| Message ID | 20200708154554.26450-1-wens@csie.org (mailing list archive) |
|---|---|
| Headers | show |
| Series | PM / OPP v2 & cpufreq backports part 2 | expand |
Hi, > -----Original Message----- > From: Chen-Yu Tsai (Moxa) [mailto:wens@csie.org] > Sent: Thursday, July 9, 2020 12:46 AM > To: iwamatsu nobuhiro(岩松 信洋 □ＳＷＣ◯ＡＣＴ) <nobuhiro1.iwamatsu@toshiba.co.jp>; pavel@denx.de > Cc: cip-dev@lists.cip-project.org; JohnsonCH.Chen@moxa.com > Subject: [PATCH 4.4.y-cip 00/23] PM / OPP v2 & cpufreq backports part 2 > > Hi everyone, > > This is part 2 of MOXA's PM / OPP / cpufreq backport series. The whole > series aims to backport patches related to PM / OPPv2 and cpufreq > which were included in the v4.4 kernel from TI's SDK. The end goal is > to include cpufreq-ti and convert am33xx to using it and OPPv2. > > Part 2 here includes patches from the v4.6 cycle, as well as additional > fixed found through Fixes tags: > > 9f8ea969d5cf PM / OPP: get/put regulators from OPP core > 7d34d56ef334 PM / OPP: Disable OPPs that aren't supported by the regulator > 655c9df96175 PM / OPP: Introduce dev_pm_opp_get_max_volt_latency() > 2174344765f4 PM / OPP: Introduce dev_pm_opp_get_max_transition_latency() > 50f8cfbd5897 PM / OPP: Parse clock-latency and voltage-tolerance for v1 bindings > d54974c2513f PM / OPP: Manage device clk > 6a0712f6f199 PM / OPP: Add dev_pm_opp_set_rate() > 896d6a4c0f41 cpufreq: dt: Convert few pr_debug/err() calls to dev_dbg/err() > 457e99e60a8f cpufreq: dt: Rename 'need_update' to 'opp_v1' > 391d9aef8145 cpufreq: dt: OPP layers handles clock-latency for V1 bindings as well > 050794aaebbb cpufreq: dt: Pass regulator name to the OPP core > 6def6ea75e6d cpufreq: dt: Unsupported OPPs are already disabled > 755b888ff098 cpufreq: dt: Reuse dev_pm_opp_get_max_transition_latency() > 78c3ba5df96c cpufreq: dt: Use dev_pm_opp_set_rate() to switch frequency > df2c8ec28e73 cpufreq: dt: No need to fetch voltage-tolerance > dd02a3d92008 cpufreq: dt: No need to allocate resources anymore > 78ecc56247f0 PM / OPP: Fix NULL pointer dereference crash when disabling OPPs > 0c717d0f9cb4 PM / OPP: Initialize regulator pointer to an error value > a5da64477ee7 PM / OPP: Fix incorrect comments > 2c2709dc6921 PM / OPP: Rename structures for clarity > b318556479cc cpufreq: dt: Drop stale comment > 21f8a99ce61b PM / OPP: Remove useless check > c5c2a97b3ac7 PM / OPP: Update voltage in case freq == old_freq > > Of these, > > b318556479cc cpufreq: dt: Drop stale comment > > was found while looking through git logs. > > 21f8a99ce61b PM / OPP: Remove useless check > c5c2a97b3ac7 PM / OPP: Update voltage in case freq == old_freq > > were found by looking for commit hashes in Fixes tags. > All other patches were included from TI's SDK. > > The patches apply cleanly on top of linux-4.4.y-cip. The last patch > involved some backporting due to path and code changes between v4.6 > and v4.18. > > Please have a look. > > I've looked at these patches and they seem to be fine. I am testing this patch as https://gitlab.com/cip-project/cip-kernel/linux-cip/-/pipelines/164602352 > Regards > ChenYu Best regards, Nobuhiro > > > Jon Hunter (1): > PM / OPP: Fix NULL pointer dereference crash when disabling OPPs > > Viresh Kumar (21): > PM / OPP: get/put regulators from OPP core > PM / OPP: Disable OPPs that aren't supported by the regulator > PM / OPP: Introduce