From patchwork Tue Jun 2 01:47:59 2015 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Stephen Boyd X-Patchwork-Id: 6526301 Return-Path: X-Original-To: patchwork-linux-arm@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 5B53AC0020 for ; Tue, 2 Jun 2015 01:52:27 +0000 (UTC) Received: from mail.kernel.org (localhost [127.0.0.1]) by mail.kernel.org (Postfix) with ESMTP id E0E2D20384 for ; Tue, 2 Jun 2015 01:52:23 +0000 (UTC) Received: from bombadil.infradead.org (bombadil.infradead.org [198.137.202.9]) (using TLSv1.2 with cipher DHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPS id 6B3D920380 for ; Tue, 2 Jun 2015 01:52:20 +0000 (UTC) Received: from localhost ([127.0.0.1] helo=bombadil.infradead.org) by bombadil.infradead.org with esmtp (Exim 4.80.1 #2 (Red Hat Linux)) id 1YzbKm-0003Sq-DC; Tue, 02 Jun 2015 01:49:40 +0000 Received: from smtp.codeaurora.org ([198.145.29.96]) by bombadil.infradead.org with esmtps (Exim 4.80.1 #2 (Red Hat Linux)) id 1YzbJr-0002we-Rs for linux-arm-kernel@lists.infradead.org; Tue, 02 Jun 2015 01:48:49 +0000 Received: from smtp.codeaurora.org (localhost [127.0.0.1]) by smtp.codeaurora.org (Postfix) with ESMTP id D5D0A140870; Tue, 2 Jun 2015 01:48:05 +0000 (UTC) Received: by smtp.codeaurora.org (Postfix, from userid 486) id C6A82140872; Tue, 2 Jun 2015 01:48:05 +0000 (UTC) X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on mail.kernel.org X-Spam-Level: X-Spam-Status: No, score=-4.2 required=5.0 tests=BAYES_00, RCVD_IN_DNSWL_MED, T_RP_MATCHES_RCVD, UNPARSEABLE_RELAY autolearn=unavailable version=3.3.1 Received: from sboyd-linux.qualcomm.com (i-global254.qualcomm.com [199.106.103.254]) (using TLSv1.1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (No client certificate requested) (Authenticated sender: sboyd@smtp.codeaurora.org) by smtp.codeaurora.org (Postfix) with ESMTPSA id A528C140865; Tue, 2 Jun 2015 01:48:02 +0000 (UTC) From: Stephen Boyd To: linux-arm-msm@vger.kernel.org Subject: [PATCH 6/6] power: avs: Add support for CPR (Core Power Reduction) Date: Mon, 1 Jun 2015 18:47:59 -0700 Message-Id: <1433209679-31389-7-git-send-email-sboyd@codeaurora.org> X-Mailer: git-send-email 2.3.0.rc1.33.g42e4583 In-Reply-To: <1433209679-31389-1-git-send-email-sboyd@codeaurora.org> References: <1433209679-31389-1-git-send-email-sboyd@codeaurora.org> X-Virus-Scanned: ClamAV using ClamSMTP X-CRM114-Version: 20100106-BlameMichelson ( TRE 0.8.0 (BSD) ) MR-646709E3 X-CRM114-CacheID: sfid-20150601_184844_112631_56A822A3 X-CRM114-Status: GOOD ( 21.44 ) X-Spam-Score: -0.0 (/) Cc: Nishanth Menon , David Collins , linux-pm@vger.kernel.org, Viresh Kumar , linux-kernel@vger.kernel.org, Mark Brown , Srinivas Kandagatla , Andy Gross , linux-arm-kernel@lists.infradead.org X-BeenThere: linux-arm-kernel@lists.infradead.org X-Mailman-Version: 2.1.18-1 Precedence: list List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , MIME-Version: 1.0 Sender: "linux-arm-kernel" Errors-To: linux-arm-kernel-bounces+patchwork-linux-arm=patchwork.kernel.org@lists.infradead.org X-Virus-Scanned: ClamAV using ClamSMTP RBCPR (RapidBridge Core Power Reduction) is a technology to reduce core power on a CPU or other device. It reads voltage settings in efuse from product test process as initial settings and populates OPPs for the device being "monitored". Each OPP corresponds to a "corner" that has a range of valid voltages for a particular frequency. While the device is running at a particular frequency, CPR monitors dynamic factors such as temperature, etc. and adjusts the voltage for that frequency accordingly to save power and meet silicon characteristic requirements. This driver is based on work by others on codeaurora.org[1]. TODO: * DT binding documentation [1] https://www.codeaurora.org/cgit/quic/la/kernel/msm-3.10/tree/drivers/regulator/cpr-regulator.c?h=msm-3.10 Cc: David Collins Signed-off-by: Stephen Boyd --- There's still some TODOs left, but it's mostly surrounding the step voltage reading part, hotplug, and DT documentation. drivers/power/avs/Kconfig | 14 + drivers/power/avs/Makefile | 1 + drivers/power/avs/qcom-cpr.c | 1983 ++++++++++++++++++++++++++++++++++++++++++ 3 files changed, 1998 insertions(+) create mode 100644 drivers/power/avs/qcom-cpr.c diff --git a/drivers/power/avs/Kconfig b/drivers/power/avs/Kconfig index 7f3d389bd601..82e604b363a6 100644 --- a/drivers/power/avs/Kconfig +++ b/drivers/power/avs/Kconfig @@ -11,6 +11,20 @@ menuconfig POWER_AVS Say Y here to enable Adaptive Voltage Scaling class support. +config QCOM_CPR + tristate "QCOM Core Power Reduction (CPR) support" + select PM_OPP + help + Say Y here to enable support for the CPR hardware found on Qualcomm + SoCs like MSM8916. + + This driver populates CPU OPPs tables and makes adjustments to the + tables based on feedback from the CPR hardware. If you want to do + CPUfrequency scaling say Y here. + + To compile this driver as a module, choose M here: the module will + be called qcom-cpr + config ROCKCHIP_IODOMAIN tristate "Rockchip IO domain support" depends on ARCH_ROCKCHIP && OF diff --git a/drivers/power/avs/Makefile b/drivers/power/avs/Makefile index ba4c7bc69225..88f4d5d49cba 100644 --- a/drivers/power/avs/Makefile +++ b/drivers/power/avs/Makefile @@ -1,2 +1,3 @@ obj-$(CONFIG_POWER_AVS_OMAP) += smartreflex.o obj-$(CONFIG_ROCKCHIP_IODOMAIN) += rockchip-io-domain.o +obj-$(CONFIG_QCOM_CPR) += qcom-cpr.o diff --git a/drivers/power/avs/qcom-cpr.c b/drivers/power/avs/qcom-cpr.c new file mode 100644 index 000000000000..b5d4c4042b8c --- /dev/null +++ b/drivers/power/avs/qcom-cpr.c @@ -0,0 +1,1983 @@ +/* + * Copyright (c) 2013-2015, The Linux Foundation. All rights reserved. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + */ + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +/* Register Offsets for RB-CPR and Bit Definitions */ + +/* RBCPR Version Register */ +#define REG_RBCPR_VERSION 0 +#define RBCPR_VER_2 0x02 + +/* RBCPR Gate Count and Target Registers */ +#define REG_RBCPR_GCNT_TARGET(n) (0x60 + 4 * n) + +#define RBCPR_GCNT_TARGET_TARGET_SHIFT 0 +#define RBCPR_GCNT_TARGET_TARGET_MASK GENMASK(11, 0) +#define RBCPR_GCNT_TARGET_GCNT_SHIFT 12 +#define RBCPR_GCNT_TARGET_GCNT_MASK GENMASK(9, 0) + +/* RBCPR Timer Control */ +#define REG_RBCPR_TIMER_INTERVAL 0x44 +#define REG_RBIF_TIMER_ADJUST 0x4c + +#define RBIF_TIMER_ADJ_CONS_UP_MASK GENMASK(3, 0) +#define RBIF_TIMER_ADJ_CONS_UP_SHIFT 0 +#define RBIF_TIMER_ADJ_CONS_DOWN_MASK GENMASK(3, 0) +#define RBIF_TIMER_ADJ_CONS_DOWN_SHIFT 4 +#define RBIF_TIMER_ADJ_CLAMP_INT_MASK GENMASK(7, 0) +#define RBIF_TIMER_ADJ_CLAMP_INT_SHIFT 8 + +/* RBCPR Config Register */ +#define REG_RBIF_LIMIT 0x48 +#define RBIF_LIMIT_CEILING_MASK GENMASK(5, 0) +#define RBIF_LIMIT_CEILING_SHIFT 6 +#define RBIF_LIMIT_FLOOR_BITS 6 +#define RBIF_LIMIT_FLOOR_MASK GENMASK(5, 0) + +#define RBIF_LIMIT_CEILING_DEFAULT RBIF_LIMIT_CEILING_MASK +#define RBIF_LIMIT_FLOOR_DEFAULT 0 + +#define REG_RBIF_SW_VLEVEL 0x94 +#define