ -2,6 +2,7 @@ config NET_DSA_MV88E6XXX tristate "Marvell 88E6xxx Ethernet switch fabric support" depends on NET_DSA + depends on PTP_1588_CLOCK_OPTIONAL select IRQ_DOMAIN select NET_DSA_TAG_EDSA select NET_DSA_TAG_DSA
drivers/net/dsa/ocelot/Kconfig +2 -0
@@ -5,6 +5,7 @@ config NET_DSA_MSCC_FELIX depends on NET_VENDOR_MICROSEMI depends on NET_VENDOR_FREESCALE depends on HAS_IOMEM + depends on PTP_1588_CLOCK_OPTIONAL select MSCC_OCELOT_SWITCH_LIB select NET_DSA_TAG_OCELOT_8021Q select NET_DSA_TAG_OCELOT
@@ -19,6 +20,7 @@ config NET_DSA_MSCC_SEVILLE depends on NET_DSA depends on NET_VENDOR_MICROSEMI depends on HAS_IOMEM + depends on PTP_1588_CLOCK_OPTIONAL select MSCC_OCELOT_SWITCH_LIB select NET_DSA_TAG_OCELOT_8021Q select NET_DSA_TAG_OCELOT
drivers/net/ethernet/amd/Kconfig +1 -1
@@ -170,11 +170,11 @@ config AMD_XGBE tristate "AMD 10GbE Ethernet driver" depends on ((OF_NET && OF_ADDRESS) || ACPI || PCI) && HAS_IOMEM depends on X86 || ARM64 || COMPILE_TEST + depends on PTP_1588_CLOCK_OPTIONAL select BITREVERSE select CRC32 select PHYLIB select AMD_XGBE_HAVE_ECC if X86 - imply PTP_1588_CLOCK help This driver supports the AMD 10GbE Ethernet device found on an AMD SoC.
drivers/net/ethernet/broadcom/Kconfig +3 -3
@@ -122,8 +122,8 @@ config SB1250_MAC config TIGON3 tristate "Broadcom Tigon3 support" depends on PCI + depends on PTP_1588_CLOCK_OPTIONAL select PHYLIB - imply PTP_1588_CLOCK help This driver supports Broadcom Tigon3 based gigabit Ethernet cards.
@@ -140,7 +140,7 @@ config TIGON3_HWMON config BNX2X tristate "Broadcom NetXtremeII 10Gb support" depends on PCI - imply PTP_1588_CLOCK + depends on PTP_1588_CLOCK_OPTIONAL select FW_LOADER select ZLIB_INFLATE select LIBCRC32C
@@ -206,7 +206,7 @@ config SYSTEMPORT config BNXT tristate "Broadcom NetXtreme-C/E support" depends on PCI - imply PTP_1588_CLOCK + depends on PTP_1588_CLOCK_OPTIONAL select FW_LOADER select LIBCRC32C select NET_DEVLINK
drivers/net/ethernet/cadence/Kconfig +1 -0
@@ -22,6 +22,7 @@ if NET_VENDOR_CADENCE config MACB tristate "Cadence MACB/GEM support" depends on HAS_DMA && COMMON_CLK + depends on PTP_1588_CLOCK_OPTIONAL select PHYLINK select CRC32 help
drivers/net/ethernet/cavium/Kconfig +2 -2
@@ -66,7 +66,7 @@ config LIQUIDIO tristate "Cavium LiquidIO support" depends on 64BIT && PCI depends on PCI - imply PTP_1588_CLOCK + depends on PTP_1588_CLOCK_OPTIONAL select FW_LOADER select LIBCRC32C select NET_DEVLINK
@@ -91,7 +91,7 @@ config OCTEON_MGMT_ETHERNET config LIQUIDIO_VF tristate "Cavium LiquidIO VF support" depends on 64BIT && PCI_MSI - imply PTP_1588_CLOCK + depends on PTP_1588_CLOCK_OPTIONAL help This driver supports Cavium LiquidIO Intelligent Server Adapter based on CN23XX chips.