dev_pm_opp_get_max_volt_latency() > PM / OPP: Introduce dev_pm_opp_get_max_transition_latency() > PM / OPP: Parse clock-latency and voltage-tolerance for v1 bindings > PM / OPP: Manage device clk > PM / OPP: Add dev_pm_opp_set_rate() > cpufreq: dt: Convert few pr_debug/err() calls to dev_dbg/err() > cpufreq: dt: Rename 'need_update' to 'opp_v1' > cpufreq: dt: OPP layers handles clock-latency for V1 bindings as well > cpufreq: dt: Pass regulator name to the OPP core > cpufreq: dt: Unsupported OPPs are already disabled > cpufreq: dt: Reuse dev_pm_opp_get_max_transition_latency() > cpufreq: dt: Use dev_pm_opp_set_rate() to switch frequency > cpufreq: dt: No need to fetch voltage-tolerance > cpufreq: dt: No need to allocate resources anymore > PM / OPP: Initialize regulator pointer to an error value > PM / OPP: Fix incorrect comments > PM / OPP: Rename structures for clarity > cpufreq: dt: Drop stale comment > PM / OPP: Remove useless check > > Waldemar Rymarkiewicz (1): > PM / OPP: Update voltage in case freq == old_freq > > drivers/base/power/opp/core.c | 1066 +++++++++++++++++++++--------- > drivers/base/power/opp/cpu.c | 22 +- > drivers/base/power/opp/debugfs.c | 85 ++- > drivers/base/power/opp/opp.h | 74 ++- > drivers/cpufreq/cpufreq-dt.c | 303 +++------ > include/linux/pm_opp.h | 27 + > 6 files changed, 973 insertions(+), 604 deletions(-) > > -- > 2.27.0 -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#4914): https://lists.cip-project.org/g/cip-dev/message/4914 Mute This Topic: https://lists.cip-project.org/mt/75379082/4520428 Group Owner: cip-dev+owner@lists.cip-project.org Unsubscribe: https://lists.cip-project.org/g/cip-dev/leave/8129116/1171672734/xyzzy [patchwork-cip-dev@patchwork.kernel.org] -=-=-=-=-=-=-=-=-=-=-=-
Hi! > This is part 2 of MOXA's PM / OPP / cpufreq backport series. The whole > series aims to backport patches related to PM / OPPv2 and cpufreq > which were included in the v4.4 kernel from TI's SDK. The end goal is > to include cpufreq-ti and convert am33xx to using it and OPPv2. Ok, I went through them, and most of them are ok. OTOH lots of them are changing common code, and I don't believe all of them are neccessary for am33xx support. Could you identify changes that are not neccesary for am33xx support, and drop those? Could the fixes be moved closer to patch they fix (ideally merged together)? Best regards, Pavel > Part 2 here includes patches from the v4.6 cycle, as well as additional > fixed found through Fixes tags: > > 9f8ea969d5cf PM / OPP: get/put regulators from OPP core > 7d34d56ef334 PM / OPP: Disable OPPs that aren't supported by the regulator > 655c9df96175 PM / OPP: Introduce dev_pm_opp_get_max_volt_latency() > 2174344765f4 PM / OPP: Introduce dev_pm_opp_get_max_transition_latency() > 50f8cfbd5897 PM / OPP: Parse clock-latency and voltage-tolerance for v1 bindings > d54974c2513f PM / OPP: Manage device clk > 6a0712f6f199 PM / OPP: Add dev_pm_opp_set_rate() > 896d6a4c0f41 cpufreq: dt: Convert few pr_debug/err() calls to dev_dbg/err() > 457e99e60a8f cpufreq: dt: Rename 'need_update' to 'opp_v1' > 391d9aef8145 cpufreq: dt: OPP layers handles clock-latency for V1 bindings as well > 050794aaebbb cpufreq: dt: Pass regulator name to the OPP core > 6def6ea75e6d cpufreq: dt: Unsupported OPPs are already disabled > 755b888ff098 cpufreq: dt: Reuse dev_pm_opp_get_max_transition_latency() > 78c3ba5df96c