RBIF_SW_VLEVEL_DEFAULT 0x20 + +#define REG_RBCPR_STEP_QUOT 0x80 +#define RBCPR_STEP_QUOT_STEPQUOT_MASK GENMASK(7, 0) +#define RBCPR_STEP_QUOT_IDLE_CLK_MASK GENMASK(3, 0) +#define RBCPR_STEP_QUOT_IDLE_CLK_SHIFT 8 + +/* RBCPR Control Register */ +#define REG_RBCPR_CTL 0x90 + +#define RBCPR_CTL_LOOP_EN BIT(0) +#define RBCPR_CTL_TIMER_EN BIT(3) +#define RBCPR_CTL_SW_AUTO_CONT_ACK_EN BIT(5) +#define RBCPR_CTL_SW_AUTO_CONT_NACK_DN_EN BIT(6) +#define RBCPR_CTL_COUNT_MODE BIT(10) +#define RBCPR_CTL_UP_THRESHOLD_MASK GENMASK(3, 0) +#define RBCPR_CTL_UP_THRESHOLD_SHIFT 24 +#define RBCPR_CTL_DN_THRESHOLD_MASK GENMASK(3, 0) +#define RBCPR_CTL_DN_THRESHOLD_SHIFT 28 + +/* RBCPR Ack/Nack Response */ +#define REG_RBIF_CONT_ACK_CMD 0x98 +#define REG_RBIF_CONT_NACK_CMD 0x9c + +/* RBCPR Result status Register */ +#define REG_RBCPR_RESULT_0 0xa0 + +#define RBCPR_RESULT0_BUSY_SHIFT 19 +#define RBCPR_RESULT0_BUSY_MASK BIT(RBCPR_RESULT0_BUSY_SHIFT) +#define RBCPR_RESULT0_ERROR_LT0_SHIFT 18 +#define RBCPR_RESULT0_ERROR_SHIFT 6 +#define RBCPR_RESULT0_ERROR_MASK GENMASK(11, 0) +#define RBCPR_RESULT0_ERROR_STEPS_SHIFT 2 +#define RBCPR_RESULT0_ERROR_STEPS_MASK GENMASK(3, 0) +#define RBCPR_RESULT0_STEP_UP_SHIFT 1 + +/* RBCPR Interrupt Control Register */ +#define REG_RBIF_IRQ_EN(n) (0x100 + 4 * n) +#define REG_RBIF_IRQ_CLEAR 0x110 +#define REG_RBIF_IRQ_STATUS 0x114 + +#define CPR_INT_DONE BIT(0) +#define CPR_INT_MIN BIT(1) +#define CPR_INT_DOWN BIT(2) +#define CPR_INT_MID BIT(3) +#define CPR_INT_UP BIT(4) +#define CPR_INT_MAX BIT(5) +#define CPR_INT_CLAMP BIT(6) +#define CPR_INT_ALL (CPR_INT_DONE | CPR_INT_MIN | CPR_INT_DOWN | \ + CPR_INT_MID | CPR_INT_UP | CPR_INT_MAX | CPR_INT_CLAMP) +#define CPR_INT_DEFAULT (CPR_INT_UP | CPR_INT_DOWN) + +#define CPR_NUM_RING_OSC 8 + +/* RBCPR Clock Control Register */ +#define RBCPR_CLK_SEL_MASK BIT(-1) +#define RBCPR_CLK_SEL_19P2_MHZ 0 +#define RBCPR_CLK_SEL_AHB_CLK BIT(0) + +/* CPR eFuse parameters */ +#define CPR_FUSE_TARGET_QUOT_BITS_MASK GENMASK(11, 0) + +#define CPR_FUSE_MIN_QUOT_DIFF 50 + +#define SPEED_BIN_NONE UINT_MAX + +#define FUSE_REVISION_UNKNOWN (-1) +#define FUSE_MAP_NO_MATCH (-1) +#define FUSE_PARAM_MATCH_ANY 0xffffffff + +/** + * enum vdd_mx_vmin_method - Method to determine vmin for vdd-mx + * @VDD_MX_VMIN_APC: Use APC voltage + * @VDD_MX_VMIN_APC_CORNER_CEILING: Use PVS corner ceiling voltage + * @VDD_MX_VMIN_APC_SLOW_CORNER_CEILING: Use slow speed corner ceiling + * @VDD_MX_VMIN_MX_VMAX: Use specified vdd-mx-vmax voltage + * @VDD_MX_VMIN_APC_CORNER_MAP: Use APC corner mapped MX voltage + */ +enum vdd_mx_vmin_method { + VDD_MX_VMIN_APC, + VDD_MX_VMIN_APC_CORNER_CEILING, + VDD_MX_VMIN_APC_SLOW_CORNER_CEILING, + VDD_MX_VMIN_MX_VMAX, + VDD_MX_VMIN_APC_CORNER_MAP, +}; +/* TODO: Trim these above to used values */ + +enum voltage_change_dir { + NO_CHANGE, + DOWN, + UP, +}; + +struct qfprom_offset { + u16 offset; + u8 width; + u8 shift; +}; + +struct cpr_fuse { + struct qfprom_offset ring_osc; + struct qfprom_offset init_voltage; + struct qfprom_offset quotient; + struct qfprom_offset quotient_offset; +}; + +struct fuse_corner_data { + int ref_uV; + int max_uV; + int min_uV; + int max_quot_scale; + int quot_offset; + int quot_scale; + int max_volt_scale; + int vdd_mx_req; +}; + +struct cpr_fuses { + struct qfprom_offset redundant; + u8 redundant_value; + int init_voltage_step; + struct fuse_corner_data *fuse_corner_data; + struct cpr_fuse *cpr_fuse; + struct qfprom_offset *disable; +}; + +struct pvs_bin { + int *uV; +}; + +struct pvs_fuses { + struct qfprom_offset redundant; + u8 redundant_value; + struct qfprom_offset *pvs_fuse; + struct pvs_bin *pvs_bins; +}; + +struct corner_data { + unsigned int fuse_corner; + unsigned long freq; +}; + +struct freq_plan { + u32 speed_bin; + u32 pvs_version; + const struct corner_data **plan; +}; + +struct fuse_conditional_min_volt { + struct qfprom_offset redundant; + u8 expected; + int min_uV; +}; + +struct fuse_uplift_wa { + struct qfprom_offset redundant; + u8 expected; + int uV; + int *quot; + int max_uV; + int speed_bin; +}; + +struct corner_override { + u32 speed_bin; + u32 pvs_version; + int *max_uV; + int *min_uV; +}; + +struct corner_adjustment { + u32 speed_bin; + u32 pvs_version; + u32 cpr_rev; + u8 *ring_osc_idx; + int *fuse_quot; + int *fuse_quot_diff; + int *fuse_quot_min; + int *fuse_quot_offset; + int *fuse_init_uV; + int *quot; + int *init_uV; + bool disable_closed_loop; +}; + +struct cpr_desc { + unsigned int num_fuse_corners; + unsigned int num_corners; + enum vdd_mx_vmin_method vdd_mx_vmin_method; + int vdd_mx_vmax; + int min_diff_quot; + int *step_quot; + struct cpr_fuses cpr_fuses; + struct qfprom_offset fuse_revision; + struct qfprom_offset speed_bin; + struct qfprom_offset pvs_version; + struct corner_data *corner_data; + struct freq_plan *freq_plans; + size_t num_freq_plans; + struct pvs_fuses *pvs_fuses; + struct fuse_conditional_min_volt *min_volt_fuse; + struct fuse_uplift_wa *uplift_wa; + struct corner_override *corner_overrides; + size_t num_corner_overrides; + struct corner_adjustment *adjustments; + size_t num_adjustments; + bool reduce_to_fuse_uV; + bool reduce_to_corner_uV; +}; + +struct acc_desc { + unsigned int enable_reg; + u32 enable_mask; + + struct reg_default *settings; + struct reg_default *override_settings; + int num_regs_per_fuse; + + struct qfprom_offset override; + u8 override_value; +}; + +struct fuse_corner { + int min_uV; + int max_uV; + int uV; + int quot; + int step_quot; + const struct reg_default *accs; + int num_accs; + int vdd_mx_req; + unsigned long max_freq; + u8 ring_osc_idx; +}; + +struct corner { + int min_uV; + int max_uV; + int uV; + int last_uV; + int quot_adjust; + u32 save_ctl; + u32 save_irq; + unsigned long freq; + struct fuse_corner *fuse_corner; +}; + +struct cpr_drv { + unsigned int num_fuse_corners; + unsigned int num_corners; + + unsigned int nb_count; + struct notifier_block cpufreq_nb; + bool switching_opp; + struct notifier_block reg_nb; + + unsigned int ref_clk_khz; + unsigned int timer_delay_us; + unsigned int timer_cons_up; + unsigned int timer_cons_down; + unsigned int up_threshold; + unsigned int down_threshold; + unsigned int idle_clocks; + unsigned int gcnt_us; + unsigned int vdd_apc_step_up_limit; + unsigned int vdd_apc_step_down_limit; + unsigned int clamp_timer_interval; + int ceiling_max; + enum vdd_mx_vmin_method vdd_mx_vmin_method; + int vdd_mx_vmax; + + struct mutex lock; + void __iomem *base; + struct corner *corner; + struct regulator *vdd_apc; + struct regulator *vdd_mx; + struct clk *cpu_clk; + struct device *cpu_dev; + struct regmap *tcsr; + bool loop_disabled; + bool suspended; + u32 gcnt; + unsigned long flags; +#define FLAGS_IGNORE_1ST_IRQ_STATUS BIT(0) + + struct fuse_corner *fuse_corners; + struct corner *corners; +}; + +static bool cpr_is_allowed(struct cpr_drv *drv) +{ + if (drv->loop_disabled) /* || disabled in software */ + return false; + else + return true; +} + +static void