drivers/net/ethernet/freescale/Kconfig +1 -1
@@ -25,10 +25,10 @@ config FEC depends on (M523x || M527x || M5272 || M528x || M520x || M532x || \ ARCH_MXC || SOC_IMX28 || COMPILE_TEST) default ARCH_MXC || SOC_IMX28 if ARM + depends on PTP_1588_CLOCK_OPTIONAL select CRC32 select PHYLIB imply NET_SELFTESTS - imply PTP_1588_CLOCK help Say Y here if you want to use the built-in 10/100 Fast ethernet controller on some Motorola ColdFire and Freescale i.MX processors.
drivers/net/ethernet/hisilicon/Kconfig +1 -1
@@ -103,7 +103,7 @@ config HNS3_HCLGE tristate "Hisilicon HNS3 HCLGE Acceleration Engine & Compatibility Layer Support" default m depends on PCI_MSI - imply PTP_1588_CLOCK + depends on PTP_1588_CLOCK_OPTIONAL help This selects the HNS3_HCLGE network acceleration engine & its hardware compatibility layer. The engine would be used in Hisilicon hip08 family of
drivers/net/ethernet/intel/Kconfig +7 -7
@@ -58,8 +58,8 @@ config E1000 config E1000E tristate "Intel(R) PRO/1000 PCI-Express Gigabit Ethernet support" depends on PCI && (!SPARC32 || BROKEN) + depends on PTP_1588_CLOCK_OPTIONAL select CRC32 - imply PTP_1588_CLOCK help This driver supports the PCI-Express Intel(R) PRO/1000 gigabit ethernet family of adapters. For PCI or PCI-X e1000 adapters,
@@ -87,8 +87,8 @@ config E1000E_HWTS config IGB tristate "Intel(R) 82575/82576 PCI-Express Gigabit Ethernet support" depends on PCI - imply PTP_1588_CLOCK - select I2C + depends on PTP_1588_CLOCK_OPTIONAL + depends on I2C select I2C_ALGOBIT help This driver supports Intel(R) 82575/82576 gigabit ethernet family of
@@ -159,9 +159,9 @@ config IXGB config IXGBE tristate "Intel(R) 10GbE PCI Express adapters support" depends on PCI + depends on PTP_1588_CLOCK_OPTIONAL select MDIO select PHYLIB - imply PTP_1588_CLOCK help This driver supports Intel(R) 10GbE PCI Express family of adapters. For more information on how to identify your adapter, go
@@ -239,7 +239,7 @@ config IXGBEVF_IPSEC config I40E tristate "Intel(R) Ethernet Controller XL710 Family support" - imply PTP_1588_CLOCK + depends on PTP_1588_CLOCK_OPTIONAL depends on PCI select AUXILIARY_BUS help
@@ -295,11 +295,11 @@ config ICE tristate "Intel(R) Ethernet Connection E800 Series Support" default n depends on PCI_MSI + depends on PTP_1588_CLOCK_OPTIONAL select AUXILIARY_BUS select DIMLIB select NET_DEVLINK select PLDMFW - imply PTP_1588_CLOCK help This driver supports Intel(R) Ethernet Connection E800 Series of devices. For more information on how to identify your adapter, go
@@ -317,7 +317,7 @@ config FM10K tristate "Intel(R) FM10000 Ethernet Switch Host Interface Support" default n depends on PCI_MSI - imply PTP_1588_CLOCK + depends on PTP_1588_CLOCK_OPTIONAL help This driver supports Intel(R) FM10000 Ethernet Switch Host Interface. For more information on how to identify your adapter,
drivers/net/ethernet/mellanox/mlx4/Kconfig +1 -1
@@ -6,8 +6,8 @@ config MLX4_EN tristate "Mellanox Technologies 1/10/40Gbit Ethernet support" depends on PCI && NETDEVICES && ETHERNET && INET + depends on PTP_1588_CLOCK_OPTIONAL select MLX4_CORE - imply PTP_1588_CLOCK help This driver supports Mellanox Technologies ConnectX Ethernet devices.
