diff options
Diffstat (limited to 'drivers')
115 files changed, 8229 insertions, 2234 deletions
diff --git a/drivers/block/loop.c b/drivers/block/loop.c index d1f168b73634..c4fd1e45ce1e 100644 --- a/drivers/block/loop.c +++ b/drivers/block/loop.c @@ -75,6 +75,7 @@ #include <linux/sysfs.h> #include <linux/miscdevice.h> #include <linux/falloc.h> +#include <linux/uio.h> #include "loop.h" #include <asm/uaccess.h> @@ -229,13 +230,14 @@ lo_do_transfer(struct loop_device *lo, int cmd, static int __do_lo_send_write(struct file *file, u8 *buf, const int len, loff_t pos) { + struct kvec kvec = {.iov_base = buf, .iov_len = len}; + struct iov_iter from; ssize_t bw; - mm_segment_t old_fs = get_fs(); + + iov_iter_kvec(&from, ITER_KVEC | WRITE, &kvec, 1, len); file_start_write(file); - set_fs(get_ds()); - bw = file->f_op->write(file, buf, len, &pos); - set_fs(old_fs); + bw = vfs_iter_write(file, &from, &pos); file_end_write(file); if (likely(bw == len)) return 0; @@ -767,7 +769,7 @@ static int loop_set_fd(struct loop_device *lo, fmode_t mode, goto out_putf; if (!(file->f_mode & FMODE_WRITE) || !(mode & FMODE_WRITE) || - !file->f_op->write) + !file->f_op->write_iter) lo_flags |= LO_FLAGS_READ_ONLY; lo_blocksize = S_ISBLK(inode->i_mode) ? diff --git a/drivers/char/hw_random/Kconfig b/drivers/char/hw_random/Kconfig index de57b38809c7..f48cf11c655e 100644 --- a/drivers/char/hw_random/Kconfig +++ b/drivers/char/hw_random/Kconfig @@ -101,6 +101,19 @@ config HW_RANDOM_BCM2835 If unsure, say Y. +config HW_RANDOM_IPROC_RNG200 + tristate "Broadcom iProc RNG200 support" + depends on ARCH_BCM_IPROC + default HW_RANDOM + ---help--- + This driver provides kernel-side support for the RNG200 + hardware found on the Broadcom iProc SoCs. + + To compile this driver as a module, choose M here: the + module will be called iproc-rng200 + + If unsure, say Y. + config HW_RANDOM_GEODE tristate "AMD Geode HW Random Number Generator support" depends on X86_32 && PCI diff --git a/drivers/char/hw_random/Makefile b/drivers/char/hw_random/Makefile index 0b4cd57f4e24..055bb01510ad 100644 --- a/drivers/char/hw_random/Makefile +++ b/drivers/char/hw_random/Makefile @@ -28,5 +28,6 @@ obj-$(CONFIG_HW_RANDOM_POWERNV) += powernv-rng.o obj-$(CONFIG_HW_RANDOM_EXYNOS) += exynos-rng.o obj-$(CONFIG_HW_RANDOM_TPM) += tpm-rng.o obj-$(CONFIG_HW_RANDOM_BCM2835) += bcm2835-rng.o +obj-$(CONFIG_HW_RANDOM_IPROC_RNG200) += iproc-rng200.o obj-$(CONFIG_HW_RANDOM_MSM) += msm-rng.o obj-$(CONFIG_HW_RANDOM_XGENE) += xgene-rng.o diff --git a/drivers/char/hw_random/bcm63xx-rng.c b/drivers/char/hw_random/bcm63xx-rng.c index ba6a65ac023b..d1494ecd9e11 100644 --- a/drivers/char/hw_random/bcm63xx-rng.c +++ b/drivers/char/hw_random/bcm63xx-rng.c @@ -13,24 +13,37 @@ #include <linux/platform_device.h> #include <linux/hw_random.h> -#include <bcm63xx_io.h> -#include <bcm63xx_regs.h> +#define RNG_CTRL 0x00 +#define RNG_EN (1 << 0) + +#define RNG_STAT 0x04 +#define RNG_AVAIL_MASK (0xff000000) + +#define RNG_DATA 0x08 +#define RNG_THRES 0x0c +#define RNG_MASK 0x10 struct bcm63xx_rng_priv { + struct hwrng rng; struct clk *clk; void __iomem *regs; }; -#define to_rng_priv(rng) ((struct bcm63xx_rng_priv *)rng->priv) +#define to_rng_priv(rng) container_of(rng, struct bcm63xx_rng_priv, rng) static int bcm63xx_rng_init(struct hwrng *rng) { struct bcm63xx_rng_priv *priv = to_rng_priv(rng); u32 val; + int error; + + error = clk_prepare_enable(priv->clk); + if (error) + return error; - val = bcm_readl(priv->regs + RNG_CTRL); + val = __raw_readl(priv->regs + RNG_CTRL); val |= RNG_EN; - bcm_writel(val, priv->regs + RNG_CTRL); + __raw_writel(val, priv->regs + RNG_CTRL); return 0; } @@ -40,23 +53,25 @@ static void bcm63xx_rng_cleanup(struct hwrng *rng) struct bcm63xx_rng_priv *priv = to_rng_priv(rng); u32 val; - val = bcm_readl(priv->regs + RNG_CTRL); + val = __raw_readl(priv->regs + RNG_CTRL); val &= ~RNG_EN; - bcm_writel(val, priv->regs + RNG_CTRL); + __raw_writel(val, priv->regs + RNG_CTRL); + + clk_didsable_unprepare(prov->clk); } static int bcm63xx_rng_data_present(struct hwrng *rng, int wait) { struct bcm63xx_rng_priv *priv = to_rng_priv(rng); - return bcm_readl(priv->regs + RNG_STAT) & RNG_AVAIL_MASK; + return __raw_readl(priv->regs + RNG_STAT) & RNG_AVAIL_MASK; } static int bcm63xx_rng_data_read(struct hwrng *rng, u32 *data) { struct bcm63xx_rng_priv *priv = to_rng_priv(rng); - *data = bcm_readl(priv->regs + RNG_DATA); + *data = __raw_readl(priv->regs + RNG_DATA); return 4; } @@ -72,94 +87,53 @@ static int bcm63xx_rng_probe(struct platform_device *pdev) r = platform_get_resource(pdev, IORESOURCE_MEM, 0); if (!r) { dev_err(&pdev->dev, "no iomem resource\n"); - ret = -ENXIO; - goto out; + return -ENXIO; } - priv = kzalloc(sizeof(*priv), GFP_KERNEL); - if (!priv) { - dev_err(&pdev->dev, "no memory for private structure\n"); - ret = -ENOMEM; - goto out; + priv = devm_kzalloc(&pdev->dev, sizeof(*priv), GFP_KERNEL); + if (!priv) + return -ENOMEM; + + priv->rng.name = pdev->name; + priv->rng.init = bcm63xx_rng_init; + priv->rng.cleanup = bcm63xx_rng_cleanup; + prov->rng.data_present = bcm63xx_rng_data_present; + priv->rng.data_read = bcm63xx_rng_data_read; + + priv->clk = devm_clk_get(&pdev->dev, "ipsec"); + if (IS_ERR(priv->clk)) { + error = PTR_ERR(priv->clk); + dev_err(&pdev->dev, "no clock for device: %d\n", error); + return error; } - rng = kzalloc(sizeof(*rng), GFP_KERNEL); - if (!rng) { - dev_err(&pdev->dev, "no memory for rng structure\n"); - ret = -ENOMEM; - goto out_free_priv; - } - - platform_set_drvdata(pdev, rng); - rng->priv = (unsigned long)priv; - rng->name = pdev->name; - rng->init = bcm63xx_rng_init; - rng->cleanup = bcm63xx_rng_cleanup; - rng->data_present = bcm63xx_rng_data_present; - rng->data_read = bcm63xx_rng_data_read; - - clk = clk_get(&pdev->dev, "ipsec"); - if (IS_ERR(clk)) { - dev_err(&pdev->dev, "no clock for device\n"); - ret = PTR_ERR(clk); - goto out_free_rng; - } - - priv->clk = clk; - if (!devm_request_mem_region(&pdev->dev, r->start, resource_size(r), pdev->name)) { dev_err(&pdev->dev, "request mem failed"); - ret = -ENOMEM; - goto out_free_rng; + return -EBUSY; } priv->regs = devm_ioremap_nocache(&pdev->dev, r->start, resource_size(r)); if (!priv->regs) { dev_err(&pdev->dev, "ioremap failed"); - ret = -ENOMEM; - goto out_free_rng; + return -ENOMEM; } - clk_enable(clk); - - ret = hwrng_register(rng); - if (ret) { - dev_err(&pdev->dev, "failed to register rng device\n"); - goto out_clk_disable; + error = devm_hwrng_register(&pdev->dev, &priv->rng); + if (error) { + dev_err(&pdev->dev, "failed to register rng device: %d\n", + error); + return error; } dev_info(&pdev->dev, "registered RNG driver\n"); return 0; - -out_clk_disable: - clk_disable(clk); -out_free_rng: - kfree(rng); -out_free_priv: - kfree(priv); -out: - return ret; -} - -static int bcm63xx_rng_remove(struct platform_device *pdev) -{ - struct hwrng *rng = platform_get_drvdata(pdev); - struct bcm63xx_rng_priv *priv = to_rng_priv(rng); - - hwrng_unregister(rng); - clk_disable(priv->clk); - kfree(priv); - kfree(rng); - - return 0; } static struct platform_driver bcm63xx_rng_driver = { .probe = bcm63xx_rng_probe, - .remove = bcm63xx_rng_remove, .driver = { .name = "bcm63xx-rng", }, diff --git a/drivers/char/hw_random/core.c b/drivers/char/hw_random/core.c index 32a8a867f7f8..571ef61f8ea9 100644 --- a/drivers/char/hw_random/core.c +++ b/drivers/char/hw_random/core.c @@ -179,7 +179,8 @@ skip_init: add_early_randomness(rng); current_quality = rng->quality ? : default_quality; - current_quality &= 1023; + if (current_quality > 1024) + current_quality = 1024; if (current_quality == 0 && hwrng_fill) kthread_stop(hwrng_fill); @@ -536,6 +537,48 @@ void hwrng_unregister(struct hwrng *rng) } EXPORT_SYMBOL_GPL(hwrng_unregister); +static void devm_hwrng_release(struct device *dev, void *res) +{ + hwrng_unregister(*(struct hwrng **)res); +} + +static int devm_hwrng_match(struct device *dev, void *res, void *data) +{ + struct hwrng **r = res; + + if (WARN_ON(!r || !*r)) + return 0; + + return *r == data; +} + +int devm_hwrng_register(struct device *dev, struct hwrng *rng) +{ + struct hwrng **ptr; + int error; + + ptr = devres_alloc(devm_hwrng_release, sizeof(*ptr), GFP_KERNEL); + if (!ptr) + return -ENOMEM; + + error = hwrng_register(rng); + if (error) { + devres_free(ptr); + return error; + } + + *ptr = rng; + devres_add(dev, ptr); + return 0; +} +EXPORT_SYMBOL_GPL(devm_hwrng_register); + +void devm_hwrng_unregister(struct device *dev, struct hwrng *rng) +{ + devres_release(dev, devm_hwrng_release, devm_hwrng_match, rng); +} +EXPORT_SYMBOL_GPL(devm_hwrng_unregister); + static int __init hwrng_modinit(void) { return register_miscdev(); diff --git a/drivers/char/hw_random/exynos-rng.c b/drivers/char/hw_random/exynos-rng.c index fed0830bf724..dc4701fd814f 100644 --- a/drivers/char/hw_random/exynos-rng.c +++ b/drivers/char/hw_random/exynos-rng.c @@ -131,16 +131,7 @@ static int exynos_rng_probe(struct platform_device *pdev) pm_runtime_use_autosuspend(&pdev->dev); pm_runtime_enable(&pdev->dev); - return hwrng_register(&exynos_rng->rng); -} - -static int exynos_rng_remove(struct platform_device *pdev) -{ - struct exynos_rng *exynos_rng = platform_get_drvdata(pdev); - - hwrng_unregister(&exynos_rng->rng); - - return 0; + return devm_hwrng_register(&pdev->dev, &exynos_rng->rng); } #ifdef CONFIG_PM @@ -172,7 +163,6 @@ static struct platform_driver exynos_rng_driver = { .pm = &exynos_rng_pm_ops, }, .probe = exynos_rng_probe, - .remove = exynos_rng_remove, }; module_platform_driver(exynos_rng_driver); diff --git a/drivers/char/hw_random/iproc-rng200.c b/drivers/char/hw_random/iproc-rng200.c new file mode 100644 index 000000000000..3eaf7cb96d36 --- /dev/null +++ b/drivers/char/hw_random/iproc-rng200.c @@ -0,0 +1,239 @@ +/* +* Copyright (C) 2015 Broadcom Corporation +* +* This program is free software; you can redistribute it and/or +* modify it under the terms of the GNU General Public License as +* published by the Free Software Foundation version 2. +* +* This program is distributed "as is" WITHOUT ANY WARRANTY of any +* kind, whether express or implied; without even the implied warranty +* of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +* GNU General Public License for more details. +*/ +/* + * DESCRIPTION: The Broadcom iProc RNG200 Driver + */ + +#include <linux/hw_random.h> +#include <linux/init.h> +#include <linux/io.h> +#include <linux/kernel.h> +#include <linux/module.h> +#include <linux/of_address.h> +#include <linux/of_platform.h> +#include <linux/platform_device.h> +#include <linux/delay.h> + +/* Registers */ +#define RNG_CTRL_OFFSET 0x00 +#define RNG_CTRL_RNG_RBGEN_MASK 0x00001FFF +#define RNG_CTRL_RNG_RBGEN_ENABLE 0x00000001 +#define RNG_CTRL_RNG_RBGEN_DISABLE 0x00000000 + +#define RNG_SOFT_RESET_OFFSET 0x04 +#define RNG_SOFT_RESET 0x00000001 + +#define RBG_SOFT_RESET_OFFSET 0x08 +#define RBG_SOFT_RESET 0x00000001 + +#define RNG_INT_STATUS_OFFSET 0x18 +#define RNG_INT_STATUS_MASTER_FAIL_LOCKOUT_IRQ_MASK 0x80000000 +#define RNG_INT_STATUS_STARTUP_TRANSITIONS_MET_IRQ_MASK 0x00020000 +#define RNG_INT_STATUS_NIST_FAIL_IRQ_MASK 0x00000020 +#define RNG_INT_STATUS_TOTAL_BITS_COUNT_IRQ_MASK 0x00000001 + +#define RNG_FIFO_DATA_OFFSET 0x20 + +#define RNG_FIFO_COUNT_OFFSET 0x24 +#define RNG_FIFO_COUNT_RNG_FIFO_COUNT_MASK 0x000000FF + +struct iproc_rng200_dev { + struct hwrng rng; + void __iomem *base; +}; + +#define to_rng_priv(rng) container_of(rng, struct iproc_rng200_dev, rng) + +static void iproc_rng200_restart(void __iomem *rng_base) +{ + uint32_t val; + + /* Disable RBG */ + val = ioread32(rng_base + RNG_CTRL_OFFSET); + val &= ~RNG_CTRL_RNG_RBGEN_MASK; + val |= RNG_CTRL_RNG_RBGEN_DISABLE; + iowrite32(val, rng_base + RNG_CTRL_OFFSET); + + /* Clear all interrupt status */ + iowrite32(0xFFFFFFFFUL, rng_base + RNG_INT_STATUS_OFFSET); + + /* Reset RNG and RBG */ + val = ioread32(rng_base + RBG_SOFT_RESET_OFFSET); + val |= RBG_SOFT_RESET; + iowrite32(val, rng_base + RBG_SOFT_RESET_OFFSET); + + val = ioread32(rng_base + RNG_SOFT_RESET_OFFSET); + val |= RNG_SOFT_RESET; + iowrite32(val, rng_base + RNG_SOFT_RESET_OFFSET); + + val = ioread32(rng_base + RNG_SOFT_RESET_OFFSET); + val &= ~RNG_SOFT_RESET; + iowrite32(val, rng_base + RNG_SOFT_RESET_OFFSET); + + val = ioread32(rng_base + RBG_SOFT_RESET_OFFSET); + val &= ~RBG_SOFT_RESET; + iowrite32(val, rng_base + RBG_SOFT_RESET_OFFSET); + + /* Enable RBG */ + val = ioread32(rng_base + RNG_CTRL_OFFSET); + val &= ~RNG_CTRL_RNG_RBGEN_MASK; + val |= RNG_CTRL_RNG_RBGEN_ENABLE; + iowrite32(val, rng_base + RNG_CTRL_OFFSET); +} + +static int iproc_rng200_read(struct hwrng *rng, void *buf, size_t max, + bool wait) +{ + struct iproc_rng200_dev *priv = to_rng_priv(rng); + uint32_t num_remaining = max; + uint32_t status; + + #define MAX_RESETS_PER_READ 1 + uint32_t num_resets = 0; + + #define MAX_IDLE_TIME (1 * HZ) + unsigned long idle_endtime = jiffies + MAX_IDLE_TIME; + + while ((num_remaining > 0) && time_before(jiffies, idle_endtime)) { + + /* Is RNG sane? If not, reset it. */ + status = ioread32(priv->base + RNG_INT_STATUS_OFFSET); + if ((status & (RNG_INT_STATUS_MASTER_FAIL_LOCKOUT_IRQ_MASK | + RNG_INT_STATUS_NIST_FAIL_IRQ_MASK)) != 0) { + + if (num_resets >= MAX_RESETS_PER_READ) + return max - num_remaining; + + iproc_rng200_restart(priv->base); + num_resets++; + } + + /* Are there any random numbers available? */ + if ((ioread32(priv->base + RNG_FIFO_COUNT_OFFSET) & + RNG_FIFO_COUNT_RNG_FIFO_COUNT_MASK) > 0) { + + if (num_remaining >= sizeof(uint32_t)) { + /* Buffer has room to store entire word */ + *(uint32_t *)buf = ioread32(priv->base + + RNG_FIFO_DATA_OFFSET); + buf += sizeof(uint32_t); + num_remaining -= sizeof(uint32_t); + } else { + /* Buffer can only store partial word */ + uint32_t rnd_number = ioread32(priv->base + + RNG_FIFO_DATA_OFFSET); + memcpy(buf, &rnd_number, num_remaining); + buf += num_remaining; + num_remaining = 0; + } + + /* Reset the IDLE timeout */ + idle_endtime = jiffies + MAX_IDLE_TIME; + } else { + if (!wait) + /* Cannot wait, return immediately */ + return max - num_remaining; + + /* Can wait, give others chance to run */ + usleep_range(min(num_remaining * 10, 500U), 500); + } + } + + return max - num_remaining; +} + +static int iproc_rng200_init(struct hwrng *rng) +{ + struct iproc_rng200_dev *priv = to_rng_priv(rng); + uint32_t val; + + /* Setup RNG. */ + val = ioread32(priv->base + RNG_CTRL_OFFSET); + val &= ~RNG_CTRL_RNG_RBGEN_MASK; + val |= RNG_CTRL_RNG_RBGEN_ENABLE; + iowrite32(val, priv->base + RNG_CTRL_OFFSET); + + return 0; +} + +static void iproc_rng200_cleanup(struct hwrng *rng) +{ + struct iproc_rng200_dev *priv = to_rng_priv(rng); + uint32_t val; + + /* Disable RNG hardware */ + val = ioread32(priv->base + RNG_CTRL_OFFSET); + val &= ~RNG_CTRL_RNG_RBGEN_MASK; + val |= RNG_CTRL_RNG_RBGEN_DISABLE; + iowrite32(val, priv->base + RNG_CTRL_OFFSET); +} + +static int iproc_rng200_probe(struct platform_device *pdev) +{ + struct iproc_rng200_dev *priv; + struct resource *res; + struct device *dev = &pdev->dev; + int ret; + + priv = devm_kzalloc(dev, sizeof(*priv), GFP_KERNEL); + if (!priv) + return -ENOMEM; + + /* Map peripheral */ + res = platform_get_resource(pdev, IORESOURCE_MEM, 0); + if (!res) { + dev_err(dev, "failed to get rng resources\n"); + return -EINVAL; + } + + priv->base = devm_ioremap_resource(dev, res); + if (IS_ERR(priv->base)) { + dev_err(dev, "failed to remap rng regs\n"); + return PTR_ERR(priv->base); + } + + priv->rng.name = "iproc-rng200", + priv->rng.read = iproc_rng200_read, + priv->rng.init = iproc_rng200_init, + priv->rng.cleanup = iproc_rng200_cleanup, + + /* Register driver */ + ret = devm_hwrng_register(dev, &priv->rng); + if (ret) { + dev_err(dev, "hwrng registration failed\n"); + return ret; + } + + dev_info(dev, "hwrng registered\n"); + + return 0; +} + +static const struct of_device_id iproc_rng200_of_match[] = { + { .compatible = "brcm,iproc-rng200", }, + {}, +}; +MODULE_DEVICE_TABLE(of, iproc_rng200_of_match); + +static struct platform_driver iproc_rng200_driver = { + .driver = { + .name = "iproc-rng200", + .of_match_table = iproc_rng200_of_match, + }, + .probe = iproc_rng200_probe, +}; +module_platform_driver(iproc_rng200_driver); + +MODULE_AUTHOR("Broadcom"); +MODULE_DESCRIPTION("iProc RNG200 Random Number Generator driver"); +MODULE_LICENSE("GPL v2"); diff --git a/drivers/char/hw_random/msm-rng.c b/drivers/char/hw_random/msm-rng.c index cea1c703d62f..96fb986402eb 100644 --- a/drivers/char/hw_random/msm-rng.c +++ b/drivers/char/hw_random/msm-rng.c @@ -157,7 +157,7 @@ static int msm_rng_probe(struct platform_device *pdev) rng->hwrng.cleanup = msm_rng_cleanup, rng->hwrng.read = msm_rng_read, - ret = hwrng_register(&rng->hwrng); + ret = devm_hwrng_register(&pdev->dev, &rng->hwrng); if (ret) { dev_err(&pdev->dev, "failed to register hwrng\n"); return ret; @@ -166,14 +166,6 @@ static int msm_rng_probe(struct platform_device *pdev) return 0; } -static int msm_rng_remove(struct platform_device *pdev) -{ - struct msm_rng *rng = platform_get_drvdata(pdev); - - hwrng_unregister(&rng->hwrng); - return 0; -} - static const struct of_device_id msm_rng_of_match[] = { { .compatible = "qcom,prng", }, {} @@ -182,7 +174,6 @@ MODULE_DEVICE_TABLE(of, msm_rng_of_match); static struct platform_driver msm_rng_driver = { .probe = msm_rng_probe, - .remove = msm_rng_remove, .driver = { .name = KBUILD_MODNAME, .of_match_table = of_match_ptr(msm_rng_of_match), diff --git a/drivers/char/hw_random/octeon-rng.c b/drivers/char/hw_random/octeon-rng.c index be1c3f607398..6234a4a19b56 100644 --- a/drivers/char/hw_random/octeon-rng.c +++ b/drivers/char/hw_random/octeon-rng.c @@ -105,7 +105,7 @@ static int octeon_rng_probe(struct platform_device *pdev) return 0; } -static int __exit octeon_rng_remove(struct platform_device *pdev) +static int octeon_rng_remove(struct platform_device *pdev) { struct hwrng *rng = platform_get_drvdata(pdev); @@ -119,7 +119,7 @@ static struct platform_driver octeon_rng_driver = { .name = "octeon_rng", }, .probe = octeon_rng_probe, - .remove = __exit_p(octeon_rng_remove), + .remove = octeon_rng_remove, }; module_platform_driver(octeon_rng_driver); diff --git a/drivers/char/hw_random/omap-rng.c b/drivers/char/hw_random/omap-rng.c index d14dcf788f17..8a1432e8bb80 100644 --- a/drivers/char/hw_random/omap-rng.c +++ b/drivers/char/hw_random/omap-rng.c @@ -236,7 +236,7 @@ static int omap4_rng_init(struct omap_rng_dev *priv) u32 val; /* Return if RNG is already running. */ - if (omap_rng_read(priv, RNG_CONFIG_REG) & RNG_CONTROL_ENABLE_TRNG_MASK) + if (omap_rng_read(priv, RNG_CONTROL_REG) & RNG_CONTROL_ENABLE_TRNG_MASK) return 0; val = RNG_CONFIG_MIN_REFIL_CYCLES << RNG_CONFIG_MIN_REFIL_CYCLES_SHIFT; @@ -262,7 +262,7 @@ static void omap4_rng_cleanup(struct omap_rng_dev *priv) val = omap_rng_read(priv, RNG_CONTROL_REG); val &= ~RNG_CONTROL_ENABLE_TRNG_MASK; - omap_rng_write(priv, RNG_CONFIG_REG, val); + omap_rng_write(priv, RNG_CONTROL_REG, val); } static irqreturn_t omap4_rng_irq(int irq, void *dev_id) @@ -408,7 +408,7 @@ err_ioremap: return ret; } -static int __exit omap_rng_remove(struct platform_device *pdev) +static int omap_rng_remove(struct platform_device *pdev) { struct omap_rng_dev *priv = platform_get_drvdata(pdev); @@ -422,9 +422,7 @@ static int __exit omap_rng_remove(struct platform_device *pdev) return 0; } -#ifdef CONFIG_PM_SLEEP - -static int omap_rng_suspend(struct device *dev) +static int __maybe_unused omap_rng_suspend(struct device *dev) { struct omap_rng_dev *priv = dev_get_drvdata(dev); @@ -434,7 +432,7 @@ static int omap_rng_suspend(struct device *dev) return 0; } -static int omap_rng_resume(struct device *dev) +static int __maybe_unused omap_rng_resume(struct device *dev) { struct omap_rng_dev *priv = dev_get_drvdata(dev); @@ -445,22 +443,15 @@ static int omap_rng_resume(struct device *dev) } static SIMPLE_DEV_PM_OPS(omap_rng_pm, omap_rng_suspend, omap_rng_resume); -#define OMAP_RNG_PM (&omap_rng_pm) - -#else - -#define OMAP_RNG_PM NULL - -#endif static struct platform_driver omap_rng_driver = { .driver = { .name = "omap_rng", - .pm = OMAP_RNG_PM, + .pm = &omap_rng_pm, .of_match_table = of_match_ptr(omap_rng_of_match), }, .probe = omap_rng_probe, - .remove = __exit_p(omap_rng_remove), + .remove = omap_rng_remove, }; module_platform_driver(omap_rng_driver); diff --git a/drivers/char/hw_random/pseries-rng.c b/drivers/char/hw_random/pseries-rng.c index bcf86f91800a..63ce51d09af1 100644 --- a/drivers/char/hw_random/pseries-rng.c +++ b/drivers/char/hw_random/pseries-rng.c @@ -61,13 +61,13 @@ static struct hwrng pseries_rng = { .read = pseries_rng_read, }; -static int __init pseries_rng_probe(struct vio_dev *dev, +static int pseries_rng_probe(struct vio_dev *dev, const struct vio_device_id *id) { return hwrng_register(&pseries_rng); } -static int __exit pseries_rng_remove(struct vio_dev *dev) +static int pseries_rng_remove(struct vio_dev *dev) { hwrng_unregister(&pseries_rng); return 0; diff --git a/drivers/char/hw_random/xgene-rng.c b/drivers/char/hw_random/xgene-rng.c index 23caa05380a8..c37cf754a985 100644 --- a/drivers/char/hw_random/xgene-rng.c +++ b/drivers/char/hw_random/xgene-rng.c @@ -21,6 +21,7 @@ * */ +#include <linux/acpi.h> #include <linux/clk.h> #include <linux/delay.h> #include <linux/hw_random.h> @@ -310,6 +311,14 @@ static int xgene_rng_init(struct hwrng *rng) return 0; } +#ifdef CONFIG_ACPI +static const struct acpi_device_id xgene_rng_acpi_match[] = { + { "APMC0D18", }, + { } +}; +MODULE_DEVICE_TABLE(acpi, xgene_rng_acpi_match); +#endif + static struct hwrng xgene_rng_func = { .name = "xgene-rng", .init = xgene_rng_init, @@ -415,6 +424,7 @@ static struct platform_driver xgene_rng_driver = { .driver = { .name = "xgene-rng", .of_match_table = xgene_rng_of_match, + .acpi_match_table = ACPI_PTR(xgene_rng_acpi_match), }, }; diff --git a/drivers/char/mem.c b/drivers/char/mem.c index 9c4fd7a8e2e5..6b1721f978c2 100644 --- a/drivers/char/mem.c +++ b/drivers/char/mem.c @@ -607,16 +607,16 @@ static ssize_t write_null(struct file *file, const char __user *buf, return count; } -static ssize_t aio_read_null(struct kiocb *iocb, const struct iovec *iov, - unsigned long nr_segs, loff_t pos) +static ssize_t read_iter_null(struct kiocb *iocb, struct iov_iter *to) { return 0; } -static ssize_t aio_write_null(struct kiocb *iocb, const struct iovec *iov, - unsigned long nr_segs, loff_t pos) +static ssize_t write_iter_null(struct kiocb *iocb, struct iov_iter *from) { - return iov_length(iov, nr_segs); + size_t count = iov_iter_count(from); + iov_iter_advance(from, count); + return count; } static int pipe_to_null(struct pipe_inode_info *info, struct pipe_buffer *buf, @@ -718,7 +718,7 @@ static int open_port(struct inode *inode, struct file *filp) #define zero_lseek null_lseek #define full_lseek null_lseek #define write_zero write_null -#define aio_write_zero aio_write_null +#define write_iter_zero write_iter_null #define open_mem open_port #define open_kmem open_mem @@ -750,8 +750,8 @@ static const struct file_operations null_fops = { .llseek = null_lseek, .read = read_null, .write = write_null, - .aio_read = aio_read_null, - .aio_write = aio_write_null, + .read_iter = read_iter_null, + .write_iter = write_iter_null, .splice_write = splice_write_null, }; @@ -764,10 +764,9 @@ static const struct file_operations __maybe_unused port_fops = { static const struct file_operations zero_fops = { .llseek = zero_lseek, - .read = new_sync_read, .write = write_zero, .read_iter = read_iter_zero, - .aio_write = aio_write_zero, + .write_iter = write_iter_zero, .mmap = mmap_zero, #ifndef CONFIG_MMU .mmap_capabilities = zero_mmap_capabilities, @@ -776,7 +775,6 @@ static const struct file_operations zero_fops = { static const struct file_operations full_fops = { .llseek = full_lseek, - .read = new_sync_read, .read_iter = read_iter_zero, .write = write_full, }; diff --git a/drivers/char/raw.c b/drivers/char/raw.c index 6e29bf2db536..5fc291c6157e 100644 --- a/drivers/char/raw.c +++ b/drivers/char/raw.c @@ -282,9 +282,7 @@ static long raw_ctl_compat_ioctl(struct file *file, unsigned int cmd, #endif static const struct file_operations raw_fops = { - .read = new_sync_read, .read_iter = blkdev_read_iter, - .write = new_sync_write, .write_iter = blkdev_write_iter, .fsync = blkdev_fsync, .open = raw_open, diff --git a/drivers/char/tpm/Kconfig b/drivers/char/tpm/Kconfig index 9d4e37549eb2..3b84a8b1bfbe 100644 --- a/drivers/char/tpm/Kconfig +++ b/drivers/char/tpm/Kconfig @@ -25,13 +25,14 @@ menuconfig TCG_TPM if TCG_TPM config TCG_TIS - tristate "TPM Interface Specification 1.2 Interface" + tristate "TPM Interface Specification 1.2 Interface / TPM 2.0 FIFO Interface" depends on X86 ---help--- If you have a TPM security chip that is compliant with the - TCG TIS 1.2 TPM specification say Yes and it will be accessible - from within Linux. To compile this driver as a module, choose - M here; the module will be called tpm_tis. + TCG TIS 1.2 TPM specification (TPM1.2) or the TCG PTP FIFO + specification (TPM2.0) say Yes and it will be accessible from + within Linux. To compile this driver as a module, choose M here; + the module will be called tpm_tis. config TCG_TIS_I2C_ATMEL tristate "TPM Interface Specification 1.2 Interface (I2C - Atmel)" @@ -100,16 +101,6 @@ config TCG_IBMVTPM will be accessible from within Linux. To compile this driver as a module, choose M here; the module will be called tpm_ibmvtpm. -config TCG_TIS_I2C_ST33 - tristate "TPM Interface Specification 1.2 Interface (I2C - STMicroelectronics)" - depends on I2C - depends on GPIOLIB - ---help--- - If you have a TPM security chip from STMicroelectronics working with - an I2C bus say Yes and it will be accessible from within Linux. - To compile this driver as a module, choose M here; the module will be - called tpm_i2c_stm_st33. - config TCG_XEN tristate "XEN TPM Interface" depends on TCG_TPM && XEN @@ -131,4 +122,5 @@ config TCG_CRB from within Linux. To compile this driver as a module, choose M here; the module will be called tpm_crb. +source "drivers/char/tpm/st33zp24/Kconfig" endif # TCG_TPM diff --git a/drivers/char/tpm/Makefile b/drivers/char/tpm/Makefile index 990cf183931d..56e8f1f3dc7e 100644 --- a/drivers/char/tpm/Makefile +++ b/drivers/char/tpm/Makefile @@ -20,6 +20,6 @@ obj-$(CONFIG_TCG_NSC) += tpm_nsc.o obj-$(CONFIG_TCG_ATMEL) += tpm_atmel.o obj-$(CONFIG_TCG_INFINEON) += tpm_infineon.o obj-$(CONFIG_TCG_IBMVTPM) += tpm_ibmvtpm.o -obj-$(CONFIG_TCG_TIS_I2C_ST33) += tpm_i2c_stm_st33.o +obj-$(CONFIG_TCG_TIS_ST33ZP24) += st33zp24/ obj-$(CONFIG_TCG_XEN) += xen-tpmfront.o obj-$(CONFIG_TCG_CRB) += tpm_crb.o diff --git a/drivers/char/tpm/st33zp24/Kconfig b/drivers/char/tpm/st33zp24/Kconfig new file mode 100644 index 000000000000..09cb727864f0 --- /dev/null +++ b/drivers/char/tpm/st33zp24/Kconfig @@ -0,0 +1,30 @@ +config TCG_TIS_ST33ZP24 + tristate "STMicroelectronics TPM Interface Specification 1.2 Interface" + depends on GPIOLIB + ---help--- + STMicroelectronics ST33ZP24 core driver. It implements the core + TPM1.2 logic and hooks into the TPM kernel APIs. Physical layers will + register against it. + + To compile this driver as a module, choose m here. The module will be called + tpm_st33zp24. + +config TCG_TIS_ST33ZP24_I2C + tristate "TPM 1.2 ST33ZP24 I2C support" + depends on TCG_TIS_ST33ZP24 + depends on I2C + ---help--- + This module adds support for the STMicroelectronics TPM security chip + ST33ZP24 with i2c interface. + To compile this driver as a module, choose M here; the module will be + called tpm_st33zp24_i2c. + +config TCG_TIS_ST33ZP24_SPI + tristate "TPM 1.2 ST33ZP24 SPI support" + depends on TCG_TIS_ST33ZP24 + depends on SPI + ---help--- + This module adds support for the STMicroelectronics TPM security chip + ST33ZP24 with spi interface. + To compile this driver as a module, choose M here; the module will be + called tpm_st33zp24_spi. diff --git a/drivers/char/tpm/st33zp24/Makefile b/drivers/char/tpm/st33zp24/Makefile new file mode 100644 index 000000000000..74a722e5e068 --- /dev/null +++ b/drivers/char/tpm/st33zp24/Makefile @@ -0,0 +1,12 @@ +# +# Makefile for ST33ZP24 TPM 1.2 driver +# + +tpm_st33zp24-objs = st33zp24.o +obj-$(CONFIG_TCG_TIS_ST33ZP24) += tpm_st33zp24.o + +tpm_st33zp24_i2c-objs = i2c.o +obj-$(CONFIG_TCG_TIS_ST33ZP24_I2C) += tpm_st33zp24_i2c.o + +tpm_st33zp24_spi-objs = spi.o +obj-$(CONFIG_TCG_TIS_ST33ZP24_SPI) += tpm_st33zp24_spi.o diff --git a/drivers/char/tpm/st33zp24/i2c.c b/drivers/char/tpm/st33zp24/i2c.c new file mode 100644 index 000000000000..ad1ee180e0c2 --- /dev/null +++ b/drivers/char/tpm/st33zp24/i2c.c @@ -0,0 +1,276 @@ +/* + * STMicroelectronics TPM I2C Linux driver for TPM ST33ZP24 + * Copyright (C) 2009 - 2015 STMicroelectronics + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * 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. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, see <http://www.gnu.org/licenses/>. + */ + +#include <linux/module.h> +#include <linux/i2c.h> +#include <linux/gpio.h> +#include <linux/of_irq.h> +#include <linux/of_gpio.h> +#include <linux/tpm.h> +#include <linux/platform_data/st33zp24.h> + +#include "st33zp24.h" + +#define TPM_DUMMY_BYTE 0xAA + +struct st33zp24_i2c_phy { + struct i2c_client *client; + u8 buf[TPM_BUFSIZE + 1]; + int io_lpcpd; +}; + +/* + * write8_reg + * Send byte to the TIS register according to the ST33ZP24 I2C protocol. + * @param: tpm_register, the tpm tis register where the data should be written + * @param: tpm_data, the tpm_data to write inside the tpm_register + * @param: tpm_size, The length of the data + * @return: Returns negative errno, or else the number of bytes written. + */ +static int write8_reg(void *phy_id, u8 tpm_register, u8 *tpm_data, int tpm_size) +{ + struct st33zp24_i2c_phy *phy = phy_id; + + phy->buf[0] = tpm_register; + memcpy(phy->buf + 1, tpm_data, tpm_size); + return i2c_master_send(phy->client, phy->buf, tpm_size + 1); +} /* write8_reg() */ + +/* + * read8_reg + * Recv byte from the TIS register according to the ST33ZP24 I2C protocol. + * @param: tpm_register, the tpm tis register where the data should be read + * @param: tpm_data, the TPM response + * @param: tpm_size, tpm TPM response size to read. + * @return: number of byte read successfully: should be one if success. + */ +static int read8_reg(void *phy_id, u8 tpm_register, u8 *tpm_data, int tpm_size) +{ + struct st33zp24_i2c_phy *phy = phy_id; + u8 status = 0; + u8 data; + + data = TPM_DUMMY_BYTE; + status = write8_reg(phy, tpm_register, &data, 1); + if (status == 2) + status = i2c_master_recv(phy->client, tpm_data, tpm_size); + return status; +} /* read8_reg() */ + +/* + * st33zp24_i2c_send + * Send byte to the TIS register according to the ST33ZP24 I2C protocol. + * @param: phy_id, the phy description + * @param: tpm_register, the tpm tis register where the data should be written + * @param: tpm_data, the tpm_data to write inside the tpm_register + * @param: tpm_size, the length of the data + * @return: number of byte written successfully: should be one if success. + */ +static int st33zp24_i2c_send(void *phy_id, u8 tpm_register, u8 *tpm_data, + int tpm_size) +{ + return write8_reg(phy_id, tpm_register | TPM_WRITE_DIRECTION, tpm_data, + tpm_size); +} + +/* + * st33zp24_i2c_recv + * Recv byte from the TIS register according to the ST33ZP24 I2C protocol. + * @param: phy_id, the phy description + * @param: tpm_register, the tpm tis register where the data should be read + * @param: tpm_data, the TPM response + * @param: tpm_size, tpm TPM response size to read. + * @return: number of byte read successfully: should be one if success. + */ +static int st33zp24_i2c_recv(void *phy_id, u8 tpm_register, u8 *tpm_data, + int tpm_size) +{ + return read8_reg(phy_id, tpm_register, tpm_data, tpm_size); +} + +static const struct st33zp24_phy_ops i2c_phy_ops = { + .send = st33zp24_i2c_send, + .recv = st33zp24_i2c_recv, +}; + +#ifdef CONFIG_OF +static int st33zp24_i2c_of_request_resources(struct st33zp24_i2c_phy *phy) +{ + struct device_node *pp; + struct i2c_client *client = phy->client; + int gpio; + int ret; + + pp = client->dev.of_node; + if (!pp) { + dev_err(&client->dev, "No platform data\n"); + return -ENODEV; + } + + /* Get GPIO from device tree */ + gpio = of_get_named_gpio(pp, "lpcpd-gpios", 0); + if (gpio < 0) { + dev_err(&client->dev, + "Failed to retrieve lpcpd-gpios from dts.\n"); + phy->io_lpcpd = -1; + /* + * lpcpd pin is not specified. This is not an issue as + * power management can be also managed by TPM specific + * commands. So leave with a success status code. + */ + return 0; + } + /* GPIO request and configuration */ + ret = devm_gpio_request_one(&client->dev, gpio, + GPIOF_OUT_INIT_HIGH, "TPM IO LPCPD"); + if (ret) { + dev_err(&client->dev, "Failed to request lpcpd pin\n"); + return -ENODEV; + } + phy->io_lpcpd = gpio; + + return 0; +} +#else +static int st33zp24_i2c_of_request_resources(struct st33zp24_i2c_phy *phy) +{ + return -ENODEV; +} +#endif + +static int st33zp24_i2c_request_resources(struct i2c_client *client, + struct st33zp24_i2c_phy *phy) +{ + struct st33zp24_platform_data *pdata; + int ret; + + pdata = client->dev.platform_data; + if (!pdata) { + dev_err(&client->dev, "No platform data\n"); + return -ENODEV; + } + + /* store for late use */ + phy->io_lpcpd = pdata->io_lpcpd; + + if (gpio_is_valid(pdata->io_lpcpd)) { + ret = devm_gpio_request_one(&client->dev, + pdata->io_lpcpd, GPIOF_OUT_INIT_HIGH, + "TPM IO_LPCPD"); + if (ret) { + dev_err(&client->dev, "Failed to request lpcpd pin\n"); + return ret; + } + } + + return 0; +} + +/* + * st33zp24_i2c_probe initialize the TPM device + * @param: client, the i2c_client drescription (TPM I2C description). + * @param: id, the i2c_device_id struct. + * @return: 0 in case of success. + * -1 in other case. + */ +static int st33zp24_i2c_probe(struct i2c_client *client, + const struct i2c_device_id *id) +{ + int ret; + struct st33zp24_platform_data *pdata; + struct st33zp24_i2c_phy *phy; + + if (!client) { + pr_info("%s: i2c client is NULL. Device not accessible.\n", + __func__); + return -ENODEV; + } + + if (!i2c_check_functionality(client->adapter, I2C_FUNC_I2C)) { + dev_info(&client->dev, "client not i2c capable\n"); + return -ENODEV; + } + + phy = devm_kzalloc(&client->dev, sizeof(struct st33zp24_i2c_phy), + GFP_KERNEL); + if (!phy) + return -ENOMEM; + + phy->client = client; + pdata = client->dev.platform_data; + if (!pdata && client->dev.of_node) { + ret = st33zp24_i2c_of_request_resources(phy); + if (ret) + return ret; + } else if (pdata) { + ret = st33zp24_i2c_request_resources(client, phy); + if (ret) + return ret; + } + + return st33zp24_probe(phy, &i2c_phy_ops, &client->dev, client->irq, + phy->io_lpcpd); +} + +/* + * st33zp24_i2c_remove remove the TPM device + * @param: client, the i2c_client description (TPM I2C description). + * @return: 0 in case of success. + */ +static int st33zp24_i2c_remove(struct i2c_client *client) +{ + struct tpm_chip *chip = i2c_get_clientdata(client); + + return st33zp24_remove(chip); +} + +static const struct i2c_device_id st33zp24_i2c_id[] = { + {TPM_ST33_I2C, 0}, + {} +}; +MODULE_DEVICE_TABLE(i2c, st33zp24_i2c_id); + +#ifdef CONFIG_OF +static const struct of_device_id of_st33zp24_i2c_match[] = { + { .compatible = "st,st33zp24-i2c", }, + {} +}; +MODULE_DEVICE_TABLE(of, of_st33zp24_i2c_match); +#endif + +static SIMPLE_DEV_PM_OPS(st33zp24_i2c_ops, st33zp24_pm_suspend, + st33zp24_pm_resume); + +static struct i2c_driver st33zp24_i2c_driver = { + .driver = { + .owner = THIS_MODULE, + .name = TPM_ST33_I2C, + .pm = &st33zp24_i2c_ops, + .of_match_table = of_match_ptr(of_st33zp24_i2c_match), + }, + .probe = st33zp24_i2c_probe, + .remove = st33zp24_i2c_remove, + .id_table = st33zp24_i2c_id +}; + +module_i2c_driver(st33zp24_i2c_driver); + +MODULE_AUTHOR("TPM support (TPMsupport@list.st.com)"); +MODULE_DESCRIPTION("STM TPM 1.2 I2C ST33 Driver"); +MODULE_VERSION("1.3.0"); +MODULE_LICENSE("GPL"); diff --git a/drivers/char/tpm/st33zp24/spi.c b/drivers/char/tpm/st33zp24/spi.c new file mode 100644 index 000000000000..f0184a1b0c1c --- /dev/null +++ b/drivers/char/tpm/st33zp24/spi.c @@ -0,0 +1,399 @@ +/* + * STMicroelectronics TPM SPI Linux driver for TPM ST33ZP24 + * Copyright (C) 2009 - 2015 STMicroelectronics + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * 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. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, see <http://www.gnu.org/licenses/>. + */ + +#include <linux/module.h> +#include <linux/spi/spi.h> +#include <linux/gpio.h> +#include <linux/of_irq.h> +#include <linux/of_gpio.h> +#include <linux/tpm.h> +#include <linux/platform_data/st33zp24.h> + +#include "st33zp24.h" + +#define TPM_DATA_FIFO 0x24 +#define TPM_INTF_CAPABILITY 0x14 + +#define TPM_DUMMY_BYTE 0x00 + +#define MAX_SPI_LATENCY 15 +#define LOCALITY0 0 + +#define ST33ZP24_OK 0x5A +#define ST33ZP24_UNDEFINED_ERR 0x80 +#define ST33ZP24_BADLOCALITY 0x81 +#define ST33ZP24_TISREGISTER_UKNOWN 0x82 +#define ST33ZP24_LOCALITY_NOT_ACTIVATED 0x83 +#define ST33ZP24_HASH_END_BEFORE_HASH_START 0x84 +#define ST33ZP24_BAD_COMMAND_ORDER 0x85 +#define ST33ZP24_INCORECT_RECEIVED_LENGTH 0x86 +#define ST33ZP24_TPM_FIFO_OVERFLOW 0x89 +#define ST33ZP24_UNEXPECTED_READ_FIFO 0x8A +#define ST33ZP24_UNEXPECTED_WRITE_FIFO 0x8B +#define ST33ZP24_CMDRDY_SET_WHEN_PROCESSING_HASH_END 0x90 +#define ST33ZP24_DUMMY_BYTES 0x00 + +/* + * TPM command can be up to 2048 byte, A TPM response can be up to + * 1024 byte. + * Between command and response, there are latency byte (up to 15 + * usually on st33zp24 2 are enough). + * + * Overall when sending a command and expecting an answer we need if + * worst case: + * 2048 (for the TPM command) + 1024 (for the TPM answer). We need + * some latency byte before the answer is available (max 15). + * We have 2048 + 1024 + 15. + */ +#define ST33ZP24_SPI_BUFFER_SIZE (TPM_BUFSIZE + (TPM_BUFSIZE / 2) +\ + MAX_SPI_LATENCY) + + +struct st33zp24_spi_phy { + struct spi_device *spi_device; + struct spi_transfer spi_xfer; + u8 tx_buf[ST33ZP24_SPI_BUFFER_SIZE]; + u8 rx_buf[ST33ZP24_SPI_BUFFER_SIZE]; + + int io_lpcpd; + int latency; +}; + +static int st33zp24_status_to_errno(u8 code) +{ + switch (code) { + case ST33ZP24_OK: + return 0; + case ST33ZP24_UNDEFINED_ERR: + case ST33ZP24_BADLOCALITY: + case ST33ZP24_TISREGISTER_UKNOWN: + case ST33ZP24_LOCALITY_NOT_ACTIVATED: + case ST33ZP24_HASH_END_BEFORE_HASH_START: + case ST33ZP24_BAD_COMMAND_ORDER: + case ST33ZP24_UNEXPECTED_READ_FIFO: + case ST33ZP24_UNEXPECTED_WRITE_FIFO: + case ST33ZP24_CMDRDY_SET_WHEN_PROCESSING_HASH_END: + return -EPROTO; + case ST33ZP24_INCORECT_RECEIVED_LENGTH: + case ST33ZP24_TPM_FIFO_OVERFLOW: + return -EMSGSIZE; + case ST33ZP24_DUMMY_BYTES: + return -ENOSYS; + } + return code; +} + +/* + * st33zp24_spi_send + * Send byte to the TIS register according to the ST33ZP24 SPI protocol. + * @param: phy_id, the phy description + * @param: tpm_register, the tpm tis register where the data should be written + * @param: tpm_data, the tpm_data to write inside the tpm_register + * @param: tpm_size, The length of the data + * @return: should be zero if success else a negative error code. + */ +static int st33zp24_spi_send(void *phy_id, u8 tpm_register, u8 *tpm_data, + int tpm_size) +{ + u8 data = 0; + int total_length = 0, nbr_dummy_bytes = 0, ret = 0; + struct st33zp24_spi_phy *phy = phy_id; + struct spi_device *dev = phy->spi_device; + u8 *tx_buf = (u8 *)phy->spi_xfer.tx_buf; + u8 *rx_buf = phy->spi_xfer.rx_buf; + + /* Pre-Header */ + data = TPM_WRITE_DIRECTION | LOCALITY0; + memcpy(tx_buf + total_length, &data, sizeof(data)); + total_length++; + data = tpm_register; + memcpy(tx_buf + total_length, &data, sizeof(data)); + total_length++; + + if (tpm_size > 0 && tpm_register == TPM_DATA_FIFO) { + tx_buf[total_length++] = tpm_size >> 8; + tx_buf[total_length++] = tpm_size; + } + + memcpy(&tx_buf[total_length], tpm_data, tpm_size); + total_length += tpm_size; + + nbr_dummy_bytes = phy->latency; + memset(&tx_buf[total_length], TPM_DUMMY_BYTE, nbr_dummy_bytes); + + phy->spi_xfer.len = total_length + nbr_dummy_bytes; + + ret = spi_sync_transfer(dev, &phy->spi_xfer, 1); + if (ret == 0) + ret = rx_buf[total_length + nbr_dummy_bytes - 1]; + + return st33zp24_status_to_errno(ret); +} /* st33zp24_spi_send() */ + +/* + * read8_recv + * Recv byte from the TIS register according to the ST33ZP24 SPI protocol. + * @param: phy_id, the phy description + * @param: tpm_register, the tpm tis register where the data should be read + * @param: tpm_data, the TPM response + * @param: tpm_size, tpm TPM response size to read. + * @return: should be zero if success else a negative error code. + */ +static int read8_reg(void *phy_id, u8 tpm_register, u8 *tpm_data, int tpm_size) +{ + u8 data = 0; + int total_length = 0, nbr_dummy_bytes, ret; + struct st33zp24_spi_phy *phy = phy_id; + struct spi_device *dev = phy->spi_device; + u8 *tx_buf = (u8 *)phy->spi_xfer.tx_buf; + u8 *rx_buf = phy->spi_xfer.rx_buf; + + /* Pre-Header */ + data = LOCALITY0; + memcpy(tx_buf + total_length, &data, sizeof(data)); + total_length++; + data = tpm_register; + memcpy(tx_buf + total_length, &data, sizeof(data)); + total_length++; + + nbr_dummy_bytes = phy->latency; + memset(&tx_buf[total_length], TPM_DUMMY_BYTE, + nbr_dummy_bytes + tpm_size); + + phy->spi_xfer.len = total_length + nbr_dummy_bytes + tpm_size; + + /* header + status byte + size of the data + status byte */ + ret = spi_sync_transfer(dev, &phy->spi_xfer, 1); + if (tpm_size > 0 && ret == 0) { + ret = rx_buf[total_length + nbr_dummy_bytes - 1]; + + memcpy(tpm_data, rx_buf + total_length + nbr_dummy_bytes, + tpm_size); + } + + return ret; +} /* read8_reg() */ + +/* + * st33zp24_spi_recv + * Recv byte from the TIS register according to the ST33ZP24 SPI protocol. + * @param: phy_id, the phy description + * @param: tpm_register, the tpm tis register where the data should be read + * @param: tpm_data, the TPM response + * @param: tpm_size, tpm TPM response size to read. + * @return: number of byte read successfully: should be one if success. + */ +static int st33zp24_spi_recv(void *phy_id, u8 tpm_register, u8 *tpm_data, + int tpm_size) +{ + int ret; + + ret = read8_reg(phy_id, tpm_register, tpm_data, tpm_size); + if (!st33zp24_status_to_errno(ret)) + return tpm_size; + return ret; +} /* st33zp24_spi_recv() */ + +static int evaluate_latency(void *phy_id) +{ + struct st33zp24_spi_phy *phy = phy_id; + int latency = 1, status = 0; + u8 data = 0; + + while (!status && latency < MAX_SPI_LATENCY) { + phy->latency = latency; + status = read8_reg(phy_id, TPM_INTF_CAPABILITY, &data, 1); + latency++; + } + return latency - 1; +} /* evaluate_latency() */ + +static const struct st33zp24_phy_ops spi_phy_ops = { + .send = st33zp24_spi_send, + .recv = st33zp24_spi_recv, +}; + +#ifdef CONFIG_OF +static int tpm_stm_spi_of_request_resources(struct st33zp24_spi_phy *phy) +{ + struct device_node *pp; + struct spi_device *dev = phy->spi_device; + int gpio; + int ret; + + pp = dev->dev.of_node; + if (!pp) { + dev_err(&dev->dev, "No platform data\n"); + return -ENODEV; + } + + /* Get GPIO from device tree */ + gpio = of_get_named_gpio(pp, "lpcpd-gpios", 0); + if (gpio < 0) { + dev_err(&dev->dev, + "Failed to retrieve lpcpd-gpios from dts.\n"); + phy->io_lpcpd = -1; + /* + * lpcpd pin is not specified. This is not an issue as + * power management can be also managed by TPM specific + * commands. So leave with a success status code. + */ + return 0; + } + /* GPIO request and configuration */ + ret = devm_gpio_request_one(&dev->dev, gpio, + GPIOF_OUT_INIT_HIGH, "TPM IO LPCPD"); + if (ret) { + dev_err(&dev->dev, "Failed to request lpcpd pin\n"); + return -ENODEV; + } + phy->io_lpcpd = gpio; + + return 0; +} +#else +static int tpm_stm_spi_of_request_resources(struct st33zp24_spi_phy *phy) +{ + return -ENODEV; +} +#endif + +static int tpm_stm_spi_request_resources(struct spi_device *dev, + struct st33zp24_spi_phy *phy) +{ + struct st33zp24_platform_data *pdata; + int ret; + + pdata = dev->dev.platform_data; + if (!pdata) { + dev_err(&dev->dev, "No platform data\n"); + return -ENODEV; + } + + /* store for late use */ + phy->io_lpcpd = pdata->io_lpcpd; + + if (gpio_is_valid(pdata->io_lpcpd)) { + ret = devm_gpio_request_one(&dev->dev, + pdata->io_lpcpd, GPIOF_OUT_INIT_HIGH, + "TPM IO_LPCPD"); + if (ret) { + dev_err(&dev->dev, "%s : reset gpio_request failed\n", + __FILE__); + return ret; + } + } + + return 0; +} + +/* + * tpm_st33_spi_probe initialize the TPM device + * @param: dev, the spi_device drescription (TPM SPI description). + * @return: 0 in case of success. + * or a negative value describing the error. + */ +static int +tpm_st33_spi_probe(struct spi_device *dev) +{ + int ret; + struct st33zp24_platform_data *pdata; + struct st33zp24_spi_phy *phy; + + /* Check SPI platform functionnalities */ + if (!dev) { + pr_info("%s: dev is NULL. Device is not accessible.\n", + __func__); + return -ENODEV; + } + + phy = devm_kzalloc(&dev->dev, sizeof(struct st33zp24_spi_phy), + GFP_KERNEL); + if (!phy) + return -ENOMEM; + + phy->spi_device = dev; + pdata = dev->dev.platform_data; + if (!pdata && dev->dev.of_node) { + ret = tpm_stm_spi_of_request_resources(phy); + if (ret) + return ret; + } else if (pdata) { + ret = tpm_stm_spi_request_resources(dev, phy); + if (ret) + return ret; + } + + phy->spi_xfer.tx_buf = phy->tx_buf; + phy->spi_xfer.rx_buf = phy->rx_buf; + + phy->latency = evaluate_latency(phy); + if (phy->latency <= 0) + return -ENODEV; + + return st33zp24_probe(phy, &spi_phy_ops, &dev->dev, dev->irq, + phy->io_lpcpd); +} + +/* + * tpm_st33_spi_remove remove the TPM device + * @param: client, the spi_device drescription (TPM SPI description). + * @return: 0 in case of success. + */ +static int tpm_st33_spi_remove(struct spi_device *dev) +{ + struct tpm_chip *chip = spi_get_drvdata(dev); + + return st33zp24_remove(chip); +} + +static const struct spi_device_id st33zp24_spi_id[] = { + {TPM_ST33_SPI, 0}, + {} +}; +MODULE_DEVICE_TABLE(spi, st33zp24_spi_id); + +#ifdef CONFIG_OF +static const struct of_device_id of_st33zp24_spi_match[] = { + { .compatible = "st,st33zp24-spi", }, + {} +}; +MODULE_DEVICE_TABLE(of, of_st33zp24_spi_match); +#endif + +static SIMPLE_DEV_PM_OPS(st33zp24_spi_ops, st33zp24_pm_suspend, + st33zp24_pm_resume); + +static struct spi_driver tpm_st33_spi_driver = { + .driver = { + .owner = THIS_MODULE, + .name = TPM_ST33_SPI, + .pm = &st33zp24_spi_ops, + .of_match_table = of_match_ptr(of_st33zp24_spi_match), + }, + .probe = tpm_st33_spi_probe, + .remove = tpm_st33_spi_remove, + .id_table = st33zp24_spi_id, +}; + +module_spi_driver(tpm_st33_spi_driver); + +MODULE_AUTHOR("TPM support (TPMsupport@list.st.com)"); +MODULE_DESCRIPTION("STM TPM 1.2 SPI ST33 Driver"); +MODULE_VERSION("1.3.0"); +MODULE_LICENSE("GPL"); diff --git a/drivers/char/tpm/st33zp24/st33zp24.c b/drivers/char/tpm/st33zp24/st33zp24.c new file mode 100644 index 000000000000..8d626784cd8d --- /dev/null +++ b/drivers/char/tpm/st33zp24/st33zp24.c @@ -0,0 +1,698 @@ +/* + * STMicroelectronics TPM Linux driver for TPM ST33ZP24 + * Copyright (C) 2009 - 2015 STMicroelectronics + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * 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. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, see <http://www.gnu.org/licenses/>. + */ + +#include <linux/module.h> +#include <linux/fs.h> +#include <linux/miscdevice.h> +#include <linux/kernel.h> +#include <linux/delay.h> +#include <linux/wait.h> +#include <linux/freezer.h> +#include <linux/string.h> +#include <linux/interrupt.h> +#include <linux/gpio.h> +#include <linux/sched.h> +#include <linux/uaccess.h> +#include <linux/io.h> +#include <linux/slab.h> + +#include "../tpm.h" +#include "st33zp24.h" + +#define TPM_ACCESS 0x0 +#define TPM_STS 0x18 +#define TPM_DATA_FIFO 0x24 +#define TPM_INTF_CAPABILITY 0x14 +#define TPM_INT_STATUS 0x10 +#define TPM_INT_ENABLE 0x08 + +#define LOCALITY0 0 + +enum st33zp24_access { + TPM_ACCESS_VALID = 0x80, + TPM_ACCESS_ACTIVE_LOCALITY = 0x20, + TPM_ACCESS_REQUEST_PENDING = 0x04, + TPM_ACCESS_REQUEST_USE = 0x02, +}; + +enum st33zp24_status { + TPM_STS_VALID = 0x80, + TPM_STS_COMMAND_READY = 0x40, + TPM_STS_GO = 0x20, + TPM_STS_DATA_AVAIL = 0x10, + TPM_STS_DATA_EXPECT = 0x08, +}; + +enum st33zp24_int_flags { + TPM_GLOBAL_INT_ENABLE = 0x80, + TPM_INTF_CMD_READY_INT = 0x080, + TPM_INTF_FIFO_AVALAIBLE_INT = 0x040, + TPM_INTF_WAKE_UP_READY_INT = 0x020, + TPM_INTF_LOCALITY_CHANGE_INT = 0x004, + TPM_INTF_STS_VALID_INT = 0x002, + TPM_INTF_DATA_AVAIL_INT = 0x001, +}; + +enum tis_defaults { + TIS_SHORT_TIMEOUT = 750, + TIS_LONG_TIMEOUT = 2000, +}; + +struct st33zp24_dev { + struct tpm_chip *chip; + void *phy_id; + const struct st33zp24_phy_ops *ops; + u32 intrs; + int io_lpcpd; +}; + +/* + * clear_interruption clear the pending interrupt. + * @param: tpm_dev, the tpm device device. + * @return: the interrupt status value. + */ +static u8 clear_interruption(struct st33zp24_dev *tpm_dev) +{ + u8 interrupt; + + tpm_dev->ops->recv(tpm_dev->phy_id, TPM_INT_STATUS, &interrupt, 1); + tpm_dev->ops->send(tpm_dev->phy_id, TPM_INT_STATUS, &interrupt, 1); + return interrupt; +} /* clear_interruption() */ + +/* + * st33zp24_cancel, cancel the current command execution or + * set STS to COMMAND READY. + * @param: chip, the tpm_chip description as specified in driver/char/tpm/tpm.h + */ +static void st33zp24_cancel(struct tpm_chip *chip) +{ + struct st33zp24_dev *tpm_dev; + u8 data; + + tpm_dev = (struct st33zp24_dev *)TPM_VPRIV(chip); + + data = TPM_STS_COMMAND_READY; + tpm_dev->ops->send(tpm_dev->phy_id, TPM_STS, &data, 1); +} /* st33zp24_cancel() */ + +/* + * st33zp24_status return the TPM_STS register + * @param: chip, the tpm chip description + * @return: the TPM_STS register value. + */ +static u8 st33zp24_status(struct tpm_chip *chip) +{ + struct st33zp24_dev *tpm_dev; + u8 data; + + tpm_dev = (struct st33zp24_dev *)TPM_VPRIV(chip); + + tpm_dev->ops->recv(tpm_dev->phy_id, TPM_STS, &data, 1); + return data; +} /* st33zp24_status() */ + +/* + * check_locality if the locality is active + * @param: chip, the tpm chip description + * @return: the active locality or -EACCESS. + */ +static int check_locality(struct tpm_chip *chip) +{ + struct st33zp24_dev *tpm_dev; + u8 data; + u8 status; + + tpm_dev = (struct st33zp24_dev *)TPM_VPRIV(chip); + + status = tpm_dev->ops->recv(tpm_dev->phy_id, TPM_ACCESS, &data, 1); + if (status && (data & + (TPM_ACCESS_ACTIVE_LOCALITY | TPM_ACCESS_VALID)) == + (TPM_ACCESS_ACTIVE_LOCALITY | TPM_ACCESS_VALID)) + return chip->vendor.locality; + + return -EACCES; +} /* check_locality() */ + +/* + * request_locality request the TPM locality + * @param: chip, the chip description + * @return: the active locality or negative value. + */ +static int request_locality(struct tpm_chip *chip) +{ + unsigned long stop; + long ret; + struct st33zp24_dev *tpm_dev; + u8 data; + + if (check_locality(chip) == chip->vendor.locality) + return chip->vendor.locality; + + tpm_dev = (struct st33zp24_dev *)TPM_VPRIV(chip); + + data = TPM_ACCESS_REQUEST_USE; + ret = tpm_dev->ops->send(tpm_dev->phy_id, TPM_ACCESS, &data, 1); + if (ret < 0) + return ret; + + stop = jiffies + chip->vendor.timeout_a; + + /* Request locality is usually effective after the request */ + do { + if (check_locality(chip) >= 0) + return chip->vendor.locality; + msleep(TPM_TIMEOUT); + } while (time_before(jiffies, stop)); + + /* could not get locality */ + return -EACCES; +} /* request_locality() */ + +/* + * release_locality release the active locality + * @param: chip, the tpm chip description. + */ +static void release_locality(struct tpm_chip *chip) +{ + struct st33zp24_dev *tpm_dev; + u8 data; + + tpm_dev = (struct st33zp24_dev *)TPM_VPRIV(chip); + data = TPM_ACCESS_ACTIVE_LOCALITY; + + tpm_dev->ops->send(tpm_dev->phy_id, TPM_ACCESS, &data, 1); +} + +/* + * get_burstcount return the burstcount value + * @param: chip, the chip description + * return: the burstcount or negative value. + */ +static int get_burstcount(struct tpm_chip *chip) +{ + unsigned long stop; + int burstcnt, status; + u8 tpm_reg, temp; + struct st33zp24_dev *tpm_dev; + + tpm_dev = (struct st33zp24_dev *)TPM_VPRIV(chip); + + stop = jiffies + chip->vendor.timeout_d; + do { + tpm_reg = TPM_STS + 1; + status = tpm_dev->ops->recv(tpm_dev->phy_id, tpm_reg, &temp, 1); + if (status < 0) + return -EBUSY; + + tpm_reg = TPM_STS + 2; + burstcnt = temp; + status = tpm_dev->ops->recv(tpm_dev->phy_id, tpm_reg, &temp, 1); + if (status < 0) + return -EBUSY; + + burstcnt |= temp << 8; + if (burstcnt) + return burstcnt; + msleep(TPM_TIMEOUT); + } while (time_before(jiffies, stop)); + return -EBUSY; +} /* get_burstcount() */ + + +/* + * wait_for_tpm_stat_cond + * @param: chip, chip description + * @param: mask, expected mask value + * @param: check_cancel, does the command expected to be canceled ? + * @param: canceled, did we received a cancel request ? + * @return: true if status == mask or if the command is canceled. + * false in other cases. + */ +static bool wait_for_tpm_stat_cond(struct tpm_chip *chip, u8 mask, + bool check_cancel, bool *canceled) +{ + u8 status = chip->ops->status(chip); + + *canceled = false; + if ((status & mask) == mask) + return true; + if (check_cancel && chip->ops->req_canceled(chip, status)) { + *canceled = true; + return true; + } + return false; +} + +/* + * wait_for_stat wait for a TPM_STS value + * @param: chip, the tpm chip description + * @param: mask, the value mask to wait + * @param: timeout, the timeout + * @param: queue, the wait queue. + * @param: check_cancel, does the command can be cancelled ? + * @return: the tpm status, 0 if success, -ETIME if timeout is reached. + */ +static int wait_for_stat(struct tpm_chip *chip, u8 mask, unsigned long timeout, + wait_queue_head_t *queue, bool check_cancel) +{ + unsigned long stop; + int ret = 0; + bool canceled = false; + bool condition; + u32 cur_intrs; + u8 status; + struct st33zp24_dev *tpm_dev; + + tpm_dev = (struct st33zp24_dev *)TPM_VPRIV(chip); + + /* check current status */ + status = st33zp24_status(chip); + if ((status & mask) == mask) + return 0; + + stop = jiffies + timeout; + + if (chip->vendor.irq) { + cur_intrs = tpm_dev->intrs; + clear_interruption(tpm_dev); + enable_irq(chip->vendor.irq); + + do { + if (ret == -ERESTARTSYS && freezing(current)) + clear_thread_flag(TIF_SIGPENDING); + + timeout = stop - jiffies; + if ((long) timeout <= 0) + return -1; + + ret = wait_event_interruptible_timeout(*queue, + cur_intrs != tpm_dev->intrs, + timeout); + clear_interruption(tpm_dev); + condition = wait_for_tpm_stat_cond(chip, mask, + check_cancel, &canceled); + if (ret >= 0 && condition) { + if (canceled) + return -ECANCELED; + return 0; + } + } while (ret == -ERESTARTSYS && freezing(current)); + + disable_irq_nosync(chip->vendor.irq); + + } else { + do { + msleep(TPM_TIMEOUT); + status = chip->ops->status(chip); + if ((status & mask) == mask) + return 0; + } while (time_before(jiffies, stop)); + } + + return -ETIME; +} /* wait_for_stat() */ + +/* + * recv_data receive data + * @param: chip, the tpm chip description + * @param: buf, the buffer where the data are received + * @param: count, the number of data to receive + * @return: the number of bytes read from TPM FIFO. + */ +static int recv_data(struct tpm_chip *chip, u8 *buf, size_t count) +{ + int size = 0, burstcnt, len, ret; + struct st33zp24_dev *tpm_dev; + + tpm_dev = (struct st33zp24_dev *)TPM_VPRIV(chip); + + while (size < count && + wait_for_stat(chip, + TPM_STS_DATA_AVAIL | TPM_STS_VALID, + chip->vendor.timeout_c, + &chip->vendor.read_queue, true) == 0) { + burstcnt = get_burstcount(chip); + if (burstcnt < 0) + return burstcnt; + len = min_t(int, burstcnt, count - size); + ret = tpm_dev->ops->recv(tpm_dev->phy_id, TPM_DATA_FIFO, + buf + size, len); + if (ret < 0) + return ret; + + size += len; + } + return size; +} + +/* + * tpm_ioserirq_handler the serirq irq handler + * @param: irq, the tpm chip description + * @param: dev_id, the description of the chip + * @return: the status of the handler. + */ +static irqreturn_t tpm_ioserirq_handler(int irq, void *dev_id) +{ + struct tpm_chip *chip = dev_id; + struct st33zp24_dev *tpm_dev; + + tpm_dev = (struct st33zp24_dev *)TPM_VPRIV(chip); + + tpm_dev->intrs++; + wake_up_interruptible(&chip->vendor.read_queue); + disable_irq_nosync(chip->vendor.irq); + + return IRQ_HANDLED; +} /* tpm_ioserirq_handler() */ + +/* + * st33zp24_send send TPM commands through the I2C bus. + * + * @param: chip, the tpm_chip description as specified in driver/char/tpm/tpm.h + * @param: buf, the buffer to send. + * @param: count, the number of bytes to send. + * @return: In case of success the number of bytes sent. + * In other case, a < 0 value describing the issue. + */ +static int st33zp24_send(struct tpm_chip *chip, unsigned char *buf, + size_t len) +{ + u32 status, i, size, ordinal; + int burstcnt = 0; + int ret; + u8 data; + struct st33zp24_dev *tpm_dev; + + if (!chip) + return -EBUSY; + if (len < TPM_HEADER_SIZE) + return -EBUSY; + + tpm_dev = (struct st33zp24_dev *)TPM_VPRIV(chip); + + ret = request_locality(chip); + if (ret < 0) + return ret; + + status = st33zp24_status(chip); + if ((status & TPM_STS_COMMAND_READY) == 0) { + st33zp24_cancel(chip); + if (wait_for_stat + (chip, TPM_STS_COMMAND_READY, chip->vendor.timeout_b, + &chip->vendor.read_queue, false) < 0) { + ret = -ETIME; + goto out_err; + } + } + + for (i = 0; i < len - 1;) { + burstcnt = get_burstcount(chip); + if (burstcnt < 0) + return burstcnt; + size = min_t(int, len - i - 1, burstcnt); + ret = tpm_dev->ops->send(tpm_dev->phy_id, TPM_DATA_FIFO, + buf + i, size); + if (ret < 0) + goto out_err; + + i += size; + } + + status = st33zp24_status(chip); + if ((status & TPM_STS_DATA_EXPECT) == 0) { + ret = -EIO; + goto out_err; + } + + ret = tpm_dev->ops->send(tpm_dev->phy_id, TPM_DATA_FIFO, + buf + len - 1, 1); + if (ret < 0) + goto out_err; + + status = st33zp24_status(chip); + if ((status & TPM_STS_DATA_EXPECT) != 0) { + ret = -EIO; + goto out_err; + } + + data = TPM_STS_GO; + ret = tpm_dev->ops->send(tpm_dev->phy_id, TPM_STS, &data, 1); + if (ret < 0) + goto out_err; + + if (chip->vendor.irq) { + ordinal = be32_to_cpu(*((__be32 *) (buf + 6))); + + ret = wait_for_stat(chip, TPM_STS_DATA_AVAIL | TPM_STS_VALID, + tpm_calc_ordinal_duration(chip, ordinal), + &chip->vendor.read_queue, false); + if (ret < 0) + goto out_err; + } + + return len; +out_err: + st33zp24_cancel(chip); + release_locality(chip); + return ret; +} + +/* + * st33zp24_recv received TPM response through TPM phy. + * @param: chip, the tpm_chip description as specified in driver/char/tpm/tpm.h. + * @param: buf, the buffer to store datas. + * @param: count, the number of bytes to send. + * @return: In case of success the number of bytes received. + * In other case, a < 0 value describing the issue. + */ +static int st33zp24_recv(struct tpm_chip *chip, unsigned char *buf, + size_t count) +{ + int size = 0; + int expected; + + if (!chip) + return -EBUSY; + + if (count < TPM_HEADER_SIZE) { + size = -EIO; + goto out; + } + + size = recv_data(chip, buf, TPM_HEADER_SIZE); + if (size < TPM_HEADER_SIZE) { + dev_err(&chip->dev, "Unable to read header\n"); + goto out; + } + + expected = be32_to_cpu(*(__be32 *)(buf + 2)); + if (expected > count) { + size = -EIO; + goto out; + } + + size += recv_data(chip, &buf[TPM_HEADER_SIZE], + expected - TPM_HEADER_SIZE); + if (size < expected) { + dev_err(&chip->dev, "Unable to read remainder of result\n"); + size = -ETIME; + } + +out: + st33zp24_cancel(chip); + release_locality(chip); + return size; +} + +/* + * st33zp24_req_canceled + * @param: chip, the tpm_chip description as specified in driver/char/tpm/tpm.h. + * @param: status, the TPM status. + * @return: Does TPM ready to compute a new command ? true. + */ +static bool st33zp24_req_canceled(struct tpm_chip *chip, u8 status) +{ + return (status == TPM_STS_COMMAND_READY); +} + +static const struct tpm_class_ops st33zp24_tpm = { + .send = st33zp24_send, + .recv = st33zp24_recv, + .cancel = st33zp24_cancel, + .status = st33zp24_status, + .req_complete_mask = TPM_STS_DATA_AVAIL | TPM_STS_VALID, + .req_complete_val = TPM_STS_DATA_AVAIL | TPM_STS_VALID, + .req_canceled = st33zp24_req_canceled, +}; + +/* + * st33zp24_probe initialize the TPM device + * @param: client, the i2c_client drescription (TPM I2C description). + * @param: id, the i2c_device_id struct. + * @return: 0 in case of success. + * -1 in other case. + */ +int st33zp24_probe(void *phy_id, const struct st33zp24_phy_ops *ops, + struct device *dev, int irq, int io_lpcpd) +{ + int ret; + u8 intmask = 0; + struct tpm_chip *chip; + struct st33zp24_dev *tpm_dev; + + chip = tpmm_chip_alloc(dev, &st33zp24_tpm); + if (IS_ERR(chip)) + return PTR_ERR(chip); + + tpm_dev = devm_kzalloc(dev, sizeof(struct st33zp24_dev), + GFP_KERNEL); + if (!tpm_dev) + return -ENOMEM; + + TPM_VPRIV(chip) = tpm_dev; + tpm_dev->phy_id = phy_id; + tpm_dev->ops = ops; + + chip->vendor.timeout_a = msecs_to_jiffies(TIS_SHORT_TIMEOUT); + chip->vendor.timeout_b = msecs_to_jiffies(TIS_LONG_TIMEOUT); + chip->vendor.timeout_c = msecs_to_jiffies(TIS_SHORT_TIMEOUT); + chip->vendor.timeout_d = msecs_to_jiffies(TIS_SHORT_TIMEOUT); + + chip->vendor.locality = LOCALITY0; + + if (irq) { + /* INTERRUPT Setup */ + init_waitqueue_head(&chip->vendor.read_queue); + tpm_dev->intrs = 0; + + if (request_locality(chip) != LOCALITY0) { + ret = -ENODEV; + goto _tpm_clean_answer; + } + + clear_interruption(tpm_dev); + ret = devm_request_irq(dev, irq, tpm_ioserirq_handler, + IRQF_TRIGGER_HIGH, "TPM SERIRQ management", + chip); + if (ret < 0) { + dev_err(&chip->dev, "TPM SERIRQ signals %d not available\n", + irq); + goto _tpm_clean_answer; + } + + intmask |= TPM_INTF_CMD_READY_INT + | TPM_INTF_STS_VALID_INT + | TPM_INTF_DATA_AVAIL_INT; + + ret = tpm_dev->ops->send(tpm_dev->phy_id, TPM_INT_ENABLE, + &intmask, 1); + if (ret < 0) + goto _tpm_clean_answer; + + intmask = TPM_GLOBAL_INT_ENABLE; + ret = tpm_dev->ops->send(tpm_dev->phy_id, (TPM_INT_ENABLE + 3), + &intmask, 1); + if (ret < 0) + goto _tpm_clean_answer; + + chip->vendor.irq = irq; + + disable_irq_nosync(chip->vendor.irq); + + tpm_gen_interrupt(chip); + } + + tpm_get_timeouts(chip); + tpm_do_selftest(chip); + + return tpm_chip_register(chip); +_tpm_clean_answer: + dev_info(&chip->dev, "TPM initialization fail\n"); + return ret; +} +EXPORT_SYMBOL(st33zp24_probe); + +/* + * st33zp24_remove remove the TPM device + * @param: tpm_data, the tpm phy. + * @return: 0 in case of success. + */ +int st33zp24_remove(struct tpm_chip *chip) +{ + tpm_chip_unregister(chip); + return 0; +} +EXPORT_SYMBOL(st33zp24_remove); + +#ifdef CONFIG_PM_SLEEP +/* + * st33zp24_pm_suspend suspend the TPM device + * @param: tpm_data, the tpm phy. + * @param: mesg, the power management message. + * @return: 0 in case of success. + */ +int st33zp24_pm_suspend(struct device *dev) +{ + struct tpm_chip *chip = dev_get_drvdata(dev); + struct st33zp24_dev *tpm_dev; + int ret = 0; + + tpm_dev = (struct st33zp24_dev *)TPM_VPRIV(chip); + + if (gpio_is_valid(tpm_dev->io_lpcpd)) + gpio_set_value(tpm_dev->io_lpcpd, 0); + else + ret = tpm_pm_suspend(dev); + + return ret; +} /* st33zp24_pm_suspend() */ +EXPORT_SYMBOL(st33zp24_pm_suspend); + +/* + * st33zp24_pm_resume resume the TPM device + * @param: tpm_data, the tpm phy. + * @return: 0 in case of success. + */ +int st33zp24_pm_resume(struct device *dev) +{ + struct tpm_chip *chip = dev_get_drvdata(dev); + struct st33zp24_dev *tpm_dev; + int ret = 0; + + tpm_dev = (struct st33zp24_dev *)TPM_VPRIV(chip); + + if (gpio_is_valid(tpm_dev->io_lpcpd)) { + gpio_set_value(tpm_dev->io_lpcpd, 1); + ret = wait_for_stat(chip, + TPM_STS_VALID, chip->vendor.timeout_b, + &chip->vendor.read_queue, false); + } else { + ret = tpm_pm_resume(dev); + if (!ret) + tpm_do_selftest(chip); + } + return ret; +} /* st33zp24_pm_resume() */ +EXPORT_SYMBOL(st33zp24_pm_resume); +#endif + +MODULE_AUTHOR("TPM support (TPMsupport@list.st.com)"); +MODULE_DESCRIPTION("ST33ZP24 TPM 1.2 driver"); +MODULE_VERSION("1.3.0"); +MODULE_LICENSE("GPL"); diff --git a/drivers/char/tpm/st33zp24/st33zp24.h b/drivers/char/tpm/st33zp24/st33zp24.h new file mode 100644 index 000000000000..c207cebf67dd --- /dev/null +++ b/drivers/char/tpm/st33zp24/st33zp24.h @@ -0,0 +1,37 @@ +/* + * STMicroelectronics TPM Linux driver for TPM ST33ZP24 + * Copyright (C) 2009 - 2015 STMicroelectronics + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * 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. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, see <http://www.gnu.org/licenses/>. + */ + +#ifndef __LOCAL_ST33ZP24_H__ +#define __LOCAL_ST33ZP24_H__ + +#define TPM_WRITE_DIRECTION 0x80 +#define TPM_BUFSIZE 2048 + +struct st33zp24_phy_ops { + int (*send)(void *phy_id, u8 tpm_register, u8 *tpm_data, int tpm_size); + int (*recv)(void *phy_id, u8 tpm_register, u8 *tpm_data, int tpm_size); +}; + +#ifdef CONFIG_PM_SLEEP +int st33zp24_pm_suspend(struct device *dev); +int st33zp24_pm_resume(struct device *dev); +#endif + +int st33zp24_probe(void *phy_id, const struct st33zp24_phy_ops *ops, + struct device *dev, int irq, int io_lpcpd); +int st33zp24_remove(struct tpm_chip *chip); +#endif /* __LOCAL_ST33ZP24_H__ */ diff --git a/drivers/char/tpm/tpm-chip.c b/drivers/char/tpm/tpm-chip.c index e096e9cddb40..283f00a7f036 100644 --- a/drivers/char/tpm/tpm-chip.c +++ b/drivers/char/tpm/tpm-chip.c @@ -170,6 +170,41 @@ static void tpm_dev_del_device(struct tpm_chip *chip) device_unregister(&chip->dev); } +static int tpm1_chip_register(struct tpm_chip *chip) +{ + int rc; + + if (chip->flags & TPM_CHIP_FLAG_TPM2) + return 0; + + rc = tpm_sysfs_add_device(chip); + if (rc) + return rc; + + rc = tpm_add_ppi(chip); + if (rc) { + tpm_sysfs_del_device(chip); + return rc; + } + + chip->bios_dir = tpm_bios_log_setup(chip->devname); + + return 0; +} + +static void tpm1_chip_unregister(struct tpm_chip *chip) +{ + if (chip->flags & TPM_CHIP_FLAG_TPM2) + return; + + if (chip->bios_dir) + tpm_bios_log_teardown(chip->bios_dir); + + tpm_remove_ppi(chip); + + tpm_sysfs_del_device(chip); +} + /* * tpm_chip_register() - create a character device for the TPM chip * @chip: TPM chip to use. @@ -185,22 +220,13 @@ int tpm_chip_register(struct tpm_chip *chip) { int rc; - /* Populate sysfs for TPM1 devices. */ - if (!(chip->flags & TPM_CHIP_FLAG_TPM2)) { - rc = tpm_sysfs_add_device(chip); - if (rc) - goto del_misc; - - rc = tpm_add_ppi(chip); - if (rc) - goto del_sysfs; - - chip->bios_dir = tpm_bios_log_setup(chip->devname); - } + rc = tpm1_chip_register(chip); + if (rc) + return rc; rc = tpm_dev_add_device(chip); if (rc) - return rc; + goto out_err; /* Make the chip available. */ spin_lock(&driver_lock); @@ -210,10 +236,8 @@ int tpm_chip_register(struct tpm_chip *chip) chip->flags |= TPM_CHIP_FLAG_REGISTERED; return 0; -del_sysfs: - tpm_sysfs_del_device(chip); -del_misc: - tpm_dev_del_device(chip); +out_err: + tpm1_chip_unregister(chip); return rc; } EXPORT_SYMBOL_GPL(tpm_chip_register); @@ -238,13 +262,7 @@ void tpm_chip_unregister(struct tpm_chip *chip) spin_unlock(&driver_lock); synchronize_rcu(); - if (!(chip->flags & TPM_CHIP_FLAG_TPM2)) { - if (chip->bios_dir) - tpm_bios_log_teardown(chip->bios_dir); - tpm_remove_ppi(chip); - tpm_sysfs_del_device(chip); - } - + tpm1_chip_unregister(chip); tpm_dev_del_device(chip); } EXPORT_SYMBOL_GPL(tpm_chip_unregister); diff --git a/drivers/char/tpm/tpm_i2c_stm_st33.c b/drivers/char/tpm/tpm_i2c_stm_st33.c deleted file mode 100644 index 612845b36c29..000000000000 --- a/drivers/char/tpm/tpm_i2c_stm_st33.c +++ /dev/null @@ -1,911 +0,0 @@ -/* - * STMicroelectronics TPM I2C Linux driver for TPM ST33ZP24 - * Copyright (C) 2009, 2010, 2014 STMicroelectronics - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License as published by - * the Free Software Foundation; either version 2 of the License, or - * (at your option) any later version. - * - * 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. - * - * You should have received a copy of the GNU General Public License - * along with this program; if not, see <http://www.gnu.org/licenses/>. - * - * STMicroelectronics version 1.2.1, Copyright (C) 2014 - * STMicroelectronics comes with ABSOLUTELY NO WARRANTY. - * This is free software, and you are welcome to redistribute it - * under certain conditions. - * - * @Author: Christophe RICARD tpmsupport@st.com - * - * @File: tpm_stm_st33_i2c.c - * - * @Synopsis: - * 09/15/2010: First shot driver tpm_tis driver for - * lpc is used as model. - */ - -#include <linux/pci.h> -#include <linux/module.h> -#include <linux/platform_device.h> -#include <linux/i2c.h> -#include <linux/fs.h> -#include <linux/miscdevice.h> -#include <linux/kernel.h> -#include <linux/delay.h> -#include <linux/wait.h> -#include <linux/freezer.h> -#include <linux/string.h> -#include <linux/interrupt.h> -#include <linux/sysfs.h> -#include <linux/gpio.h> -#include <linux/sched.h> -#include <linux/uaccess.h> -#include <linux/io.h> -#include <linux/slab.h> -#include <linux/of_irq.h> -#include <linux/of_gpio.h> - -#include <linux/platform_data/tpm_stm_st33.h> -#include "tpm.h" - -#define TPM_ACCESS 0x0 -#define TPM_STS 0x18 -#define TPM_HASH_END 0x20 -#define TPM_DATA_FIFO 0x24 -#define TPM_HASH_DATA 0x24 -#define TPM_HASH_START 0x28 -#define TPM_INTF_CAPABILITY 0x14 -#define TPM_INT_STATUS 0x10 -#define TPM_INT_ENABLE 0x08 - -#define TPM_DUMMY_BYTE 0xAA -#define TPM_WRITE_DIRECTION 0x80 -#define TPM_HEADER_SIZE 10 -#define TPM_BUFSIZE 2048 - -#define LOCALITY0 0 - - -enum stm33zp24_access { - TPM_ACCESS_VALID = 0x80, - TPM_ACCESS_ACTIVE_LOCALITY = 0x20, - TPM_ACCESS_REQUEST_PENDING = 0x04, - TPM_ACCESS_REQUEST_USE = 0x02, -}; - -enum stm33zp24_status { - TPM_STS_VALID = 0x80, - TPM_STS_COMMAND_READY = 0x40, - TPM_STS_GO = 0x20, - TPM_STS_DATA_AVAIL = 0x10, - TPM_STS_DATA_EXPECT = 0x08, -}; - -enum stm33zp24_int_flags { - TPM_GLOBAL_INT_ENABLE = 0x80, - TPM_INTF_CMD_READY_INT = 0x080, - TPM_INTF_FIFO_AVALAIBLE_INT = 0x040, - TPM_INTF_WAKE_UP_READY_INT = 0x020, - TPM_INTF_LOCALITY_CHANGE_INT = 0x004, - TPM_INTF_STS_VALID_INT = 0x002, - TPM_INTF_DATA_AVAIL_INT = 0x001, -}; - -enum tis_defaults { - TIS_SHORT_TIMEOUT = 750, - TIS_LONG_TIMEOUT = 2000, -}; - -struct tpm_stm_dev { - struct i2c_client *client; - struct tpm_chip *chip; - u8 buf[TPM_BUFSIZE + 1]; - u32 intrs; - int io_lpcpd; -}; - -/* - * write8_reg - * Send byte to the TIS register according to the ST33ZP24 I2C protocol. - * @param: tpm_register, the tpm tis register where the data should be written - * @param: tpm_data, the tpm_data to write inside the tpm_register - * @param: tpm_size, The length of the data - * @return: Returns negative errno, or else the number of bytes written. - */ -static int write8_reg(struct tpm_stm_dev *tpm_dev, u8 tpm_register, - u8 *tpm_data, u16 tpm_size) -{ - tpm_dev->buf[0] = tpm_register; - memcpy(tpm_dev->buf + 1, tpm_data, tpm_size); - return i2c_master_send(tpm_dev->client, tpm_dev->buf, tpm_size + 1); -} /* write8_reg() */ - -/* - * read8_reg - * Recv byte from the TIS register according to the ST33ZP24 I2C protocol. - * @param: tpm_register, the tpm tis register where the data should be read - * @param: tpm_data, the TPM response - * @param: tpm_size, tpm TPM response size to read. - * @return: number of byte read successfully: should be one if success. - */ -static int read8_reg(struct tpm_stm_dev *tpm_dev, u8 tpm_register, - u8 *tpm_data, int tpm_size) -{ - u8 status = 0; - u8 data; - - data = TPM_DUMMY_BYTE; - status = write8_reg(tpm_dev, tpm_register, &data, 1); - if (status == 2) - status = i2c_master_recv(tpm_dev->client, tpm_data, tpm_size); - return status; -} /* read8_reg() */ - -/* - * I2C_WRITE_DATA - * Send byte to the TIS register according to the ST33ZP24 I2C protocol. - * @param: tpm_dev, the chip description - * @param: tpm_register, the tpm tis register where the data should be written - * @param: tpm_data, the tpm_data to write inside the tpm_register - * @param: tpm_size, The length of the data - * @return: number of byte written successfully: should be one if success. - */ -#define I2C_WRITE_DATA(tpm_dev, tpm_register, tpm_data, tpm_size) \ - (write8_reg(tpm_dev, tpm_register | \ - TPM_WRITE_DIRECTION, tpm_data, tpm_size)) - -/* - * I2C_READ_DATA - * Recv byte from the TIS register according to the ST33ZP24 I2C protocol. - * @param: tpm_dev, the chip description - * @param: tpm_register, the tpm tis register where the data should be read - * @param: tpm_data, the TPM response - * @param: tpm_size, tpm TPM response size to read. - * @return: number of byte read successfully: should be one if success. - */ -#define I2C_READ_DATA(tpm_dev, tpm_register, tpm_data, tpm_size) \ - (read8_reg(tpm_dev, tpm_register, tpm_data, tpm_size)) - -/* - * clear_interruption - * clear the TPM interrupt register. - * @param: tpm, the chip description - * @return: the TPM_INT_STATUS value - */ -static u8 clear_interruption(struct tpm_stm_dev *tpm_dev) -{ - u8 interrupt; - - I2C_READ_DATA(tpm_dev, TPM_INT_STATUS, &interrupt, 1); - I2C_WRITE_DATA(tpm_dev, TPM_INT_STATUS, &interrupt, 1); - return interrupt; -} /* clear_interruption() */ - -/* - * tpm_stm_i2c_cancel, cancel is not implemented. - * @param: chip, the tpm_chip description as specified in driver/char/tpm/tpm.h - */ -static void tpm_stm_i2c_cancel(struct tpm_chip *chip) -{ - struct tpm_stm_dev *tpm_dev; - u8 data; - - tpm_dev = (struct tpm_stm_dev *)TPM_VPRIV(chip); - - data = TPM_STS_COMMAND_READY; - I2C_WRITE_DATA(tpm_dev, TPM_STS, &data, 1); -} /* tpm_stm_i2c_cancel() */ - -/* - * tpm_stm_spi_status return the TPM_STS register - * @param: chip, the tpm chip description - * @return: the TPM_STS register value. - */ -static u8 tpm_stm_i2c_status(struct tpm_chip *chip) -{ - struct tpm_stm_dev *tpm_dev; - u8 data; - - tpm_dev = (struct tpm_stm_dev *)TPM_VPRIV(chip); - - I2C_READ_DATA(tpm_dev, TPM_STS, &data, 1); - return data; -} /* tpm_stm_i2c_status() */ - - -/* - * check_locality if the locality is active - * @param: chip, the tpm chip description - * @return: the active locality or -EACCESS. - */ -static int check_locality(struct tpm_chip *chip) -{ - struct tpm_stm_dev *tpm_dev; - u8 data; - u8 status; - - tpm_dev = (struct tpm_stm_dev *)TPM_VPRIV(chip); - - status = I2C_READ_DATA(tpm_dev, TPM_ACCESS, &data, 1); - if (status && (data & - (TPM_ACCESS_ACTIVE_LOCALITY | TPM_ACCESS_VALID)) == - (TPM_ACCESS_ACTIVE_LOCALITY | TPM_ACCESS_VALID)) - return chip->vendor.locality; - - return -EACCES; -} /* check_locality() */ - -/* - * request_locality request the TPM locality - * @param: chip, the chip description - * @return: the active locality or EACCESS. - */ -static int request_locality(struct tpm_chip *chip) -{ - unsigned long stop; - long ret; - struct tpm_stm_dev *tpm_dev; - u8 data; - - if (check_locality(chip) == chip->vendor.locality) - return chip->vendor.locality; - - tpm_dev = (struct tpm_stm_dev *)TPM_VPRIV(chip); - - data = TPM_ACCESS_REQUEST_USE; - ret = I2C_WRITE_DATA(tpm_dev, TPM_ACCESS, &data, 1); - if (ret < 0) - goto end; - - stop = jiffies + chip->vendor.timeout_a; - - /* Request locality is usually effective after the request */ - do { - if (check_locality(chip) >= 0) - return chip->vendor.locality; - msleep(TPM_TIMEOUT); - } while (time_before(jiffies, stop)); - ret = -EACCES; -end: - return ret; -} /* request_locality() */ - -/* - * release_locality release the active locality - * @param: chip, the tpm chip description. - */ -static void release_locality(struct tpm_chip *chip) -{ - struct tpm_stm_dev *tpm_dev; - u8 data; - - tpm_dev = (struct tpm_stm_dev *)TPM_VPRIV(chip); - data = TPM_ACCESS_ACTIVE_LOCALITY; - - I2C_WRITE_DATA(tpm_dev, TPM_ACCESS, &data, 1); -} - -/* - * get_burstcount return the burstcount address 0x19 0x1A - * @param: chip, the chip description - * return: the burstcount. - */ -static int get_burstcount(struct tpm_chip *chip) -{ - unsigned long stop; - int burstcnt, status; - u8 tpm_reg, temp; - struct tpm_stm_dev *tpm_dev; - - tpm_dev = (struct tpm_stm_dev *)TPM_VPRIV(chip); - - stop = jiffies + chip->vendor.timeout_d; - do { - tpm_reg = TPM_STS + 1; - status = I2C_READ_DATA(tpm_dev, tpm_reg, &temp, 1); - if (status < 0) - goto end; - - tpm_reg = tpm_reg + 1; - burstcnt = temp; - status = I2C_READ_DATA(tpm_dev, tpm_reg, &temp, 1); - if (status < 0) - goto end; - - burstcnt |= temp << 8; - if (burstcnt) - return burstcnt; - msleep(TPM_TIMEOUT); - } while (time_before(jiffies, stop)); - -end: - return -EBUSY; -} /* get_burstcount() */ - -static bool wait_for_tpm_stat_cond(struct tpm_chip *chip, u8 mask, - bool check_cancel, bool *canceled) -{ - u8 status = chip->ops->status(chip); - - *canceled = false; - if ((status & mask) == mask) - return true; - if (check_cancel && chip->ops->req_canceled(chip, status)) { - *canceled = true; - return true; - } - return false; -} - -/* - * interrupt_to_status - * @param: irq_mask, the irq mask value to wait - * @return: the corresponding tpm_sts value - */ -static u8 interrupt_to_status(u8 irq_mask) -{ - u8 status = 0; - - if ((irq_mask & TPM_INTF_STS_VALID_INT) == TPM_INTF_STS_VALID_INT) - status |= TPM_STS_VALID; - if ((irq_mask & TPM_INTF_DATA_AVAIL_INT) == TPM_INTF_DATA_AVAIL_INT) - status |= TPM_STS_DATA_AVAIL; - if ((irq_mask & TPM_INTF_CMD_READY_INT) == TPM_INTF_CMD_READY_INT) - status |= TPM_STS_COMMAND_READY; - - return status; -} /* status_to_interrupt() */ - -/* - * wait_for_stat wait for a TPM_STS value - * @param: chip, the tpm chip description - * @param: mask, the value mask to wait - * @param: timeout, the timeout - * @param: queue, the wait queue. - * @param: check_cancel, does the command can be cancelled ? - * @return: the tpm status, 0 if success, -ETIME if timeout is reached. - */ -static int wait_for_stat(struct tpm_chip *chip, u8 mask, unsigned long timeout, - wait_queue_head_t *queue, bool check_cancel) -{ - unsigned long stop; - int ret; - bool canceled = false; - bool condition; - u32 cur_intrs; - u8 interrupt, status; - struct tpm_stm_dev *tpm_dev; - - tpm_dev = (struct tpm_stm_dev *)TPM_VPRIV(chip); - - /* check current status */ - status = tpm_stm_i2c_status(chip); - if ((status & mask) == mask) - return 0; - - stop = jiffies + timeout; - - if (chip->vendor.irq) { - cur_intrs = tpm_dev->intrs; - interrupt = clear_interruption(tpm_dev); - enable_irq(chip->vendor.irq); - -again: - timeout = stop - jiffies; - if ((long) timeout <= 0) - return -1; - - ret = wait_event_interruptible_timeout(*queue, - cur_intrs != tpm_dev->intrs, timeout); - - interrupt |= clear_interruption(tpm_dev); - status = interrupt_to_status(interrupt); - condition = wait_for_tpm_stat_cond(chip, mask, - check_cancel, &canceled); - - if (ret >= 0 && condition) { - if (canceled) - return -ECANCELED; - return 0; - } - if (ret == -ERESTARTSYS && freezing(current)) { - clear_thread_flag(TIF_SIGPENDING); - goto again; - } - disable_irq_nosync(chip->vendor.irq); - - } else { - do { - msleep(TPM_TIMEOUT); - status = chip->ops->status(chip); - if ((status & mask) == mask) - return 0; - } while (time_before(jiffies, stop)); - } - - return -ETIME; -} /* wait_for_stat() */ - -/* - * recv_data receive data - * @param: chip, the tpm chip description - * @param: buf, the buffer where the data are received - * @param: count, the number of data to receive - * @return: the number of bytes read from TPM FIFO. - */ -static int recv_data(struct tpm_chip *chip, u8 *buf, size_t count) -{ - int size = 0, burstcnt, len, ret; - struct tpm_stm_dev *tpm_dev; - - tpm_dev = (struct tpm_stm_dev *)TPM_VPRIV(chip); - - while (size < count && - wait_for_stat(chip, - TPM_STS_DATA_AVAIL | TPM_STS_VALID, - chip->vendor.timeout_c, - &chip->vendor.read_queue, true) == 0) { - burstcnt = get_burstcount(chip); - if (burstcnt < 0) - return burstcnt; - len = min_t(int, burstcnt, count - size); - ret = I2C_READ_DATA(tpm_dev, TPM_DATA_FIFO, buf + size, len); - if (ret < 0) - return ret; - - size += len; - } - return size; -} - -/* - * tpm_ioserirq_handler the serirq irq handler - * @param: irq, the tpm chip description - * @param: dev_id, the description of the chip - * @return: the status of the handler. - */ -static irqreturn_t tpm_ioserirq_handler(int irq, void *dev_id) -{ - struct tpm_chip *chip = dev_id; - struct tpm_stm_dev *tpm_dev; - - tpm_dev = (struct tpm_stm_dev *)TPM_VPRIV(chip); - - tpm_dev->intrs++; - wake_up_interruptible(&chip->vendor.read_queue); - disable_irq_nosync(chip->vendor.irq); - - return IRQ_HANDLED; -} /* tpm_ioserirq_handler() */ - - -/* - * tpm_stm_i2c_send send TPM commands through the I2C bus. - * - * @param: chip, the tpm_chip description as specified in driver/char/tpm/tpm.h - * @param: buf, the buffer to send. - * @param: count, the number of bytes to send. - * @return: In case of success the number of bytes sent. - * In other case, a < 0 value describing the issue. - */ -static int tpm_stm_i2c_send(struct tpm_chip *chip, unsigned char *buf, - size_t len) -{ - u32 status, i, size; - int burstcnt = 0; - int ret; - u8 data; - struct i2c_client *client; - struct tpm_stm_dev *tpm_dev; - - if (!chip) - return -EBUSY; - if (len < TPM_HEADER_SIZE) - return -EBUSY; - - tpm_dev = (struct tpm_stm_dev *)TPM_VPRIV(chip); - client = tpm_dev->client; - - client->flags = 0; - - ret = request_locality(chip); - if (ret < 0) - return ret; - - status = tpm_stm_i2c_status(chip); - if ((status & TPM_STS_COMMAND_READY) == 0) { - tpm_stm_i2c_cancel(chip); - if (wait_for_stat - (chip, TPM_STS_COMMAND_READY, chip->vendor.timeout_b, - &chip->vendor.read_queue, false) < 0) { - ret = -ETIME; - goto out_err; - } - } - - for (i = 0; i < len - 1;) { - burstcnt = get_burstcount(chip); - if (burstcnt < 0) - return burstcnt; - size = min_t(int, len - i - 1, burstcnt); - ret = I2C_WRITE_DATA(tpm_dev, TPM_DATA_FIFO, buf + i, size); - if (ret < 0) - goto out_err; - - i += size; - } - - status = tpm_stm_i2c_status(chip); - if ((status & TPM_STS_DATA_EXPECT) == 0) { - ret = -EIO; - goto out_err; - } - - ret = I2C_WRITE_DATA(tpm_dev, TPM_DATA_FIFO, buf + len - 1, 1); - if (ret < 0) - goto out_err; - - status = tpm_stm_i2c_status(chip); - if ((status & TPM_STS_DATA_EXPECT) != 0) { - ret = -EIO; - goto out_err; - } - - data = TPM_STS_GO; - I2C_WRITE_DATA(tpm_dev, TPM_STS, &data, 1); - - return len; -out_err: - tpm_stm_i2c_cancel(chip); - release_locality(chip); - return ret; -} - -/* - * tpm_stm_i2c_recv received TPM response through the I2C bus. - * @param: chip, the tpm_chip description as specified in driver/char/tpm/tpm.h. - * @param: buf, the buffer to store datas. - * @param: count, the number of bytes to send. - * @return: In case of success the number of bytes received. - * In other case, a < 0 value describing the issue. - */ -static int tpm_stm_i2c_recv(struct tpm_chip *chip, unsigned char *buf, - size_t count) -{ - int size = 0; - int expected; - - if (!chip) - return -EBUSY; - - if (count < TPM_HEADER_SIZE) { - size = -EIO; - goto out; - } - - size = recv_data(chip, buf, TPM_HEADER_SIZE); - if (size < TPM_HEADER_SIZE) { - dev_err(chip->pdev, "Unable to read header\n"); - goto out; - } - - expected = be32_to_cpu(*(__be32 *)(buf + 2)); - if (expected > count) { - size = -EIO; - goto out; - } - - size += recv_data(chip, &buf[TPM_HEADER_SIZE], - expected - TPM_HEADER_SIZE); - if (size < expected) { - dev_err(chip->pdev, "Unable to read remainder of result\n"); - size = -ETIME; - goto out; - } - -out: - chip->ops->cancel(chip); - release_locality(chip); - return size; -} - -static bool tpm_stm_i2c_req_canceled(struct tpm_chip *chip, u8 status) -{ - return (status == TPM_STS_COMMAND_READY); -} - -static const struct tpm_class_ops st_i2c_tpm = { - .send = tpm_stm_i2c_send, - .recv = tpm_stm_i2c_recv, - .cancel = tpm_stm_i2c_cancel, - .status = tpm_stm_i2c_status, - .req_complete_mask = TPM_STS_DATA_AVAIL | TPM_STS_VALID, - .req_complete_val = TPM_STS_DATA_AVAIL | TPM_STS_VALID, - .req_canceled = tpm_stm_i2c_req_canceled, -}; - -#ifdef CONFIG_OF -static int tpm_stm_i2c_of_request_resources(struct tpm_chip *chip) -{ - struct device_node *pp; - struct tpm_stm_dev *tpm_dev = (struct tpm_stm_dev *)TPM_VPRIV(chip); - struct i2c_client *client = tpm_dev->client; - int gpio; - int ret; - - pp = client->dev.of_node; - if (!pp) { - dev_err(chip->pdev, "No platform data\n"); - return -ENODEV; - } - - /* Get GPIO from device tree */ - gpio = of_get_named_gpio(pp, "lpcpd-gpios", 0); - if (gpio < 0) { - dev_err(chip->pdev, "Failed to retrieve lpcpd-gpios from dts.\n"); - tpm_dev->io_lpcpd = -1; - /* - * lpcpd pin is not specified. This is not an issue as - * power management can be also managed by TPM specific - * commands. So leave with a success status code. - */ - return 0; - } - /* GPIO request and configuration */ - ret = devm_gpio_request_one(&client->dev, gpio, - GPIOF_OUT_INIT_HIGH, "TPM IO LPCPD"); - if (ret) { - dev_err(chip->pdev, "Failed to request lpcpd pin\n"); - return -ENODEV; - } - tpm_dev->io_lpcpd = gpio; - - return 0; -} -#else -static int tpm_stm_i2c_of_request_resources(struct tpm_chip *chip) -{ - return -ENODEV; -} -#endif - -static int tpm_stm_i2c_request_resources(struct i2c_client *client, - struct tpm_chip *chip) -{ - struct st33zp24_platform_data *pdata; - struct tpm_stm_dev *tpm_dev = (struct tpm_stm_dev *)TPM_VPRIV(chip); - int ret; - - pdata = client->dev.platform_data; - if (!pdata) { - dev_err(chip->pdev, "No platform data\n"); - return -ENODEV; - } - - /* store for late use */ - tpm_dev->io_lpcpd = pdata->io_lpcpd; - - if (gpio_is_valid(pdata->io_lpcpd)) { - ret = devm_gpio_request_one(&client->dev, - pdata->io_lpcpd, GPIOF_OUT_INIT_HIGH, - "TPM IO_LPCPD"); - if (ret) { - dev_err(chip->pdev, "%s : reset gpio_request failed\n", - __FILE__); - return ret; - } - } - - return 0; -} - -/* - * tpm_stm_i2c_probe initialize the TPM device - * @param: client, the i2c_client drescription (TPM I2C description). - * @param: id, the i2c_device_id struct. - * @return: 0 in case of success. - * -1 in other case. - */ -static int -tpm_stm_i2c_probe(struct i2c_client *client, const struct i2c_device_id *id) -{ - int ret; - u8 intmask = 0; - struct tpm_chip *chip; - struct st33zp24_platform_data *platform_data; - struct tpm_stm_dev *tpm_dev; - - if (!client) { - pr_info("%s: i2c client is NULL. Device not accessible.\n", - __func__); - return -ENODEV; - } - - if (!i2c_check_functionality(client->adapter, I2C_FUNC_I2C)) { - dev_info(&client->dev, "client not i2c capable\n"); - return -ENODEV; - } - - tpm_dev = devm_kzalloc(&client->dev, sizeof(struct tpm_stm_dev), - GFP_KERNEL); - if (!tpm_dev) - return -ENOMEM; - - chip = tpmm_chip_alloc(&client->dev, &st_i2c_tpm); - if (IS_ERR(chip)) - return PTR_ERR(chip); - - TPM_VPRIV(chip) = tpm_dev; - tpm_dev->client = client; - - platform_data = client->dev.platform_data; - if (!platform_data && client->dev.of_node) { - ret = tpm_stm_i2c_of_request_resources(chip); - if (ret) - goto _tpm_clean_answer; - } else if (platform_data) { - ret = tpm_stm_i2c_request_resources(client, chip); - if (ret) - goto _tpm_clean_answer; - } - - chip->vendor.timeout_a = msecs_to_jiffies(TIS_SHORT_TIMEOUT); - chip->vendor.timeout_b = msecs_to_jiffies(TIS_LONG_TIMEOUT); - chip->vendor.timeout_c = msecs_to_jiffies(TIS_SHORT_TIMEOUT); - chip->vendor.timeout_d = msecs_to_jiffies(TIS_SHORT_TIMEOUT); - - chip->vendor.locality = LOCALITY0; - - if (client->irq) { - /* INTERRUPT Setup */ - init_waitqueue_head(&chip->vendor.read_queue); - tpm_dev->intrs = 0; - - if (request_locality(chip) != LOCALITY0) { - ret = -ENODEV; - goto _tpm_clean_answer; - } - - clear_interruption(tpm_dev); - ret = devm_request_irq(&client->dev, client->irq, - tpm_ioserirq_handler, - IRQF_TRIGGER_HIGH, - "TPM SERIRQ management", chip); - if (ret < 0) { - dev_err(chip->pdev, "TPM SERIRQ signals %d not available\n", - client->irq); - goto _tpm_clean_answer; - } - - intmask |= TPM_INTF_CMD_READY_INT - | TPM_INTF_STS_VALID_INT - | TPM_INTF_DATA_AVAIL_INT; - - ret = I2C_WRITE_DATA(tpm_dev, TPM_INT_ENABLE, &intmask, 1); - if (ret < 0) - goto _tpm_clean_answer; - - intmask = TPM_GLOBAL_INT_ENABLE; - ret = I2C_WRITE_DATA(tpm_dev, (TPM_INT_ENABLE + 3), - &intmask, 1); - if (ret < 0) - goto _tpm_clean_answer; - - chip->vendor.irq = client->irq; - - disable_irq_nosync(chip->vendor.irq); - - tpm_gen_interrupt(chip); - } - - tpm_get_timeouts(chip); - tpm_do_selftest(chip); - - return tpm_chip_register(chip); -_tpm_clean_answer: - dev_info(chip->pdev, "TPM I2C initialisation fail\n"); - return ret; -} - -/* - * tpm_stm_i2c_remove remove the TPM device - * @param: client, the i2c_client description (TPM I2C description). - * @return: 0 in case of success. - */ -static int tpm_stm_i2c_remove(struct i2c_client *client) -{ - struct tpm_chip *chip = - (struct tpm_chip *) i2c_get_clientdata(client); - - if (chip) - tpm_chip_unregister(chip); - - return 0; -} - -#ifdef CONFIG_PM_SLEEP -/* - * tpm_stm_i2c_pm_suspend suspend the TPM device - * @param: client, the i2c_client drescription (TPM I2C description). - * @param: mesg, the power management message. - * @return: 0 in case of success. - */ -static int tpm_stm_i2c_pm_suspend(struct device *dev) -{ - struct st33zp24_platform_data *pin_infos = dev->platform_data; - int ret = 0; - - if (gpio_is_valid(pin_infos->io_lpcpd)) - gpio_set_value(pin_infos->io_lpcpd, 0); - else - ret = tpm_pm_suspend(dev); - - return ret; -} /* tpm_stm_i2c_suspend() */ - -/* - * tpm_stm_i2c_pm_resume resume the TPM device - * @param: client, the i2c_client drescription (TPM I2C description). - * @return: 0 in case of success. - */ -static int tpm_stm_i2c_pm_resume(struct device *dev) -{ - struct tpm_chip *chip = dev_get_drvdata(dev); - struct st33zp24_platform_data *pin_infos = dev->platform_data; - - int ret = 0; - - if (gpio_is_valid(pin_infos->io_lpcpd)) { - gpio_set_value(pin_infos->io_lpcpd, 1); - ret = wait_for_stat(chip, - TPM_STS_VALID, chip->vendor.timeout_b, - &chip->vendor.read_queue, false); - } else { - ret = tpm_pm_resume(dev); - if (!ret) - tpm_do_selftest(chip); - } - return ret; -} /* tpm_stm_i2c_pm_resume() */ -#endif - -static const struct i2c_device_id tpm_stm_i2c_id[] = { - {TPM_ST33_I2C, 0}, - {} -}; -MODULE_DEVICE_TABLE(i2c, tpm_stm_i2c_id); - -#ifdef CONFIG_OF -static const struct of_device_id of_st33zp24_i2c_match[] = { - { .compatible = "st,st33zp24-i2c", }, - {} -}; -MODULE_DEVICE_TABLE(of, of_st33zp24_i2c_match); -#endif - -static SIMPLE_DEV_PM_OPS(tpm_stm_i2c_ops, tpm_stm_i2c_pm_suspend, - tpm_stm_i2c_pm_resume); - -static struct i2c_driver tpm_stm_i2c_driver = { - .driver = { - .owner = THIS_MODULE, - .name = TPM_ST33_I2C, - .pm = &tpm_stm_i2c_ops, - .of_match_table = of_match_ptr(of_st33zp24_i2c_match), - }, - .probe = tpm_stm_i2c_probe, - .remove = tpm_stm_i2c_remove, - .id_table = tpm_stm_i2c_id -}; - -module_i2c_driver(tpm_stm_i2c_driver); - -MODULE_AUTHOR("Christophe Ricard (tpmsupport@st.com)"); -MODULE_DESCRIPTION("STM TPM I2C ST33 Driver"); -MODULE_VERSION("1.2.1"); -MODULE_LICENSE("GPL"); diff --git a/drivers/char/tpm/tpm_infineon.c b/drivers/char/tpm/tpm_infineon.c index 29ba520ac24d..6c488e635fdd 100644 --- a/drivers/char/tpm/tpm_infineon.c +++ b/drivers/char/tpm/tpm_infineon.c @@ -591,27 +591,8 @@ static void tpm_inf_pnp_remove(struct pnp_dev *dev) } } -static int tpm_inf_pnp_suspend(struct pnp_dev *dev, pm_message_t pm_state) -{ - struct tpm_chip *chip = pnp_get_drvdata(dev); - int rc; - if (chip) { - u8 savestate[] = { - 0, 193, /* TPM_TAG_RQU_COMMAND */ - 0, 0, 0, 10, /* blob length (in bytes) */ - 0, 0, 0, 152 /* TPM_ORD_SaveState */ - }; - dev_info(&dev->dev, "saving TPM state\n"); - rc = tpm_inf_send(chip, savestate, sizeof(savestate)); - if (rc < 0) { - dev_err(&dev->dev, "error while saving TPM state\n"); - return rc; - } - } - return 0; -} - -static int tpm_inf_pnp_resume(struct pnp_dev *dev) +#ifdef CONFIG_PM_SLEEP +static int tpm_inf_resume(struct device *dev) { /* Re-configure TPM after suspending */ tpm_config_out(ENABLE_REGISTER_PAIR, TPM_INF_ADDR); @@ -625,16 +606,19 @@ static int tpm_inf_pnp_resume(struct pnp_dev *dev) tpm_config_out(DISABLE_REGISTER_PAIR, TPM_INF_ADDR); /* disable RESET, LP and IRQC */ tpm_data_out(RESET_LP_IRQC_DISABLE, CMD); - return tpm_pm_resume(&dev->dev); + return tpm_pm_resume(dev); } +#endif +static SIMPLE_DEV_PM_OPS(tpm_inf_pm, tpm_pm_suspend, tpm_inf_resume); static struct pnp_driver tpm_inf_pnp_driver = { .name = "tpm_inf_pnp", .id_table = tpm_inf_pnp_tbl, .probe = tpm_inf_pnp_probe, - .suspend = tpm_inf_pnp_suspend, - .resume = tpm_inf_pnp_resume, - .remove = tpm_inf_pnp_remove + .remove = tpm_inf_pnp_remove, + .driver = { + .pm = &tpm_inf_pm, + } }; module_pnp_driver(tpm_inf_pnp_driver); diff --git a/drivers/crypto/Kconfig b/drivers/crypto/Kconfig index 2fb0fdfc87df..800bf41718e1 100644 --- a/drivers/crypto/Kconfig +++ b/drivers/crypto/Kconfig @@ -391,7 +391,7 @@ config CRYPTO_DEV_ATMEL_SHA config CRYPTO_DEV_CCP bool "Support for AMD Cryptographic Coprocessor" - depends on (X86 && PCI) || ARM64 + depends on ((X86 && PCI) || (ARM64 && (OF_ADDRESS || ACPI))) && HAS_IOMEM default n help The AMD Cryptographic Coprocessor provides hardware support @@ -436,4 +436,26 @@ config CRYPTO_DEV_QCE hardware. To compile this driver as a module, choose M here. The module will be called qcrypto. +config CRYPTO_DEV_VMX + bool "Support for VMX cryptographic acceleration instructions" + depends on PPC64 + default n + help + Support for VMX cryptographic acceleration instructions. + +source "drivers/crypto/vmx/Kconfig" + +config CRYPTO_DEV_IMGTEC_HASH + depends on MIPS || COMPILE_TEST + tristate "Imagination Technologies hardware hash accelerator" + select CRYPTO_ALGAPI + select CRYPTO_MD5 + select CRYPTO_SHA1 + select CRYPTO_SHA256 + select CRYPTO_HASH + help + This driver interfaces with the Imagination Technologies + hardware hash accelerator. Supporting MD5/SHA1/SHA224/SHA256 + hashing algorithms. + endif # CRYPTO_HW diff --git a/drivers/crypto/Makefile b/drivers/crypto/Makefile index 3924f93d5774..fb84be7e6be5 100644 --- a/drivers/crypto/Makefile +++ b/drivers/crypto/Makefile @@ -6,6 +6,7 @@ obj-$(CONFIG_CRYPTO_DEV_CCP) += ccp/ obj-$(CONFIG_CRYPTO_DEV_FSL_CAAM) += caam/ obj-$(CONFIG_CRYPTO_DEV_GEODE) += geode-aes.o obj-$(CONFIG_CRYPTO_DEV_HIFN_795X) += hifn_795x.o +obj-$(CONFIG_CRYPTO_DEV_IMGTEC_HASH) += img-hash.o obj-$(CONFIG_CRYPTO_DEV_IXP4XX) += ixp4xx_crypto.o obj-$(CONFIG_CRYPTO_DEV_MV_CESA) += mv_cesa.o obj-$(CONFIG_CRYPTO_DEV_MXS_DCP) += mxs-dcp.o @@ -25,3 +26,4 @@ obj-$(CONFIG_CRYPTO_DEV_TALITOS) += talitos.o obj-$(CONFIG_CRYPTO_DEV_UX500) += ux500/ obj-$(CONFIG_CRYPTO_DEV_QAT) += qat/ obj-$(CONFIG_CRYPTO_DEV_QCE) += qce/ +obj-$(CONFIG_CRYPTO_DEV_VMX) += vmx/ diff --git a/drivers/crypto/amcc/crypto4xx_core.c b/drivers/crypto/amcc/crypto4xx_core.c index d02b77150070..3b28e8c3de28 100644 --- a/drivers/crypto/amcc/crypto4xx_core.c +++ b/drivers/crypto/amcc/crypto4xx_core.c @@ -1155,7 +1155,7 @@ struct crypto4xx_alg_common crypto4xx_alg[] = { /** * Module Initialization Routine */ -static int __init crypto4xx_probe(struct platform_device *ofdev) +static int crypto4xx_probe(struct platform_device *ofdev) { int rc; struct resource res; @@ -1263,7 +1263,7 @@ err_alloc_dev: return rc; } -static int __exit crypto4xx_remove(struct platform_device *ofdev) +static int crypto4xx_remove(struct platform_device *ofdev) { struct device *dev = &ofdev->dev; struct crypto4xx_core_device *core_dev = dev_get_drvdata(dev); @@ -1291,7 +1291,7 @@ static struct platform_driver crypto4xx_driver = { .of_match_table = crypto4xx_match, }, .probe = crypto4xx_probe, - .remove = __exit_p(crypto4xx_remove), + .remove = crypto4xx_remove, }; module_platform_driver(crypto4xx_driver); diff --git a/drivers/crypto/atmel-aes.c b/drivers/crypto/atmel-aes.c index 6597aac9905d..0f9a9dc06a83 100644 --- a/drivers/crypto/atmel-aes.c +++ b/drivers/crypto/atmel-aes.c @@ -315,10 +315,10 @@ static int atmel_aes_crypt_dma(struct atmel_aes_dev *dd, dd->dma_size = length; - if (!