cpr_write(struct cpr_drv *drv, u32 offset, u32 value) +{ + writel_relaxed(value, drv->base + offset); +} + +static u32 cpr_read(struct cpr_drv *drv, u32 offset) +{ + return readl_relaxed(drv->base + offset); +} + +static void +cpr_masked_write(struct cpr_drv *drv, u32 offset, u32 mask, u32 value) +{ + u32 val; + + val = readl_relaxed(drv->base + offset); + val &= ~mask; + val |= value & mask; + writel_relaxed(val, drv->base + offset); +} + +static void cpr_irq_clr(struct cpr_drv *drv) +{ + cpr_write(drv, REG_RBIF_IRQ_CLEAR, CPR_INT_ALL); +} + +static void cpr_irq_clr_nack(struct cpr_drv *drv) +{ + cpr_irq_clr(drv); + cpr_write(drv, REG_RBIF_CONT_NACK_CMD, 1); +} + +static void cpr_irq_clr_ack(struct cpr_drv *drv) +{ + cpr_irq_clr(drv); + cpr_write(drv, REG_RBIF_CONT_ACK_CMD, 1); +} + +static void cpr_irq_set(struct cpr_drv *drv, u32 int_bits) +{ + cpr_write(drv, REG_RBIF_IRQ_EN(0), int_bits); +} + +static void cpr_ctl_modify(struct cpr_drv *drv, u32 mask, u32 value) +{ + cpr_masked_write(drv, REG_RBCPR_CTL, mask, value); +} + +static void cpr_ctl_enable(struct cpr_drv *drv, struct corner *corner) +{ + u32 val, mask; + + if (drv->suspended) + return; + + /* Program Consecutive Up & Down */ + val = drv->timer_cons_down << RBIF_TIMER_ADJ_CONS_DOWN_SHIFT; + val |= drv->timer_cons_up << RBIF_TIMER_ADJ_CONS_UP_SHIFT; + mask = RBIF_TIMER_ADJ_CONS_UP_MASK | RBIF_TIMER_ADJ_CONS_DOWN_MASK; + cpr_masked_write(drv, REG_RBIF_TIMER_ADJUST, mask, val); + cpr_masked_write(drv, REG_RBCPR_CTL, + RBCPR_CTL_SW_AUTO_CONT_NACK_DN_EN | + RBCPR_CTL_SW_AUTO_CONT_ACK_EN, + corner->save_ctl); + cpr_irq_set(drv, corner->save_irq); + + if (cpr_is_allowed(drv) /*&& drv->vreg_enabled */ && + corner->max_uV > corner->min_uV) + val = RBCPR_CTL_LOOP_EN; + else + val = 0; + cpr_ctl_modify(drv, RBCPR_CTL_LOOP_EN, val); +} + +static void cpr_ctl_disable(struct cpr_drv *drv) +{ + if (drv->suspended) + return; + + cpr_irq_set(drv, 0); + cpr_ctl_modify(drv, RBCPR_CTL_SW_AUTO_CONT_NACK_DN_EN | + RBCPR_CTL_SW_AUTO_CONT_ACK_EN, 0); + cpr_masked_write(drv, REG_RBIF_TIMER_ADJUST, + RBIF_TIMER_ADJ_CONS_UP_MASK | + RBIF_TIMER_ADJ_CONS_DOWN_MASK, 0); + cpr_irq_clr(drv); + cpr_write(drv, REG_RBIF_CONT_ACK_CMD, 1); + cpr_write(drv, REG_RBIF_CONT_NACK_CMD, 1); + cpr_ctl_modify(drv, RBCPR_CTL_LOOP_EN, 0); +} + +static bool cpr_ctl_is_enabled(struct cpr_drv *drv) +{ + u32 reg_val; + + reg_val = cpr_read(drv, REG_RBCPR_CTL); + return reg_val & RBCPR_CTL_LOOP_EN; +} + +static bool cpr_ctl_is_busy(struct cpr_drv *drv) +{ + u32 reg_val; + + reg_val = cpr_read(drv, REG_RBCPR_RESULT_0); + return reg_val & RBCPR_RESULT0_BUSY_MASK; +} + +static void cpr_corner_save(struct cpr_drv *drv, struct corner *corner) +{ + corner->save_ctl = cpr_read(drv, REG_RBCPR_CTL); + corner->save_irq = cpr_read(drv, REG_RBIF_IRQ_EN(0)); +} + +static void cpr_corner_restore(struct cpr_drv *drv, struct corner *corner) +{ + u32 gcnt, ctl, irq, ro_sel, step_quot; + struct fuse_corner *fuse = corner->fuse_corner; + int i; + + ro_sel = fuse->ring_osc_idx; + gcnt = drv->gcnt; + gcnt |= fuse->quot - corner->quot_adjust; + + /* Program the step quotient and idle clocks */ + step_quot = drv->idle_clocks << RBCPR_STEP_QUOT_IDLE_CLK_SHIFT; + step_quot |= fuse->step_quot; + cpr_write(drv, REG_RBCPR_STEP_QUOT, step_quot); + + /* Clear the target quotient value and gate count of all ROs */ + for (i = 0; i < CPR_NUM_RING_OSC; i++) + cpr_write(drv, REG_RBCPR_GCNT_TARGET(i), 0); + + cpr_write(drv, REG_RBCPR_GCNT_TARGET(ro_sel), gcnt); + ctl = corner->save_ctl; + cpr_write(drv, REG_RBCPR_CTL, ctl); + irq = corner->save_irq; + cpr_irq_set(drv, irq); + pr_debug("gcnt = 0x%08x, ctl = 0x%08x, irq = 0x%08x\n", gcnt, ctl, irq); +} + +static void cpr_corner_switch(struct cpr_drv *drv, struct corner *corner) +{ + if (drv->corner == corner) + return; + + cpr_corner_restore(drv, corner); +} + +static int +cpr_mx_get(struct cpr_drv *drv, struct fuse_corner *fuse, int apc_volt) +{ + int vdd_mx; + struct fuse_corner *highest_fuse; + + highest_fuse = &drv->fuse_corners[drv->num_fuse_corners - 1]; + + switch (drv->vdd_mx_vmin_method) { + case VDD_MX_VMIN_APC: + vdd_mx = apc_volt; + break; + case VDD_MX_VMIN_APC_CORNER_CEILING: + vdd_mx = fuse->max_uV; + break; + case VDD_MX_VMIN_APC_SLOW_CORNER_CEILING: + vdd_mx = highest_fuse->max_uV; + break; + case VDD_MX_VMIN_MX_VMAX: + vdd_mx = drv->vdd_mx_vmax; + break; + case VDD_MX_VMIN_APC_CORNER_MAP: + vdd_mx = fuse->vdd_mx_req; + break; + default: + BUG(); + } + + return vdd_mx; +} + +static void cpr_set_acc(struct regmap *tcsr, struct fuse_corner *f, + struct fuse_corner *end) +{ + if (f < end) { + for (f += 1; f <= end; f++) + regmap_multi_reg_write(tcsr, f->accs, f->num_accs); + } else { + for (f -= 1; f >= end; f--) + regmap_multi_reg_write(tcsr, f->accs, f->num_accs); + } +} + +static int cpr_pre_voltage(struct cpr_drv *drv, + struct fuse_corner *fuse_corner, + enum voltage_change_dir dir, int vdd_mx_vmin) +{ + int ret = 0; + struct fuse_corner *prev_fuse_corner = drv->corner->fuse_corner; + + if (drv->tcsr && dir == DOWN) + cpr_set_acc(drv->tcsr, prev_fuse_corner, fuse_corner); + + if (vdd_mx_vmin && dir == UP) + ret = qcom_rpm_set_corner(drv->vdd_mx, vdd_mx_vmin); + + return ret; +} + +static int cpr_post_voltage(struct cpr_drv *drv, + struct fuse_corner *fuse_corner, + enum voltage_change_dir dir, int vdd_mx_vmin) +{ + int ret = 0; + struct fuse_corner *prev_fuse_corner = drv->corner->fuse_corner; + + if (drv->tcsr && dir == UP) + cpr_set_acc(drv->tcsr, prev_fuse_corner, fuse_corner); + + if (vdd_mx_vmin && dir == DOWN) + ret = qcom_rpm_set_corner(drv->vdd_mx, vdd_mx_vmin); + + return ret; +} + +static int cpr_regulator_notifier(struct notifier_block *nb, + unsigned long event, void *d) +{ + struct cpr_drv *drv = container_of(nb, struct cpr_drv, reg_nb); + u32 val, mask; + int last_uV, new_uV; + + switch (event) { + case REGULATOR_EVENT_VOLTAGE_CHANGE: + new_uV = (int)(uintptr_t)d; + break; + default: + return 0; + } + + mutex_lock(&drv->lock); + + last_uV = drv->corner->last_uV; + + if (drv->switching_opp) { + goto unlock; + } else if (last_uV < new_uV) { + /* Disable auto nack down */ + mask = RBCPR_CTL_SW_AUTO_CONT_NACK_DN_EN; + val = 0; + } else if (last_uV > new_uV) { + /* Restore default threshold for UP */ + mask = RBCPR_CTL_UP_THRESHOLD_MASK; + mask <<= RBCPR_CTL_UP_THRESHOLD_SHIFT; + val = drv->up_threshold; + val <<= RBCPR_CTL_UP_THRESHOLD_SHIFT; + } else { /* Somehow it's the same? */ + goto unlock; + } + + cpr_ctl_modify(drv, mask, val); + + /* Re-enable default interrupts */ + cpr_irq_set(drv, CPR_INT_DEFAULT); + + /* Ack */ + cpr_irq_clr_ack(drv); + + /* Save register values for the corner */ + cpr_corner_save(drv, drv->corner); + drv->corner->last_uV = new_uV; +unlock: + mutex_unlock(&drv->lock); + + return 0; +} + +static int cpr_scale(struct cpr_drv *drv, enum voltage_change_dir dir) +{ + u32 val, error_steps, reg_mask; + int last_uV, new_uV, step_uV; + struct corner *corner; + struct fuse_corner *fuse; + + //step_uV = regulator_get_linear_step(drv->vdd_apc); + step_uV = 12500; /*TODO: Get step volt here */ + corner = drv->corner; + fuse = corner->fuse_corner; + + val = cpr_read(drv, REG_RBCPR_RESULT_0); + + error_steps = val >> RBCPR_RESULT0_ERROR_STEPS_SHIFT; + error_steps &= RBCPR_RESULT0_ERROR_STEPS_MASK; + last_uV = corner->last_uV; + + if (dir == UP) { + if (drv->clamp_timer_interval && + error_steps < drv->up_threshold) { + /* + * Handle the case where another measurement started + * after the interrupt was triggered due to a core + * exiting from power collapse. + */ + error_steps = max(drv->up_threshold, + drv->vdd_apc_step_up_limit); + } + + if (last_uV >= corner->max_uV) { + cpr_irq_clr_nack(drv); + + /* Maximize the UP threshold */ + reg_mask = RBCPR_CTL_UP_THRESHOLD_MASK; + reg_mask <<= RBCPR_CTL_UP_THRESHOLD_SHIFT; + val = reg_mask; + cpr_ctl_modify(drv, reg_mask, val); + + /* Disable UP interrupt */ + cpr_irq_set(drv, CPR_INT_DEFAULT & ~CPR_INT_UP); + + return 0; + } + + if (error_steps > drv->vdd_apc_step_up_limit) + error_steps = drv->vdd_apc_step_up_limit; + + /* Calculate new voltage */ + new_uV = last_uV + error_steps * step_uV; + if (new_uV > corner->max_uV) + new_uV = corner->max_uV; + } else if (dir == DOWN) { + if (drv->clamp_timer_interval + && error_steps < drv->down_threshold) { + /* + * Handle the case where another measurement started + * after the interrupt was triggered due to a core + * exiting from power collapse. + */ + error_steps = max(drv->down_threshold, + drv->vdd_apc_step_down_limit); + } + + if (last_uV <= corner->min_uV) { + cpr_irq_clr_nack(drv); + + /* Enable auto nack down */ + reg_mask = RBCPR_CTL_SW_AUTO_CONT_NACK_DN_EN; + val = RBCPR_CTL_SW_AUTO_CONT_NACK_DN_EN; + + cpr_ctl_modify(drv, reg_mask, val); + + /* Disable DOWN interrupt */ + cpr_irq_set(drv, CPR_INT_DEFAULT & ~CPR_INT_DOWN); + + return 0; + } + + if (error_steps > drv->vdd_apc_step_down_limit) + error_steps = drv->vdd_apc_step_down_limit; + + /* Calculate new voltage */ + new_uV = last_uV - error_steps * step_uV; + if (new_uV < corner->min_uV) + new_uV = corner->min_uV; + } + + return new_uV; +} + +static irqreturn_t cpr_irq_handler(int irq, void *dev) +{ + struct cpr_drv *drv = dev; + u32 val; + int new_uV = 0; + struct corner *corner; + + mutex_lock(&drv->lock); + + val = cpr_read(drv, REG_RBIF_IRQ_STATUS); + if (drv->flags & FLAGS_IGNORE_1ST_IRQ_STATUS) + val = cpr_read(drv, REG_RBIF_IRQ_STATUS); + + pr_info("IRQ_STATUS = %#02x\n", val); + + if (!cpr_ctl_is_enabled(drv)) { + pr_debug("CPR is disabled\n"); + goto unlock; + } else if (cpr_ctl_is_busy(drv) && !drv->clamp_timer_interval) { + pr_debug("CPR measurement is not ready\n"); + goto unlock; + } else if (!cpr_is_allowed(drv)) { + val = cpr_read(drv, REG_RBCPR_CTL); + pr_err_ratelimited("Interrupt broken? RBCPR_CTL = %#02x\n", + val); + goto unlock; + } + + /* Following sequence of handling is as per each IRQ's priority */ + if (val & CPR_INT_UP) { + new_uV = cpr_scale(drv, UP); + } else if (val & CPR_INT_DOWN) { + new_uV = cpr_scale(drv, DOWN); + } else if (val & CPR_INT_MIN) { + cpr_irq_clr_nack(drv); + } else if (val & CPR_INT_MAX) { + cpr_irq_clr_nack(drv); + } else if (val & CPR_INT_MID) { + /* RBCPR_CTL_SW_AUTO_CONT_ACK_EN is enabled */ + pr_debug("IRQ occurred for Mid Flag\n"); + } else { + pr_debug("IRQ occurred for unknown flag (%#08x)\n", val); + } + + /* Save register values for the corner */ + corner = drv->corner; + cpr_corner_save(drv, corner); +unlock: + mutex_unlock(&drv->lock); + + if (new_uV) + dev_pm_opp_adjust_voltage(drv->cpu_dev, corner->freq, new_uV); + + return IRQ_HANDLED; +} + +/* + * TODO: Register for hotplug notifier and turn on/off CPR when CPUs are offline + */ +static int cpr_enable(struct cpr_drv *drv) +{ + int ret; + + /* Enable dependency power before vdd_apc */ + if (drv->vdd_mx) { + ret = regulator_enable(drv->vdd_mx); + if (ret) + return ret; + } + + ret = regulator_enable(drv->vdd_apc); + if (ret) + return ret; + + mutex_lock(&drv->lock); + //drv->vreg_enabled = true; + if (cpr_is_allowed(drv) && drv->corner) { + cpr_irq_clr(drv); + cpr_corner_restore(drv, drv->corner); + cpr_ctl_enable(drv, drv->corner); + } + mutex_unlock(&drv->lock); + pr_info("CPR is enableed!\n"); + + return 0; +} + +static int cpr_disable(struct cpr_drv *drv) +{ + int ret; + + ret = regulator_disable(drv->vdd_apc); + if (ret) + return ret; + + if (drv->vdd_mx) + ret = regulator_disable(drv->vdd_mx); + if (ret) + return ret; + + mutex_lock(&drv->lock); + //drv->vreg_enabled = false; + if (cpr_is_allowed(drv)) + cpr_ctl_disable(drv); + mutex_unlock(&drv->lock); + + return 0; +} + +#ifdef CONFIG_PM_SLEEP +static int cpr_suspend(struct device *dev) +{ + struct cpr_drv *drv = platform_get_drvdata(to_platform_device(dev)); + + if (cpr_is_allowed(drv)) { + mutex_lock(&drv->lock); + cpr_ctl_disable(drv); + cpr_irq_clr(drv); + drv->suspended = true; + mutex_unlock(&drv->lock); + } + + return 0; +} + +static int cpr_resume(struct device *dev) +{ + struct cpr_drv *drv = platform_get_drvdata(to_platform_device(dev)); + + if (cpr_is_allowed(drv)) { + mutex_lock(&drv->lock); + drv->suspended = false; + cpr_irq_clr(drv); + cpr_ctl_enable(drv, drv->corner); + mutex_unlock(&drv->lock); + } + + return 0; +} +#endif + +static SIMPLE_DEV_PM_OPS(cpr_pm_ops, cpr_suspend, cpr_resume); + +static int cpr_config(struct cpr_drv *drv) +{ + int i; + u32 val, gcnt; + struct corner *corner; + + /* Disable interrupt and CPR */ + cpr_write(drv, REG_RBIF_IRQ_EN(0), 0); + cpr_write(drv, REG_RBCPR_CTL, 0); + + /* Program the default HW Ceiling, Floor and vlevel */ + val = RBIF_LIMIT_CEILING_DEFAULT << RBIF_LIMIT_CEILING_SHIFT; + val |= RBIF_LIMIT_FLOOR_DEFAULT; + cpr_write(drv, REG_RBIF_LIMIT, val); + cpr_write(drv, REG_RBIF_SW_VLEVEL, RBIF_SW_VLEVEL_DEFAULT); + + /* Clear the target quotient value and gate count of all ROs */ + for (i = 0; i < CPR_NUM_RING_OSC; i++) + cpr_write(drv, REG_RBCPR_GCNT_TARGET(i), 0); + + /* Init and save gcnt */ + gcnt = (drv->ref_clk_khz * drv->gcnt_us) / 1000; + gcnt = gcnt & RBCPR_GCNT_TARGET_GCNT_MASK; + gcnt <<= RBCPR_GCNT_TARGET_GCNT_SHIFT; + drv->gcnt = gcnt; + + /* Program the delay count for the timer */ + val = (drv->ref_clk_khz * drv->timer_delay_us) / 1000; + cpr_write(drv, REG_RBCPR_TIMER_INTERVAL, val); + pr_info("Timer count: 0x%0x (for %d us)\n", val, drv->timer_delay_us); + + /* Program Consecutive Up & Down */ + val = drv->timer_cons_down << RBIF_TIMER_ADJ_CONS_DOWN_SHIFT; + val |= drv->timer_cons_up << RBIF_TIMER_ADJ_CONS_UP_SHIFT; + val |= drv->clamp_timer_interval << RBIF_TIMER_ADJ_CLAMP_INT_SHIFT; + cpr_write(drv, REG_RBIF_TIMER_ADJUST, val); + + /* Program the control register */ + val = drv->up_threshold << RBCPR_CTL_UP_THRESHOLD_SHIFT; + val |= drv->down_threshold << RBCPR_CTL_DN_THRESHOLD_SHIFT; + val |= RBCPR_CTL_TIMER_EN | RBCPR_CTL_COUNT_MODE; + val |= RBCPR_CTL_SW_AUTO_CONT_ACK_EN; + cpr_write(drv, REG_RBCPR_CTL, val); + + for (i = 0; i < drv->num_corners; i++) { + corner = &drv->corners[i]; + corner->save_ctl = val; + corner->save_irq = CPR_INT_DEFAULT; + } + + cpr_irq_set(drv, CPR_INT_DEFAULT); + + val = cpr_read(drv, REG_RBCPR_VERSION); + if (val <= RBCPR_VER_2) + drv->flags |= FLAGS_IGNORE_1ST_IRQ_STATUS; + + return 0; +} + +/* Called twice for each CPU in policy, one pre and one post event */ +static int +cpr_cpufreq_notifier(struct notifier_block *nb, unsigned long event, void *f) +{ + struct cpr_drv *drv = container_of(nb, struct cpr_drv, cpufreq_nb); + struct cpufreq_freqs *freqs = f; + unsigned long old = freqs->old * 1000; + unsigned long new = freqs->new * 1000; + struct corner *corner, *end; + enum voltage_change_dir dir; + int ret = 0, new_uV; + int vdd_mx_vmin = 0; + struct fuse_corner *fuse_corner; + + /* Determine direction */ + if (old > new) + dir = DOWN; + else if (old < new) + dir = UP; + else + dir = NO_CHANGE; + + /* Determine new corner we're going to */ + corner = drv->corners; + end = &corner[drv->num_corners - 1]; + for (; corner <= end; corner++) + if (corner->freq == new) + break; + + if (corner > end) + return -EINVAL; + + fuse_corner = corner->fuse_corner; + + if (cpr_is_allowed(drv)) { + new_uV = corner->last_uV; + } else { + new_uV = corner->uV; + } + + if (dir != NO_CHANGE && drv->vdd_mx) + vdd_mx_vmin = cpr_mx_get(drv, fuse_corner, new_uV); + + mutex_lock(&drv->lock); + if (event == CPUFREQ_PRECHANGE) { + if (drv->nb_count++) + goto unlock; + + pr_info("Pre change [%ld] %p @ %lu?\n", corner - drv->corners, + corner, corner->freq); + if (cpr_is_allowed(drv)) + cpr_ctl_disable(drv); + + ret = cpr_pre_voltage(drv, fuse_corner, dir, vdd_mx_vmin); + if (ret) + goto unlock; + + drv->switching_opp = true; + } + + if (event == CPUFREQ_POSTCHANGE) { + if (--drv->nb_count) + goto unlock; + + pr_info("Post change [%ld] %p @ %lu?\n", corner - drv->corners, + corner, corner->freq); + + ret = cpr_post_voltage(drv, fuse_corner, dir, vdd_mx_vmin); + if (ret) + goto unlock; + + if (cpr_is_allowed(drv) /* && drv->vreg_enabled */) { + cpr_irq_clr(drv); + cpr_corner_switch(drv, corner); + cpr_ctl_enable(drv, corner); + } + + drv->corner = corner; + drv->switching_opp = false; + } +unlock: + mutex_unlock(&drv->lock); + + return ret; +} + +static u32 +cpr_read_efuse(struct nvmem_device *qfprom, const struct qfprom_offset *efuse) +{ + u64 buffer = 0; + size_t bytes; + int ret; + + bytes = DIV_ROUND_UP(efuse->width + efuse->shift, BITS_PER_BYTE); + ret = nvmem_device_read(qfprom, efuse->offset, bytes, &buffer); + + buffer >>= efuse->shift; + buffer &= GENMASK(efuse->width - 1, 0); + + return buffer; +} + +static void +cpr_populate_ring_osc_idx(const struct cpr_fuse *fuses, struct cpr_drv *drv, + struct nvmem_device *qfprom) +{ + struct fuse_corner *fuse = drv->fuse_corners; + struct fuse_corner *end = fuse + drv->num_fuse_corners; + + for (; fuse < end; fuse++, fuses++) + fuse->ring_osc_idx = cpr_read_efuse(qfprom, &fuses->ring_osc); +} + +static const struct corner_adjustment *cpr_find_adjustment(u32 speed_bin, + u32 pvs_version, u32 cpr_rev, const struct cpr_desc *desc, + const struct cpr_drv *drv) +{ + int i, j; + u32 val, ro; + struct corner_adjustment *a; + + for (i = 0; i < desc->num_adjustments; i++) { + a = &desc->adjustments[i]; + + if (a->speed_bin != speed_bin && + a->speed_bin != FUSE_PARAM_MATCH_ANY) + continue; + if (a->pvs_version != pvs_version && + a->pvs_version != FUSE_PARAM_MATCH_ANY) + continue; + if (a->cpr_rev != cpr_rev && + a->cpr_rev != FUSE_PARAM_MATCH_ANY) + continue; + for (j = 0; j < drv->num_fuse_corners; j++) { + val = a->ring_osc_idx[j]; + ro = drv->fuse_corners[j].ring_osc_idx; + if (val != ro && val != FUSE_PARAM_MATCH_ANY) + break; + } + if (j == drv->num_fuse_corners) + return a; + } + + return NULL; +} + +static void cpr_fuse_corner_init(struct cpr_drv *drv, + const struct cpr_desc *desc, + struct nvmem_device *qfprom, + const struct cpr_fuse *fuses, u32 speed, + const struct corner_adjustment *adjustments, + const struct acc_desc *acc_desc) +{ + int i; + unsigned int idx = 0; + unsigned int step_volt; + int steps, step_size_uv; + const struct fuse_corner_data *fdata; + struct fuse_corner *fuse, *end, *prev; + const struct qfprom_offset *pvs_efuse; + const struct qfprom_offset *init_v_efuse; + const struct qfprom_offset *redun; + const struct fuse_conditional_min_volt *min_v; + const struct fuse_uplift_wa *up; + bool do_min_v = false, do_uplift = false; + const int *pvs_uV = NULL; + const int *adj_uV, *adj_quot, *adj_min, *min_diff_quot; + const int *step_quot; + int uV, diff; + u32 bits, bin; + u32 min_uV; + u8 expected; + const struct reg_default *accs; + + redun = &acc_desc->override; + expected = acc_desc->override_value; + if (redun->width && cpr_read_efuse(qfprom, redun) == expected) + accs = acc_desc->override_settings; + else + accs = acc_desc->settings; + + /* Figure out if we should apply workarounds */ + min_v = desc->min_volt_fuse; + do_min_v = min_v && + cpr_read_efuse(qfprom, &min_v->redundant) == min_v->expected; + if (do_min_v) + min_uV = min_v->min_uV; + + up = desc->uplift_wa; + if (!do_min_v && up) + if (cpr_read_efuse(qfprom, &up->redundant) == up->expected) + do_uplift = up->speed_bin == speed; + + adj_uV = adjustments ? adjustments->fuse_init_uV : NULL; + adj_quot = adjustments ? adjustments->fuse_quot : NULL; + adj_min = adjustments ? adjustments->fuse_quot_min : NULL; + min_diff_quot = adjustments ? adjustments->fuse_quot_diff : NULL; + fuse = drv->fuse_corners; + end = &fuse[drv->num_fuse_corners - 1]; + fdata = desc->cpr_fuses.fuse_corner_data; + step_quot = desc->step_quot; + + /* + * The initial voltage for each fuse corner may be determined by one of + * two ways. Either initial voltages are encoded for each fuse corner + * in a dedicated fuse per fuse corner (fuses::init_voltage), or we + * use the PVS bin fuse to use a table of initial voltages (pvs_uV). + */ + if (fuses->init_voltage.width) { + //step_volt = regulator_get_linear_step(drv->vdd_apc); + step_volt = 12500; /* TODO: Replace with ^ when apc_reg ready */ + step_size_uv = desc->cpr_fuses.init_voltage_step; + } else { + redun = &desc->pvs_fuses->redundant; + expected = desc->pvs_fuses->redundant_value; + if (redun->width) + idx = !!