drivers/net/ethernet/mellanox/mlx5/core/Kconfig +1 -1
@@ -10,7 +10,7 @@ config MLX5_CORE select NET_DEVLINK depends on VXLAN || !VXLAN depends on MLXFW || !MLXFW - depends on PTP_1588_CLOCK || !PTP_1588_CLOCK + depends on PTP_1588_CLOCK_OPTIONAL depends on PCI_HYPERV_INTERFACE || !PCI_HYPERV_INTERFACE help Core driver for low level functionality of the ConnectX-4 and
drivers/net/ethernet/mellanox/mlxsw/Kconfig +1 -1
@@ -58,10 +58,10 @@ config MLXSW_SPECTRUM depends on NET_IPGRE || NET_IPGRE=n depends on IPV6_GRE || IPV6_GRE=n depends on VXLAN || VXLAN=n + depends on PTP_1588_CLOCK_OPTIONAL select GENERIC_ALLOCATOR select PARMAN select OBJAGG - imply PTP_1588_CLOCK select NET_PTP_CLASSIFY if PTP_1588_CLOCK default m help
drivers/net/ethernet/mscc/Kconfig +2 -0
@@ -13,6 +13,7 @@ if NET_VENDOR_MICROSEMI # Users should depend on NET_SWITCHDEV, HAS_IOMEM, BRIDGE config MSCC_OCELOT_SWITCH_LIB + depends on PTP_1588_CLOCK_OPTIONAL select NET_DEVLINK select REGMAP_MMIO select PACKING
@@ -24,6 +25,7 @@ config MSCC_OCELOT_SWITCH_LIB config MSCC_OCELOT_SWITCH tristate "Ocelot switch driver" + depends on PTP_1588_CLOCK_OPTIONAL depends on BRIDGE || BRIDGE=n depends on NET_SWITCHDEV depends on HAS_IOMEM
drivers/net/ethernet/oki-semi/pch_gbe/Kconfig +1 -0
@@ -6,6 +6,7 @@ config PCH_GBE tristate "OKI SEMICONDUCTOR IOH(ML7223/ML7831) GbE" depends on PCI && (X86_32 || COMPILE_TEST) + depends on PTP_1588_CLOCK select MII select PTP_1588_CLOCK_PCH select NET_PTP_CLASSIFY
drivers/net/ethernet/pensando/Kconfig +1 -1
@@ -20,7 +20,7 @@ if NET_VENDOR_PENSANDO config IONIC tristate "Pensando Ethernet IONIC Support" depends on 64BIT && PCI - depends on PTP_1588_CLOCK || !PTP_1588_CLOCK + depends on PTP_1588_CLOCK_OPTIONAL select NET_DEVLINK select DIMLIB help
drivers/net/ethernet/qlogic/Kconfig +1 -1
@@ -99,7 +99,7 @@ config QED_SRIOV config QEDE tristate "QLogic QED 25/40/100Gb Ethernet NIC" depends on QED - imply PTP_1588_CLOCK + depends on PTP_1588_CLOCK_OPTIONAL help This enables the support for Marvell FastLinQ adapters family, ethernet driver.
drivers/net/ethernet/renesas/Kconfig +1 -1
@@ -32,11 +32,11 @@ config SH_ETH config RAVB tristate "Renesas Ethernet AVB support" depends on ARCH_RENESAS || COMPILE_TEST + depends on PTP_1588_CLOCK_OPTIONAL select CRC32 select MII select MDIO_BITBANG select PHYLIB - imply PTP_1588_CLOCK help Renesas Ethernet AVB device driver. This driver supports the following SoCs:
drivers/net/ethernet/samsung/Kconfig +1 -1
@@ -20,9 +20,9 @@ if NET_VENDOR_SAMSUNG config SXGBE_ETH tristate "Samsung 10G/2.5G/1G SXGBE Ethernet driver" depends on HAS_IOMEM && HAS_DMA + depends on PTP_1588_CLOCK_OPTIONAL select PHYLIB select CRC32 - imply PTP_1588_CLOCK help This is the driver for the SXGBE 10G Ethernet IP block found on Samsung platforms.
drivers/net/ethernet/sfc/Kconfig +1 -1
@@ -19,9 +19,9 @@ if NET_VENDOR_SOLARFLARE config SFC tristate "Solarflare SFC9000/SFC9100/EF100-family support" depends on PCI + depends on PTP_1588_CLOCK_OPTIONAL select MDIO select CRC32 - imply PTP_1588_CLOCK help This driver supports 10/40-gigabit Ethernet cards based on the Solarflare SFC9000-family and SFC9100-family controllers.
