| Message ID | 20240821121539.374343-9-wojciech.drewek@intel.com (mailing list archive) |
|---|---|
| State | Awaiting Upstream |
| Delegated to: | Netdev Maintainers |
| Headers | show |
| Series | Add support for Rx timestamping for both ice and iavf drivers | expand |
| Context | Check | Description |
|---|---|---|
| netdev/tree_selection | success | Guessing tree name failed - patch did not apply |
From: Wojciech Drewek <wojciech.drewek@intel.com> Date: Wed, 21 Aug 2024 14:15:33 +0200 > From: Jacob Keller <jacob.e.keller@intel.com> > > The Rx timestamps reported by hardware may only have 32 bits of storage > for nanosecond time. These timestamps cannot be directly reported to the > Linux stack, as it expects 64bits of time. > > To handle this, the timestamps must be extended using an algorithm that > calculates the corrected 64bit timestamp by comparison between the PHC > time and the timestamp. This algorithm requires the PHC time to be > captured within ~2 seconds of when the timestamp was captured. > > Instead of trying to read the PHC time in the Rx hotpath, the algorithm > relies on a cached value that is periodically updated. > > Keep this cached time up to date by using the PTP .do_aux_work kthread > function. > > The iavf_ptp_do_aux_work will reschedule itself about twice a second, > and will check whether or not the cached PTP time needs to be updated. > If so, it issues a VIRTCHNL_OP_1588_PTP_GET_TIME to request the time > from the PF. The jitter and latency involved with this command aren't > important, because the cached time just needs to be kept up to date > within about ~2 seconds. > > Signed-off-by: Jacob Keller <jacob.e.keller@intel.com> > Reviewed-by: Wojciech Drewek <wojciech.drewek@intel.com> > Reviewed-by: Simon Horman <horms@kernel.org> > Co-developed-by: Mateusz Polchlopek <mateusz.polchlopek@intel.com> > Signed-off-by: Mateusz Polchlopek <mateusz.polchlopek@intel.com> > Signed-off-by: Wojciech Drewek <wojciech.drewek@intel.com> > --- > drivers/net/ethernet/intel/iavf/iavf_ptp.c | 52 ++++++++++++++++++++++ > drivers/net/ethernet/intel/iavf/iavf_ptp.h | 1 + > 2 files changed, 53 insertions(+) > > diff --git a/drivers/net/ethernet/intel/iavf/iavf_ptp.c b/drivers/net/ethernet/intel/iavf/iavf_ptp.c > index d709d381958f..7124a717cd03 100644 > --- a/drivers/net/ethernet/intel/iavf/iavf_ptp.c > +++ b/drivers/net/ethernet/intel/iavf/iavf_ptp.c > @@ -153,6 +153,55 @@ static int iavf_ptp_gettimex64(struct ptp_clock_info *info, > return iavf_read_phc_indirect(adapter, ts, sts); > } > > +/** > + * iavf_ptp_cache_phc_time - Cache PHC time for performing timestamp extension > + * @adapter: private adapter structure > + * > + * Periodically cache the PHC time in order to allow for timestamp extension. > + * This is required because the Tx and Rx timestamps only contain 32bits of > + * nanoseconds. Timestamp extension allows calculating the corrected 64bit > + * timestamp. This algorithm relies on the cached time being within ~1 second > + * of the timestamp. > + */ > +static void iavf_ptp_cache_phc_time(struct iavf_adapter *adapter) > +{ > + if (time_is_before_jiffies(adapter->ptp.cached_phc_updated + HZ)) { Also invert the condition to avoid +1 indent level? > + /* The response from virtchnl will store the time into > + * cached_phc_time. > + */ > + iavf_send_phc_read(adapter); > + } > +} > + > +/** > + * iavf_ptp_do_aux_work - Perform periodic work required for PTP support > + * @info: PTP clock info structure > + * > + * Handler to take care of periodic work required for PTP operation. This > + * includes the following tasks: > + * > + * 1) updating cached_phc_time > + * > + * cached_phc_time is used by the Tx and Rx timestamp flows in order to > + * perform timestamp extension, by carefully comparing the timestamp > + * 32bit nanosecond timestamps and determining the corrected 64bit > + * timestamp value to report to userspace. This algorithm only works if > + * the cached_phc_time is within ~1 second of the Tx or Rx timestamp > + * event. This task periodically reads the PHC time and stores it, to > + * ensure that timestamp extension operates correctly. > + * > + * Returns: time in jiffies until the periodic task should be re-scheduled. > + */ > +long iavf_ptp_do_aux_work(struct ptp_clock_info *info) Why isn't it static since it's used only within this one file? > +{ > + struct iavf_adapter *adapter = iavf_clock_to_adapter(info); > + > + iavf_ptp_cache_phc_time(adapter); > + > + /* Check work about twice a second */ > + return msecs_to_jiffies(500); Thanks, Olek
