| Message ID | 20230405063144.36231-1-glipus@gmail.com (mailing list archive) |
|---|---|
| State | RFC |
| Delegated to: | Netdev Maintainers |
| Headers | show |
| Series | [RFC,v3,1/5] Add NDOs for hardware timestamp get/set | expand |
On Wed, Apr 05, 2023 at 12:31:44AM -0600, Maxim Georgiev wrote: > Current NIC driver API demands drivers supporting hardware timestamping > to implement handling logic for SIOCGHWTSTAMP/SIOCSHWTSTAMP IOCTLs. > Handling these IOCTLs requires dirivers to implement request parameter > structure translation between user and kernel address spaces, handling > possible translation failures, etc. This translation code is pretty much > identical across most of the NIC drivers that support SIOCGHWTSTAMP/ > SIOCSHWTSTAMP. > This patch extends NDO functiuon set with ndo_hwtstamp_get/set > functions, implements SIOCGHWTSTAMP/SIOCSHWTSTAMP IOCTL translation > to ndo_hwtstamp_get/set function calls including parameter structure > translation and translation error handling. > > This patch is sent out as RFC. > It still pending on basic testing. > > Suggested-by: Jakub Kicinski <kuba@kernel.org> > Signed-off-by: Maxim Georgiev <glipus@gmail.com> > --- > Changes in v3: > - Moved individual driver conversions to separate patches > > Changes in v2: > - Introduced kernel_hwtstamp_config structure > - Added netlink_ext_ack* and kernel_hwtstamp_config* as NDO hw timestamp > function parameters > - Reodered function variable declarations in dev_hwtstamp() > - Refactored error handling logic in dev_hwtstamp() > - Split dev_hwtstamp() into GET and SET versions > - Changed net_hwtstamp_validate() to accept struct hwtstamp_config * > as a parameter > --- > include/linux/net_tstamp.h | 8 ++++++++ > include/linux/netdevice.h | 16 ++++++++++++++++ > net/core/dev_ioctl.c | 36 ++++++++++++++++++++++++++++++++++-- > 3 files changed, 58 insertions(+), 2 deletions(-) > > diff --git a/include/linux/net_tstamp.h b/include/linux/net_tstamp.h > index fd67f3cc0c4b..063260475e77 100644 > --- a/include/linux/net_tstamp.h > +++ b/include/linux/net_tstamp.h > @@ -30,4 +30,12 @@ static inline void hwtstamp_config_to_kernel(struct kernel_hwtstamp_config *kern > kernel_cfg->rx_filter = cfg->rx_filter; > } > > +static inline void hwtstamp_kernel_to_config(struct hwtstamp_config *cfg, > + const struct kernel_hwtstamp_config *kernel_cfg) The reason why I suggested the name "hwtstamp_config_from_kernel()" was to not break apart "hwtstamp" and "config", which together form the name of one structure (hwtstamp_config). > +{ > + cfg->flags = kernel_cfg->flags; > + cfg->tx_type = kernel_cfg->tx_type; > + cfg->rx_filter = kernel_cfg->rx_filter; > +} > + > #endif /* _LINUX_NET_TIMESTAMPING_H_ */ > diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h > index a740be3bb911..8356002d0ac0 100644 > --- a/include/linux/netdevice.h > +++ b/include/linux/netdevice.h > @@ -57,6 +57,7 @@ > struct netpoll_info; > struct device; > struct ethtool_ops; > +struct kernel_hwtstamp_config; > struct phy_device; > struct dsa_port; > struct ip_tunnel_parm; > @@ -1412,6 +1413,15 @@ struct netdev_net_notifier { > * Get hardware timestamp based on normal/adjustable time or free running > * cycle counter. This function is required if physical clock supports a > * free running cycle counter. > + * int (*ndo_hwtstamp_get)(struct net_device *dev, > + * struct kernel_hwtstamp_config *kernel_config, > + * struct netlink_ext_ack *extack); > + * Get hardware timestamping parameters currently configured for NIC > + * device. > + * int (*ndo_hwtstamp_set)(struct net_device *dev, > + * struct