drivers/net/ethernet/stmicro/stmmac/Kconfig +1 -1
@@ -2,12 +2,12 @@ config STMMAC_ETH tristate "STMicroelectronics Multi-Gigabit Ethernet driver" depends on HAS_IOMEM && HAS_DMA + depends on PTP_1588_CLOCK_OPTIONAL select MII select PCS_XPCS select PAGE_POOL select PHYLINK select CRC32 - imply PTP_1588_CLOCK select RESET_CONTROLLER help This is the driver for the Ethernet IPs built around a
drivers/net/phy/Kconfig +2 -0
@@ -236,6 +236,7 @@ config MICROCHIP_T1_PHY config MICROSEMI_PHY tristate "Microsemi PHYs" depends on MACSEC || MACSEC=n + depends on PTP_1588_CLOCK_OPTIONAL || !NETWORK_PHY_TIMESTAMPING select CRYPTO_LIB_AES if MACSEC help Currently supports VSC8514, VSC8530, VSC8531, VSC8540 and VSC8541 PHYs
@@ -253,6 +254,7 @@ config NATIONAL_PHY config NXP_C45_TJA11XX_PHY tristate "NXP C45 TJA11XX PHYs" + depends on PTP_1588_CLOCK_OPTIONAL help Enable support for NXP C45 TJA11XX PHYs. Currently supports only the TJA1103 PHY.
drivers/ptp/Kconfig +14 -1
@@ -8,6 +8,7 @@ menu "PTP clock support" config PTP_1588_CLOCK tristate "PTP clock support" depends on NET && POSIX_TIMERS + default ETHERNET select PPS select NET_PTP_CLASSIFY help
@@ -26,6 +27,18 @@ config PTP_1588_CLOCK To compile this driver as a module, choose M here: the module will be called ptp. +config PTP_1588_CLOCK_OPTIONAL + tristate + default y if PTP_1588_CLOCK=n + default PTP_1588_CLOCK + help + Drivers that can optionally use the PTP_1588_CLOCK framework + should depend on this symbol to prevent them from being built + into vmlinux while the PTP support itself is in a loadable + module. + If PTP support is disabled, this dependency will still be + met, and drivers refer to dummy helpers. + config PTP_1588_CLOCK_DTE tristate "Broadcom DTE as PTP clock" depends on PTP_1588_CLOCK
@@ -91,7 +104,7 @@ config PTP_1588_CLOCK_PCH tristate "Intel PCH EG20T as PTP clock" depends on X86_32 || COMPILE_TEST depends on HAS_IOMEM && NET - imply PTP_1588_CLOCK + depends on PTP_1588_CLOCK help This driver adds support for using the PCH EG20T as a PTP clock. The hardware supports time stamping of PTP packets
drivers/ptp/ptp_vclock.c +2 -0
@@ -149,6 +149,7 @@ void ptp_vclock_unregister(struct ptp_vclock *vclock)kfree (vclock); } +#if IS_BUILTIN (CONFIG_PTP_1588_CLOCK )int ptp_get_vclocks_index (int pclock_index,int **vclock_index) {char name[PTP_CLOCK_NAME_LEN ] ="" ;
@@ -217,3 +218,4 @@ void ptp_convert_timestamp(struct skb_shared_hwtstamps *hwtstamps, hwtstamps->hwtstamp = ns_to_ktime(ns); }EXPORT_SYMBOL (ptp_convert_timestamp); +#endif
include/linux/ptp_clock_kernel.h +28 -20
67 lines not on this page · raw message
Signature · 1 line
2.29.2
Quoted 5 lines · Arnd Bergmann
Since this should cover the dependencies correctly, the
IS_REACHABLE()hack in the header is no longer needed now, and can be turned back into a normalIS_ENABLED()check. Any driver that gets the dependency wrong will now cause a link time failure rather than being unable to use PTP support when that is in a loadable module.
yes!
Quoted 3 lines
Changes in v3:
- rewrite to introduce a newPTP_1588_CLOCK_OPTIONALsymbol
- use it for all driver, not just Intel's
Thanks for following up.