cpufreq: dt: Use dev_pm_opp_set_rate() to switch frequency > df2c8ec28e73 cpufreq: dt: No need to fetch voltage-tolerance > dd02a3d92008 cpufreq: dt: No need to allocate resources anymore > 78ecc56247f0 PM / OPP: Fix NULL pointer dereference crash when disabling OPPs > 0c717d0f9cb4 PM / OPP: Initialize regulator pointer to an error value > a5da64477ee7 PM / OPP: Fix incorrect comments > 2c2709dc6921 PM / OPP: Rename structures for clarity > b318556479cc cpufreq: dt: Drop stale comment > 21f8a99ce61b PM / OPP: Remove useless check > c5c2a97b3ac7 PM / OPP: Update voltage in case freq == old_freq > > Of these, > > b318556479cc cpufreq: dt: Drop stale comment > > was found while looking through git logs. > > 21f8a99ce61b PM / OPP: Remove useless check > c5c2a97b3ac7 PM / OPP: Update voltage in case freq == old_freq > > were found by looking for commit hashes in Fixes tags. > All other patches were included from TI's SDK. > > The patches apply cleanly on top of linux-4.4.y-cip. The last patch > involved some backporting due to path and code changes between v4.6 > and v4.18. > > Please have a look. > > > Regards > ChenYu > > > Jon Hunter (1): > PM / OPP: Fix NULL pointer dereference crash when disabling OPPs > > Viresh Kumar (21): > PM / OPP: get/put regulators from OPP core > PM / OPP: Disable OPPs that aren't supported by the regulator > PM / OPP: Introduce dev_pm_opp_get_max_volt_latency() > PM / OPP: Introduce dev_pm_opp_get_max_transition_latency() > PM / OPP: Parse clock-latency and voltage-tolerance for v1 bindings > PM / OPP: Manage device clk > PM / OPP: Add dev_pm_opp_set_rate() > cpufreq: dt: Convert few pr_debug/err() calls to dev_dbg/err() > cpufreq: dt: Rename 'need_update' to 'opp_v1' > cpufreq: dt: OPP layers handles clock-latency for V1 bindings as well > cpufreq: dt: Pass regulator name to the OPP core > cpufreq: dt: Unsupported OPPs are already disabled > cpufreq: dt: Reuse dev_pm_opp_get_max_transition_latency() > cpufreq: dt: Use dev_pm_opp_set_rate() to switch frequency > cpufreq: dt: No need to fetch voltage-tolerance > cpufreq: dt: No need to allocate resources anymore > PM / OPP: Initialize regulator pointer to an error value > PM / OPP: Fix incorrect comments > PM / OPP: Rename structures for clarity > cpufreq: dt: Drop stale comment > PM / OPP: Remove useless check > > Waldemar Rymarkiewicz (1): > PM / OPP: Update voltage in case freq == old_freq > > drivers/base/power/opp/core.c | 1066 +++++++++++++++++++++--------- > drivers/base/power/opp/cpu.c | 22 +- > drivers/base/power/opp/debugfs.c | 85 ++- > drivers/base/power/opp/opp.h | 74 ++- > drivers/cpufreq/cpufreq-dt.c | 303 +++------ > include/linux/pm_opp.h | 27 + > 6 files changed, 973 insertions(+), 604 deletions(-) >
On Wed, Jul 15, 2020 at 1:33 AM Pavel Machek <pavel@ucw.cz> wrote: > > Hi! > > > This is part 2 of MOXA's PM / OPP / cpufreq backport series. The whole > > series aims to backport patches related to PM / OPPv2 and cpufreq > > which were included in the v4.4 kernel from TI's SDK. The end goal is > > to include cpufreq-ti and convert am33xx to using it and OPPv2. > > Ok, I went through them, and most of them are ok. > > OTOH lots of them are changing common code, and I don't believe all of > them are neccessary for am33xx support. > > Could you identify changes that are not neccesary for am33xx support, > and drop those? So I tried just