kernel_hwtstamp_config *kernel_config, > + * struct netlink_ext_ack *extack); > + * Set hardware timestamping parameters for NIC device. > */ > struct net_device_ops { > int (*ndo_init)(struct net_device *dev); > @@ -1646,6 +1656,12 @@ struct net_device_ops { > ktime_t (*ndo_get_tstamp)(struct net_device *dev, > const struct skb_shared_hwtstamps *hwtstamps, > bool cycles); > + int (*ndo_hwtstamp_get)(struct net_device *dev, > + struct kernel_hwtstamp_config *kernel_config, > + struct netlink_ext_ack *extack); > + int (*ndo_hwtstamp_set)(struct net_device *dev, > + struct kernel_hwtstamp_config *kernel_config, > + struct netlink_ext_ack *extack); > }; > > struct xdp_metadata_ops { > diff --git a/net/core/dev_ioctl.c b/net/core/dev_ioctl.c > index 6d772837eb3f..736f310a0661 100644 > --- a/net/core/dev_ioctl.c > +++ b/net/core/dev_ioctl.c > @@ -254,11 +254,30 @@ static int dev_eth_ioctl(struct net_device *dev, > > static int dev_get_hwtstamp(struct net_device *dev, struct ifreq *ifr) > { > - return dev_eth_ioctl(dev, ifr, SIOCGHWTSTAMP); > + const struct net_device_ops *ops = dev->netdev_ops; > + struct kernel_hwtstamp_config kernel_cfg; Should we zero-initialize kernel_cfg (" = {}"), in case the driver does not bother to populate, say, "flags"? > + struct hwtstamp_config config; > + int err; > + > + if (!ops->ndo_hwtstamp_get) > + return dev_eth_ioctl(dev, ifr, SIOCGHWTSTAMP); > + > + if (!netif_device_present(dev)) > + return -ENODEV; > + > + err = ops->ndo_hwtstamp_get(dev, &kernel_cfg, NULL); > + if (err) > + return err; > + > + hwtstamp_kernel_to_config(&config, &kernel_cfg); > + if (copy_to_user(ifr->ifr_data, &config, sizeof(config))) > + return -EFAULT; > + return 0; > } > > static int dev_set_hwtstamp(struct net_device *dev, struct ifreq *ifr) > { > + const struct net_device_ops *ops = dev->netdev_ops; > struct netdev_notifier_hwtstamp_info info = { > .info.dev = dev, > }; > @@ -288,7 +307,20 @@ static int dev_set_hwtstamp(struct net_device *dev, struct ifreq *ifr) > return err; > } > > - return dev_eth_ioctl(dev, ifr, SIOCSHWTSTAMP); > + if (!ops->ndo_hwtstamp_set) > + return dev_eth_ioctl(dev, ifr, SIOCSHWTSTAMP); > + > + if (!netif_device_present(dev)) > + return -ENODEV; > + > + err = ops->ndo_hwtstamp_set(dev, &kernel_cfg, NULL); > + if (err) > + return err; > + > + hwtstamp_kernel_to_config(&cfg, &kernel_cfg); Cosmetic blank line here > + if (copy_to_user(ifr->ifr_data, &cfg, sizeof(cfg))) > + return -EFAULT; and here? (and in equivalent positions in dev_get_hwtstamp()) > + return 0; > } > > static int dev_siocbond(struct net_device *dev, > -- > 2.39.2 >
On Wed, Apr 5, 2023 at 6:31 AM Vladimir Oltean <vladimir.oltean@nxp.com> wrote: > > On Wed, Apr 05, 2023 at 12:31:44AM -0600, Maxim Georgiev wrote: > > Current NIC driver API demands drivers supporting hardware timestamping > > to implement handling logic for SIOCGHWTSTAMP/SIOCSHWTSTAMP IOCTLs. > > Handling these IOCTLs requires dirivers to implement request parameter > > structure translation between user and kernel address spaces, handling > > possible translation failures, etc. This translation code is pretty much > > identical across most of the NIC drivers that support SIOCGHWTSTAMP/ > > SIOCSHWTSTAMP. > > This patch extends NDO functiuon set with ndo_hwtstamp_get/set > > functions, implements SIOCGHWTSTAMP/SIOCSHWTSTAMP IOCTL translation > > to ndo_hwtstamp_get/set function calls including parameter structure > > translation and translation error handling. > > > > This patch is sent out as RFC. > > It still pending on basic testing. > > > > Suggested-by: Jakub Kicinski <kuba@kernel.org> > > Signed-off-by: Maxim Georgiev <glipus@gmail.com> > > --- > > Changes in v3: > > - Moved individual driver conversions to separate patches > > > > Changes in v2: > > - Introduced kernel_hwtstamp_config structure > > - Added netlink_ext_ack* and kernel_hwtstamp_config* as NDO hw timestamp > > function parameters > > - Reodered function variable declarations in dev_hwtstamp() > > - Refactored error handling logic in dev_hwtstamp() > > - Split dev_hwtstamp() into GET and SET versions > > - Changed net_hwtstamp_validate() to accept struct hwtstamp_config * > > as a parameter > > --- > > include/linux/net_tstamp.h | 8 ++++++++ > > include/linux/netdevice.h | 16 ++++++++++++++++ > > net/core/dev_ioctl.c | 36 ++++++++++++++++++++++++++++++++++-- > > 3 files changed, 58 insertions(+), 2 deletions(-) > > > > diff --git a/include/linux/net_tstamp.h b/include/linux/net_tstamp.h > > index fd67f3cc0c4b..063260475e77 100644 > > --- a/include/linux/net_tstamp.h > > +++ b/include/linux/net_tstamp.h > > @@ -30,4 +30,12 @@ static inline void hwtstamp_config_to_kernel(struct kernel_hwtstamp_config *kern > > kernel_cfg->rx_filter = cfg->rx_filter; > > } > > > > +static inline void hwtstamp_kernel_to_config(struct hwtstamp_config *cfg, > > + const struct kernel_hwtstamp_config *kernel_cfg) > > The reason why I suggested the name "hwtstamp_config_from_kernel()" was > to not break apart "hwtstamp" and "config", which together form the name > of one structure (hwtstamp_config). > Sorry I missed you referred the exact function name in your previous comments. Will be happy to rename. > > +{ > > + cfg->flags = kernel_cfg->flags; > > + cfg->tx_type = kernel_cfg->tx_type; > > + cfg->rx_filter = kernel_cfg->rx_filter; > > +} > > + > > #endif /* _LINUX_NET_TIMESTAMPING_H_ */ > > diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h > > index a740be3bb911..8356002d0ac0 100644 > > --- a/include/linux/netdevice.h > > +++ b/include/linux/netdevice.h > > @@ -57,6 +57,7 @@ > > struct netpoll_info; > > struct device; > > struct ethtool_ops; > > +struct kernel_hwtstamp_config; > > struct phy_device; > > struct dsa_port; > > struct ip_tunnel_parm; > > @@ -1412,6 +1413,15 @@ struct netdev_net_notifier { > > * Get hardware timestamp based on normal/adjustable time or free running > > * cycle counter. This function is required if physical clock supports a > > * free running cycle counter. > > + * int (*ndo_hwtstamp_get)(struct net_device *dev, > > + * struct kernel_hwtstamp_config *kernel_config, > > + * struct netlink_ext_ack *extack); > > + * Get hardware timestamping parameters currently configured for NIC > > + * device. > > + * int (*ndo_hwtstamp_set)(struct net_device *dev, > > + * struct kernel_hwtstamp_config *kernel_config, > > + * struct netlink_ext_ack *extack); > > + * Set hardware timestamping parameters for NIC device. > > */ > > struct net_device_ops { > > int (*ndo_init)(struct net_device *dev); > > @@ -1646,6 +1656,12 @@ struct net_device_ops { > > ktime_t (*ndo_get_tstamp)(struct net_device *dev, > > const struct skb_shared_hwtstamps *hwtstamps, > > bool cycles); > > + int (*ndo_hwtstamp_get)(struct net_device *dev, > > + struct kernel_hwtstamp_config *kernel_config, > > + struct netlink_ext_ack *extack); > > + int (*ndo_hwtstamp_set)(struct net_device *dev, > > + struct kernel_hwtstamp_config *kernel_config, > > + struct netlink_ext_ack *extack); > > }; > > > > struct xdp_metadata_ops { > > diff --git a/net/core/dev_ioctl.c b/net/core/dev_ioctl.c > > index 6d772837eb3f..736f310a0661 100644 > > --- a/net/core/dev_ioctl.c > > +++ b/net/core/dev_ioctl.c > > @@ -254,11 +254,30 @@ static int dev_eth_ioctl(struct net_device *dev, > > > > static int dev_get_hwtstamp(struct net_device *dev, struct ifreq *ifr) > > { > > - return dev_eth_ioctl(dev, ifr, SIOCGHWTSTAMP); > > + const struct net_device_ops *ops = dev->netdev_ops; > > + struct kernel_hwtstamp_config kernel_cfg; > > Should we zero-initialize kernel_cfg (" = {}"), in case the driver does > not bother to populate, say, "flags"? I've added kernel_cfg zero-initialization in [RFC PATCH v3 2/5]. But your comment makes sense. Let me move the initialization to this patch. > > > + struct hwtstamp_config config; > > + int err; > > + > > + if (!ops->ndo_hwtstamp_get) > > + return dev_eth_ioctl(dev, ifr, SIOCGHWTSTAMP); > > + > > + if (!netif_device_present(dev)) > > + return -ENODEV; > > + > > + err = ops->ndo_hwtstamp_get(dev, &kernel_cfg, NULL); > > + if (err) > > + return err; > > + > > + hwtstamp_kernel_to_config(&config, &kernel_cfg); > > + if (copy_to_user(ifr->ifr_data, &config, sizeof(config))) > > + return -EFAULT; > > + return 0; > > } > > > > static int dev_set_hwtstamp(struct net_device *dev, struct ifreq *ifr) > > { > > + const struct net_device_ops *ops = dev->netdev_ops; > > struct netdev_notifier_hwtstamp_info info = { > > .info.dev = dev, > > }; > > @@ -288,7 +307,20 @@ static int dev_set_hwtstamp(struct net_device *dev, struct ifreq *ifr) > > return err; > > } > > > > - return dev_eth_ioctl(dev, ifr, SIOCSHWTSTAMP); > > + if (!ops->ndo_hwtstamp_set) > > + return dev_eth_ioctl(dev, ifr, SIOCSHWTSTAMP); > > + > > + if (!netif_device_present(dev)) > > + return -ENODEV; > > + > > + err = ops->ndo_hwtstamp_set(dev, &kernel_cfg, NULL); > > + if (err) > > + return err; > > + > > + hwtstamp_kernel_to_config(&cfg, &kernel_cfg); > > Cosmetic blank line here > > > + if (copy_to_user(ifr->ifr_data, &cfg, sizeof(cfg))) > > + return -EFAULT; > > and here? > > (and in equivalent positions in dev_get_hwtstamp()) Thank you for pointing it out, I'll add the blank lines around these ifs. > > > + return 0; > > } > > > > static int dev_siocbond(struct net_device *dev, > > -- > > 2.39.2 > > Vladimir, thank you for the feedback!
On Wed, 5 Apr 2023 09:54:27 -0600 Max Georgiev <glipus@gmail.com> wrote: > On Wed, Apr 5, 2023 at 6:31 AM Vladimir Oltean <vladimir.oltean@nxp.com> > wrote: > > > > On Wed, Apr 05, 2023 at 12:31:44AM -0600, Maxim Georgiev wrote: > > > Current NIC driver API demands drivers supporting hardware timestamping > > > to implement handling logic for SIOCGHWTSTAMP/SIOCSHWTSTAMP IOCTLs. > > > Handling these IOCTLs requires dirivers to implement request parameter > > > structure translation between user and kernel address spaces, handling > > > possible translation failures, etc. This translation code is pretty much > > > identical across most of the NIC drivers that support SIOCGHWTSTAMP/ > > > SIOCSHWTSTAMP. > > > This patch extends NDO functiuon set with ndo_hwtstamp_get/set > > > functions, implements SIOCGHWTSTAMP/SIOCSHWTSTAMP IOCTL translation > > > to ndo_hwtstamp_get/set function calls including parameter structure > > > translation and translation error handling. > > > > > > This patch is sent out as RFC. > > > It still pending on basic testing. Just wondering about the status of this patch series. Do you want hardware testing before v4? As there were several reviews, I was waiting for the next version before doing any testing but if you ask for it to move forward I can deal with it. Also, I have cadence macb MAC which supports time stamping, I can adapt your last patch on e1000e to macb driver but I would appreciate if you do it beforehand.