Quoted 15 lines
drivers/net/ethernet/intel/Kconfig +1
-1@@ -58,8 +58,8 @@ config E1000 config E1000E tristate "Intel(R) PRO/1000 PCI-Express Gigabit Ethernet support" depends on PCI && (!SPARC32 || BROKEN) + depends on PTP_1588_CLOCK_OPTIONAL select CRC32 - imply PTP_1588_CLOCK help This driver supports the PCI-Express Intel(R) PRO/1000 gigabit ethernet family of adapters. For PCI or PCI-X e1000 adapters,
drivers/net/ethernet/intel/Kconfig config E1000E_HWTS · line 91
Quoted 7 lines
drivers/net/ethernet/intel/Kconfig +2
-2@@ -87,8 +87,8 @@ config E1000E_HWTS config IGB tristate "Intel(R) 82575/82576 PCI-Express Gigabit Ethernet support" depends on PCI - imply PTP_1588_CLOCK - select I2C + depends on PTP_1588_CLOCK_OPTIONAL + depends on I2C
This little i2c bit sneaks in, but I guess you considered any possible trouble with it?
Quoted 11 lines · Richard Cochran
drivers/net/ethernet/intel/Kconfig config E1000E_HWTS · line 91
Quoted 8 lines
drivers/net/ethernet/intel/Kconfig +2
-2@@ -87,8 +87,8 @@ config E1000E_HWTS config IGB tristate "Intel(R) 82575/82576 PCI-Express Gigabit Ethernet support" depends on PCI - imply PTP_1588_CLOCK - select I2C + depends on PTP_1588_CLOCK_OPTIONAL + depends on I2CThis little i2c bit sneaks in, but I guess you considered any possible trouble with it?
Good catch!
I did need this with v2, as it was causing a circular dependency against (IIRC) CONFIG_MLXSW_I2C, but I'm fairly sure it's not needed any more after everything else uses 'depends on' now.
I'm happy to resend a v4 without that change, as it doesn't belong in here, or we just leave it because it is correct after all, depending on what the Intel ethernet people prefer.
Quoted 1 line
Acked-by Richard Cochran <[email protected]>
Thanks,
Arnd
Quoted 46 lines
-----Original Message-----
From: Arnd Bergmann <[email protected]>
Sent: Wednesday, August 04, 2021 8:21 AM
To: Richard Cochran <[email protected]>
Cc: David S. Miller <[email protected]>; Jakub Kicinski <[email protected]>; Keller, Jacob E <[email protected]>; Arnd Bergmann <[email protected]>; Andrew Lunn <[email protected]>; Vivien Didelot <[email protected]>; Florian Fainelli <[email protected]>; Vladimir Oltean <[email protected]>; Claudiu Manoil <[email protected]>; Alexandre Belloni <[email protected]>; Microchip Linux Driver Support <[email protected]>; Nicolas Ferre <[email protected]>; Claudiu Beznea <[email protected]>; Yisen Zhuang <[email protected]>; Salil Mehta <[email protected]>; Brandeburg, Jesse <[email protected]>; Nguyen, Anthony L <[email protected]>; Tariq Toukan <[email protected]>; Saeed Mahameed <[email protected]>; Leon Romanovsky <[email protected]>; Jiri Pirko <[email protected]>; Ido Schimmel <[email protected]>; Shannon Nelson <[email protected]>; drivers at pensando.io; Sergei Shtylyov <[email protected]>; Edward Cree <[email protected]>; Martin Habets <[email protected]>; Giuseppe Cavallaro <[email protected]>; Alexandre Torgue <[email protected]>; Jose Abreu <[email protected]>; Heiner Kallweit <[email protected]>; Russell King <[email protected]>; Yangbo Lu <[email protected]>; Randy Dunlap <[email protected]>; Simon Horman <[email protected]>; Networking <[email protected]>; Linux Kernel Mailing List <[email protected]>; Intel Wired LAN <intel- wired-lan at lists.osuosl.org>
Subject: Re: [PATCH net-next v3] ethernet: fixPTP_1588_CLOCKdependenciesQuoted 11 lines · Richard Cochran
drivers/net/ethernet/intel/Kconfig config E1000E_HWTS · line 91
Quoted 8 lines
drivers/net/ethernet/intel/Kconfig +2
-2@@ -87,8 +87,8 @@ config E1000E_HWTS config IGB tristate "Intel(R) 82575/82576 PCI-Express Gigabit Ethernet support" depends on PCI - imply PTP_1588_CLOCK - select I2C + depends on PTP_1588_CLOCK_OPTIONAL + depends on I2CThis little i2c bit sneaks in, but I guess you considered any possible trouble with it?
Good catch!
