Daniel,
On 02/03/2016 04:22 AM, Daniel Thompson wrote:
> On 01/02/16 22:41, Joshua Henderson wrote:
>> Add support for the hardware pseudo and true random number generator
>> peripheral found on PIC32.
>>
>> Signed-off-by: Joshua Henderson <joshua.henderson@xxxxxxxxxxxxx>
>> Signed-off-by: Purna Chandra Mandal <purna.mandal@xxxxxxxxxxxxx>
>> ---
>> drivers/char/hw_random/Kconfig | 13 +++
>> drivers/char/hw_random/Makefile | 1 +
>> drivers/char/hw_random/pic32-rng.c | 152 ++++++++++++++++++++++++++++++++++++
>> 3 files changed, 166 insertions(+)
>> create mode 100644 drivers/char/hw_random/pic32-rng.c
>>
>> diff --git a/drivers/char/hw_random/Kconfig b/drivers/char/hw_random/Kconfig
>> index dbf2271..3ab0c46 100644
>> --- a/drivers/char/hw_random/Kconfig
>> +++ b/drivers/char/hw_random/Kconfig
>> @@ -381,6 +381,19 @@ config HW_RANDOM_STM32
>>
>> If unsure, say N.
>>
>> +config HW_RANDOM_PIC32
>> + tristate "Microchip PIC32 Random Number Generator support"
>> + depends on HW_RANDOM && MACH_PIC32
>> + default y
>> + ---help---
>> + This driver provides kernel-side support for the Random Number
>> + Generator hardware found on a PIC32.
>> +
>> + To compile this driver as a module, choose M here. the
>> + module will be called pic32-rng.
>> +
>> + If unsure, say Y.
>> +
>> endif # HW_RANDOM
>>
>> config UML_RANDOM
>> diff --git a/drivers/char/hw_random/Makefile b/drivers/char/hw_random/Makefile
>> index 5ad3976..f5a6fa7 100644
>> --- a/drivers/char/hw_random/Makefile
>> +++ b/drivers/char/hw_random/Makefile
>> @@ -33,3 +33,4 @@ obj-$(CONFIG_HW_RANDOM_MSM) += msm-rng.o
>> obj-$(CONFIG_HW_RANDOM_ST) += st-rng.o
>> obj-$(CONFIG_HW_RANDOM_XGENE) += xgene-rng.o
>> obj-$(CONFIG_HW_RANDOM_STM32) += stm32-rng.o
>> +obj-$(CONFIG_HW_RANDOM_PIC32) += pic32-rng.o
>> diff --git a/drivers/char/hw_random/pic32-rng.c b/drivers/char/hw_random/pic32-rng.c
>> new file mode 100644
>> index 0000000..d2f32c4
>> --- /dev/null
>> +++ b/drivers/char/hw_random/pic32-rng.c
>> @@ -0,0 +1,152 @@
>> +/*
>> + * PIC32 RNG driver
>> + *
>> + * Joshua Henderson <joshua.henderson@xxxxxxxxxxxxx>
>> + * Copyright (C) 2016 Microchip Technology Inc. All rights reserved.
>> + *
>> + * This program is free software; you can distribute it and/or modify it
>> + * under the terms of the GNU General Public License (Version 2) as
>> + * published by the Free Software Foundation.
>> + *
>> + * This program is distributed in the hope 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/module.h>
>> +#include <linux/delay.h>
>> +#include <linux/slab.h>
>> +#include <linux/err.h>
>> +#include <linux/io.h>
>> +#include <linux/hw_random.h>
>> +#include <linux/platform_device.h>
>> +#include <linux/of.h>
>> +#include <linux/of_device.h>
>> +#include <linux/clk.h>
>> +#include <linux/clkdev.h>
>> +
>> +#define RNGCON 0x04
>> +#define TRNGEN BIT(8)
>> +#define PRNGEN BIT(9)
>> +#define PRNGCONT BIT(10)
>> +#define TRNGMOD BIT(11)
>> +#define SEEDLOAD BIT(12)
>> +#define RNGPOLY1 0x08
>> +#define RNGPOLY2 0x0C
>> +#define RNGNUMGEN1 0x10
>> +#define RNGNUMGEN2 0x14
>> +#define RNGSEED1 0x18
>> +#define RNGSEED2 0x1C
>> +#define RNGRCNT 0x20
>> +#define RCNT_MASK 0x7F
>> +
>> +struct pic32_rng {
>> + void __iomem *base;
>> + struct hwrng rng;
>> + struct clk *clk;
>> +};
>> +
>> +static int pic32_rng_read(struct hwrng *rng, void *buf, size_t max,
>> + bool wait)
>> +{
>> + struct pic32_rng *prng = container_of(rng, struct pic32_rng, rng);
>> + u64 *data = buf;
>> +
>> + *data = ((u64)readl(prng->base + RNGNUMGEN2) << 32) +
>> + readl(prng->base + RNGNUMGEN1);
>> + return 4;
>> +}
>
> I have a number of questions at this point (although #3 is by far the most important).