(dd->flags & AES_FLAGS_FAST)) { - dma_sync_single_for_device(dd->dev, dma_addr_in, length, - DMA_TO_DEVICE); - } + dma_sync_single_for_device(dd->dev, dma_addr_in, length, + DMA_TO_DEVICE); + dma_sync_single_for_device(dd->dev, dma_addr_out, length, + DMA_FROM_DEVICE); if (dd->flags & AES_FLAGS_CFB8) { dd->dma_lch_in.dma_conf.dst_addr_width = @@ -391,6 +391,11 @@ static int atmel_aes_crypt_cpu_start(struct atmel_aes_dev *dd) { dd->flags &= ~AES_FLAGS_DMA; + dma_sync_single_for_cpu(dd->dev, dd->dma_addr_in, + dd->dma_size, DMA_TO_DEVICE); + dma_sync_single_for_cpu(dd->dev, dd->dma_addr_out, + dd->dma_size, DMA_FROM_DEVICE); + /* use cache buffers */ dd->nb_in_sg = atmel_aes_sg_length(dd->req, dd->in_sg); if (!dd->nb_in_sg) @@ -459,6 +464,9 @@ static int atmel_aes_crypt_dma_start(struct atmel_aes_dev *dd) dd->flags |= AES_FLAGS_FAST; } else { + dma_sync_single_for_cpu(dd->dev, dd->dma_addr_in, + dd->dma_size, DMA_TO_DEVICE); + /* use cache buffers */ count = atmel_aes_sg_copy(&dd->in_sg, &dd->in_offset, dd->buf_in, dd->buflen, dd->total, 0); @@ -619,7 +627,7 @@ static int atmel_aes_crypt_dma_stop(struct atmel_aes_dev *dd) dma_unmap_sg(dd->dev, dd->out_sg, 1, DMA_FROM_DEVICE); dma_unmap_sg(dd->dev, dd->in_sg, 1, DMA_TO_DEVICE); } else { - dma_sync_single_for_device(dd->dev, dd->dma_addr_out, + dma_sync_single_for_cpu(dd->dev, dd->dma_addr_out, dd->dma_size, DMA_FROM_DEVICE); /* copy data */ @@ -1246,6 +1254,11 @@ static void atmel_aes_get_cap(struct atmel_aes_dev *dd) /* keep only major version number */ switch (dd->hw_version & 0xff0) { + case 0x200: + dd->caps.has_dualbuff = 1; + dd->caps.has_cfb64 = 1; + dd->caps.max_burst_size = 4; + break; case 0x130: dd->caps.has_dualbuff = 1; dd->caps.has_cfb64 = 1; @@ -1336,6 +1349,7 @@ static int atmel_aes_probe(struct platform_device *pdev) platform_set_drvdata(pdev, aes_dd); INIT_LIST_HEAD(&aes_dd->list); + spin_lock_init(&aes_dd->lock); tasklet_init(&aes_dd->done_task, atmel_aes_done_task, (unsigned long)aes_dd); @@ -1374,7 +1388,7 @@ static int atmel_aes_probe(struct platform_device *pdev) /* Initializing the clock */ aes_dd->iclk = clk_get(&pdev->dev, "aes_clk"); if (IS_ERR(aes_dd->iclk)) { - dev_err(dev, "clock intialization failed.\n"); + dev_err(dev, "clock initialization failed.\n"); err = PTR_ERR(aes_dd->iclk); goto clk_err; } diff --git a/drivers/crypto/atmel-sha.c b/drivers/crypto/atmel-sha.c index 34db04addc18..5b35433c5399 100644 --- a/drivers/crypto/atmel-sha.c +++ b/drivers/crypto/atmel-sha.c @@ -163,8 +163,20 @@ static size_t atmel_sha_append_sg(struct atmel_sha_reqctx *ctx) count = min(ctx->sg->length - ctx->offset, ctx->total); count = min(count, ctx->buflen - ctx->bufcnt); - if (count <= 0) - break; + if (count <= 0) { + /* + * Check if count <= 0 because the buffer is full or + * because the sg length is 0. In the latest case, + * check if there is another sg in the list, a 0 length + * sg doesn't necessarily mean the end of the sg list. + */ + if ((ctx->sg->length == 0) && !sg_is_last(ctx->sg)) { + ctx->sg = sg_next(ctx->sg); + continue; + } else { + break; + } + } scatterwalk_map_and_copy(ctx->buffer + ctx->bufcnt, ctx->sg, ctx->offset, count, 0); @@ -420,14 +432,8 @@ static int atmel_sha_xmit_dma(struct atmel_sha_dev *dd, dma_addr_t dma_addr1, dev_dbg(dd->dev, "xmit_dma: digcnt: 0x%llx 0x%llx, length: %d, final: %d\n", ctx->digcnt[1], ctx->digcnt[0], length1, final); - if (ctx->flags & (SHA_FLAGS_SHA1 | SHA_FLAGS_SHA224 | - SHA_FLAGS_SHA256)) { - dd->dma_lch_in.dma_conf.src_maxburst = 16; - dd->dma_lch_in.dma_conf.dst_maxburst = 16; - } else { - dd->dma_lch_in.dma_conf.src_maxburst = 32; - dd->dma_lch_in.dma_conf.dst_maxburst = 32; - } + dd->dma_lch_in.dma_conf.src_maxburst = 16; + dd->dma_lch_in.dma_conf.dst_maxburst = 16; dmaengine_slave_config(dd->dma_lch_in.chan, &dd->dma_lch_in.dma_conf); @@ -529,7 +535,7 @@ static int atmel_sha_update_dma_slow(struct atmel_sha_dev *dd) if (final) atmel_sha_fill_padding(ctx, 0); - if (final || (ctx->bufcnt == ctx->buflen && ctx->total)) { + if (final || (ctx->bufcnt == ctx->buflen)) { count = ctx->bufcnt; ctx->bufcnt = 0; return atmel_sha_xmit_dma_map(dd, ctx, count, final); @@ -1266,6 +1272,12 @@ static void atmel_sha_get_cap(struct atmel_sha_dev *dd) /* keep only major version number */ switch (dd->hw_version & 0xff0) { + case 0x420: + dd->caps.has_dma = 1; + dd->caps.has_dualbuff = 1; + dd->caps.has_sha224 = 1; + dd->caps.has_sha_384_512 = 1; + break; case 0x410: dd->caps.has_dma = 1; dd->caps.has_dualbuff = 1; @@ -1349,6 +1361,7 @@ static int atmel_sha_probe(struct platform_device *pdev) platform_set_drvdata(pdev, sha_dd); INIT_LIST_HEAD(&sha_dd->list); + spin_lock_init(&sha_dd->lock); tasklet_init(&sha_dd->done_task, atmel_sha_done_task, (unsigned long)sha_dd); @@ -1385,7 +1398,7 @@ static int atmel_sha_probe(struct platform_device *pdev) /* Initializing the clock */ sha_dd->iclk = clk_get(&pdev->dev, "sha_clk"); if (IS_ERR(sha_dd->iclk)) { - dev_err(dev, "clock intialization failed.\n"); + dev_err(dev, "clock initialization failed.\n"); err = PTR_ERR(sha_dd->iclk); goto clk_err; } diff --git a/drivers/crypto/atmel-tdes.c b/drivers/crypto/atmel-tdes.c index 258772d9b22f..ca2999709eb4 100644 --- a/drivers/crypto/atmel-tdes.c +++ b/drivers/crypto/atmel-tdes.c @@ -1370,6 +1370,7 @@ static int atmel_tdes_probe(struct platform_device *pdev) platform_set_drvdata(pdev, tdes_dd); INIT_LIST_HEAD(&tdes_dd->list); + spin_lock_init(&tdes_dd->lock); tasklet_init(&tdes_dd->done_task, atmel_tdes_done_task, (unsigned long)tdes_dd); @@ -1408,7 +1409,7 @@ static int atmel_tdes_probe(struct platform_device *pdev) /* Initializing the clock */ tdes_dd->iclk = clk_get(&pdev->dev, "tdes_clk"); if (IS_ERR(tdes_dd->iclk)) { - dev_err(dev, "clock intialization failed.\n"); + dev_err(dev, "clock initialization failed.\n"); err = PTR_ERR(tdes_dd->iclk); goto clk_err; } diff --git a/drivers/crypto/caam/caamhash.c b/drivers/crypto/caam/caamhash.c index f347ab7eea95..ba0532efd3ae 100644 --- a/drivers/crypto/caam/caamhash.c +++ b/drivers/crypto/caam/caamhash.c @@ -1172,6 +1172,7 @@ static int ahash_final_no_ctx(struct ahash_request *req) return -ENOMEM; } + edesc->sec4_sg_bytes = 0; sh_len = desc_len(sh_desc); desc = edesc->hw_desc; init_job_desc_shared(desc, ptr, sh_len, HDR_SHARE_DEFER | HDR_REVERSE); diff --git a/drivers/crypto/caam/caamrng.c b/drivers/crypto/caam/caamrng.c index ae31e555793c..26a544b505f1 100644 --- a/drivers/crypto/caam/caamrng.c +++ b/drivers/crypto/caam/caamrng.c @@ -52,7 +52,7 @@ /* length of descriptors */ #define DESC_JOB_O_LEN (CAAM_CMD_SZ * 2 + CAAM_PTR_SZ * 2) -#define DESC_RNG_LEN (10 * CAAM_CMD_SZ) +#define DESC_RNG_LEN (4 * CAAM_CMD_SZ) /* Buffer, its dma address and lock */ struct buf_data { @@ -90,8 +90,8 @@ static inline void rng_unmap_ctx(struct caam_rng_ctx *ctx) struct device *jrdev = ctx->jrdev; if (ctx->sh_desc_dma) - dma_unmap_single(jrdev, ctx->sh_desc_dma, DESC_RNG_LEN, - DMA_TO_DEVICE); + dma_unmap_single(jrdev, ctx->sh_desc_dma, + desc_bytes(ctx->sh_desc), DMA_TO_DEVICE); rng_unmap_buf(jrdev, &ctx->bufs[0]); rng_unmap_buf(jrdev, &ctx->bufs[1]); } diff --git a/drivers/crypto/ccp/Makefile b/drivers/crypto/ccp/Makefile index 7f592d8d07bb..55a1f3951578 100644 --- a/drivers/crypto/ccp/Makefile +++ b/drivers/crypto/ccp/Makefile @@ -1,11 +1,6 @@ obj-$(CONFIG_CRYPTO_DEV_CCP_DD) += ccp.o -ccp-objs := ccp-dev.o ccp-ops.o -ifdef CONFIG_X86 -ccp-objs += ccp-pci.o -endif -ifdef CONFIG_ARM64 -ccp-objs += ccp-platform.o -endif +ccp-objs := ccp-dev.o ccp-ops.o ccp-platform.o +ccp-$(CONFIG_PCI) += ccp-pci.o obj-$(CONFIG_CRYPTO_DEV_CCP_CRYPTO) += ccp-crypto.o ccp-crypto-objs := ccp-crypto-main.o \ diff --git a/drivers/crypto/ccp/ccp-crypto-aes-cmac.c b/drivers/crypto/ccp/ccp-crypto-aes-cmac.c index 8e162ad82085..ea7e8446956a 100644 --- a/drivers/crypto/ccp/ccp-crypto-aes-cmac.c +++ b/drivers/crypto/ccp/ccp-crypto-aes-cmac.c @@ -23,7 +23,6 @@ #include "ccp-crypto.h" - static int ccp_aes_cmac_complete(struct crypto_async_request *async_req, int ret) { @@ -38,11 +37,13 @@ static int ccp_aes_cmac_complete(struct crypto_async_request *async_req, if (rctx->hash_rem) { /* Save remaining data to buffer */ unsigned int offset = rctx->nbytes - rctx->hash_rem; + scatterwalk_map_and_copy(rctx->buf, rctx->src, offset, rctx->hash_rem, 0); rctx->buf_count = rctx->hash_rem; - } else + } else { rctx->buf_count = 0; + } /* Update result area if supplied */ if (req->result) @@ -202,7 +203,7 @@ static int ccp_aes_cmac_digest(struct ahash_request *req) } static int ccp_aes_cmac_setkey(struct crypto_ahash *tfm, const u8 *key, - unsigned int key_len) + unsigned int key_len) { struct ccp_ctx *ctx = crypto_tfm_ctx(crypto_ahash_tfm(tfm)); struct ccp_crypto_ahash_alg *alg = @@ -292,7 +293,8 @@ static int ccp_aes_cmac_cra_init(struct crypto_tfm *tfm) crypto_ahash_set_reqsize(ahash, sizeof(struct ccp_aes_cmac_req_ctx)); cipher_tfm = crypto_alloc_cipher("aes", 0, - CRYPTO_ALG_ASYNC | CRYPTO_ALG_NEED_FALLBACK); + CRYPTO_ALG_ASYNC | + CRYPTO_ALG_NEED_FALLBACK); if (IS_ERR(cipher_tfm)) { pr_warn("could not load aes cipher driver\n"); return PTR_ERR(cipher_tfm); @@ -354,7 +356,7 @@ int ccp_register_aes_cmac_algs(struct list_head *head) ret = crypto_register_ahash(alg); if (ret) { pr_err("%s ahash algorithm registration error (%d)\n", - base->cra_name, ret); + base->cra_name, ret); kfree(ccp_alg); return ret; } diff --git a/drivers/crypto/ccp/ccp-crypto-aes-xts.c b/drivers/crypto/ccp/ccp-crypto-aes-xts.c index 0cc5594b7de3..52c7395cb8d8 100644 --- a/drivers/crypto/ccp/ccp-crypto-aes-xts.c +++ b/drivers/crypto/ccp/ccp-crypto-aes-xts.c @@ -21,7 +21,6 @@ #include "ccp-crypto.h" - struct ccp_aes_xts_def { const char *name; const char *drv_name; @@ -216,7 +215,6 @@ static void ccp_aes_xts_cra_exit(struct crypto_tfm *tfm) ctx->u.aes.tfm_ablkcipher = NULL; } - static int ccp_register_aes_xts_alg(struct list_head *head, const struct ccp_aes_xts_def *def) { @@ -255,7 +253,7 @@ static int ccp_register_aes_xts_alg(struct list_head *head, ret = crypto_register_alg(alg); if (ret) { pr_err("%s ablkcipher algorithm registration error (%d)\n", - alg->cra_name, ret); + alg->cra_name, ret); kfree(ccp_alg); return ret; } diff --git a/drivers/crypto/ccp/ccp-crypto-aes.c b/drivers/crypto/ccp/ccp-crypto-aes.c index e46490db0f63..7984f910884d 100644 --- a/drivers/crypto/ccp/ccp-crypto-aes.c +++ b/drivers/crypto/ccp/ccp-crypto-aes.c @@ -22,7 +22,6 @@ #include "ccp-crypto.h" - static int ccp_aes_complete(struct crypto_async_request *async_req, int ret) { struct ablkcipher_request *req = ablkcipher_request_cast(async_req); @@ -345,7 +344,7 @@ static int ccp_register_aes_alg(struct list_head *head, ret = crypto_register_alg(alg); if (ret) { pr_err("%s ablkcipher algorithm registration error (%d)\n", - alg->cra_name, ret); + alg->cra_name, ret); kfree(ccp_alg); return ret; } diff --git a/drivers/crypto/ccp/ccp-crypto-main.c b/drivers/crypto/ccp/ccp-crypto-main.c index 4d4e016d755b..bdec01ec608f 100644 --- a/drivers/crypto/ccp/ccp-crypto-main.c +++ b/drivers/crypto/ccp/ccp-crypto-main.c @@ -33,7 +33,6 @@ static unsigned int sha_disable; module_param(sha_disable, uint, 0444); MODULE_PARM_DESC(sha_disable, "Disable use of SHA - any non-zero value"); - /* List heads for the supported algorithms */ static LIST_HEAD(hash_algs); static LIST_HEAD(cipher_algs); @@ -48,6 +47,7 @@ struct ccp_crypto_queue { struct list_head *backlog; unsigned int cmd_count; }; + #define CCP_CRYPTO_MAX_QLEN 100 static struct ccp_crypto_queue req_queue; @@ -77,7 +77,6 @@ struct ccp_crypto_cpu { int err; }; - static inline bool ccp_crypto_success(int err) { if (err && (err != -EINPROGRESS) && (err != -EBUSY)) @@ -143,7 +142,7 @@ static void ccp_crypto_complete(void *data, int err) int ret; if (err == -EINPROGRESS) { - /* Only propogate the -EINPROGRESS if necessary */ + /* Only propagate the -EINPROGRESS if necessary */ if (crypto_cmd->ret == -EBUSY) { crypto_cmd->ret = -EINPROGRESS; req->complete(req, -EINPROGRESS); diff --git a/drivers/crypto/ccp/ccp-crypto-sha.c b/drivers/crypto/ccp/ccp-crypto-sha.c index 96531571f7cf..507b34e0cc19 100644 --- a/drivers/crypto/ccp/ccp-crypto-sha.c +++ b/drivers/crypto/ccp/ccp-crypto-sha.c @@ -23,7 +23,6 @@ #include "ccp-crypto.h" - static int ccp_sha_complete(struct crypto_async_request *async_req, int ret) { struct ahash_request *req = ahash_request_cast(async_req); @@ -37,11 +36,13 @@ static int ccp_sha_complete(struct crypto_async_request *async_req, int ret) if (rctx->hash_rem) { /* Save remaining data to buffer */ unsigned int offset = rctx->nbytes - rctx->hash_rem; + scatterwalk_map_and_copy(rctx->buf, rctx->src, offset, rctx->hash_rem, 0); rctx->buf_count = rctx->hash_rem; - } else + } else { rctx->buf_count = 0; + } /* Update result area if supplied */ if (req->result) @@ -227,8 +228,9 @@ static int ccp_sha_setkey(struct crypto_ahash *tfm, const u8 *key, } key_len = digest_size; - } else + } else { memcpy(ctx->u.sha.key, key, key_len); + } for (i = 0; i < block_size; i++) { ctx->u.sha.ipad[i] = ctx->u.sha.key[i] ^ 0x36; @@ -355,7 +357,7 @@ static int ccp_register_hmac_alg(struct list_head *head, ret = crypto_register_ahash(alg); if (ret) { pr_err("%s ahash algorithm registration error (%d)\n", - base->cra_name, ret); + base->cra_name, ret); kfree(ccp_alg); return ret; } @@ -410,7 +412,7 @@ static int ccp_register_sha_alg(struct list_head *head, ret = crypto_register_ahash(alg); if (ret) { pr_err("%s ahash algorithm registration error (%d)\n", - base->cra_name, ret); + base->cra_name, ret); kfree(ccp_alg); return ret; } diff --git a/drivers/crypto/ccp/ccp-crypto.h b/drivers/crypto/ccp/ccp-crypto.h index 9aa4ae184f7f..76a96f0f44c6 100644 --- a/drivers/crypto/ccp/ccp-crypto.h +++ b/drivers/crypto/ccp/ccp-crypto.h @@ -13,7 +13,6 @@ #ifndef __CCP_CRYPTO_H__ #define __CCP_CRYPTO_H__ - #include <linux/list.h> #include <linux/wait.h> #include <linux/pci.h> @@ -25,7 +24,6 @@ #include <crypto/hash.h> #include <crypto/sha.h> - #define CCP_CRA_PRIORITY 300 struct ccp_crypto_ablkcipher_alg { @@ -68,7 +66,6 @@ static inline struct ccp_crypto_ahash_alg * return container_of(ahash_alg, struct ccp_crypto_ahash_alg, alg); } - /***** AES related defines *****/ struct ccp_aes_ctx { /* Fallback cipher for XTS with unsupported unit sizes */ diff --git a/drivers/crypto/ccp/ccp-dev.c b/drivers/crypto/ccp/ccp-dev.c index ca29c120b85f..861bacc1bb94 100644 --- a/drivers/crypto/ccp/ccp-dev.c +++ b/drivers/crypto/ccp/ccp-dev.c @@ -37,7 +37,6 @@ struct ccp_tasklet_data { struct ccp_cmd *cmd; }; - static struct ccp_device *ccp_dev; static inline struct ccp_device *ccp_get_device(void) { @@ -296,11 +295,9 @@ struct ccp_device *ccp_alloc_struct(struct device *dev) { struct ccp_device *ccp; - ccp = kzalloc(sizeof(*ccp), GFP_KERNEL); - if (ccp == NULL) { - dev_err(dev, "unable to allocate device struct\n"); + ccp = devm_kzalloc(dev, sizeof(*ccp), GFP_KERNEL); + if (!ccp) return NULL; - } ccp->dev = dev; INIT_LIST_HEAD(&ccp->cmd); diff --git a/drivers/crypto/ccp/ccp-dev.h b/drivers/crypto/ccp/ccp-dev.h index 62ff35a6b9ec..6ff89031fb96 100644 --- a/drivers/crypto/ccp/ccp-dev.h +++ b/drivers/crypto/ccp/ccp-dev.h @@ -21,7 +21,7 @@ #include <linux/wait.h> #include <linux/dmapool.h> #include <linux/hw_random.h> - +#include <linux/bitops.h> #define MAX_DMAPOOL_NAME_LEN 32 @@ -33,7 +33,6 @@ #define CACHE_NONE 0x00 #define CACHE_WB_NO_ALLOC 0xb7 - /****** Register Mappings ******/ #define Q_MASK_REG 0x000 #define TRNG_OUT_REG 0x00c @@ -54,8 +53,8 @@ #define CMD_Q_CACHE_BASE 0x228 #define CMD_Q_CACHE_INC 0x20 -#define CMD_Q_ERROR(__qs) ((__qs) & 0x0000003f); -#define CMD_Q_DEPTH(__qs) (((__qs) >> 12) & 0x0000000f); +#define CMD_Q_ERROR(__qs) ((__qs) & 0x0000003f) +#define CMD_Q_DEPTH(__qs) (((__qs) >> 12) & 0x0000000f) /****** REQ0 Related Values ******/ #define REQ0_WAIT_FOR_WRITE 0x00000004 @@ -103,7 +102,6 @@ /****** REQ6 Related Values ******/ #define REQ6_MEMTYPE_SHIFT 16 - /****** Key Storage Block ******/ #define KSB_START 77 #define KSB_END 127 @@ -114,7 +112,7 @@ #define CCP_JOBID_MASK 0x0000003f #define CCP_DMAPOOL_MAX_SIZE 64 -#define CCP_DMAPOOL_ALIGN (1 << 5) +#define CCP_DMAPOOL_ALIGN BIT(5) #define CCP_REVERSE_BUF_SIZE 64 @@ -142,7 +140,6 @@ #define CCP_ECC_RESULT_OFFSET 60 #define CCP_ECC_RESULT_SUCCESS 0x0001 - struct ccp_device; struct ccp_cmd; @@ -261,7 +258,6 @@ struct ccp_device { unsigned int axcache; }; - int ccp_pci_init(void); void ccp_pci_exit(void); diff --git a/drivers/crypto/ccp/ccp-ops.c b/drivers/crypto/ccp/ccp-ops.c index 8729364261d7..71f2e3c89424 100644 --- a/drivers/crypto/ccp/ccp-ops.c +++ b/drivers/crypto/ccp/ccp-ops.c @@ -27,7 +27,6 @@ #include "ccp-dev.h" - enum ccp_memtype { CCP_MEMTYPE_SYSTEM = 0, CCP_MEMTYPE_KSB, @@ -515,7 +514,6 @@ static int ccp_init_sg_workarea(struct ccp_sg_workarea *wa, struct device *dev, if (!wa->dma_count) return -ENOMEM; - return 0; } @@ -763,8 +761,9 @@ static void ccp_prepare_data(struct ccp_data *src, struct ccp_data *dst, sg_dst_len = sg_dma_len(dst->sg_wa.sg) - dst->sg_wa.sg_used; sg_dst_len = min_t(u64, src->sg_wa.bytes_left, sg_dst_len); op_len = min(sg_src_len, sg_dst_len); - } else + } else { op_len = sg_src_len; + } /* The data operation length will be at least block_size in length * or the smaller of available sg room remaining for the source or @@ -1131,9 +1130,9 @@ static int ccp_run_aes_cmd(struct ccp_cmd_queue *cmd_q, struct ccp_cmd *cmd) if (ret) goto e_ctx; - if (in_place) + if (in_place) { dst = src; - else { + } else { ret = ccp_init_data(&dst, cmd_q, aes->dst, aes->src_len, AES_BLOCK_SIZE, DMA_FROM_DEVICE); if (ret) @@ -1304,9 +1303,9 @@ static int ccp_run_xts_aes_cmd(struct ccp_cmd_queue *cmd_q, if (ret) goto e_ctx; - if (in_place) + if (in_place) { dst = src; - else { + } else { ret = ccp_init_data(&dst, cmd_q, xts->dst, xts->src_len, unit_size, DMA_FROM_DEVICE); if (ret) @@ -1451,8 +1450,9 @@ static int ccp_run_sha_cmd(struct ccp_cmd_queue *cmd_q, struct ccp_cmd *cmd) goto e_ctx; } memcpy(ctx.address, init, CCP_SHA_CTXSIZE); - } else + } else { ccp_set_dm_area(&ctx, 0, sha->ctx, 0, sha->ctx_len); + } ret = ccp_copy_to_ksb(cmd_q, &ctx, op.jobid, op.ksb_ctx, CCP_PASSTHRU_BYTESWAP_256BIT); @@ -1732,9 +1732,9 @@ static int ccp_run_passthru_cmd(struct ccp_cmd_queue *cmd_q, if (ret) goto e_mask; - if (in_place) + if (in_place) { dst = src; - else { + } else { ret = ccp_init_data(&dst, cmd_q, pt->dst, pt->src_len, CCP_PASSTHRU_MASKSIZE, DMA_FROM_DEVICE); if (ret) @@ -1974,7 +1974,7 @@ static int ccp_run_ecc_pm_cmd(struct ccp_cmd_queue *cmd_q, struct ccp_cmd *cmd) src.address += CCP_ECC_OPERAND_SIZE; /* Set the first point Z coordianate to 1 */ - *(src.address) = 0x01; + *src.address = 0x01; src.address += CCP_ECC_OPERAND_SIZE; if (ecc->function == CCP_ECC_FUNCTION_PADD_384BIT) { @@ -1989,7 +1989,7 @@ static int ccp_run_ecc_pm_cmd(struct ccp_cmd_queue *cmd_q, struct ccp_cmd *cmd) src.address += CCP_ECC_OPERAND_SIZE; /* Set the second point Z coordianate to 1 */ - *(src.address) = 0x01; + *src.address = 0x01; src.address += CCP_ECC_OPERAND_SIZE; } else { /* Copy the Domain "a" parameter */ diff --git a/drivers/crypto/ccp/ccp-pci.c b/drivers/crypto/ccp/ccp-pci.c index 7f89c946adfe..af190d4795a8 100644 --- a/drivers/crypto/ccp/ccp-pci.c +++ b/drivers/crypto/ccp/ccp-pci.c @@ -174,11 +174,10 @@ static int ccp_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id) if (!ccp) goto e_err; - ccp_pci = kzalloc(sizeof(*ccp_pci), GFP_KERNEL); - if (!ccp_pci) { - ret = -ENOMEM; - goto e_free1; - } + ccp_pci = devm_kzalloc(dev, sizeof(*ccp_pci), GFP_KERNEL); + if (!ccp_pci) + goto e_err; + ccp->dev_specific = ccp_pci; ccp->get_irq = ccp_get_irqs; ccp->free_irq = ccp_free_irqs; @@ -186,7 +185,7 @@ static int ccp_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id) ret = pci_request_regions(pdev, "ccp"); if (ret) { dev_err(dev, "pci_request_regions failed (%d)\n", ret); - goto e_free2; + goto e_err; } ret = pci_enable_device(pdev); @@ -204,7 +203,7 @@ static int ccp_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id) ret = -EIO; ccp->io_map = pci_iomap(pdev, bar, 0); - if (ccp->io_map == NULL) { + if (!ccp->io_map) { dev_err(dev, "pci_iomap failed\n"); goto e_device; } @@ -239,12 +238,6 @@ e_device: e_regions: pci_release_regions(pdev); -e_free2: - kfree(ccp_pci); - -e_free1: - kfree(ccp); - e_err: dev_notice(dev, "initialization failed\n"); return ret; @@ -266,8 +259,6 @@ static void ccp_pci_remove(struct pci_dev *pdev) pci_release_regions(pdev); - kfree(ccp); - dev_notice(dev, "disabled\n"); } diff --git a/drivers/crypto/ccp/ccp-platform.c b/drivers/crypto/ccp/ccp-platform.c index 8c50bad25f7e..b1c20b2b5647 100644 --- a/drivers/crypto/ccp/ccp-platform.c +++ b/drivers/crypto/ccp/ccp-platform.c @@ -23,9 +23,15 @@ #include <linux/delay.h> #include <linux/ccp.h> #include <linux/of.h> +#include <linux/of_address.h> +#include <linux/acpi.h> #include "ccp-dev.h" +struct ccp_platform { + int use_acpi; + int coherent; +}; static int ccp_get_irq(struct ccp_device *ccp) { @@ -84,10 +90,64 @@ static struct resource *ccp_find_mmio_area(struct ccp_device *ccp) return NULL; } +#ifdef CONFIG_ACPI +static int ccp_acpi_support(struct ccp_device *ccp) +{ + struct ccp_platform *ccp_platform = ccp->dev_specific; + struct acpi_device *adev = ACPI_COMPANION(ccp->dev); + acpi_handle handle; + acpi_status status; + unsigned long long data; + int cca; + + /* Retrieve the device cache coherency value */ + handle = adev->handle; + do { + status = acpi_evaluate_integer(handle, "_CCA", NULL, &data); + if (!ACPI_FAILURE(status)) { + cca = data; + break; + } + } while (!ACPI_FAILURE(status)); + + if (ACPI_FAILURE(status)) { + dev_err(ccp->dev, "error obtaining acpi coherency value\n"); + return -EINVAL; + } + + ccp_platform->coherent = !!cca; + + return 0; +} +#else /* CONFIG_ACPI */ +static int ccp_acpi_support(struct ccp_device *ccp) +{ + return -EINVAL; +} +#endif + +#ifdef CONFIG_OF +static int ccp_of_support(struct ccp_device *ccp) +{ + struct ccp_platform *ccp_platform = ccp->dev_specific; + + ccp_platform->coherent = of_dma_is_coherent(ccp->dev->of_node); + + return 0; +} +#else +static int ccp_of_support(struct ccp_device *ccp) +{ + return -EINVAL; +} +#endif + static int ccp_platform_probe(struct platform_device *pdev) { struct ccp_device *ccp; + struct ccp_platform *ccp_platform; struct device *dev = &pdev->dev; + struct acpi_device *adev = ACPI_COMPANION(dev); struct resource *ior; int ret; @@ -96,24 +156,40 @@ static int ccp_platform_probe(struct platform_device *pdev) if (!ccp) goto e_err; - ccp->dev_specific = NULL; + ccp_platform = devm_kzalloc(dev, sizeof(*ccp_platform), GFP_KERNEL); + if (!ccp_platform) + goto e_err; + + ccp->dev_specific = ccp_platform; ccp->get_irq = ccp_get_irqs; ccp->free_irq = ccp_free_irqs; + ccp_platform->use_acpi = (!adev || acpi_disabled) ? 0 : 1; + ior = ccp_find_mmio_area(ccp); ccp->io_map = devm_ioremap_resource(dev, ior); if (IS_ERR(ccp->io_map)) { ret = PTR_ERR(ccp->io_map); - goto e_free; + goto e_err; } ccp->io_regs = ccp->io_map; if (!dev->dma_mask) dev->dma_mask = &dev->coherent_dma_mask; - *(dev->dma_mask) = DMA_BIT_MASK(48); - dev->coherent_dma_mask = DMA_BIT_MASK(48); + ret = dma_set_mask_and_coherent(dev, DMA_BIT_MASK(48)); + if (ret) { + dev_err(dev, "dma_set_mask_and_coherent failed (%d)\n", ret); + goto e_err; + } + + if (ccp_platform->use_acpi) + ret = ccp_acpi_support(ccp); + else + ret = ccp_of_support(ccp); + if (ret) + goto e_err; - if (of_property_read_bool(dev->of_node, "dma-coherent")) + if (ccp_platform->coherent) ccp->axcache = CACHE_WB_NO_ALLOC; else ccp->axcache = CACHE_NONE; @@ -122,15 +198,12 @@ static int ccp_platform_probe(struct platform_device *pdev) ret = ccp_init(ccp); if (ret) - goto e_free; + goto e_err; dev_notice(dev, "enabled\n"); return 0; -e_free: - kfree(ccp); - e_err: dev_notice(dev, "initialization failed\n"); return ret; @@ -143,8 +216,6 @@ static int ccp_platform_remove(struct platform_device *pdev) ccp_destroy(ccp); - kfree(ccp); - dev_notice(dev, "disabled\n"); return 0; @@ -200,15 +271,29 @@ static int ccp_platform_resume(struct platform_device *pdev) } #endif -static const struct of_device_id ccp_platform_ids[] = { +#ifdef CONFIG_ACPI +static const struct acpi_device_id ccp_acpi_match[] = { + { "AMDI0C00", 0 }, + { }, +}; +#endif + +#ifdef CONFIG_OF +static const struct of_device_id ccp_of_match[] = { { .compatible = "amd,ccp-seattle-v1a" }, { }, }; +#endif static struct platform_driver ccp_platform_driver = { .driver = { .name = "AMD Cryptographic Coprocessor", - .of_match_table = ccp_platform_ids, +#ifdef CONFIG_ACPI + .acpi_match_table = ccp_acpi_match, +#endif +#ifdef CONFIG_OF + .of_match_table = ccp_of_match, +#endif }, .probe = ccp_platform_probe, .remove = ccp_platform_remove, diff --git a/drivers/crypto/img-hash.c b/drivers/crypto/img-hash.c new file mode 100644 index 000000000000..ad47d0d61098 --- /dev/null +++ b/drivers/crypto/img-hash.c @@ -0,0 +1,1029 @@ +/* + * Copyright (c) 2014 Imagination Technologies + * Authors: Will Thomas, James Hartley + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as published + * by the Free Software Foundation. + * + * Interface structure taken from omap-sham driver + */ + +#include <linux/clk.h> +#include <linux/dmaengine.h> +#include <linux/interrupt.h> +#include <linux/io.h> +#include <linux/kernel.h> +#include <linux/module.h> +#include <linux/of_device.h> +#include <linux/platform_device.h> +#include <linux/scatterlist.h> + +#include <crypto/internal/hash.h> +#include <crypto/md5.h> +#include <crypto/sha.h> + +#define CR_RESET 0 +#define CR_RESET_SET 1 +#define CR_RESET_UNSET 0 + +#define CR_MESSAGE_LENGTH_H 0x4 +#define CR_MESSAGE_LENGTH_L 0x8 + +#define CR_CONTROL 0xc +#define CR_CONTROL_BYTE_ORDER_3210 0 +#define CR_CONTROL_BYTE_ORDER_0123 1 +#define CR_CONTROL_BYTE_ORDER_2310 2 +#define CR_CONTROL_BYTE_ORDER_1032 3 +#define CR_CONTROL_BYTE_ORDER_SHIFT 8 +#define CR_CONTROL_ALGO_MD5 0 +#define CR_CONTROL_ALGO_SHA1 1 +#define CR_CONTROL_ALGO_SHA224 2 +#define CR_CONTROL_ALGO_SHA256 3 + +#define CR_INTSTAT 0x10 +#define CR_INTENAB 0x14 +#define CR_INTCLEAR 0x18 +#define CR_INT_RESULTS_AVAILABLE BIT(0) +#define CR_INT_NEW_RESULTS_SET BIT(1) +#define CR_INT_RESULT_READ_ERR BIT(2) +#define CR_INT_MESSAGE_WRITE_ERROR BIT(3) +#define CR_INT_STATUS BIT(8) + +#define CR_RESULT_QUEUE 0x1c +#define CR_RSD0 0x40 +#define CR_CORE_REV 0x50 +#define CR_CORE_DES1 0x60 +#define CR_CORE_DES2 0x70 + +#define DRIVER_FLAGS_BUSY BIT(0) +#define DRIVER_FLAGS_FINAL BIT(1) +#define DRIVER_FLAGS_DMA_ACTIVE BIT(2) +#define DRIVER_FLAGS_OUTPUT_READY BIT(3) +#define DRIVER_FLAGS_INIT BIT(4) +#define DRIVER_FLAGS_CPU BIT(5) +#define DRIVER_FLAGS_DMA_READY BIT(6) +#define DRIVER_FLAGS_ERROR BIT(7) +#define DRIVER_FLAGS_SG BIT(8) +#define DRIVER_FLAGS_SHA1 BIT(18) +#define DRIVER_FLAGS_SHA224 BIT(19) +#define DRIVER_FLAGS_SHA256 BIT(20) +#define DRIVER_FLAGS_MD5 BIT(21) + +#define IMG_HASH_QUEUE_LENGTH 20 +#define IMG_HASH_DMA_THRESHOLD 64 + +#ifdef __LITTLE_ENDIAN +#define IMG_HASH_BYTE_ORDER CR_CONTROL_BYTE_ORDER_3210 +#else +#define IMG_HASH_BYTE_ORDER CR_CONTROL_BYTE_ORDER_0123 +#endif + +struct img_hash_dev; + +struct img_hash_request_ctx { + struct img_hash_dev *hdev; + u8 digest[SHA256_DIGEST_SIZE] __aligned(sizeof(u32)); + unsigned long flags; + size_t digsize; + + dma_addr_t dma_addr; + size_t dma_ct; + + /* sg root */ + struct scatterlist *sgfirst; + /* walk state */ + struct scatterlist *sg; + size_t nents; + size_t offset; + unsigned int total; + size_t sent; + + unsigned long op; + + size_t bufcnt; + u8 buffer[0] __aligned(sizeof(u32)); + struct ahash_request fallback_req; +}; + +struct img_hash_ctx { + struct img_hash_dev *hdev; + unsigned long flags; + struct crypto_ahash *fallback; +}; + +struct img_hash_dev { + struct list_head list; + struct device *dev; + struct clk *hash_clk; + struct clk *sys_clk; + void __iomem *io_base; + + phys_addr_t bus_addr; + void __iomem *cpu_addr; + + spinlock_t lock; + int err; + struct tasklet_struct done_task; + struct tasklet_struct dma_task; + + unsigned long flags; + struct crypto_queue queue; + struct ahash_request *req; + + struct dma_chan *dma_lch; +}; + +struct img_hash_drv { + struct list_head dev_list; + spinlock_t lock; +}; + +static struct img_hash_drv img_hash = { + .dev_list = LIST_HEAD_INIT(img_hash.dev_list), + .lock = __SPIN_LOCK_UNLOCKED(img_hash.lock), +}; + +static inline u32 img_hash_read(struct img_hash_dev *hdev, u32 offset) +{ + return readl_relaxed(hdev->io_base + offset); +} + +static inline void img_hash_write(struct img_hash_dev *hdev, + u32 offset, u32 value) +{ + writel_relaxed(value, hdev->io_base + offset); +} + +static inline u32 img_hash_read_result_queue(struct img_hash_dev *hdev) +{ + return be32_to_cpu(img_hash_read(hdev, CR_RESULT_QUEUE)); +} + +static void img_hash_start(struct img_hash_dev *hdev, bool dma) +{ + struct img_hash_request_ctx *ctx = ahash_request_ctx(hdev->req); + u32 cr = IMG_HASH_BYTE_ORDER << CR_CONTROL_BYTE_ORDER_SHIFT; + + if (ctx->flags & DRIVER_FLAGS_MD5) + cr |= CR_CONTROL_ALGO_MD5; + else if (ctx->flags & DRIVER_FLAGS_SHA1) + cr |= CR_CONTROL_ALGO_SHA1; + else if (ctx->flags & DRIVER_FLAGS_SHA224) + cr |= CR_CONTROL_ALGO_SHA224; + else if (ctx->flags & DRIVER_FLAGS_SHA256) + cr |= CR_CONTROL_ALGO_SHA256; + dev_dbg(hdev->dev, "Starting hash process\n"); + img_hash_write(hdev, CR_CONTROL, cr); + + /* + * The hardware block requires two cycles between writing the control + * register and writing the first word of data in non DMA mode, to + * ensure the first data write is not grouped in burst with the control + * register write a read is issued to 'flush' the bus. + */ + if (!dma) + img_hash_read(hdev, CR_CONTROL); +} + +static int img_hash_xmit_cpu(struct img_hash_dev *hdev, const u8 *buf, + size_t length, int final) +{ + u32 count, len32; + const u32 *buffer = (const u32 *)buf; + + dev_dbg(hdev->dev, "xmit_cpu: length: %zu bytes\n", length); + + if (final) + hdev->flags |= DRIVER_FLAGS_FINAL; + + len32 = DIV_ROUND_UP(length, sizeof(u32)); + + for (count = 0; count < len32; count++) + writel_relaxed(buffer[count], hdev->cpu_addr); + + return -EINPROGRESS; +} + +static void img_hash_dma_callback(void *data) +{ + struct img_hash_dev *hdev = (struct img_hash_dev *)data; + struct img_hash_request_ctx *ctx = ahash_request_ctx(hdev->req); + + if (ctx->bufcnt) { + img_hash_xmit_cpu(hdev, ctx->buffer, ctx->bufcnt, 0); + ctx->bufcnt = 0; + } + if (ctx->sg) + tasklet_schedule(&hdev->dma_task); +} + +static int img_hash_xmit_dma(struct img_hash_dev *hdev, struct scatterlist *sg) +{ + struct dma_async_tx_descriptor *desc; + struct img_hash_request_ctx *ctx = ahash_request_ctx(hdev->req); + + ctx->dma_ct = dma_map_sg(hdev->dev, sg, 1, DMA_MEM_TO_DEV); + if (ctx->dma_ct == 0) { + dev_err(hdev->dev, "Invalid DMA sg\n"); + hdev->err = -EINVAL; + return -EINVAL; + } + + desc = dmaengine_prep_slave_sg(hdev->dma_lch, + sg, + ctx->dma_ct, + DMA_MEM_TO_DEV, + DMA_PREP_INTERRUPT | DMA_CTRL_ACK); + if (!desc) { + dev_err(hdev->dev, "Null DMA descriptor\n"); + hdev->err = -EINVAL; + dma_unmap_sg(hdev->dev, sg, 1, DMA_MEM_TO_DEV); + return -EINVAL; + } + desc->callback = img_hash_dma_callback; + desc->callback_param = hdev; + dmaengine_submit(desc); + dma_async_issue_pending(hdev->dma_lch); + + return 0; +} + +static int img_hash_write_via_cpu(struct img_hash_dev *hdev) +{ + struct img_hash_request_ctx *ctx = ahash_request_ctx(hdev->req); + + ctx->bufcnt = sg_copy_to_buffer(hdev->req->src, sg_nents(ctx->sg), + ctx->buffer, hdev->req->nbytes); + + ctx->total = hdev->req->nbytes; + ctx->bufcnt = 0; + + hdev->flags |= (DRIVER_FLAGS_CPU | DRIVER_FLAGS_FINAL); + + img_hash_start(hdev, false); + + return img_hash_xmit_cpu(hdev, ctx->buffer, ctx->total, 1); +} + +static int img_hash_finish(struct ahash_request *req) +{ + struct img_hash_request_ctx *ctx = ahash_request_ctx(req); + + if (!req->result) + return -EINVAL; + + memcpy(req->result, ctx->digest, ctx->digsize); + + return 0; +} + +static void img_hash_copy_hash(struct ahash_request *req) +{ + struct img_hash_request_ctx *ctx = ahash_request_ctx(req); + u32 *hash = (u32 *)ctx->digest; + int i; + + for (i = (ctx->digsize / sizeof(u32)) - 1; i >= 0; i--) + hash[i] = img_hash_read_result_queue(ctx->hdev); +} + +static void img_hash_finish_req(struct ahash_request *req, int err) +{ + struct img_hash_request_ctx *ctx = ahash_request_ctx(req); + struct img_hash_dev *hdev = ctx->hdev; + + if (!err) { + img_hash_copy_hash(req); + if (DRIVER_FLAGS_FINAL & hdev->flags) + err = img_hash_finish(req); + } else { + dev_warn(hdev->dev, "Hash failed with error %d\n", err); + ctx->flags |= DRIVER_FLAGS_ERROR; + } + + hdev->flags &= ~(DRIVER_FLAGS_DMA_READY | DRIVER_FLAGS_OUTPUT_READY | + DRIVER_FLAGS_CPU | DRIVER_FLAGS_BUSY | DRIVER_FLAGS_FINAL); + + if (req->base.complete) + req->base.complete(&req->base, err); +} + +static int img_hash_write_via_dma(struct img_hash_dev *hdev) +{ + struct img_hash_request_ctx *ctx = ahash_request_ctx(hdev->req); + + img_hash_start(hdev, true); + + dev_dbg(hdev->dev, "xmit dma size: %d\n", ctx->total); + + if (!ctx->total) + hdev->flags |= DRIVER_FLAGS_FINAL; + + hdev->flags |= DRIVER_FLAGS_DMA_ACTIVE | DRIVER_FLAGS_FINAL; + + tasklet_schedule(&hdev->dma_task); + + return -EINPROGRESS; +} + +static int img_hash_dma_init(struct img_hash_dev *hdev) +{ + struct dma_slave_config dma_conf; + int err = -EINVAL; + + hdev->dma_lch = dma_request_slave_channel(hdev->dev, "tx"); + if (!hdev->dma_lch) { + dev_err(hdev->dev, "Couldn't aquire a slave DMA channel.\n"); + return -EBUSY; + } + dma_conf.direction = DMA_MEM_TO_DEV; + dma_conf.dst_addr = hdev->bus_addr; + dma_conf.dst_addr_width = DMA_SLAVE_BUSWIDTH_4_BYTES; + dma_conf.dst_maxburst = 16; + dma_conf.device_fc = false; + + err = dmaengine_slave_config(hdev->dma_lch, &dma_conf); + if (err) { + dev_err(hdev->dev, "Couldn't configure DMA slave.\n"); + dma_release_channel(hdev->dma_lch); + return err; + } + + return 0; +} + +static void img_hash_dma_task(unsigned long d) +{ + struct img_hash_dev *hdev = (struct img_hash_dev *)d; + struct img_hash_request_ctx *ctx = ahash_request_ctx(hdev->req); + u8 *addr; + size_t nbytes, bleft, wsend, len, tbc; + struct scatterlist tsg; + + if (!ctx->sg) + return; + + addr = sg_virt(ctx->sg); + nbytes = ctx->sg->length - ctx->offset; + + /* + * The hash accelerator does not support a data valid mask. This means + * that if each dma (i.e. per page) is not a multiple of 4 bytes, the + * padding bytes in the last word written by that dma would erroneously + * be included in the hash. To avoid this we round down the transfer, + * and add the excess to the start of the next dma. It does not matter + * that the final dma may not be a multiple of 4 bytes as the hashing + * block is programmed to accept the correct number of bytes. + */ + + bleft = nbytes % 4; + wsend = (nbytes / 4); + + if (wsend) { + sg_init_one(&tsg, addr + ctx->offset, wsend * 4); + if (img_hash_xmit_dma(hdev, &tsg)) { + dev_err(hdev->dev, "DMA failed, falling back to CPU"); + ctx->flags |= DRIVER_FLAGS_CPU; + hdev->err = 0; + img_hash_xmit_cpu(hdev, addr + ctx->offset, + wsend * 4, 0); + ctx->sent += wsend * 4; + wsend = 0; + } else { + ctx->sent += wsend * 4; + } + } + + if (bleft) { + ctx->bufcnt = sg_pcopy_to_buffer(ctx->sgfirst, ctx->nents, + ctx->buffer, bleft, ctx->sent); + tbc = 0; + ctx->sg = sg_next(ctx->sg); + while (ctx->sg && (ctx->bufcnt < 4)) { + len = ctx->sg->length; + if (likely(len > (4 - ctx->bufcnt))) + len = 4 - ctx->bufcnt; + tbc = sg_pcopy_to_buffer(ctx->sgfirst, ctx->nents, + ctx->buffer + ctx->bufcnt, len, + ctx->sent + ctx->bufcnt); + ctx->bufcnt += tbc; + if (tbc >= ctx->sg->length) { + ctx->sg = sg_next(ctx->sg); + tbc = 0; + } + } + + ctx->sent += ctx->bufcnt; + ctx->offset = tbc; + + if (!wsend) + img_hash_dma_callback(hdev); + } else { + ctx->offset = 0; + ctx->sg = sg_next(ctx->sg); + } +} + +static int img_hash_write_via_dma_stop(struct img_hash_dev *hdev) +{ + struct img_hash_request_ctx *ctx = ahash_request_ctx(hdev->req); + + if (ctx->flags & DRIVER_FLAGS_SG) + dma_unmap_sg(hdev->dev, ctx->sg, ctx->dma_ct, DMA_TO_DEVICE); + + return 0; +} + +static int img_hash_process_data(struct img_hash_dev *hdev) +{ + struct ahash_request *req = hdev->req; + struct img_hash_request_ctx *ctx = ahash_request_ctx(req); + int err = 0; + + ctx->bufcnt = 0; + + if (req->nbytes >= IMG_HASH_DMA_THRESHOLD) { + dev_dbg(hdev->dev, "process data request(%d bytes) using DMA\n", + req->nbytes); + err = img_hash_write_via_dma(hdev); + } else { + dev_dbg(hdev->dev, "process data request(%d bytes) using CPU\n", + req->nbytes); + err = img_hash_write_via_cpu(hdev); + } + return err; +} + +static int img_hash_hw_init(struct img_hash_dev *hdev) +{ + unsigned long long nbits; + u32 u, l; + + img_hash_write(hdev, CR_RESET, CR_RESET_SET); + img_hash_write(hdev, CR_RESET, CR_RESET_UNSET); + img_hash_write(hdev, CR_INTENAB, CR_INT_NEW_RESULTS_SET); + + nbits = (u64)hdev->req->nbytes << 3; + u = nbits >> 32; + l = nbits; + img_hash_write(hdev, CR_MESSAGE_LENGTH_H, u); + img_hash_write(hdev, CR_MESSAGE_LENGTH_L, l); + + if (!(DRIVER_FLAGS_INIT & hdev->flags)) { + hdev->flags |= DRIVER_FLAGS_INIT; + hdev->err = 0; + } + dev_dbg(hdev->dev, "hw initialized, nbits: %llx\n", nbits); + return 0; +} + +static int img_hash_init(struct ahash_request *req) +{ + struct crypto_ahash *tfm = crypto_ahash_reqtfm(req); + struct img_hash_request_ctx *rctx = ahash_request_ctx(req); + struct img_hash_ctx *ctx = crypto_ahash_ctx(tfm); + + ahash_request_set_tfm(&rctx->fallback_req, ctx->fallback); + rctx->fallback_req.base.flags = req->base.flags + & CRYPTO_TFM_REQ_MAY_SLEEP; + + return crypto_ahash_init(&rctx->fallback_req); +} + +static int img_hash_handle_queue(struct img_hash_dev *hdev, + struct ahash_request *req) +{ + struct crypto_async_request *async_req, *backlog; + struct img_hash_request_ctx *ctx; + unsigned long flags; + int err = 0, res = 0; + + spin_lock_irqsave(&hdev->lock, flags); + + if (req) + res = ahash_enqueue_request(&hdev->queue, req); + + if (DRIVER_FLAGS_BUSY & hdev->flags) { + spin_unlock_irqrestore(&hdev->lock, flags); + return res; + } + + backlog = crypto_get_backlog(&hdev->queue); + async_req = crypto_dequeue_request(&hdev->queue); + if (async_req) + hdev->flags |= DRIVER_FLAGS_BUSY; + + spin_unlock_irqrestore(&hdev->lock, flags); + + if (!async_req) + return res; + + if (backlog) + backlog->complete(backlog, -EINPROGRESS); + + req = ahash_request_cast(async_req); + hdev->req = req; + + ctx = ahash_request_ctx(req); + + dev_info(hdev->dev, "processing req, op: %lu, bytes: %d\n", + ctx->op, req->nbytes); + + err = img_hash_hw_init(hdev); + + if (!err) + err = img_hash_process_data(hdev); + + if (err != -EINPROGRESS) { + /* done_task will not finish so do it here */ + img_hash_finish_req(req, err); + } + return res; +} + +static int img_hash_update(struct ahash_request *req) +{ + struct img_hash_request_ctx *rctx = ahash_request_ctx(req); + struct crypto_ahash *tfm = crypto_ahash_reqtfm(req); + struct img_hash_ctx *ctx = crypto_ahash_ctx(tfm); + + ahash_request_set_tfm(&rctx->fallback_req, ctx->fallback); + rctx->fallback_req.base.flags = req->base.flags + & CRYPTO_TFM_REQ_MAY_SLEEP; + rctx->fallback_req.nbytes = req->nbytes; + rctx->fallback_req.src = req->src; + + return crypto_ahash_update(&rctx->fallback_req); +} + +static int img_hash_final(struct ahash_request *req) +{ + struct img_hash_request_ctx *rctx = ahash_request_ctx(req); + struct crypto_ahash *tfm = crypto_ahash_reqtfm(req); + struct img_hash_ctx *ctx = crypto_ahash_ctx(tfm); + + ahash_request_set_tfm(&rctx->fallback_req, ctx->fallback); + rctx->fallback_req.base.flags = req->base.flags + & CRYPTO_TFM_REQ_MAY_SLEEP; + rctx->fallback_req.result = req->result; + + return crypto_ahash_final(&rctx->fallback_req); +} + +static int img_hash_finup(struct ahash_request *req) +{ + struct img_hash_request_ctx *rctx = ahash_request_ctx(req); + struct crypto_ahash *tfm = crypto_ahash_reqtfm(req); + struct img_hash_ctx *ctx = crypto_ahash_ctx(tfm); + + ahash_request_set_tfm(&rctx->fallback_req, ctx->fallback); + rctx->fallback_req.base.flags = req->base.flags + & CRYPTO_TFM_REQ_MAY_SLEEP; + rctx->fallback_req.nbytes = req->nbytes; + rctx->fallback_req.src = req->src; + rctx->fallback_req.result = req->result; + + return crypto_ahash_finup(&rctx->fallback_req); +} + +static int img_hash_digest(struct ahash_request *req) +{ + struct crypto_ahash *tfm = crypto_ahash_reqtfm(req); + struct img_hash_ctx *tctx = crypto_ahash_ctx(tfm); + struct img_hash_request_ctx *ctx = ahash_request_ctx(req); + struct img_hash_dev *hdev = NULL; + struct img_hash_dev *tmp; + int err; + + spin_lock(&img_hash.lock); + if (!tctx->hdev) { + list_for_each_entry(tmp, &img_hash.dev_list, list) { + hdev = tmp; + break; + } + tctx->hdev = hdev; + + } else { + hdev = tctx->hdev; + } + + spin_unlock(&img_hash.lock); + ctx->hdev = hdev; + ctx->flags = 0; + ctx->digsize = crypto_ahash_digestsize(tfm); + + switch (ctx->digsize) { + case SHA1_DIGEST_SIZE: + ctx->flags |= DRIVER_FLAGS_SHA1; + break; + case SHA256_DIGEST_SIZE: + ctx->flags |= DRIVER_FLAGS_SHA256; + break; + case SHA224_DIGEST_SIZE: + ctx->flags |= DRIVER_FLAGS_SHA224; + break; + case MD5_DIGEST_SIZE: + ctx->flags |= DRIVER_FLAGS_MD5; + break; + default: + return -EINVAL; + } + + ctx->bufcnt = 0; + ctx->offset = 0; + ctx->sent = 0; + ctx->total = req->nbytes; + ctx->sg = req->src; + ctx->sgfirst = req->src; + ctx->nents = sg_nents(ctx->sg); + + err = img_hash_handle_queue(tctx->hdev, req); + + return err; +} + +static int img_hash_cra_init(struct crypto_tfm *tfm) +{ + struct img_hash_ctx *ctx = crypto_tfm_ctx(tfm); + const char *alg_name = crypto_tfm_alg_name(tfm); + int err = -ENOMEM; + + ctx->fallback = crypto_alloc_ahash(alg_name, 0, + CRYPTO_ALG_NEED_FALLBACK); + if (IS_ERR(ctx->fallback)) { + pr_err("img_hash: Could not load fallback driver.\n"); + err = PTR_ERR(ctx->fallback); + goto err; + } + crypto_ahash_set_reqsize(__crypto_ahash_cast(tfm), + sizeof(struct img_hash_request_ctx) + + IMG_HASH_DMA_THRESHOLD); + + return 0; + +err: + return err; +} + +static void img_hash_cra_exit(struct crypto_tfm *tfm) +{ + struct img_hash_ctx *tctx = crypto_tfm_ctx(tfm); + + crypto_free_ahash(tctx->fallback); +} + +static irqreturn_t img_irq_handler(int irq, void *dev_id) +{ + struct img_hash_dev *hdev = dev_id; + u32 reg; + + reg = img_hash_read(hdev, CR_INTSTAT); + img_hash_write(hdev, CR_INTCLEAR, reg); + + if (reg & CR_INT_NEW_RESULTS_SET) { + dev_dbg(hdev->dev, "IRQ CR_INT_NEW_RESULTS_SET\n"); + if (DRIVER_FLAGS_BUSY & hdev->flags) { + hdev->flags |= DRIVER_FLAGS_OUTPUT_READY; + if (!(DRIVER_FLAGS_CPU & hdev->flags)) + hdev->flags |= DRIVER_FLAGS_DMA_READY; + tasklet_schedule(&hdev->done_task); + } else { + dev_warn(hdev->dev, + "HASH interrupt when no active requests.\n"); + } + } else if (reg & CR_INT_RESULTS_AVAILABLE) { + dev_warn(hdev->dev, + "IRQ triggered before the hash had completed\n"); + } else if (reg & CR_INT_RESULT_READ_ERR) { + dev_warn(hdev->dev, + "Attempt to read from an empty result queue\n"); + } else if (reg & CR_INT_MESSAGE_WRITE_ERROR) { + dev_warn(hdev->dev, + "Data written before the hardware was configured\n"); + } + return IRQ_HANDLED; +} + +static struct ahash_alg img_algs[] = { + { + .init = img_hash_init, + .update = img_hash_update, + .final = img_hash_final, + .finup = img_hash_finup, + .digest = img_hash_digest, + .halg = { + .digestsize = MD5_DIGEST_SIZE, + .base = { + .cra_name = "md5", + .cra_driver_name = "img-md5", + .cra_priority = 300, + .cra_flags = + CRYPTO_ALG_ASYNC | + CRYPTO_ALG_NEED_FALLBACK, + .cra_blocksize = MD5_HMAC_BLOCK_SIZE, + .cra_ctxsize = sizeof(struct img_hash_ctx), + .cra_init = img_hash_cra_init, + .cra_exit = img_hash_cra_exit, + .cra_module = THIS_MODULE, + } + } + }, + { + .init = img_hash_init, + .update = img_hash_update, + .final = img_hash_final, + .finup = img_hash_finup, + .digest = img_hash_digest, + .halg = { + .digestsize = SHA1_DIGEST_SIZE, + .base = { + .cra_name = "sha1", + .cra_driver_name = "img-sha1", + .cra_priority = 300, + .cra_flags = + CRYPTO_ALG_ASYNC | + CRYPTO_ALG_NEED_FALLBACK, + .cra_blocksize = SHA1_BLOCK_SIZE, + .cra_ctxsize = sizeof(struct img_hash_ctx), + .cra_init = img_hash_cra_init, + .cra_exit = img_hash_cra_exit, + .cra_module = THIS_MODULE, + } + } + }, + { + .init = img_hash_init, + .update = img_hash_update, + .final = img_hash_final, + .finup = img_hash_finup, + .digest = img_hash_digest, + .halg = { + .digestsize = SHA224_DIGEST_SIZE, + .base = { + .cra_name = "sha224", + .cra_driver_name = "img-sha224", + .cra_priority = 300, + .cra_flags = + CRYPTO_ALG_ASYNC | + CRYPTO_ALG_NEED_FALLBACK, + .cra_blocksize = SHA224_BLOCK_SIZE, + .cra_ctxsize = sizeof(struct img_hash_ctx), + .cra_init = img_hash_cra_init, + .cra_exit = img_hash_cra_exit, + .cra_module = THIS_MODULE, + } + } + }, + { + .init = img_hash_init, + .update = img_hash_update, + .final = img_hash_final, + .finup = img_hash_finup, + .digest = img_hash_digest, + .halg = { + .digestsize = SHA256_DIGEST_SIZE, + .base = { + .cra_name = "sha256", + .cra_driver_name = "img-sha256", + .cra_priority = 300, + .cra_flags = + CRYPTO_ALG_ASYNC | + CRYPTO_ALG_NEED_FALLBACK, + .cra_blocksize = SHA256_BLOCK_SIZE, + .cra_ctxsize = sizeof(struct img_hash_ctx), + .cra_init = img_hash_cra_init, + .cra_exit = img_hash_cra_exit, + .cra_module = THIS_MODULE, + } + } + } +}; + +static int img_register_algs(struct img_hash_dev *hdev) +{ + int i, err; + + for (i = 0; i < ARRAY_SIZE(img_algs); i++) { + err = crypto_register_ahash(&img_algs[i]); + if (err) + goto err_reg; + } + return 0; + +err_reg: + for (; i--; ) + crypto_unregister_ahash(&img_algs[i]); + + return err; +} + +static int img_unregister_algs(struct img_hash_dev *hdev) +{ + int i; + + for (i = 0; i < ARRAY_SIZE(img_algs); i++) + crypto_unregister_ahash(&img_algs[i]); + return 0; +} + +static void img_hash_done_task(unsigned long data) +{ + struct img_hash_dev *hdev = (struct img_hash_dev *)data; + int err = 0; + + if (hdev->err == -EINVAL) { + err = hdev->err; + goto finish; + } + + if (!