Quoted 8 lines
I did need this with v2, as it was causing a circular dependency against (IIRC)
CONFIG_MLXSW_I2C, but I'm fairly sure it's not needed any more after everything else uses 'depends on' now.I'm happy to resend a v4 without that change, as it doesn't belong in here, or we just leave it because it is correct after all, depending on what the Intel ethernet people prefer.
I'm fine with keeping it in.
Quoted 5 lines
Quoted 1 line
Acked-by Richard Cochran <[email protected]>Thanks,
Arnd
Quoted 89 lines
-----Original Message-----
From: Arnd Bergmann <[email protected]>
Sent: Wednesday, August 04, 2021 4:57 AM
To: David S. Miller <[email protected]>; Jakub Kicinski <[email protected]>; Richard Cochran <[email protected]>; Keller, Jacob E <[email protected]>
Cc: Arnd Bergmann <[email protected]>; Andrew Lunn <[email protected]>; Vivien Didelot <[email protected]>; Florian Fainelli <[email protected]>; Vladimir Oltean <[email protected]>; Claudiu Manoil <[email protected]>; Alexandre Belloni <[email protected]>; UNGLinuxDriver at microchip.com; Nicolas Ferre <[email protected]>; Claudiu Beznea <[email protected]>; Yisen Zhuang <[email protected]>; Salil Mehta <[email protected]>; Brandeburg, Jesse <[email protected]>; Nguyen, Anthony L <[email protected]>; Tariq Toukan <[email protected]>; Saeed Mahameed <[email protected]>; Leon Romanovsky <[email protected]>; Jiri Pirko <[email protected]>; Ido Schimmel <[email protected]>; Shannon Nelson <[email protected]>; drivers at pensando.io; Sergei Shtylyov <[email protected]>; Edward Cree <[email protected]>; Martin Habets <[email protected]>; Giuseppe Cavallaro <[email protected]>; Alexandre Torgue <[email protected]>; Jose Abreu <[email protected]>; Heiner Kallweit <[email protected]>; Russell King <[email protected]>; Yangbo Lu <[email protected]>; Randy Dunlap <[email protected]>; Simon Horman <[email protected]>; netdev at vger.kernel.org; linux- kernel at vger.kernel.org; intel-wired-lan at lists.osuosl.org
Subject: [PATCH net-next v3] ethernet: fixPTP_1588_CLOCKdependenciesFrom: Arnd Bergmann <[email protected]>
The 'imply' keyword does not do what most people think it does, it only politely asks
Kconfigto turn on another symbol, but does not prevent it from being disabled manually or built as a loadable module when the user is built-in. In the ICE driver, the latter now causes a link failure:aarch64-linux-ld:
drivers/net/ethernet/intel/ice/ice_main.o: in function `ice_eth_ioctl':
ice_main.c:(.text+0x13b0): undefined reference toice_ptp_get_ts_config' ice_main.c:(.text+0x13b0): relocation truncated to fit: R_AARCH64_CALL26 against undefined symbolice_ptp_get_ts_config'
aarch64-linux-ld: ice_main.c:(.text+0x13bc): undefined reference to `ice_ptp_set_ts_config'
ice_main.c:(.text+0x13bc): relocation truncated to fit:R_AARCH64_CALL26against undefined symbol `ice_ptp_set_ts_config'
aarch64-linux-ld:drivers/net/ethernet/intel/ice/ice_main.o: in function `ice_prepare_for_reset':
ice_main.c:(.text+0x31fc): undefined reference toice_ptp_release' ice_main.c:(.text+0x31fc): relocation truncated to fit: R_AARCH64_CALL26 against undefined symbolice_ptp_release'
aarch64-linux-ld:drivers/net/ethernet/intel/ice/ice_main.o: in function `ice_rebuild':This is a recurring problem in many drivers, and we have discussed it several times befores, without reaching a consensus. I'm providing a link to the previous email thread for reference, which discusses some related problems.