On Thu, Apr 20, 2023 at 8:16 AM Köry Maincent <kory.maincent@bootlin.com> wrote: > > On Wed, 5 Apr 2023 09:54:27 -0600 > Max Georgiev <glipus@gmail.com> wrote: > > > On Wed, Apr 5, 2023 at 6:31 AM Vladimir Oltean <vladimir.oltean@nxp.com> > > wrote: > > > > > > On Wed, Apr 05, 2023 at 12:31:44AM -0600, Maxim Georgiev wrote: > > > > Current NIC driver API demands drivers supporting hardware timestamping > > > > to implement handling logic for SIOCGHWTSTAMP/SIOCSHWTSTAMP IOCTLs. > > > > Handling these IOCTLs requires dirivers to implement request parameter > > > > structure translation between user and kernel address spaces, handling > > > > possible translation failures, etc. This translation code is pretty much > > > > identical across most of the NIC drivers that support SIOCGHWTSTAMP/ > > > > SIOCSHWTSTAMP. > > > > This patch extends NDO functiuon set with ndo_hwtstamp_get/set > > > > functions, implements SIOCGHWTSTAMP/SIOCSHWTSTAMP IOCTL translation > > > > to ndo_hwtstamp_get/set function calls including parameter structure > > > > translation and translation error handling. > > > > > > > > This patch is sent out as RFC. > > > > It still pending on basic testing. > > > Just wondering about the status of this patch series. > Do you want hardware testing before v4? > As there were several reviews, I was waiting for the next version before doing > any testing but if you ask for it to move forward I can deal with it. > Also, I have cadence macb MAC which supports time stamping, I can adapt your > last patch on e1000e to macb driver but I would appreciate if you do it > beforehand. I've updated the patches in the stack based on the feedback you folks kindly shared for v3. I'm currently testing the combination of the changes in dev_ioctl.c and netdevsim driver changes - almost done with that. Let me finish this testing, then update the patch descriptions, and I'll be ready to send it out for review. Regarding e1000e conversion patch: I don't have access to any NICs with hw timestamping support. I was going to drop the e1000e patch from v4 unless someone volunteered to test it on real hardware. But that will have to wait till the rest of the stack is reviewed and at least preliminary accepted, right?
On Thu, 20 Apr 2023 12:04:06 -0600 Max Georgiev <glipus@gmail.com> wrote: > I've updated the patches in the stack based on the feedback you folks > kindly shared > for v3. I'm currently testing the combination of the changes in > dev_ioctl.c and netdevsim > driver changes - almost done with that. Let me finish this testing, > then update the patch > descriptions, and I'll be ready to send it out for review. > > Regarding e1000e conversion patch: I don't have access to any NICs > with hw timestamping > support. I was going to drop the e1000e patch from v4 unless someone > volunteered to test it on real hardware. But that will have to wait till the > rest of the stack is reviewed and > at least preliminary accepted, right? Great, thanks for the status. Indeed, if no one raise its hand to test the e1000e conversion you can drop it.
diff --git a/include/linux/net_tstamp.h b/include/linux/net_tstamp.h index fd67f3cc0c4b..063260475e77 100644 --- a/include/linux/net_tstamp.h +++ b/include/linux/net_tstamp.h @@ -30,4 +30,12 @@ static inline void hwtstamp_config_to_kernel(struct kernel_hwtstamp_config *kern kernel_cfg->rx_filter = cfg->rx_filter; } +static inline void hwtstamp_kernel_to_config(struct hwtstamp_config *cfg, + const struct kernel_hwtstamp_config *kernel_cfg) +{ + cfg->flags = kernel_cfg->flags; + cfg->tx_type = kernel_cfg->tx_type; + cfg->rx_filter = kernel_cfg->rx_filter; +} + #endif /* _LINUX_NET_TIMESTAMPING_H_ */ diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h index a740be3bb911..8356002d0ac0 100644 --- a/include/linux/netdevice.h +++ b/include/linux/netdevice.h @@ -57,6 +57,7 @@ struct netpoll_info; struct device; struct ethtool_ops; +struct kernel_hwtstamp_config; struct phy_device; struct dsa_port; struct ip_tunnel_parm; @@ -1412,6 +1413,15 @@ struct netdev_net_notifier { * Get hardware timestamp based on normal/adjustable time or free running * cycle