Re: [PATCH v5 8/8] ARM: tegra: Add firmware calls required for suspend-resume
From: Dmitry Osipenko <digetx@gmail.com>
Date: 2019-02-22 18:26:28
Also in:
linux-tegra, lkml
22.02.2019 20:59, Dmitry Osipenko пишет:
quoted hunk ↗ jump to hunk
In order to resume CPU from suspend via trusted Foundations firmware, the LP1/LP2 boot vectors and CPU caches need to be set up using the firmware calls. Signed-off-by: Dmitry Osipenko <digetx@gmail.com> --- arch/arm/mach-tegra/pm.c | 53 ++++++++++++++++++----------- arch/arm/mach-tegra/reset-handler.S | 26 ++++++++++++++ arch/arm/mach-tegra/sleep.S | 3 +- 3 files changed, 61 insertions(+), 21 deletions(-)diff --git a/arch/arm/mach-tegra/pm.c b/arch/arm/mach-tegra/pm.c index 66c8cd63dd86..f209f59e0daf 100644 --- a/arch/arm/mach-tegra/pm.c +++ b/arch/arm/mach-tegra/pm.c@@ -33,6 +33,7 @@ #include <soc/tegra/pmc.h> #include <asm/cacheflush.h> +#include <asm/firmware.h> #include <asm/idmap.h> #include <asm/proc-fns.h> #include <asm/smp_plat.h>@@ -160,6 +161,28 @@ int tegra_cpu_do_idle(void) static int tegra_sleep_cpu(unsigned long v2p) { + /* + * L2 cache disabling using kernel API only allowed when all + * secondary CPU's are offline. Cache have to be disabled with + * MMU-on if cache maintenance is done via Trusted Foundations + * firmware. Note that CPUIDLE won't ever enter powergate on Tegra30 + * if any of secondary CPU's is online and this is the LP2-idle + * code-path only for Tegra20/30. + */ + if (trusted_foundations_registered()) + outer_disable(); + + /* + * Note that besides of setting up CPU reset vector this firmware + * call may also do the following, depending on the FW version: + * 1) Disable L2. But this doesn't matter since we already + * disabled the L2. + * 2) Disable D-cache. This need to be taken into account in + * particular by the tegra_disable_clean_inv_dcache() which + * shall avoid the re-disable. + */ + call_firmware_op(prepare_idle, TF_PM_MODE_LP2); + setup_mm_for_reboot(); tegra_sleep_cpu_finish(v2p);@@ -196,24 +219,13 @@ void tegra_idle_lp2_last(void) cpu_cluster_pm_enter(); suspend_cpu_complex(); - /* - * L2 cache disabling using kernel API only allowed when all - * secondary CPU's are offline. Cache have to be disabled early - * if cache maintenance is done via Trusted Foundations firmware. - * Note that CPUIDLE won't ever enter powergate on Tegra30 if any - * of secondary CPU's is online and this is the LP2 codepath only - * for Tegra20/30. - */ - if (trusted_foundations_registered()) - outer_disable(); - cpu_suspend(PHYS_OFFSET - PAGE_OFFSET, &tegra_sleep_cpu); /* * Resume L2 cache if it wasn't re-enabled early during resume, * which is the case for Tegra30 that has to re-enable the cache * via firmware call. In other cases cache is already enabled and - * hence re-enabling is a no-op. + * hence re-enabling is a no-op. This is always a no-op on Tegra114+. */ outer_resume();@@ -235,6 +247,15 @@ enum tegra_suspend_mode tegra_pm_validate_suspend_mode( static int tegra_sleep_core(unsigned long v2p) { + /* + * Cache have to be disabled with MMU-on if cache maintenance is done + * via Trusted Foundations firmware. This is a no-op on Tegra114+. + */ + if (trusted_foundations_registered()) + outer_disable(); + + call_firmware_op(prepare_idle, TF_PM_MODE_LP1); + setup_mm_for_reboot(); tegra_sleep_core_finish(v2p);@@ -360,14 +381,6 @@ static int tegra_suspend_enter(suspend_state_t state) break; } - /* - * Cache have to be disabled early if cache maintenance is done - * via Trusted Foundations firmware. Otherwise this is a no-op, - * like on Tegra114+. - */ - if (trusted_foundations_registered()) - outer_disable(); - cpu_suspend(PHYS_OFFSET - PAGE_OFFSET, tegra_sleep_func);
Seems I messed up the rebasing a tad. Will send another version. _______________________________________________ linux-arm-kernel mailing list linux-arm-kernel@lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-arm-kernel