backporting the ti-cpufreq driver. However as the am33xx OPP table wasn't completely converted to the v2 format, missing the clock-latency-ns property, it is unusable in this standalone state. A couple of patches for PM/OPP are needed: 0764c604c812 PM / OPP: Expose _of_get_opp_desc_node as dev_pm_opp API 50f8cfbd5897 PM / OPP: Parse clock-latency and voltage-tolerance for v1 bindings I suppose this would be preferred over backporting all the OPP / PM cleanups? > Could the fixes be moved closer to patch they fix (ideally merged > together)? Sure. They are currently listed in roughly the same order as mainline. > Best regards, > Pavel > > > > Part 2 here includes patches from the v4.6 cycle, as well as additional > > fixed found through Fixes tags: > > > > 9f8ea969d5cf PM / OPP: get/put regulators from OPP core > > 7d34d56ef334 PM / OPP: Disable OPPs that aren't supported by the regulator > > 655c9df96175 PM / OPP: Introduce dev_pm_opp_get_max_volt_latency() > > 2174344765f4 PM / OPP: Introduce dev_pm_opp_get_max_transition_latency() > > 50f8cfbd5897 PM / OPP: Parse clock-latency and voltage-tolerance for v1 bindings > > d54974c2513f PM / OPP: Manage device clk > > 6a0712f6f199 PM / OPP: Add dev_pm_opp_set_rate() > > 896d6a4c0f41 cpufreq: dt: Convert few pr_debug/err() calls to dev_dbg/err() > > 457e99e60a8f cpufreq: dt: Rename 'need_update' to 'opp_v1' > > 391d9aef8145 cpufreq: dt: OPP layers handles clock-latency for V1 bindings as well > > 050794aaebbb cpufreq: dt: Pass regulator name to the OPP core > > 6def6ea75e6d cpufreq: dt: Unsupported OPPs are already disabled > > 755b888ff098 cpufreq: dt: Reuse dev_pm_opp_get_max_transition_latency() > > 78c3ba5df96c cpufreq: dt: Use dev_pm_opp_set_rate() to switch frequency > > df2c8ec28e73 cpufreq: dt: No need to fetch voltage-tolerance > > dd02a3d92008 cpufreq: dt: No need to allocate resources anymore > > 78ecc56247f0 PM / OPP: Fix NULL pointer dereference crash when disabling OPPs > > 0c717d0f9cb4 PM / OPP: Initialize regulator pointer to an error value > > a5da64477ee7 PM / OPP: Fix incorrect comments > > 2c2709dc6921 PM / OPP: Rename structures for clarity > > b318556479cc cpufreq: dt: Drop stale comment > > 21f8a99ce61b PM / OPP: Remove useless check > > c5c2a97b3ac7 PM / OPP: Update voltage in case freq == old_freq > > > > Of these, > > > > b318556479cc cpufreq: dt: Drop stale comment > > > > was found while looking through git logs. > > > > 21f8a99ce61b PM / OPP: Remove useless check > > c5c2a97b3ac7 PM / OPP: Update voltage in case freq == old_freq > > > > were found by looking for commit hashes in Fixes tags. > > All other patches were included from TI's SDK. > > > > The patches apply cleanly on top of linux-4.4.y-cip. The last patch > > involved some backporting due to path and code changes between v4.6 > > and v4.18. > > > > Please have a look. > > > > > > Regards > > ChenYu > > > > > > Jon Hunter (1): > > PM / OPP: Fix NULL pointer dereference crash when disabling OPPs > > > > Viresh Kumar (21): > > PM / OPP: get/put regulators from OPP core > > PM / OPP: Disable OPPs that aren't supported by the regulator > > PM / OPP: Introduce dev_pm_opp_get_max_volt_latency() > > PM / OPP: Introduce dev_pm_opp_get_max_transition_latency() > > PM / OPP: Parse clock-latency and voltage-tolerance for v1 bindings > > PM / OPP: Manage device clk > > PM / OPP: Add dev_pm_opp_set_rate() > > cpufreq: dt: Convert few pr_debug/err() calls to dev_dbg/err() > > cpufreq: dt: Rename 'need_update' to 'opp_v1' > > cpufreq: dt: OPP layers handles clock-latency for V1 bindings as well > > cpufreq: dt: Pass regulator name to the OPP core > > cpufreq: dt: Unsupported OPPs are already disabled > > cpufreq: dt: Reuse dev_pm_opp_get_max_transition_latency() > > cpufreq: dt: Use dev_pm_opp_set_rate() to switch frequency > > cpufreq: dt: No need to fetch voltage-tolerance > > cpufreq: dt: No need to allocate resources anymore > > PM / OPP: Initialize regulator pointer to an error value > > PM / OPP: Fix incorrect comments > > PM / OPP: Rename structures for clarity > > cpufreq: dt: Drop stale comment > > PM / OPP: Remove useless check > > > > Waldemar Rymarkiewicz (1): > > PM / OPP: Update voltage in case freq == old_freq > > > > drivers/base/power/opp/core.c | 1066 +++++++++++++++++++++--------- > > drivers/base/power/opp/cpu.c | 22 +- > > drivers/base/power/opp/debugfs.c | 85 ++- > > drivers/base/power/opp/opp.h | 74 ++- > > drivers/cpufreq/cpufreq-dt.c | 303 +++------ > > include/linux/pm_opp.h | 27 + > > 6 files changed, 973 insertions(+), 604 deletions(-) > > > > > -- > DENX Software Engineering GmbH, Managing Director: Wolfgang Denk > HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany > -=-=-=-=-=-=-=-=-=-=-=- Links: You receive all messages sent to this group. View/Reply Online (#5331): https://lists.cip-project.org/g/cip-dev/message/5331 Mute This Topic: https://lists.cip-project.org/mt/75379082/4520428 Group Owner: cip-dev+owner@lists.cip-project.org Unsubscribe: https://lists.cip-project.org/g/cip-dev/leave/8129116/1171672734/xyzzy [patchwork-cip-dev@patchwork.kernel.org] -=-=-=-=-=-=-=-=-=-=-=-
Hi! > > > This is part 2 of MOXA's PM / OPP / cpufreq backport series. The whole > > > series aims to backport patches related to PM / OPPv2 and cpufreq > > > which were included in the v4.4 kernel from TI's SDK. The end goal is > > > to include cpufreq-ti and convert am33xx to using it and OPPv2. > > > > Ok, I went through them, and most of them are ok. > > > > OTOH lots of them are changing common code, and I don't believe all of > > them are neccessary for am33xx support. > > > > Could you identify changes that are not neccesary for am33xx support, > > and drop those? > > So I tried just backporting the ti-cpufreq driver. However as the am33xx > OPP table wasn't completely converted to the v2 format, missing the > clock-latency-ns property, it is unusable in this standalone state. > > A couple of patches for PM/OPP are needed: > > 0764c604c812 PM / OPP: Expose _of_get_opp_desc_node as dev_pm_opp API > 50f8cfbd5897 PM / OPP: Parse clock-latency and voltage-tolerance > for v1 bindings > > I suppose this would be preferred over backporting all the OPP / PM > cleanups? Yes, that seems to be better solution than the previous series. Best regards, Pavel