>
> 1. If the random number of 42-bit shouldn't this be return 5; ?
You are right. However, this will essentially be changed to 8 in response to moving to just the TRNG in comments below.
>
> 2. If you really do mean return 4; then why mess about with all the
> shifting. You can just read and discard the upper 32-bits.
>
> 3. Why are you using the PRNG at all? The PRNG is a simple LSFR and
> I don't think it ever gets reseeded at any point (meaning we get
> almost no useful entropy from it at all).
>
> Far better to read directly from the TRNG.
Makes sense. Typically we just use the TRNG to seed the LSFR, but I suppose there is no real reason to do this if max speed is not a concern (above 24Mbps which is max TRNG speed). I'll convert the driver to just drop usage of the PRNG and only use the TRNG.
>
>
>> +
>> +static int pic32_rng_probe(struct platform_device *pdev)
>> +{
>> + struct pic32_rng *prng;
>> + struct resource *res;
>> + u32 v, t;
>> + int ret;
>> +
>> + prng = devm_kzalloc(&pdev->dev, sizeof(*prng), GFP_KERNEL);
>> + if (!prng)
>> + return -ENOMEM;
>> +
>> + res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
>> + prng->base = devm_ioremap_resource(&pdev->dev, res);
>> + if (IS_ERR(prng->base))
>> + return PTR_ERR(prng->base);
>> +
>> + prng->clk = devm_clk_get(&pdev->dev, NULL);
>> + if (IS_ERR(prng->clk))
>> + return PTR_ERR(prng->clk);
>> +
>> + clk_prepare_enable(prng->clk);
>> +
>> + /* enable TRNG in enhanced mode */
>> + v = readl(prng->base + RNGCON);
>> + v &= ~(TRNGEN | PRNGEN | 0xff);
>> + v |= TRNGMOD;
>> + writel(v | TRNGEN, prng->base + RNGCON);
>> +
>> + /* wait for valid seed */
>> + usleep_range(100, 200);
>> + t = readl(prng->base + RNGRCNT) & RCNT_MASK;
>> + if (t < 0x2A)
>> + dev_warn(&pdev->dev, "seed not generated.\n");
>> +
>> + /* load initial seed */
>> + writel(v | SEEDLOAD, prng->base + RNGCON);
>> +
>> + /* load initial polynomial: 42bit poly */
>> + writel(0x00c00003, prng->base + RNGPOLY1);
>> + writel(0x00000000, prng->base + RNGPOLY2);
>> +
>> + /* start PRNG to generate 42bit random */
>> + v |= 0x2A | PRNGCONT | PRNGEN;
>> + writel(v, prng->base + RNGCON);
>> +
>> + prng->rng.name = pdev->name;
>> + prng->rng.read = pic32_rng_read;
>> +
>> + ret = hwrng_register(&prng->rng);
>> + if (ret)
>> + goto err_register;
>> +
>> + platform_set_drvdata(pdev, prng);
>> +
>> + return 0;
>> +
>> +err_register:
>> + return ret;
>
> Looks like we leak clock references on this error path.
Good catch. Will fix.
>
>
>> +
>> +static int pic32_rng_remove(struct platform_device *pdev)
>> +{
>> + struct pic32_rng *rng = platform_get_drvdata(pdev);
>> +
>> + hwrng_unregister(&rng->rng);
>> + writel(0, rng->base + RNGCON);
>> + clk_disable_unprepare(rng->clk);
>> + return 0;
>> +}
>> +
>> +static const struct of_device_id pic32_rng_of_match[] = {
>> + { .compatible = "microchip,pic32mzda-rng", },
>> + { /* sentinel */ }
>> +};
>> +MODULE_DEVICE_TABLE(of, pic32_rng_of_match);
>> +
>> +static struct platform_driver pic32_rng_driver = {
>> + .probe = pic32_rng_probe,
>> + .remove = pic32_rng_remove,
>> + .driver = {
>> + .name = "pic32-rng",
>> + .owner = THIS_MODULE,
>> + .of_match_table = of_match_ptr(pic32_rng_of_match),
>> + },
>> +};
>> +
>> +module_platform_driver(pic32_rng_driver);
>> +
>> +MODULE_LICENSE("GPL");
>> +MODULE_AUTHOR("Joshua Henderson <joshua.henderson@xxxxxxxxxxxxx>");
>> +MODULE_DESCRIPTION("Microchip PIC32 RNG Driver");
>>
>
After a bit of testing, converting the driver to be pure TRNG does give good results with rngtest. Good call. Will submit an update.
Thanks,
Josh
