On Fri, Jun 12, 2020 at 11:22 AM Jan Kiszka <jan.kis...@siemens.com> wrote: > > On 12.06.20 13:41, Jouko Haapaluoma via Xenomai wrote: > > This patch set adds basic I-pipe support for the SAMA5D3 SoC > > > > A few mainline patches need to be applied to the tcb_clksrc driver so > > that it can be used as the only clocksource. A patch from RT_PREEMPT > > was also merged to support higher than 32k clkevent frequency. > > > > Additional patch was created to increase the clkevent max_delta to 32-bit > > when TCB has 32-bit channels. This makes it possible to always use high > > resolution clkevent when 32-bit channels are available. > > > > After that the ipipe_timer and clksrc support can be added into tcb_clksrc. > > > > Pinctrl-at91 also requires usage of the ipipe_handle_demuxed_irq() > > > > Tested with latest ipipe-arm branch ipipe/master (Linux 4.19.82) > > > > Alexandre Belloni (4): > > ARM: at91: move SoC specific definitions to SoC folder > > clocksource/drivers/tcb_clksrc: Stop depending on atmel_tclib > > clocksource/drivers/tcb_clksrc: Use tcb as sched_clock > > clocksource/drivers/timer-atmel-pit: Rework Kconfig option > > > > Benedikt Spranger (1): > > clocksource: TCLIB: Allow higher clock rates for clock events > > > > Jouko Haapaluoma (3): > > tcb_clksrc: Always use high resolution with 32-bit channels > > tcb_clksrc: Add ipipe_timer and ipipe_tsc support > > pinctrl-at91: Add ipipe pipeline support > > > > Tuomo Jauhiainen (1): > > mach-at91: Add IPIPE_ARM_KUSER_TSC for SAMA5D3 > > > > arch/arm/mach-at91/Kconfig | 1 + > > drivers/clocksource/Kconfig | 5 +- > > drivers/clocksource/tcb_clksrc.c | 237 > > +++++++++++++++------ > > drivers/misc/Kconfig | 20 +- > > drivers/misc/atmel_tclib.c | 2 +- > > drivers/pinctrl/pinctrl-at91.c | 5 +- > > drivers/pwm/pwm-atmel-tcb.c | 2 +- > > include/{linux/atmel_tc.h => soc/at91/atmel_tcb.h} | 4 +- > > 8 files changed, 189 insertions(+), 87 deletions(-) > > rename include/{linux/atmel_tc.h => soc/at91/atmel_tcb.h} (99%) > > > > Great, thanks for the patches! I assume Greg will look into them in details. > > Regarding the mainline patches: Are they simply missing in 4.19 because > they were not backported as fixes (feature patches)? Are they safe to be > applied on a stable branch, i.e. side-effect free for other targets? > > Jan > > -- > Siemens AG, Corporate Technology, CT RDA IOT SES-DE > Corporate Competence Center Embedded Linux
Found them, I'll apply them and run a few tests.