Hi everyone, This is part 2 of MOXA's PM / OPP / cpufreq backport series. The whole series aims to backport patches related to PM / OPPv2 and cpufreq which were included in the v4.4 kernel from TI's SDK. The end goal is to include cpufreq-ti and convert am33xx to using it and OPPv2. Part 2 here includes patches from the v4.6 cycle, as well as additional fixed found through Fixes tags: 9f8ea969d5cf PM / OPP: get/put regulators from OPP core 7d34d56ef334 PM / OPP: Disable OPPs that aren't supported by the regulator 655c9df96175 PM / OPP: Introduce dev_pm_opp_get_max_volt_latency() 2174344765f4 PM / OPP: Introduce dev_pm_opp_get_max_transition_latency() 50f8cfbd5897 PM / OPP: Parse clock-latency and voltage-tolerance for v1 bindings d54974c2513f PM / OPP: Manage device clk 6a0712f6f199 PM / OPP: Add dev_pm_opp_set_rate() 896d6a4c0f41 cpufreq: dt: Convert few pr_debug/err() calls to dev_dbg/err() 457e99e60a8f cpufreq: dt: Rename 'need_update' to 'opp_v1' 391d9aef8145 cpufreq: dt: OPP layers handles clock-latency for V1 bindings as well 050794aaebbb cpufreq: dt: Pass regulator name to the OPP core 6def6ea75e6d cpufreq: dt: Unsupported OPPs are already disabled 755b888ff098 cpufreq: dt: Reuse dev_pm_opp_get_max_transition_latency() 78c3ba5df96c cpufreq: dt: Use dev_pm_opp_set_rate() to switch frequency df2c8ec28e73 cpufreq: dt: No need to fetch voltage-tolerance dd02a3d92008 cpufreq: dt: No need to allocate resources anymore 78ecc56247f0 PM / OPP: Fix NULL pointer dereference crash when disabling OPPs 0c717d0f9cb4 PM / OPP: Initialize regulator pointer to an error value a5da64477ee7 PM / OPP: Fix incorrect comments 2c2709dc6921 PM / OPP: Rename structures for clarity b318556479cc cpufreq: dt: Drop stale comment 21f8a99ce61b PM / OPP: Remove useless check c5c2a97b3ac7 PM / OPP: Update voltage in case freq == old_freq Of these, b318556479cc cpufreq: dt: Drop stale comment was found while looking through git logs. 21f8a99ce61b PM / OPP: Remove useless check c5c2a97b3ac7 PM / OPP: Update voltage in case freq == old_freq were found by looking for commit hashes in Fixes tags. All other patches were included from TI's SDK. The patches apply cleanly on top of linux-4.4.y-cip. The last patch involved some backporting due to path and code changes between v4.6 and v4.18. Please have a look. Regards ChenYu Jon Hunter (1): PM / OPP: Fix NULL pointer dereference crash when disabling OPPs Viresh Kumar (21): PM / OPP: get/put regulators from OPP core PM / OPP: Disable OPPs that aren't supported by the regulator PM / OPP: Introduce dev_pm_opp_get_max_volt_latency() PM / OPP: Introduce dev_pm_opp_get_max_transition_latency() PM / OPP: Parse clock-latency and voltage-tolerance for v1 bindings PM / OPP: Manage device clk PM / OPP: Add dev_pm_opp_set_rate() cpufreq: dt: Convert few pr_debug/err() calls to dev_dbg/err() cpufreq: dt: Rename 'need_update' to 'opp_v1' cpufreq: dt: OPP layers handles clock-latency for V1 bindings as well cpufreq: dt: Pass regulator name to the OPP core cpufreq: dt: Unsupported OPPs are already disabled cpufreq: dt: Reuse dev_pm_opp_get_max_transition_latency() cpufreq: dt: Use dev_pm_opp_set_rate() to switch frequency cpufreq: dt: No need to fetch voltage-tolerance cpufreq: dt: No need to allocate resources anymore PM / OPP: Initialize regulator pointer to an error value PM / OPP: Fix incorrect comments PM / OPP: Rename structures for clarity cpufreq: dt: Drop stale comment PM / OPP: Remove useless check Waldemar Rymarkiewicz (1): PM / OPP: Update voltage in case freq == old_freq drivers/base/power/opp/core.c | 1066 +++++++++++++++++++++--------- drivers/base/power/opp/cpu.c | 22 +- drivers/base/power/opp/debugfs.c | 85 ++- drivers/base/power/opp/opp.h | 74 ++- drivers/cpufreq/cpufreq-dt.c | 303 +++------ include/linux/pm_opp.h | 27 + 6 files changed, 973 insertions(+), 604 deletions(-)