counter. This function is required if physical clock supports a * free running cycle counter. + * int (*ndo_hwtstamp_get)(struct net_device *dev, + * struct kernel_hwtstamp_config *kernel_config, + * struct netlink_ext_ack *extack); + * Get hardware timestamping parameters currently configured for NIC + * device. + * int (*ndo_hwtstamp_set)(struct net_device *dev, + * struct kernel_hwtstamp_config *kernel_config, + * struct netlink_ext_ack *extack); + * Set hardware timestamping parameters for NIC device. */ struct net_device_ops { int (*ndo_init)(struct net_device *dev); @@ -1646,6 +1656,12 @@ struct net_device_ops { ktime_t (*ndo_get_tstamp)(struct net_device *dev, const struct skb_shared_hwtstamps *hwtstamps, bool cycles); + int (*ndo_hwtstamp_get)(struct net_device *dev, + struct kernel_hwtstamp_config *kernel_config, + struct netlink_ext_ack *extack); + int (*ndo_hwtstamp_set)(struct net_device *dev, + struct kernel_hwtstamp_config *kernel_config, + struct netlink_ext_ack *extack); }; struct xdp_metadata_ops { diff --git a/net/core/dev_ioctl.c b/net/core/dev_ioctl.c index 6d772837eb3f..736f310a0661 100644 --- a/net/core/dev_ioctl.c +++ b/net/core/dev_ioctl.c @@ -254,11 +254,30 @@ static int dev_eth_ioctl(struct net_device *dev, static int dev_get_hwtstamp(struct net_device *dev, struct ifreq *ifr) { - return dev_eth_ioctl(dev, ifr, SIOCGHWTSTAMP); + const struct net_device_ops *ops = dev->netdev_ops; + struct kernel_hwtstamp_config kernel_cfg; + struct hwtstamp_config config; + int err; + + if (!ops->ndo_hwtstamp_get) + return dev_eth_ioctl(dev, ifr, SIOCGHWTSTAMP); + + if (!netif_device_present(dev)) + return -ENODEV; + + err = ops->ndo_hwtstamp_get(dev, &kernel_cfg, NULL); + if (err) + return err; + + hwtstamp_kernel_to_config(&config, &kernel_cfg); + if (copy_to_user(ifr->ifr_data, &config, sizeof(config))) + return -EFAULT; + return 0; } static int dev_set_hwtstamp(struct net_device *dev, struct ifreq *ifr) { + const struct net_device_ops *ops = dev->netdev_ops; struct netdev_notifier_hwtstamp_info info = { .info.dev = dev, }; @@ -288,7 +307,20 @@ static int dev_set_hwtstamp(struct net_device *dev, struct ifreq *ifr) return err; } - return dev_eth_ioctl(dev, ifr, SIOCSHWTSTAMP); + if (!ops->ndo_hwtstamp_set) + return dev_eth_ioctl(dev, ifr, SIOCSHWTSTAMP); + + if (!netif_device_present(dev)) + return -ENODEV; + + err = ops->ndo_hwtstamp_set(dev, &kernel_cfg, NULL); + if (err) + return err; + + hwtstamp_kernel_to_config(&cfg, &kernel_cfg); + if (copy_to_user(ifr->ifr_data, &cfg, sizeof(cfg))) + return -EFAULT; + return 0; } static int dev_siocbond(struct net_device *dev,
Current NIC driver API demands drivers supporting hardware timestamping to implement handling logic for SIOCGHWTSTAMP/SIOCSHWTSTAMP IOCTLs. Handling these IOCTLs requires dirivers to implement request parameter structure translation between user and kernel address spaces, handling possible translation failures, etc. This translation code is pretty much identical across most of the NIC drivers that support SIOCGHWTSTAMP/ SIOCSHWTSTAMP. This patch extends NDO functiuon set with ndo_hwtstamp_get/set functions, implements SIOCGHWTSTAMP/SIOCSHWTSTAMP IOCTL translation to ndo_hwtstamp_get/set function calls including parameter structure translation and translation error handling. This patch is sent out as RFC. It still pending on basic testing. Suggested-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: Maxim Georgiev <glipus@gmail.com> --- Changes in v3: - Moved individual driver conversions to separate patches Changes in v2: - Introduced kernel_hwtstamp_config structure - Added netlink_ext_ack* and kernel_hwtstamp_config* as NDO hw timestamp function parameters - Reodered function variable declarations in dev_hwtstamp() - Refactored error handling logic in dev_hwtstamp() - Split dev_hwtstamp() into GET and SET versions - Changed net_hwtstamp_validate() to accept struct hwtstamp_config * as a parameter --- include/linux/net_tstamp.h | 8 ++++++++ include/linux/netdevice.h | 16 ++++++++++++++++ net/core/dev_ioctl.c | 36 ++++++++++++++++++++++++++++++++++-- 3 files changed, 58 insertions(+), 2 deletions(-)