(DRIVER_FLAGS_BUSY & hdev->flags)) { + img_hash_handle_queue(hdev, NULL); + return; + } + + if (DRIVER_FLAGS_CPU & hdev->flags) { + if (DRIVER_FLAGS_OUTPUT_READY & hdev->flags) { + hdev->flags &= ~DRIVER_FLAGS_OUTPUT_READY; + goto finish; + } + } else if (DRIVER_FLAGS_DMA_READY & hdev->flags) { + if (DRIVER_FLAGS_DMA_ACTIVE & hdev->flags) { + hdev->flags &= ~DRIVER_FLAGS_DMA_ACTIVE; + img_hash_write_via_dma_stop(hdev); + if (hdev->err) { + err = hdev->err; + goto finish; + } + } + if (DRIVER_FLAGS_OUTPUT_READY & hdev->flags) { + hdev->flags &= ~(DRIVER_FLAGS_DMA_READY | + DRIVER_FLAGS_OUTPUT_READY); + goto finish; + } + } + return; + +finish: + img_hash_finish_req(hdev->req, err); +} + +static const struct of_device_id img_hash_match[] = { + { .compatible = "img,hash-accelerator" }, + {} +}; +MODULE_DEVICE_TABLE(of, img_hash_match); + +static int img_hash_probe(struct platform_device *pdev) +{ + struct img_hash_dev *hdev; + struct device *dev = &pdev->dev; + struct resource *hash_res; + int irq; + int err; + + hdev = devm_kzalloc(dev, sizeof(*hdev), GFP_KERNEL); + if (hdev == NULL) + return -ENOMEM; + + spin_lock_init(&hdev->lock); + + hdev->dev = dev; + + platform_set_drvdata(pdev, hdev); + + INIT_LIST_HEAD(&hdev->list); + + tasklet_init(&hdev->done_task, img_hash_done_task, (unsigned long)hdev); + tasklet_init(&hdev->dma_task, img_hash_dma_task, (unsigned long)hdev); + + crypto_init_queue(&hdev->queue, IMG_HASH_QUEUE_LENGTH); + + /* Register bank */ + hash_res = platform_get_resource(pdev, IORESOURCE_MEM, 0); + + hdev->io_base = devm_ioremap_resource(dev, hash_res); + if (IS_ERR(hdev->io_base)) { + err = PTR_ERR(hdev->io_base); + dev_err(dev, "can't ioremap, returned %d\n", err); + + goto res_err; + } + + /* Write port (DMA or CPU) */ + hash_res = platform_get_resource(pdev, IORESOURCE_MEM, 1); + hdev->cpu_addr = devm_ioremap_resource(dev, hash_res); + if (IS_ERR(hdev->cpu_addr)) { + dev_err(dev, "can't ioremap write port\n"); + err = PTR_ERR(hdev->cpu_addr); + goto res_err; + } + hdev->bus_addr = hash_res->start; + + irq = platform_get_irq(pdev, 0); + if (irq < 0) { + dev_err(dev, "no IRQ resource info\n"); + err = irq; + goto res_err; + } + + err = devm_request_irq(dev, irq, img_irq_handler, 0, + dev_name(dev), hdev); + if (err) { + dev_err(dev, "unable to request irq\n"); + goto res_err; + } + dev_dbg(dev, "using IRQ channel %d\n", irq); + + hdev->hash_clk = devm_clk_get(&pdev->dev, "hash"); + if (IS_ERR(hdev->hash_clk)) { + dev_err(dev, "clock initialization failed.\n"); + err = PTR_ERR(hdev->hash_clk); + goto res_err; + } + + hdev->sys_clk = devm_clk_get(&pdev->dev, "sys"); + if (IS_ERR(hdev->sys_clk)) { + dev_err(dev, "clock initialization failed.\n"); + err = PTR_ERR(hdev->sys_clk); + goto res_err; + } + + err = clk_prepare_enable(hdev->hash_clk); + if (err) + goto res_err; + + err = clk_prepare_enable(hdev->sys_clk); + if (err) + goto clk_err; + + err = img_hash_dma_init(hdev); + if (err) + goto dma_err; + + dev_dbg(dev, "using %s for DMA transfers\n", + dma_chan_name(hdev->dma_lch)); + + spin_lock(&img_hash.lock); + list_add_tail(&hdev->list, &img_hash.dev_list); + spin_unlock(&img_hash.lock); + + err = img_register_algs(hdev); + if (err) + goto err_algs; + dev_dbg(dev, "Img MD5/SHA1/SHA224/SHA256 Hardware accelerator initialized\n"); + + return 0; + +err_algs: + spin_lock(&img_hash.lock); + list_del(&hdev->list); + spin_unlock(&img_hash.lock); + dma_release_channel(hdev->dma_lch); +dma_err: + clk_disable_unprepare(hdev->sys_clk); +clk_err: + clk_disable_unprepare(hdev->hash_clk); +res_err: + tasklet_kill(&hdev->done_task); + tasklet_kill(&hdev->dma_task); + + return err; +} + +static int img_hash_remove(struct platform_device *pdev) +{ + static struct img_hash_dev *hdev; + + hdev = platform_get_drvdata(pdev); + spin_lock(&img_hash.lock); + list_del(&hdev->list); + spin_unlock(&img_hash.lock); + + img_unregister_algs(hdev); + + tasklet_kill(&hdev->done_task); + tasklet_kill(&hdev->dma_task); + + dma_release_channel(hdev->dma_lch); + + clk_disable_unprepare(hdev->hash_clk); + clk_disable_unprepare(hdev->sys_clk); + + return 0; +} + +static struct platform_driver img_hash_driver = { + .probe = img_hash_probe, + .remove = img_hash_remove, + .driver = { + .name = "img-hash-accelerator", + .of_match_table = of_match_ptr(img_hash_match), + } +}; +module_platform_driver(img_hash_driver); + +MODULE_LICENSE("GPL v2"); +MODULE_DESCRIPTION("Imgtec SHA1/224/256 & MD5 hw accelerator driver"); +MODULE_AUTHOR("Will Thomas."); +MODULE_AUTHOR("James Hartley <james.hartley@imgtec.com>"); diff --git a/drivers/crypto/mxs-dcp.c b/drivers/crypto/mxs-dcp.c index 829d6394fb33..59ed54e464a9 100644 --- a/drivers/crypto/mxs-dcp.c +++ b/drivers/crypto/mxs-dcp.c @@ -153,7 +153,7 @@ static int mxs_dcp_start_dma(struct dcp_async_ctx *actx) struct dcp *sdcp = global_sdcp; const int chan = actx->chan; uint32_t stat; - int ret; + unsigned long ret; struct dcp_dma_desc *desc = &sdcp->coh->desc[actx->chan]; dma_addr_t desc_phys = dma_map_single(sdcp->dev, desc, sizeof(*desc), diff --git a/drivers/crypto/omap-aes.c b/drivers/crypto/omap-aes.c index 42f95a4326b0..9a28b7e07c71 100644 --- a/drivers/crypto/omap-aes.c +++ b/drivers/crypto/omap-aes.c @@ -554,15 +554,23 @@ static int omap_aes_crypt_dma_stop(struct omap_aes_dev *dd) return err; } -static int omap_aes_check_aligned(struct scatterlist *sg) +static int omap_aes_check_aligned(struct scatterlist *sg, int total) { + int len = 0; + while (sg) { if (!IS_ALIGNED(sg->offset, 4)) return -1; if (!IS_ALIGNED(sg->length, AES_BLOCK_SIZE)) return -1; + + len += sg->length; sg = sg_next(sg); } + + if (len != total) + return -1; + return 0; } @@ -633,8 +641,8 @@ static int omap_aes_handle_queue(struct omap_aes_dev *dd, dd->in_sg = req->src; dd->out_sg = req->dst; - if (omap_aes_check_aligned(dd->in_sg) || - omap_aes_check_aligned(dd->out_sg)) { + if (omap_aes_check_aligned(dd->in_sg, dd->total) || + omap_aes_check_aligned(dd->out_sg, dd->total)) { if (omap_aes_copy_sgs(dd)) pr_err("Failed to copy SGs for unaligned cases\n"); dd->sgs_copied = 1; diff --git a/drivers/crypto/omap-sham.c b/drivers/crypto/omap-sham.c index 3c76696ee578..4d63e0d4da9a 100644 --- a/drivers/crypto/omap-sham.c +++ b/drivers/crypto/omap-sham.c @@ -640,6 +640,7 @@ static size_t omap_sham_append_sg(struct omap_sham_reqctx *ctx) while (ctx->sg) { vaddr = kmap_atomic(sg_page(ctx->sg)); + vaddr += ctx->sg->offset; count = omap_sham_append_buffer(ctx, vaddr + ctx->offset, @@ -1945,6 +1946,7 @@ static int omap_sham_probe(struct platform_device *pdev) dd->flags |= dd->pdata->flags; pm_runtime_enable(dev); + pm_runtime_irq_safe(dev); pm_runtime_get_sync(dev); rev = omap_sham_read(dd, SHA_REG_REV(dd)); pm_runtime_put_sync(&pdev->dev); diff --git a/drivers/crypto/qat/qat_common/adf_accel_devices.h b/drivers/crypto/qat/qat_common/adf_accel_devices.h index 19c0efa29ab3..f22ce7169fa5 100644 --- a/drivers/crypto/qat/qat_common/adf_accel_devices.h +++ b/drivers/crypto/qat/qat_common/adf_accel_devices.h @@ -52,7 +52,6 @@ #include <linux/io.h> #include "adf_cfg_common.h" -#define PCI_VENDOR_ID_INTEL 0x8086 #define ADF_DH895XCC_DEVICE_NAME "dh895xcc" #define ADF_DH895XCC_PCI_DEVICE_ID 0x435 #define ADF_PCI_MAX_BARS 3 diff --git a/drivers/crypto/qat/qat_common/adf_accel_engine.c b/drivers/crypto/qat/qat_common/adf_accel_engine.c index c77453b900a3..7f8b66c915ed 100644 --- a/drivers/crypto/qat/qat_common/adf_accel_engine.c +++ b/drivers/crypto/qat/qat_common/adf_accel_engine.c @@ -60,36 +60,40 @@ int adf_ae_fw_load(struct adf_accel_dev *accel_dev) if (request_firmware(&loader_data->uof_fw, hw_device->fw_name, &accel_dev->accel_pci_dev.pci_dev->dev)) { - pr_err("QAT: Failed to load firmware %s\n", hw_device->fw_name); + dev_err(&GET_DEV(accel_dev), "Failed to load firmware %s\n", + hw_device->fw_name); return -EFAULT; } uof_size = loader_data->uof_fw->size; uof_addr = (void *)loader_data->uof_fw->data; if (qat_uclo_map_uof_obj(loader_data->fw_loader, uof_addr, uof_size)) { - pr_err("QAT: Failed to map UOF\n"); + dev_err(&GET_DEV(accel_dev), "Failed to map UOF\n"); goto out_err; } if (qat_uclo_wr_all_uimage(loader_data->fw_loader)) { - pr_err("QAT: Failed to map UOF\n"); + dev_err(&GET_DEV(accel_dev), "Failed to map UOF\n"); goto out_err; } return 0; out_err: - release_firmware(loader_data->uof_fw); + adf_ae_fw_release(accel_dev); return -EFAULT; } -int adf_ae_fw_release(struct adf_accel_dev *accel_dev) +void adf_ae_fw_release(struct adf_accel_dev *accel_dev) { struct adf_fw_loader_data *loader_data = accel_dev->fw_loader; - release_firmware(loader_data->uof_fw); qat_uclo_del_uof_obj(loader_data->fw_loader); qat_hal_deinit(loader_data->fw_loader); + + if (loader_data->uof_fw) + release_firmware(loader_data->uof_fw); + + loader_data->uof_fw = NULL; loader_data->fw_loader = NULL; - return 0; } int adf_ae_start(struct adf_accel_dev *accel_dev) @@ -104,8 +108,9 @@ int adf_ae_start(struct adf_accel_dev *accel_dev) ae_ctr++; } } - pr_info("QAT: qat_dev%d started %d acceleration engines\n", - accel_dev->accel_id, ae_ctr); + dev_info(&GET_DEV(accel_dev), + "qat_dev%d started %d acceleration engines\n", + accel_dev->accel_id, ae_ctr); return 0; } @@ -121,8 +126,9 @@ int adf_ae_stop(struct adf_accel_dev *accel_dev) ae_ctr++; } } - pr_info("QAT: qat_dev%d stopped %d acceleration engines\n", - accel_dev->accel_id, ae_ctr); + dev_info(&GET_DEV(accel_dev), + "qat_dev%d stopped %d acceleration engines\n", + accel_dev->accel_id, ae_ctr); return 0; } @@ -147,12 +153,12 @@ int adf_ae_init(struct adf_accel_dev *accel_dev) accel_dev->fw_loader = loader_data; if (qat_hal_init(accel_dev)) { - pr_err("QAT: Failed to init the AEs\n"); + dev_err(&GET_DEV(accel_dev), "Failed to init the AEs\n"); kfree(loader_data); return -EFAULT; } if (adf_ae_reset(accel_dev, 0)) { - pr_err("QAT: Failed to reset the AEs\n"); + dev_err(&GET_DEV(accel_dev), "Failed to reset the AEs\n"); qat_hal_deinit(loader_data->fw_loader); kfree(loader_data); return -EFAULT; @@ -162,6 +168,9 @@ int adf_ae_init(struct adf_accel_dev *accel_dev) int adf_ae_shutdown(struct adf_accel_dev *accel_dev) { + struct adf_fw_loader_data *loader_data = accel_dev->fw_loader; + + qat_hal_deinit(loader_data->fw_loader); kfree(accel_dev->fw_loader); accel_dev->fw_loader = NULL; return 0; diff --git a/drivers/crypto/qat/qat_common/adf_aer.c b/drivers/crypto/qat/qat_common/adf_aer.c index fa1fef824de2..2dbc733b8ab2 100644 --- a/drivers/crypto/qat/qat_common/adf_aer.c +++ b/drivers/crypto/qat/qat_common/adf_aer.c @@ -60,14 +60,14 @@ static pci_ers_result_t adf_error_detected(struct pci_dev *pdev, { struct adf_accel_dev *accel_dev = adf_devmgr_pci_to_accel_dev(pdev); - pr_info("QAT: Acceleration driver hardware error detected.\n"); + dev_info(&pdev->dev, "Acceleration driver hardware error detected.\n"); if (!accel_dev) { - pr_err("QAT: Can't find acceleration device\n"); + dev_err(&pdev->dev, "Can't find acceleration device\n"); return PCI_ERS_RESULT_DISCONNECT; } if (state == pci_channel_io_perm_failure) { - pr_err("QAT: Can't recover from device error\n"); + dev_err(&pdev->dev, "Can't recover from device error\n"); return PCI_ERS_RESULT_DISCONNECT; } @@ -88,10 +88,12 @@ static void adf_dev_restore(struct adf_accel_dev *accel_dev) struct pci_dev *parent = pdev->bus->self; uint16_t bridge_ctl = 0; - pr_info("QAT: Resetting device qat_dev%d\n", accel_dev->accel_id); + dev_info(&GET_DEV(accel_dev), "Resetting device qat_dev%d\n", + accel_dev->accel_id); if (!pci_wait_for_pending_transaction(pdev)) - pr_info("QAT: Transaction still in progress. Proceeding\n"); + dev_info(&GET_DEV(accel_dev), + "Transaction still in progress. Proceeding\n"); pci_read_config_word(parent, PCI_BRIDGE_CONTROL, &bridge_ctl); bridge_ctl |= PCI_BRIDGE_CTL_BUS_RESET; @@ -158,7 +160,8 @@ static int adf_dev_aer_schedule_reset(struct adf_accel_dev *accel_dev, unsigned long timeout = wait_for_completion_timeout( &reset_data->compl, wait_jiffies); if (!timeout) { - pr_err("QAT: Reset device timeout expired\n"); + dev_err(&GET_DEV(accel_dev), + "Reset device timeout expired\n"); ret = -EFAULT; } kfree(reset_data); @@ -184,8 +187,8 @@ static pci_ers_result_t adf_slot_reset(struct pci_dev *pdev) static void adf_resume(struct pci_dev *pdev) { - pr_info("QAT: Acceleration driver reset completed\n"); - pr_info("QAT: Device is up and runnig\n"); + dev_info(&pdev->dev, "Acceleration driver reset completed\n"); + dev_info(&pdev->dev, "Device is up and runnig\n"); } static struct pci_error_handlers adf_err_handler = { @@ -236,7 +239,7 @@ EXPORT_SYMBOL_GPL(adf_disable_aer); int adf_init_aer(void) { device_reset_wq = create_workqueue("qat_device_reset_wq"); - return (device_reset_wq == NULL) ? -EFAULT : 0; + return !device_reset_wq ? -EFAULT : 0; } void adf_exit_aer(void) diff --git a/drivers/crypto/qat/qat_common/adf_cfg.c b/drivers/crypto/qat/qat_common/adf_cfg.c index de16da9070a5..ab65bc274561 100644 --- a/drivers/crypto/qat/qat_common/adf_cfg.c +++ b/drivers/crypto/qat/qat_common/adf_cfg.c @@ -142,7 +142,8 @@ int adf_cfg_dev_add(struct adf_accel_dev *accel_dev) dev_cfg_data, &qat_dev_cfg_fops); if (!dev_cfg_data->debug) { - pr_err("QAT: Failed to create qat cfg debugfs entry.\n"); + dev_err(&GET_DEV(accel_dev), + "Failed to create qat cfg debugfs entry.\n"); kfree(dev_cfg_data); accel_dev->cfg = NULL; return -EFAULT; @@ -305,7 +306,7 @@ int adf_cfg_add_key_value_param(struct adf_accel_dev *accel_dev, snprintf(key_val->val, ADF_CFG_MAX_VAL_LEN_IN_BYTES, "0x%lx", (unsigned long)val); } else { - pr_err("QAT: Unknown type given.\n"); + dev_err(&GET_DEV(accel_dev), "Unknown type given.\n"); kfree(key_val); return -1; } diff --git a/drivers/crypto/qat/qat_common/adf_cfg_strings.h b/drivers/crypto/qat/qat_common/adf_cfg_strings.h index c7ac758ebc90..13575111382c 100644 --- a/drivers/crypto/qat/qat_common/adf_cfg_strings.h +++ b/drivers/crypto/qat/qat_common/adf_cfg_strings.h @@ -59,7 +59,7 @@ #define ADF_RING_SYM_TX "RingSymTx" #define ADF_RING_RND_TX "RingNrbgTx" #define ADF_RING_ASYM_RX "RingAsymRx" -#define ADF_RING_SYM_RX "RinSymRx" +#define ADF_RING_SYM_RX "RingSymRx" #define ADF_RING_RND_RX "RingNrbgRx" #define ADF_RING_DC_TX "RingTx" #define ADF_RING_DC_RX "RingRx" @@ -69,15 +69,15 @@ #define ADF_DC "Dc" #define ADF_ETRMGR_COALESCING_ENABLED "InterruptCoalescingEnabled" #define ADF_ETRMGR_COALESCING_ENABLED_FORMAT \ - ADF_ETRMGR_BANK"%d"ADF_ETRMGR_COALESCING_ENABLED + ADF_ETRMGR_BANK "%d" ADF_ETRMGR_COALESCING_ENABLED #define ADF_ETRMGR_COALESCE_TIMER "InterruptCoalescingTimerNs" #define ADF_ETRMGR_COALESCE_TIMER_FORMAT \ - ADF_ETRMGR_BANK"%d"ADF_ETRMGR_COALESCE_TIMER + ADF_ETRMGR_BANK "%d" ADF_ETRMGR_COALESCE_TIMER #define ADF_ETRMGR_COALESCING_MSG_ENABLED "InterruptCoalescingNumResponses" #define ADF_ETRMGR_COALESCING_MSG_ENABLED_FORMAT \ - ADF_ETRMGR_BANK"%d"ADF_ETRMGR_COALESCING_MSG_ENABLED + ADF_ETRMGR_BANK "%d" ADF_ETRMGR_COALESCING_MSG_ENABLED #define ADF_ETRMGR_CORE_AFFINITY "CoreAffinity" #define ADF_ETRMGR_CORE_AFFINITY_FORMAT \ - ADF_ETRMGR_BANK"%d"ADF_ETRMGR_CORE_AFFINITY + ADF_ETRMGR_BANK "%d" ADF_ETRMGR_CORE_AFFINITY #define ADF_ACCEL_STR "Accelerator%d" #endif diff --git a/drivers/crypto/qat/qat_common/adf_common_drv.h b/drivers/crypto/qat/qat_common/adf_common_drv.h index a62e485c8786..0666ee6a3360 100644 --- a/drivers/crypto/qat/qat_common/adf_common_drv.h +++ b/drivers/crypto/qat/qat_common/adf_common_drv.h @@ -115,7 +115,7 @@ int adf_dev_restarted_notify(struct adf_accel_dev *accel_dev); int adf_ae_init(struct adf_accel_dev *accel_dev); int adf_ae_shutdown(struct adf_accel_dev *accel_dev); int adf_ae_fw_load(struct adf_accel_dev *accel_dev); -int adf_ae_fw_release(struct adf_accel_dev *accel_dev); +void adf_ae_fw_release(struct adf_accel_dev *accel_dev); int adf_ae_start(struct adf_accel_dev *accel_dev); int adf_ae_stop(struct adf_accel_dev *accel_dev); diff --git a/drivers/crypto/qat/qat_common/adf_ctl_drv.c b/drivers/crypto/qat/qat_common/adf_ctl_drv.c index 74207a6f0516..cb5f066e93a6 100644 --- a/drivers/crypto/qat/qat_common/adf_ctl_drv.c +++ b/drivers/crypto/qat/qat_common/adf_ctl_drv.c @@ -77,14 +77,14 @@ struct adf_ctl_drv_info { struct class *drv_class; }; -static struct adf_ctl_drv_info adt_ctl_drv; +static struct adf_ctl_drv_info adf_ctl_drv; static void adf_chr_drv_destroy(void) { - device_destroy(adt_ctl_drv.drv_class, MKDEV(adt_ctl_drv.major, 0)); - cdev_del(&adt_ctl_drv.drv_cdev); - class_destroy(adt_ctl_drv.drv_class); - unregister_chrdev_region(MKDEV(adt_ctl_drv.major, 0), 1); + device_destroy(adf_ctl_drv.drv_class, MKDEV(adf_ctl_drv.major, 0)); + cdev_del(&adf_ctl_drv.drv_cdev); + class_destroy(adf_ctl_drv.drv_class); + unregister_chrdev_region(MKDEV(adf_ctl_drv.major, 0), 1); } static int adf_chr_drv_create(void) @@ -97,20 +97,20 @@ static int adf_chr_drv_create(void) return -EFAULT; } - adt_ctl_drv.drv_class = class_create(THIS_MODULE, DEVICE_NAME); - if (IS_ERR(adt_ctl_drv.drv_class)) { + adf_ctl_drv.drv_class = class_create(THIS_MODULE, DEVICE_NAME); + if (IS_ERR(adf_ctl_drv.drv_class)) { pr_err("QAT: class_create failed for adf_ctl\n"); goto err_chrdev_unreg; } - adt_ctl_drv.major = MAJOR(dev_id); - cdev_init(&adt_ctl_drv.drv_cdev, &adf_ctl_ops); - if (cdev_add(&adt_ctl_drv.drv_cdev, dev_id, 1)) { + adf_ctl_drv.major = MAJOR(dev_id); + cdev_init(&adf_ctl_drv.drv_cdev, &adf_ctl_ops); + if (cdev_add(&adf_ctl_drv.drv_cdev, dev_id, 1)) { pr_err("QAT: cdev add failed\n"); goto err_class_destr; } - drv_device = device_create(adt_ctl_drv.drv_class, NULL, - MKDEV(adt_ctl_drv.major, 0), + drv_device = device_create(adf_ctl_drv.drv_class, NULL, + MKDEV(adf_ctl_drv.major, 0), NULL, DEVICE_NAME); if (IS_ERR(drv_device)) { pr_err("QAT: failed to create device\n"); @@ -118,9 +118,9 @@ static int adf_chr_drv_create(void) } return 0; err_cdev_del: - cdev_del(&adt_ctl_drv.drv_cdev); + cdev_del(&adf_ctl_drv.drv_cdev); err_class_destr: - class_destroy(adt_ctl_drv.drv_class); + class_destroy(adf_ctl_drv.drv_class); err_chrdev_unreg: unregister_chrdev_region(dev_id, 1); return -EFAULT; @@ -159,14 +159,16 @@ static int adf_add_key_value_data(struct adf_accel_dev *accel_dev, if (adf_cfg_add_key_value_param(accel_dev, section, key_val->key, (void *)val, key_val->type)) { - pr_err("QAT: failed to add keyvalue.\n"); + dev_err(&GET_DEV(accel_dev), + "failed to add hex keyvalue.\n"); return -EFAULT; } } else { if (adf_cfg_add_key_value_param(accel_dev, section, key_val->key, key_val->val, key_val->type)) { - pr_err("QAT: failed to add keyvalue.\n"); + dev_err(&GET_DEV(accel_dev), + "failed to add keyvalue.\n"); return -EFAULT; } } @@ -185,12 +187,14 @@ static int adf_copy_key_value_data(struct adf_accel_dev *accel_dev, while (section_head) { if (copy_from_user(§ion, (void __user *)section_head, sizeof(*section_head))) { - pr_err("QAT: failed to copy section info\n"); + dev_err(&GET_DEV(accel_dev), + "failed to copy section info\n"); goto out_err; } if (adf_cfg_section_add(accel_dev, section.name)) { - pr_err("QAT: failed to add section.\n"); + dev_err(&GET_DEV(accel_dev), + "failed to add section.\n"); goto out_err; } @@ -199,7 +203,8 @@ static int adf_copy_key_value_data(struct adf_accel_dev *accel_dev, while (params_head) { if (copy_from_user(&key_val, (void __user *)params_head, sizeof(key_val))) { - pr_err("QAT: Failed to copy keyvalue.\n"); + dev_err(&GET_DEV(accel_dev), + "Failed to copy keyvalue.\n"); goto out_err; } if (adf_add_key_value_data(accel_dev, section.name, @@ -258,8 +263,9 @@ static int adf_ctl_is_device_in_use(int id) if (id == dev->accel_id || id == ADF_CFG_ALL_DEVICES) { if (adf_devmgr_in_reset(dev) || adf_dev_in_use(dev)) { - pr_info("QAT: device qat_dev%d is busy\n", - dev->accel_id); + dev_info(&GET_DEV(dev), + "device qat_dev%d is busy\n", + dev->accel_id); return -EBUSY; } } @@ -280,7 +286,8 @@ static int adf_ctl_stop_devices(uint32_t id) continue; if (adf_dev_stop(accel_dev)) { - pr_err("QAT: Failed to stop qat_dev%d\n", id); + dev_err(&GET_DEV(accel_dev), + "Failed to stop qat_dev%d\n", id); ret = -EFAULT; } else { adf_dev_shutdown(accel_dev); @@ -343,17 +350,20 @@ static int adf_ctl_ioctl_dev_start(struct file *fp, unsigned int cmd, } if (!adf_dev_started(accel_dev)) { - pr_info("QAT: Starting acceleration device qat_dev%d.\n", - ctl_data->device_id); + dev_info(&GET_DEV(accel_dev), + "Starting acceleration device qat_dev%d.\n", + ctl_data->device_id); ret = adf_dev_init(accel_dev); if (!ret) ret = adf_dev_start(accel_dev); } else { - pr_info("QAT: Acceleration device qat_dev%d already started.\n", - ctl_data->device_id); + dev_info(&GET_DEV(accel_dev), + "Acceleration device qat_dev%d already started.\n", + ctl_data->device_id); } if (ret) { - pr_err("QAT: Failed to start qat_dev%d\n", ctl_data->device_id); + dev_err(&GET_DEV(accel_dev), "Failed to start qat_dev%d\n", + ctl_data->device_id); adf_dev_stop(accel_dev); adf_dev_shutdown(accel_dev); } @@ -408,7 +418,7 @@ static int adf_ctl_ioctl_get_status(struct file *fp, unsigned int cmd, if (copy_to_user((void __user *)arg, &dev_info, sizeof(struct adf_dev_status_info))) { - pr_err("QAT: failed to copy status.\n"); + dev_err(&GET_DEV(accel_dev), "failed to copy status.\n"); return -EFAULT; } return 0; diff --git a/drivers/crypto/qat/qat_common/adf_dev_mgr.c b/drivers/crypto/qat/qat_common/adf_dev_mgr.c index 4a0a829d4500..3f0ff9e7d840 100644 --- a/drivers/crypto/qat/qat_common/adf_dev_mgr.c +++ b/drivers/crypto/qat/qat_common/adf_dev_mgr.c @@ -67,7 +67,8 @@ int adf_devmgr_add_dev(struct adf_accel_dev *accel_dev) struct list_head *itr; if (num_devices == ADF_MAX_DEVICES) { - pr_err("QAT: Only support up to %d devices\n", ADF_MAX_DEVICES); + dev_err(&GET_DEV(accel_dev), "Only support up to %d devices\n", + ADF_MAX_DEVICES); return -EFAULT; } diff --git a/drivers/crypto/qat/qat_common/adf_init.c b/drivers/crypto/qat/qat_common/adf_init.c index 8f0ca498ab87..245f43237a2d 100644 --- a/drivers/crypto/qat/qat_common/adf_init.c +++ b/drivers/crypto/qat/qat_common/adf_init.c @@ -124,12 +124,12 @@ int adf_dev_init(struct adf_accel_dev *accel_dev) if (!hw_data) { dev_err(&GET_DEV(accel_dev), - "QAT: Failed to init device - hw_data not set\n"); + "Failed to init device - hw_data not set\n"); return -EFAULT; } if (!test_bit(ADF_STATUS_CONFIGURED, &accel_dev->status)) { - pr_info("QAT: Device not configured\n"); + dev_err(&GET_DEV(accel_dev), "Device not configured\n"); return -EFAULT; } @@ -151,20 +151,21 @@ int adf_dev_init(struct adf_accel_dev *accel_dev) hw_data->enable_ints(accel_dev); if (adf_ae_init(accel_dev)) { - pr_err("QAT: Failed to initialise Acceleration Engine\n"); + dev_err(&GET_DEV(accel_dev), + "Failed to initialise Acceleration Engine\n"); return -EFAULT; } set_bit(ADF_STATUS_AE_INITIALISED, &accel_dev->status); if (adf_ae_fw_load(accel_dev)) { - pr_err("QAT: Failed to load acceleration FW\n"); - adf_ae_fw_release(accel_dev); + dev_err(&GET_DEV(accel_dev), + "Failed to load acceleration FW\n"); return -EFAULT; } set_bit(ADF_STATUS_AE_UCODE_LOADED, &accel_dev->status); if (hw_data->alloc_irq(accel_dev)) { - pr_err("QAT: Failed to allocate interrupts\n"); + dev_err(&GET_DEV(accel_dev), "Failed to allocate interrupts\n"); return -EFAULT; } set_bit(ADF_STATUS_IRQ_ALLOCATED, &accel_dev->status); @@ -179,8 +180,9 @@ int adf_dev_init(struct adf_accel_dev *accel_dev) if (!service->admin) continue; if (service->event_hld(accel_dev, ADF_EVENT_INIT)) { - pr_err("QAT: Failed to initialise service %s\n", - service->name); + dev_err(&GET_DEV(accel_dev), + "Failed to initialise service %s\n", + service->name); return -EFAULT; } set_bit(accel_dev->accel_id, &service->init_status); @@ -190,8 +192,9 @@ int adf_dev_init(struct adf_accel_dev *accel_dev) if (service->admin) continue; if (service->event_hld(accel_dev, ADF_EVENT_INIT)) { - pr_err("QAT: Failed to initialise service %s\n", - service->name); + dev_err(&GET_DEV(accel_dev), + "Failed to initialise service %s\n", + service->name); return -EFAULT; } set_bit(accel_dev->accel_id, &service->init_status); @@ -221,7 +224,7 @@ int adf_dev_start(struct adf_accel_dev *accel_dev) set_bit(ADF_STATUS_STARTING, &accel_dev->status); if (adf_ae_start(accel_dev)) { - pr_err("QAT: AE Start Failed\n"); + dev_err(&GET_DEV(accel_dev), "AE Start Failed\n"); return -EFAULT; } set_bit(ADF_STATUS_AE_STARTED, &accel_dev->status); @@ -231,8 +234,9 @@ int adf_dev_start(struct adf_accel_dev *accel_dev) if (!service->admin) continue; if (service->event_hld(accel_dev, ADF_EVENT_START)) { - pr_err("QAT: Failed to start service %s\n", - service->name); + dev_err(&GET_DEV(accel_dev), + "Failed to start service %s\n", + service->name); return -EFAULT; } set_bit(accel_dev->accel_id, &service->start_status); @@ -242,8 +246,9 @@ int adf_dev_start(struct adf_accel_dev *accel_dev) if (service->admin) continue; if (service->event_hld(accel_dev, ADF_EVENT_START)) { - pr_err("QAT: Failed to start service %s\n", - service->name); + dev_err(&GET_DEV(accel_dev), + "Failed to start service %s\n", + service->name); return -EFAULT; } set_bit(accel_dev->accel_id, &service->start_status); @@ -253,7 +258,8 @@ int adf_dev_start(struct adf_accel_dev *accel_dev) set_bit(ADF_STATUS_STARTED, &accel_dev->status); if (qat_algs_register()) { - pr_err("QAT: Failed to register crypto algs\n"); + dev_err(&GET_DEV(accel_dev), + "Failed to register crypto algs\n"); set_bit(ADF_STATUS_STARTING, &accel_dev->status); clear_bit(ADF_STATUS_STARTED, &accel_dev->status); return -EFAULT; @@ -287,7 +293,8 @@ int adf_dev_stop(struct adf_accel_dev *accel_dev) clear_bit(ADF_STATUS_STARTED, &accel_dev->status); if (qat_algs_unregister()) - pr_err("QAT: Failed to unregister crypto algs\n"); + dev_err(&GET_DEV(accel_dev), + "Failed to unregister crypto algs\n"); list_for_each(list_itr, &service_table) { service = list_entry(list_itr, struct service_hndl, list); @@ -310,8 +317,9 @@ int adf_dev_stop(struct adf_accel_dev *accel_dev) if (!test_bit(accel_dev->accel_id, &service->start_status)) continue; if (service->event_hld(accel_dev, ADF_EVENT_STOP)) - pr_err("QAT: Failed to shutdown service %s\n", - service->name); + dev_err(&GET_DEV(accel_dev), + "Failed to shutdown service %s\n", + service->name); else clear_bit(accel_dev->accel_id, &service->start_status); } @@ -321,7 +329,7 @@ int adf_dev_stop(struct adf_accel_dev *accel_dev) if (test_bit(ADF_STATUS_AE_STARTED, &accel_dev->status)) { if (adf_ae_stop(accel_dev)) - pr_err("QAT: failed to stop AE\n"); + dev_err(&GET_DEV(accel_dev), "failed to stop AE\n"); else clear_bit(ADF_STATUS_AE_STARTED, &accel_dev->status); } @@ -350,16 +358,14 @@ void adf_dev_shutdown(struct adf_accel_dev *accel_dev) } if (test_bit(ADF_STATUS_AE_UCODE_LOADED, &accel_dev->status)) { - if (adf_ae_fw_release(accel_dev)) - pr_err("QAT: Failed to release the ucode\n"); - else - clear_bit(ADF_STATUS_AE_UCODE_LOADED, - &accel_dev->status); + adf_ae_fw_release(accel_dev); + clear_bit(ADF_STATUS_AE_UCODE_LOADED, &accel_dev->status); } if (test_bit(ADF_STATUS_AE_INITIALISED, &accel_dev->status)) { if (adf_ae_shutdown(accel_dev)) - pr_err("QAT: Failed to shutdown Accel Engine\n"); + dev_err(&GET_DEV(accel_dev), + "Failed to shutdown Accel Engine\n"); else clear_bit(ADF_STATUS_AE_INITIALISED, &accel_dev->status); @@ -372,8 +378,9 @@ void adf_dev_shutdown(struct adf_accel_dev *accel_dev) if (!test_bit(accel_dev->accel_id, &service->init_status)) continue; if (service->event_hld(accel_dev, ADF_EVENT_SHUTDOWN)) - pr_err("QAT: Failed to shutdown service %s\n", - service->name); + dev_err(&GET_DEV(accel_dev), + "Failed to shutdown service %s\n", + service->name); else clear_bit(accel_dev->accel_id, &service->init_status); } @@ -384,8 +391,9 @@ void adf_dev_shutdown(struct adf_accel_dev *accel_dev) if (!test_bit(accel_dev->accel_id, &service->init_status)) continue; if (service->event_hld(accel_dev, ADF_EVENT_SHUTDOWN)) - pr_err("QAT: Failed to shutdown service %s\n", - service->name); + dev_err(&GET_DEV(accel_dev), + "Failed to shutdown service %s\n", + service->name); else clear_bit(accel_dev->accel_id, &service->init_status); } @@ -419,16 +427,18 @@ int adf_dev_restarting_notify(struct adf_accel_dev *accel_dev) if (service->admin) continue; if (service->event_hld(accel_dev, ADF_EVENT_RESTARTING)) - pr_err("QAT: Failed to restart service %s.\n", - service->name); + dev_err(&GET_DEV(accel_dev), + "Failed to restart service %s.\n", + service->name); } list_for_each(list_itr, &service_table) { service = list_entry(list_itr, struct service_hndl, list); if (!service->admin) continue; if (service->event_hld(accel_dev, ADF_EVENT_RESTARTING)) - pr_err("QAT: Failed to restart service %s.\n", - service->name); + dev_err(&GET_DEV(accel_dev), + "Failed to restart service %s.\n", + service->name); } return 0; } @@ -443,16 +453,18 @@ int adf_dev_restarted_notify(struct adf_accel_dev *accel_dev) if (service->admin) continue; if (service->event_hld(accel_dev, ADF_EVENT_RESTARTED)) - pr_err("QAT: Failed to restart service %s.\n", - service->name); + dev_err(&GET_DEV(accel_dev), + "Failed to restart service %s.\n", + service->name); } list_for_each(list_itr, &service_table) { service = list_entry(list_itr, struct service_hndl, list); if (!service->admin) continue; if (service->event_hld(accel_dev, ADF_EVENT_RESTARTED)) - pr_err("QAT: Failed to restart service %s.\n", - service->name); + dev_err(&GET_DEV(accel_dev), + "Failed to restart service %s.\n", + service->name); } return 0; } diff --git a/drivers/crypto/qat/qat_common/adf_transport.c b/drivers/crypto/qat/qat_common/adf_transport.c index 7dd54aaee9fa..ccec327489da 100644 --- a/drivers/crypto/qat/qat_common/adf_transport.c +++ b/drivers/crypto/qat/qat_common/adf_transport.c @@ -195,7 +195,7 @@ static int adf_init_ring(struct adf_etr_ring_data *ring) memset(ring->base_addr, 0x7F, ring_size_bytes); /* The base_addr has to be aligned to the size of the buffer */ if (adf_check_ring_alignment(ring->dma_addr, ring_size_bytes)) { - pr_err("QAT: Ring address not aligned\n"); + dev_err(&GET_DEV(accel_dev), "Ring address not aligned\n"); dma_free_coherent(&GET_DEV(accel_dev), ring_size_bytes, ring->base_addr, ring->dma_addr); return -EFAULT; @@ -242,32 +242,33 @@ int adf_create_ring(struct adf_accel_dev *accel_dev, const char *section, int ret; if (bank_num >= GET_MAX_BANKS(accel_dev)) { - pr_err("QAT: Invalid bank number\n"); + dev_err(&GET_DEV(accel_dev), "Invalid bank number\n"); return -EFAULT; } if (msg_size > ADF_MSG_SIZE_TO_BYTES(ADF_MAX_MSG_SIZE)) { - pr_err("QAT: Invalid msg size\n"); + dev_err(&GET_DEV(accel_dev), "Invalid msg size\n"); return -EFAULT; } if (ADF_MAX_INFLIGHTS(adf_verify_ring_size(msg_size, num_msgs), ADF_BYTES_TO_MSG_SIZE(msg_size)) < 2) { - pr_err("QAT: Invalid ring size for given msg size\n"); + dev_err(&GET_DEV(accel_dev), + "Invalid ring size for given msg size\n"); return -EFAULT; } if (adf_cfg_get_param_value(accel_dev, section, ring_name, val)) { - pr_err("QAT: Section %s, no such entry : %s\n", - section, ring_name); + dev_err(&GET_DEV(accel_dev), "Section %s, no such entry : %s\n", + section, ring_name); return -EFAULT; } if (kstrtouint(val, 10, &ring_num)) { - pr_err("QAT: Can't get ring number\n"); + dev_err(&GET_DEV(accel_dev), "Can't get ring number\n"); return -EFAULT; } bank = &transport_data->banks[bank_num]; if (adf_reserve_ring(bank, ring_num)) { - pr_err("QAT: Ring %d, %s already exists.\n", - ring_num, ring_name); + dev_err(&GET_DEV(accel_dev), "Ring %d, %s already exists.\n", + ring_num, ring_name); return -EFAULT; } ring = &bank->rings[ring_num]; @@ -287,7 +288,8 @@ int adf_create_ring(struct adf_accel_dev *accel_dev, const char *section, accel_dev->hw_device->hw_arb_ring_enable(ring); if (adf_ring_debugfs_add(ring, ring_name)) { - pr_err("QAT: Couldn't add ring debugfs entry\n"); + dev_err(&GET_DEV(accel_dev), + "Couldn't add ring debugfs entry\n"); ret = -EFAULT; goto err; } @@ -428,7 +430,8 @@ static int adf_init_bank(struct adf_accel_dev *accel_dev, goto err; } else { if (i < hw_data->tx_rx_gap) { - pr_err("QAT: Invalid tx rings mask config\n"); + dev_err(&GET_DEV(accel_dev), + "Invalid tx rings mask config\n"); goto err; } tx_ring = &bank->rings[i - hw_data->tx_rx_gap]; @@ -436,7 +439,8 @@ static int adf_init_bank(struct adf_accel_dev *accel_dev, } } if (adf_bank_debugfs_add(bank)) { - pr_err("QAT: Failed to add bank debugfs entry\n"); + dev_err(&GET_DEV(accel_dev), + "Failed to add bank debugfs entry\n"); goto err; } @@ -492,7 +496,8 @@ int adf_init_etr_data(struct adf_accel_dev *accel_dev) etr_data->debug = debugfs_create_dir("transport", accel_dev->debugfs_dir); if (!etr_data->debug) { - pr_err("QAT: Unable to create transport debugfs entry\n"); + dev_err(&GET_DEV(accel_dev), + "Unable to create transport debugfs entry\n"); ret = -ENOENT; goto err_bank_debug; } diff --git a/drivers/crypto/qat/qat_common/adf_transport_debug.c b/drivers/crypto/qat/qat_common/adf_transport_debug.c index 6b6974553514..e41986967294 100644 --- a/drivers/crypto/qat/qat_common/adf_transport_debug.c +++ b/drivers/crypto/qat/qat_common/adf_transport_debug.c @@ -100,6 +100,8 @@ static int adf_ring_show(struct seq_file *sfile, void *v) empty = READ_CSR_E_STAT(csr, bank->bank_number); seq_puts(sfile, "------- Ring configuration -------\n"); + seq_printf(sfile, "ring name: %s\n", + ring->ring_debug->ring_name); seq_printf(sfile, "ring num %d, bank num %d\n", ring->ring_number, ring->bank->bank_number); seq_printf(sfile, "head %x, tail %x, empty: %d\n", diff --git a/drivers/crypto/qat/qat_common/icp_qat_hw.h b/drivers/crypto/qat/qat_common/icp_qat_hw.h index 68f191b653b0..121d5e6e46ca 100644 --- a/drivers/crypto/qat/qat_common/icp_qat_hw.h +++ b/drivers/crypto/qat/qat_common/icp_qat_hw.h @@ -145,7 +145,7 @@ struct icp_qat_hw_auth_setup { }; #define QAT_HW_DEFAULT_ALIGNMENT 8 -#define QAT_HW_ROUND_UP(val, n) (((val) + ((n)-1)) & (~(n-1))) +#define QAT_HW_ROUND_UP(val, n) (((val) + ((n) - 1)) & (~(n - 1))) #define ICP_QAT_HW_NULL_STATE1_SZ 32 #define ICP_QAT_HW_MD5_STATE1_SZ 16 #define ICP_QAT_HW_SHA1_STATE1_SZ 20 diff --git a/drivers/crypto/qat/qat_common/qat_crypto.c b/drivers/crypto/qat/qat_common/qat_crypto.c index 828f2a686aab..3bd705ca5973 100644 --- a/drivers/crypto/qat/qat_common/qat_crypto.c +++ b/drivers/crypto/qat/qat_common/qat_crypto.c @@ -110,13 +110,13 @@ struct qat_crypto_instance *qat_crypto_get_instance_node(int node) list_for_each(itr, adf_devmgr_get_head()) { accel_dev = list_entry(itr, struct adf_accel_dev, list); if ((node == dev_to_node(&GET_DEV(accel_dev)) || - dev_to_node(&GET_DEV(accel_dev)) < 0) - && adf_dev_started(accel_dev)) + dev_to_node(&GET_DEV(accel_dev)) < 0) && + adf_dev_started(accel_dev)) break; accel_dev = NULL; } if (!accel_dev) { - pr_err("QAT: Could not find device on node %d\n", node); + pr_err("QAT: Could not find a device on node %d\n", node); accel_dev = adf_devmgr_get_first(); } if (!accel_dev || !adf_dev_started(accel_dev)) @@ -137,7 +137,8 @@ struct qat_crypto_instance *qat_crypto_get_instance_node(int node) if (atomic_add_return(1, &inst_best->refctr) == 1) { if (adf_dev_get(accel_dev)) { atomic_dec(&inst_best->refctr); - pr_err("QAT: Could increment dev refctr\n"); + dev_err(&GET_DEV(accel_dev), + "Could not increment dev refctr\n"); return NULL; } } diff --git a/drivers/crypto/qat/qat_common/qat_hal.c b/drivers/crypto/qat/qat_common/qat_hal.c index b818c19713bf..274ff7e9de6e 100644 --- a/drivers/crypto/qat/qat_common/qat_hal.c +++ b/drivers/crypto/qat/qat_common/qat_hal.c @@ -434,8 +434,8 @@ static void qat_hal_reset_timestamp(struct icp_qat_fw_loader_handle *handle) SET_GLB_CSR(handle, MISC_CONTROL, misc_ctl | MC_TIMESTAMP_ENABLE); } -#define ESRAM_AUTO_TINIT (1<<2) -#define ESRAM_AUTO_TINIT_DONE (1<<3) +#define ESRAM_AUTO_TINIT BIT(2) +#define ESRAM_AUTO_TINIT_DONE BIT(3) #define ESRAM_AUTO_INIT_USED_CYCLES (1640) #define ESRAM_AUTO_INIT_CSR_OFFSET 0xC1C static int qat_hal_init_esram(struct icp_qat_fw_loader_handle *handle) @@ -718,7 +718,7 @@ int qat_hal_init(struct adf_accel_dev *accel_dev) handle->hal_handle->ae_max_num = max_en_ae_id + 1; /* take all AEs out of reset */ if (qat_hal_clr_reset(handle)) { - pr_err("QAT: qat_hal_clr_reset error\n"); + dev_err(&GET_DEV(accel_dev), "qat_hal_clr_reset error\n"); goto out_err; } if (qat_hal_clear_gpr(handle)) diff --git a/drivers/crypto/qat/qat_dh895xcc/adf_admin.c b/drivers/crypto/qat/qat_dh895xcc/adf_admin.c index 53c491b59f07..e4666065c399 100644 --- a/drivers/crypto/qat/qat_dh895xcc/adf_admin.c +++ b/drivers/crypto/qat/qat_dh895xcc/adf_admin.c @@ -93,7 +93,8 @@ int adf_put_admin_msg_sync(struct adf_accel_dev *accel_dev, memcpy(out, admin->virt_addr + offset + ADF_ADMINMSG_LEN, ADF_ADMINMSG_LEN); else - pr_err("QAT: Failed to send admin msg to accelerator\n"); + dev_err(&GET_DEV(accel_dev), + "Failed to send admin msg to accelerator\n"); mutex_unlock(&admin->lock); return received ? 0 : -EFAULT; diff --git a/drivers/crypto/qat/qat_dh895xcc/adf_dh895xcc_hw_data.c b/drivers/crypto/qat/qat_dh895xcc/adf_dh895xcc_hw_data.c index 6a735d5c0e37..b1386922d7a2 100644 --- a/drivers/crypto/qat/qat_dh895xcc/adf_dh895xcc_hw_data.c +++ b/drivers/crypto/qat/qat_dh895xcc/adf_dh895xcc_hw_data.c @@ -150,7 +150,8 @@ void adf_get_arbiter_mapping(struct adf_accel_dev *accel_dev, *arb_map_config = thrd_to_arb_map_sku6; break; default: - pr_err("QAT: The configuration doesn't match any SKU"); + dev_err(&GET_DEV(accel_dev), + "The configuration doesn't match any SKU"); *arb_map_config = NULL; } } diff --git a/drivers/crypto/qat/qat_dh895xcc/adf_dh895xcc_hw_data.h b/drivers/crypto/qat/qat_dh895xcc/adf_dh895xcc_hw_data.h index 01e0be21e93a..25269a9f24a2 100644 --- a/drivers/crypto/qat/qat_dh895xcc/adf_dh895xcc_hw_data.h +++ b/drivers/crypto/qat/qat_dh895xcc/adf_dh895xcc_hw_data.h @@ -73,11 +73,11 @@ /* Error detection and correction */ #define ADF_DH895XCC_AE_CTX_ENABLES(i) (i * 0x1000 + 0x20818) #define ADF_DH895XCC_AE_MISC_CONTROL(i) (i * 0x1000 + 0x20960) -#define ADF_DH895XCC_ENABLE_AE_ECC_ERR (1 << 28) -#define ADF_DH895XCC_ENABLE_AE_ECC_PARITY_CORR (1 << 24 | 1 << 12) +#define ADF_DH895XCC_ENABLE_AE_ECC_ERR BIT(28) +#define ADF_DH895XCC_ENABLE_AE_ECC_PARITY_CORR (BIT(24) | BIT(12)) #define ADF_DH895XCC_UERRSSMSH(i) (i * 0x4000 + 0x18) #define ADF_DH895XCC_CERRSSMSH(i) (i * 0x4000 + 0x10) -#define ADF_DH895XCC_ERRSSMSH_EN (1 << 3) +#define ADF_DH895XCC_ERRSSMSH_EN BIT(3) /* Admin Messages Registers */ #define ADF_DH895XCC_ADMINMSGUR_OFFSET (0x3A000 + 0x574) diff --git a/drivers/crypto/qat/qat_dh895xcc/adf_drv.c b/drivers/crypto/qat/qat_dh895xcc/adf_drv.c index 8ffdb95c9804..9decea2779c6 100644 --- a/drivers/crypto/qat/qat_dh895xcc/adf_drv.c +++ b/drivers/crypto/qat/qat_dh895xcc/adf_drv.c @@ -236,7 +236,7 @@ static int adf_probe(struct pci_dev *pdev, const struct pci_device_id *ent) } accel_dev = kzalloc_node(sizeof(*accel_dev), GFP_KERNEL, - dev_to_node(&pdev->dev)); + dev_to_node(&pdev->dev)); if (!accel_dev) return -ENOMEM; @@ -379,7 +379,7 @@ out_err: return ret; } -static void __exit adf_remove(struct pci_dev *pdev) +static void adf_remove(struct pci_dev *pdev) { struct adf_accel_dev *accel_dev = adf_devmgr_pci_to_accel_dev(pdev); diff --git a/drivers/crypto/qat/qat_dh895xcc/adf_isr.c b/drivers/crypto/qat/qat_dh895xcc/adf_isr.c index fe8f89697ad8..0d03c109c2d3 100644 --- a/drivers/crypto/qat/qat_dh895xcc/adf_isr.c +++ b/drivers/crypto/qat/qat_dh895xcc/adf_isr.c @@ -73,7 +73,7 @@ static int adf_enable_msix(struct adf_accel_dev *accel_dev) if (pci_enable_msix_exact(pci_dev_info->pci_dev, pci_dev_info->msix_entries.entries, msix_num_entries)) { - pr_err("QAT: Failed to enable MSIX IRQ\n"); + dev_err(&GET_DEV(accel_dev), "Failed to enable MSIX IRQ\n"); return -EFAULT; } return 0; @@ -97,7 +97,8 @@ static irqreturn_t adf_msix_isr_ae(int irq, void *dev_ptr) { struct adf_accel_dev *accel_dev = dev_ptr; - pr_info("QAT: qat_dev%d spurious AE interrupt\n", accel_dev->accel_id); + dev_info(&GET_DEV(accel_dev), "qat_dev%d spurious AE interrupt\n", + accel_dev->accel_id); return IRQ_HANDLED; } @@ -121,8 +122,9 @@ static int adf_request_irqs(struct adf_accel_dev *accel_dev) ret = request_irq(msixe[i].vector, adf_msix_isr_bundle, 0, name, bank); if (ret) { - pr_err("QAT: failed to enable irq %d for %s\n", - msixe[i].vector, name); + dev_err(&GET_DEV(accel_dev), + "failed to enable irq %d for %s\n", + msixe[i].vector, name); return ret; } @@ -136,8 +138,9 @@ static int adf_request_irqs(struct adf_accel_dev *accel_dev) "qat%d-ae-cluster", accel_dev->accel_id); ret = request_irq(msixe[i].vector, adf_msix_isr_ae, 0, name, accel_dev); if (ret) { - pr_err("QAT: failed to enable irq %d, for %s\n", - msixe[i].vector, name); + dev_err(&GET_DEV(accel_dev), + "failed to enable irq %d, for %s\n", + msixe[i].vector, name); return ret; } return ret; diff --git a/drivers/crypto/sahara.c b/drivers/crypto/sahara.c index 290a7f0a681f..6be377f6b9e7 100644 --- a/drivers/crypto/sahara.c +++ b/drivers/crypto/sahara.c @@ -479,6 +479,7 @@ static int sahara_hw_descriptor_create(struct sahara_dev *dev) struct scatterlist *sg; int ret; int i, j; + int idx = 0; /* Copy new key if necessary */ if (ctx->flags & FLAGS_NEW_KEY) { @@ -486,17 +487,20 @@ static int sahara_hw_descriptor_create(struct sahara_dev *dev) ctx->flags &= ~FLAGS_NEW_KEY; if (dev->flags & FLAGS_CBC) { - dev->hw_desc[0]->len1 = AES_BLOCK_SIZE; - dev->hw_desc[0]->p1 = dev->iv_phys_base; + dev->hw_desc[idx]->len1 = AES_BLOCK_SIZE; + dev->hw_desc[idx]->p1 = dev->iv_phys_base; } else { - dev->hw_desc[0]->len1 = 0; - dev->hw_desc[0]->p1 = 0; + dev->hw_desc[idx]->len1 = 0; + dev->hw_desc[idx]->p1 = 0; } - dev->hw_desc[0]->len2 = ctx->keylen; - dev->hw_desc[0]->p2 = dev->key_phys_base; - dev->hw_desc[0]->next = dev->hw_phys_desc[1]; + dev->hw_desc[idx]->len2 = ctx->keylen; + dev->hw_desc[idx]->p2 = dev->key_phys_base; + dev->hw_desc[idx]->next = dev->hw_phys_desc[1]; + + dev->hw_desc[idx]->hdr = sahara_aes_key_hdr(dev); + + idx++; } - dev->hw_desc[0]->hdr = sahara_aes_key_hdr(dev); dev->nb_in_sg = sahara_sg_length(dev->in_sg, dev->total); dev->nb_out_sg = sahara_sg_length(dev->out_sg, dev->total); @@ -520,7 +524,7 @@ static int sahara_hw_descriptor_create(struct sahara_dev *dev) } /* Create input links */ - dev->hw_desc[1]->p1 = dev->hw_phys_link[0]; + dev->hw_desc[idx]->p1 = dev->hw_phys_link[0]; sg = dev->in_sg; for (i = 0; i < dev->nb_in_sg; i++) { dev->hw_link[i]->len = sg->length; @@ -534,7 +538,7 @@ static int sahara_hw_descriptor_create(struct sahara_dev *dev) } /* Create output links */ - dev->hw_desc[1]->p2 = dev->hw_phys_link[i]; + dev->hw_desc[idx]->p2 = dev->hw_phys_link[i]; sg = dev->out_sg; for (j = i; j < dev->nb_out_sg + i; j++) { dev->hw_link[j]->len = sg->length; @@ -548,10 +552,10 @@ static int sahara_hw_descriptor_create(struct sahara_dev *dev) } /* Fill remaining fields of hw_desc[1] */ - dev->hw_desc[1]->hdr = sahara_aes_data_link_hdr(dev); - dev->hw_desc[1]->len1 = dev->total; - dev->hw_desc[1]->len2 = dev->total; - dev->hw_desc[1]->next = 0; + dev->hw_desc[idx]->hdr = sahara_aes_data_link_hdr(dev); + dev->hw_desc[idx]->len1 = dev->total; + dev->hw_desc[idx]->len2 = dev->total; + dev->hw_desc[idx]->next = 0; sahara_dump_descriptors(dev); sahara_dump_links(dev); @@ -576,6 +580,7 @@ static int sahara_aes_process(struct ablkcipher_request *req) struct sahara_ctx *ctx; struct sahara_aes_reqctx *rctx; int ret; + unsigned long timeout; /* Request is ready to be dispatched by the device */ dev_dbg(dev->device, @@ -601,10 +606,12 @@ static int sahara_aes_process(struct ablkcipher_request *req) reinit_completion(&dev->dma_completion); ret = sahara_hw_descriptor_create(dev); + if (ret) + return -EINVAL; - ret = wait_for_completion_timeout(&dev->dma_completion, + timeout = wait_for_completion_timeout(&dev->dma_completion, msecs_to_jiffies(SAHARA_TIMEOUT_MS)); - if (!ret) { + if (!timeout) { dev_err(dev->device, "AES timeout\n"); return -ETIMEDOUT; } @@ -1044,7 +1051,8 @@ static int sahara_sha_process(struct ahash_request *req) { struct sahara_dev *dev = dev_ptr; struct sahara_sha_reqctx *rctx = ahash_request_ctx(req); - int ret = -EINPROGRESS; + int ret; + unsigned long timeout; ret = sahara_sha_prepare_request(req); if (!ret) @@ -1070,9 +1078,9 @@ static int sahara_sha_process(struct ahash_request *req) sahara_write(dev, dev->hw_phys_desc[0], SAHARA_REG_DAR); - ret = wait_for_completion_timeout(&dev->dma_completion, + timeout = wait_for_completion_timeout(&dev->dma_completion, msecs_to_jiffies(SAHARA_TIMEOUT_MS)); - if (!ret) { + if (!timeout) { dev_err(dev->device, "SHA timeout\n"); return -ETIMEDOUT; } @@ -1092,15 +1100,20 @@ static int sahara_queue_manage(void *data) { struct sahara_dev *dev = (struct sahara_dev *)data; struct crypto_async_request *async_req; + struct crypto_async_request *backlog; int ret = 0; do { __set_current_state(TASK_INTERRUPTIBLE); mutex_lock(&dev->queue_mutex); + backlog = crypto_get_backlog(&dev->queue); async_req = crypto_dequeue_request(&dev->queue); mutex_unlock(&dev->queue_mutex); + if (backlog) + backlog->complete(backlog, -EINPROGRESS); + if (async_req) { if (crypto_tfm_alg_type(async_req->tfm) == CRYPTO_ALG_TYPE_AHASH) { diff --git a/drivers/crypto/talitos.c b/drivers/crypto/talitos.c index ebbae8d3ce0d..857414afa29a 100644 --- a/drivers/crypto/talitos.c +++ b/drivers/crypto/talitos.c @@ -637,8 +637,6 @@ static void talitos_unregister_rng(struct device *dev) #define TALITOS_MAX_KEY_SIZE 96 #define TALITOS_MAX_IV_LENGTH 16 /* max of AES_BLOCK_SIZE, DES3_EDE_BLOCK_SIZE */ -#define MD5_BLOCK_SIZE 64 - struct talitos_ctx { struct device *dev; int ch; @@ -2195,7 +2193,7 @@ static struct talitos_alg_template driver_algs[] = { .halg.base = { .cra_name = "md5", .cra_driver_name = "md5-talitos", - .cra_blocksize = MD5_BLOCK_SIZE, + .cra_blocksize = MD5_HMAC_BLOCK_SIZE, .cra_flags = CRYPTO_ALG_TYPE_AHASH | CRYPTO_ALG_ASYNC, } @@ -2285,7 +2283,7 @@ static struct talitos_alg_template driver_algs[] = { .halg.base = { .cra_name = "hmac(md5)", .cra_driver_name = "hmac-md5-talitos", - .cra_blocksize = MD5_BLOCK_SIZE, + .cra_blocksize = MD5_HMAC_BLOCK_SIZE, .cra_flags = CRYPTO_ALG_TYPE_AHASH | CRYPTO_ALG_ASYNC, } @@ -2706,20 +2704,16 @@ static int talitos_probe(struct platform_device *ofdev) goto err_out; } + priv->fifo_len = roundup_pow_of_two(priv->chfifo_len); + for (i = 0; i < priv->num_channels; i++) { priv->chan[i].reg = priv->reg + TALITOS_CH_STRIDE * (i + 1); if (!priv->irq[1] || !(i & 1)) priv->chan[i].reg += TALITOS_CH_BASE_OFFSET; - } - for (i = 0; i < priv->num_channels; i++) { spin_lock_init(&priv->chan[i].head_lock); spin_lock_init(&priv->chan[i].tail_lock); - } - priv->fifo_len = roundup_pow_of_two(priv->chfifo_len); - - for (i = 0; i < priv->num_channels; i++) { priv->chan[i].fifo = kzalloc(sizeof(struct talitos_request) * priv->fifo_len, GFP_KERNEL); if (!priv->chan[i].fifo) { @@ -2727,11 +2721,10 @@ static int talitos_probe(struct platform_device *ofdev) err = -ENOMEM; goto err_out; } - } - for (i = 0; i < priv->num_channels; i++) atomic_set(&priv->chan[i].submit_count, -(priv->chfifo_len - 1)); + } dma_set_mask(dev, DMA_BIT_MASK(36)); diff --git a/drivers/crypto/ux500/hash/hash_core.c b/drivers/crypto/ux500/hash/hash_core.c index 187a8fd7eee7..5f5f360628fc 100644 --- a/drivers/crypto/ux500/hash/hash_core.c +++ b/drivers/crypto/ux500/hash/hash_core.c @@ -184,7 +184,7 @@ static int hash_set_dma_transfer(struct hash_ctx *ctx, struct scatterlist *sg, direction, DMA_CTRL_ACK | DMA_PREP_INTERRUPT); if (!desc) { dev_err(ctx->device->dev, - "%s: device_prep_slave_sg() failed!\n", __func__); + "%s: dmaengine_prep_slave_sg() failed!