(cpr_read_efuse(qfprom, redun) == expected); + + pvs_efuse = &desc->pvs_fuses->pvs_fuse[idx]; + bin = cpr_read_efuse(qfprom, pvs_efuse); + pvs_uV = desc->pvs_fuses->pvs_bins[bin].uV; + } + + /* Populate fuse_corner voltage and ring_osc_idx members */ + prev = NULL; + for (i = 0; fuse <= end; fuse++, fuses++, i++) { + if (pvs_uV) { + uV = pvs_uV[i]; + } else { + init_v_efuse = &fuses->init_voltage; + bits = cpr_read_efuse(qfprom, init_v_efuse); + /* Not two's complement.. instead highest bit is sign */ + steps = bits & BIT(init_v_efuse->width - 1) ? -1 : 1; + steps *= bits & ~BIT(init_v_efuse->width - 1); + + uV = fdata[i].ref_uV + steps * step_size_uv; + uV = DIV_ROUND_UP(uV, step_volt) * step_volt; + } + + if (adj_uV) + uV += adj_uV[i]; + + fuse->min_uV = fdata[i].min_uV; + fuse->max_uV = fdata[i].max_uV; + + if (do_min_v) { + if (fuse->max_uV < min_uV) { + fuse->max_uV = min_uV; + fuse->min_uV = min_uV; + } else if (fuse->min_uV < min_uV) { + fuse->min_uV = min_uV; + } + } + + fuse->uV = clamp(uV, fuse->min_uV, fuse->max_uV); + + if (fuse == end) { + if (do_uplift) { + end->uV += up->uV; + end->uV = clamp(end->uV, 0, up->max_uV); + } + /* + * Allow the highest fuse corner's PVS voltage to + * define the ceiling voltage for that corner in order + * to support SoC's in which variable ceiling values + * are required. + */ + end->max_uV = max(end->max_uV, end->uV); + } + + /* Unpack the target quotient by scaling. */ + fuse->quot = cpr_read_efuse(qfprom, &fuses->quotient); + fuse->quot *= fdata[i].quot_scale; + fuse->quot += fdata[i].quot_offset; + + if (adj_quot) { + fuse->quot += adj_quot[i]; + + if (prev && min_diff_quot) { + diff = min_diff_quot[i]; + if (fuse->quot - prev->quot <= diff) + fuse->quot = prev->quot + adj_min[i]; + } + prev = fuse; + } + + if (do_uplift) + fuse->quot += up->quot[i]; + + fuse->step_quot = step_quot[fuse->ring_osc_idx]; + + fuse->accs = accs; + fuse->num_accs = acc_desc->num_regs_per_fuse; + accs += acc_desc->num_regs_per_fuse; + + fuse->vdd_mx_req = fdata[i].vdd_mx_req; + } + + /* + * Restrict all fuse corner PVS voltages based upon per corner + * ceiling and floor voltages. + */ + for (fuse = drv->fuse_corners, i = 0; fuse <= end; fuse++, i++) { + if (fuse->uV > fuse->max_uV) + fuse->uV = fuse->max_uV; + else if (fuse->uV < fuse->min_uV) + fuse->uV = fuse->min_uV; + + pr_info("fuse corner %d: [%d %d %d] RO%d quot %d squot %d\n", i, + fuse->min_uV, fuse->uV, fuse->max_uV, + fuse->ring_osc_idx, fuse->quot, + fuse->step_quot); + } + + drv->ceiling_max = end->max_uV; +} + +static int cpr_populate_opps(struct device_node *of_node, struct cpr_drv *drv, + const struct corner_data **plan) +{ + int i, j, ret, cpu; + struct device *cpu_dev; + struct device_node *np; + struct corner *corner; + const struct corner_data *p; + + for (i = 0; (np = of_parse_phandle(of_node, "qcom,cpr-cpus", i)); i++) { + for_each_possible_cpu(cpu) + if (arch_find_n_match_cpu_physical_id(np, cpu, NULL)) + break; + + of_node_put(np); + if (cpu >= nr_cpu_ids) { + pr_err("Failed to find logical CPU for %s\n", np->name); + return -EINVAL; + } + + cpu_dev = get_cpu_device(cpu); + if (!cpu_dev) + return -EINVAL; + + for (j = 0, corner = drv->corners; plan[j]; j++, corner++) { + p = plan[j]; + ret = dev_pm_opp_add(cpu_dev, p->freq, corner->uV); + if (ret) + return ret; + corner->freq = p->freq; + } + + /* + * Keep cpu_dev and its regulator and clock for monitoring + * voltage changes and updating OPPs later. + */ + if (i == 0) { + drv->cpu_dev = cpu_dev; + drv->vdd_apc = devm_regulator_get(cpu_dev, "cpu"); + if (IS_ERR(drv->vdd_apc)) + return PTR_ERR(drv->vdd_apc); + drv->cpu_clk = devm_clk_get(cpu_dev, NULL); + if (IS_ERR(drv->cpu_clk)) + return PTR_ERR(drv->cpu_clk); + } + } + + return 0; +} + +static const struct corner_data ** +find_freq_plan(const struct cpr_desc *desc, u32 speed_bin, u32 pvs_version) +{ + int i; + const struct freq_plan *p; + + for (i = 0; i < desc->num_freq_plans; i++) { + p = &desc->freq_plans[i]; + + if (p->speed_bin != speed_bin && + p->speed_bin != FUSE_PARAM_MATCH_ANY) + continue; + if (p->pvs_version != pvs_version && + p->pvs_version != FUSE_PARAM_MATCH_ANY) + continue; + + return p->plan; + } + + return NULL; + +} + +static struct corner_override *find_corner_override(const struct cpr_desc *desc, + u32 speed_bin, u32 pvs_version) +{ + int i; + struct corner_override *o; + + for (i = 0; i < desc->num_corner_overrides; i++) { + o = &desc->corner_overrides[i]; + + if (o->speed_bin != speed_bin && + o->speed_bin != FUSE_PARAM_MATCH_ANY) + continue; + if (o->pvs_version != pvs_version && + o->pvs_version != FUSE_PARAM_MATCH_ANY) + continue; + + return o; + } + + return NULL; + +} + +static void cpr_corner_init(struct cpr_drv *drv, const struct cpr_desc *desc, + const struct cpr_fuse *fuses, u32 speed_bin, + u32 pvs_version, struct nvmem_device *qfprom, + const struct corner_adjustment *adjustments, + const struct corner_data **plan) +{ + int i, fnum, quot_diff, scaling; + struct fuse_corner *fuse, *prev_fuse; + struct corner *corner, *end; + const struct corner_data *cdata, *p; + const struct fuse_corner_data *fdata; + bool apply_scaling = false; + const int *adj_quot, *adj_volt, *adj_quot_offset; + const struct qfprom_offset *quot_offset; + unsigned long freq_corner, freq_diff, freq_diff_mhz; + unsigned long freq_high, freq_low; + int volt_high; + u64 temp, temp_limit; + int step_volt = 12500; /* TODO: Get from regulator APIs */ + const struct corner_override *override; + + corner = drv->corners; + end = &corner[drv->num_corners - 1]; + cdata = desc->corner_data; + fdata = desc->cpr_fuses.fuse_corner_data; + adj_quot = adjustments ? adjustments->quot : NULL; + adj_volt = adjustments ? adjustments->init_uV : NULL; + adj_quot_offset = adjustments ? adjustments->fuse_quot_offset : NULL; + + override = find_corner_override(desc, speed_bin, pvs_version); + + /* + * Store maximum frequency for each fuse corner based on the frequency + * plan + */ + for (i = 0; plan[i]; i++) { + p = plan[i]; + freq_corner = p->freq; + fnum = p->fuse_corner; + fuse = &drv->fuse_corners[fnum]; + if (freq_corner > fuse->max_freq) + fuse->max_freq = freq_corner; + + } + + /* + * Get the quotient adjustment scaling factor, according to: + * + * scaling = min(1000 * (QUOT(corner_N) - QUOT(corner_N-1)) + * / (freq(corner_N) - freq(corner_N-1)), max_factor) + * + * QUOT(corner_N): quotient read from fuse for fuse corner N + * QUOT(corner_N-1): quotient read from fuse for fuse corner (N - 1) + * freq(corner_N): max frequency in MHz supported by fuse corner N + * freq(corner_N-1): max frequency in MHz supported by fuse corner + * (N - 1) + * + * Then walk through the corners mapped to each fuse corner + * and calculate the quotient adjustment for each one using the + * following formula: + * + * quot_adjust = (freq_max - freq_corner) * scaling / 1000 + * + * freq_max: max frequency in MHz supported by the fuse corner + * freq_corner: frequency in MHz corresponding to the corner + * scaling: calculated from above equation + * + * + * + + + * | v | + * q | f c o | f c + * u | c l | c + * o | f t | f + * t | c a | c + * | c f g | c f + * | e | + * +--------------- +---------------- + * 0 1 2 3 4 5 6 0 1 2 3 4 5 6 + * corner corner + * + * c = corner + * f = fuse corner + * + */ + for (apply_scaling = false, i = 0; corner <= end; corner++, i++) { + freq_corner = cdata[i].freq; + fnum = cdata[i].fuse_corner; + fuse = &drv->fuse_corners[fnum]; + if (fnum) + prev_fuse = &drv->fuse_corners[fnum - 1]; + else + prev_fuse = NULL; + + corner->fuse_corner = fuse; + corner->uV = fuse->uV; + if (prev_fuse && cdata[i - 1].freq == prev_fuse->max_freq) { + quot_offset = &fuses[fnum].quotient_offset; + if (quot_offset->width) { + quot_diff = cpr_read_efuse(qfprom, quot_offset); + quot_diff *= fdata->quot_scale; + if (adj_quot_offset) + quot_diff += adj_quot_offset[fnum]; + } else { + quot_diff = fuse->quot - prev_fuse->quot; + } + + freq_diff = fuse->max_freq - prev_fuse->max_freq; + freq_diff /= 1000000; /* Convert to MHz */ + scaling = 1000 * quot_diff / freq_diff; + scaling = min(scaling, fdata[fnum].max_quot_scale); + + apply_scaling = true; + } else if (freq_corner == fuse->max_freq) { + /* This is a fuse corner; don't scale anything */ + apply_scaling = false; + } + + if (apply_scaling) { + freq_diff = fuse->max_freq - freq_corner; + freq_diff_mhz = freq_diff / 1000000; + corner->quot_adjust = scaling * freq_diff_mhz / 1000; + + freq_high = fuse->max_freq; + freq_low = fuse->max_freq; + volt_high = fuse->uV; + + /* + if (freq_high > freq_low && volt_high > volt_low && + freq_high > freq_corner) + */ + + temp = freq_diff * (fuse->uV - prev_fuse->uV); + do_div(temp, freq_high - freq_low); + + /* + * max_volt_scale has units of uV/MHz while freq values + * have units of Hz. Divide by 1000000 to convert to. + */ + temp_limit = freq_diff * fdata[fnum].max_volt_scale; + do_div(temp_limit, 1000000); + + corner->uV = volt_high - min(temp, temp_limit); + corner->uV = roundup(corner->uV, step_volt); + } + + if (adj_quot) + corner->quot_adjust -= adj_quot[i]; + + if (adj_volt) + corner->uV += adj_volt[i]; + + /* Load per corner ceiling and floor voltages if they exist. */ + if (override) { + corner->max_uV = override->max_uV[i]; + corner->min_uV = override->min_uV[i]; + } else { + corner->max_uV = fuse->max_uV; + corner->min_uV = fuse->min_uV; + } + + if (drv->ceiling_max < corner->max_uV) + drv->ceiling_max = corner->max_uV; + + corner->uV = clamp(corner->uV, corner->min_uV, corner->max_uV); + corner->last_uV = corner->uV; + + /* Reduce the ceiling voltage if needed */ + if (desc->reduce_to_corner_uV && corner->uV < corner->max_uV) + corner->max_uV = corner->uV; + else if (desc->reduce_to_fuse_uV && fuse->uV < corner->max_uV) + corner->max_uV = max(corner->min_uV, fuse->uV); + + pr_info("corner %d: [%d %d %d] quot %d\n", i, + corner->min_uV, corner->uV, corner->max_uV, + fuse->quot - corner->quot_adjust); + } +} + +static const struct cpr_fuse * +cpr_get_fuses(const struct cpr_desc *desc, struct nvmem_device *qfprom) +{ + u32 expected = desc->cpr_fuses.redundant_value; + const struct qfprom_offset *fuse = &desc->cpr_fuses.redundant; + unsigned int idx; + + idx = !!(fuse->width && cpr_read_efuse(qfprom, fuse) == expected); + + return &desc->cpr_fuses.cpr_fuse[idx * desc->num_fuse_corners]; +} + +static bool cpr_is_close_loop_disabled(struct cpr_drv *drv, + const struct cpr_desc *desc, struct nvmem_device *qfprom, + const struct cpr_fuse *fuses, + const struct corner_adjustment *adj) +{ + const struct qfprom_offset *disable; + unsigned int idx; + struct fuse_corner *highest_fuse, *second_highest_fuse; + int min_diff_quot, diff_quot; + + if (adj && adj->disable_closed_loop) + return true; + + if (!desc->cpr_fuses.disable) + return false; + + /* + * Are the fuses the redundant ones? This avoids reading the fuse + * redundant bit again + */ + idx = !!(fuses == desc->cpr_fuses.cpr_fuse); + disable = &desc->cpr_fuses.disable[idx]; + + if (cpr_read_efuse(qfprom, disable)) + return true; + + if (!fuses->quotient_offset.width) { + /* + * Check if the target quotients for the highest two fuse + * corners are too close together. + */ + highest_fuse = &drv->fuse_corners[drv->num_fuse_corners - 1]; + second_highest_fuse = highest_fuse - 1; + + min_diff_quot = desc->min_diff_quot; + diff_quot = highest_fuse->quot - second_highest_fuse->quot; + + return diff_quot < min_diff_quot; + } + + return false; +} + +static int cpr_init_parameters(struct platform_device *pdev, + struct cpr_drv *drv) +{ + struct device_node *of_node = pdev->dev.of_node; + int ret; + + ret = of_property_read_u32(of_node, "qcom,cpr-ref-clk", + &drv->ref_clk_khz); + if (ret) + return ret; + ret = of_property_read_u32(of_node, "qcom,cpr-timer-delay-us", + &drv->timer_delay_us); + if (ret) + return ret; + ret = of_property_read_u32(of_node, "qcom,cpr-timer-cons-up", + &drv->timer_cons_up); + if (ret) + return ret; + ret = of_property_read_u32(of_node, "qcom,cpr-timer-cons-down", + &drv->timer_cons_down); + if (ret) + return ret; + drv->timer_cons_down &= RBIF_TIMER_ADJ_CONS_DOWN_MASK; + + ret = of_property_read_u32(of_node, "qcom,cpr-up-threshold", + &drv->up_threshold); + drv->up_threshold &= RBCPR_CTL_UP_THRESHOLD_MASK; + if (ret) + return ret; + + ret = of_property_read_u32(of_node, "qcom,cpr-down-threshold", + &drv->down_threshold); + drv->down_threshold &= RBCPR_CTL_DN_THRESHOLD_MASK; + if (ret) + return ret; + + ret = of_property_read_u32(of_node, "qcom,cpr-idle-clocks", + &drv->idle_clocks); + drv->idle_clocks &= RBCPR_STEP_QUOT_IDLE_CLK_MASK; + if (ret) + return ret; + + ret = of_property_read_u32(of_node, "qcom,cpr-gcnt-us", &drv->gcnt_us); + if (ret) + return ret; + ret = of_property_read_u32(of_node, "qcom,vdd-apc-step-up-limit", + &drv->vdd_apc_step_up_limit); + if (ret) + return ret; + ret = of_property_read_u32(of_node, "qcom,vdd-apc-step-down-limit", + &drv->vdd_apc_step_down_limit); + if (ret) + return ret; + + ret = of_property_read_u32(of_node, "qcom,cpr-clamp-timer-interval", + &drv->clamp_timer_interval); + if (ret && ret != -EINVAL) + return ret; + + drv->clamp_timer_interval = min_t(unsigned int, + drv->clamp_timer_interval, + RBIF_TIMER_ADJ_CLAMP_INT_MASK); + + pr_info("up threshold = %u, down threshold = %u\n", + drv->up_threshold, drv->down_threshold); + + return 0; +} + +static int cpr_init_and_enable_corner(struct cpr_drv *drv) +{ + unsigned long rate; + struct corner *end; + + end = &drv->corners[drv->num_corners - 1]; + rate = clk_get_rate(drv->cpu_clk); + + for (drv->corner = drv->corners; drv->corner <= end; drv->corner++) + if (drv->corner->freq == rate) + break; + + if (drv->corner > end) + return -EINVAL; + + return cpr_enable(drv); +} + +static struct corner_data msm8916_corner_data[] = { + /* [corner] -> { fuse corner, freq } */ + { 0, 200000000 }, + { 0, 400000000 }, + { 1, 533330000 }, + { 1, 800000000 }, + { 2, 998400000 }, + { 2, 1094400000 }, + { 2, 1152000000 }, + { 2, 1209600000 }, + { 2, 1363200000 }, +}; + +static const struct cpr_desc msm8916_desc = { + .num_fuse_corners = 3, + .vdd_mx_vmin_method = VDD_MX_VMIN_APC_CORNER_MAP, + .min_diff_quot = CPR_FUSE_MIN_QUOT_DIFF, + .step_quot = (int []){ 26, 26, 26, 26, 26, 26, 26, 26 }, + .cpr_fuses = { + .init_voltage_step = 10000, + .fuse_corner_data = (struct fuse_corner_data[]){ + /* ref_uV max_uV min_uV max_q q_off q_scl v_scl mx */ + { 1050000, 1050000, 1050000, 0, 0, 1, 0, 3 }, + { 1150000, 1150000, 1050000, 0, 0, 1, 0, 4 }, + { 1350000, 1350000, 1162500, 650, 0, 1, 0, 6 }, + }, + .cpr_fuse = (struct