| Message ID | 4958885.31r3eYUQgx@rjwysocki.net |
|---|---|
| Headers | show |
| Series | thermal: core: Use lists of trips for trip crossing detection and handling | expand |
On Wed, Oct 16, 2024 at 1:37 PM Rafael J. Wysocki <rjw@rjwysocki.net> wrote: > > Hi Everyone, > > This is a continuation of > > https://lore.kernel.org/linux-pm/4985597.31r3eYUQgx@rjwysocki.net/ > > derived from patches [3-7/8] in the following patch series: > > https://lore.kernel.org/linux-pm/4920970.GXAFRqVoOG@rjwysocki.net/ > > but mostly rewritten. > > It is based on the observation that putting trip points on sorted lists > allows to reduce overhead related to the handling of them in some cases. > Namely, it avoids the need to walk all trips in a thermal zone every > time the zone temperature is updated (including invalid ones) and > generally leads to cleaner code. > > Patches [01-08/10] are preliminary, patch [09/10] makes the key changes, > and patch [10/10] is a super-cosmetic cleanup on top of the rest. > > Please refer to the individual patch changelogs for details. This material is on the thermal-core-experimental branch in linux-pm.git along with https://lore.kernel.org/linux-pm/2215082.irdbgypaU6@rjwysocki.net/ and https://lore.kernel.org/linux-pm/4985597.31r3eYUQgx@rjwysocki.net/ which are also present in the thermal-core-testing branch.
Hi Rafael, On 10/21/24 12:16, Rafael J. Wysocki wrote: > On Wed, Oct 16, 2024 at 1:37 PM Rafael J. Wysocki <rjw@rjwysocki.net> wrote: >> >> Hi Everyone, >> >> This is a continuation of >> >> https://lore.kernel.org/linux-pm/4985597.31r3eYUQgx@rjwysocki.net/ >> >> derived from patches [3-7/8] in the following patch series: >> >> https://lore.kernel.org/linux-pm/4920970.GXAFRqVoOG@rjwysocki.net/ >> >> but mostly rewritten. >> >> It is based on the observation that putting trip points on sorted lists >> allows to reduce overhead related to the handling of them in some cases. >> Namely, it avoids the need to walk all trips in a thermal zone every >> time the zone temperature is updated (including invalid ones) and >> generally leads to cleaner code. >> >> Patches [01-08/10] are preliminary, patch [09/10] makes the key changes, >> and patch [10/10] is a super-cosmetic cleanup on top of the rest. >> >> Please refer to the individual patch changelogs for details. > > This material is on the thermal-core-experimental branch in > linux-pm.git along with > > https://lore.kernel.org/linux-pm/2215082.irdbgypaU6@rjwysocki.net/ > > and > > https://lore.kernel.org/linux-pm/4985597.31r3eYUQgx@rjwysocki.net/ > > which are also present in the thermal-core-testing branch. If it's not too late, I can review that stuff tomorrow for you. Regards, Lukasz
Hi Lukasz, On Wed, Oct 23, 2024 at 12:51 AM Lukasz Luba <lukasz.luba@arm.com> wrote: > > Hi Rafael, > > On 10/21/24 12:16, Rafael J. Wysocki wrote: > > On Wed, Oct 16, 2024 at 1:37 PM Rafael J. Wysocki <rjw@rjwysocki.net> wrote: > >> > >> Hi Everyone, > >> > >> This is a continuation of > >> > >> https://lore.kernel.org/linux-pm/4985597.31r3eYUQgx@rjwysocki.net/ > >> > >> derived from patches [3-7/8] in the following patch series: > >> > >> https://lore.kernel.org/linux-pm/4920970.GXAFRqVoOG@rjwysocki.net/ > >> > >> but mostly rewritten. > >> > >> It is based on the observation that putting trip points on sorted lists > >> allows to reduce overhead related to the handling of them in some cases. > >> Namely, it avoids the need to walk all trips in a thermal zone every > >> time the zone temperature is updated (including invalid ones) and > >> generally leads to cleaner code. > >> > >> Patches [01-08/10] are preliminary, patch [09/10] makes the key changes, > >> and patch [10/10] is a super-cosmetic cleanup on top of the rest. > >> > >> Please refer to the individual patch changelogs for details. > > > > This material is on the thermal-core-experimental branch in > > linux-pm.git along with > > > > https://lore.kernel.org/linux-pm/2215082.irdbgypaU6@rjwysocki.net/ > > > > and > > > > https://lore.kernel.org/linux-pm/4985597.31r3eYUQgx@rjwysocki.net/ > > > > which are also present in the thermal-core-testing branch. > > > If it's not too late, I can review that stuff tomorrow for you. No, it is not too late, please review it! And thanks a lot for all of the reviews so far!