\n", __func__); return -EFAULT; } diff --git a/drivers/crypto/vmx/Kconfig b/drivers/crypto/vmx/Kconfig new file mode 100644 index 000000000000..771babf16aa0 --- /dev/null +++ b/drivers/crypto/vmx/Kconfig @@ -0,0 +1,8 @@ +config CRYPTO_DEV_VMX_ENCRYPT + tristate "Encryption acceleration support on P8 CPU" + depends on PPC64 && CRYPTO_DEV_VMX + default y + help + Support for VMX cryptographic acceleration instructions on Power8 CPU. + This module supports acceleration for AES and GHASH in hardware. If you + choose 'M' here, this module will be called vmx-crypto. diff --git a/drivers/crypto/vmx/Makefile b/drivers/crypto/vmx/Makefile new file mode 100644 index 000000000000..c699c6e6c82e --- /dev/null +++ b/drivers/crypto/vmx/Makefile @@ -0,0 +1,19 @@ +obj-$(CONFIG_CRYPTO_DEV_VMX_ENCRYPT) += vmx-crypto.o +vmx-crypto-objs := vmx.o aesp8-ppc.o ghashp8-ppc.o aes.o aes_cbc.o aes_ctr.o ghash.o + +ifeq ($(CONFIG_CPU_LITTLE_ENDIAN),y) +TARGET := linux-ppc64le +else +TARGET := linux-pcc64 +endif + +quiet_cmd_perl = PERL $@ + cmd_perl = $(PERL) $(<) $(TARGET) > $(@) + +$(src)/aesp8-ppc.S: $(src)/aesp8-ppc.pl + $(call cmd,perl) + +$(src)/ghashp8-ppc.S: $(src)/ghashp8-ppc.pl + $(call cmd,perl) + +.PRECIOUS: $(obj)/aesp8-ppc.S $(obj)/ghashp8-ppc.S diff --git a/drivers/crypto/vmx/aes.c b/drivers/crypto/vmx/aes.c new file mode 100644 index 000000000000..ab300ea19434 --- /dev/null +++ b/drivers/crypto/vmx/aes.c @@ -0,0 +1,139 @@ +/** + * AES routines supporting VMX instructions on the Power 8 + * + * Copyright (C) 2015 International Business Machines Inc. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 only. + * + * 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. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. + * + * Author: Marcelo Henrique Cerri <mhcerri@br.ibm.com> + */ + +#include <linux/types.h> +#include <linux/err.h> +#include <linux/crypto.h> +#include <linux/delay.h> +#include <linux/hardirq.h> +#include <asm/switch_to.h> +#include <crypto/aes.h> + +#include "aesp8-ppc.h" + +struct p8_aes_ctx { + struct crypto_cipher *fallback; + struct aes_key enc_key; + struct aes_key dec_key; +}; + +static int p8_aes_init(struct crypto_tfm *tfm) +{ + const char *alg; + struct crypto_cipher *fallback; + struct p8_aes_ctx *ctx = crypto_tfm_ctx(tfm); + + if (!(alg = crypto_tfm_alg_name(tfm))) { + printk(KERN_ERR "Failed to get algorithm name.\n"); + return -ENOENT; + } + + fallback = crypto_alloc_cipher(alg, 0 ,CRYPTO_ALG_NEED_FALLBACK); + if (IS_ERR(fallback)) { + printk(KERN_ERR "Failed to allocate transformation for '%s': %ld\n", + alg, PTR_ERR(fallback)); + return PTR_ERR(fallback); + } + printk(KERN_INFO "Using '%s' as fallback implementation.\n", + crypto_tfm_alg_driver_name((struct crypto_tfm *) fallback)); + + crypto_cipher_set_flags(fallback, + crypto_cipher_get_flags((struct crypto_cipher *) tfm)); + ctx->fallback = fallback; + + return 0; +} + +static void p8_aes_exit(struct crypto_tfm *tfm) +{ + struct p8_aes_ctx *ctx = crypto_tfm_ctx(tfm); + + if (ctx->fallback) { + crypto_free_cipher(ctx->fallback); + ctx->fallback = NULL; + } +} + +static int p8_aes_setkey(struct crypto_tfm *tfm, const u8 *key, + unsigned int keylen) +{ + int ret; + struct p8_aes_ctx *ctx = crypto_tfm_ctx(tfm); + + pagefault_disable(); + enable_kernel_altivec(); + ret = aes_p8_set_encrypt_key(key, keylen * 8, &ctx->enc_key); + ret += aes_p8_set_decrypt_key(key, keylen * 8, &ctx->dec_key); + pagefault_enable(); + + ret += crypto_cipher_setkey(ctx->fallback, key, keylen); + return ret; +} + +static void p8_aes_encrypt(struct crypto_tfm *tfm, u8 *dst, const u8 *src) +{ + struct p8_aes_ctx *ctx = crypto_tfm_ctx(tfm); + + if (in_interrupt()) { + crypto_cipher_encrypt_one(ctx->fallback, dst, src); + } else { + pagefault_disable(); + enable_kernel_altivec(); + aes_p8_encrypt(src, dst, &ctx->enc_key); + pagefault_enable(); + } +} + +static void p8_aes_decrypt(struct crypto_tfm *tfm, u8 *dst, const u8 *src) +{ + struct p8_aes_ctx *ctx = crypto_tfm_ctx(tfm); + + if (in_interrupt()) { + crypto_cipher_decrypt_one(ctx->fallback, dst, src); + } else { + pagefault_disable(); + enable_kernel_altivec(); + aes_p8_decrypt(src, dst, &ctx->dec_key); + pagefault_enable(); + } +} + +struct crypto_alg p8_aes_alg = { + .cra_name = "aes", + .cra_driver_name = "p8_aes", + .cra_module = THIS_MODULE, + .cra_priority = 1000, + .cra_type = NULL, + .cra_flags = CRYPTO_ALG_TYPE_CIPHER | CRYPTO_ALG_NEED_FALLBACK, + .cra_alignmask = 0, + .cra_blocksize = AES_BLOCK_SIZE, + .cra_ctxsize = sizeof(struct p8_aes_ctx), + .cra_init = p8_aes_init, + .cra_exit = p8_aes_exit, + .cra_cipher = { + .cia_min_keysize = AES_MIN_KEY_SIZE, + .cia_max_keysize = AES_MAX_KEY_SIZE, + .cia_setkey = p8_aes_setkey, + .cia_encrypt = p8_aes_encrypt, + .cia_decrypt = p8_aes_decrypt, + }, +}; + diff --git a/drivers/crypto/vmx/aes_cbc.c b/drivers/crypto/vmx/aes_cbc.c new file mode 100644 index 000000000000..1a559b7dddb5 --- /dev/null +++ b/drivers/crypto/vmx/aes_cbc.c @@ -0,0 +1,184 @@ +/** + * AES CBC routines supporting VMX instructions on the Power 8 + * + * Copyright (C) 2015 International Business Machines Inc. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 only. + * + * 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. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. + * + * Author: Marcelo Henrique Cerri <mhcerri@br.ibm.com> + */ + +#include <linux/types.h> +#include <linux/err.h> +#include <linux/crypto.h> +#include <linux/delay.h> +#include <linux/hardirq.h> +#include <asm/switch_to.h> +#include <crypto/aes.h> +#include <crypto/scatterwalk.h> + +#include "aesp8-ppc.h" + +struct p8_aes_cbc_ctx { + struct crypto_blkcipher *fallback; + struct aes_key enc_key; + struct aes_key dec_key; +}; + +static int p8_aes_cbc_init(struct crypto_tfm *tfm) +{ + const char *alg; + struct crypto_blkcipher *fallback; + struct p8_aes_cbc_ctx *ctx = crypto_tfm_ctx(tfm); + + if (!(alg = crypto_tfm_alg_name(tfm))) { + printk(KERN_ERR "Failed to get algorithm name.\n"); + return -ENOENT; + } + + fallback = crypto_alloc_blkcipher(alg, 0 ,CRYPTO_ALG_NEED_FALLBACK); + if (IS_ERR(fallback)) { + printk(KERN_ERR "Failed to allocate transformation for '%s': %ld\n", + alg, PTR_ERR(fallback)); + return PTR_ERR(fallback); + } + printk(KERN_INFO "Using '%s' as fallback implementation.\n", + crypto_tfm_alg_driver_name((struct crypto_tfm *) fallback)); + + crypto_blkcipher_set_flags(fallback, + crypto_blkcipher_get_flags((struct crypto_blkcipher *) tfm)); + ctx->fallback = fallback; + + return 0; +} + +static void p8_aes_cbc_exit(struct crypto_tfm *tfm) +{ + struct p8_aes_cbc_ctx *ctx = crypto_tfm_ctx(tfm); + + if (ctx->fallback) { + crypto_free_blkcipher(ctx->fallback); + ctx->fallback = NULL; + } +} + +static int p8_aes_cbc_setkey(struct crypto_tfm *tfm, const u8 *key, + unsigned int keylen) +{ + int ret; + struct p8_aes_cbc_ctx *ctx = crypto_tfm_ctx(tfm); + + pagefault_disable(); + enable_kernel_altivec(); + ret = aes_p8_set_encrypt_key(key, keylen * 8, &ctx->enc_key); + ret += aes_p8_set_decrypt_key(key, keylen * 8, &ctx->dec_key); + pagefault_enable(); + + ret += crypto_blkcipher_setkey(ctx->fallback, key, keylen); + return ret; +} + +static int p8_aes_cbc_encrypt(struct blkcipher_desc *desc, + struct scatterlist *dst, struct scatterlist *src, + unsigned int nbytes) +{ + int ret; + struct blkcipher_walk walk; + struct p8_aes_cbc_ctx *ctx = crypto_tfm_ctx( + crypto_blkcipher_tfm(desc->tfm)); + struct blkcipher_desc fallback_desc = { + .tfm = ctx->fallback, + .info = desc->info, + .flags = desc->flags + }; + + if (in_interrupt()) { + ret = crypto_blkcipher_encrypt(&fallback_desc, dst, src, nbytes); + } else { + pagefault_disable(); + enable_kernel_altivec(); + + blkcipher_walk_init(&walk, dst, src, nbytes); + ret = blkcipher_walk_virt(desc, &walk); + while ((nbytes = walk.nbytes)) { + aes_p8_cbc_encrypt(walk.src.virt.addr, walk.dst.virt.addr, + nbytes & AES_BLOCK_MASK, &ctx->enc_key, walk.iv, 1); + nbytes &= AES_BLOCK_SIZE - 1; + ret = blkcipher_walk_done(desc, &walk, nbytes); + } + + pagefault_enable(); + } + + return ret; +} + +static int p8_aes_cbc_decrypt(struct blkcipher_desc *desc, + struct scatterlist *dst, struct scatterlist *src, + unsigned int nbytes) +{ + int ret; + struct blkcipher_walk walk; + struct p8_aes_cbc_ctx *ctx = crypto_tfm_ctx( + crypto_blkcipher_tfm(desc->tfm)); + struct blkcipher_desc fallback_desc = { + .tfm = ctx->fallback, + .info = desc->info, + .flags = desc->flags + }; + + if (in_interrupt()) { + ret = crypto_blkcipher_decrypt(&fallback_desc, dst, src, nbytes); + } else { + pagefault_disable(); + enable_kernel_altivec(); + + blkcipher_walk_init(&walk, dst, src, nbytes); + ret = blkcipher_walk_virt(desc, &walk); + while ((nbytes = walk.nbytes)) { + aes_p8_cbc_encrypt(walk.src.virt.addr, walk.dst.virt.addr, + nbytes & AES_BLOCK_MASK, &ctx->dec_key, walk.iv, 0); + nbytes &= AES_BLOCK_SIZE - 1; + ret = blkcipher_walk_done(desc, &walk, nbytes); + } + + pagefault_enable(); + } + + return ret; +} + + +struct crypto_alg p8_aes_cbc_alg = { + .cra_name = "cbc(aes)", + .cra_driver_name = "p8_aes_cbc", + .cra_module = THIS_MODULE, + .cra_priority = 1000, + .cra_type = &crypto_blkcipher_type, + .cra_flags = CRYPTO_ALG_TYPE_BLKCIPHER | CRYPTO_ALG_NEED_FALLBACK, + .cra_alignmask = 0, + .cra_blocksize = AES_BLOCK_SIZE, + .cra_ctxsize = sizeof(struct p8_aes_cbc_ctx), + .cra_init = p8_aes_cbc_init, + .cra_exit = p8_aes_cbc_exit, + .cra_blkcipher = { + .ivsize = 0, + .min_keysize = AES_MIN_KEY_SIZE, + .max_keysize = AES_MAX_KEY_SIZE, + .setkey = p8_aes_cbc_setkey, + .encrypt = p8_aes_cbc_encrypt, + .decrypt = p8_aes_cbc_decrypt, + }, +}; + diff --git a/drivers/crypto/vmx/aes_ctr.c b/drivers/crypto/vmx/aes_ctr.c new file mode 100644 index 000000000000..96dbee4bf4a6 --- /dev/null +++ b/drivers/crypto/vmx/aes_ctr.c @@ -0,0 +1,167 @@ +/** + * AES CTR routines supporting VMX instructions on the Power 8 + * + * Copyright (C) 2015 International Business Machines Inc. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 only. + * + * 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. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. + * + * Author: Marcelo Henrique Cerri <mhcerri@br.ibm.com> + */ + +#include <linux/types.h> +#include <linux/err.h> +#include <linux/crypto.h> +#include <linux/delay.h> +#include <linux/hardirq.h> +#include <asm/switch_to.h> +#include <crypto/aes.h> +#include <crypto/scatterwalk.h> +#include "aesp8-ppc.h" + +struct p8_aes_ctr_ctx { + struct crypto_blkcipher *fallback; + struct aes_key enc_key; +}; + +static int p8_aes_ctr_init(struct crypto_tfm *tfm) +{ + const char *alg; + struct crypto_blkcipher *fallback; + struct p8_aes_ctr_ctx *ctx = crypto_tfm_ctx(tfm); + + if (!(alg = crypto_tfm_alg_name(tfm))) { + printk(KERN_ERR "Failed to get algorithm name.\n"); + return -ENOENT; + } + + fallback = crypto_alloc_blkcipher(alg, 0 ,CRYPTO_ALG_NEED_FALLBACK); + if (IS_ERR(fallback)) { + printk(KERN_ERR "Failed to allocate transformation for '%s': %ld\n", + alg, PTR_ERR(fallback)); + return PTR_ERR(fallback); + } + printk(KERN_INFO "Using '%s' as fallback implementation.\n", + crypto_tfm_alg_driver_name((struct crypto_tfm *) fallback)); + + crypto_blkcipher_set_flags(fallback, + crypto_blkcipher_get_flags((struct crypto_blkcipher *) tfm)); + ctx->fallback = fallback; + + return 0; +} + +static void p8_aes_ctr_exit(struct crypto_tfm *tfm) +{ + struct p8_aes_ctr_ctx *ctx = crypto_tfm_ctx(tfm); + + if (ctx->fallback) { + crypto_free_blkcipher(ctx->fallback); + ctx->fallback = NULL; + } +} + +static int p8_aes_ctr_setkey(struct crypto_tfm *tfm, const u8 *key, + unsigned int keylen) +{ + int ret; + struct p8_aes_ctr_ctx *ctx = crypto_tfm_ctx(tfm); + + pagefault_disable(); + enable_kernel_altivec(); + ret = aes_p8_set_encrypt_key(key, keylen * 8, &ctx->enc_key); + pagefault_enable(); + + ret += crypto_blkcipher_setkey(ctx->fallback, key, keylen); + return ret; +} + +static void p8_aes_ctr_final(struct p8_aes_ctr_ctx *ctx, + struct blkcipher_walk *walk) +{ + u8 *ctrblk = walk->iv; + u8 keystream[AES_BLOCK_SIZE]; + u8 *src = walk->src.virt.addr; + u8 *dst = walk->dst.virt.addr; + unsigned int nbytes = walk->nbytes; + + pagefault_disable(); + enable_kernel_altivec(); + aes_p8_encrypt(ctrblk, keystream, &ctx->enc_key); + pagefault_enable(); + + crypto_xor(keystream, src, nbytes); + memcpy(dst, keystream, nbytes); + crypto_inc(ctrblk, AES_BLOCK_SIZE); +} + +static int p8_aes_ctr_crypt(struct blkcipher_desc *desc, + struct scatterlist *dst, struct scatterlist *src, + unsigned int nbytes) +{ + int ret; + struct blkcipher_walk walk; + struct p8_aes_ctr_ctx *ctx = crypto_tfm_ctx( + crypto_blkcipher_tfm(desc->tfm)); + struct blkcipher_desc fallback_desc = { + .tfm = ctx->fallback, + .info = desc->info, + .flags = desc->flags + }; + + if (in_interrupt()) { + ret = crypto_blkcipher_encrypt(&fallback_desc, dst, src, nbytes); + } else { + blkcipher_walk_init(&walk, dst, src, nbytes); + ret = blkcipher_walk_virt_block(desc, &walk, AES_BLOCK_SIZE); + while ((nbytes = walk.nbytes) >= AES_BLOCK_SIZE) { + pagefault_disable(); + enable_kernel_altivec(); + aes_p8_ctr32_encrypt_blocks(walk.src.virt.addr, walk.dst.virt.addr, + (nbytes & AES_BLOCK_MASK)/AES_BLOCK_SIZE, &ctx->enc_key, walk.iv); + pagefault_enable(); + + crypto_inc(walk.iv, AES_BLOCK_SIZE); + nbytes &= AES_BLOCK_SIZE - 1; + ret = blkcipher_walk_done(desc, &walk, nbytes); + } + if (walk.nbytes) { + p8_aes_ctr_final(ctx, &walk); + ret = blkcipher_walk_done(desc, &walk, 0); + } + } + + return ret; +} + +struct crypto_alg p8_aes_ctr_alg = { + .cra_name = "ctr(aes)", + .cra_driver_name = "p8_aes_ctr", + .cra_module = THIS_MODULE, + .cra_priority = 1000, + .cra_type = &crypto_blkcipher_type, + .cra_flags = CRYPTO_ALG_TYPE_BLKCIPHER | CRYPTO_ALG_NEED_FALLBACK, + .cra_alignmask = 0, + .cra_blocksize = 1, + .cra_ctxsize = sizeof(struct p8_aes_ctr_ctx), + .cra_init = p8_aes_ctr_init, + .cra_exit = p8_aes_ctr_exit, + .cra_blkcipher = { + .ivsize = 0, + .min_keysize = AES_MIN_KEY_SIZE, + .max_keysize = AES_MAX_KEY_SIZE, + .setkey = p8_aes_ctr_setkey, + .encrypt = p8_aes_ctr_crypt, + .decrypt = p8_aes_ctr_crypt, + }, +}; diff --git a/drivers/crypto/vmx/aesp8-ppc.h b/drivers/crypto/vmx/aesp8-ppc.h new file mode 100644 index 000000000000..e963945a83e1 --- /dev/null +++ b/drivers/crypto/vmx/aesp8-ppc.h @@ -0,0 +1,20 @@ +#include <linux/types.h> +#include <crypto/aes.h> + +#define AES_BLOCK_MASK (~(AES_BLOCK_SIZE-1)) + +struct aes_key { + u8 key[AES_MAX_KEYLENGTH]; + int rounds; +}; + +int aes_p8_set_encrypt_key(const u8 *userKey, const int bits, + struct aes_key *key); +int aes_p8_set_decrypt_key(const u8 *userKey, const int bits, + struct aes_key *key); +void aes_p8_encrypt(const u8 *in, u8 *out, const struct aes_key *key); +void aes_p8_decrypt(const u8 *in, u8 *out,const struct aes_key *key); +void aes_p8_cbc_encrypt(const u8 *in, u8 *out, size_t len, + const struct aes_key *key, u8 *iv, const int enc); +void aes_p8_ctr32_encrypt_blocks(const u8 *in, u8 *out, + size_t len, const struct aes_key *key, const u8 *iv); diff --git a/drivers/crypto/vmx/aesp8-ppc.pl b/drivers/crypto/vmx/aesp8-ppc.pl new file mode 100644 index 000000000000..6c5c20c6108e --- /dev/null +++ b/drivers/crypto/vmx/aesp8-ppc.pl @@ -0,0 +1,1930 @@ +#!/usr/bin/env perl +# +# ==================================================================== +# Written by Andy Polyakov <appro@openssl.org> for the OpenSSL +# project. The module is, however, dual licensed under OpenSSL and +# CRYPTOGAMS licenses depending on where you obtain it. For further +# details see http://www.openssl.org/~appro/cryptogams/. +# ==================================================================== +# +# This module implements support for AES instructions as per PowerISA +# specification version 2.07, first implemented by POWER8 processor. +# The module is endian-agnostic in sense that it supports both big- +# and little-endian cases. Data alignment in parallelizable modes is +# handled with VSX loads and stores, which implies MSR.VSX flag being +# set. It should also be noted that ISA specification doesn't prohibit +# alignment exceptions for these instructions on page boundaries. +# Initially alignment was handled in pure AltiVec/VMX way [when data +# is aligned programmatically, which in turn guarantees exception- +# free execution], but it turned to hamper performance when vcipher +# instructions are interleaved. It's reckoned that eventual +# misalignment penalties at page boundaries are in average lower +# than additional overhead in pure AltiVec approach. + +$flavour = shift; + +if ($flavour =~ /64/) { + $SIZE_T =8; + $LRSAVE =2*$SIZE_T; + $STU ="stdu"; + $POP ="ld"; + $PUSH ="std"; + $UCMP ="cmpld"; + $SHL ="sldi"; +} elsif ($flavour =~ /32/) { + $SIZE_T =4; + $LRSAVE =$SIZE_T; + $STU ="stwu"; + $POP ="lwz"; + $PUSH ="stw"; + $UCMP ="cmplw"; + $SHL ="slwi"; +} else { die "nonsense $flavour"; } + +$LITTLE_ENDIAN = ($flavour=~/le$/) ? $SIZE_T : 0; + +$0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1; +( $xlate="${dir}ppc-xlate.pl" and -f $xlate ) or +( $xlate="${dir}../../perlasm/ppc-xlate.pl" and -f $xlate) or +die "can't locate ppc-xlate.pl"; + +open STDOUT,"| $^X $xlate $flavour ".shift || die "can't call $xlate: $!"; + +$FRAME=8*$SIZE_T; +$prefix="aes_p8"; + +$sp="r1"; +$vrsave="r12"; + +######################################################################### +{{{ # Key setup procedures # +my ($inp,$bits,$out,$ptr,$cnt,$rounds)=map("r$_",(3..8)); +my ($zero,$in0,$in1,$key,$rcon,$mask,$tmp)=map("v$_",(0..6)); +my ($stage,$outperm,$outmask,$outhead,$outtail)=map("v$_",(7..11)); + +$code.=<<___; +.machine "any" + +.text + +.align 7 +rcon: +.long 0x01000000, 0x01000000, 0x01000000, 0x01000000 ?rev +.long 0x1b000000, 0x1b000000, 0x1b000000, 0x1b000000 ?rev +.long 0x0d0e0f0c, 0x0d0e0f0c, 0x0d0e0f0c, 0x0d0e0f0c ?rev +.long 0,0,0,0 ?asis +Lconsts: + mflr r0 + bcl 20,31,\$+4 + mflr $ptr #vvvvv "distance between . and rcon + addi $ptr,$ptr,-0x48 + mtlr r0 + blr + .long 0 + .byte 0,12,0x14,0,0,0,0,0 +.asciz "AES for PowerISA 2.07, CRYPTOGAMS by <appro\@openssl.org>" + +.globl .${prefix}_set_encrypt_key +Lset_encrypt_key: + mflr r11 + $PUSH r11,$LRSAVE($sp) + + li $ptr,-1 + ${UCMP}i $inp,0 + beq- Lenc_key_abort # if ($inp==0) return -1; + ${UCMP}i $out,0 + beq- Lenc_key_abort # if ($out==0) return -1; + li $ptr,-2 + cmpwi $bits,128 + blt- Lenc_key_abort + cmpwi $bits,256 + bgt- Lenc_key_abort + andi. r0,$bits,0x3f + bne- Lenc_key_abort + + lis r0,0xfff0 + mfspr $vrsave,256 + mtspr 256,r0 + + bl Lconsts + mtlr r11 + + neg r9,$inp + lvx $in0,0,$inp + addi $inp,$inp,15 # 15 is not typo + lvsr $key,0,r9 # borrow $key + li r8,0x20 + cmpwi $bits,192 + lvx $in1,0,$inp + le?vspltisb $mask,0x0f # borrow $mask + lvx $rcon,0,$ptr + le?vxor $key,$key,$mask # adjust for byte swap + lvx $mask,r8,$ptr + addi $ptr,$ptr,0x10 + vperm $in0,$in0,$in1,$key # align [and byte swap in LE] + li $cnt,8 + vxor $zero,$zero,$zero + mtctr $cnt + + ?lvsr $outperm,0,$out + vspltisb $outmask,-1 + lvx $outhead,0,$out + ?vperm $outmask,$zero,$outmask,$outperm + + blt Loop128 + addi $inp,$inp,8 + beq L192 + addi $inp,$inp,8 + b L256 + +.align 4 +Loop128: + vperm $key,$in0,$in0,$mask # rotate-n-splat + vsldoi $tmp,$zero,$in0,12 # >>32 + vperm $outtail,$in0,$in0,$outperm # rotate + vsel $stage,$outhead,$outtail,$outmask + vmr $outhead,$outtail + vcipherlast $key,$key,$rcon + stvx $stage,0,$out + addi $out,$out,16 + + vxor $in0,$in0,$tmp + vsldoi $tmp,$zero,$tmp,12 # >>32 + vxor $in0,$in0,$tmp + vsldoi $tmp,$zero,$tmp,12 # >>32 + vxor $in0,$in0,$tmp + vadduwm $rcon,$rcon,$rcon + vxor $in0,$in0,$key + bdnz Loop128 + + lvx $rcon,0,$ptr # last two round keys + + vperm $key,$in0,$in0,$mask # rotate-n-splat + vsldoi $tmp,$zero,$in0,12 # >>32 + vperm $outtail,$in0,$in0,$outperm # rotate + vsel $stage,$outhead,$outtail,$outmask + vmr $outhead,$outtail + vcipherlast $key,$key,$rcon + stvx $stage,0,$out + addi $out,$out,16 + + vxor $in0,$in0,$tmp + vsldoi $tmp,$zero,$tmp,12 # >>32 + vxor $in0,$in0,$tmp + vsldoi $tmp,$zero,$tmp,12 # >>32 + vxor $in0,$in0,$tmp + vadduwm $rcon,$rcon,$rcon + vxor $in0,$in0,$key + + vperm $key,$in0,$in0,$mask # rotate-n-splat + vsldoi $tmp,$zero,$in0,12 # >>32 + vperm $outtail,$in0,$in0,$outperm # rotate + vsel $stage,$outhead,$outtail,$outmask + vmr $outhead,$outtail + vcipherlast $key,$key,$rcon + stvx $stage,0,$out + addi $out,$out,16 + + vxor $in0,$in0,$tmp + vsldoi $tmp,$zero,$tmp,12 # >>32 + vxor $in0,$in0,$tmp + vsldoi $tmp,$zero,$tmp,12 # >>32 + vxor $in0,$in0,$tmp + vxor $in0,$in0,$key + vperm $outtail,$in0,$in0,$outperm # rotate + vsel $stage,$outhead,$outtail,$outmask + vmr $outhead,$outtail + stvx $stage,0,$out + + addi $inp,$out,15 # 15 is not typo + addi $out,$out,0x50 + + li $rounds,10 + b Ldone + +.align 4 +L192: + lvx $tmp,0,$inp + li $cnt,4 + vperm $outtail,$in0,$in0,$outperm # rotate + vsel $stage,$outhead,$outtail,$outmask + vmr $outhead,$outtail + stvx $stage,0,$out + addi $out,$out,16 + vperm $in1,$in1,$tmp,$key # align [and byte swap in LE] + vspltisb $key,8 # borrow $key + mtctr $cnt + vsububm $mask,$mask,$key # adjust the mask + +Loop192: + vperm $key,$in1,$in1,$mask # roate-n-splat + vsldoi $tmp,$zero,$in0,12 # >>32 + vcipherlast $key,$key,$rcon + + vxor $in0,$in0,$tmp + vsldoi $tmp,$zero,$tmp,12 # >>32 + vxor $in0,$in0,$tmp + vsldoi $tmp,$zero,$tmp,12 # >>32 + vxor $in0,$in0,$tmp + + vsldoi $stage,$zero,$in1,8 + vspltw $tmp,$in0,3 + vxor $tmp,$tmp,$in1 + vsldoi $in1,$zero,$in1,12 # >>32 + vadduwm $rcon,$rcon,$rcon + vxor $in1,$in1,$tmp + vxor $in0,$in0,$key + vxor $in1,$in1,$key + vsldoi $stage,$stage,$in0,8 + + vperm $key,$in1,$in1,$mask # rotate-n-splat + vsldoi $tmp,$zero,$in0,12 # >>32 + vperm $outtail,$stage,$stage,$outperm # rotate + vsel $stage,$outhead,$outtail,$outmask + vmr $outhead,$outtail + vcipherlast $key,$key,$rcon + stvx $stage,0,$out + addi $out,$out,16 + + vsldoi $stage,$in0,$in1,8 + vxor $in0,$in0,$tmp + vsldoi $tmp,$zero,$tmp,12 # >>32 + vperm $outtail,$stage,$stage,$outperm # rotate + vsel $stage,$outhead,$outtail,$outmask + vmr $outhead,$outtail + vxor $in0,$in0,$tmp + vsldoi $tmp,$zero,$tmp,12 # >>32 + vxor $in0,$in0,$tmp + stvx $stage,0,$out + addi $out,$out,16 + + vspltw $tmp,$in0,3 + vxor $tmp,$tmp,$in1 + vsldoi $in1,$zero,$in1,12 # >>32 + vadduwm $rcon,$rcon,$rcon + vxor $in1,$in1,$tmp + vxor $in0,$in0,$key + vxor $in1,$in1,$key + vperm $outtail,$in0,$in0,$outperm # rotate + vsel $stage,$outhead,$outtail,$outmask + vmr $outhead,$outtail + stvx $stage,0,$out + addi $inp,$out,15 # 15 is not typo + addi $out,$out,16 + bdnz Loop192 + + li $rounds,12 + addi $out,$out,0x20 + b Ldone + +.align 4 +L256: + lvx $tmp,0,$inp + li $cnt,7 + li $rounds,14 + vperm $outtail,$in0,$in0,$outperm # rotate + vsel $stage,$outhead,$outtail,$outmask + vmr $outhead,$outtail + stvx $stage,0,$out + addi $out,$out,16 + vperm $in1,$in1,$tmp,$key # align [and byte swap in LE] + mtctr $cnt + +Loop256: + vperm $key,$in1,$in1,$mask # rotate-n-splat + vsldoi $tmp,$zero,$in0,12 # >>32 + vperm $outtail,$in1,$in1,$outperm # rotate + vsel $stage,$outhead,$outtail,$outmask + vmr $outhead,$outtail + vcipherlast $key,$key,$rcon + stvx $stage,0,$out + addi $out,$out,16 + + vxor $in0,$in0,$tmp + vsldoi $tmp,$zero,$tmp,12 # >>32 + vxor $in0,$in0,$tmp + vsldoi $tmp,$zero,$tmp,12 # >>32 + vxor $in0,$in0,$tmp + vadduwm $rcon,$rcon,$rcon + vxor $in0,$in0,$key + vperm $outtail,$in0,$in0,$outperm # rotate + vsel $stage,$outhead,$outtail,$outmask + vmr $outhead,$outtail + stvx $stage,0,$out + addi $inp,$out,15 # 15 is not typo + addi $out,$out,16 + bdz Ldone + + vspltw $key,$in0,3 # just splat + vsldoi $tmp,$zero,$in1,12 # >>32 + vsbox $key,$key + + vxor $in1,$in1,$tmp + vsldoi $tmp,$zero,$tmp,12 # >>32 + vxor $in1,$in1,$tmp + vsldoi $tmp,$zero,$tmp,12 # >>32 + vxor $in1,$in1,$tmp + + vxor $in1,$in1,$key + b Loop256 + +.align 4 +Ldone: + lvx $in1,0,$inp # redundant in aligned case + vsel $in1,$outhead,$in1,$outmask + stvx $in1,0,$inp + li $ptr,0 + mtspr 256,$vrsave + stw $rounds,0($out) + +Lenc_key_abort: + mr r3,$ptr + blr + .long 0 + .byte 0,12,0x14,1,0,0,3,0 + .long 0 +.size .${prefix}_set_encrypt_key,.-.${prefix}_set_encrypt_key + +.globl .${prefix}_set_decrypt_key + $STU $sp,-$FRAME($sp) + mflr r10 + $PUSH r10,$FRAME+$LRSAVE($sp) + bl Lset_encrypt_key + mtlr r10 + + cmpwi r3,0 + bne- Ldec_key_abort + + slwi $cnt,$rounds,4 + subi $inp,$out,240 # first round key + srwi $rounds,$rounds,1 + add $out,$inp,$cnt # last round key + mtctr $rounds + +Ldeckey: + lwz r0, 0($inp) + lwz r6, 4($inp) + lwz r7, 8($inp) + lwz r8, 12($inp) + addi $inp,$inp,16 + lwz r9, 0($out) + lwz r10,4($out) + lwz r11,8($out) + lwz r12,12($out) + stw r0, 0($out) + stw r6, 4($out) + stw r7, 8($out) + stw r8, 12($out) + subi $out,$out,16 + stw r9, -16($inp) + stw r10,-12($inp) + stw r11,-8($inp) + stw r12,-4($inp) + bdnz Ldeckey + + xor r3,r3,r3 # return value +Ldec_key_abort: + addi $sp,$sp,$FRAME + blr + .long 0 + .byte 0,12,4,1,0x80,0,3,0 + .long 0 +.size .${prefix}_set_decrypt_key,.-.${prefix}_set_decrypt_key +___ +}}} +######################################################################### +{{{ # Single block en- and decrypt procedures # +sub gen_block () { +my $dir = shift; +my $n = $dir eq "de" ? "n" : ""; +my ($inp,$out,$key,$rounds,$idx)=map("r$_",(3..7)); + +$code.=<<___; +.globl .${prefix}_${dir}crypt + lwz $rounds,240($key) + lis r0,0xfc00 + mfspr $vrsave,256 + li $idx,15 # 15 is not typo + mtspr 256,r0 + + lvx v0,0,$inp + neg r11,$out + lvx v1,$idx,$inp + lvsl v2,0,$inp # inpperm + le?vspltisb v4,0x0f + ?lvsl v3,0,r11 # outperm + le?vxor v2,v2,v4 + li $idx,16 + vperm v0,v0,v1,v2 # align [and byte swap in LE] + lvx v1,0,$key + ?lvsl v5,0,$key # keyperm + srwi $rounds,$rounds,1 + lvx v2,$idx,$key + addi $idx,$idx,16 + subi $rounds,$rounds,1 + ?vperm v1,v1,v2,v5 # align round key + + vxor v0,v0,v1 + lvx v1,$idx,$key + addi $idx,$idx,16 + mtctr $rounds + +Loop_${dir}c: + ?vperm v2,v2,v1,v5 + v${n}cipher v0,v0,v2 + lvx v2,$idx,$key + addi $idx,$idx,16 + ?vperm v1,v1,v2,v5 + v${n}cipher v0,v0,v1 + lvx v1,$idx,$key + addi $idx,$idx,16 + bdnz Loop_${dir}c + + ?vperm v2,v2,v1,v5 + v${n}cipher v0,v0,v2 + lvx v2,$idx,$key + ?vperm v1,v1,v2,v5 + v${n}cipherlast v0,v0,v1 + + vspltisb v2,-1 + vxor v1,v1,v1 + li $idx,15 # 15 is not typo + ?vperm v2,v1,v2,v3 # outmask + le?vxor v3,v3,v4 + lvx v1,0,$out # outhead + vperm v0,v0,v0,v3 # rotate [and byte swap in LE] + vsel v1,v1,v0,v2 + lvx v4,$idx,$out + stvx v1,0,$out + vsel v0,v0,v4,v2 + stvx v0,$idx,$out + + mtspr 256,$vrsave + blr + .long 0 + .byte 0,12,0x14,0,0,0,3,0 + .long 0 +.size .${prefix}_${dir}crypt,.-.${prefix}_${dir}crypt +___ +} +&gen_block("en"); +&gen_block("de"); +}}} +######################################################################### +{{{ # CBC en- and decrypt procedures # +my ($inp,$out,$len,$key,$ivp,$enc,$rounds,$idx)=map("r$_",(3..10)); +my ($rndkey0,$rndkey1,$inout,$tmp)= map("v$_",(0..3)); +my ($ivec,$inptail,$inpperm,$outhead,$outperm,$outmask,$keyperm)= + map("v$_",(4..10)); +$code.=<<___; +.globl .${prefix}_cbc_encrypt + ${UCMP}i $len,16 + bltlr- + + cmpwi $enc,0 # test direction + lis r0,0xffe0 + mfspr $vrsave,256 + mtspr 256,r0 + + li $idx,15 + vxor $rndkey0,$rndkey0,$rndkey0 + le?vspltisb $tmp,0x0f + + lvx $ivec,0,$ivp # load [unaligned] iv + lvsl $inpperm,0,$ivp + lvx $inptail,$idx,$ivp + le?vxor $inpperm,$inpperm,$tmp + vperm $ivec,$ivec,$inptail,$inpperm + + neg r11,$inp + ?lvsl $keyperm,0,$key # prepare for unaligned key + lwz $rounds,240($key) + + lvsr $inpperm,0,r11 # prepare for unaligned load + lvx $inptail,0,$inp + addi $inp,$inp,15 # 15 is not typo + le?vxor $inpperm,$inpperm,$tmp + + ?lvsr $outperm,0,$out # prepare for unaligned store + vspltisb $outmask,-1 + lvx $outhead,0,$out + ?vperm $outmask,$rndkey0,$outmask,$outperm + le?vxor $outperm,$outperm,$tmp + + srwi $rounds,$rounds,1 + li $idx,16 + subi $rounds,$rounds,1 + beq Lcbc_dec + +Lcbc_enc: + vmr $inout,$inptail + lvx $inptail,0,$inp + addi $inp,$inp,16 + mtctr $rounds + subi $len,$len,16 # len-=16 + + lvx $rndkey0,0,$key + vperm $inout,$inout,$inptail,$inpperm + lvx $rndkey1,$idx,$key + addi $idx,$idx,16 + ?vperm $rndkey0,$rndkey0,$rndkey1,$keyperm + vxor $inout,$inout,$rndkey0 + lvx $rndkey0,$idx,$key + addi $idx,$idx,16 + vxor $inout,$inout,$ivec + +Loop_cbc_enc: + ?vperm $rndkey1,$rndkey1,$rndkey0,$keyperm + vcipher $inout,$inout,$rndkey1 + lvx $rndkey1,$idx,$key + addi $idx,$idx,16 + ?vperm $rndkey0,$rndkey0,$rndkey1,$keyperm + vcipher $inout,$inout,$rndkey0 + lvx $rndkey0,$idx,$key + addi $idx,$idx,16 + bdnz Loop_cbc_enc + + ?vperm $rndkey1,$rndkey1,$rndkey0,$keyperm + vcipher $inout,$inout,$rndkey1 + lvx $rndkey1,$idx,$key + li $idx,16 + ?vperm $rndkey0,$rndkey0,$rndkey1,$keyperm + vcipherlast $ivec,$inout,$rndkey0 + ${UCMP}i $len,16 + + vperm $tmp,$ivec,$ivec,$outperm + vsel $inout,$outhead,$tmp,$outmask + vmr $outhead,$tmp + stvx $inout,0,$out + addi $out,$out,16 + bge Lcbc_enc + + b Lcbc_done + +.align 4 +Lcbc_dec: + ${UCMP}i $len,128 + bge _aesp8_cbc_decrypt8x + vmr $tmp,$inptail + lvx $inptail,0,$inp + addi $inp,$inp,16 + mtctr $rounds + subi $len,$len,16 # len-=16 + + lvx $rndkey0,0,$key + vperm $tmp,$tmp,$inptail,$inpperm + lvx $rndkey1,$idx,$key + addi $idx,$idx,16 + ?vperm $rndkey0,$rndkey0,$rndkey1,$keyperm + vxor $inout,$tmp,$rndkey0 + lvx $rndkey0,$idx,$key + addi $idx,$idx,16 + +Loop_cbc_dec: + ?vperm $rndkey1,$rndkey1,$rndkey0,$keyperm + vncipher $inout,$inout,$rndkey1 + lvx $rndkey1,$idx,$key + addi $idx,$idx,16 + ?vperm $rndkey0,$rndkey0,$rndkey1,$keyperm + vncipher $inout,$inout,$rndkey0 + lvx $rndkey0,$idx,$key + addi $idx,$idx,16 + bdnz Loop_cbc_dec + + ?vperm $rndkey1,$rndkey1,$rndkey0,$keyperm + vncipher $inout,$inout,$rndkey1 + lvx $rndkey1,$idx,$key + li $idx,16 + ?vperm $rndkey0,$rndkey0,$rndkey1,$keyperm + vncipherlast $inout,$inout,$rndkey0 + ${UCMP}i $len,16 + + vxor $inout,$inout,$ivec + vmr $ivec,$tmp + vperm $tmp,$inout,$inout,$outperm + vsel $inout,$outhead,$tmp,$outmask + vmr $outhead,$tmp + stvx $inout,0,$out + addi $out,$out,16 + bge Lcbc_dec + +Lcbc_done: + addi $out,$out,-1 + lvx $inout,0,$out # redundant in aligned case + vsel $inout,$outhead,$inout,$outmask + stvx $inout,0,$out + + neg $enc,$ivp # write [unaligned] iv + li $idx,15 # 15 is not typo + vxor $rndkey0,$rndkey0,$rndkey0 + vspltisb $outmask,-1 + le?vspltisb $tmp,0x0f + ?lvsl $outperm,0,$enc + ?vperm $outmask,$rndkey0,$outmask,$outperm + le?vxor $outperm,$outperm,$tmp + lvx $outhead,0,$ivp + vperm $ivec,$ivec,$ivec,$outperm + vsel $inout,$outhead,$ivec,$outmask + lvx $inptail,$idx,$ivp + stvx $inout,0,$ivp + vsel $inout,$ivec,$inptail,$outmask + stvx $inout,$idx,$ivp + + mtspr 256,$vrsave + blr + .long 0 + .byte 0,12,0x14,0,0,0,6,0 + .long 0 +___ +######################################################################### +{{ # Optimized CBC decrypt procedure # +my $key_="r11"; +my ($x00,$x10,$x20,$x30,$x40,$x50,$x60,$x70)=map("r$_",(0,8,26..31)); +my ($in0, $in1, $in2, $in3, $in4, $in5, $in6, $in7 )=map("v$_",(0..3,10..13)); +my ($out0,$out1,$out2,$out3,$out4,$out5,$out6,$out7)=map("v$_",(14..21)); +my $rndkey0="v23"; # v24-v25 rotating buffer for first found keys + # v26-v31 last 6 round keys +my ($tmp,$keyperm)=($in3,$in4); # aliases with "caller", redundant assignment + +$code.=<<___; +.align 5 +_aesp8_cbc_decrypt8x: + $STU $sp,-`($FRAME+21*16+6*$SIZE_T)`($sp) + li r10,`$FRAME+8*16+15` + li r11,`$FRAME+8*16+31` + stvx v20,r10,$sp # ABI says so + addi r10,r10,32 + stvx v21,r11,$sp + addi r11,r11,32 + stvx v22,r10,$sp + addi r10,r10,32 + stvx v23,r11,$sp + addi r11,r11,32 + stvx v24,r10,$sp + addi r10,r10,32 + stvx v25,r11,$sp + addi r11,r11,32 + stvx v26,r10,$sp + addi r10,r10,32 + stvx v27,r11,$sp + addi r11,r11,32 + stvx v28,r10,$sp + addi r10,r10,32 + stvx v29,r11,$sp + addi r11,r11,32 + stvx v30,r10,$sp + stvx v31,r11,$sp + li r0,-1 + stw $vrsave,`$FRAME+21*16-4`($sp) # save vrsave + li $x10,0x10 + $PUSH r26,`$FRAME+21*16+0*$SIZE_T`($sp) + li $x20,0x20 + $PUSH r27,`$FRAME+21*16+1*$SIZE_T`($sp) + li $x30,0x30 + $PUSH r28,`$FRAME+21*16+2*$SIZE_T`($sp) + li $x40,0x40 + $PUSH r29,`$FRAME+21*16+3*$SIZE_T`($sp) + li $x50,0x50 + $PUSH r30,`$FRAME+21*16+4*$SIZE_T`($sp) + li $x60,0x60 + $PUSH r31,`$FRAME+21*16+5*$SIZE_T`($sp) + li $x70,0x70 + mtspr 256,r0 + + subi $rounds,$rounds,3 # -4 in total + subi $len,$len,128 # bias + + lvx $rndkey0,$x00,$key # load key schedule + lvx v30,$x10,$key + addi $key,$key,0x20 + lvx v31,$x00,$key + ?vperm $rndkey0,$rndkey0,v30,$keyperm + addi $key_,$sp,$FRAME+15 + mtctr $rounds + +Load_cbc_dec_key: + ?vperm v24,v30,v31,$keyperm + lvx v30,$x10,$key + addi $key,$key,0x20 + stvx v24,$x00,$key_ # off-load round[1] + ?vperm v25,v31,v30,$keyperm + lvx v31,$x00,$key + stvx v25,$x10,$key_ # off-load round[2] + addi $key_,$key_,0x20 + bdnz Load_cbc_dec_key + + lvx v26,$x10,$key + ?vperm v24,v30,v31,$keyperm + lvx v27,$x20,$key + stvx v24,$x00,$key_ # off-load round[3] + ?vperm v25,v31,v26,$keyperm + lvx v28,$x30,$key + stvx v25,$x10,$key_ # off-load round[4] + addi $key_,$sp,$FRAME+15 # rewind $key_ + ?vperm v26,v26,v27,$keyperm + lvx v29,$x40,$key + ?vperm v27,v27,v28,$keyperm + lvx v30,$x50,$key + ?vperm v28,v28,v29,$keyperm + lvx v31,$x60,$key + ?vperm v29,v29,v30,$keyperm + lvx $out0,$x70,$key # borrow $out0 + ?vperm v30,v30,v31,$keyperm + lvx v24,$x00,$key_ # pre-load round[1] + ?vperm v31,v31,$out0,$keyperm + lvx v25,$x10,$key_ # pre-load round[2] + + #lvx $inptail,0,$inp # "caller" already did this + #addi $inp,$inp,15 # 15 is not typo + subi $inp,$inp,15 # undo "caller" + + le?li $idx,8 + lvx_u $in0,$x00,$inp # load first 8 "words" + le?lvsl $inpperm,0,$idx + le?vspltisb $tmp,0x0f + lvx_u $in1,$x10,$inp + le?vxor $inpperm,$inpperm,$tmp # transform for lvx_u/stvx_u + lvx_u $in2,$x20,$inp + le?vperm $in0,$in0,$in0,$inpperm + lvx_u $in3,$x30,$inp + le?vperm $in1,$in1,$in1,$inpperm + lvx_u $in4,$x40,$inp + le?vperm $in2,$in2,$in2,$inpperm + vxor $out0,$in0,$rndkey0 + lvx_u $in5,$x50,$inp + le?vperm $in3,$in3,$in3,$inpperm + vxor $out1,$in1,$rndkey0 + lvx_u $in6,$x60,$inp + le?vperm $in4,$in4,$in4,$inpperm + vxor $out2,$in2,$rndkey0 + lvx_u $in7,$x70,$inp + addi $inp,$inp,0x80 + le?vperm $in5,$in5,$in5,$inpperm + vxor $out3,$in3,$rndkey0 + le?vperm $in6,$in6,$in6,$inpperm + vxor $out4,$in4,$rndkey0 + le?vperm $in7,$in7,$in7,$inpperm + vxor $out5,$in5,$rndkey0 + vxor $out6,$in6,$rndkey0 + vxor $out7,$in7,$rndkey0 + + mtctr $rounds + b Loop_cbc_dec8x +.align 5 +Loop_cbc_dec8x: + vncipher $out0,$out0,v24 + vncipher $out1,$out1,v24 + vncipher $out2,$out2,v24 + vncipher $out3,$out3,v24 + vncipher $out4,$out4,v24 + vncipher $out5,$out5,v24 + vncipher $out6,$out6,v24 + vncipher $out7,$out7,v24 + lvx v24,$x20,$key_ # round[3] + addi $key_,$key_,0x20 + + vncipher $out0,$out0,v25 + vncipher $out1,$out1,v25 + vncipher $out2,$out2,v25 + vncipher $out3,$out3,v25 + vncipher $out4,$out4,v25 + vncipher $out5,$out5,v25 + vncipher $out6,$out6,v25 + vncipher $out7,$out7,v25 + lvx v25,$x10,$key_ # round[4] + bdnz Loop_cbc_dec8x + + subic $len,$len,128 # $len-=128 + vncipher $out0,$out0,v24 + vncipher $out1,$out1,v24 + vncipher $out2,$out2,v24 + vncipher $out3,$out3,v24 + vncipher $out4,$out4,v24 + vncipher $out5,$out5,v24 + vncipher $out6,$out6,v24 + vncipher $out7,$out7,v24 + + subfe. r0,r0,r0 # borrow?-1:0 + vncipher $out0,$out0,v25 + vncipher $out1,$out1,v25 + vncipher $out2,$out2,v25 + vncipher $out3,$out3,v25 + vncipher $out4,$out4,v25 + vncipher $out5,$out5,v25 + vncipher $out6,$out6,v25 + vncipher $out7,$out7,v25 + + and r0,r0,$len + vncipher $out0,$out0,v26 + vncipher $out1,$out1,v26 + vncipher $out2,$out2,v26 + vncipher $out3,$out3,v26 + vncipher $out4,$out4,v26 + vncipher $out5,$out5,v26 + vncipher $out6,$out6,v26 + vncipher $out7,$out7,v26 + + add $inp,$inp,r0 # $inp is adjusted in such + # way that at exit from the + # loop inX-in7 are loaded + # with last "words" + vncipher $out0,$out0,v27 + vncipher $out1,$out1,v27 + vncipher $out2,$out2,v27 + vncipher $out3,$out3,v27 + vncipher $out4,$out4,v27 + vncipher $out5,$out5,v27 + vncipher $out6,$out6,v27 + vncipher $out7,$out7,v27 + + addi $key_,$sp,$FRAME+15 # rewind $key_ + vncipher $out0,$out0,v28 + vncipher $out1,$out1,v28 + vncipher $out2,$out2,v28 + vncipher $out3,$out3,v28 + vncipher $out4,$out4,v28 + vncipher $out5,$out5,v28 + vncipher $out6,$out6,v28 + vncipher $out7,$out7,v28 + lvx v24,$x00,$key_ # re-pre-load round[1] + + vncipher $out0,$out0,v29 + vncipher $out1,$out1,v29 + vncipher $out2,$out2,v29 + vncipher $out3,$out3,v29 + vncipher $out4,$out4,v29 + vncipher $out5,$out5,v29 + vncipher $out6,$out6,v29 + vncipher $out7,$out7,v29 + lvx v25,$x10,$key_ # re-pre-load round[2] + + vncipher $out0,$out0,v30 + vxor $ivec,$ivec,v31 # xor with last round key + vncipher $out1,$out1,v30 + vxor $in0,$in0,v31 + vncipher $out2,$out2,v30 + vxor $in1,$in1,v31 + vncipher $out3,$out3,v30 + vxor $in2,$in2,v31 + vncipher $out4,$out4,v30 + vxor $in3,$in3,v31 + vncipher $out5,$out5,v30 + vxor $in4,$in4,v31 + vncipher $out6,$out6,v30 + vxor $in5,$in5,v31 + vncipher $out7,$out7,v30 + vxor $in6,$in6,v31 + + vncipherlast $out0,$out0,$ivec + vncipherlast $out1,$out1,$in0 + lvx_u $in0,$x00,$inp # load next input block + vncipherlast $out2,$out2,$in1 + lvx_u $in1,$x10,$inp + vncipherlast $out3,$out3,$in2 + le?vperm $in0,$in0,$in0,$inpperm + lvx_u $in2,$x20,$inp + vncipherlast $out4,$out4,$in3 + le?vperm $in1,$in1,$in1,$inpperm + lvx_u $in3,$x30,$inp + vncipherlast $out5,$out5,$in4 + le?vperm $in2,$in2,$in2,$inpperm + lvx_u $in4,$x40,$inp + vncipherlast $out6,$out6,$in5 + le?vperm $in3,$in3,$in3,$inpperm + lvx_u $in5,$x50,$inp + vncipherlast $out7,$out7,$in6 + le?vperm $in4,$in4,$in4,$inpperm + lvx_u $in6,$x60,$inp + vmr $ivec,$in7 + le?vperm $in5,$in5,$in5,$inpperm + lvx_u $in7,$x70,$inp + addi $inp,$inp,0x80 + + le?vperm $out0,$out0,$out0,$inpperm + le?vperm $out1,$out1,$out1,$inpperm + stvx_u $out0,$x00,$out + le?vperm $in6,$in6,$in6,$inpperm + vxor $out0,$in0,$rndkey0 + le?vperm $out2,$out2,$out2,$inpperm + stvx_u $out1,$x10,$out + le?vperm $in7,$in7,$in7,$inpperm + vxor $out1,$in1,$rndkey0 + le?vperm $out3,$out3,$out3,$inpperm + stvx_u $out2,$x20,$out + vxor $out2,$in2,$rndkey0 + le?vperm $out4,$out4,$out4,$inpperm + stvx_u $out3,$x30,$out + vxor $out3,$in3,$rndkey0 + le?vperm $out5,$out5,$out5,$inpperm + stvx_u $out4,$x40,$out + vxor $out4,$in4,$rndkey0 + le?vperm $out6,$out6,$out6,$inpperm + stvx_u $out5,$x50,$out + vxor $out5,$in5,$rndkey0 + le?vperm $out7,$out7,$out7,$inpperm + stvx_u $out6,$x60,$out + vxor $out6,$in6,$rndkey0 + stvx_u $out7,$x70,$out + addi $out,$out,0x80 + vxor $out7,$in7,$rndkey0 + + mtctr $rounds + beq Loop_cbc_dec8x # did $len-=128 borrow? + + addic. $len,$len,128 + beq Lcbc_dec8x_done + nop + nop + +Loop_cbc_dec8x_tail: # up to 7 "words" tail... + vncipher $out1,$out1,v24 + vncipher $out2,$out2,v24 + vncipher $out3,$out3,v24 + vncipher $out4,$out4,v24 + vncipher $out5,$out5,v24 + vncipher $out6,$out6,v24 + vncipher $out7,$out7,v24 + lvx v24,$x20,$key_ # round[3] + addi $key_,$key_,0x20 + + vncipher $out1,$out1,v25 + vncipher $out2,$out2,v25 + vncipher $out3,$out3,v25 + vncipher $out4,$out4,v25 + vncipher $out5,$out5,v25 + vncipher $out6,$out6,v25 + vncipher $out7,$out7,v25 + lvx v25,$x10,$key_ # round[4] + bdnz Loop_cbc_dec8x_tail + + vncipher $out1,$out1,v24 + vncipher $out2,$out2,v24 + vncipher $out3,$out3,v24 + vncipher $out4,$out4,v24 + vncipher $out5,$out5,v24 + vncipher $out6,$out6,v24 + vncipher $out7,$out7,v24 + + vncipher $out1,$out1,v25 + vncipher $out2,$out2,v25 + vncipher $out3,$out3,v25 + vncipher $out4,$out4,v25 + vncipher $out5,$out5,v25 + vncipher $out6,$out6,v25 + vncipher $out7,$out7,v25 + + vncipher $out1,$out1,v26 + vncipher $out2,$out2,v26 + vncipher $out3,$out3,v26 + vncipher $out4,$out4,v26 + vncipher $out5,$out5,v26 + vncipher $out6,$out6,v26 + vncipher $out7,$out7,v26 + + vncipher $out1,$out1,v27 + vncipher $out2,$out2,v27 + vncipher $out3,$out3,v27 + vncipher $out4,$out4,v27 + vncipher $out5,$out5,v27 + vncipher $out6,$out6,v27 + vncipher $out7,$out7,v27 + + vncipher $out1,$out1,v28 + vncipher $out2,$out2,v28 + vncipher $out3,$out3,v28 + vncipher $out4,$out4,v28 + vncipher $out5,$out5,v28 + vncipher $out6,$out6,v28 + vncipher $out7,$out7,v28 + + vncipher $out1,$out1,v29 + vncipher $out2,$out2,v29 + vncipher $out3,$out3,v29 + vncipher $out4,$out4,v29 + vncipher $out5,$out5,v29 + vncipher $out6,$out6,v29 + vncipher $out7,$out7,v29 + + vncipher $out1,$out1,v30 + vxor $ivec,$ivec,v31 # last round key + vncipher $out2,$out2,v30 + vxor $in1,$in1,v31 + vncipher $out3,$out3,v30 + vxor $in2,$in2,v31 + vncipher $out4,$out4,v30 + vxor $in3,$in3,v31 + vncipher $out5,$out5,v30 + vxor $in4,$in4,v31 + vncipher $out6,$out6,v30 + vxor $in5,$in5,v31 + vncipher $out7,$out7,v30 + vxor $in6,$in6,v31 + + cmplwi $len,32 # switch($len) + blt Lcbc_dec8x_one + nop + beq Lcbc_dec8x_two + cmplwi $len,64 + blt Lcbc_dec8x_three + nop + beq Lcbc_dec8x_four + cmplwi $len,96 + blt Lcbc_dec8x_five + nop + beq Lcbc_dec8x_six + +Lcbc_dec8x_seven: + vncipherlast $out1,$out1,$ivec + vncipherlast $out2,$out2,$in1 + vncipherlast $out3,$out3,$in2 + vncipherlast $out4,$out4,$in3 + vncipherlast $out5,$out5,$in4 + vncipherlast $out6,$out6,$in5 + vncipherlast $out7,$out7,$in6 + vmr $ivec,$in7 + + le?vperm $out1,$out1,$out1,$inpperm + le?vperm $out2,$out2,$out2,$inpperm + stvx_u $out1,$x00,$out + le?vperm $out3,$out3,$out3,$inpperm + stvx_u $out2,$x10,$out + le?vperm $out4,$out4,$out4,$inpperm + stvx_u $out3,$x20,$out + le?vperm $out5,$out5,$out5,$inpperm + stvx_u $out4,$x30,$out + le?vperm $out6,$out6,$out6,$inpperm + stvx_u $out5,$x40,$out + le?vperm $out7,$out7,$out7,$inpperm + stvx_u $out6,$x50,$out + stvx_u $out7,$x60,$out + addi $out,$out,0x70 + b Lcbc_dec8x_done + +.align 5 +Lcbc_dec8x_six: + vncipherlast $out2,$out2,$ivec + vncipherlast $out3,$out3,$in2 + vncipherlast $out4,$out4,$in3 + vncipherlast $out5,$out5,$in4 + vncipherlast $out6,$out6,$in5 + vncipherlast $out7,$out7,$in6 + vmr $ivec,$in7 + + le?vperm $out2,$out2,$out2,$inpperm + le?vperm $out3,$out3,$out3,$inpperm + stvx_u $out2,$x00,$out + le?vperm $out4,$out4,$out4,$inpperm + stvx_u $out3,$x10,$out + le?vperm $out5,$out5,$out5,$inpperm + stvx_u $out4,$x20,$out + le?vperm $out6,$out6,$out6,$inpperm + stvx_u $out5,$x30,$out + le?vperm $out7,$out7,$out7,$inpperm + stvx_u $out6,$x40,$out + stvx_u $out7,$x50,$out + addi $out,$out,0x60 + b Lcbc_dec8x_done + +.align 5 +Lcbc_dec8x_five: + vncipherlast $out3,$out3,$ivec + vncipherlast $out4,$out4,$in3 + vncipherlast $out5,$out5,$in4 + vncipherlast $out6,$out6,$in5 + vncipherlast $out7,$out7,$in6 + vmr $ivec,$in7 + + le?vperm $out3,$out3,$out3,$inpperm + le?vperm $out4,$out4,$out4,$inpperm + stvx_u $out3,$x00,$out + le?vperm $out5,$out5,$out5,$inpperm + stvx_u $out4,$x10,$out + le?vperm $out6,$out6,$out6,$inpperm + stvx_u $out5,$x20,$out + le?vperm $out7,$out7,$out7,$inpperm + stvx_u $out6,$x30,$out + stvx_u $out7,$x40,$out + addi $out,$out,0x50 + b Lcbc_dec8x_done + +.align 5 +Lcbc_dec8x_four: + vncipherlast $out4,$out4,$ivec + vncipherlast $out5,$out5,$in4 + vncipherlast $out6,$out6,$in5 + vncipherlast $out7,$out7,$in6 + vmr $ivec,$in7 + + le?vperm $out4,$out4,$out4,$inpperm + le?vperm $out5,$out5,$out5,$inpperm + stvx_u $out4,$x00,$out + le?vperm $out6,$out6,$out6,$inpperm + stvx_u $out5,$x10,$out + le?vperm $out7,$out7,$out7,$inpperm + stvx_u $out6,$x20,$out + stvx_u $out7,$x30,$out + addi $out,$out,0x40 + b Lcbc_dec8x_done + +.align 5 +Lcbc_dec8x_three: + vncipherlast $out5,$out5,$ivec + vncipherlast $out6,$out6,$in5 + vncipherlast $out7,$out7,$in6 + vmr $ivec,$in7 + + le?vperm $out5,$out5,$out5,$inpperm + le?vperm $out6,$out6,$out6,$inpperm + stvx_u $out5,$x00,$out + le?vperm $out7,$out7,$out7,$inpperm + stvx_u $out6,$x10,$out + stvx_u $out7,$x20,$out + addi $out,$out,0x30 + b Lcbc_dec8x_done + +.align 5 +Lcbc_dec8x_two: + vncipherlast $out6,$out6,$ivec + vncipherlast $out7,$out7,$in6 + vmr $ivec,$in7 + + le?vperm $out6,$out6,$out6,$inpperm + le?vperm $out7,$out7,$out7,$inpperm + stvx_u $out6,$x00,$out + stvx_u $out7,$x10,$out + addi $out,$out,0x20 + b Lcbc_dec8x_done + +.align 5 +Lcbc_dec8x_one: + vncipherlast $out7,$out7,$ivec + vmr $ivec,$in7 + + le?vperm $out7,$out7,$out7,$inpperm + stvx_u $out7,0,$out + addi $out,$out,0x10 + +Lcbc_dec8x_done: + le?vperm $ivec,$ivec,$ivec,$inpperm + stvx_u $ivec,0,$ivp # write [unaligned] iv + + li r10,`$FRAME+15` + li r11,`$FRAME+31` + stvx $inpperm,r10,$sp # wipe copies of round keys + addi r10,r10,32 + stvx $inpperm,r11,$sp + addi r11,r11,32 + stvx $inpperm,r10,$sp + addi r10,r10,32 + stvx $inpperm,r11,$sp + addi r11,r11,32 + stvx $inpperm,r10,$sp + addi r10,r10,32 + stvx $inpperm,r11,$sp + addi r11,r11,32 + stvx $inpperm,r10,$sp + addi r10,r10,32 + stvx $inpperm,r11,$sp + addi r11,r11,32 + + mtspr 256,$vrsave + lvx v20,r10,$sp # ABI says so + addi r10,r10,32 + lvx v21,r11,$sp + addi r11,r11,32 + lvx v22,r10,$sp + addi r10,r10,32 + lvx v23,r11,$sp + addi r11,r11,32 + lvx v24,r10,$sp + addi r10,r10,32 + lvx v25,r11,$sp + addi r11,r11,32 + lvx v26,r10,$sp + addi r10,r10,32 + lvx v27,r11,$sp + addi r11,r11,32 + lvx v28,r10,$sp + addi r10,r10,32 + lvx v29,r11,$sp + addi r11,r11,32 + lvx v30,r10,$sp + lvx v31,r11,$sp + $POP r26,`$FRAME+21*16+0*$SIZE_T`($sp) + $POP r27,`$FRAME+21*16+1*$SIZE_T`($sp) + $POP r28,`$FRAME+21*16+2*$SIZE_T`($sp) + $POP r29,`$FRAME+21*16+3*$SIZE_T`($sp) + $POP r30,`$FRAME+21*16+4*$SIZE_T`($sp) + $POP r31,`$FRAME+21*16+5*$SIZE_T`($sp) + addi $sp,$sp,`$FRAME+21*16+6*$SIZE_T` + blr + .long 0 + .byte 0,12,0x14,0,0x80,6,6,0 + .long 0 +.size .${prefix}_cbc_encrypt,.-.${prefix}_cbc_encrypt +___ +}} }}} + +######################################################################### +{{{ # CTR procedure[s] # +my ($inp,$out,$len,$key,$ivp,$x10,$rounds,$idx)=map("r$_",(3..10)); +my ($rndkey0,$rndkey1,$inout,$tmp)= map("v$_",(0..3)); +my ($ivec,$inptail,$inpperm,$outhead,$outperm,$outmask,$keyperm,$one)= + map("v$_",(4..11)); +my $dat=$tmp; + +$code.=<<___; +.globl .${prefix}_ctr32_encrypt_blocks + ${UCMP}i $len,1 + bltlr- + + lis r0,0xfff0 + mfspr $vrsave,256 + mtspr 256,r0 + + li $idx,15 + vxor $rndkey0,$rndkey0,$rndkey0 + le?vspltisb $tmp,0x0f + + lvx $ivec,0,$ivp # load [unaligned] iv + lvsl $inpperm,0,$ivp + lvx $inptail,$idx,$ivp + vspltisb $one,1 + le?vxor $inpperm,$inpperm,$tmp + vperm $ivec,$ivec,$inptail,$inpperm + vsldoi $one,$rndkey0,$one,1 + + neg r11,$inp + ?lvsl $keyperm,0,$key # prepare for unaligned key + lwz $rounds,240($key) + + lvsr $inpperm,0,r11 # prepare for unaligned load + lvx $inptail,0,$inp + addi $inp,$inp,15 # 15 is not typo + le?vxor $inpperm,$inpperm,$tmp + + srwi $rounds,$rounds,1 + li $idx,16 + subi $rounds,$rounds,1 + + ${UCMP}i $len,8 + bge _aesp8_ctr32_encrypt8x + + ?lvsr $outperm,0,$out # prepare for unaligned store + vspltisb $outmask,-1 + lvx $outhead,0,$out + ?vperm $outmask,$rndkey0,$outmask,$outperm + le?vxor $outperm,$outperm,$tmp + + lvx $rndkey0,0,$key + mtctr $rounds + lvx $rndkey1,$idx,$key + addi $idx,$idx,16 + ?vperm $rndkey0,$rndkey0,$rndkey1,$keyperm + vxor $inout,$ivec,$rndkey0 + lvx $rndkey0,$idx,$key + addi $idx,$idx,16 + b Loop_ctr32_enc + +.align 5 +Loop_ctr32_enc: + ?vperm $rndkey1,$rndkey1,$rndkey0,$keyperm + vcipher $inout,$inout,$rndkey1 + lvx $rndkey1,$idx,$key + addi $idx,$idx,16 + ?vperm $rndkey0,$rndkey0,$rndkey1,$keyperm + vcipher $inout,$inout,$rndkey0 + lvx $rndkey0,$idx,$key + addi $idx,$idx,16 + bdnz Loop_ctr32_enc + + vadduwm $ivec,$ivec,$one + vmr $dat,$inptail + lvx $inptail,0,$inp + addi $inp,$inp,16 + subic. $len,$len,1 # blocks-- + + ?vperm $rndkey1,$rndkey1,$rndkey0,$keyperm + vcipher $inout,$inout,$rndkey1 + lvx $rndkey1,$idx,$key + vperm $dat,$dat,$inptail,$inpperm + li $idx,16 + ?vperm $rndkey1,$rndkey0,$rndkey1,$keyperm + lvx $rndkey0,0,$key + vxor $dat,$dat,$rndkey1 # last round key + vcipherlast $inout,$inout,$dat + + lvx $rndkey1,$idx,$key + addi $idx,$idx,16 + vperm $inout,$inout,$inout,$outperm + vsel $dat,$outhead,$inout,$outmask + mtctr $rounds + ?vperm $rndkey0,$rndkey0,$rndkey1,$keyperm + vmr $outhead,$inout + vxor $inout,$ivec,$rndkey0 + lvx $rndkey0,$idx,$key + addi $idx,$idx,16 + stvx $dat,0,$out + addi $out,$out,16 + bne Loop_ctr32_enc + + addi $out,$out,-1 + lvx $inout,0,$out # redundant in aligned case + vsel $inout,$outhead,$inout,$outmask + stvx $inout,0,$out + + mtspr 256,$vrsave + blr + .long 0 + .byte 0,12,0x14,0,0,0,6,0 + .long 0 +___ +######################################################################### +{{ # Optimized CTR procedure # +my $key_="r11"; +my ($x00,$x10,$x20,$x30,$x40,$x50,$x60,$x70)=map("r$_",(0,8,26..31)); +my ($in0, $in1, $in2, $in3, $in4, $in5, $in6, $in7 )=map("v$_",(0..3,10,12..14)); +my ($out0,$out1,$out2,$out3,$out4,$out5,$out6,$out7)=map("v$_",(15..22)); +my $rndkey0="v23"; # v24-v25 rotating buffer for first found keys + # v26-v31 last 6 round keys +my ($tmp,$keyperm)=($in3,$in4); # aliases with "caller", redundant assignment +my ($two,$three,$four)=($outhead,$outperm,$outmask); + +$code.