cpr_fuse[]){ + { + .ring_osc = { 222, 3, 6}, + .init_voltage = { 220, 6, 2 }, + .quotient = { 221, 12, 2 }, + }, + { + .ring_osc = { 222, 3, 6}, + .init_voltage = { 218, 6, 2 }, + .quotient = { 219, 12, 0 }, + }, + { + .ring_osc = { 222, 3, 6}, + .init_voltage = { 216, 6, 0 }, + .quotient = { 216, 12, 6 }, + }, + }, + .disable = &(struct qfprom_offset){ 223, 1, 1 }, + }, + .speed_bin = { 12, 3, 2 }, + .pvs_version = { 6, 2, 7 }, + .corner_data = msm8916_corner_data, + .num_corners = ARRAY_SIZE(msm8916_corner_data), + .num_freq_plans = 3, + .freq_plans = (struct freq_plan[]){ + { + .speed_bin = 0, + .pvs_version = 0, + .plan = (const struct corner_data* []){ + msm8916_corner_data + 0, + msm8916_corner_data + 1, + msm8916_corner_data + 2, + msm8916_corner_data + 3, + msm8916_corner_data + 4, + msm8916_corner_data + 5, + msm8916_corner_data + 6, + msm8916_corner_data + 7, + NULL + }, + }, + { + .speed_bin = 0, + .pvs_version = 1, + .plan = (const struct corner_data* []){ + msm8916_corner_data + 0, + msm8916_corner_data + 1, + msm8916_corner_data + 2, + msm8916_corner_data + 3, + msm8916_corner_data + 4, + msm8916_corner_data + 5, + msm8916_corner_data + 6, + msm8916_corner_data + 7, + NULL + }, + }, + { + .speed_bin = 2, + .pvs_version = 0, + .plan = (const struct corner_data* []){ + msm8916_corner_data + 0, + msm8916_corner_data + 1, + msm8916_corner_data + 2, + msm8916_corner_data + 3, + msm8916_corner_data + 4, + msm8916_corner_data + 5, + msm8916_corner_data + 6, + msm8916_corner_data + 7, + msm8916_corner_data + 8, + NULL + }, + }, + }, +}; + +static const struct acc_desc msm8916_acc_desc = { + .settings = (struct reg_default[]){ + { 0xf000, 0 }, + { 0xf000, 0x100 }, + { 0xf000, 0x101 } + }, + .override_settings = (struct reg_default[]){ + { 0xf000, 0 }, + { 0xf000, 0x100 }, + { 0xf000, 0x100 } + }, + .num_regs_per_fuse = 1, + .override = { 6, 1, 4 }, + .override_value = 1, +}; + +static const struct of_device_id cpr_descs[] = { + { .compatible = "qcom,qfprom-msm8916", .data = &msm8916_desc }, + { } +}; + +static const struct of_device_id acc_descs[] = { + { .compatible = "qcom,tcsr-msm8916", .data = &msm8916_acc_desc }, + { } +}; + +static int cpr_probe(struct platform_device *pdev) +{ + struct resource *res; + struct device *dev = &pdev->dev; + struct cpr_drv *drv; + const struct cpr_fuse *cpr_fuses; + const struct corner_adjustment *adj; + const struct corner_data **plan; + size_t len; + int irq, ret; + const struct cpr_desc *desc; + const struct acc_desc *acc_desc; + const struct of_device_id *match; + struct device_node *np; + struct nvmem_device *qfprom; + u32 cpr_rev = FUSE_REVISION_UNKNOWN; + u32 speed_bin = SPEED_BIN_NONE; + u32 pvs_version = 0; + + np = of_parse_phandle(dev->of_node, "nvmem", 0); + if (!np) + return -ENODEV; + + match = of_match_node(cpr_descs, np); + if (!match) + return -EINVAL; + desc = match->data; + + qfprom = nvmem_device_get(dev, "qfprom"); + if (IS_ERR(qfprom)) + return PTR_ERR(qfprom); + + len = sizeof(*drv) + + sizeof(*drv->fuse_corners) * desc->num_fuse_corners + + sizeof(*drv->corners) * desc->num_corners; + + drv = devm_kzalloc(dev, len, GFP_KERNEL); + if (!drv) + return -ENOMEM; + + np = of_parse_phandle(dev->of_node, "acc-syscon", 0); + if (!np) + return -ENODEV; + + match = of_match_node(acc_descs, np); + if (!match) + return -EINVAL; + + acc_desc = match->data; + drv->tcsr = syscon_node_to_regmap(np); + if (IS_ERR(drv->tcsr)) + return PTR_ERR(drv->tcsr); + + drv->num_fuse_corners = desc->num_fuse_corners; + drv->num_corners = desc->num_corners; + drv->fuse_corners = (struct fuse_corner *)(drv + 1); + drv->corners = (struct corner *)(drv->fuse_corners + + drv->num_fuse_corners); + mutex_init(&drv->lock); + + res = platform_get_resource(pdev, IORESOURCE_MEM, 0); + drv->base = devm_ioremap_resource(&pdev->dev, res); + if (IS_ERR(drv->base)) + return PTR_ERR(drv->base); + + irq = platform_get_irq(pdev, 0); + if (irq < 0) + return -EINVAL; + + drv->vdd_mx = devm_regulator_get(dev, "vdd-mx"); + if (IS_ERR(drv->vdd_mx)) + return PTR_ERR(drv->vdd_mx); + + drv->vdd_mx_vmin_method = desc->vdd_mx_vmin_method; + drv->vdd_mx_vmax = desc->vdd_mx_vmax; + + if (desc->fuse_revision.width) + cpr_rev = cpr_read_efuse(qfprom, &desc->fuse_revision); + if (desc->speed_bin.width) + speed_bin = cpr_read_efuse(qfprom, &desc->speed_bin); + if (desc->pvs_version.width) + pvs_version = cpr_read_efuse(qfprom, &desc->pvs_version); + + plan = find_freq_plan(desc, speed_bin, pvs_version); + if (!plan) + return -EINVAL; + + cpr_fuses = cpr_get_fuses(desc, qfprom); + cpr_populate_ring_osc_idx(cpr_fuses, drv, qfprom); + + adj = cpr_find_adjustment(speed_bin, pvs_version, cpr_rev, desc, drv); + + cpr_fuse_corner_init(drv, desc, qfprom, cpr_fuses, speed_bin, adj, + acc_desc); + cpr_corner_init(drv, desc, cpr_fuses, speed_bin, pvs_version, qfprom, + adj, plan); + + ret = cpr_populate_opps(dev->of_node, drv, plan); + if (ret) + return ret; + + drv->loop_disabled = cpr_is_close_loop_disabled(drv, desc, qfprom, + cpr_fuses, adj); + pr_info("CPR closed loop is %sabled\n", + drv->loop_disabled ? "dis" : "en"); + + ret = cpr_init_parameters(pdev, drv); + if (ret) + return ret; + + /* Configure CPR HW but keep it disabled */ + ret = cpr_config(drv); + if (ret) + return ret; + + /* Enable ACC if required */ + if (acc_desc->enable_mask) + regmap_update_bits(drv->tcsr, acc_desc->enable_reg, + acc_desc->enable_mask, + acc_desc->enable_mask); + + ret = devm_request_threaded_irq(&pdev->dev, irq, NULL, + cpr_irq_handler, IRQF_ONESHOT | IRQF_TRIGGER_RISING, + "cpr", drv); + if (ret) + return ret; + + ret = cpr_init_and_enable_corner(drv); + if (ret) + return ret; + + drv->reg_nb.notifier_call = cpr_regulator_notifier; + ret = regulator_register_notifier(drv->vdd_apc, &drv->reg_nb); + if (ret) + return ret; + + drv->cpufreq_nb.notifier_call = cpr_cpufreq_notifier; + ret = cpufreq_register_notifier(&drv->cpufreq_nb, + CPUFREQ_TRANSITION_NOTIFIER); + if (ret) { + regulator_unregister_notifier(drv->vdd_apc, &drv->reg_nb); + return ret; + } + + /* + * Ensure that enable state accurately reflects the case in which CPR + * is permanently disabled. + */ + //cpr_vreg->enable &= !cpr_vreg->loop_disabled; + + platform_set_drvdata(pdev, drv); + + return 0; +} + +static int cpr_remove(struct platform_device *pdev) +{ + struct cpr_drv *drv = platform_get_drvdata(pdev); + + if (cpr_is_allowed(drv)) { + cpr_ctl_disable(drv); + cpr_irq_set(drv, 0); + } + + return 0; +} + +static const struct of_device_id cpr_match_table[] = { + { .compatible = "qcom,cpr" }, + { } +}; +MODULE_DEVICE_TABLE(of, cpr_match_table); + +static struct platform_driver cpr_driver = { + .probe = cpr_probe, + .remove = cpr_remove, + .driver = { + .name = "qcom-cpr", + .of_match_table = cpr_match_table, + .pm = &cpr_pm_ops, + }, +}; +module_platform_driver(cpr_driver); + +MODULE_DESCRIPTION("Core Power Reduction (CPR) driver"); +MODULE_LICENSE("GPL v2"); +MODULE_ALIAS("platform:qcom-cpr");