To solve the dependency issue better than the 'imply' keyword, introduce a separate
Kconfigsymbol "CONFIG_PTP_1588_CLOCK_OPTIONAL" that any driver can depend on if it is able to use PTP support when available, but works fine without it. WheneverCONFIG_PTP_1588_CLOCK=m, those drivers are then prevented from being built-in, the same way as with a 'depends onPTP_1588_CLOCK|| !PTP_1588_CLOCK' dependency that does the same trick, but that can be rather confusing when you first see it.Since this should cover the dependencies correctly, the
IS_REACHABLE()hack in the header is no longer needed now, and can be turned back into a normalIS_ENABLED()check. Any driver that gets the dependency wrong will now cause a link time failure rather than being unable to use PTP support when that is in a loadable module.However, the two recently added
ptp_get_vclocks_index()andptp_convert_timestamp()interfaces are only called from builtin code with ethtool and socket timestamps, so keep the current behavior by stubbing those out completely when PTP is in a loadable module. This should be addressed properly in a follow-up.As Richard suggested, we may want to actually turn PTP support into a 'bool' option later on, preventing it from being a loadable module altogether, which would be one way to solve the problem with the ethtool interface.
Fixes 06c16d89d2cb ("ice: register 1588 PTP clock device object for E810 devices")Link:
https://lore.kernel.org/netdev/CAK8P3a06enZOf=XyZ+zcAwBczv41UuCTz+=0FMf
2gBz1_cOnZQ at mail.gmail.com/
Quoted 7 lines
i_0BZzf9a9RfPTX3ppSwu9WZXwat mail.gmail.com/arnd at kernel.org/
Cc Richard Cochran <[email protected]>Signed-off-by Arnd Bergmann <[email protected]>---
Thanks for the "_OPTIONAL" symbol, I definitely find that more readable than the "depends on A || !A".
drivers/net/ethernet/oki-semi/pch_gbe/Kconfig line 12
Quoted 13 lines
drivers/net/ethernet/oki-semi/pch_gbe/Kconfig +1
-0@@ -6,6 +6,7 @@ config PCH_GBE tristate "OKI SEMICONDUCTOR IOH(ML7223/ML7831) GbE" depends on PCI && (X86_32 || COMPILE_TEST) + depends on PTP_1588_CLOCK select MII select PTP_1588_CLOCK_PCH select NET_PTP_CLASSIFY
I did notice this one driver which now directly depends on PTP_1558_CLOCK, but I suspect that's because it actually doesn't work if you disable PTP?
Quoted 8 lines · Jacob E
Quoted 8 lines
drivers/net/ethernet/oki-semi/pch_gbe/Kconfig +0
-0@@ -6,6 +6,7 @@ config PCH_GBE tristate "OKI SEMICONDUCTOR IOH(ML7223/ML7831) GbE"
Quoted 8 lines
drivers/net/ethernet/oki-semi/pch_gbe/Kconfig line 12
Quoted 5 lines
drivers/net/ethernet/oki-semi/pch_gbe/Kconfig +1
-0depends on PCI && (X86_32 || COMPILE_TEST) + depends on PTP_1588_CLOCK select MII select PTP_1588_CLOCK_PCH select NET_PTP_CLASSIFYI did notice this one driver which now directly depends on
PTP_1558_CLOCK, but I suspect that's because it actually doesn't work if you disable PTP?
Yes, it's the 'select PTP_1588_CLOCK_PCH' and 'select NET_PTP_CLASSIFY' that actually need it.
Arnd
Quoted 78 lines · Arnd Bergmann
From: Arnd Bergmann <[email protected]>
The 'imply' keyword does not do what most people think it does, it only politely asks
Kconfigto turn on another symbol, but does not prevent it from being disabled manually or built as a loadable module when the user is built-in. In the ICE driver, the latter now causes a link failure:aarch64-linux-ld:
drivers/net/ethernet/intel/ice/ice_main.o: in function `ice_eth_ioctl':
ice_main.c:(.text+0x13b0): undefined reference to `ice_ptp_get_ts_config'
ice_main.c:(.text+0x13b0): relocation truncated to fit:R_AARCH64_CALL26against undefined symbol `ice_ptp_get_ts_config'
aarch64-linux-ld: ice_main.c:(.text+0x13bc): undefined reference toice_ptp_set_ts_config' ice_main.c:(.text+0x13bc): relocation truncated to fit: R_AARCH64_CALL26 against undefined symbolice_ptp_set_ts_config'
aarch64-linux-ld:drivers/net/ethernet/intel/ice/ice_main.o: in function `ice_prepare_for_reset':
ice_main.c:(.text+0x31fc): undefined reference to `ice_ptp_release'
ice_main.c:(.text+0x31fc): relocation truncated to fit:R_AARCH64_CALL26against undefined symbol `ice_ptp_release'
aarch64-linux-ld:drivers/net/ethernet/intel/ice/ice_main.o: in function `ice_rebuild':This is a recurring problem in many drivers, and we have discussed it several times befores, without reaching a consensus. I'm providing a link to the previous email thread for reference, which discusses some related problems.