=<<___; +.align 5 +_aesp8_ctr32_encrypt8x: + $STU $sp,-`($FRAME+21*16+6*$SIZE_T)`($sp) + li r10,`$FRAME+8*16+15` + li r11,`$FRAME+8*16+31` + stvx v20,r10,$sp # ABI says so + addi r10,r10,32 + stvx v21,r11,$sp + addi r11,r11,32 + stvx v22,r10,$sp + addi r10,r10,32 + stvx v23,r11,$sp + addi r11,r11,32 + stvx v24,r10,$sp + addi r10,r10,32 + stvx v25,r11,$sp + addi r11,r11,32 + stvx v26,r10,$sp + addi r10,r10,32 + stvx v27,r11,$sp + addi r11,r11,32 + stvx v28,r10,$sp + addi r10,r10,32 + stvx v29,r11,$sp + addi r11,r11,32 + stvx v30,r10,$sp + stvx v31,r11,$sp + li r0,-1 + stw $vrsave,`$FRAME+21*16-4`($sp) # save vrsave + li $x10,0x10 + $PUSH r26,`$FRAME+21*16+0*$SIZE_T`($sp) + li $x20,0x20 + $PUSH r27,`$FRAME+21*16+1*$SIZE_T`($sp) + li $x30,0x30 + $PUSH r28,`$FRAME+21*16+2*$SIZE_T`($sp) + li $x40,0x40 + $PUSH r29,`$FRAME+21*16+3*$SIZE_T`($sp) + li $x50,0x50 + $PUSH r30,`$FRAME+21*16+4*$SIZE_T`($sp) + li $x60,0x60 + $PUSH r31,`$FRAME+21*16+5*$SIZE_T`($sp) + li $x70,0x70 + mtspr 256,r0 + + subi $rounds,$rounds,3 # -4 in total + + lvx $rndkey0,$x00,$key # load key schedule + lvx v30,$x10,$key + addi $key,$key,0x20 + lvx v31,$x00,$key + ?vperm $rndkey0,$rndkey0,v30,$keyperm + addi $key_,$sp,$FRAME+15 + mtctr $rounds + +Load_ctr32_enc_key: + ?vperm v24,v30,v31,$keyperm + lvx v30,$x10,$key + addi $key,$key,0x20 + stvx v24,$x00,$key_ # off-load round[1] + ?vperm v25,v31,v30,$keyperm + lvx v31,$x00,$key + stvx v25,$x10,$key_ # off-load round[2] + addi $key_,$key_,0x20 + bdnz Load_ctr32_enc_key + + lvx v26,$x10,$key + ?vperm v24,v30,v31,$keyperm + lvx v27,$x20,$key + stvx v24,$x00,$key_ # off-load round[3] + ?vperm v25,v31,v26,$keyperm + lvx v28,$x30,$key + stvx v25,$x10,$key_ # off-load round[4] + addi $key_,$sp,$FRAME+15 # rewind $key_ + ?vperm v26,v26,v27,$keyperm + lvx v29,$x40,$key + ?vperm v27,v27,v28,$keyperm + lvx v30,$x50,$key + ?vperm v28,v28,v29,$keyperm + lvx v31,$x60,$key + ?vperm v29,v29,v30,$keyperm + lvx $out0,$x70,$key # borrow $out0 + ?vperm v30,v30,v31,$keyperm + lvx v24,$x00,$key_ # pre-load round[1] + ?vperm v31,v31,$out0,$keyperm + lvx v25,$x10,$key_ # pre-load round[2] + + vadduwm $two,$one,$one + subi $inp,$inp,15 # undo "caller" + $SHL $len,$len,4 + + vadduwm $out1,$ivec,$one # counter values ... + vadduwm $out2,$ivec,$two + vxor $out0,$ivec,$rndkey0 # ... xored with rndkey[0] + le?li $idx,8 + vadduwm $out3,$out1,$two + vxor $out1,$out1,$rndkey0 + le?lvsl $inpperm,0,$idx + vadduwm $out4,$out2,$two + vxor $out2,$out2,$rndkey0 + le?vspltisb $tmp,0x0f + vadduwm $out5,$out3,$two + vxor $out3,$out3,$rndkey0 + le?vxor $inpperm,$inpperm,$tmp # transform for lvx_u/stvx_u + vadduwm $out6,$out4,$two + vxor $out4,$out4,$rndkey0 + vadduwm $out7,$out5,$two + vxor $out5,$out5,$rndkey0 + vadduwm $ivec,$out6,$two # next counter value + vxor $out6,$out6,$rndkey0 + vxor $out7,$out7,$rndkey0 + + mtctr $rounds + b Loop_ctr32_enc8x +.align 5 +Loop_ctr32_enc8x: + vcipher $out0,$out0,v24 + vcipher $out1,$out1,v24 + vcipher $out2,$out2,v24 + vcipher $out3,$out3,v24 + vcipher $out4,$out4,v24 + vcipher $out5,$out5,v24 + vcipher $out6,$out6,v24 + vcipher $out7,$out7,v24 +Loop_ctr32_enc8x_middle: + lvx v24,$x20,$key_ # round[3] + addi $key_,$key_,0x20 + + vcipher $out0,$out0,v25 + vcipher $out1,$out1,v25 + vcipher $out2,$out2,v25 + vcipher $out3,$out3,v25 + vcipher $out4,$out4,v25 + vcipher $out5,$out5,v25 + vcipher $out6,$out6,v25 + vcipher $out7,$out7,v25 + lvx v25,$x10,$key_ # round[4] + bdnz Loop_ctr32_enc8x + + subic r11,$len,256 # $len-256, borrow $key_ + vcipher $out0,$out0,v24 + vcipher $out1,$out1,v24 + vcipher $out2,$out2,v24 + vcipher $out3,$out3,v24 + vcipher $out4,$out4,v24 + vcipher $out5,$out5,v24 + vcipher $out6,$out6,v24 + vcipher $out7,$out7,v24 + + subfe r0,r0,r0 # borrow?-1:0 + vcipher $out0,$out0,v25 + vcipher $out1,$out1,v25 + vcipher $out2,$out2,v25 + vcipher $out3,$out3,v25 + vcipher $out4,$out4,v25 + vcipher $out5,$out5,v25 + vcipher $out6,$out6,v25 + vcipher $out7,$out7,v25 + + and r0,r0,r11 + addi $key_,$sp,$FRAME+15 # rewind $key_ + vcipher $out0,$out0,v26 + vcipher $out1,$out1,v26 + vcipher $out2,$out2,v26 + vcipher $out3,$out3,v26 + vcipher $out4,$out4,v26 + vcipher $out5,$out5,v26 + vcipher $out6,$out6,v26 + vcipher $out7,$out7,v26 + lvx v24,$x00,$key_ # re-pre-load round[1] + + subic $len,$len,129 # $len-=129 + vcipher $out0,$out0,v27 + addi $len,$len,1 # $len-=128 really + vcipher $out1,$out1,v27 + vcipher $out2,$out2,v27 + vcipher $out3,$out3,v27 + vcipher $out4,$out4,v27 + vcipher $out5,$out5,v27 + vcipher $out6,$out6,v27 + vcipher $out7,$out7,v27 + lvx v25,$x10,$key_ # re-pre-load round[2] + + vcipher $out0,$out0,v28 + lvx_u $in0,$x00,$inp # load input + vcipher $out1,$out1,v28 + lvx_u $in1,$x10,$inp + vcipher $out2,$out2,v28 + lvx_u $in2,$x20,$inp + vcipher $out3,$out3,v28 + lvx_u $in3,$x30,$inp + vcipher $out4,$out4,v28 + lvx_u $in4,$x40,$inp + vcipher $out5,$out5,v28 + lvx_u $in5,$x50,$inp + vcipher $out6,$out6,v28 + lvx_u $in6,$x60,$inp + vcipher $out7,$out7,v28 + lvx_u $in7,$x70,$inp + addi $inp,$inp,0x80 + + vcipher $out0,$out0,v29 + le?vperm $in0,$in0,$in0,$inpperm + vcipher $out1,$out1,v29 + le?vperm $in1,$in1,$in1,$inpperm + vcipher $out2,$out2,v29 + le?vperm $in2,$in2,$in2,$inpperm + vcipher $out3,$out3,v29 + le?vperm $in3,$in3,$in3,$inpperm + vcipher $out4,$out4,v29 + le?vperm $in4,$in4,$in4,$inpperm + vcipher $out5,$out5,v29 + le?vperm $in5,$in5,$in5,$inpperm + vcipher $out6,$out6,v29 + le?vperm $in6,$in6,$in6,$inpperm + vcipher $out7,$out7,v29 + le?vperm $in7,$in7,$in7,$inpperm + + add $inp,$inp,r0 # $inp is adjusted in such + # way that at exit from the + # loop inX-in7 are loaded + # with last "words" + subfe. r0,r0,r0 # borrow?-1:0 + vcipher $out0,$out0,v30 + vxor $in0,$in0,v31 # xor with last round key + vcipher $out1,$out1,v30 + vxor $in1,$in1,v31 + vcipher $out2,$out2,v30 + vxor $in2,$in2,v31 + vcipher $out3,$out3,v30 + vxor $in3,$in3,v31 + vcipher $out4,$out4,v30 + vxor $in4,$in4,v31 + vcipher $out5,$out5,v30 + vxor $in5,$in5,v31 + vcipher $out6,$out6,v30 + vxor $in6,$in6,v31 + vcipher $out7,$out7,v30 + vxor $in7,$in7,v31 + + bne Lctr32_enc8x_break # did $len-129 borrow? + + vcipherlast $in0,$out0,$in0 + vcipherlast $in1,$out1,$in1 + vadduwm $out1,$ivec,$one # counter values ... + vcipherlast $in2,$out2,$in2 + vadduwm $out2,$ivec,$two + vxor $out0,$ivec,$rndkey0 # ... xored with rndkey[0] + vcipherlast $in3,$out3,$in3 + vadduwm $out3,$out1,$two + vxor $out1,$out1,$rndkey0 + vcipherlast $in4,$out4,$in4 + vadduwm $out4,$out2,$two + vxor $out2,$out2,$rndkey0 + vcipherlast $in5,$out5,$in5 + vadduwm $out5,$out3,$two + vxor $out3,$out3,$rndkey0 + vcipherlast $in6,$out6,$in6 + vadduwm $out6,$out4,$two + vxor $out4,$out4,$rndkey0 + vcipherlast $in7,$out7,$in7 + vadduwm $out7,$out5,$two + vxor $out5,$out5,$rndkey0 + le?vperm $in0,$in0,$in0,$inpperm + vadduwm $ivec,$out6,$two # next counter value + vxor $out6,$out6,$rndkey0 + le?vperm $in1,$in1,$in1,$inpperm + vxor $out7,$out7,$rndkey0 + mtctr $rounds + + vcipher $out0,$out0,v24 + stvx_u $in0,$x00,$out + le?vperm $in2,$in2,$in2,$inpperm + vcipher $out1,$out1,v24 + stvx_u $in1,$x10,$out + le?vperm $in3,$in3,$in3,$inpperm + vcipher $out2,$out2,v24 + stvx_u $in2,$x20,$out + le?vperm $in4,$in4,$in4,$inpperm + vcipher $out3,$out3,v24 + stvx_u $in3,$x30,$out + le?vperm $in5,$in5,$in5,$inpperm + vcipher $out4,$out4,v24 + stvx_u $in4,$x40,$out + le?vperm $in6,$in6,$in6,$inpperm + vcipher $out5,$out5,v24 + stvx_u $in5,$x50,$out + le?vperm $in7,$in7,$in7,$inpperm + vcipher $out6,$out6,v24 + stvx_u $in6,$x60,$out + vcipher $out7,$out7,v24 + stvx_u $in7,$x70,$out + addi $out,$out,0x80 + + b Loop_ctr32_enc8x_middle + +.align 5 +Lctr32_enc8x_break: + cmpwi $len,-0x60 + blt Lctr32_enc8x_one + nop + beq Lctr32_enc8x_two + cmpwi $len,-0x40 + blt Lctr32_enc8x_three + nop + beq Lctr32_enc8x_four + cmpwi $len,-0x20 + blt Lctr32_enc8x_five + nop + beq Lctr32_enc8x_six + cmpwi $len,0x00 + blt Lctr32_enc8x_seven + +Lctr32_enc8x_eight: + vcipherlast $out0,$out0,$in0 + vcipherlast $out1,$out1,$in1 + vcipherlast $out2,$out2,$in2 + vcipherlast $out3,$out3,$in3 + vcipherlast $out4,$out4,$in4 + vcipherlast $out5,$out5,$in5 + vcipherlast $out6,$out6,$in6 + vcipherlast $out7,$out7,$in7 + + le?vperm $out0,$out0,$out0,$inpperm + le?vperm $out1,$out1,$out1,$inpperm + stvx_u $out0,$x00,$out + le?vperm $out2,$out2,$out2,$inpperm + stvx_u $out1,$x10,$out + le?vperm $out3,$out3,$out3,$inpperm + stvx_u $out2,$x20,$out + le?vperm $out4,$out4,$out4,$inpperm + stvx_u $out3,$x30,$out + le?vperm $out5,$out5,$out5,$inpperm + stvx_u $out4,$x40,$out + le?vperm $out6,$out6,$out6,$inpperm + stvx_u $out5,$x50,$out + le?vperm $out7,$out7,$out7,$inpperm + stvx_u $out6,$x60,$out + stvx_u $out7,$x70,$out + addi $out,$out,0x80 + b Lctr32_enc8x_done + +.align 5 +Lctr32_enc8x_seven: + vcipherlast $out0,$out0,$in1 + vcipherlast $out1,$out1,$in2 + vcipherlast $out2,$out2,$in3 + vcipherlast $out3,$out3,$in4 + vcipherlast $out4,$out4,$in5 + vcipherlast $out5,$out5,$in6 + vcipherlast $out6,$out6,$in7 + + le?vperm $out0,$out0,$out0,$inpperm + le?vperm $out1,$out1,$out1,$inpperm + stvx_u $out0,$x00,$out + le?vperm $out2,$out2,$out2,$inpperm + stvx_u $out1,$x10,$out + le?vperm $out3,$out3,$out3,$inpperm + stvx_u $out2,$x20,$out + le?vperm $out4,$out4,$out4,$inpperm + stvx_u $out3,$x30,$out + le?vperm $out5,$out5,$out5,$inpperm + stvx_u $out4,$x40,$out + le?vperm $out6,$out6,$out6,$inpperm + stvx_u $out5,$x50,$out + stvx_u $out6,$x60,$out + addi $out,$out,0x70 + b Lctr32_enc8x_done + +.align 5 +Lctr32_enc8x_six: + vcipherlast $out0,$out0,$in2 + vcipherlast $out1,$out1,$in3 + vcipherlast $out2,$out2,$in4 + vcipherlast $out3,$out3,$in5 + vcipherlast $out4,$out4,$in6 + vcipherlast $out5,$out5,$in7 + + le?vperm $out0,$out0,$out0,$inpperm + le?vperm $out1,$out1,$out1,$inpperm + stvx_u $out0,$x00,$out + le?vperm $out2,$out2,$out2,$inpperm + stvx_u $out1,$x10,$out + le?vperm $out3,$out3,$out3,$inpperm + stvx_u $out2,$x20,$out + le?vperm $out4,$out4,$out4,$inpperm + stvx_u $out3,$x30,$out + le?vperm $out5,$out5,$out5,$inpperm + stvx_u $out4,$x40,$out + stvx_u $out5,$x50,$out + addi $out,$out,0x60 + b Lctr32_enc8x_done + +.align 5 +Lctr32_enc8x_five: + vcipherlast $out0,$out0,$in3 + vcipherlast $out1,$out1,$in4 + vcipherlast $out2,$out2,$in5 + vcipherlast $out3,$out3,$in6 + vcipherlast $out4,$out4,$in7 + + le?vperm $out0,$out0,$out0,$inpperm + le?vperm $out1,$out1,$out1,$inpperm + stvx_u $out0,$x00,$out + le?vperm $out2,$out2,$out2,$inpperm + stvx_u $out1,$x10,$out + le?vperm $out3,$out3,$out3,$inpperm + stvx_u $out2,$x20,$out + le?vperm $out4,$out4,$out4,$inpperm + stvx_u $out3,$x30,$out + stvx_u $out4,$x40,$out + addi $out,$out,0x50 + b Lctr32_enc8x_done + +.align 5 +Lctr32_enc8x_four: + vcipherlast $out0,$out0,$in4 + vcipherlast $out1,$out1,$in5 + vcipherlast $out2,$out2,$in6 + vcipherlast $out3,$out3,$in7 + + le?vperm $out0,$out0,$out0,$inpperm + le?vperm $out1,$out1,$out1,$inpperm + stvx_u $out0,$x00,$out + le?vperm $out2,$out2,$out2,$inpperm + stvx_u $out1,$x10,$out + le?vperm $out3,$out3,$out3,$inpperm + stvx_u $out2,$x20,$out + stvx_u $out3,$x30,$out + addi $out,$out,0x40 + b Lctr32_enc8x_done + +.align 5 +Lctr32_enc8x_three: + vcipherlast $out0,$out0,$in5 + vcipherlast $out1,$out1,$in6 + vcipherlast $out2,$out2,$in7 + + le?vperm $out0,$out0,$out0,$inpperm + le?vperm $out1,$out1,$out1,$inpperm + stvx_u $out0,$x00,$out + le?vperm $out2,$out2,$out2,$inpperm + stvx_u $out1,$x10,$out + stvx_u $out2,$x20,$out + addi $out,$out,0x30 + b Lcbc_dec8x_done + +.align 5 +Lctr32_enc8x_two: + vcipherlast $out0,$out0,$in6 + vcipherlast $out1,$out1,$in7 + + le?vperm $out0,$out0,$out0,$inpperm + le?vperm $out1,$out1,$out1,$inpperm + stvx_u $out0,$x00,$out + stvx_u $out1,$x10,$out + addi $out,$out,0x20 + b Lcbc_dec8x_done + +.align 5 +Lctr32_enc8x_one: + vcipherlast $out0,$out0,$in7 + + le?vperm $out0,$out0,$out0,$inpperm + stvx_u $out0,0,$out + addi $out,$out,0x10 + +Lctr32_enc8x_done: + li r10,`$FRAME+15` + li r11,`$FRAME+31` + stvx $inpperm,r10,$sp # wipe copies of round keys + addi r10,r10,32 + stvx $inpperm,r11,$sp + addi r11,r11,32 + stvx $inpperm,r10,$sp + addi r10,r10,32 + stvx $inpperm,r11,$sp + addi r11,r11,32 + stvx $inpperm,r10,$sp + addi r10,r10,32 + stvx $inpperm,r11,$sp + addi r11,r11,32 + stvx $inpperm,r10,$sp + addi r10,r10,32 + stvx $inpperm,r11,$sp + addi r11,r11,32 + + mtspr 256,$vrsave + lvx v20,r10,$sp # ABI says so + addi r10,r10,32 + lvx v21,r11,$sp + addi r11,r11,32 + lvx v22,r10,$sp + addi r10,r10,32 + lvx v23,r11,$sp + addi r11,r11,32 + lvx v24,r10,$sp + addi r10,r10,32 + lvx v25,r11,$sp + addi r11,r11,32 + lvx v26,r10,$sp + addi r10,r10,32 + lvx v27,r11,$sp + addi r11,r11,32 + lvx v28,r10,$sp + addi r10,r10,32 + lvx v29,r11,$sp + addi r11,r11,32 + lvx v30,r10,$sp + lvx v31,r11,$sp + $POP r26,`$FRAME+21*16+0*$SIZE_T`($sp) + $POP r27,`$FRAME+21*16+1*$SIZE_T`($sp) + $POP r28,`$FRAME+21*16+2*$SIZE_T`($sp) + $POP r29,`$FRAME+21*16+3*$SIZE_T`($sp) + $POP r30,`$FRAME+21*16+4*$SIZE_T`($sp) + $POP r31,`$FRAME+21*16+5*$SIZE_T`($sp) + addi $sp,$sp,`$FRAME+21*16+6*$SIZE_T` + blr + .long 0 + .byte 0,12,0x14,0,0x80,6,6,0 + .long 0 +.size .${prefix}_ctr32_encrypt_blocks,.-.${prefix}_ctr32_encrypt_blocks +___ +}} }}} + +my $consts=1; +foreach(split("\n",$code)) { + s/\`([^\`]*)\`/eval($1)/geo; + + # constants table endian-specific conversion + if ($consts && m/\.(long|byte)\s+(.+)\s+(\?[a-z]*)$/o) { + my $conv=$3; + my @bytes=(); + + # convert to endian-agnostic format + if ($1 eq "long") { + foreach (split(/,\s*/,$2)) { + my $l = /^0/?oct:int; + push @bytes,($l>>24)&0xff,($l>>16)&0xff,($l>>8)&0xff,$l&0xff; + } + } else { + @bytes = map(/^0/?oct:int,split(/,\s*/,$2)); + } + + # little-endian conversion + if ($flavour =~ /le$/o) { + SWITCH: for($conv) { + /\?inv/ && do { @bytes=map($_^0xf,@bytes); last; }; + /\?rev/ && do { @bytes=reverse(@bytes); last; }; + } + } + + #emit + print ".byte\t",join(',',map (sprintf("0x%02x",$_),@bytes)),"\n"; + next; + } + $consts=0 if (m/Lconsts:/o); # end of table + + # instructions prefixed with '?' are endian-specific and need + # to be adjusted accordingly... + if ($flavour =~ /le$/o) { # little-endian + s/le\?//o or + s/be\?/#be#/o or + s/\?lvsr/lvsl/o or + s/\?lvsl/lvsr/o or + s/\?(vperm\s+v[0-9]+,\s*)(v[0-9]+,\s*)(v[0-9]+,\s*)(v[0-9]+)/$1$3$2$4/o or + s/\?(vsldoi\s+v[0-9]+,\s*)(v[0-9]+,)\s*(v[0-9]+,\s*)([0-9]+)/$1$3$2 16-$4/o or + s/\?(vspltw\s+v[0-9]+,\s*)(v[0-9]+,)\s*([0-9])/$1$2 3-$3/o; + } else { # big-endian + s/le\?/#le#/o or + s/be\?//o or + s/\?([a-z]+)/$1/o; + } + + print $_,"\n"; +} + +close STDOUT; diff --git a/drivers/crypto/vmx/ghash.c b/drivers/crypto/vmx/ghash.c new file mode 100644 index 000000000000..d0ffe277af5c --- /dev/null +++ b/drivers/crypto/vmx/ghash.c @@ -0,0 +1,214 @@ +/** + * GHASH routines supporting VMX instructions on the Power 8 + * + * Copyright (C) 2015 International Business Machines Inc. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 only. + * + * 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. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. + * + * Author: Marcelo Henrique Cerri <mhcerri@br.ibm.com> + */ + +#include <linux/types.h> +#include <linux/err.h> +#include <linux/crypto.h> +#include <linux/delay.h> +#include <linux/hardirq.h> +#include <asm/switch_to.h> +#include <crypto/aes.h> +#include <crypto/scatterwalk.h> +#include <crypto/internal/hash.h> +#include <crypto/b128ops.h> + +#define IN_INTERRUPT in_interrupt() + +#define GHASH_BLOCK_SIZE (16) +#define GHASH_DIGEST_SIZE (16) +#define GHASH_KEY_LEN (16) + +void gcm_init_p8(u128 htable[16], const u64 Xi[2]); +void gcm_gmult_p8(u64 Xi[2], const u128 htable[16]); +void gcm_ghash_p8(u64 Xi[2], const u128 htable[16], + const u8 *in,size_t len); + +struct p8_ghash_ctx { + u128 htable[16]; + struct crypto_shash *fallback; +}; + +struct p8_ghash_desc_ctx { + u64 shash[2]; + u8 buffer[GHASH_DIGEST_SIZE]; + int bytes; + struct shash_desc fallback_desc; +}; + +static int p8_ghash_init_tfm(struct crypto_tfm *tfm) +{ + const char *alg; + struct crypto_shash *fallback; + struct crypto_shash *shash_tfm = __crypto_shash_cast(tfm); + struct p8_ghash_ctx *ctx = crypto_tfm_ctx(tfm); + + if (!(alg = crypto_tfm_alg_name(tfm))) { + printk(KERN_ERR "Failed to get algorithm name.\n"); + return -ENOENT; + } + + fallback = crypto_alloc_shash(alg, 0 ,CRYPTO_ALG_NEED_FALLBACK); + if (IS_ERR(fallback)) { + printk(KERN_ERR "Failed to allocate transformation for '%s': %ld\n", + alg, PTR_ERR(fallback)); + return PTR_ERR(fallback); + } + printk(KERN_INFO "Using '%s' as fallback implementation.\n", + crypto_tfm_alg_driver_name(crypto_shash_tfm(fallback))); + + crypto_shash_set_flags(fallback, + crypto_shash_get_flags((struct crypto_shash *) tfm)); + ctx->fallback = fallback; + + shash_tfm->descsize = sizeof(struct p8_ghash_desc_ctx) + + crypto_shash_descsize(fallback); + + return 0; +} + +static void p8_ghash_exit_tfm(struct crypto_tfm *tfm) +{ + struct p8_ghash_ctx *ctx = crypto_tfm_ctx(tfm); + + if (ctx->fallback) { + crypto_free_shash(ctx->fallback); + ctx->fallback = NULL; + } +} + +static int p8_ghash_init(struct shash_desc *desc) +{ + struct p8_ghash_ctx *ctx = crypto_tfm_ctx(crypto_shash_tfm(desc->tfm)); + struct p8_ghash_desc_ctx *dctx = shash_desc_ctx(desc); + + dctx->bytes = 0; + memset(dctx->shash, 0, GHASH_DIGEST_SIZE); + dctx->fallback_desc.tfm = ctx->fallback; + dctx->fallback_desc.flags = desc->flags; + return crypto_shash_init(&dctx->fallback_desc); +} + +static int p8_ghash_setkey(struct crypto_shash *tfm, const u8 *key, + unsigned int keylen) +{ + struct p8_ghash_ctx *ctx = crypto_tfm_ctx(crypto_shash_tfm(tfm)); + + if (keylen != GHASH_KEY_LEN) + return -EINVAL; + + pagefault_disable(); + enable_kernel_altivec(); + enable_kernel_fp(); + gcm_init_p8(ctx->htable, (const u64 *) key); + pagefault_enable(); + return crypto_shash_setkey(ctx->fallback, key, keylen); +} + +static int p8_ghash_update(struct shash_desc *desc, + const u8 *src, unsigned int srclen) +{ + unsigned int len; + struct p8_ghash_ctx *ctx = crypto_tfm_ctx(crypto_shash_tfm(desc->tfm)); + struct p8_ghash_desc_ctx *dctx = shash_desc_ctx(desc); + + if (IN_INTERRUPT) { + return crypto_shash_update(&dctx->fallback_desc, src, srclen); + } else { + if (dctx->bytes) { + if (dctx->bytes + srclen < GHASH_DIGEST_SIZE) { + memcpy(dctx->buffer + dctx->bytes, src, srclen); + dctx->bytes += srclen; + return 0; + } + memcpy(dctx->buffer + dctx->bytes, src, + GHASH_DIGEST_SIZE - dctx->bytes); + pagefault_disable(); + enable_kernel_altivec(); + enable_kernel_fp(); + gcm_ghash_p8(dctx->shash, ctx->htable, dctx->buffer, + GHASH_DIGEST_SIZE); + pagefault_enable(); + src += GHASH_DIGEST_SIZE - dctx->bytes; + srclen -= GHASH_DIGEST_SIZE - dctx->bytes; + dctx->bytes = 0; + } + len = srclen & ~(GHASH_DIGEST_SIZE - 1); + if (len) { + pagefault_disable(); + enable_kernel_altivec(); + enable_kernel_fp(); + gcm_ghash_p8(dctx->shash, ctx->htable, src, len); + pagefault_enable(); + src += len; + srclen -= len; + } + if (srclen) { + memcpy(dctx->buffer, src, srclen); + dctx->bytes = srclen; + } + return 0; + } +} + +static int p8_ghash_final(struct shash_desc *desc, u8 *out) +{ + int i; + struct p8_ghash_ctx *ctx = crypto_tfm_ctx(crypto_shash_tfm(desc->tfm)); + struct p8_ghash_desc_ctx *dctx = shash_desc_ctx(desc); + + if (IN_INTERRUPT) { + return crypto_shash_final(&dctx->fallback_desc, out); + } else { + if (dctx->bytes) { + for (i = dctx->bytes; i < GHASH_DIGEST_SIZE; i++) + dctx->buffer[i] = 0; + pagefault_disable(); + enable_kernel_altivec(); + enable_kernel_fp(); + gcm_ghash_p8(dctx->shash, ctx->htable, dctx->buffer, + GHASH_DIGEST_SIZE); + pagefault_enable(); + dctx->bytes = 0; + } + memcpy(out, dctx->shash, GHASH_DIGEST_SIZE); + return 0; + } +} + +struct shash_alg p8_ghash_alg = { + .digestsize = GHASH_DIGEST_SIZE, + .init = p8_ghash_init, + .update = p8_ghash_update, + .final = p8_ghash_final, + .setkey = p8_ghash_setkey, + .descsize = sizeof(struct p8_ghash_desc_ctx), + .base = { + .cra_name = "ghash", + .cra_driver_name = "p8_ghash", + .cra_priority = 1000, + .cra_flags = CRYPTO_ALG_TYPE_SHASH | CRYPTO_ALG_NEED_FALLBACK, + .cra_blocksize = GHASH_BLOCK_SIZE, + .cra_ctxsize = sizeof(struct p8_ghash_ctx), + .cra_module = THIS_MODULE, + .cra_init = p8_ghash_init_tfm, + .cra_exit = p8_ghash_exit_tfm, + }, +}; diff --git a/drivers/crypto/vmx/ghashp8-ppc.pl b/drivers/crypto/vmx/ghashp8-ppc.pl new file mode 100644 index 000000000000..0a6f899839dd --- /dev/null +++ b/drivers/crypto/vmx/ghashp8-ppc.pl @@ -0,0 +1,228 @@ +#!/usr/bin/env perl +# +# ==================================================================== +# Written by Andy Polyakov <appro@openssl.org> for the OpenSSL +# project. The module is, however, dual licensed under OpenSSL and +# CRYPTOGAMS licenses depending on where you obtain it. For further +# details see http://www.openssl.org/~appro/cryptogams/. +# ==================================================================== +# +# GHASH for for PowerISA v2.07. +# +# July 2014 +# +# Accurate performance measurements are problematic, because it's +# always virtualized setup with possibly throttled processor. +# Relative comparison is therefore more informative. This initial +# version is ~2.1x slower than hardware-assisted AES-128-CTR, ~12x +# faster than "4-bit" integer-only compiler-generated 64-bit code. +# "Initial version" means that there is room for futher improvement. + +$flavour=shift; +$output =shift; + +if ($flavour =~ /64/) { + $SIZE_T=8; + $LRSAVE=2*$SIZE_T; + $STU="stdu"; + $POP="ld"; + $PUSH="std"; +} elsif ($flavour =~ /32/) { + $SIZE_T=4; + $LRSAVE=$SIZE_T; + $STU="stwu"; + $POP="lwz"; + $PUSH="stw"; +} else { die "nonsense $flavour"; } + +$0 =~ m/(.*[\/\\])[^\/\\]+$/; $dir=$1; +( $xlate="${dir}ppc-xlate.pl" and -f $xlate ) or +( $xlate="${dir}../../perlasm/ppc-xlate.pl" and -f $xlate) or +die "can't locate ppc-xlate.pl"; + +open STDOUT,"| $^X $xlate $flavour $output" || die "can't call $xlate: $!"; + +my ($Xip,$Htbl,$inp,$len)=map("r$_",(3..6)); # argument block + +my ($Xl,$Xm,$Xh,$IN)=map("v$_",(0..3)); +my ($zero,$t0,$t1,$t2,$xC2,$H,$Hh,$Hl,$lemask)=map("v$_",(4..12)); +my $vrsave="r12"; + +$code=<<___; +.machine "any" + +.text + +.globl .gcm_init_p8 + lis r0,0xfff0 + li r8,0x10 + mfspr $vrsave,256 + li r9,0x20 + mtspr 256,r0 + li r10,0x30 + lvx_u $H,0,r4 # load H + + vspltisb $xC2,-16 # 0xf0 + vspltisb $t0,1 # one + vaddubm $xC2,$xC2,$xC2 # 0xe0 + vxor $zero,$zero,$zero + vor $xC2,$xC2,$t0 # 0xe1 + vsldoi $xC2,$xC2,$zero,15 # 0xe1... + vsldoi $t1,$zero,$t0,1 # ...1 + vaddubm $xC2,$xC2,$xC2 # 0xc2... + vspltisb $t2,7 + vor $xC2,$xC2,$t1 # 0xc2....01 + vspltb $t1,$H,0 # most significant byte + vsl $H,$H,$t0 # H<<=1 + vsrab $t1,$t1,$t2 # broadcast carry bit + vand $t1,$t1,$xC2 + vxor $H,$H,$t1 # twisted H + + vsldoi $H,$H,$H,8 # twist even more ... + vsldoi $xC2,$zero,$xC2,8 # 0xc2.0 + vsldoi $Hl,$zero,$H,8 # ... and split + vsldoi $Hh,$H,$zero,8 + + stvx_u $xC2,0,r3 # save pre-computed table + stvx_u $Hl,r8,r3 + stvx_u $H, r9,r3 + stvx_u $Hh,r10,r3 + + mtspr 256,$vrsave + blr + .long 0 + .byte 0,12,0x14,0,0,0,2,0 + .long 0 +.size .gcm_init_p8,.-.gcm_init_p8 + +.globl .gcm_gmult_p8 + lis r0,0xfff8 + li r8,0x10 + mfspr $vrsave,256 + li r9,0x20 + mtspr 256,r0 + li r10,0x30 + lvx_u $IN,0,$Xip # load Xi + + lvx_u $Hl,r8,$Htbl # load pre-computed table + le?lvsl $lemask,r0,r0 + lvx_u $H, r9,$Htbl + le?vspltisb $t0,0x07 + lvx_u $Hh,r10,$Htbl + le?vxor $lemask,$lemask,$t0 + lvx_u $xC2,0,$Htbl + le?vperm $IN,$IN,$IN,$lemask + vxor $zero,$zero,$zero + + vpmsumd $Xl,$IN,$Hl # H.lo·Xi.lo + vpmsumd $Xm,$IN,$H # H.hi·Xi.lo+H.lo·Xi.hi + vpmsumd $Xh,$IN,$Hh # H.hi·Xi.hi + + vpmsumd $t2,$Xl,$xC2 # 1st phase + + vsldoi $t0,$Xm,$zero,8 + vsldoi $t1,$zero,$Xm,8 + vxor $Xl,$Xl,$t0 + vxor $Xh,$Xh,$t1 + + vsldoi $Xl,$Xl,$Xl,8 + vxor $Xl,$Xl,$t2 + + vsldoi $t1,$Xl,$Xl,8 # 2nd phase + vpmsumd $Xl,$Xl,$xC2 + vxor $t1,$t1,$Xh + vxor $Xl,$Xl,$t1 + + le?vperm $Xl,$Xl,$Xl,$lemask + stvx_u $Xl,0,$Xip # write out Xi + + mtspr 256,$vrsave + blr + .long 0 + .byte 0,12,0x14,0,0,0,2,0 + .long 0 +.size .gcm_gmult_p8,.-.gcm_gmult_p8 + +.globl .gcm_ghash_p8 + lis r0,0xfff8 + li r8,0x10 + mfspr $vrsave,256 + li r9,0x20 + mtspr 256,r0 + li r10,0x30 + lvx_u $Xl,0,$Xip # load Xi + + lvx_u $Hl,r8,$Htbl # load pre-computed table + le?lvsl $lemask,r0,r0 + lvx_u $H, r9,$Htbl + le?vspltisb $t0,0x07 + lvx_u $Hh,r10,$Htbl + le?vxor $lemask,$lemask,$t0 + lvx_u $xC2,0,$Htbl + le?vperm $Xl,$Xl,$Xl,$lemask + vxor $zero,$zero,$zero + + lvx_u $IN,0,$inp + addi $inp,$inp,16 + subi $len,$len,16 + le?vperm $IN,$IN,$IN,$lemask + vxor $IN,$IN,$Xl + b Loop + +.align 5 +Loop: + subic $len,$len,16 + vpmsumd $Xl,$IN,$Hl # H.lo·Xi.lo + subfe. r0,r0,r0 # borrow?-1:0 + vpmsumd $Xm,$IN,$H # H.hi·Xi.lo+H.lo·Xi.hi + and r0,r0,$len + vpmsumd $Xh,$IN,$Hh # H.hi·Xi.hi + add $inp,$inp,r0 + + vpmsumd $t2,$Xl,$xC2 # 1st phase + + vsldoi $t0,$Xm,$zero,8 + vsldoi $t1,$zero,$Xm,8 + vxor $Xl,$Xl,$t0 + vxor $Xh,$Xh,$t1 + + vsldoi $Xl,$Xl,$Xl,8 + vxor $Xl,$Xl,$t2 + lvx_u $IN,0,$inp + addi $inp,$inp,16 + + vsldoi $t1,$Xl,$Xl,8 # 2nd phase + vpmsumd $Xl,$Xl,$xC2 + le?vperm $IN,$IN,$IN,$lemask + vxor $t1,$t1,$Xh + vxor $IN,$IN,$t1 + vxor $IN,$IN,$Xl + beq Loop # did $len-=16 borrow? + + vxor $Xl,$Xl,$t1 + le?vperm $Xl,$Xl,$Xl,$lemask + stvx_u $Xl,0,$Xip # write out Xi + + mtspr 256,$vrsave + blr + .long 0 + .byte 0,12,0x14,0,0,0,4,0 + .long 0 +.size .gcm_ghash_p8,.-.gcm_ghash_p8 + +.asciz "GHASH for PowerISA 2.07, CRYPTOGAMS by <appro\@openssl.org>" +.align 2 +___ + +foreach (split("\n",$code)) { + if ($flavour =~ /le$/o) { # little-endian + s/le\?//o or + s/be\?/#be#/o; + } else { + s/le\?/#le#/o or + s/be\?//o; + } + print $_,"\n"; +} + +close STDOUT; # enforce flush diff --git a/drivers/crypto/vmx/ppc-xlate.pl b/drivers/crypto/vmx/ppc-xlate.pl new file mode 100644 index 000000000000..a59188494af8 --- /dev/null +++ b/drivers/crypto/vmx/ppc-xlate.pl @@ -0,0 +1,207 @@ +#!/usr/bin/env perl + +# PowerPC assembler distiller by <appro>. + +my $flavour = shift; +my $output = shift; +open STDOUT,">$output" || die "can't open $output: $!"; + +my %GLOBALS; +my $dotinlocallabels=($flavour=~/linux/)?1:0; + +################################################################ +# directives which need special treatment on different platforms +################################################################ +my $globl = sub { + my $junk = shift; + my $name = shift; + my $global = \$GLOBALS{$name}; + my $ret; + + $name =~ s|^[\.\_]||; + + SWITCH: for ($flavour) { + /aix/ && do { $name = ".$name"; + last; + }; + /osx/ && do { $name = "_$name"; + last; + }; + /linux/ + && do { $ret = "_GLOBAL($name)"; + last; + }; + } + + $ret = ".globl $name\nalign 5\n$name:" if (!$ret); + $$global = $name; + $ret; +}; +my $text = sub { + my $ret = ($flavour =~ /aix/) ? ".csect\t.text[PR],7" : ".text"; + $ret = ".abiversion 2\n".$ret if ($flavour =~ /linux.*64le/); + $ret; +}; +my $machine = sub { + my $junk = shift; + my $arch = shift; + if ($flavour =~ /osx/) + { $arch =~ s/\"//g; + $arch = ($flavour=~/64/) ? "ppc970-64" : "ppc970" if ($arch eq "any"); + } + ".machine $arch"; +}; +my $size = sub { + if ($flavour =~ /linux/) + { shift; + my $name = shift; $name =~ s|^[\.\_]||; + my $ret = ".size $name,.-".($flavour=~/64$/?".":"").$name; + $ret .= "\n.size .$name,.-.$name" if ($flavour=~/64$/); + $ret; + } + else + { ""; } +}; +my $asciz = sub { + shift; + my $line = join(",",@_); + if ($line =~ /^"(.*)"$/) + { ".byte " . join(",",unpack("C*",$1),0) . "\n.align 2"; } + else + { ""; } +}; +my $quad = sub { + shift; + my @ret; + my ($hi,$lo); + for (@_) { + if (/^0x([0-9a-f]*?)([0-9a-f]{1,8})$/io) + { $hi=$1?"0x$1":"0"; $lo="0x$2"; } + elsif (/^([0-9]+)$/o) + { $hi=$1>>32; $lo=$1&0xffffffff; } # error-prone with 32-bit perl + else + { $hi=undef; $lo=$_; } + + if (defined($hi)) + { push(@ret,$flavour=~/le$/o?".long\t$lo,$hi":".long\t$hi,$lo"); } + else + { push(@ret,".quad $lo"); } + } + join("\n",@ret); +}; + +################################################################ +# simplified mnemonics not handled by at least one assembler +################################################################ +my $cmplw = sub { + my $f = shift; + my $cr = 0; $cr = shift if ($#_>1); + # Some out-of-date 32-bit GNU assembler just can't handle cmplw... + ($flavour =~ /linux.*32/) ? + " .long ".sprintf "0x%x",31<<26|$cr<<23|$_[0]<<16|$_[1]<<11|64 : + " cmplw ".join(',',$cr,@_); +}; +my $bdnz = sub { + my $f = shift; + my $bo = $f=~/[\+\-]/ ? 16+9 : 16; # optional "to be taken" hint + " bc $bo,0,".shift; +} if ($flavour!~/linux/); +my $bltlr = sub { + my $f = shift; + my $bo = $f=~/\-/ ? 12+2 : 12; # optional "not to be taken" hint + ($flavour =~ /linux/) ? # GNU as doesn't allow most recent hints + " .long ".sprintf "0x%x",19<<26|$bo<<21|16<<1 : + " bclr $bo,0"; +}; +my $bnelr = sub { + my $f = shift; + my $bo = $f=~/\-/ ? 4+2 : 4; # optional "not to be taken" hint + ($flavour =~ /linux/) ? # GNU as doesn't allow most recent hints + " .long ".sprintf "0x%x",19<<26|$bo<<21|2<<16|16<<1 : + " bclr $bo,2"; +}; +my $beqlr = sub { + my $f = shift; + my $bo = $f=~/-/ ? 12+2 : 12; # optional "not to be taken" hint + ($flavour =~ /linux/) ? # GNU as doesn't allow most recent hints + " .long ".sprintf "0x%X",19<<26|$bo<<21|2<<16|16<<1 : + " bclr $bo,2"; +}; +# GNU assembler can't handle extrdi rA,rS,16,48, or when sum of last two +# arguments is 64, with "operand out of range" error. +my $extrdi = sub { + my ($f,$ra,$rs,$n,$b) = @_; + $b = ($b+$n)&63; $n = 64-$n; + " rldicl $ra,$rs,$b,$n"; +}; +my $vmr = sub { + my ($f,$vx,$vy) = @_; + " vor $vx,$vy,$vy"; +}; + +# PowerISA 2.06 stuff +sub vsxmem_op { + my ($f, $vrt, $ra, $rb, $op) = @_; + " .long ".sprintf "0x%X",(31<<26)|($vrt<<21)|($ra<<16)|($rb<<11)|($op*2+1); +} +# made-up unaligned memory reference AltiVec/VMX instructions +my $lvx_u = sub { vsxmem_op(@_, 844); }; # lxvd2x +my $stvx_u = sub { vsxmem_op(@_, 972); }; # stxvd2x +my $lvdx_u = sub { vsxmem_op(@_, 588); }; # lxsdx +my $stvdx_u = sub { vsxmem_op(@_, 716); }; # stxsdx +my $lvx_4w = sub { vsxmem_op(@_, 780); }; # lxvw4x +my $stvx_4w = sub { vsxmem_op(@_, 908); }; # stxvw4x + +# PowerISA 2.07 stuff +sub vcrypto_op { + my ($f, $vrt, $vra, $vrb, $op) = @_; + " .long ".sprintf "0x%X",(4<<26)|($vrt<<21)|($vra<<16)|($vrb<<11)|$op; +} +my $vcipher = sub { vcrypto_op(@_, 1288); }; +my $vcipherlast = sub { vcrypto_op(@_, 1289); }; +my $vncipher = sub { vcrypto_op(@_, 1352); }; +my $vncipherlast= sub { vcrypto_op(@_, 1353); }; +my $vsbox = sub { vcrypto_op(@_, 0, 1480); }; +my $vshasigmad = sub { my ($st,$six)=splice(@_,-2); vcrypto_op(@_, $st<<4|$six, 1730); }; +my $vshasigmaw = sub { my ($st,$six)=splice(@_,-2); vcrypto_op(@_, $st<<4|$six, 1666); }; +my $vpmsumb = sub { vcrypto_op(@_, 1032); }; +my $vpmsumd = sub { vcrypto_op(@_, 1224); }; +my $vpmsubh = sub { vcrypto_op(@_, 1096); }; +my $vpmsumw = sub { vcrypto_op(@_, 1160); }; +my $vaddudm = sub { vcrypto_op(@_, 192); }; + +my $mtsle = sub { + my ($f, $arg) = @_; + " .long ".sprintf "0x%X",(31<<26)|($arg<<21)|(147*2); +}; + +print "#include <asm/ppc_asm.h>\n" if $flavour =~ /linux/; + +while($line=<>) { + + $line =~ s|[#!;].*$||; # get rid of asm-style comments... + $line =~ s|/\*.*\*/||; # ... and C-style comments... + $line =~ s|^\s+||; # ... and skip white spaces in beginning... + $line =~ s|\s+$||; # ... and at the end + + { + $line =~ s|\b\.L(\w+)|L$1|g; # common denominator for Locallabel + $line =~ s|\bL(\w+)|\.L$1|g if ($dotinlocallabels); + } + + { + $line =~ s|^\s*(\.?)(\w+)([\.\+\-]?)\s*||; + my $c = $1; $c = "\t" if ($c eq ""); + my $mnemonic = $2; + my $f = $3; + my $opcode = eval("\$$mnemonic"); + $line =~ s/\b(c?[rf]|v|vs)([0-9]+)\b/$2/g if ($c ne "." and $flavour !~ /osx/); + if (ref($opcode) eq 'CODE') { $line = &$opcode($f,split(',',$line)); } + elsif ($mnemonic) { $line = $c.$mnemonic.$f."\t".$line; } + } + + print $line if ($line); + print "\n"; +} + +close STDOUT; diff --git a/drivers/crypto/vmx/vmx.c b/drivers/crypto/vmx/vmx.c new file mode 100644 index 000000000000..44d8d5cfe40d --- /dev/null +++ b/drivers/crypto/vmx/vmx.c @@ -0,0 +1,88 @@ +/** + * Routines supporting VMX instructions on the Power 8 + * + * Copyright (C) 2015 International Business Machines Inc. + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2 only. + * + * 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. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. + * + * Author: Marcelo Henrique Cerri <mhcerri@br.ibm.com> + */ + +#include <linux/module.h> +#include <linux/moduleparam.h> +#include <linux/types.h> +#include <linux/err.h> +#include <linux/crypto.h> +#include <asm/cputable.h> +#include <crypto/internal/hash.h> + +extern struct shash_alg p8_ghash_alg; +extern struct crypto_alg p8_aes_alg; +extern struct crypto_alg p8_aes_cbc_alg; +extern struct crypto_alg p8_aes_ctr_alg; +static struct crypto_alg *algs[] = { + &p8_aes_alg, + &p8_aes_cbc_alg, + &p8_aes_ctr_alg, + NULL, +}; + +int __init p8_init(void) +{ + int ret = 0; + struct crypto_alg **alg_it; + + if (!(cur_cpu_spec->cpu_user_features2 & PPC_FEATURE2_VEC_CRYPTO)) + return -ENODEV; + + for (alg_it = algs; *alg_it; alg_it++) { + ret = crypto_register_alg(*alg_it); + printk(KERN_INFO "crypto_register_alg '%s' = %d\n", + (*alg_it)->cra_name, ret); + if (ret) { + for (alg_it--; alg_it >= algs; alg_it--) + crypto_unregister_alg(*alg_it); + break; + } + } + if (ret) + return ret; + + ret = crypto_register_shash(&p8_ghash_alg); + if (ret) { + for (alg_it = algs; *alg_it; alg_it++) + crypto_unregister_alg(*alg_it); + } + return ret; +} + +void __exit p8_exit(void) +{ + struct crypto_alg **alg_it; + + for (alg_it = algs; *alg_it; alg_it++) { + printk(KERN_INFO "Removing '%s'\n", (*alg_it)->cra_name); + crypto_unregister_alg(*alg_it); + } + crypto_unregister_shash(&p8_ghash_alg); +} + +module_init(p8_init); +module_exit(p8_exit); + +MODULE_AUTHOR("Marcelo Cerri<mhcerri@br.ibm.com>"); +MODULE_DESCRIPTION("IBM VMX cryptogaphic acceleration instructions support on Power 8"); +MODULE_LICENSE("GPL"); +MODULE_VERSION("1.0.0"); + diff --git a/drivers/dma/of-dma.c b/drivers/dma/of-dma.c index ca31f1b45366..cbd4a8aff120 100644 --- a/drivers/dma/of-dma.c +++ b/drivers/dma/of-dma.c @@ -194,6 +194,7 @@ struct dma_chan *of_dma_request_slave_channel(struct device_node *np, return ERR_PTR(ret_no_channel); } +EXPORT_SYMBOL_GPL(of_dma_request_slave_channel); /** * of_dma_simple_xlate - Simple DMA engine translation function diff --git a/drivers/infiniband/hw/ipath/ipath_file_ops.c b/drivers/infiniband/hw/ipath/ipath_file_ops.c index aed8afee56da..450d15965005 100644 --- a/drivers/infiniband/hw/ipath/ipath_file_ops.c +++ b/drivers/infiniband/hw/ipath/ipath_file_ops.c @@ -42,6 +42,7 @@ #include <linux/io.h> #include <linux/jiffies.h> #include <linux/cpu.h> +#include <linux/uio.h> #include <asm/pgtable.h> #include "ipath_kernel.h" @@ -52,15 +53,19 @@ static int ipath_open(struct inode *, struct file *); static int ipath_close(struct inode *, struct file *); static ssize_t ipath_write(struct file *, const char __user *, size_t, loff_t *); -static ssize_t ipath_writev(struct kiocb *, const struct iovec *, - unsigned long , loff_t); +static ssize_t ipath_write_iter(struct kiocb *, struct iov_iter *from); static unsigned int ipath_poll(struct file *, struct poll_table_struct *); static int ipath_mmap(struct file *, struct vm_area_struct *); +/* + * This is really, really weird shit - write() and writev() here + * have completely unrelated semantics. Sucky userland ABI, + * film at 11. + */ static const struct file_operations ipath_file_ops = { .owner = THIS_MODULE, .write = ipath_write, - .aio_write = ipath_writev, + .write_iter = ipath_write_iter, .open = ipath_open, .release = ipath_close, .poll = ipath_poll, @@ -2413,18 +2418,17 @@ bail: return ret; } -static ssize_t ipath_writev(struct kiocb *iocb, const struct iovec *iov, - unsigned long dim, loff_t off) +static ssize_t ipath_write_iter(struct kiocb *iocb, struct iov_iter *from) { struct file *filp = iocb->ki_filp; struct ipath_filedata *fp = filp->private_data; struct ipath_portdata *pd = port_fp(filp); struct ipath_user_sdma_queue *pq = fp->pq; - if (!dim) + if (!iter_is_iovec(from) || !from->nr_segs) return -EINVAL; - return ipath_user_sdma_writev(pd->port_dd, pq, iov, dim); + return ipath_user_sdma_writev(pd->port_dd, pq, from->iov, from->nr_segs); } static struct class *ipath_class; diff --git a/drivers/infiniband/hw/qib/qib_file_ops.c b/drivers/infiniband/hw/qib/qib_file_ops.c index 14046f5a37fa..9ea6c440a00c 100644 --- a/drivers/infiniband/hw/qib/qib_file_ops.c +++ b/drivers/infiniband/hw/qib/qib_file_ops.c @@ -43,6 +43,7 @@ #include <asm/pgtable.h> #include <linux/delay.h> #include <linux/export.h> +#include <linux/uio.h> #include "qib.h" #include "qib_common.h" @@ -54,15 +55,19 @@ static int qib_open(struct inode *, struct file *); static int qib_close(struct inode *, struct file *); static ssize_t qib_write(struct file *, const char __user *, size_t, loff_t *); -static ssize_t qib_aio_write(struct kiocb *, const struct iovec *, - unsigned long, loff_t); +static ssize_t qib_write_iter(struct kiocb *, struct iov_iter *); static unsigned int qib_poll(struct file *, struct poll_table_struct *); static int qib_mmapf(struct file *, struct vm_area_struct *); +/* + * This is really, really weird shit - write() and writev() here + * have completely unrelated semantics. Sucky userland ABI, + * film at 11. + */ static const struct file_operations qib_file_ops = { .owner = THIS_MODULE, .write = qib_write, - .aio_write = qib_aio_write, + .write_iter = qib_write_iter, .open = qib_open, .release = qib_close, .poll = qib_poll, @@ -2248,17 +2253,16 @@ bail: return ret; } -static ssize_t qib_aio_write(struct kiocb *iocb, const struct iovec *iov, - unsigned long dim, loff_t off) +static ssize_t qib_write_iter(struct kiocb *iocb, struct iov_iter *from) { struct qib_filedata *fp = iocb->ki_filp->private_data; struct qib_ctxtdata *rcd = ctxt_fp(iocb->ki_filp); struct qib_user_sdma_queue *pq = fp->pq; - if (!dim || !pq) + if (!iter_is_iovec(from) || !from->nr_segs || !pq) return -EINVAL; - - return qib_user_sdma_writev(rcd, pq, iov, dim); + + return qib_user_sdma_writev(rcd, pq, from->iov, from->nr_segs); } static struct class *qib_class; diff --git a/drivers/leds/Kconfig b/drivers/leds/Kconfig index 25b320d64e26..966b9605f5f0 100644 --- a/drivers/leds/Kconfig +++ b/drivers/leds/Kconfig @@ -526,6 +526,14 @@ config LEDS_VERSATILE This option enabled support for the LEDs on the ARM Versatile and RealView boards. Say Y to enabled these. +config LEDS_PM8941_WLED + tristate "LED support for the Qualcomm PM8941 WLED block" + depends on LEDS_CLASS + select REGMAP + help + This option enables support for the 'White' LED block + on Qualcomm PM8941 PMICs. + comment "LED Triggers" source "drivers/leds/trigger/Kconfig" diff --git a/drivers/leds/Makefile b/drivers/leds/Makefile index cbba921b6f1c..bf4609338e10 100644 --- a/drivers/leds/Makefile +++ b/drivers/leds/Makefile @@ -58,6 +58,7 @@ obj-$(CONFIG_LEDS_BLINKM) += leds-blinkm.o obj-$(CONFIG_LEDS_SYSCON) += leds-syscon.o obj-$(CONFIG_LEDS_VERSATILE) += leds-versatile.o obj-$(CONFIG_LEDS_MENF21BMC) += leds-menf21bmc.o +obj-$(CONFIG_LEDS_PM8941_WLED) += leds-pm8941-wled.o # LED SPI Drivers obj-$(CONFIG_LEDS_DAC124S085) += leds-dac124s085.o diff --git a/drivers/leds/led-class-flash.c b/drivers/leds/led-class-flash.c index 4a19fd44f93f..3b2573411a37 100644 --- a/drivers/leds/led-class-flash.c +++ b/drivers/leds/led-class-flash.c @@ -216,75 +216,6 @@ static ssize_t flash_fault_show(struct device *dev, } static DEVICE_ATTR_RO(flash_fault); -static ssize_t available_sync_leds_show(struct device *dev, - struct device_attribute *attr, char *buf) -{ - struct led_classdev *led_cdev = dev_get_drvdata(dev); - struct led_classdev_flash *fled_cdev = lcdev_to_flcdev(led_cdev); - char *pbuf = buf; - int i, buf_len; - - buf_len = sprintf(pbuf, "[0: none] "); - pbuf += buf_len; - - for (i = 0; i < fled_cdev->num_sync_leds; ++i) { - buf_len = sprintf(pbuf, "[%d: %s] ", i + 1, - fled_cdev->sync_leds[i]->led_cdev.name); - pbuf += buf_len; - } - - return sprintf(buf, "%s\n", buf); -} -static DEVICE_ATTR_RO(available_sync_leds); - -static ssize_t flash_sync_strobe_store(struct device *dev, - struct device_attribute *attr, const char *buf, size_t size) -{ - struct led_classdev *led_cdev = dev_get_drvdata(dev); - struct led_classdev_flash *fled_cdev = lcdev_to_flcdev(led_cdev); - unsigned long led_id; - ssize_t ret; - - mutex_lock(&led_cdev->led_access); - - if (led_sysfs_is_disabled(led_cdev)) { - ret = -EBUSY; - goto unlock; - } - - ret = kstrtoul(buf, 10, &led_id); - if (ret) - goto unlock; - - if (led_id > fled_cdev->num_sync_leds) { - ret = -ERANGE; - goto unlock; - } - - fled_cdev->sync_led_id = led_id; - - ret = size; -unlock: - mutex_unlock(&led_cdev->led_access); - return ret; -} - -static ssize_t flash_sync_strobe_show(struct device *dev, - struct device_attribute *attr, char *buf) -{ - struct led_classdev *led_cdev = dev_get_drvdata(dev); - struct led_classdev_flash *fled_cdev = lcdev_to_flcdev(led_cdev); - int sled_id = fled_cdev->sync_led_id; - char *sync_led_name = "none"; - - if (fled_cdev->sync_led_id > 0) - sync_led_name = (char *) - fled_cdev->sync_leds[sled_id - 1]->led_cdev.name; - - return sprintf(buf, "[%d: %s]\n", sled_id, sync_led_name); -} -static DEVICE_ATTR_RW(flash_sync_strobe); - static struct attribute *led_flash_strobe_attrs[] = { &dev_attr_flash_strobe.attr, NULL, @@ -307,12 +238,6 @@ static struct attribute *led_flash_fault_attrs[] = { NULL, }; -static struct attribute *led_flash_sync_strobe_attrs[] = { - &dev_attr_available_sync_leds.attr, - &dev_attr_flash_sync_strobe.attr, - NULL, -}; - static const struct attribute_group led_flash_strobe_group = { .attrs = led_flash_strobe_attrs, }; @@ -329,10 +254,6 @@ static const struct attribute_group led_flash_fault_group = { .attrs = led_flash_fault_attrs, }; -static const struct attribute_group led_flash_sync_strobe_group = { - .attrs = led_flash_sync_strobe_attrs, -}; - static void led_flash_resume(struct led_classdev *led_cdev) { struct led_classdev_flash *fled_cdev = lcdev_to_flcdev(led_cdev); @@ -361,9 +282,6 @@ static void led_flash_init_sysfs_groups(struct led_classdev_flash *fled_cdev) if (ops->fault_get) flash_groups[num_sysfs_groups++] = &led_flash_fault_group; - if (led_cdev->flags & LED_DEV_CAP_SYNC_STROBE) - flash_groups[num_sysfs_groups++] = &led_flash_sync_strobe_group; - led_cdev->groups = flash_groups; } diff --git a/drivers/leds/led-class.c b/drivers/leds/led-class.c index 795ec994c663..728681debdbe 100644 --- a/drivers/leds/led-class.c +++ b/drivers/leds/led-class.c @@ -212,6 +212,31 @@ static const struct dev_pm_ops leds_class_dev_pm_ops = { .resume = led_resume, }; +static int match_name(struct device *dev, const void *data) +{ + if (!dev_name(dev)) + return 0; + return !strcmp(dev_name(dev), (char *)data); +} + +static int led_classdev_next_name(const char *init_name, char *name, + size_t len) +{ + unsigned int i = 0; + int ret = 0; + + strlcpy(name, init_name, len); + + while (class_find_device(leds_class, NULL, name, match_name) && + (ret < len)) + ret = snprintf(name, len, "%s_%u", init_name, ++i); + + if (ret >= len) + return -ENOMEM; + + return i; +} + /** * led_classdev_register - register a new object of led_classdev class. * @parent: The device to register. @@ -219,12 +244,22 @@ static const struct dev_pm_ops leds_class_dev_pm_ops = { */ int led_classdev_register(struct device *parent, struct led_classdev *led_cdev) { + char name[64]; + int ret; + + ret = led_classdev_next_name(led_cdev->name, name, sizeof(name)); + if (ret < 0) + return ret; + led_cdev->dev = device_create_with_groups(leds_class, parent, 0, - led_cdev, led_cdev->groups, - "%s", led_cdev->name); + led_cdev, led_cdev->groups, "%s", name); if (IS_ERR(led_cdev->dev)) return PTR_ERR(led_cdev->dev); + if (ret) + dev_warn(parent, "Led %s renamed to %s due to name collision", + led_cdev->name, dev_name(led_cdev->dev)); + #ifdef CONFIG_LEDS_TRIGGERS init_rwsem(&led_cdev->trigger_lock); #endif @@ -288,6 +323,63 @@ void led_classdev_unregister(struct led_classdev *led_cdev) } EXPORT_SYMBOL_GPL(led_classdev_unregister); +static void devm_led_classdev_release(struct device *dev, void *res) +{ + led_classdev_unregister(*(struct led_classdev **)res); +} + +/** + * devm_led_classdev_register - resource managed led_classdev_register() + * @parent: The device to register. + * @led_cdev: the led_classdev structure for this device. + */ +int devm_led_classdev_register(struct device *parent, + struct led_classdev *led_cdev) +{ + struct led_classdev **dr; + int rc; + + dr = devres_alloc(devm_led_classdev_release, sizeof(*dr), GFP_KERNEL); + if (!dr) + return -ENOMEM; + + rc = led_classdev_register(parent, led_cdev); + if (rc) { + devres_free(dr); + return rc; + } + + *dr = led_cdev; + devres_add(parent, dr); + + return 0; +} +EXPORT_SYMBOL_GPL(devm_led_classdev_register); + +static int devm_led_classdev_match(struct device *dev, void *res, void *data) +{ + struct led_cdev **p = res; + + if (WARN_ON(!p || !