| Message ID | 20230305205226.773025-1-marex@denx.de (mailing list archive) |
|---|---|
| State | Handled Elsewhere, archived |
| Headers | show |
| Series | power: supply: Ratelimit no data debug output | expand |
Hi, On 3/5/23 21:52, Marek Vasut wrote: > Reduce the amount of output this dev_dbg() statement emits into logs, > otherwise if system software polls the sysfs entry for data and keeps > getting -ENODATA, it could end up filling the logs up. > > This does in fact make systemd journald choke, since during boot the > sysfs power supply entries are polled and if journald starts at the > same time, the journal is just being repeatedly filled up, and the > system stops on trying to start journald without booting any further. > > Signed-off-by: Marek Vasut <marex@denx.de> > --- > Cc: Hans de Goede <hdegoede@redhat.com> > Cc: Sebastian Reichel <sre@kernel.org> > Cc: linux-pm@vger.kernel.org Thanks, patch looks good to me: Reviewed-by: Hans de Goede <hdegoede@redhat.com> Regards, Hans > --- > drivers/power/supply/power_supply_sysfs.c | 3 ++- > 1 file changed, 2 insertions(+), 1 deletion(-) > > diff --git a/drivers/power/supply/power_supply_sysfs.c b/drivers/power/supply/power_supply_sysfs.c > index c228205e09538..4bbb3053eef44 100644 > --- a/drivers/power/supply/power_supply_sysfs.c > +++ b/drivers/power/supply/power_supply_sysfs.c > @@ -285,7 +285,8 @@ static ssize_t power_supply_show_property(struct device *dev, > > if (ret < 0) { > if (ret == -ENODATA) > - dev_dbg(dev, "driver has no data for `%s' property\n", > + dev_dbg_ratelimited(dev, > + "driver has no data for `%s' property\n", > attr->attr.name); > else if (ret != -ENODEV && ret != -EAGAIN) > dev_err_ratelimited(dev,
On 3/5/23 21:57, Hans de Goede wrote: > Hi, Hi, > On 3/5/23 21:52, Marek Vasut wrote: >> Reduce the amount of output this dev_dbg() statement emits into logs, >> otherwise if system software polls the sysfs entry for data and keeps >> getting -ENODATA, it could end up filling the logs up. >> >> This does in fact make systemd journald choke, since during boot the >> sysfs power supply entries are polled and if journald starts at the >> same time, the journal is just being repeatedly filled up, and the >> system stops on trying to start journald without booting any further. >> >> Signed-off-by: Marek Vasut <marex@denx.de> >> --- >> Cc: Hans de Goede <hdegoede@redhat.com> >> Cc: Sebastian Reichel <sre@kernel.org> >> Cc: linux-pm@vger.kernel.org > > Thanks, patch looks good to me: > > Reviewed-by: Hans de Goede <hdegoede@redhat.com> Any news ? Maybe this could be applied now ?
Hi, On Thu, May 11, 2023 at 01:46:14AM +0200, Marek Vasut wrote: > On 3/5/23 21:57, Hans de Goede wrote: > > On 3/5/23 21:52, Marek Vasut wrote: > > > Reduce the amount of output this dev_dbg() statement emits into logs, > > > otherwise if system software polls the sysfs entry for data and keeps > > > getting -ENODATA, it could end up filling the logs up. > > > > > > This does in fact make systemd journald choke, since during boot the > > > sysfs power supply entries are polled and if journald starts at the > > > same time, the journal is just being repeatedly filled up, and the > > > system stops on trying to start journald without booting any further. > > > > > > Signed-off-by: Marek Vasut <marex@denx.de> > > > --- > > > Cc: Hans de Goede <hdegoede@redhat.com> > > > Cc: Sebastian Reichel <sre@kernel.org> > > > Cc: linux-pm@vger.kernel.org > > > > Thanks, patch looks good to me: > > > > Reviewed-by: Hans de Goede <hdegoede@redhat.com> > > Any news ? Maybe this could be applied now ? Sorry for the delay; I applied it to my fixes branch now. -- Sebastian
On 5/11/23 20:48, Sebastian Reichel wrote: > Hi, Hi, > On Thu, May 11, 2023 at 01:46:14AM +0200, Marek Vasut wrote: >> On 3/5/23 21:57, Hans de Goede wrote: >>> On 3/5/23 21:52, Marek Vasut wrote: >>>> Reduce the amount of output this dev_dbg() statement emits into logs, >>>> otherwise if system software polls the sysfs entry for data and keeps >>>> getting -ENODATA, it could end up filling the logs up. >>>> >>>> This does in fact make systemd journald choke, since during boot the >>>> sysfs power supply entries are polled and if journald starts at the >>>> same time, the journal is just being repeatedly filled up, and the >>>> system stops on trying to start journald without booting any further. >>>> >>>> Signed-off-by: Marek Vasut <marex@denx.de> >>>> --- >>>> Cc: Hans de Goede <hdegoede@redhat.com> >>>> Cc: Sebastian Reichel <sre@kernel.org> >>>> Cc: linux-pm@vger.kernel.org >>> >>> Thanks, patch looks good to me: >>> >>> Reviewed-by: Hans de Goede <hdegoede@redhat.com> >> >> Any news ? Maybe this could be applied now ? > > Sorry for the delay; I applied it to my fixes branch now. Thanks !
diff --git a/drivers/power/supply/power_supply_sysfs.c b/drivers/power/supply/power_supply_sysfs.c index c228205e09538..4bbb3053eef44 100644 --- a/drivers/power/supply/power_supply_sysfs.c +++ b/drivers/power/supply/power_supply_sysfs.c @@ -285,7 +285,8 @@ static ssize_t power_supply_show_property(struct device *dev, if (ret < 0) { if (ret == -ENODATA) - dev_dbg(dev, "driver has no data for `%s' property\n", + dev_dbg_ratelimited(dev, + "driver has no data for `%s' property\n", attr->attr.name); else if (ret != -ENODEV && ret != -EAGAIN) dev_err_ratelimited(dev,
Reduce the amount of output this dev_dbg() statement emits into logs, otherwise if system software polls the sysfs entry for data and keeps getting -ENODATA, it could end up filling the logs up. This does in fact make systemd journald choke, since during boot the sysfs power supply entries are polled and if journald starts at the same time, the journal is just being repeatedly filled up, and the system stops on trying to start journald without booting any further. Signed-off-by: Marek Vasut <marex@denx.de> --- Cc: Hans de Goede <hdegoede@redhat.com> Cc: Sebastian Reichel <sre@kernel.org> Cc: linux-pm@vger.kernel.org --- drivers/power/supply/power_supply_sysfs.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-)