To solve the dependency issue better than the 'imply' keyword, introduce a separate
Kconfigsymbol "CONFIG_PTP_1588_CLOCK_OPTIONAL" that any driver can depend on if it is able to use PTP support when available, but works fine without it. WheneverCONFIG_PTP_1588_CLOCK=m, those drivers are then prevented from being built-in, the same way as with a 'depends onPTP_1588_CLOCK|| !PTP_1588_CLOCK' dependency that does the same trick, but that can be rather confusing when you first see it.Since this should cover the dependencies correctly, the
IS_REACHABLE()hack in the header is no longer needed now, and can be turned back into a normalIS_ENABLED()check. Any driver that gets the dependency wrong will now cause a link time failure rather than being unable to use PTP support when that is in a loadable module.However, the two recently added
ptp_get_vclocks_index()andptp_convert_timestamp()interfaces are only called from builtin code with ethtool and socket timestamps, so keep the current behavior by stubbing those out completely when PTP is in a loadable module. This should be addressed properly in a follow-up.As Richard suggested, we may want to actually turn PTP support into a 'bool' option later on, preventing it from being a loadable module altogether, which would be one way to solve the problem with the ethtool interface.
Fixes 06c16d89d2cb ("ice: register 1588 PTP clock device object for E810 devices")Link https://lore.kernel.org/netdev/CAK8P3a06enZOf=XyZ+zcAwBczv41UuCTz+=0FMf2gBz1_cOnZQ at mail.gmail.com/Link https://lore.kernel.org/netdev/CAK8P3a3=eOxE-K25754+fB_-i_0BZzf9a9RfPTX3ppSwu9WZXw at mail.gmail.com/Link https://lore.kernel.org/netdev/20210726084540.3282344-1-arnd at kernel.org/Cc Richard Cochran <[email protected]>Signed-off-by Arnd Bergmann <[email protected]>---
Changes in v3:
- rewrite to introduce a newPTP_1588_CLOCK_OPTIONALsymbol
- use it for all driver, not just Intel's
- changeIS_REACHABLE()toIS_ENABLED()in the headerChanges in v2:
- include a missing patch hunk
- link to a previous discussion with Richard Cochran--- drivers/net/dsa/mv88e6xxx/Kconfig | 1 + drivers/net/dsa/ocelot/Kconfig | 2 + drivers/net/ethernet/amd/Kconfig | 2 +- drivers/net/ethernet/broadcom/Kconfig | 6 +-- drivers/net/ethernet/cadence/Kconfig | 1 + drivers/net/ethernet/cavium/Kconfig | 4 +- drivers/net/ethernet/freescale/Kconfig | 2 +- drivers/net/ethernet/hisilicon/Kconfig | 2 +- drivers/net/ethernet/intel/Kconfig | 14 +++--- drivers/net/ethernet/mellanox/mlx4/Kconfig | 2 +- .../net/ethernet/mellanox/mlx5/core/Kconfig | 2 +- drivers/net/ethernet/mellanox/mlxsw/Kconfig | 2 +- drivers/net/ethernet/mscc/Kconfig | 2 + drivers/net/ethernet/oki-semi/pch_gbe/Kconfig | 1 + drivers/net/ethernet/pensando/Kconfig | 2 +-
[...]
Quoted 7 lines
drivers/net/ethernet/qlogic/Kconfig | 2 +-drivers/net/ethernet/pensando/Kconfig +0
-0
drivers/net/ethernet/pensando/Kconfig if NET_VENDOR_PENSANDO · line 26
Quoted 8 lines
drivers/net/ethernet/pensando/Kconfig +1
-1@@ -20,7 +20,7 @@ if NET_VENDOR_PENSANDO config IONIC tristate "Pensando Ethernet IONIC Support" depends on 64BIT && PCI - depends on PTP_1588_CLOCK || !PTP_1588_CLOCK + depends on PTP_1588_CLOCK_OPTIONAL select NET_DEVLINK select DIMLIB help