From patchwork Fri Mar 4 12:23:05 2016 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Jon Hunter X-Patchwork-Id: 8503541 Return-Path: X-Original-To: patchwork-linux-pm@patchwork.kernel.org Delivered-To: patchwork-parsemail@patchwork2.web.kernel.org Received: from mail.kernel.org (mail.kernel.org [198.145.29.136]) by patchwork2.web.kernel.org (Postfix) with ESMTP id 8EE01C0553 for ; Fri, 4 Mar 2016 12:23:30 +0000 (UTC) Received: from mail.kernel.org (localhost [127.0.0.1]) by mail.kernel.org (Postfix) with ESMTP id 0637C201EF for ; Fri, 4 Mar 2016 12:23:29 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 550A9201DD for ; Fri, 4 Mar 2016 12:23:27 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751431AbcCDMXZ (ORCPT ); Fri, 4 Mar 2016 07:23:25 -0500 Received: from hqemgate14.nvidia.com ([216.228.121.143]:15070 "EHLO hqemgate14.nvidia.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751066AbcCDMXY (ORCPT ); Fri, 4 Mar 2016 07:23:24 -0500 Received: from hqnvupgp08.nvidia.com (Not Verified[216.228.121.13]) by hqemgate14.nvidia.com id ; Fri, 04 Mar 2016 04:23:06 -0800 Received: from hqemhub02.nvidia.com ([172.20.150.31]) by hqnvupgp08.nvidia.com (PGP Universal service); Fri, 04 Mar 2016 04:22:59 -0800 X-PGP-Universal: processed; by hqnvupgp08.nvidia.com on Fri, 04 Mar 2016 04:22:59 -0800 Received: from jonathanh-lm.nvidia.com (172.20.144.16) by hqemhub02.nvidia.com (172.20.150.31) with Microsoft SMTP Server (TLS) id 8.3.406.0; Fri, 4 Mar 2016 04:23:22 -0800 From: Jon Hunter To: Stephen Warren , Thierry Reding , Alexandre Courbot CC: Rob Herring , Pawel Moll , Mark Rutland , Ian Campbell , Kumar Gala , "Rafael J. Wysocki" , Kevin Hilman , Ulf Hansson , linux-tegra@vger.kernel.org, devicetree@vger.kernel.org, linux-pm@vger.kernel.org, Jon Hunter Subject: [PATCH V7 3/4] soc: tegra: pmc: Add generic PM domain support Date: Fri, 4 Mar 2016 12:23:05 +0000 Message-ID: <1457094186-15786-4-git-send-email-jonathanh@nvidia.com> X-Mailer: git-send-email 2.1.4 In-Reply-To: <1457094186-15786-1-git-send-email-jonathanh@nvidia.com> References: <1457094186-15786-1-git-send-email-jonathanh@nvidia.com> X-NVConfidentiality: public MIME-Version: 1.0 Sender: linux-pm-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-pm@vger.kernel.org X-Spam-Status: No, score=-6.9 required=5.0 tests=BAYES_00, RCVD_IN_DNSWL_HI, RP_MATCHES_RCVD, UNPARSEABLE_RELAY autolearn=unavailable version=3.3.1 X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on mail.kernel.org X-Virus-Scanned: ClamAV using ClamSMTP Adds generic PM support to the PMC driver where the PM domains are populated from device-tree and the PM domain consumer devices are bound to their relevant PM domains via device-tree as well. Update the tegra_powergate_sequence_power_up() API so that internally it calls the same tegra_powergate_xxx functions that are used by the tegra generic power domain code for consistency. To ensure that the Tegra power domains (a.k.a powergates) cannot be controlled via both the legacy tegra_powergate_xxx functions as well as the generic PM domain framework, add a bit map for available powergates that can be controlled via the legacy powergate functions. Move the majority of the tegra_powergate_remove_clamping() function to a sub-function, so that this can be used by both the legacy and generic power domain code. Currently, the power