On 21.08.2024 16:43, Alexander Lobakin wrote: > From: Wojciech Drewek <wojciech.drewek@intel.com> > Date: Wed, 21 Aug 2024 14:15:33 +0200 > >> From: Jacob Keller <jacob.e.keller@intel.com> >> >> The Rx timestamps reported by hardware may only have 32 bits of storage >> for nanosecond time. These timestamps cannot be directly reported to the >> Linux stack, as it expects 64bits of time. >> >> To handle this, the timestamps must be extended using an algorithm that >> calculates the corrected 64bit timestamp by comparison between the PHC >> time and the timestamp. This algorithm requires the PHC time to be >> captured within ~2 seconds of when the timestamp was captured. >> >> Instead of trying to read the PHC time in the Rx hotpath, the algorithm >> relies on a cached value that is periodically updated. >> >> Keep this cached time up to date by using the PTP .do_aux_work kthread >> function. >> >> The iavf_ptp_do_aux_work will reschedule itself about twice a second, >> and will check whether or not the cached PTP time needs to be updated. >> If so, it issues a VIRTCHNL_OP_1588_PTP_GET_TIME to request the time >> from the PF. The jitter and latency involved with this command aren't >> important, because the cached time just needs to be kept up to date >> within about ~2 seconds. >> >> Signed-off-by: Jacob Keller <jacob.e.keller@intel.com> >> Reviewed-by: Wojciech Drewek <wojciech.drewek@intel.com> >> Reviewed-by: Simon Horman <horms@kernel.org> >> Co-developed-by: Mateusz Polchlopek <mateusz.polchlopek@intel.com> >> Signed-off-by: Mateusz Polchlopek <mateusz.polchlopek@intel.com> >> Signed-off-by: Wojciech Drewek <wojciech.drewek@intel.com> >> --- >> drivers/net/ethernet/intel/iavf/iavf_ptp.c | 52 ++++++++++++++++++++++ >> drivers/net/ethernet/intel/iavf/iavf_ptp.h | 1 + >> 2 files changed, 53 insertions(+) >> >> diff --git a/drivers/net/ethernet/intel/iavf/iavf_ptp.c b/drivers/net/ethernet/intel/iavf/iavf_ptp.c >> index d709d381958f..7124a717cd03 100644 >> --- a/drivers/net/ethernet/intel/iavf/iavf_ptp.c >> +++ b/drivers/net/ethernet/intel/iavf/iavf_ptp.c >> @@ -153,6 +153,55 @@ static int iavf_ptp_gettimex64(struct ptp_clock_info *info, >> return iavf_read_phc_indirect(adapter, ts, sts); >> } >> >> +/** >> + * iavf_ptp_cache_phc_time - Cache PHC time for performing timestamp extension >> + * @adapter: private adapter structure >> + * >> + * Periodically cache the PHC time in order to allow for timestamp extension. >> + * This is required because the Tx and Rx timestamps only contain 32bits of >> + * nanoseconds. Timestamp extension allows calculating the corrected 64bit >> + * timestamp. This algorithm relies on the cached time being within ~1 second >> + * of the timestamp. >> + */ >> +static void iavf_ptp_cache_phc_time(struct iavf_adapter *adapter) >> +{ >> + if (time_is_before_jiffies(adapter->ptp.cached_phc_updated + HZ)) { > > Also invert the condition to avoid +1 indent level? Agree > >> + /* The response from virtchnl will store the time into >> + * cached_phc_time. >> + */ >> + iavf_send_phc_read(adapter); >> + } >> +} >> + >> +/** >> + * iavf_ptp_do_aux_work - Perform periodic work required for PTP support >> + * @info: PTP clock info structure >> + * >> + * Handler to take care of periodic work required for PTP operation. This >> + * includes the following tasks: >> + * >> + * 1) updating cached_phc_time >> + * >> + * cached_phc_time is used by the Tx and Rx timestamp flows in order to >> + * perform timestamp extension, by carefully comparing the timestamp >> + * 32bit nanosecond timestamps and determining the corrected 64bit >> + * timestamp value to report to userspace. This algorithm only works if >> + * the cached_phc_time is within ~1 second of the Tx or Rx timestamp >> + * event. This task periodically reads the PHC time and stores it, to >> + * ensure that timestamp extension operates correctly. >> + * >> + * Returns: time in jiffies until the periodic task should be re-scheduled. >> + */ >> +long iavf_ptp_do_aux_work(struct ptp_clock_info *info) > > Why isn't it static since it's used only within this one file? Agree > >> +{ >> + struct iavf_adapter *adapter = iavf_clock_to_adapter(info); >> + >> + iavf_ptp_cache_phc_time(adapter); >> + >> + /* Check work about twice a second */ >> + return msecs_to_jiffies(500); > > Thanks, > Olek
diff --git a/drivers/net/ethernet/intel/iavf/iavf_ptp.c b/drivers/net/ethernet/intel/iavf/iavf_ptp.c index d709d381958f..7124a717cd03 100644 --- a/drivers/net/ethernet/intel/iavf/iavf_ptp.c +++ b/drivers/net/ethernet/intel/iavf/iavf_ptp.c @@ -153,6 +153,55 @@ static int iavf_ptp_gettimex64(struct ptp_clock_info *info, return iavf_read_phc_indirect(adapter, ts, sts); } +/** + * iavf_ptp_cache_phc_time - Cache PHC time for performing timestamp extension + * @adapter: private adapter structure + * + * Periodically cache the PHC time in order to allow for timestamp extension. + * This is required because the Tx and Rx timestamps only contain 32bits of + * nanoseconds. Timestamp extension allows calculating the corrected 64bit + * timestamp. This algorithm relies on the cached time being within ~1 second + * of the timestamp. + */ +static void iavf_ptp_cache_phc_time(struct iavf_adapter *adapter) +{ + if (time_is_before_jiffies(adapter->ptp.cached_phc_updated + HZ)) { + /* The response from virtchnl will store the time into + * cached_phc_time. + */ + iavf_send_phc_read(adapter); + } +} + +/** + * iavf_ptp_do_aux_work - Perform periodic work required for PTP support + * @info: PTP clock info structure + * + * Handler to take care of periodic work required for PTP operation. This + * includes the following tasks: + * + * 1) updating cached_phc_time + * + * cached_phc_time is used by the Tx and Rx timestamp flows in order to + * perform timestamp extension, by carefully comparing the timestamp + * 32bit nanosecond timestamps and determining the corrected 64bit + * timestamp value to report to userspace. This algorithm only works if + * the cached_phc_time is within ~1 second of the Tx or Rx timestamp + * event. This task periodically reads the PHC time and stores it, to + * ensure that timestamp extension operates correctly. + * + * Returns: time in jiffies until the periodic task should be re-scheduled. + */ +long iavf_ptp_do_aux_work(struct ptp_clock_info *info) +{ + struct iavf_adapter *adapter = iavf_clock_to_adapter(info); + + iavf_ptp_cache_phc_time(adapter); + + /* Check work about twice a second */ + return msecs_to_jiffies(500); +} + /** * iavf_ptp_register_clock - Register a new PTP for userspace * @adapter: private adapter structure @@ -172,6 +221,7 @@ static int iavf_ptp_register_clock(struct iavf_adapter *adapter) dev_driver_string(dev), dev_name(dev)); ptp_info->owner = THIS_MODULE; ptp_info->gettimex64 = iavf_ptp_gettimex64; + ptp_info->do_aux_work = iavf_ptp_do_aux_work; adapter->ptp.clock = ptp_clock_register(ptp_info, dev); if (IS_ERR(adapter->ptp.clock)) { @@ -211,6 +261,8 @@ void iavf_ptp_init(struct iavf_adapter *adapter) return; } + ptp_schedule_worker(adapter->ptp.clock, 0); + adapter->ptp.initialized = true; } diff --git a/drivers/net/ethernet/intel/iavf/iavf_ptp.h b/drivers/net/ethernet/intel/iavf/iavf_ptp.h index 0bb4bddc1495..88f2a7bc4506 100644 --- a/drivers/net/ethernet/intel/iavf/iavf_ptp.h +++ b/drivers/net/ethernet/intel/iavf/iavf_ptp.h @@ -14,5 +14,6 @@ void iavf_ptp_release(struct iavf_adapter *adapter); void iavf_ptp_process_caps(struct iavf_adapter *adapter); bool iavf_ptp_cap_supported(const struct iavf_adapter *adapter, u32 cap); void iavf_virtchnl_send_ptp_cmd(struct iavf_adapter *adapter); +long iavf_ptp_do_aux_work(struct ptp_clock_info *ptp); #endif /* _IAVF_PTP_H_ */