*p)) + return 0; + + return *p == data; +} + +/** + * devm_led_classdev_unregister() - resource managed led_classdev_unregister() + * @parent: The device to unregister. + * @led_cdev: the led_classdev structure for this device. + */ +void devm_led_classdev_unregister(struct device *dev, + struct led_classdev *led_cdev) +{ + WARN_ON(devres_release(dev, + devm_led_classdev_release, + devm_led_classdev_match, led_cdev)); +} +EXPORT_SYMBOL_GPL(devm_led_classdev_unregister); + static int __init leds_init(void) { leds_class = class_create(THIS_MODULE, "leds"); diff --git a/drivers/leds/leds-lp8501.c b/drivers/leds/leds-lp8501.c index 00f068b0fa6f..d3098e395fff 100644 --- a/drivers/leds/leds-lp8501.c +++ b/drivers/leds/leds-lp8501.c @@ -406,6 +406,6 @@ static struct i2c_driver lp8501_driver = { module_i2c_driver(lp8501_driver); -MODULE_DESCRIPTION("Texas Instruments LP8501 LED drvier"); +MODULE_DESCRIPTION("Texas Instruments LP8501 LED driver"); MODULE_AUTHOR("Milo Kim"); MODULE_LICENSE("GPL"); diff --git a/drivers/leds/leds-lp8860.c b/drivers/leds/leds-lp8860.c index 840e93f3ab3e..8c2b7fbe2392 100644 --- a/drivers/leds/leds-lp8860.c +++ b/drivers/leds/leds-lp8860.c @@ -391,11 +391,13 @@ static int lp8860_probe(struct i2c_client *client, } } - led->enable_gpio = devm_gpiod_get(&client->dev, "enable"); - if (IS_ERR(led->enable_gpio)) - led->enable_gpio = NULL; - else - gpiod_direction_output(led->enable_gpio, 0); + led->enable_gpio = devm_gpiod_get_optional(&client->dev, + "enable", GPIOD_OUT_LOW); + if (IS_ERR(led->enable_gpio)) { + ret = PTR_ERR(led->enable_gpio); + dev_err(&client->dev, "Failed to get enable gpio: %d\n", ret); + return ret; + } led->regulator = devm_regulator_get(&client->dev, "vled"); if (IS_ERR(led->regulator)) @@ -486,6 +488,6 @@ static struct i2c_driver lp8860_driver = { }; module_i2c_driver(lp8860_driver); -MODULE_DESCRIPTION("Texas Instruments LP8860 LED drvier"); +MODULE_DESCRIPTION("Texas Instruments LP8860 LED driver"); MODULE_AUTHOR("Dan Murphy <dmurphy@ti.com>"); MODULE_LICENSE("GPL"); diff --git a/drivers/leds/leds-pca963x.c b/drivers/leds/leds-pca963x.c index f110b4c456ba..bee3e1ab27fd 100644 --- a/drivers/leds/leds-pca963x.c +++ b/drivers/leds/leds-pca963x.c @@ -289,7 +289,7 @@ pca963x_dt_init(struct i2c_client *client, struct pca963x_chipdef *chip) return ERR_PTR(-ENOMEM); for_each_child_of_node(np, child) { - struct led_info led; + struct led_info led = {}; u32 reg; int res; diff --git a/drivers/leds/leds-pm8941-wled.c b/drivers/leds/leds-pm8941-wled.c new file mode 100644 index 000000000000..bf64a593fbf1 --- /dev/null +++ b/drivers/leds/leds-pm8941-wled.c @@ -0,0 +1,435 @@ +/* Copyright (c) 2015, Sony Mobile Communications, AB. + * + * 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 <linux/kernel.h> +#include <linux/leds.h> +#include <linux/module.h> +#include <linux/of.h> +#include <linux/of_device.h> +#include <linux/regmap.h> + +#define PM8941_WLED_REG_VAL_BASE 0x40 +#define PM8941_WLED_REG_VAL_MAX 0xFFF + +#define PM8941_WLED_REG_MOD_EN 0x46 +#define PM8941_WLED_REG_MOD_EN_BIT BIT(7) +#define PM8941_WLED_REG_MOD_EN_MASK BIT(7) + +#define PM8941_WLED_REG_SYNC 0x47 +#define PM8941_WLED_REG_SYNC_MASK 0x07 +#define PM8941_WLED_REG_SYNC_LED1 BIT(0) +#define PM8941_WLED_REG_SYNC_LED2 BIT(1) +#define PM8941_WLED_REG_SYNC_LED3 BIT(2) +#define PM8941_WLED_REG_SYNC_ALL 0x07 +#define PM8941_WLED_REG_SYNC_CLEAR 0x00 + +#define PM8941_WLED_REG_FREQ 0x4c +#define PM8941_WLED_REG_FREQ_MASK 0x0f + +#define PM8941_WLED_REG_OVP 0x4d +#define PM8941_WLED_REG_OVP_MASK 0x03 + +#define PM8941_WLED_REG_BOOST 0x4e +#define PM8941_WLED_REG_BOOST_MASK 0x07 + +#define PM8941_WLED_REG_SINK 0x4f +#define PM8941_WLED_REG_SINK_MASK 0xe0 +#define PM8941_WLED_REG_SINK_SHFT 0x05 + +/* Per-'string' registers below */ +#define PM8941_WLED_REG_STR_OFFSET 0x10 + +#define PM8941_WLED_REG_STR_MOD_EN_BASE 0x60 +#define PM8941_WLED_REG_STR_MOD_MASK BIT(7) +#define PM8941_WLED_REG_STR_MOD_EN BIT(7) + +#define PM8941_WLED_REG_STR_SCALE_BASE 0x62 +#define PM8941_WLED_REG_STR_SCALE_MASK 0x1f + +#define PM8941_WLED_REG_STR_MOD_SRC_BASE 0x63 +#define PM8941_WLED_REG_STR_MOD_SRC_MASK 0x01 +#define PM8941_WLED_REG_STR_MOD_SRC_INT 0x00 +#define PM8941_WLED_REG_STR_MOD_SRC_EXT 0x01 + +#define PM8941_WLED_REG_STR_CABC_BASE 0x66 +#define PM8941_WLED_REG_STR_CABC_MASK BIT(7) +#define PM8941_WLED_REG_STR_CABC_EN BIT(7) + +struct pm8941_wled_config { + u32 i_boost_limit; + u32 ovp; + u32 switch_freq; + u32 num_strings; + u32 i_limit; + bool cs_out_en; + bool ext_gen; + bool cabc_en; +}; + +struct pm8941_wled { + struct regmap *regmap; + u16 addr; + + struct led_classdev cdev; + + struct pm8941_wled_config cfg; +}; + +static int pm8941_wled_set(struct led_classdev *cdev, + enum led_brightness value) +{ + struct pm8941_wled *wled; + u8 ctrl = 0; + u16 val; + int rc; + int i; + + wled = container_of(cdev, struct pm8941_wled, cdev); + + if (value != 0) + ctrl = PM8941_WLED_REG_MOD_EN_BIT; + + val = value * PM8941_WLED_REG_VAL_MAX / LED_FULL; + + rc = regmap_update_bits(wled->regmap, + wled->addr + PM8941_WLED_REG_MOD_EN, + PM8941_WLED_REG_MOD_EN_MASK, ctrl); + if (rc) + return rc; + + for (i = 0; i < wled->cfg.num_strings; ++i) { + u8 v[2] = { val & 0xff, (val >> 8) & 0xf }; + + rc = regmap_bulk_write(wled->regmap, + wled->addr + PM8941_WLED_REG_VAL_BASE + 2 * i, + v, 2); + if (rc) + return rc; + } + + rc = regmap_update_bits(wled->regmap, + wled->addr + PM8941_WLED_REG_SYNC, + PM8941_WLED_REG_SYNC_MASK, PM8941_WLED_REG_SYNC_ALL); + if (rc) + return rc; + + rc = regmap_update_bits(wled->regmap, + wled->addr + PM8941_WLED_REG_SYNC, + PM8941_WLED_REG_SYNC_MASK, PM8941_WLED_REG_SYNC_CLEAR); + return rc; +} + +static void pm8941_wled_set_brightness(struct led_classdev *cdev, + enum led_brightness value) +{ + if (pm8941_wled_set(cdev, value)) { + dev_err(cdev->dev, "Unable to set brightness\n"); + return; + } + cdev->brightness = value; +} + +static int pm8941_wled_setup(struct pm8941_wled *wled) +{ + int rc; + int i; + + rc = regmap_update_bits(wled->regmap, + wled->addr + PM8941_WLED_REG_OVP, + PM8941_WLED_REG_OVP_MASK, wled->cfg.ovp); + if (rc) + return rc; + + rc = regmap_update_bits(wled->regmap, + wled->addr + PM8941_WLED_REG_BOOST, + PM8941_WLED_REG_BOOST_MASK, wled->cfg.i_boost_limit); + if (rc) + return rc; + + rc = regmap_update_bits(wled->regmap, + wled->addr + PM8941_WLED_REG_FREQ, + PM8941_WLED_REG_FREQ_MASK, wled->cfg.switch_freq); + if (rc) + return rc; + + if (wled->cfg.cs_out_en) { + u8 all = (BIT(wled->cfg.num_strings) - 1) + << PM8941_WLED_REG_SINK_SHFT; + + rc = regmap_update_bits(wled->regmap, + wled->addr + PM8941_WLED_REG_SINK, + PM8941_WLED_REG_SINK_MASK, all); + if (rc) + return rc; + } + + for (i = 0; i < wled->cfg.num_strings; ++i) { + u16 addr = wled->addr + PM8941_WLED_REG_STR_OFFSET * i; + + rc = regmap_update_bits(wled->regmap, + addr + PM8941_WLED_REG_STR_MOD_EN_BASE, + PM8941_WLED_REG_STR_MOD_MASK, + PM8941_WLED_REG_STR_MOD_EN); + if (rc) + return rc; + + if (wled->cfg.ext_gen) { + rc = regmap_update_bits(wled->regmap, + addr + PM8941_WLED_REG_STR_MOD_SRC_BASE, + PM8941_WLED_REG_STR_MOD_SRC_MASK, + PM8941_WLED_REG_STR_MOD_SRC_EXT); + if (rc) + return rc; + } + + rc = regmap_update_bits(wled->regmap, + addr + PM8941_WLED_REG_STR_SCALE_BASE, + PM8941_WLED_REG_STR_SCALE_MASK, + wled->cfg.i_limit); + if (rc) + return rc; + + rc = regmap_update_bits(wled->regmap, + addr + PM8941_WLED_REG_STR_CABC_BASE, + PM8941_WLED_REG_STR_CABC_MASK, + wled->cfg.cabc_en ? + PM8941_WLED_REG_STR_CABC_EN : 0); + if (rc) + return rc; + } + + return 0; +} + +static const struct pm8941_wled_config pm8941_wled_config_defaults = { + .i_boost_limit = 3, + .i_limit = 20, + .ovp = 2, + .switch_freq = 5, + .num_strings = 0, + .cs_out_en = false, + .ext_gen = false, + .cabc_en = false, +}; + +struct pm8941_wled_var_cfg { + const u32 *values; + u32 (*fn)(u32); + int size; +}; + +static const u32 pm8941_wled_i_boost_limit_values[] = { + 105, 385, 525, 805, 980, 1260, 1400, 1680, +}; + +static const struct pm8941_wled_var_cfg pm8941_wled_i_boost_limit_cfg = { + .values = pm8941_wled_i_boost_limit_values, + .size = ARRAY_SIZE(pm8941_wled_i_boost_limit_values), +}; + +static const u32 pm8941_wled_ovp_values[] = { + 35, 32, 29, 27, +}; + +static const struct pm8941_wled_var_cfg pm8941_wled_ovp_cfg = { + .values = pm8941_wled_ovp_values, + .size = ARRAY_SIZE(pm8941_wled_ovp_values), +}; + +static u32 pm8941_wled_num_strings_values_fn(u32 idx) +{ + return idx + 1; +} + +static const struct pm8941_wled_var_cfg pm8941_wled_num_strings_cfg = { + .fn = pm8941_wled_num_strings_values_fn, + .size = 3, +}; + +static u32 pm8941_wled_switch_freq_values_fn(u32 idx) +{ + return 19200 / (2 * (1 + idx)); +} + +static const struct pm8941_wled_var_cfg pm8941_wled_switch_freq_cfg = { + .fn = pm8941_wled_switch_freq_values_fn, + .size = 16, +}; + +static const struct pm8941_wled_var_cfg pm8941_wled_i_limit_cfg = { + .size = 26, +}; + +static u32 pm8941_wled_values(const struct pm8941_wled_var_cfg *cfg, u32 idx) +{ + if (idx >= cfg->size) + return UINT_MAX; + if (cfg->fn) + return cfg->fn(idx); + if (cfg->values) + return cfg->values[idx]; + return idx; +} + +static int pm8941_wled_configure(struct pm8941_wled *wled, struct device *dev) +{ + struct pm8941_wled_config *cfg = &wled->cfg; + u32 val; + int rc; + u32 c; + int i; + int j; + + const struct { + const char *name; + u32 *val_ptr; + const struct pm8941_wled_var_cfg *cfg; + } u32_opts[] = { + { + "qcom,current-boost-limit", + &cfg->i_boost_limit, + .cfg = &pm8941_wled_i_boost_limit_cfg, + }, + { + "qcom,current-limit", + &cfg->i_limit, + .cfg = &pm8941_wled_i_limit_cfg, + }, + { + "qcom,ovp", + &cfg->ovp, + .cfg = &pm8941_wled_ovp_cfg, + }, + { + "qcom,switching-freq", + &cfg->switch_freq, + .cfg = &pm8941_wled_switch_freq_cfg, + }, + { + "qcom,num-strings", + &cfg->num_strings, + .cfg = &pm8941_wled_num_strings_cfg, + }, + }; + const struct { + const char *name; + bool *val_ptr; + } bool_opts[] = { + { "qcom,cs-out", &cfg->cs_out_en, }, + { "qcom,ext-gen", &cfg->ext_gen, }, + { "qcom,cabc", &cfg->cabc_en, }, + }; + + rc = of_property_read_u32(dev->of_node, "reg", &val); + if (rc || val > 0xffff) { + dev_err(dev, "invalid IO resources\n"); + return rc ? rc : -EINVAL; + } + wled->addr = val; + + rc = of_property_read_string(dev->of_node, "label", &wled->cdev.name); + if (rc) + wled->cdev.name = dev->of_node->name; + + wled->cdev.default_trigger = of_get_property(dev->of_node, + "linux,default-trigger", NULL); + + *cfg = pm8941_wled_config_defaults; + for (i = 0; i < ARRAY_SIZE(u32_opts); ++i) { + rc = of_property_read_u32(dev->of_node, u32_opts[i].name, &val); + if (rc == -EINVAL) { + continue; + } else if (rc) { + dev_err(dev, "error reading '%s'\n", u32_opts[i].name); + return rc; + } + + c = UINT_MAX; + for (j = 0; c != val; j++) { + c = pm8941_wled_values(u32_opts[i].cfg, j); + if (c == UINT_MAX) { + dev_err(dev, "invalid value for '%s'\n", + u32_opts[i].name); + return -EINVAL; + } + } + + dev_dbg(dev, "'%s' = %u\n", u32_opts[i].name, c); + *u32_opts[i].val_ptr = j; + } + + for (i = 0; i < ARRAY_SIZE(bool_opts); ++i) { + if (of_property_read_bool(dev->of_node, bool_opts[i].name)) + *bool_opts[i].val_ptr = true; + } + + cfg->num_strings = cfg->num_strings + 1; + + return 0; +} + +static int pm8941_wled_probe(struct platform_device *pdev) +{ + struct pm8941_wled *wled; + struct regmap *regmap; + int rc; + + regmap = dev_get_regmap(pdev->dev.parent, NULL); + if (!regmap) { + dev_err(&pdev->dev, "Unable to get regmap\n"); + return -EINVAL; + } + + wled = devm_kzalloc(&pdev->dev, sizeof(*wled), GFP_KERNEL); + if (!wled) + return -ENOMEM; + + wled->regmap = regmap; + + rc = pm8941_wled_configure(wled, &pdev->dev); + if (rc) + return rc; + + rc = pm8941_wled_setup(wled); + if (rc) + return rc; + + wled->cdev.brightness_set = pm8941_wled_set_brightness; + + rc = devm_led_classdev_register(&pdev->dev, &wled->cdev); + if (rc) + return rc; + + platform_set_drvdata(pdev, wled); + + return 0; +}; + +static const struct of_device_id pm8941_wled_match_table[] = { + { .compatible = "qcom,pm8941-wled" }, + {} +}; +MODULE_DEVICE_TABLE(of, pm8941_wled_match_table); + +static struct platform_driver pm8941_wled_driver = { + .probe = pm8941_wled_probe, + .driver = { + .name = "pm8941-wled", + .of_match_table = pm8941_wled_match_table, + }, +}; + +module_platform_driver(pm8941_wled_driver); + +MODULE_DESCRIPTION("pm8941 wled driver"); +MODULE_LICENSE("GPL v2"); +MODULE_ALIAS("platform:pm8941-wled"); diff --git a/drivers/leds/leds-pwm.c b/drivers/leds/leds-pwm.c index f668500a2157..1d07e3e83d29 100644 --- a/drivers/leds/leds-pwm.c +++ b/drivers/leds/leds-pwm.c @@ -121,9 +121,6 @@ static int led_pwm_add(struct device *dev, struct led_pwm_priv *priv, return ret; } - if (child) - led_data->period = pwm_get_period(led_data->pwm); - led_data->can_sleep = pwm_can_sleep(led_data->pwm); if (led_data->can_sleep) INIT_WORK(&led_data->work, led_pwm_work); diff --git a/drivers/mtd/ubi/attach.c b/drivers/mtd/ubi/attach.c index 9d2e16f3150a..68eea5befaf1 100644 --- a/drivers/mtd/ubi/attach.c +++ b/drivers/mtd/ubi/attach.c @@ -410,7 +410,7 @@ int ubi_compare_lebs(struct ubi_device *ubi, const struct ubi_ainf_peb *aeb, second_is_newer = !second_is_newer; } else { dbg_bld("PEB %d CRC is OK", pnum); - bitflips = !!err; + bitflips |= !!err; } mutex_unlock(&ubi->buf_mutex); @@ -1301,6 +1301,30 @@ out_ech: return err; } +static struct ubi_attach_info *alloc_ai(void) +{ + struct ubi_attach_info *ai; + + ai = kzalloc(sizeof(struct ubi_attach_info), GFP_KERNEL); + if (!ai) + return ai; + + INIT_LIST_HEAD(&ai->corr); + INIT_LIST_HEAD(&ai->free); + INIT_LIST_HEAD(&ai->erase); + INIT_LIST_HEAD(&ai->alien); + ai->volumes = RB_ROOT; + ai->aeb_slab_cache = kmem_cache_create("ubi_aeb_slab_cache", + sizeof(struct ubi_ainf_peb), + 0, 0, NULL); + if (!ai->aeb_slab_cache) { + kfree(ai); + ai = NULL; + } + + return ai; +} + #ifdef CONFIG_MTD_UBI_FASTMAP /** @@ -1313,7 +1337,7 @@ out_ech: * UBI_NO_FASTMAP denotes that no fastmap was found. * UBI_BAD_FASTMAP denotes that the found fastmap was invalid. */ -static int scan_fast(struct ubi_device *ubi, struct ubi_attach_info *ai) +static int scan_fast(struct ubi_device *ubi, struct ubi_attach_info **ai) { int err, pnum, fm_anchor = -1; unsigned long long max_sqnum = 0; @@ -1334,7 +1358,7 @@ static int scan_fast(struct ubi_device *ubi, struct ubi_attach_info *ai) cond_resched(); dbg_gen("process PEB %d", pnum); - err = scan_peb(ubi, ai, pnum, &vol_id, &sqnum); + err = scan_peb(ubi, *ai, pnum, &vol_id, &sqnum); if (err < 0) goto out_vidh; @@ -1350,7 +1374,12 @@ static int scan_fast(struct ubi_device *ubi, struct ubi_attach_info *ai) if (fm_anchor < 0) return UBI_NO_FASTMAP; - return ubi_scan_fastmap(ubi, ai, fm_anchor); + destroy_ai(*ai); + *ai = alloc_ai(); + if (!*ai) + return -ENOMEM; + + return ubi_scan_fastmap(ubi, *ai, fm_anchor); out_vidh: ubi_free_vid_hdr(ubi, vidh); @@ -1362,30 +1391,6 @@ out: #endif -static struct ubi_attach_info *alloc_ai(const char *slab_name) -{ - struct ubi_attach_info *ai; - - ai = kzalloc(sizeof(struct ubi_attach_info), GFP_KERNEL); - if (!ai) - return ai; - - INIT_LIST_HEAD(&ai->corr); - INIT_LIST_HEAD(&ai->free); - INIT_LIST_HEAD(&ai->erase); - INIT_LIST_HEAD(&ai->alien); - ai->volumes = RB_ROOT; - ai->aeb_slab_cache = kmem_cache_create(slab_name, - sizeof(struct ubi_ainf_peb), - 0, 0, NULL); - if (!ai->aeb_slab_cache) { - kfree(ai); - ai = NULL; - } - - return ai; -} - /** * ubi_attach - attach an MTD device. * @ubi: UBI device descriptor @@ -1399,7 +1404,7 @@ int ubi_attach(struct ubi_device *ubi, int force_scan) int err; struct ubi_attach_info *ai; - ai = alloc_ai("ubi_aeb_slab_cache"); + ai = alloc_ai(); if (!ai) return -ENOMEM; @@ -1413,11 +1418,11 @@ int ubi_attach(struct ubi_device *ubi, int force_scan) if (force_scan) err = scan_all(ubi, ai, 0); else { - err = scan_fast(ubi, ai); - if (err > 0) { + err = scan_fast(ubi, &ai); + if (err > 0 || mtd_is_eccerr(err)) { if (err != UBI_NO_FASTMAP) { destroy_ai(ai); - ai = alloc_ai("ubi_aeb_slab_cache2"); + ai = alloc_ai(); if (!ai) return -ENOMEM; @@ -1453,10 +1458,10 @@ int ubi_attach(struct ubi_device *ubi, int force_scan) goto out_wl; #ifdef CONFIG_MTD_UBI_FASTMAP - if (ubi->fm && ubi_dbg_chk_gen(ubi)) { + if (ubi->fm && ubi_dbg_chk_fastmap(ubi)) { struct ubi_attach_info *scan_ai; - scan_ai = alloc_ai("ubi_ckh_aeb_slab_cache"); + scan_ai = alloc_ai(); if (!scan_ai) { err = -ENOMEM; goto out_wl; diff --git a/drivers/mtd/ubi/build.c b/drivers/mtd/ubi/build.c index ba01a8d22d28..9690cf9aaef5 100644 --- a/drivers/mtd/ubi/build.c +++ b/drivers/mtd/ubi/build.c @@ -81,6 +81,7 @@ static struct mtd_dev_param __initdata mtd_dev_param[UBI_MAX_DEVICES]; #ifdef CONFIG_MTD_UBI_FASTMAP /* UBI module parameter to enable fastmap automatically on non-fastmap images */ static bool fm_autoconvert; +static bool fm_debug; #endif /* Root UBI "class" object (corresponds to '/<sysfs>/class/ubi/') */ struct class *ubi_class; @@ -154,23 +155,22 @@ static struct device_attribute dev_mtd_num = */ int ubi_volume_notify(struct ubi_device *ubi, struct ubi_volume *vol, int ntype) { + int ret; struct ubi_notification nt; ubi_do_get_device_info(ubi, &nt.di); ubi_do_get_volume_info(ubi, vol, &nt.vi); -#ifdef CONFIG_MTD_UBI_FASTMAP switch (ntype) { case UBI_VOLUME_ADDED: case UBI_VOLUME_REMOVED: case UBI_VOLUME_RESIZED: case UBI_VOLUME_RENAMED: - if (ubi_update_fastmap(ubi)) { - ubi_err(ubi, "Unable to update fastmap!"); - ubi_ro_mode(ubi); - } + ret = ubi_update_fastmap(ubi); + if (ret) + ubi_msg(ubi, "Unable to write a new fastmap: %i", ret); } -#endif + return blocking_notifier_call_chain(&ubi_notifiers, ntype, &nt); } @@ -950,8 +950,10 @@ int ubi_attach_mtd_dev(struct mtd_info *mtd, int ubi_num, if (ubi->fm_pool.max_size < UBI_FM_MIN_POOL_SIZE) ubi->fm_pool.max_size = UBI_FM_MIN_POOL_SIZE; - ubi->fm_wl_pool.max_size = UBI_FM_WL_POOL_SIZE; + ubi->fm_wl_pool.max_size = ubi->fm_pool.max_size / 2; ubi->fm_disabled = !fm_autoconvert; + if (fm_debug) + ubi_enable_dbg_chk_fastmap(ubi); if (!ubi->fm_disabled && (int)mtd_div_by_eb(ubi->mtd->size, ubi->mtd) <= UBI_FM_MAX_START) { @@ -970,8 +972,8 @@ int ubi_attach_mtd_dev(struct mtd_info *mtd, int ubi_num, mutex_init(&ubi->ckvol_mutex); mutex_init(&ubi->device_mutex); spin_lock_init(&ubi->volumes_lock); - mutex_init(&ubi->fm_mutex); - init_rwsem(&ubi->fm_sem); + init_rwsem(&ubi->fm_protect); + init_rwsem(&ubi->fm_eba_sem); ubi_msg(ubi, "attaching mtd%d", mtd->index); @@ -1115,8 +1117,11 @@ int ubi_detach_mtd_dev(int ubi_num, int anyway) ubi_msg(ubi, "detaching mtd%d", ubi->mtd->index); #ifdef CONFIG_MTD_UBI_FASTMAP /* If we don't write a new fastmap at detach time we lose all - * EC updates that have been made since the last written fastmap. */ - ubi_update_fastmap(ubi); + * EC updates that have been made since the last written fastmap. + * In case of fastmap debugging we omit the update to simulate an + * unclean shutdown. */ + if (!ubi_dbg_chk_fastmap(ubi)) + ubi_update_fastmap(ubi); #endif /* * Before freeing anything, we have to stop the background thread to @@ -1501,6 +1506,8 @@ MODULE_PARM_DESC(mtd, "MTD devices to attach. Parameter format: mtd=<name|num|pa #ifdef CONFIG_MTD_UBI_FASTMAP module_param(fm_autoconvert, bool, 0644); MODULE_PARM_DESC(fm_autoconvert, "Set this parameter to enable fastmap automatically on images without a fastmap."); +module_param(fm_debug, bool, 0); +MODULE_PARM_DESC(fm_debug, "Set this parameter to enable fastmap debugging by default. Warning, this will make fastmap slow!"); #endif MODULE_VERSION(__stringify(UBI_VERSION)); MODULE_DESCRIPTION("UBI - Unsorted Block Images"); diff --git a/drivers/mtd/ubi/cdev.c b/drivers/mtd/ubi/cdev.c index d647e504f9b1..d16fccf79179 100644 --- a/drivers/mtd/ubi/cdev.c +++ b/drivers/mtd/ubi/cdev.c @@ -455,7 +455,7 @@ static long vol_cdev_ioctl(struct file *file, unsigned int cmd, /* Validate the request */ err = -EINVAL; if (req.lnum < 0 || req.lnum >= vol->reserved_pebs || - req.bytes < 0 || req.lnum >= vol->usable_leb_size) + req.bytes < 0 || req.bytes > vol->usable_leb_size) break; err = get_exclusive(desc); diff --git a/drivers/mtd/ubi/debug.c b/drivers/mtd/ubi/debug.c index 7335c9ff9d99..b077e43b5ba9 100644 --- a/drivers/mtd/ubi/debug.c +++ b/drivers/mtd/ubi/debug.c @@ -263,7 +263,7 @@ static ssize_t dfs_file_read(struct file *file, char __user *user_buf, struct dentry *dent = file->f_path.dentry; struct ubi_device *ubi; struct ubi_debug_info *d; - char buf[3]; + char buf[8]; int val; ubi = ubi_get_device(ubi_num); @@ -275,12 +275,30 @@ static ssize_t dfs_file_read(struct file *file, char __user *user_buf, val = d->chk_gen; else if (dent == d->dfs_chk_io) val = d->chk_io; + else if (dent == d->dfs_chk_fastmap) + val = d->chk_fastmap; else if (dent == d->dfs_disable_bgt) val = d->disable_bgt; else if (dent == d->dfs_emulate_bitflips) val = d->emulate_bitflips; else if (dent == d->dfs_emulate_io_failures) val = d->emulate_io_failures; + else if (dent == d->dfs_emulate_power_cut) { + snprintf(buf, sizeof(buf), "%u\n", d->emulate_power_cut); + count = simple_read_from_buffer(user_buf, count, ppos, + buf, strlen(buf)); + goto out; + } else if (dent == d->dfs_power_cut_min) { + snprintf(buf, sizeof(buf), "%u\n", d->power_cut_min); + count = simple_read_from_buffer(user_buf, count, ppos, + buf, strlen(buf)); + goto out; + } else if (dent == d->dfs_power_cut_max) { + snprintf(buf, sizeof(buf), "%u\n", d->power_cut_max); + count = simple_read_from_buffer(user_buf, count, ppos, + buf, strlen(buf)); + goto out; + } else { count = -EINVAL; goto out; @@ -309,7 +327,7 @@ static ssize_t dfs_file_write(struct file *file, const char __user *user_buf, struct ubi_device *ubi; struct ubi_debug_info *d; size_t buf_size; - char buf[8]; + char buf[8] = {0}; int val; ubi = ubi_get_device(ubi_num); @@ -323,6 +341,21 @@ static ssize_t dfs_file_write(struct file *file, const char __user *user_buf, goto out; } + if (dent == d->dfs_power_cut_min) { + if (kstrtouint(buf, 0, &d->power_cut_min) != 0) + count = -EINVAL; + goto out; + } else if (dent == d->dfs_power_cut_max) { + if (kstrtouint(buf, 0, &d->power_cut_max) != 0) + count = -EINVAL; + goto out; + } else if (dent == d->dfs_emulate_power_cut) { + if (kstrtoint(buf, 0, &val) != 0) + count = -EINVAL; + d->emulate_power_cut = val; + goto out; + } + if (buf[0] == '1') val = 1; else if (buf[0] == '0') @@ -336,6 +369,8 @@ static ssize_t dfs_file_write(struct file *file, const char __user *user_buf, d->chk_gen = val; else if (dent == d->dfs_chk_io) d->chk_io = val; + else if (dent == d->dfs_chk_fastmap) + d->chk_fastmap = val; else if (dent == d->dfs_disable_bgt) d->disable_bgt = val; else if (dent == d->dfs_emulate_bitflips) @@ -406,6 +441,13 @@ int ubi_debugfs_init_dev(struct ubi_device *ubi) goto out_remove; d->dfs_chk_io = dent; + fname = "chk_fastmap"; + dent = debugfs_create_file(fname, S_IWUSR, d->dfs_dir, (void *)ubi_num, + &dfs_fops); + if (IS_ERR_OR_NULL(dent)) + goto out_remove; + d->dfs_chk_fastmap = dent; + fname = "tst_disable_bgt"; dent = debugfs_create_file(fname, S_IWUSR, d->dfs_dir, (void *)ubi_num, &dfs_fops); @@ -427,6 +469,27 @@ int ubi_debugfs_init_dev(struct ubi_device *ubi) goto out_remove; d->dfs_emulate_io_failures = dent; + fname = "tst_emulate_power_cut"; + dent = debugfs_create_file(fname, S_IWUSR, d->dfs_dir, (void *)ubi_num, + &dfs_fops); + if (IS_ERR_OR_NULL(dent)) + goto out_remove; + d->dfs_emulate_power_cut = dent; + + fname = "tst_emulate_power_cut_min"; + dent = debugfs_create_file(fname, S_IWUSR, d->dfs_dir, (void *)ubi_num, + &dfs_fops); + if (IS_ERR_OR_NULL(dent)) + goto out_remove; + d->dfs_power_cut_min = dent; + + fname = "tst_emulate_power_cut_max"; + dent = debugfs_create_file(fname, S_IWUSR, d->dfs_dir, (void *)ubi_num, + &dfs_fops); + if (IS_ERR_OR_NULL(dent)) + goto out_remove; + d->dfs_power_cut_max = dent; + return 0; out_remove: @@ -447,3 +510,36 @@ void ubi_debugfs_exit_dev(struct ubi_device *ubi) if (IS_ENABLED(CONFIG_DEBUG_FS)) debugfs_remove_recursive(ubi->dbg.dfs_dir); } + +/** + * ubi_dbg_power_cut - emulate a power cut if it is time to do so + * @ubi: UBI device description object + * @caller: Flags set to indicate from where the function is being called + * + * Returns non-zero if a power cut was emulated, zero if not. + */ +int ubi_dbg_power_cut(struct ubi_device *ubi, int caller) +{ + unsigned int range; + + if ((ubi->dbg.emulate_power_cut & caller) == 0) + return 0; + + if (ubi->dbg.power_cut_counter == 0) { + ubi->dbg.power_cut_counter = ubi->dbg.power_cut_min; + + if (ubi->dbg.power_cut_max > ubi->dbg.power_cut_min) { + range = ubi->dbg.power_cut_max - ubi->dbg.power_cut_min; + ubi->dbg.power_cut_counter += prandom_u32() % range; + } + return 0; + } + + ubi->dbg.power_cut_counter--; + if (ubi->dbg.power_cut_counter) + return 0; + + ubi_msg(ubi, "XXXXXXXXXXXXXXX emulating a power cut XXXXXXXXXXXXXXXX"); + ubi_ro_mode(ubi); + return 1; +} diff --git a/drivers/mtd/ubi/debug.h b/drivers/mtd/ubi/debug.h index cba89fcd1587..eb8985e5c178 100644 --- a/drivers/mtd/ubi/debug.h +++ b/drivers/mtd/ubi/debug.h @@ -127,4 +127,16 @@ static inline int ubi_dbg_chk_gen(const struct ubi_device *ubi) { return ubi->dbg.chk_gen; } + +static inline int ubi_dbg_chk_fastmap(const struct ubi_device *ubi) +{ + return ubi->dbg.chk_fastmap; +} + +static inline void ubi_enable_dbg_chk_fastmap(struct ubi_device *ubi) +{ + ubi->dbg.chk_fastmap = 1; +} + +int ubi_dbg_power_cut(struct ubi_device *ubi, int caller); #endif /* !__UBI_DEBUG_H__ */ diff --git a/drivers/mtd/ubi/eba.c b/drivers/mtd/ubi/eba.c index 16e34b37d134..51bca035cd83 100644 --- a/drivers/mtd/ubi/eba.c +++ b/drivers/mtd/ubi/eba.c @@ -340,9 +340,9 @@ int ubi_eba_unmap_leb(struct ubi_device *ubi, struct ubi_volume *vol, dbg_eba("erase LEB %d:%d, PEB %d", vol_id, lnum, pnum); - down_read(&ubi->fm_sem); + down_read(&ubi->fm_eba_sem); vol->eba_tbl[lnum] = UBI_LEB_UNMAPPED; - up_read(&ubi->fm_sem); + up_read(&ubi->fm_eba_sem); err = ubi_wl_put_peb(ubi, vol_id, lnum, pnum, 0); out_unlock: @@ -567,6 +567,7 @@ retry: new_pnum = ubi_wl_get_peb(ubi); if (new_pnum < 0) { ubi_free_vid_hdr(ubi, vid_hdr); + up_read(&ubi->fm_eba_sem); return new_pnum; } @@ -577,13 +578,16 @@ retry: if (err && err != UBI_IO_BITFLIPS) { if (err > 0) err = -EIO; + up_read(&ubi->fm_eba_sem); goto out_put; } vid_hdr->sqnum = cpu_to_be64(ubi_next_sqnum(ubi)); err = ubi_io_write_vid_hdr(ubi, new_pnum, vid_hdr); - if (err) + if (err) { + up_read(&ubi->fm_eba_sem); goto write_error; + } data_size = offset + len; mutex_lock(&ubi->buf_mutex); @@ -592,8 +596,10 @@ retry: /* Read everything before the area where the write failure happened */ if (offset > 0) { err = ubi_io_read_data(ubi, ubi->peb_buf, pnum, 0, offset); - if (err && err != UBI_IO_BITFLIPS) + if (err && err != UBI_IO_BITFLIPS) { + up_read(&ubi->fm_eba_sem); goto out_unlock; + } } memcpy(ubi->peb_buf + offset, buf, len); @@ -601,15 +607,15 @@ retry: err = ubi_io_write_data(ubi, ubi->peb_buf, new_pnum, 0, data_size); if (err) { mutex_unlock(&ubi->buf_mutex); + up_read(&ubi->fm_eba_sem); goto write_error; } mutex_unlock(&ubi->buf_mutex); ubi_free_vid_hdr(ubi, vid_hdr); - down_read(&ubi->fm_sem); vol->eba_tbl[lnum] = new_pnum; - up_read(&ubi->fm_sem); + up_read(&ubi->fm_eba_sem); ubi_wl_put_peb(ubi, vol_id, lnum, pnum, 1); ubi_msg(ubi, "data was successfully recovered"); @@ -704,6 +710,7 @@ retry: if (pnum < 0) { ubi_free_vid_hdr(ubi, vid_hdr); leb_write_unlock(ubi, vol_id, lnum); + up_read(&ubi->fm_eba_sem); return pnum; } @@ -714,6 +721,7 @@ retry: if (err) { ubi_warn(ubi, "failed to write VID header to LEB %d:%d, PEB %d", vol_id, lnum, pnum); + up_read(&ubi->fm_eba_sem); goto write_error; } @@ -722,13 +730,13 @@ retry: if (err) { ubi_warn(ubi, "failed to write %d bytes at offset %d of LEB %d:%d, PEB %d", len, offset, vol_id, lnum, pnum); + up_read(&ubi->fm_eba_sem); goto write_error; } } - down_read(&ubi->fm_sem); vol->eba_tbl[lnum] = pnum; - up_read(&ubi->fm_sem); + up_read(&ubi->fm_eba_sem); leb_write_unlock(ubi, vol_id, lnum); ubi_free_vid_hdr(ubi, vid_hdr); @@ -825,6 +833,7 @@ retry: if (pnum < 0) { ubi_free_vid_hdr(ubi, vid_hdr); leb_write_unlock(ubi, vol_id, lnum); + up_read(&ubi->fm_eba_sem); return pnum; } @@ -835,6 +844,7 @@ retry: if (err) { ubi_warn(ubi, "failed to write VID header to LEB %d:%d, PEB %d", vol_id, lnum, pnum); + up_read(&ubi->fm_eba_sem); goto write_error; } @@ -842,13 +852,13 @@ retry: if (err) { ubi_warn(ubi, "failed to write %d bytes of data to PEB %d", len, pnum); + up_read(&ubi->fm_eba_sem); goto write_error; } ubi_assert(vol->eba_tbl[lnum] < 0); - down_read(&ubi->fm_sem); vol->eba_tbl[lnum] = pnum; - up_read(&ubi->fm_sem); + up_read(&ubi->fm_eba_sem); leb_write_unlock(ubi, vol_id, lnum); ubi_free_vid_hdr(ubi, vid_hdr); @@ -900,7 +910,7 @@ write_error: int ubi_eba_atomic_leb_change(struct ubi_device *ubi, struct ubi_volume *vol, int lnum, const void *buf, int len) { - int err, pnum, tries = 0, vol_id = vol->vol_id; + int err, pnum, old_pnum, tries = 0, vol_id = vol->vol_id; struct ubi_vid_hdr *vid_hdr; uint32_t crc; @@ -943,6 +953,7 @@ retry: pnum = ubi_wl_get_peb(ubi); if (pnum < 0) { err = pnum; + up_read(&ubi->fm_eba_sem); goto out_leb_unlock; } @@ -953,6 +964,7 @@ retry: if (err) { ubi_warn(ubi, "failed to write VID header to LEB %d:%d, PEB %d", vol_id, lnum, pnum); + up_read(&ubi->fm_eba_sem); goto write_error; } @@ -960,19 +972,20 @@ retry: if (err) { ubi_warn(ubi, "failed to write %d bytes of data to PEB %d", len, pnum); + up_read(&ubi->fm_eba_sem); goto write_error; } - if (vol->eba_tbl[lnum] >= 0) { - err = ubi_wl_put_peb(ubi, vol_id, lnum, vol->eba_tbl[lnum], 0); + old_pnum = vol->eba_tbl[lnum]; + vol->eba_tbl[lnum] = pnum; + up_read(&ubi->fm_eba_sem); + + if (old_pnum >= 0) { + err = ubi_wl_put_peb(ubi, vol_id, lnum, old_pnum, 0); if (err) goto out_leb_unlock; } - down_read(&ubi->fm_sem); - vol->eba_tbl[lnum] = pnum; - up_read(&ubi->fm_sem); - out_leb_unlock: leb_write_unlock(ubi, vol_id, lnum); out_mutex: @@ -1218,9 +1231,9 @@ int ubi_eba_copy_leb(struct ubi_device *ubi, int from, int to, } ubi_assert(vol->eba_tbl[lnum] == from); - down_read(&ubi->fm_sem); + down_read(&ubi->fm_eba_sem); vol->eba_tbl[lnum] = to; - up_read(&ubi->fm_sem); + up_read(&ubi->fm_eba_sem); out_unlock_buf: mutex_unlock(&ubi->buf_mutex); @@ -1419,7 +1432,8 @@ int ubi_eba_init(struct ubi_device *ubi, struct ubi_attach_info *ai) * during re-size. */ ubi_move_aeb_to_list(av, aeb, &ai->erase); - vol->eba_tbl[aeb->lnum] = aeb->pnum; + else + vol->eba_tbl[aeb->lnum] = aeb->pnum; } } diff --git a/drivers/mtd/ubi/fastmap-wl.c b/drivers/mtd/ubi/fastmap-wl.c new file mode 100644 index 000000000000..b2a665398bca --- /dev/null +++ b/drivers/mtd/ubi/fastmap-wl.c @@ -0,0 +1,362 @@ +/* + * Copyright (c) 2012 Linutronix GmbH + * Copyright (c) 2014 sigma star gmbh + * Author: Richard Weinberger <richard@nod.at> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; version 2. + * + * 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. + * + */ + +/** + * update_fastmap_work_fn - calls ubi_update_fastmap from a work queue + * @wrk: the work description object + */ +static void update_fastmap_work_fn(struct work_struct *wrk) +{ + struct ubi_device *ubi = container_of(wrk, struct ubi_device, fm_work); + + ubi_update_fastmap(ubi); + spin_lock(&ubi->wl_lock); + ubi->fm_work_scheduled = 0; + spin_unlock(&ubi->wl_lock); +} + +/** + * find_anchor_wl_entry - find wear-leveling entry to used as anchor PEB. + * @root: the RB-tree where to look for + */ +static struct ubi_wl_entry *find_anchor_wl_entry(struct rb_root *root) +{ + struct rb_node *p; + struct ubi_wl_entry *e, *victim = NULL; + int max_ec = UBI_MAX_ERASECOUNTER; + + ubi_rb_for_each_entry(p, e, root, u.rb) { + if (e->pnum < UBI_FM_MAX_START && e->ec < max_ec) { + victim = e; + max_ec = e->ec; + } + } + + return victim; +} + +/** + * return_unused_pool_pebs - returns unused PEB to the free tree. + * @ubi: UBI device description object + * @pool: fastmap pool description object + */ +static void return_unused_pool_pebs(struct ubi_device *ubi, + struct ubi_fm_pool *pool) +{ + int i; + struct ubi_wl_entry *e; + + for (i = pool->used; i < pool->size; i++) { + e = ubi->lookuptbl[pool->pebs[i]]; + wl_tree_add(e, &ubi->free); + ubi->free_count++; + } +} + +static int anchor_pebs_avalible(struct rb_root *root) +{ + struct rb_node *p; + struct ubi_wl_entry *e; + + ubi_rb_for_each_entry(p, e, root, u.rb) + if (e->pnum < UBI_FM_MAX_START) + return 1; + + return 0; +} + +/** + * ubi_wl_get_fm_peb - find a physical erase block with a given maximal number. + * @ubi: UBI device description object + * @anchor: This PEB will be used as anchor PEB by fastmap + * + * The function returns a physical erase block with a given maximal number + * and removes it from the wl subsystem. + * Must be called with wl_lock held! + */ +struct ubi_wl_entry *ubi_wl_get_fm_peb(struct ubi_device *ubi, int anchor) +{ + struct ubi_wl_entry *e = NULL; + + if (!ubi->free.rb_node || (ubi->free_count - ubi->beb_rsvd_pebs < 1)) + goto out; + + if (anchor) + e = find_anchor_wl_entry(&ubi->free); + else + e = find_mean_wl_entry(ubi, &ubi->free); + + if (!e) + goto out; + + self_check_in_wl_tree(ubi, e, &ubi->free); + + /* remove it from the free list, + * the wl subsystem does no longer know this erase block */ + rb_erase(&e->u.rb, &ubi->free); + ubi->free_count--; +out: + return e; +} + +/** + * ubi_refill_pools - refills all fastmap PEB pools. + * @ubi: UBI device description object + */ +void ubi_refill_pools(struct ubi_device *ubi) +{ + struct ubi_fm_pool *wl_pool = &ubi->fm_wl_pool; + struct ubi_fm_pool *pool = &ubi->fm_pool; + struct ubi_wl_entry *e; + int enough; + + spin_lock(&ubi->wl_lock); + + return_unused_pool_pebs(ubi, wl_pool); + return_unused_pool_pebs(ubi, pool); + + wl_pool->size = 0; + pool->size = 0; + + for (;;) { + enough = 0; + if (pool->size < pool->max_size) { + if (!ubi->free.rb_node) + break; + + e = wl_get_wle(ubi); + if (!e) + break; + + pool->pebs[pool->size] = e->pnum; + pool->size++; + } else + enough++; + + if (wl_pool->size < wl_pool->max_size) { + if (!ubi->free.rb_node || + (ubi->free_count - ubi->beb_rsvd_pebs < 5)) + break; + + e = find_wl_entry(ubi, &ubi->free, WL_FREE_MAX_DIFF); + self_check_in_wl_tree(ubi, e, &ubi->free); + rb_erase(&e->u.rb, &ubi->free); + ubi->free_count--; + + wl_pool->pebs[wl_pool->size] = e->pnum; + wl_pool->size++; + } else + enough++; + + if (enough == 2) + break; + } + + wl_pool->used = 0; + pool->used = 0; + + spin_unlock(&ubi->wl_lock); +} + +/** + * ubi_wl_get_peb - get a physical eraseblock. + * @ubi: UBI device description object + * + * This function returns a physical eraseblock in case of success and a + * negative error code in case of failure. + * Returns with ubi->fm_eba_sem held in read mode! + */ +int ubi_wl_get_peb(struct ubi_device *ubi) +{ + int ret, retried = 0; + struct ubi_fm_pool *pool = &ubi->fm_pool; + struct ubi_fm_pool *wl_pool = &ubi->fm_wl_pool; + +again: + down_read(&ubi->fm_eba_sem); + spin_lock(&ubi->wl_lock); + + /* We check here also for the WL pool because at this point we can + * refill the WL pool synchronous. */ + if (pool->used == pool->size || wl_pool->used == wl_pool->size) { + spin_unlock(&ubi->wl_lock); + up_read(&ubi->fm_eba_sem); + ret = ubi_update_fastmap(ubi); + if (ret) { + ubi_msg(ubi, "Unable to write a new fastmap: %i", ret); + down_read(&ubi->fm_eba_sem); + return -ENOSPC; + } + down_read(&ubi->fm_eba_sem); + spin_lock(&ubi->wl_lock); + } + + if (pool->used == pool->size) { + spin_unlock(&ubi->wl_lock); + if (retried) { + ubi_err(ubi, "Unable to get a free PEB from user WL pool"); + ret = -ENOSPC; + goto out; + } + retried = 1; + up_read(&ubi->fm_eba_sem); + goto again; + } + + ubi_assert(pool->used < pool->size); + ret = pool->pebs[pool->used++]; + prot_queue_add(ubi, ubi->lookuptbl[ret]); + spin_unlock(&ubi->wl_lock); +out: + return ret; +} + +/* get_peb_for_wl - returns a PEB to be used internally by the WL sub-system. + * + * @ubi: UBI device description object + */ +static struct ubi_wl_entry *get_peb_for_wl(struct ubi_device *ubi) +{ + struct ubi_fm_pool *pool = &ubi->fm_wl_pool; + int pnum; + + if (pool->used == pool->size) { + /* We cannot update the fastmap here because this + * function is called in atomic context. + * Let's fail here and refill/update it as soon as possible. */ + if (!ubi->fm_work_scheduled) { + ubi->fm_work_scheduled = 1; + schedule_work(&ubi->fm_work); + } + return NULL; + } + + pnum = pool->pebs[pool->used++]; + return ubi->lookuptbl[pnum]; +} + +/** + * ubi_ensure_anchor_pebs - schedule wear-leveling to produce an anchor PEB. + * @ubi: UBI device description object + */ +int ubi_ensure_anchor_pebs(struct ubi_device *ubi) +{ + struct ubi_work *wrk; + + spin_lock(&ubi->wl_lock); + if (ubi->wl_scheduled) { + spin_unlock(&ubi->wl_lock); + return 0; + } + ubi->wl_scheduled = 1; + spin_unlock(&ubi->wl_lock); + + wrk = kmalloc(sizeof(struct ubi_work), GFP_NOFS); + if (!wrk) { + spin_lock(&ubi->wl_lock); + ubi->wl_scheduled = 0; + spin_unlock(&ubi->wl_lock); + return -ENOMEM; + } + + wrk->anchor = 1; + wrk->func = &wear_leveling_worker; + schedule_ubi_work(ubi, wrk); + return 0; +} + +/** + * ubi_wl_put_fm_peb - returns a PEB used in a fastmap to the wear-leveling + * sub-system. + * see: ubi_wl_put_peb() + * + * @ubi: UBI device description object + * @fm_e: physical eraseblock to return + * @lnum: the last used logical eraseblock number for the PEB + * @torture: if this physical eraseblock has to be tortured + */ +int ubi_wl_put_fm_peb(struct ubi_device *ubi, struct ubi_wl_entry *fm_e, + int lnum, int torture) +{ + struct ubi_wl_entry *e; + int vol_id, pnum = fm_e->pnum; + + dbg_wl("PEB %d", pnum); + + ubi_assert(pnum >= 0); + ubi_assert(pnum < ubi->peb_count); + + spin_lock(&ubi->wl_lock); + e = ubi->lookuptbl[pnum]; + + /* This can happen if we recovered from a fastmap the very + * first time and writing now a new one. In this case the wl system + * has never seen any PEB used by the original fastmap. + */ + if (!e) { + e = fm_e; + ubi_assert(e->ec >= 0); + ubi->lookuptbl[pnum] = e; + } + + spin_unlock(&ubi->wl_lock); + + vol_id = lnum ? UBI_FM_DATA_VOLUME_ID : UBI_FM_SB_VOLUME_ID; + return schedule_erase(ubi, e, vol_id, lnum, torture); +} + +/** + * ubi_is_erase_work - checks whether a work is erase work. + * @wrk: The work object to be checked + */ +int ubi_is_erase_work(struct ubi_work *wrk) +{ + return wrk->func == erase_worker; +} + +static void ubi_fastmap_close(struct ubi_device *ubi) +{ + int i; + + flush_work(&ubi->fm_work); + return_unused_pool_pebs(ubi, &ubi->fm_pool); + return_unused_pool_pebs(ubi, &ubi->fm_wl_pool); + + if (ubi->fm) { + for (i = 0; i < ubi->fm->used_blocks; i++) + kfree(ubi->fm->e[i]); + } + kfree(ubi->fm); +} + +/** + * may_reserve_for_fm - tests whether a PEB shall be reserved for fastmap. + * See find_mean_wl_entry() + * + * @ubi: UBI device description object + * @e: physical eraseblock to return + * @root: RB tree to test against. + */ +static struct ubi_wl_entry *may_reserve_for_fm(struct ubi_device *ubi, + struct ubi_wl_entry *e, + struct rb_root *root) { + if (e && !ubi->fm_disabled && !ubi->fm && + e->pnum < UBI_FM_MAX_START) + e = rb_entry(rb_next(root->rb_node), + struct ubi_wl_entry, u.rb); + + return e; +} diff --git a/drivers/mtd/ubi/fastmap.c b/drivers/mtd/ubi/fastmap.c index db3defdfc3c0..02a6de2f53ee 100644 --- a/drivers/mtd/ubi/fastmap.c +++ b/drivers/mtd/ubi/fastmap.c @@ -1,5 +1,6 @@ /* * Copyright (c) 2012 Linutronix GmbH + * Copyright (c) 2014 sigma star gmbh * Author: Richard Weinberger <richard@nod.at> * * This program is free software; you can redistribute it and/or modify @@ -17,6 +18,69 @@ #include "ubi.h" /** + * init_seen - allocate memory for used for debugging. + * @ubi: UBI device description object + */ +static inline int *init_seen(struct ubi_device *ubi) +{ + int *ret; + + if (!ubi_dbg_chk_fastmap(ubi)) + return NULL; + + ret = kcalloc(ubi->peb_count, sizeof(int), GFP_KERNEL); + if (!ret) + return ERR_PTR(-ENOMEM); + + return ret; +} + +/** + * free_seen - free the seen logic integer array. + * @seen: integer array of @ubi->peb_count size + */ +static inline void free_seen(int *seen) +{ + kfree(seen); +} + +/** + * set_seen - mark a PEB as seen. + * @ubi: UBI device description object + * @pnum: The PEB to be makred as seen + * @seen: integer array of @ubi->peb_count size + */ +static inline void set_seen(struct ubi_device *ubi, int pnum, int *seen) +{ + if (!ubi_dbg_chk_fastmap(ubi) || !seen) + return; + + seen[pnum] = 1; +} + +/** + * self_check_seen - check whether all PEB have been seen by fastmap. + * @ubi: UBI device description object + * @seen: integer array of @ubi->peb_count size + */ +static int self_check_seen(struct ubi_device *ubi, int *seen) +{ + int pnum, ret = 0; + + if (!ubi_dbg_chk_fastmap(ubi) || !seen) + return 0; + + for (pnum = 0; pnum < ubi->peb_count; pnum++) { + if (!seen[pnum] && ubi->lookuptbl[pnum]) { + ubi_err(ubi, "self-check failed for PEB %d, fastmap didn't see it", pnum); + ret = -EINVAL; + } + } + + return ret; +} + +/** * ubi_calc_fm_size - calculates the fastmap size in bytes for an UBI device. * @ubi: UBI device description object */ @@ -136,14 +200,15 @@ static struct ubi_ainf_volume *add_vol(struct ubi_attach_info *ai, int vol_id, if (!av) goto out; - av->highest_lnum = av->leb_count = 0; + av->highest_lnum = av->leb_count = av->used_ebs = 0; av->vol_id = vol_id; - av->used_ebs = used_ebs; av->data_pad = data_pad; av->last_data_size = last_eb_bytes; av->compat = 0; av->vol_type = vol_type; av->root = RB_ROOT; + if (av->vol_type == UBI_STATIC_VOLUME) + av->used_ebs = used_ebs; dbg_bld("found volume (ID %i)", vol_id); @@ -362,6 +427,7 @@ static void unmap_peb(struct ubi_attach_info *ai, int pnum) aeb = rb_entry(node2, struct ubi_ainf_peb, u.rb); if (aeb->pnum == pnum) { rb_erase(&aeb->u.rb, &av->root); + av->leb_count--; kmem_cache_free(ai->aeb_slab_cache, aeb); return; } @@ -376,7 +442,6 @@ static void unmap_peb(struct ubi_attach_info *ai, int pnum) * @pebs: an array of all PEB numbers in the to be scanned pool * @pool_size: size of the pool (number of entries in @pebs) * @max_sqnum: pointer to the maximal sequence number - * @eba_orphans: list of PEBs which need to be scanned * @free: list of PEBs which are most likely free (and go into @ai->free) * * Returns 0 on success, if the pool is unusable UBI_BAD_FASTMAP is returned. @@ -384,12 +449,12 @@ static void unmap_peb(struct ubi_attach_info *ai, int pnum) */ static int scan_pool(struct ubi_device *ubi, struct ubi_attach_info *ai, int *pebs, int pool_size, unsigned long long *max_sqnum, - struct list_head *eba_orphans, struct list_head *free) + struct list_head *free) { struct ubi_vid_hdr *vh; struct ubi_ec_hdr *ech; - struct ubi_ainf_peb *new_aeb, *tmp_aeb; - int i, pnum, err, found_orphan, ret = 0; + struct ubi_ainf_peb *new_aeb; + int i, pnum, err, ret = 0; ech = kzalloc(ubi->ec_hdr_alsize, GFP_KERNEL); if (!ech) @@ -457,18 +522,6 @@ static int scan_pool(struct ubi_device *ubi, struct ubi_attach_info *ai, if (err == UBI_IO_BITFLIPS) scrub = 1; - found_orphan = 0; - list_for_each_entry(tmp_aeb, eba_orphans, u.list) { - if (tmp_aeb->pnum == pnum) { - found_orphan = 1; - break; - } - } - if (found_orphan) { - list_del(&tmp_aeb->u.list); - kmem_cache_free(ai->aeb_slab_cache, tmp_aeb); - } - new_aeb = kmem_cache_alloc(ai->aeb_slab_cache, GFP_KERNEL); if (!new_aeb) { @@ -543,10 +596,9 @@ static int ubi_attach_fastmap(struct ubi_device *ubi, struct ubi_attach_info *ai, struct ubi_fastmap_layout *fm) { - struct list_head used, eba_orphans, free; + struct list_head used, free; struct ubi_ainf_volume *av; struct ubi_ainf_peb *aeb, *tmp_aeb, *_tmp_aeb; - struct ubi_ec_hdr *ech; struct ubi_fm_sb *fmsb; struct ubi_fm_hdr *fmhdr; struct ubi_fm_scan_pool *fmpl1, *fmpl2; @@ -560,22 +612,8 @@ static int ubi_attach_fastmap(struct ubi_device *ubi, INIT_LIST_HEAD(&used); INIT_LIST_HEAD(&free); - INIT_LIST_HEAD(&eba_orphans); - INIT_LIST_HEAD(&ai->corr); - INIT_LIST_HEAD(&ai->free); - INIT_LIST_HEAD(&ai->erase); - INIT_LIST_HEAD(&ai->alien); - ai->volumes = RB_ROOT; ai->min_ec = UBI_MAX_ERASECOUNTER; - ai->aeb_slab_cache = kmem_cache_create("ubi_ainf_peb_slab", - sizeof(struct ubi_ainf_peb), - 0, 0, NULL); - if (!ai->aeb_slab_cache) { - ret = -ENOMEM; - goto fail; - } - fmsb = (struct ubi_fm_sb *)(fm_raw); ai->max_sqnum = fmsb->sqnum; fm_pos += sizeof(struct ubi_fm_sb); @@ -741,28 +779,9 @@ static int ubi_attach_fastmap(struct ubi_device *ubi, } } - /* This can happen if a PEB is already in an EBA known - * by this fastmap but the PEB itself is not in the used - * list. - * In this case the PEB can be within the fastmap pool - * or while writing the fastmap it was in the protection - * queue. - */ if (!aeb) { - aeb = kmem_cache_alloc(ai->aeb_slab_cache, - GFP_KERNEL); - if (!aeb) { - ret = -ENOMEM; - - goto fail; - } - - aeb->lnum = j; - aeb->pnum = be32_to_cpu(fm_eba->pnum[j]); - aeb->ec = -1; - aeb->scrub = aeb->copy_flag = aeb->sqnum = 0; - list_add_tail(&aeb->u.list, &eba_orphans); - continue; + ubi_err(ubi, "PEB %i is in EBA but not in used list", pnum); + goto fail_bad; } aeb->lnum = j; @@ -775,49 +794,13 @@ static int ubi_attach_fastmap(struct ubi_device *ubi, dbg_bld("inserting PEB:%i (LEB %i) to vol %i", aeb->pnum, aeb->lnum, av->vol_id); } - - ech = kzalloc(ubi->ec_hdr_alsize, GFP_KERNEL); - if (!ech) { - ret = -ENOMEM; - goto fail; - } - - list_for_each_entry_safe(tmp_aeb, _tmp_aeb, &eba_orphans, - u.list) { - int err; - - if (ubi_io_is_bad(ubi, tmp_aeb->pnum)) { - ubi_err(ubi, "bad PEB in fastmap EBA orphan list"); - ret = UBI_BAD_FASTMAP; - kfree(ech); - goto fail; - } - - err = ubi_io_read_ec_hdr(ubi, tmp_aeb->pnum, ech, 0); - if (err && err != UBI_IO_BITFLIPS) { - ubi_err(ubi, "unable to read EC header! PEB:%i err:%i", - tmp_aeb->pnum, err); - ret = err > 0 ? UBI_BAD_FASTMAP : err; - kfree(ech); - - goto fail; - } else if (err == UBI_IO_BITFLIPS) - tmp_aeb->scrub = 1; - - tmp_aeb->ec = be64_to_cpu(ech->ec); - assign_aeb_to_av(ai, tmp_aeb, av); - } - - kfree(ech); } - ret = scan_pool(ubi, ai, fmpl1->pebs, pool_size, &max_sqnum, - &eba_orphans, &free); + ret = scan_pool(ubi, ai, fmpl1->pebs, pool_size, &max_sqnum, &free); if (ret) goto fail; - ret = scan_pool(ubi, ai, fmpl2->pebs, wl_pool_size, &max_sqnum, - &eba_orphans, &free); + ret = scan_pool(ubi, ai, fmpl2->pebs, wl_pool_size, &max_sqnum, &free); if (ret) goto fail; @@ -827,8 +810,9 @@ static int ubi_attach_fastmap(struct ubi_device *ubi, list_for_each_entry_safe(tmp_aeb, _tmp_aeb, &free, u.list) list_move_tail(&tmp_aeb->u.list, &ai->free); - ubi_assert(list_empty(&used)); - ubi_assert(list_empty(&eba_orphans)); + list_for_each_entry_safe(tmp_aeb, _tmp_aeb, &used, u.list) + list_move_tail(&tmp_aeb->u.list, &ai->erase); + ubi_assert(list_empty(&free)); /* @@ -850,10 +834,6 @@ fail: list_del(&tmp_aeb->u.list); kmem_cache_free(ai->aeb_slab_cache, tmp_aeb); } - list_for_each_entry_safe(tmp_aeb, _tmp_aeb, &eba_orphans, u.list) { - list_del(&tmp_aeb->u.list); - kmem_cache_free(ai->aeb_slab_cache, tmp_aeb); - } list_for_each_entry_safe(tmp_aeb, _tmp_aeb, &free, u.list) { list_del(&tmp_aeb->u.list); kmem_cache_free(ai->aeb_slab_cache, tmp_aeb); @@ -884,7 +864,7 @@ int ubi_scan_fastmap(struct ubi_device *ubi, struct ubi_attach_info *ai, __be32 crc, tmp_crc; unsigned long long sqnum = 0; - mutex_lock(&ubi->fm_mutex); + down_write(&ubi->fm_protect); memset(ubi->fm_buf, 0, ubi->fm_size); fmsb = kmalloc(sizeof(*fmsb), GFP_KERNEL); @@ -1075,7 +1055,7 @@ int ubi_scan_fastmap(struct ubi_device *ubi, struct ubi_attach_info *ai, ubi_free_vid_hdr(ubi, vh); kfree(ech); out: - mutex_unlock(&ubi->fm_mutex); + up_write(&ubi->fm_protect); if (ret == UBI_BAD_FASTMAP) ubi_err(ubi, "Attach by fastmap failed, doing a full scan!"); return ret; @@ -1107,13 +1087,14 @@ static int ubi_write_fastmap(struct ubi_device *ubi, struct ubi_fm_ec *fec; struct ubi_fm_volhdr *fvh; struct ubi_fm_eba *feba; - struct rb_node *node; struct ubi_wl_entry *wl_e; struct ubi_volume *vol; struct ubi_vid_hdr *avhdr, *dvhdr; struct ubi_work *ubi_wrk; + struct rb_node *tmp_rb; int ret, i, j, free_peb_count, used_peb_count, vol_count; int scrub_peb_count, erase_peb_count; + int *seen_pebs = NULL; fm_raw = ubi->fm_buf; memset(ubi->fm_buf, 0, ubi->fm_size); @@ -1130,6 +1111,12 @@ static int ubi_write_fastmap(struct ubi_device *ubi, goto out_kfree; } + seen_pebs = init_seen(ubi); + if (IS_ERR(seen_pebs)) { + ret = PTR_ERR(seen_pebs); + goto out_kfree; + } + spin_lock(&ubi->volumes_lock); spin_lock(&ubi->wl_lock); @@ -1160,8 +1147,10 @@ static int ubi_write_fastmap(struct ubi_device *ubi, fmpl1->size = cpu_to_be16(ubi->fm_pool.size); fmpl1->max_size = cpu_to_be16(ubi->fm_pool.max_size); - for (i = 0; i < ubi->fm_pool.size; i++) + for (i = 0; i < ubi->fm_pool.size; i++) { fmpl1->pebs[i] = cpu_to_be32(ubi->fm_pool.pebs[i]); + set_seen(ubi, ubi->fm_pool.pebs[i], seen_pebs); + } fmpl2 = (struct ubi_fm_scan_pool *)(fm_raw + fm_pos); fm_pos += sizeof(*fmpl2); @@ -1169,14 +1158,16 @@ static int ubi_write_fastmap(struct ubi_device *ubi, fmpl2->size = cpu_to_be16(ubi->fm_wl_pool.size); fmpl2->max_size = cpu_to_be16(ubi->fm_wl_pool.max_size); - for (i = 0; i < ubi->fm_wl_pool.size; i++) + for (i = 0; i < ubi->fm_wl_pool.size; i++) { fmpl2->pebs[i] = cpu_to_be32(ubi->fm_wl_pool.pebs[i]); + set_seen(ubi, ubi->fm_wl_pool.pebs[i], seen_pebs); + } - for (node = rb_first(&ubi->free); node; node = rb_next(node)) { - wl_e = rb_entry(node, struct ubi_wl_entry, u.rb); + ubi_for_each_free_peb(ubi, wl_e, tmp_rb) { fec = (struct ubi_fm_ec *)(fm_raw + fm_pos); fec->pnum = cpu_to_be32(wl_e->pnum); + set_seen(ubi, wl_e->pnum, seen_pebs); fec->ec = cpu_to_be32(wl_e->ec); free_peb_count++; @@ -1185,11 +1176,11 @@ static int ubi_write_fastmap(struct ubi_device *ubi, } fmh->free_peb_count = cpu_to_be32(free_peb_count); - for (node = rb_first(&ubi->used); node; node = rb_next(node)) { - wl_e = rb_entry(node, struct ubi_wl_entry, u.rb); + ubi_for_each_used_peb(ubi, wl_e, tmp_rb) { fec = (struct ubi_fm_ec *)(fm_raw + fm_pos); fec->pnum = cpu_to_be32(wl_e->pnum); + set_seen(ubi, wl_e->pnum, seen_pebs); fec->ec = cpu_to_be32(wl_e->ec); used_peb_count++; @@ -1197,25 +1188,24 @@ static int ubi_write_fastmap(struct ubi_device *ubi, ubi_assert(fm_pos <= ubi->fm_size); } - for (i = 0; i < UBI_PROT_QUEUE_LEN; i++) { - list_for_each_entry(wl_e, &ubi->pq[i], u.list) { - fec = (struct ubi_fm_ec *)(fm_raw + fm_pos); + ubi_for_each_protected_peb(ubi, i, wl_e) { + fec = (struct ubi_fm_ec *)(fm_raw + fm_pos); - fec->pnum = cpu_to_be32(wl_e->pnum); - fec->ec = cpu_to_be32(wl_e->ec); + fec->pnum = cpu_to_be32(wl_e->pnum); + set_seen(ubi, wl_e->pnum, seen_pebs); + fec->ec = cpu_to_be32(wl_e->ec); - used_peb_count++; - fm_pos += sizeof(*fec); - ubi_assert(fm_pos <= ubi->fm_size); - } + used_peb_count++; + fm_pos += sizeof(*fec); + ubi_assert(fm_pos <= ubi->fm_size); } fmh->used_peb_count = cpu_to_be32(used_peb_count); - for (node = rb_first(&ubi->scrub); node; node = rb_next(node)) { - wl_e = rb_entry(node, struct ubi_wl_entry, u.rb); + ubi_for_each_scrub_peb(ubi, wl_e, tmp_rb) { fec = (struct ubi_fm_ec *)(fm_raw + fm_pos); fec->pnum = cpu_to_be32(wl_e->pnum); + set_seen(ubi, wl_e->pnum, seen_pebs); fec->ec = cpu_to_be32(wl_e->ec); scrub_peb_count++; @@ -1233,6 +1223,7 @@ static int ubi_write_fastmap(struct ubi_device *ubi, fec = (struct ubi_fm_ec *)(fm_raw + fm_pos); fec->pnum = cpu_to_be32(wl_e->pnum); + set_seen(ubi, wl_e->pnum, seen_pebs); fec->ec = cpu_to_be32(wl_e->ec); erase_peb_count++; @@ -1292,6 +1283,7 @@ static int ubi_write_fastmap(struct ubi_device *ubi, for (i = 0; i < new_fm->used_blocks; i++) { fmsb->block_loc[i] = cpu_to_be32(new_fm->e[i]->pnum); + set_seen(ubi, new_fm->e[i]->pnum, seen_pebs); fmsb->block_ec[i] = cpu_to_be32(new_fm->e[i]->ec); } @@ -1325,11 +1317,13 @@ static int ubi_write_fastmap(struct ubi_device *ubi, ubi_assert(new_fm); ubi->fm = new_fm; + ret = self_check_seen(ubi, seen_pebs); dbg_bld("fastmap written!"); out_kfree: ubi_free_vid_hdr(ubi, avhdr); ubi_free_vid_hdr(ubi, dvhdr); + free_seen(seen_pebs); out: return ret; } @@ -1384,31 +1378,87 @@ out: /** * invalidate_fastmap - destroys a fastmap. * @ubi: UBI device object - * @fm: the fastmap to be destroyed * + * This function ensures that upon next UBI attach a full scan + * is issued. We need this if UBI is about to write a new fastmap + * but is unable to do so. In this case we have two options: + * a) Make sure that the current fastmap will not be usued upon + * attach time and contine or b) fall back to RO mode to have the + * current fastmap in a valid state. * Returns 0 on success, < 0 indicates an internal error. */ -static int invalidate_fastmap(struct ubi_device *ubi, - struct ubi_fastmap_layout *fm) +static int invalidate_fastmap(struct ubi_device *ubi) { int ret; - struct ubi_vid_hdr *vh; + struct ubi_fastmap_layout *fm; + struct ubi_wl_entry *e; + struct ubi_vid_hdr *vh = NULL; - ret = erase_block(ubi, fm->e[0]->pnum); - if (ret < 0) - return ret; + if (!ubi->fm) + return 0; + + ubi->fm = NULL; + + ret = -ENOMEM; + fm = kzalloc(sizeof(*fm), GFP_KERNEL); + if (!fm) + goto out; vh = new_fm_vhdr(ubi, UBI_FM_SB_VOLUME_ID); if (!vh) - return -ENOMEM; + goto out_free_fm; - /* deleting the current fastmap SB is not enough, an old SB may exist, - * so create a (corrupted) SB such that fastmap will find it and fall - * back to scanning mode in any case */ + ret = -ENOSPC; + e = ubi_wl_get_fm_peb(ubi, 1); + if (!e) + goto out_free_fm; + + /* + * Create fake fastmap such that UBI will fall back + * to scanning mode. + */ vh->sqnum = cpu_to_be64(ubi_next_sqnum(ubi)); - ret = ubi_io_write_vid_hdr(ubi, fm->e[0]->pnum, vh); + ret = ubi_io_write_vid_hdr(ubi, e->pnum, vh); + if (ret < 0) { + ubi_wl_put_fm_peb(ubi, e, 0, 0); + goto out_free_fm; + } + + fm->used_blocks = 1; + fm->e[0] = e; + + ubi->fm = fm; +out: + ubi_free_vid_hdr(ubi, vh); return ret; + +out_free_fm: + kfree(fm); + goto out; +} + +/** + * return_fm_pebs - returns all PEBs used by a fastmap back to the + * WL sub-system. + * @ubi: UBI device object + * @fm: fastmap layout object + */ +static void return_fm_pebs(struct ubi_device *ubi, + struct ubi_fastmap_layout *fm) +{ + int i; + + if (!fm) + return; + + for (i = 0; i < fm->used_blocks; i++) { + if (fm->e[i]) { + ubi_wl_put_fm_peb(ubi, fm->e[i], i, + fm->to_be_tortured[i]); + fm->e[i] = NULL; + } + } } /** @@ -1420,45 +1470,32 @@ static int invalidate_fastmap(struct ubi_device *ubi, */ int ubi_update_fastmap(struct ubi_device *ubi) { - int ret, i; + int ret, i, j; struct ubi_fastmap_layout *new_fm, *old_fm; struct ubi_wl_entry *tmp_e; - mutex_lock(&ubi->fm_mutex); + down_write(&ubi->fm_protect); ubi_refill_pools(ubi); if (ubi->ro_mode || ubi->fm_disabled) { - mutex_unlock(&ubi->fm_mutex); + up_write(&ubi->fm_protect); return 0; } ret = ubi_ensure_anchor_pebs(ubi); if (ret) { - mutex_unlock(&ubi->fm_mutex); + up_write(&ubi->fm_protect); return ret; } new_fm = kzalloc(sizeof(*new_fm), GFP_KERNEL); if (!new_fm) { - mutex_unlock(&ubi->fm_mutex); + up_write(&ubi->fm_protect); return -ENOMEM; } new_fm->used_blocks = ubi->fm_size / ubi->leb_size; - - for (i = 0; i < new_fm->used_blocks; i++) { - new_fm->e[i] = kmem_cache_alloc(ubi_wl_entry_slab, GFP_KERNEL); - if (!new_fm->e[i]) { - while (i--) - kfree(new_fm->e[i]); - - kfree(new_fm); - mutex_unlock(&ubi->fm_mutex); - return -ENOMEM; - } - } - old_fm = ubi->fm; ubi->fm = NULL; @@ -1473,37 +1510,49 @@ int ubi_update_fastmap(struct ubi_device *ubi) tmp_e = ubi_wl_get_fm_peb(ubi, 0); spin_unlock(&ubi->wl_lock); - if (!tmp_e && !old_fm) { - int j; - ubi_err(ubi, "could not get any free erase block"); - - for (j = 1; j < i; j++) - ubi_wl_put_fm_peb(ubi, new_fm->e[j], j, 0); - - ret = -ENOSPC; - goto err; - } else if (!tmp_e && old_fm) { - ret = erase_block(ubi, old_fm->e[i]->pnum); - if (ret < 0) { - int j; - - for (j = 1; j < i; j++) - ubi_wl_put_fm_peb(ubi, new_fm->e[j], - j, 0); + if (!tmp_e) { + if (old_fm && old_fm->e[i]) { + ret = erase_block(ubi, old_fm->e[i]->pnum); + if (ret < 0) { + ubi_err(ubi, "could not erase old fastmap PEB"); + + for (j = 1; j < i; j++) { + ubi_wl_put_fm_peb(ubi, new_fm->e[j], + j, 0); + new_fm->e[j] = NULL; + } + goto err; + } + new_fm->e[i] = old_fm->e[i]; + old_fm->e[i] = NULL; + } else { + ubi_err(ubi, "could not get any free erase block"); + + for (j = 1; j < i; j++) { + ubi_wl_put_fm_peb(ubi, new_fm->e[j], j, 0); + new_fm->e[j] = NULL; + } - ubi_err(ubi, "could not erase old fastmap PEB"); + ret = -ENOSPC; goto err; } - - new_fm->e[i]->pnum = old_fm->e[i]->pnum; - new_fm->e[i]->ec = old_fm->e[i]->ec; } else { - new_fm->e[i]->pnum = tmp_e->pnum; - new_fm->e[i]->ec = tmp_e->ec; + new_fm->e[i] = tmp_e; - if (old_fm) + if (old_fm && old_fm->e[i]) { ubi_wl_put_fm_peb(ubi, old_fm->e[i], i, old_fm->to_be_tortured[i]); + old_fm->e[i] = NULL; + } + } + } + + /* Old fastmap is larger than the new one */ + if (old_fm && new_fm->used_blocks < old_fm->used_blocks) { + for (i = new_fm->used_blocks; i < old_fm->used_blocks; i++) { + ubi_wl_put_fm_peb(ubi, old_fm->e[i], i, + old_fm->to_be_tortured[i]); + old_fm->e[i] = NULL; } } @@ -1516,67 +1565,67 @@ int ubi_update_fastmap(struct ubi_device *ubi) if (!tmp_e) { ret = erase_block(ubi, old_fm->e[0]->pnum); if (ret < 0) { - int i; ubi_err(ubi, "could not erase old anchor PEB"); - for (i = 1; i < new_fm->used_blocks; i++) + for (i = 1; i < new_fm->used_blocks; i++) { ubi_wl_put_fm_peb(ubi, new_fm->e[i], i, 0); + new_fm->e[i] = NULL; + } goto err; } - - new_fm->e[0]->pnum = old_fm->e[0]->pnum; + new_fm->e[0] = old_fm->e[0]; new_fm->e[0]->ec = ret; + old_fm->e[0] = NULL; } else { /* we've got a new anchor PEB, return the old one */ ubi_wl_put_fm_peb(ubi, old_fm->e[0], 0, old_fm->to_be_tortured[0]); - - new_fm->e[0]->pnum = tmp_e->pnum; - new_fm->e[0]->ec = tmp_e->ec; + new_fm->e[0] = tmp_e; + old_fm->e[0] = NULL; } } else { if (!tmp_e) { - int i; ubi_err(ubi, "could not find any anchor PEB"); - for (i = 1; i < new_fm->used_blocks; i++) + for (i = 1; i < new_fm->used_blocks; i++) { ubi_wl_put_fm_peb(ubi, new_fm->e[i], i, 0); + new_fm->e[i] = NULL; + } ret = -ENOSPC; goto err; } - - new_fm->e[0]->pnum = tmp_e->pnum; - new_fm->e[0]->ec = tmp_e->ec; + new_fm->e[0] = tmp_e; } down_write(&ubi->work_sem); - down_write(&ubi->fm_sem); + down_write(&ubi->fm_eba_sem); ret = ubi_write_fastmap(ubi, new_fm); - up_write(&ubi->fm_sem); + up_write(&ubi->fm_eba_sem); up_write(&ubi->work_sem); if (ret) goto err; out_unlock: - mutex_unlock(&ubi->fm_mutex); + up_write(&ubi->fm_protect); kfree(old_fm); return ret; err: - kfree(new_fm); - ubi_warn(ubi, "Unable to write new fastmap, err=%i", ret); - ret = 0; - if (old_fm) { - ret = invalidate_fastmap(ubi, old_fm); - if (ret < 0) - ubi_err(ubi, "Unable to invalidiate current fastmap!"); - else if (ret) - ret = 0; + ret = invalidate_fastmap(ubi); + if (ret < 0) { + ubi_err(ubi, "Unable to invalidiate current fastmap!"); + ubi_ro_mode(ubi); + } else { + return_fm_pebs(ubi, old_fm); + return_fm_pebs(ubi, new_fm); + ret = 0; } + + kfree(new_fm); goto out_unlock; } diff --git a/drivers/mtd/ubi/io.c b/drivers/mtd/ubi/io.c index ed0bcb35472f..5bbd1f094f4e 100644 --- a/drivers/mtd/ubi/io.c +++ b/drivers/mtd/ubi/io.c @@ -859,6 +859,9 @@ int ubi_io_write_ec_hdr(struct ubi_device *ubi, int pnum, if (err) return err; + if (ubi_dbg_power_cut(ubi, POWER_CUT_EC_WRITE)) + return -EROFS; + err = ubi_io_write(ubi, ec_hdr, pnum, 0, ubi->ec_hdr_alsize); return err; } @@ -1106,6 +1109,9 @@ int ubi_io_write_vid_hdr(struct ubi_device *ubi, int pnum, if (err) return err; + if (ubi_dbg_power_cut(ubi, POWER_CUT_VID_WRITE)) + return -EROFS; + p = (char *)vid_hdr - ubi->vid_hdr_shift; err = ubi_io_write(ubi, p, pnum, ubi->vid_hdr_aloffset, ubi->vid_hdr_alsize); diff --git a/drivers/mtd/ubi/ubi-media.h b/drivers/mtd/ubi/ubi-media.h index ac2b24d1783d..d0d072e7ccd2 100644 --- a/drivers/mtd/ubi/ubi-media.h +++ b/drivers/mtd/ubi/ubi-media.h @@ -403,8 +403,6 @@ struct ubi_vtbl_record { #define UBI_FM_MIN_POOL_SIZE 8 #define UBI_FM_MAX_POOL_SIZE 256 -#define UBI_FM_WL_POOL_SIZE 25 - /** * struct ubi_fm_sb - UBI fastmap super block * @magic: fastmap super block magic number (%UBI_FM_SB_MAGIC) diff --git a/drivers/mtd/ubi/ubi.h b/drivers/mtd/ubi/ubi.h index c5be82d9d345..c998212fc680 100644 --- a/drivers/mtd/ubi/ubi.h +++ b/drivers/mtd/ubi/ubi.h @@ -151,6 +151,17 @@ enum { UBI_BAD_FASTMAP, }; +/* + * Flags for emulate_power_cut in ubi_debug_info + * + * POWER_CUT_EC_WRITE: Emulate a power cut when writing an EC header + * POWER_CUT_VID_WRITE: Emulate a power cut when writing a VID header + */ +enum { + POWER_CUT_EC_WRITE = 0x01, + POWER_CUT_VID_WRITE = 0x02, +}; + /** * struct ubi_wl_entry - wear-leveling entry. * @u.rb: link in the corresponding (free/used) RB-tree @@ -356,30 +367,48 @@ struct ubi_wl_entry; * * @chk_gen: if UBI general extra checks are enabled * @chk_io: if UBI I/O extra checks are enabled + * @chk_fastmap: if UBI fastmap extra checks are enabled * @disable_bgt: disable the background task for testing purposes * @emulate_bitflips: emulate bit-flips for testing purposes * @emulate_io_failures: emulate write/erase failures for testing purposes + * @emulate_power_cut: emulate power cut for testing purposes + * @power_cut_counter: count down for writes left until emulated power cut + * @power_cut_min: minimum number of writes before emulating a power cut + * @power_cut_max: maximum number of writes until emulating a power cut * @dfs_dir_name: name of debugfs directory containing files of this UBI device * @dfs_dir: direntry object of the UBI device debugfs directory * @dfs_chk_gen: debugfs knob to enable UBI general extra checks * @dfs_chk_io: debugfs knob to enable UBI I/O extra checks + * @dfs_chk_fastmap: debugfs knob to enable UBI fastmap extra checks * @dfs_disable_bgt: debugfs knob to disable the background task * @dfs_emulate_bitflips: debugfs knob to emulate bit-flips * @dfs_emulate_io_failures: debugfs knob to emulate write/erase failures + * @dfs_emulate_power_cut: debugfs knob to emulate power cuts + * @dfs_power_cut_min: debugfs knob for minimum writes before power cut + * @dfs_power_cut_max: debugfs knob for maximum writes until power cut */ struct ubi_debug_info { unsigned int chk_gen:1; unsigned int chk_io:1; + unsigned int chk_fastmap:1; unsigned int disable_bgt:1; unsigned int emulate_bitflips:1; unsigned int emulate_io_failures:1; + unsigned int emulate_power_cut:2; + unsigned int power_cut_counter; + unsigned int power_cut_min; + unsigned int power_cut_max; char dfs_dir_name[UBI_DFS_DIR_LEN + 1]; struct dentry *dfs_dir; struct dentry *dfs_chk_gen; struct dentry *dfs_chk_io; + struct dentry *dfs_chk_fastmap; struct dentry *dfs_disable_bgt; struct dentry *dfs_emulate_bitflips; struct dentry *dfs_emulate_io_failures; + struct dentry *dfs_emulate_power_cut; + struct dentry *dfs_power_cut_min; + struct dentry *dfs_power_cut_max; }; /** @@ -426,11 +455,13 @@ struct ubi_debug_info { * @fm_pool: in-memory data structure of the fastmap pool * @fm_wl_pool: in-memory data structure of the fastmap pool used by the WL * sub-system - * @fm_mutex: serializes ubi_update_fastmap() and protects @fm_buf + * @fm_protect: serializes ubi_update_fastmap(), protects @fm_buf and makes sure + * that critical sections cannot be interrupted by ubi_update_fastmap() * @fm_buf: vmalloc()'d buffer which holds the raw fastmap * @fm_size: fastmap size in bytes - * @fm_sem: allows ubi_update_fastmap() to block EBA table changes + * @fm_eba_sem: allows ubi_update_fastmap() to block EBA table changes * @fm_work: fastmap work queue + * @fm_work_scheduled: non-zero if fastmap work was scheduled * * @used: RB-tree of used physical eraseblocks * @erroneous: RB-tree of erroneous used physical eraseblocks @@ -442,7 +473,8 @@ struct ubi_debug_info { * @pq_head: protection queue head * @wl_lock: protects the @used, @free, @pq, @pq_head, @lookuptbl, @move_from, * @move_to, @move_to_put @erase_pending, @wl_scheduled, @works, - * @erroneous, and @erroneous_peb_count fields + * @erroneous, @erroneous_peb_count, @fm_work_scheduled, @fm_pool, + * and @fm_wl_pool fields * @move_mutex: serializes eraseblock moves * @work_sem: used to wait for all the scheduled works to finish and prevent * new works from being submitted @@ -479,7 +511,7 @@ struct ubi_debug_info { * @vid_hdr_offset: starting offset of the volume identifier header (might be * unaligned) * @vid_hdr_aloffset: starting offset of the VID header aligned to - * @hdrs_min_io_size + * @hdrs_min_io_size * @vid_hdr_shift: contains @vid_hdr_offset - @vid_hdr_aloffset * @bad_allowed: whether the MTD device admits of bad physical eraseblocks or * not @@ -532,11 +564,12 @@ struct ubi_device { struct ubi_fastmap_layout *fm; struct ubi_fm_pool fm_pool; struct ubi_fm_pool fm_wl_pool; - struct rw_semaphore fm_sem; - struct mutex fm_mutex; + struct rw_semaphore fm_eba_sem; + struct rw_semaphore fm_protect; void *fm_buf; size_t fm_size; struct work_struct fm_work; + int fm_work_scheduled; /* Wear-leveling sub-system's stuff */ struct rb_root used; @@ -868,10 +901,14 @@ int ubi_compare_lebs(struct ubi_device *ubi, const struct ubi_ainf_peb *aeb, int pnum, const struct ubi_vid_hdr *vid_hdr); /* fastmap.c */ +#ifdef CONFIG_MTD_UBI_FASTMAP size_t ubi_calc_fm_size(struct ubi_device *ubi); int ubi_update_fastmap(struct ubi_device *ubi); int ubi_scan_fastmap(struct ubi_device *ubi, struct ubi_attach_info *ai, int fm_anchor); +#else +static inline int ubi_update_fastmap(struct ubi_device *ubi) { return 0; } +#endif /* block.c */ #ifdef CONFIG_MTD_UBI_BLOCK @@ -892,6 +929,42 @@ static inline int ubiblock_remove(struct ubi_volume_info *vi) } #endif +/* + * ubi_for_each_free_peb - walk the UBI free RB tree. + * @ubi: UBI device description object + * @e: a pointer to a ubi_wl_entry to use as cursor + * @pos: a pointer to RB-tree entry type to use as a loop counter + */ +#define ubi_for_each_free_peb(ubi, e, tmp_rb) \ + ubi_rb_for_each_entry((tmp_rb), (e), &(ubi)->free, u.rb) + +/* + * ubi_for_each_used_peb - walk the UBI used RB tree. + * @ubi: UBI device description object + * @e: a pointer to a ubi_wl_entry to use as cursor + * @pos: a pointer to RB-tree entry type to use as a loop counter + */ +#define ubi_for_each_used_peb(ubi, e, tmp_rb) \ + ubi_rb_for_each_entry((tmp_rb), (e), &(ubi)->used, u.rb) + +/* + * ubi_for_each_scub_peb - walk the UBI scub RB tree. + * @ubi: UBI device description object + * @e: a pointer to a ubi_wl_entry to use as cursor + * @pos: a pointer to RB-tree entry type to use as a loop counter + */ +#define ubi_for_each_scrub_peb(ubi, e, tmp_rb) \ + ubi_rb_for_each_entry((tmp_rb), (e), &(ubi)->scrub, u.rb) + +/* + * ubi_for_each_protected_peb - walk the UBI protection queue. + * @ubi: UBI device description object + * @i: a integer used as counter + * @e: a pointer to a ubi_wl_entry to use as cursor + */ +#define ubi_for_each_protected_peb(ubi, i, e) \ + for ((i) = 0; (i) < UBI_PROT_QUEUE_LEN; (i)++) \ + list_for_each_entry((e), &(ubi->pq[(i)]), u.list) /* * ubi_rb_for_each_entry - walk an RB-tree. diff --git a/drivers/mtd/ubi/wl.c b/drivers/mtd/ubi/wl.c index 8f7bde6a85d6..16214d3d57a4 100644 --- a/drivers/mtd/ubi/wl.c +++ b/drivers/mtd/ubi/wl.c @@ -103,6 +103,7 @@ #include <linux/freezer.h> #include <linux/kthread.h> #include "ubi.h" +#include "wl.h" /* Number of physical eraseblocks reserved for wear-leveling purposes */ #define WL_RESERVED_PEBS 1 @@ -140,42 +141,6 @@ static int self_check_in_wl_tree(const struct ubi_device *ubi, static int self_check_in_pq(const struct ubi_device *ubi, struct ubi_wl_entry *e); -#ifdef CONFIG_MTD_UBI_FASTMAP -/** - * update_fastmap_work_fn - calls ubi_update_fastmap from a work queue - * @wrk: the work description object - */ -static void update_fastmap_work_fn(struct work_struct *wrk) -{ - struct ubi_device *ubi = container_of(wrk, struct ubi_device, fm_work); - ubi_update_fastmap(ubi); -} - -/** - * ubi_ubi_is_fm_block - returns 1 if a PEB is currently used in a fastmap. - * @ubi: UBI device description object - * @pnum: the to be checked PEB - */ -static int ubi_is_fm_block(struct ubi_device *ubi, int pnum) -{ - int i; - - if (!ubi->fm) - return 0; - - for (i = 0; i < ubi->fm->used_blocks; i++) - if (ubi->fm->e[i]->pnum == pnum) - return 1; - - return 0; -} -#else -static int ubi_is_fm_block(struct ubi_device *ubi, int pnum) -{ - return 0; -} -#endif - /** * wl_tree_add - add a wear-leveling entry to a WL RB-tree. * @e: the wear-leveling entry to add @@ -213,6 +178,20 @@ static void wl_tree_add(struct ubi_wl_entry *e, struct rb_root *root) } /** + * wl_tree_destroy - destroy a wear-leveling entry. + * @ubi: UBI device description object + * @e: the wear-leveling entry to add + * + * This function destroys a wear leveling entry and removes + * the reference from the lookup table. + */ +static void wl_entry_destroy(struct ubi_device *ubi, struct ubi_wl_entry *e) +{ + ubi->lookuptbl[e->pnum] = NULL; + kmem_cache_free(ubi_wl_entry_slab, e); +} + +/** * do_work - do one pending work. * @ubi: UBI device description object * @@ -260,33 +239,6 @@ static int do_work(struct ubi_device *ubi) } /** - * produce_free_peb - produce a free physical eraseblock. - * @ubi: UBI device description object - * - * This function tries to make a free PEB by means of synchronous execution of - * pending works. This may be needed if, for example the background thread is - * disabled. Returns zero in case of success and a negative error code in case - * of failure. - */ -static int produce_free_peb(struct ubi_device *ubi) -{ - int err; - - while (!ubi->free.rb_node && ubi->works_count) { - spin_unlock(&ubi->wl_lock); - - dbg_wl("do one work synchronously"); - err = do_work(ubi); - - spin_lock(&ubi->wl_lock); - if (err) - return err; - } - - return 0; -} - -/** * in_wl_tree - check if wear-leveling entry is present in a WL RB-tree. * @e: the wear-leveling entry to check * @root: the root of the tree @@ -409,119 +361,32 @@ static struct ubi_wl_entry *find_mean_wl_entry(struct ubi_device *ubi, if (last->ec - first->ec < WL_FREE_MAX_DIFF) { e = rb_entry(root->rb_node, struct ubi_wl_entry, u.rb); -#ifdef CONFIG_MTD_UBI_FASTMAP /* If no fastmap has been written and this WL entry can be used * as anchor PEB, hold it back and return the second best * WL entry such that fastmap can use the anchor PEB later. */ - if (e && !ubi->fm_disabled && !ubi->fm && - e->pnum < UBI_FM_MAX_START) - e = rb_entry(rb_next(root->rb_node), - struct ubi_wl_entry, u.rb); -#endif + e = may_reserve_for_fm(ubi, e, root); } else e = find_wl_entry(ubi, root, WL_FREE_MAX_DIFF/2); return e; } -#ifdef CONFIG_MTD_UBI_FASTMAP -/** - * find_anchor_wl_entry - find wear-leveling entry to used as anchor PEB. - * @root: the RB-tree where to look for - */ -static struct ubi_wl_entry *find_anchor_wl_entry(struct rb_root *root) -{ - struct rb_node *p; - struct ubi_wl_entry *e, *victim = NULL; - int max_ec = UBI_MAX_ERASECOUNTER; - - ubi_rb_for_each_entry(p, e, root, u.rb) { - if (e->pnum < UBI_FM_MAX_START && e->ec < max_ec) { - victim = e; - max_ec = e->ec; - } - } - - return victim; -} - -static int anchor_pebs_avalible(struct rb_root *root) -{ - struct rb_node *p; - struct ubi_wl_entry *e; - - ubi_rb_for_each_entry(p, e, root, u.rb) - if (e->pnum < UBI_FM_MAX_START) - return 1; - - return 0; -} - /** - * ubi_wl_get_fm_peb - find a physical erase block with a given maximal number. + * wl_get_wle - get a mean wl entry to be used by ubi_wl_get_peb() or + * refill_wl_user_pool(). * @ubi: UBI device description object - * @anchor: This PEB will be used as anchor PEB by fastmap * - * The function returns a physical erase block with a given maximal number - * and removes it from the wl subsystem. - * Must be called with wl_lock held! + * This function returns a a wear leveling entry in case of success and + * NULL in case of failure. */ -struct ubi_wl_entry *ubi_wl_get_fm_peb(struct ubi_device *ubi, int anchor) +static struct ubi_wl_entry *wl_get_wle(struct ubi_device *ubi) { - struct ubi_wl_entry *e = NULL; - - if (!ubi->free.rb_node || (ubi->free_count - ubi->beb_rsvd_pebs < 1)) - goto out; - - if (anchor) - e = find_anchor_wl_entry(&ubi->free); - else - e = find_mean_wl_entry(ubi, &ubi->free); - - if (!e) - goto out; - - self_check_in_wl_tree(ubi, e, &ubi->free); - - /* remove it from the free list, - * the wl subsystem does no longer know this erase block */ - rb_erase(&e->u.rb, &ubi->free); - ubi->free_count--; -out: - return e; -} -#endif - -/** - * __wl_get_peb - get a physical eraseblock. - * @ubi: UBI device description object - * - * This function returns a physical eraseblock in case of success and a - * negative error code in case of failure. - */ -static int __wl_get_peb(struct ubi_device *ubi) -{ - int err; struct ubi_wl_entry *e; -retry: - if (!ubi->free.rb_node) { - if (ubi->works_count == 0) { - ubi_err(ubi, "no free eraseblocks"); - ubi_assert(list_empty(&ubi->works)); - return -ENOSPC; - } - - err = produce_free_peb(ubi); - if (err < 0) - return err; - goto retry; - } - e = find_mean_wl_entry(ubi, &ubi->free); if (!e) { ubi_err(ubi, "no free eraseblocks"); - return -ENOSPC; + return NULL; } self_check_in_wl_tree(ubi, e, &ubi->free); @@ -533,174 +398,10 @@ retry: rb_erase(&e->u.rb, &ubi->free); ubi->free_count--; dbg_wl("PEB %d EC %d", e->pnum, e->ec); -#ifndef CONFIG_MTD_UBI_FASTMAP - /* We have to enqueue e only if fastmap is disabled, - * is fastmap enabled prot_queue_add() will be called by - * ubi_wl_get_peb() after removing e from the pool. */ - prot_queue_add(ubi, e); -#endif - return e->pnum; -} - -#ifdef CONFIG_MTD_UBI_FASTMAP -/** - * return_unused_pool_pebs - returns unused PEB to the free tree. - * @ubi: UBI device description object - * @pool: fastmap pool description object - */ -static void return_unused_pool_pebs(struct ubi_device *ubi, - struct ubi_fm_pool *pool) -{ - int i; - struct ubi_wl_entry *e; - - for (i = pool->used; i < pool->size; i++) { - e = ubi->lookuptbl[pool->pebs[i]]; - wl_tree_add(e, &ubi->free); - ubi->free_count++; - } -} - -/** - * refill_wl_pool - refills all the fastmap pool used by the - * WL sub-system. - * @ubi: UBI device description object - */ -static void refill_wl_pool(struct ubi_device *ubi) -{ - struct ubi_wl_entry *e; - struct ubi_fm_pool *pool = &ubi->fm_wl_pool; - - return_unused_pool_pebs(ubi, pool); - - for (pool->size = 0; pool->size < pool->max_size; pool->size++) { - if (!ubi->free.rb_node || - (ubi->free_count - ubi->beb_rsvd_pebs < 5)) - break; - - e = find_wl_entry(ubi, &ubi->free, WL_FREE_MAX_DIFF); - self_check_in_wl_tree(ubi, e, &ubi->free); - rb_erase(&e->u.rb, &ubi->free); - ubi->free_count--; - - pool->pebs[pool->size] = e->pnum; - } - pool->used = 0; -} - -/** - * refill_wl_user_pool - refills all the fastmap pool used by ubi_wl_get_peb. - * @ubi: UBI device description object - */ -static void refill_wl_user_pool(struct ubi_device *ubi) -{ - struct ubi_fm_pool *pool = &ubi->fm_pool; - - return_unused_pool_pebs(ubi, pool); - - for (pool->size = 0; pool->size < pool->max_size; pool->size++) { - pool->pebs[pool->size] = __wl_get_peb(ubi); - if (pool->pebs[pool->size] < 0) - break; - } - pool->used = 0; -} - -/** - * ubi_refill_pools - refills all fastmap PEB pools. - * @ubi: UBI device description object - */ -void ubi_refill_pools(struct ubi_device *ubi) -{ - spin_lock(&ubi->wl_lock); - refill_wl_pool(ubi); - refill_wl_user_pool(ubi); - spin_unlock(&ubi->wl_lock); -} - -/* ubi_wl_get_peb - works exaclty like __wl_get_peb but keeps track of - * the fastmap pool. - */ -int ubi_wl_get_peb(struct ubi_device *ubi) -{ - int ret; - struct ubi_fm_pool *pool = &ubi->fm_pool; - struct ubi_fm_pool *wl_pool = &ubi->fm_wl_pool; - - if (!pool->size || !wl_pool->size || pool->used == pool->size || - wl_pool->used == wl_pool->size) - ubi_update_fastmap(ubi); - - /* we got not a single free PEB */ - if (!pool->size) - ret = -ENOSPC; - else { - spin_lock(&ubi->wl_lock); - ret = pool->pebs[pool->used++]; - prot_queue_add(ubi, ubi->lookuptbl[ret]); - spin_unlock(&ubi->wl_lock); - } - - return ret; -} - -/* get_peb_for_wl - returns a PEB to be used internally by the WL sub-system. - * - * @ubi: UBI device description object - */ -static struct ubi_wl_entry *get_peb_for_wl(struct ubi_device *ubi) -{ - struct ubi_fm_pool *pool = &ubi->fm_wl_pool; - int pnum; - - if (pool->used == pool->size || !pool->size) { - /* We cannot update the fastmap here because this - * function is called in atomic context. - * Let's fail here and refill/update it as soon as possible. */ - schedule_work(&ubi->fm_work); - return NULL; - } else { - pnum = pool->pebs[pool->used++]; - return ubi->lookuptbl[pnum]; - } -} -#else -static struct ubi_wl_entry *get_peb_for_wl(struct ubi_device *ubi) -{ - struct ubi_wl_entry *e; - - e = find_wl_entry(ubi, &ubi->free, WL_FREE_MAX_DIFF); - self_check_in_wl_tree(ubi, e, &ubi->free); - ubi->free_count--; - ubi_assert(ubi->free_count >= 0); - rb_erase(&e->u.rb, &ubi->free); return e; } -int ubi_wl_get_peb(struct ubi_device *ubi) -{ - int peb, err; - - spin_lock(&ubi->wl_lock); - peb = __wl_get_peb(ubi); - spin_unlock(&ubi->wl_lock); - - if (peb < 0) - return peb; - - err = ubi_self_check_all_ff(ubi, peb, ubi->vid_hdr_aloffset, - ubi->peb_size - ubi->vid_hdr_aloffset); - if (err) { - ubi_err(ubi, "new PEB %d does not contain all 0xFF bytes", - peb); - return err; - } - - return peb; -} -#endif - /** * prot_queue_del - remove a physical eraseblock from the protection queue. * @ubi: UBI device description object @@ -867,17 +568,6 @@ static void schedule_ubi_work(struct ubi_device *ubi, struct ubi_work *wrk) static int erase_worker(struct ubi_device *ubi, struct ubi_work *wl_wrk, int shutdown); -#ifdef CONFIG_MTD_UBI_FASTMAP -/** - * ubi_is_erase_work - checks whether a work is erase work. - * @wrk: The work object to be checked - */ -int ubi_is_erase_work(struct ubi_work *wrk) -{ - return wrk->func == erase_worker; -} -#endif - /** * schedule_erase - schedule an erase work. * @ubi: UBI device description object @@ -895,7 +585,6 @@ static int schedule_erase(struct ubi_device *ubi, struct ubi_wl_entry *e, struct ubi_work *wl_wrk; ubi_assert(e); - ubi_assert(!ubi_is_fm_block(ubi, e->pnum)); dbg_wl("schedule erasure of PEB %d, EC %d, torture %d", e->pnum, e->ec, torture); @@ -942,51 +631,6 @@ static int do_sync_erase(struct ubi_device *ubi, struct ubi_wl_entry *e, return erase_worker(ubi, wl_wrk, 0); } -#ifdef CONFIG_MTD_UBI_FASTMAP -/** - * ubi_wl_put_fm_peb - returns a PEB used in a fastmap to the wear-leveling - * sub-system. - * see: ubi_wl_put_peb() - * - * @ubi: UBI device description object - * @fm_e: physical eraseblock to return - * @lnum: the last used logical eraseblock number for the PEB - * @torture: if this physical eraseblock has to be tortured - */ -int ubi_wl_put_fm_peb(struct ubi_device *ubi, struct ubi_wl_entry *fm_e, - int lnum, int torture) -{ - struct ubi_wl_entry *e; - int vol_id, pnum = fm_e->pnum; - - dbg_wl("PEB %d", pnum); - - ubi_assert(pnum >= 0); - ubi_assert(pnum < ubi->peb_count); - - spin_lock(&ubi->wl_lock); - e = ubi->lookuptbl[pnum]; - - /* This can happen if we recovered from a fastmap the very - * first time and writing now a new one. In this case the wl system - * has never seen any PEB used by the original fastmap. - */ - if (!e) { - e = fm_e; - ubi_assert(e->ec >= 0); - ubi->lookuptbl[pnum] = e; - } else { - e->ec = fm_e->ec; - kfree(fm_e); - } - - spin_unlock(&ubi->wl_lock); - - vol_id = lnum ? UBI_FM_DATA_VOLUME_ID : UBI_FM_SB_VOLUME_ID; - return schedule_erase(ubi, e, vol_id, lnum, torture); -} -#endif - /** * wear_leveling_worker - wear-leveling worker function. * @ubi: UBI device description object @@ -1002,7 +646,7 @@ static int wear_leveling_worker(struct ubi_device *ubi, struct ubi_work *wrk, int shutdown) { int err, scrubbing = 0, torture = 0, protect = 0, erroneous = 0; - int vol_id = -1, uninitialized_var(lnum); + int vol_id = -1, lnum = -1; #ifdef CONFIG_MTD_UBI_FASTMAP int anchor = wrk->anchor; #endif @@ -1214,7 +858,7 @@ static int wear_leveling_worker(struct ubi_device *ubi, struct ubi_work *wrk, err = do_sync_erase(ubi, e1, vol_id, lnum, 0); if (err) { if (e2) - kmem_cache_free(ubi_wl_entry_slab, e2); + wl_entry_destroy(ubi, e2); goto out_ro; } @@ -1282,8 +926,8 @@ out_error: spin_unlock(&ubi->wl_lock); ubi_free_vid_hdr(ubi, vid_hdr); - kmem_cache_free(ubi_wl_entry_slab, e1); - kmem_cache_free(ubi_wl_entry_slab, e2); + wl_entry_destroy(ubi, e1); + wl_entry_destroy(ubi, e2); out_ro: ubi_ro_mode(ubi); @@ -1369,38 +1013,6 @@ out_unlock: return err; } -#ifdef CONFIG_MTD_UBI_FASTMAP -/** - * ubi_ensure_anchor_pebs - schedule wear-leveling to produce an anchor PEB. - * @ubi: UBI device description object - */ -int ubi_ensure_anchor_pebs(struct ubi_device *ubi) -{ - struct ubi_work *wrk; - - spin_lock(&ubi->wl_lock); - if (ubi->wl_scheduled) { - spin_unlock(&ubi->wl_lock); - return 0; - } - ubi->wl_scheduled = 1; - spin_unlock(&ubi->wl_lock); - - wrk = kmalloc(sizeof(struct ubi_work), GFP_NOFS); - if (!wrk) { - spin_lock(&ubi->wl_lock); - ubi->wl_scheduled = 0; - spin_unlock(&ubi->wl_lock); - return -ENOMEM; - } - - wrk->anchor = 1; - wrk->func = &wear_leveling_worker; - schedule_ubi_work(ubi, wrk); - return 0; -} -#endif - /** * erase_worker - physical eraseblock erase worker function. * @ubi: UBI device description object @@ -1425,15 +1037,13 @@ static int erase_worker(struct ubi_device *ubi, struct ubi_work *wl_wrk, if (shutdown) { dbg_wl("cancel erasure of PEB %d EC %d", pnum, e->ec); kfree(wl_wrk); - kmem_cache_free(ubi_wl_entry_slab, e); + wl_entry_destroy(ubi, e); return 0; } dbg_wl("erase PEB %d EC %d LEB %d:%d", pnum, e->ec, wl_wrk->vol_id, wl_wrk->lnum); - ubi_assert(!ubi_is_fm_block(ubi, e->pnum)); - err = sync_erase(ubi, e, wl_wrk->torture); if (!err) { /* Fine, we've erased it successfully */ @@ -1471,7 +1081,7 @@ static int erase_worker(struct ubi_device *ubi, struct ubi_work *wl_wrk, return err; } - kmem_cache_free(ubi_wl_entry_slab, e); + wl_entry_destroy(ubi, e); if (err != -EIO) /* * If this is not %-EIO, we have no idea what to do. Scheduling @@ -1563,6 +1173,8 @@ int ubi_wl_put_peb(struct ubi_device *ubi, int vol_id, int lnum, ubi_assert(pnum >= 0); ubi_assert(pnum < ubi->peb_count); + down_read(&ubi->fm_protect); + retry: spin_lock(&ubi->wl_lock); e = ubi->lookuptbl[pnum]; @@ -1593,6 +1205,7 @@ retry: ubi_assert(!ubi->move_to_put); ubi->move_to_put = 1; spin_unlock(&ubi->wl_lock); + up_read(&ubi->fm_protect); return 0; } else { if (in_wl_tree(e, &ubi->used)) { @@ -1614,6 +1227,7 @@ retry: ubi_err(ubi, "PEB %d not found", pnum); ubi_ro_mode(ubi); spin_unlock(&ubi->wl_lock); + up_read(&ubi->fm_protect); return err; } } @@ -1627,6 +1241,7 @@ retry: spin_unlock(&ubi->wl_lock); } + up_read(&ubi->fm_protect); return err; } @@ -1758,9 +1373,10 @@ int ubi_wl_flush(struct ubi_device *ubi, int vol_id, int lnum) /** * tree_destroy - destroy an RB-tree. + * @ubi: UBI device description object * @root: the root of the tree to destroy */ -static void tree_destroy(struct rb_root *root) +static void tree_destroy(struct ubi_device *ubi, struct rb_root *root) { struct rb_node *rb; struct ubi_wl_entry *e; @@ -1782,7 +1398,7 @@ static void tree_destroy(struct rb_root *root) rb->rb_right = NULL; } - kmem_cache_free(ubi_wl_entry_slab, e); + wl_entry_destroy(ubi, e); } } } @@ -1850,6 +1466,9 @@ int ubi_thread(void *u) */ static void shutdown_work(struct ubi_device *ubi) { +#ifdef CONFIG_MTD_UBI_FASTMAP + flush_work(&ubi->fm_work); +#endif while (!list_empty(&ubi->works)) { struct ubi_work *wrk; @@ -1883,9 +1502,6 @@ int ubi_wl_init(struct ubi_device *ubi, struct ubi_attach_info *ai) init_rwsem(&ubi->work_sem); ubi->max_ec = ai->max_ec; INIT_LIST_HEAD(&ubi->works); -#ifdef CONFIG_MTD_UBI_FASTMAP - INIT_WORK(&ubi->fm_work, update_fastmap_work_fn); -#endif sprintf(ubi->bgt_name, UBI_BGT_NAME_PATTERN, ubi->ubi_num); @@ -1907,10 +1523,9 @@ int ubi_wl_init(struct ubi_device *ubi, struct ubi_attach_info *ai) e->pnum = aeb->pnum; e->ec = aeb->ec; - ubi_assert(!ubi_is_fm_block(ubi, e->pnum)); ubi->lookuptbl[e->pnum] = e; if (schedule_erase(ubi, e, aeb->vol_id, aeb->lnum, 0)) { - kmem_cache_free(ubi_wl_entry_slab, e); + wl_entry_destroy(ubi, e); goto out_free; } @@ -1928,7 +1543,6 @@ int ubi_wl_init(struct ubi_device *ubi, struct ubi_attach_info *ai) e->pnum = aeb->pnum; e->ec = aeb->ec; ubi_assert(e->ec >= 0); - ubi_assert(!ubi_is_fm_block(ubi, e->pnum)); wl_tree_add(e, &ubi->free); ubi->free_count++; @@ -1966,17 +1580,20 @@ int ubi_wl_init(struct ubi_device *ubi, struct ubi_attach_info *ai) dbg_wl("found %i PEBs", found_pebs); - if (ubi->fm) + if (ubi->fm) { ubi_assert(ubi->good_peb_count == \ found_pebs + ubi->fm->used_blocks); + + for (i = 0; i < ubi->fm->used_blocks; i++) { + e = ubi->fm->e[i]; + ubi->lookuptbl[e->pnum] = e; + } + } else ubi_assert(ubi->good_peb_count == found_pebs); reserved_pebs = WL_RESERVED_PEBS; -#ifdef CONFIG_MTD_UBI_FASTMAP - /* Reserve enough LEBs to store two fastmaps. */ - reserved_pebs += (ubi->fm_size / ubi->leb_size) * 2; -#endif + ubi_fastmap_init(ubi, &reserved_pebs); if (ubi->avail_pebs < reserved_pebs) { ubi_err(ubi, "no enough physical eraseblocks (%d, need %d)", @@ -1998,9 +1615,9 @@ int ubi_wl_init(struct ubi_device *ubi, struct ubi_attach_info *ai) out_free: shutdown_work(ubi); - tree_destroy(&ubi->used); - tree_destroy(&ubi->free); - tree_destroy(&ubi->scrub); + tree_destroy(ubi, &ubi->used); + tree_destroy(ubi, &ubi->free); + tree_destroy(ubi, &ubi->scrub); kfree(ubi->lookuptbl); return err; } @@ -2017,7 +1634,7 @@ static void protection_queue_destroy(struct ubi_device *ubi) for (i = 0; i < UBI_PROT_QUEUE_LEN; ++i) { list_for_each_entry_safe(e, tmp, &ubi->pq[i], u.list) { list_del(&e->u.list); - kmem_cache_free(ubi_wl_entry_slab, e); + wl_entry_destroy(ubi, e); } } } @@ -2029,12 +1646,13 @@ static void protection_queue_destroy(struct ubi_device *ubi) void ubi_wl_close(struct ubi_device *ubi) { dbg_wl("close the WL sub-system"); + ubi_fastmap_close(ubi); shutdown_work(ubi); protection_queue_destroy(ubi); - tree_destroy(&ubi->used); - tree_destroy(&ubi->erroneous); - tree_destroy(&ubi->free); - tree_destroy(&ubi->scrub); + tree_destroy(ubi, &ubi->used); + tree_destroy(ubi, &ubi->erroneous); + tree_destroy(ubi, &ubi->free); + tree_destroy(ubi, &ubi->scrub); kfree(ubi->lookuptbl); } @@ -2133,3 +1751,94 @@ static int self_check_in_pq(const struct ubi_device *ubi, dump_stack(); return -EINVAL; } +#ifndef CONFIG_MTD_UBI_FASTMAP +static struct ubi_wl_entry *get_peb_for_wl(struct ubi_device *ubi) +{ + struct ubi_wl_entry *e; + + e = find_wl_entry(ubi, &ubi->free, WL_FREE_MAX_DIFF); + self_check_in_wl_tree(ubi, e, &ubi->free); + ubi->free_count--; + ubi_assert(ubi->free_count >= 0); + rb_erase(&e->u.rb, &ubi->free); + + return e; +} + +/** + * produce_free_peb - produce a free physical eraseblock. + * @ubi: UBI device description object + * + * This function tries to make a free PEB by means of synchronous execution of + * pending works. This may be needed if, for example the background thread is + * disabled. Returns zero in case of success and a negative error code in case + * of failure. + */ +static int produce_free_peb(struct ubi_device *ubi) +{ + int err; + + while (!ubi->free.rb_node && ubi->works_count) { + spin_unlock(&ubi->wl_lock); + + dbg_wl("do one work synchronously"); + err = do_work(ubi); + + spin_lock(&ubi->wl_lock); + if (err) + return err; + } + + return 0; +} + +/** + * ubi_wl_get_peb - get a physical eraseblock. + * @ubi: UBI device description object + * + * This function returns a physical eraseblock in case of success and a + * negative error code in case of failure. + * Returns with ubi->fm_eba_sem held in read mode! + */ +int ubi_wl_get_peb(struct ubi_device *ubi) +{ + int err; + struct ubi_wl_entry *e; + +retry: + down_read(&ubi->fm_eba_sem); + spin_lock(&ubi->wl_lock); + if (!ubi->free.rb_node) { + if (ubi->works_count == 0) { + ubi_err(ubi, "no free eraseblocks"); + ubi_assert(list_empty(&ubi->works)); + spin_unlock(&ubi->wl_lock); + return -ENOSPC; + } + + err = produce_free_peb(ubi); + if (err < 0) { + spin_unlock(&ubi->wl_lock); + return err; + } + spin_unlock(&ubi->wl_lock); + up_read(&ubi->fm_eba_sem); + goto retry; + + } + e = wl_get_wle(ubi); + prot_queue_add(ubi, e); + spin_unlock(&ubi->wl_lock); + + err = ubi_self_check_all_ff(ubi, e->pnum, ubi->vid_hdr_aloffset, + ubi->peb_size - ubi->vid_hdr_aloffset); + if (err) { + ubi_err(ubi, "new PEB %d does not contain all 0xFF bytes", e->pnum); + return err; + } + + return e->pnum; +} +#else +#include "fastmap-wl.c" +#endif diff --git a/drivers/mtd/ubi/wl.h b/drivers/mtd/ubi/wl.h new file mode 100644 index 000000000000..bd1f07e5ce9a --- /dev/null +++ b/drivers/mtd/ubi/wl.h @@ -0,0 +1,28 @@ +#ifndef UBI_WL_H +#define UBI_WL_H +#ifdef CONFIG_MTD_UBI_FASTMAP +static int anchor_pebs_avalible(struct rb_root *root); +static void update_fastmap_work_fn(struct work_struct *wrk); +static struct ubi_wl_entry *find_anchor_wl_entry(struct rb_root *root); +static struct ubi_wl_entry *get_peb_for_wl(struct ubi_device *ubi); +static void ubi_fastmap_close(struct ubi_device *ubi); +static inline void ubi_fastmap_init(struct ubi_device *ubi, int *count) +{ + /* Reserve enough LEBs to store two fastmaps. */ + *count += (ubi->fm_size / ubi->leb_size) * 2; + INIT_WORK(&ubi->fm_work, update_fastmap_work_fn); +} +static struct ubi_wl_entry *may_reserve_for_fm(struct ubi_device *ubi, + struct ubi_wl_entry *e, + struct rb_root *root); +#else /* !CONFIG_MTD_UBI_FASTMAP */ +static struct ubi_wl_entry *get_peb_for_wl(struct ubi_device *ubi); +static inline void ubi_fastmap_close(struct ubi_device *ubi) { } +static inline void ubi_fastmap_init(struct ubi_device *ubi, int *count) { } +static struct ubi_wl_entry *may_reserve_for_fm(struct ubi_device *ubi, + struct ubi_wl_entry *e, + struct rb_root *root) { + return e; +} +#endif /* CONFIG_MTD_UBI_FASTMAP */ +#endif /* UBI_WL_H */ diff --git a/drivers/net/macvtap.c b/drivers/net/macvtap.c index 8362aef0c15e..9c91ff872485 100644 --- a/drivers/net/macvtap.c +++ b/drivers/net/macvtap.c @@ -1118,8 +1118,6 @@ static const struct file_operations macvtap_fops = { .owner = THIS_MODULE, .open = macvtap_open, .release = macvtap_release, - .read = new_sync_read, - .write = new_sync_write, .read_iter = macvtap_read_iter, .write_iter = macvtap_write_iter, .poll = macvtap_poll, diff --git a/drivers/net/tun.c b/drivers/net/tun.c index b96b94cee760..e470ae59d405 100644 --- a/drivers/net/tun.c +++ b/drivers/net/tun.c @@ -2223,8 +2223,6 @@ static void tun_chr_show_fdinfo(struct seq_file *m, struct file *f) static const struct file_operations tun_fops = { .owner = THIS_MODULE, .llseek = no_llseek, - .read = new_sync_read, - .write = new_sync_write, .read_iter = tun_chr_read_iter, .write_iter = tun_chr_write_iter, .poll = tun_chr_poll, diff --git a/drivers/staging/android/ashmem.c b/drivers/staging/android/ashmem.c index d140b733940c..c5c037ccf32c 100644 --- a/drivers/staging/android/ashmem.c +++ b/drivers/staging/android/ashmem.c @@ -310,7 +310,7 @@ static ssize_t ashmem_read(struct file *file, char __user *buf, * be destroyed until all references to the file are dropped and * ashmem_release is called. */ - ret = asma->file->f_op->read(asma->file, buf, len, pos); + ret = __vfs_read(asma->file, buf, len, pos); if (ret >= 0) { /** Update backing file pos, since f_ops->read() doesn't */ asma->file->f_pos = *pos; diff --git a/drivers/staging/comedi/drivers/serial2002.c b/drivers/staging/comedi/drivers/serial2002.c index ad35ed6e93f0..304ebff119ee 100644 --- a/drivers/staging/comedi/drivers/serial2002.c +++ b/drivers/staging/comedi/drivers/serial2002.c @@ -108,24 +108,16 @@ static int serial2002_tty_write(struct file *f, unsigned char *buf, int count) { const char __user *p = (__force const char __user *)buf; int result; + loff_t offset = 0; mm_segment_t oldfs; oldfs = get_fs(); set_fs(KERNEL_DS); - f->f_pos = 0; - result = f->f_op->write(f, p, count, &f->f_pos); + result = __vfs_write(f, p, count, &offset); set_fs(oldfs); return result; } -static int serial2002_tty_readb(struct file *f, unsigned char *buf) -{ - char __user *p = (__force char __user *)buf; - - f->f_pos = 0; - return f->f_op->read(f, p, 1, &f->f_pos); -} - static void serial2002_tty_read_poll_wait(struct file *f, int timeout) { struct poll_wqueues table; @@ -161,13 +153,15 @@ static int serial2002_tty_read(struct file *f, int timeout) result = -1; if (!IS_ERR(f)) { mm_segment_t oldfs; + char __user *p = (__force char __user *)&ch; + loff_t offset = 0; oldfs = get_fs(); set_fs(KERNEL_DS); if (f->f_op->poll) { serial2002_tty_read_poll_wait(f, timeout); - if (serial2002_tty_readb(f, &ch) == 1) + if (__vfs_read(f, p, 1, &offset) == 1) result = ch; } else { /* Device does not support poll, busy wait */ @@ -178,7 +172,7 @@ static int serial2002_tty_read(struct file *f, int timeout) if (retries >= timeout) break; - if (serial2002_tty_readb(f, &ch) == 1) { + if (__vfs_read(f, p, 1, &offset) == 1) { result = ch; break; } diff --git a/drivers/staging/lustre/lustre/llite/file.c b/drivers/staging/lustre/lustre/llite/file.c index 85e74d18d1c7..529062ea112b 100644 --- a/drivers/staging/lustre/lustre/llite/file.c +++ b/drivers/staging/lustre/lustre/llite/file.c @@ -3135,9 +3135,7 @@ int ll_inode_permission(struct inode *inode, int mask) /* -o localflock - only provides locally consistent flock locks */ struct file_operations ll_file_operations = { - .read = new_sync_read, .read_iter = ll_file_read_iter, - .write = new_sync_write, .write_iter = ll_file_write_iter, .unlocked_ioctl = ll_file_ioctl, .open = ll_file_open, @@ -3150,9 +3148,7 @@ struct file_operations ll_file_operations = { }; struct file_operations ll_file_operations_flock = { - .read = new_sync_read, .read_iter = ll_file_read_iter, - .write = new_sync_write, .write_iter = ll_file_write_iter, .unlocked_ioctl = ll_file_ioctl, .open = ll_file_open, @@ -3168,9 +3164,7 @@ struct file_operations ll_file_operations_flock = { /* These are for -o noflock - to return ENOSYS on flock calls */ struct file_operations ll_file_operations_noflock = { - .read = new_sync_read, .read_iter = ll_file_read_iter, - .write = new_sync_write, .write_iter = ll_file_write_iter, .unlocked_ioctl = ll_file_ioctl, .open = ll_file_open, diff --git a/drivers/staging/lustre/lustre/llite/llite_internal.h b/drivers/staging/lustre/lustre/llite/llite_internal.h index 2af1d7286250..e7422f5c9c6f 100644 --- a/drivers/staging/lustre/lustre/llite/llite_internal.h +++ b/drivers/staging/lustre/lustre/llite/llite_internal.h @@ -938,10 +938,8 @@ struct ll_cl_context { }; struct vvp_thread_info { - struct iovec vti_local_iov; struct vvp_io_args vti_args; struct ra_io_arg vti_ria; - struct kiocb vti_kiocb; struct ll_cl_context vti_io_ctx; }; diff --git a/drivers/usb/gadget/function/f_fs.c b/drivers/usb/gadget/function/f_fs.c index a12315a78248..6bdb57069044 100644 --- a/drivers/usb/gadget/function/f_fs.c +++ b/drivers/usb/gadget/function/f_fs.c @@ -1061,8 +1061,6 @@ static const struct file_operations ffs_epfile_operations = { .llseek = no_llseek, .open = ffs_epfile_open, - .write = new_sync_write, - .read = new_sync_read, .write_iter = ffs_epfile_write_iter, .read_iter = ffs_epfile_read_iter, .release = ffs_epfile_release, diff --git a/drivers/usb/gadget/legacy/inode.c b/drivers/usb/gadget/legacy/inode.c index 662ef2c1c62b..6af58c6dba5e 100644 --- a/drivers/usb/gadget/legacy/inode.c +++ b/drivers/usb/gadget/legacy/inode.c @@ -699,8 +699,6 @@ static const struct file_operations ep_io_operations = { .open = ep_open, .release = ep_release, .llseek = no_llseek, - .read = new_sync_read, - .write = new_sync_write, .unlocked_ioctl = ep_ioctl, .read_iter = ep_read_iter, .write_iter = ep_write_iter, |