domains are not removed once added because this is not yet supported by the PM domains framework. An error message will be displayed if we are unable to instantiate a power domain. This is based upon work by Thierry Reding and Vince Hsu . Signed-off-by: Jon Hunter --- drivers/soc/tegra/pmc.c | 485 ++++++++++++++++++++++++++++++++++++++++++------ include/soc/tegra/pmc.h | 1 + 2 files changed, 425 insertions(+), 61 deletions(-) diff --git a/drivers/soc/tegra/pmc.c b/drivers/soc/tegra/pmc.c index 08966c26d65c..bb173456bbff 100644 --- a/drivers/soc/tegra/pmc.c +++ b/drivers/soc/tegra/pmc.c @@ -31,10 +31,13 @@ #include #include #include +#include #include +#include #include #include #include +#include #include #include @@ -102,6 +105,16 @@ #define GPU_RG_CNTRL 0x2d4 +struct tegra_powergate { + struct generic_pm_domain genpd; + struct tegra_pmc *pmc; + unsigned int id; + struct clk **clks; + unsigned int num_clks; + struct reset_control **resets; + unsigned int num_resets; +}; + struct tegra_pmc_soc { unsigned int num_powergates; const char *const *powergates; @@ -132,6 +145,7 @@ struct tegra_pmc_soc { * @cpu_pwr_good_en: CPU power good signal is enabled * @lp0_vec_phys: physical base address of the LP0 warm boot code * @lp0_vec_size: size of the LP0 warm boot code + * @powergates_available: Bitmap of available power gates * @powergates_lock: mutex for power gate register access */ struct tegra_pmc { @@ -156,6 +170,7 @@ struct tegra_pmc { bool cpu_pwr_good_en; u32 lp0_vec_phys; u32 lp0_vec_size; + DECLARE_BITMAP(powergates_available, TEGRA_POWERGATE_MAX); struct mutex powergates_lock; }; @@ -165,6 +180,12 @@ static struct tegra_pmc *pmc = &(struct tegra_pmc) { .suspend_mode = TEGRA_SUSPEND_NONE, }; +static inline struct tegra_powergate * +to_powergate(struct generic_pm_domain *domain) +{ + return container_of(domain, struct tegra_powergate, genpd); +} + static u32 tegra_pmc_readl(unsigned long offset) { return readl(pmc->base + offset); @@ -188,6 +209,31 @@ static inline bool tegra_powergate_is_valid(int id) return (pmc->soc && pmc->soc->powergates[id]); } +static inline bool tegra_powergate_is_available(int id) +{ + return test_bit(id, pmc->powergates_available); +} + +static int tegra_powergate_lookup(struct tegra_pmc *pmc, const char *name) +{ + unsigned int i; + + if (!pmc || !pmc->soc || !name) + return -EINVAL; + + for (i = 0; i < pmc->soc->num_powergates; i++) { + if (!tegra_powergate_is_valid(i)) + continue; + + if (!strcmp(name, pmc->soc->powergates[i])) + return i; + } + + dev_err(pmc->dev, "powergate %s not found\n", name); + + return -ENODEV; +} + /** * tegra_powergate_set() - set the state of a partition * @id: partition ID @@ -218,13 +264,219 @@ static int tegra_powergate_set(unsigned int id, bool new_state) return err; } +static int __tegra_powergate_remove_clamping(unsigned int id) +{ + u32 mask; + + mutex_lock(&pmc->powergates_lock); + + /* + * On Tegra124 and later, the clamps for the GPU are controlled by a + * separate register (with different semantics). + */ + if (id == TEGRA_POWERGATE_3D) { + if (pmc->soc->has_gpu_clamps) { + tegra_pmc_writel(0, GPU_RG_CNTRL); + goto out; + } + } + + /* + * Tegra 2 has a bug where PCIE and VDE clamping masks are + * swapped relatively to the partition ids + */ + if (id == TEGRA_POWERGATE_VDEC) + mask = (1 << TEGRA_POWERGATE_PCIE); + else if (id == TEGRA_POWERGATE_PCIE) + mask = (1 << TEGRA_POWERGATE_VDEC); + else + mask = (1 << id); + + tegra_pmc_writel(mask, REMOVE_CLAMPING); + +out: + mutex_unlock(&pmc->powergates_lock); + + return 0; +} + +static void tegra_powergate_disable_clocks(struct tegra_powergate *pg) +{ + unsigned int i; + + for (i = 0; i < pg->num_clks; i++) + clk_disable_unprepare(pg->clks[i]); +} + +static int tegra_powergate_enable_clocks(struct tegra_powergate *pg) +{ + unsigned int i; + int err; + + for (i = 0; i < pg->num_clks; i++) { + err = clk_prepare_enable(pg->clks[i]); + if (err) + goto out; + } + + return 0; + +out: + while (i--) + clk_disable_unprepare(pg->clks[i]); + + return err; +} + +static int tegra_powergate_reset_assert(struct tegra_powergate *pg) +{ + unsigned int i; + int err; + + for (i = 0; i < pg->num_resets; i++) { + err = reset_control_assert(pg->resets[i]); + if (err) + return err; + } + + return 0; +} + +static int tegra_powergate_reset_deassert(struct tegra_powergate *pg) +{ + unsigned int i; + int err; + + for (i = 0; i < pg->num_resets; i++) { + err = reset_control_deassert(pg->resets[i]); + if (err) + return err; + } + + return 0; +} + +static int tegra_powergate_power_up(struct tegra_powergate *pg, + bool disable_clocks) +{ + int err; + + err = tegra_powergate_reset_assert(pg); + if (err) + return err; + + usleep_range(10, 20); + + err = tegra_powergate_set(pg->id, true); + if (err < 0) + return err; + + usleep_range(10, 20); + + err = tegra_powergate_enable_clocks(pg); + if (err) + goto disable_clks; + + usleep_range(10, 20); + + err = __tegra_powergate_remove_clamping(pg->id); + if (err) + goto disable_clks; + + usleep_range(10, 20); + + err = tegra_powergate_reset_deassert(pg); + if (err) + goto powergate_off; + + usleep_range(10, 20); + + if (disable_clocks) + tegra_powergate_disable_clocks(pg); + + return 0; + +disable_clks: + tegra_powergate_disable_clocks(pg); + usleep_range(10, 20); +powergate_off: + tegra_powergate_set(pg->id, false); + + return err; +} + +static int tegra_powergate_power_down(struct tegra_powergate *pg) +{ + int err; + + err = tegra_powergate_enable_clocks(pg); + if (err) + return err; + + usleep_range(10, 20); + + err = tegra_powergate_reset_assert(pg); + if (err) + goto disable_clks; + + usleep_range(10, 20); + + tegra_powergate_disable_clocks(pg); + + usleep_range(10, 20); + + err = tegra_powergate_set(pg->id, false); + if (err) + goto assert_resets; + + return 0; + +assert_resets: + tegra_powergate_enable_clocks(pg); + usleep_range(10, 20); + tegra_powergate_reset_deassert(pg); + usleep_range(10, 20); +disable_clks: + tegra_powergate_disable_clocks(pg); + + return err; +} + +static int tegra_genpd_power_on(struct generic_pm_domain *domain) +{ + struct tegra_powergate *pg = to_powergate(domain); + struct tegra_pmc *pmc = pg->pmc; + int err; + + err = tegra_powergate_power_up(pg, true); + if (err) + dev_err(pmc->dev, "failed to turn on PM domain %s: %d\n", + pg->genpd.name, err); + + return err; +} + +static int tegra_genpd_power_off(struct generic_pm_domain *domain) +{ + struct tegra_powergate *pg = to_powergate(domain); + struct tegra_pmc *pmc = pg->pmc; + int err; + + err = tegra_powergate_power_down(pg); + if (err) + dev_err(pmc->dev, "failed to turn off PM domain %s: %d\n", + pg->genpd.name, err); + + return err; +} + /** * tegra_powergate_power_on() - power on partition * @id: partition ID */ int tegra_powergate_power_on(unsigned int id) { - if (!tegra_powergate_is_valid(id)) + if (!tegra_powergate_is_available(id)) return -EINVAL; return tegra_powergate_set(id, true); @@ -236,7 +488,7 @@ int tegra_powergate_power_on(unsigned int id) */ int tegra_powergate_power_off(unsigned int id) { - if (!tegra_powergate_is_valid(id)) + if (!tegra_powergate_is_available(id)) return -EINVAL; return tegra_powergate_set(id, false); @@ -267,41 +519,10 @@ int tegra_powergate_is_powered(unsigned int id) */ int tegra_powergate_remove_clamping(unsigned int id) { - u32 mask; - - if (!tegra_powergate_is_valid(id)) + if (!tegra_powergate_is_available(id)) return -EINVAL; - mutex_lock(&pmc->powergates_lock); - - /* - * On Tegra124 and later, the clamps for the GPU are controlled by a - * separate register (with different semantics). - */ - if (id == TEGRA_POWERGATE_3D) { - if (pmc->soc->has_gpu_clamps) { - tegra_pmc_writel(0, GPU_RG_CNTRL); - goto out; - } - } - - /* - * Tegra 2 has a bug where PCIE and VDE clamping masks are - * swapped relatively to the partition ids - */ - if (id == TEGRA_POWERGATE_VDEC) - mask = (1 << TEGRA_POWERGATE_PCIE); - else if (id == TEGRA_POWERGATE_PCIE) - mask = (1 << TEGRA_POWERGATE_VDEC); - else - mask = (1 << id); - - tegra_pmc_writel(mask, REMOVE_CLAMPING); - -out: - mutex_unlock(&pmc->powergates_lock); - - return 0; + return __tegra_powergate_remove_clamping(id); } EXPORT_SYMBOL(tegra_powergate_remove_clamping); @@ -316,35 +537,20 @@ EXPORT_SYMBOL(tegra_powergate_remove_clamping); int tegra_powergate_sequence_power_up(unsigned int id, struct clk *clk, struct reset_control *rst) { - int ret; - - reset_control_assert(rst); - - ret = tegra_powergate_power_on(id); - if (ret) - goto err_power; - - ret = clk_prepare_enable(clk); - if (ret) - goto err_clk; - - usleep_range(10, 20); + struct tegra_powergate pg; + int err; - ret = tegra_powergate_remove_clamping(id); - if (ret) - goto err_clamp; + pg.id = id; + pg.clks = &clk; + pg.num_clks = 1; + pg.resets = &rst; + pg.num_resets = 1; - usleep_range(10, 20); - reset_control_deassert(rst); - - return 0; + err = tegra_powergate_power_up(&pg, false); + if (err) + pr_err("failed to turn on partition %d: %d\n", id, err); -err_clamp: - clk_disable_unprepare(clk); -err_clk: - tegra_powergate_power_off(id); -err_power: - return ret; + return err; } EXPORT_SYMBOL(tegra_powergate_sequence_power_up); @@ -486,6 +692,155 @@ static int tegra_powergate_debugfs_init(void) return 0; } +static int tegra_powergate_of_get_clks(struct tegra_powergate *pg, + struct device_node *np) +{ + struct clk *clk; + unsigned int i, count; + int err; + + count = of_count_phandle_with_args(np, "clocks", "#clock-cells"); + if (count == 0) + return -ENODEV; + + pg->clks = kcalloc(count, sizeof(clk), GFP_KERNEL); + if (!pg->clks) + return -ENOMEM; + + for (i = 0; i < count; i++) { + pg->clks[i] = of_clk_get(np, i); + if (IS_ERR(pg->clks[i])) { + err = PTR_ERR(pg->clks[i]); + goto err; + } + } + + pg->num_clks = count; + + return 0; + +err: + while (i--) + clk_put(pg->clks[i]); + kfree(pg->clks); + + return err; +} + +static int tegra_powergate_of_get_resets(struct tegra_powergate *pg, + struct device_node *np) +{ + struct reset_control *rst; + unsigned int i, count; + int err; + + count = of_count_phandle_with_args(np, "resets", "#reset-cells"); + if (count == 0) + return -ENODEV; + + pg->resets = kcalloc(count, sizeof(rst), GFP_KERNEL); + if (!pg->resets) + return -ENOMEM; + + for (i = 0; i < count; i++) { + pg->resets[i] = of_reset_control_get_by_index(np, i); + if (IS_ERR(pg->resets[i])) { + err = PTR_ERR(pg->resets[i]); + goto error; + } + } + + pg->num_resets = count; + + return 0; + +error: + while (i--) + reset_control_put(pg->resets[i]); + kfree(pg->resets); + + return err; +} + +static void tegra_powergate_add(struct tegra_pmc *pmc, struct device_node *np) +{ + struct tegra_powergate *pg; + bool off; + int id; + + pg = kzalloc(sizeof(*pg), GFP_KERNEL); + if (!pg) + goto error; + + id = tegra_powergate_lookup(pmc, np->name); + if (id < 0) + goto free_mem; + + /* + * Clear the bit for this powergate so it cannot be managed + * directly via the legacy APIs for controlling powergates. + */ + clear_bit(id, pmc->powergates_available); + + pg->id = id; + pg->genpd.name = np->name; + pg->genpd.power_off = tegra_genpd_power_off; + pg->genpd.power_on = tegra_genpd_power_on; + pg->pmc = pmc; + + if (tegra_powergate_of_get_clks(pg, np)) + goto set_available; + + if (tegra_powergate_of_get_resets(pg, np)) + goto remove_clks; + + off = !tegra_powergate_is_powered(pg->id); + + pm_genpd_init(&pg->genpd, NULL, off); + + if (of_genpd_add_provider_simple(np, &pg->genpd)) + goto remove_resets; + + dev_dbg(pmc->dev, "added power domain %s\n", pg->genpd.name); + + return; + +remove_resets: + while (pg->num_resets--) + reset_control_put(pg->resets[pg->num_resets]); + kfree(pg->resets); + +remove_clks: + while (pg->num_clks--) + clk_put(pg->clks[pg->num_clks]); + kfree(pg->clks); + +set_available: + set_bit(id, pmc->powergates_available); + +free_mem: + kfree(pg); + +error: + dev_err(pmc->dev, "failed to create power domain for %s\n", np->name); +} + +static void tegra_powergate_init(struct tegra_pmc *pmc) +{ + struct device_node *np, *child; + + np = of_get_child_by_name(pmc->dev->of_node, "powergates"); + if (!np) + return; + + for_each_child_of_node(np, child) { + tegra_powergate_add(pmc, child); + of_node_put(child); + } + + of_node_put(np); +} + static int tegra_io_rail_prepare(unsigned int id, unsigned long *request, unsigned long *status, unsigned int *bit) { @@ -887,6 +1242,8 @@ static int tegra_pmc_probe(struct platform_device *pdev) return err; } + tegra_powergate_init(pmc); + mutex_lock(&pmc->powergates_lock); iounmap(pmc->base); pmc->base = base; @@ -1120,6 +1477,7 @@ static int __init tegra_pmc_early_init(void) const struct of_device_id *match; struct device_node *np; struct resource regs; + unsigned int i; bool invert; u32 value; @@ -1169,6 +1527,11 @@ static int __init tegra_pmc_early_init(void) return -ENXIO; } + /* Create a bit-map of the available and valid partitions */ + for (i = 0; i < pmc->soc->num_powergates; i++) + if (pmc->soc->powergates[i]) + set_bit(i, pmc->powergates_available); + mutex_init(&pmc->powergates_lock); /* diff --git a/include/soc/tegra/pmc.h b/include/soc/tegra/pmc.h index 07e332dd44fb..e9e53473a63e 100644 --- a/include/soc/tegra/pmc.h +++ b/include/soc/tegra/pmc.h @@ -72,6 +72,7 @@ int tegra_pmc_cpu_remove_clamping(unsigned int cpuid); #define TEGRA_POWERGATE_AUD 27 #define TEGRA_POWERGATE_DFD 28 #define TEGRA_POWERGATE_VE2 29 +#define TEGRA_POWERGATE_MAX TEGRA_POWERGATE_VE2 #define TEGRA_POWERGATE_3D0 TEGRA_POWERGATE_3D