* [PATCH 0/3] MT67XX random number generator support @ 2019-05-08 3:58 neal.liu 2019-05-08 3:58 ` [PATCH 1/3] soc: mediatek: add SMC fid table for SIP interface neal.liu ` (2 more replies) 0 siblings, 3 replies; 12+ messages in thread From: neal.liu @ 2019-05-08 3:58 UTC (permalink / raw) To: mpm, herbert, robh+dt, mark.rutland, matthias.bgg Cc: wsd_upstream, linux-crypto, linux-kernel, devicetree, linux-mediatek These patch series introduce a generic rng driver for Trustzone based kernel driver which would like to communicate with ATF SIP services. Patch #1 initials SMC fid table for Mediatek SIP interfaces and adds HWRNG related SMC call. Patch #2..3 adds mt67xx-rng kernel driver for Trustzone based SoCs. For Mediatek SoCs on ARMv8 with TrustZone enabled, peripherals like entropy sources is not accessible from normal world (linux) and rather accessible from secure world (ATF/TEE) only. This driver aims to provide a generic interface to ATF rng service. Neal Liu (3): soc: mediatek: add SMC fid table for SIP interface dt-bindings: rng: update bindings for MT67xx SoCs hwrng: add mt67xx-rng driver Documentation/devicetree/bindings/rng/mtk-rng.txt | 13 ++- drivers/char/hw_random/Kconfig | 16 ++++ drivers/char/hw_random/Makefile | 1 + drivers/char/hw_random/mt67xx-rng.c | 104 +++++++++++++++++++++ include/linux/soc/mediatek/mtk_sip_svc.h | 55 +++++++++++ 5 files changed, 186 insertions(+), 3 deletions(-) create mode 100644 drivers/char/hw_random/mt67xx-rng.c create mode 100644 include/linux/soc/mediatek/mtk_sip_svc.h -- 1.7.9.5 ^ permalink raw reply [flat|nested] 12+ messages in thread
* [PATCH 1/3] soc: mediatek: add SMC fid table for SIP interface 2019-05-08 3:58 [PATCH 0/3] MT67XX random number generator support neal.liu @ 2019-05-08 3:58 ` neal.liu 2019-05-08 3:58 ` [PATCH 2/3] dt-bindings: rng: update bindings for MT67xx SoCs neal.liu 2019-05-08 3:58 ` [PATCH 3/3] hwrng: add mt67xx-rng driver neal.liu 2 siblings, 0 replies; 12+ messages in thread From: neal.liu @ 2019-05-08 3:58 UTC (permalink / raw) To: mpm, herbert, robh+dt, mark.rutland, matthias.bgg Cc: wsd_upstream, linux-crypto, linux-kernel, devicetree, linux-mediatek, Neal Liu From: Neal Liu <neal.liu@mediatek.com> 1. Add a header file to provide SIP interface to ARM Trusted Firmware(ATF) 2. Add hwrng SMC fid Signed-off-by: Neal Liu <neal.liu@mediatek.com> --- include/linux/soc/mediatek/mtk_sip_svc.h | 55 ++++++++++++++++++++++++++++++ 1 file changed, 55 insertions(+) create mode 100644 include/linux/soc/mediatek/mtk_sip_svc.h diff --git a/include/linux/soc/mediatek/mtk_sip_svc.h b/include/linux/soc/mediatek/mtk_sip_svc.h new file mode 100644 index 0000000..ac73f68 --- /dev/null +++ b/include/linux/soc/mediatek/mtk_sip_svc.h @@ -0,0 +1,55 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * Copyright (c) 2019 MediaTek Inc. + */ + +#ifndef _MTK_SECURE_API_H_ +#define _MTK_SECURE_API_H_ + +#include <linux/kernel.h> + + +/* Error Code */ +#define SIP_SVC_E_SUCCESS 0 +#define SIP_SVC_E_NOT_SUPPORTED -1 +#define SIP_SVC_E_INVALID_PARAMS -2 +#define SIP_SVC_E_INVALID_RANGE -3 +#define SIP_SVC_E_PERMISSION_DENY -4 + +#ifdef CONFIG_ARM64 +#define MTK_SIP_SMC_AARCH_BIT 0x40000000 +#else +#define MTK_SIP_SMC_AARCH_BIT 0x00000000 +#endif + + +/******************************************************************************* + * Defines for Mediatek runtime services func ids + ******************************************************************************/ + +/* Debug feature and ATF related SMC call */ + +/* CPU operations related SMC call */ + +/* SPM related SMC call */ + +/* Low power related SMC call */ + +/* AMMS related SMC call */ + +/* Security related SMC call */ +/* HWRNG */ +#define MTK_SIP_KERNEL_GET_RND \ + (0x82000206 | MTK_SIP_SMC_AARCH_BIT) + +/* Storage Encryption related SMC call */ + +/* Platform related SMC call */ + +/* Pheripheral related SMC call */ + +/* MM related SMC call */ + + +#endif /* _MTK_SECURE_API_H_ */ + -- 1.7.9.5 ^ permalink raw reply related [flat|nested] 12+ messages in thread
* [PATCH 2/3] dt-bindings: rng: update bindings for MT67xx SoCs 2019-05-08 3:58 [PATCH 0/3] MT67XX random number generator support neal.liu 2019-05-08 3:58 ` [PATCH 1/3] soc: mediatek: add SMC fid table for SIP interface neal.liu @ 2019-05-08 3:58 ` neal.liu 2019-05-08 3:58 ` [PATCH 3/3] hwrng: add mt67xx-rng driver neal.liu 2 siblings, 0 replies; 12+ messages in thread From: neal.liu @ 2019-05-08 3:58 UTC (permalink / raw) To: mpm, herbert, robh+dt, mark.rutland, matthias.bgg Cc: wsd_upstream, linux-crypto, linux-kernel, devicetree, linux-mediatek, Neal Liu From: Neal Liu <neal.liu@mediatek.com> Document the binding used by the Mediatek MT67xx SoCs random number generator with TrustZone enabled. Signed-off-by: Neal Liu <neal.liu@mediatek.com> --- Documentation/devicetree/bindings/rng/mtk-rng.txt | 13 ++++++++++--- 1 file changed, 10 insertions(+), 3 deletions(-) diff --git a/Documentation/devicetree/bindings/rng/mtk-rng.txt b/Documentation/devicetree/bindings/rng/mtk-rng.txt index 2bc89f1..5f40316 100644 --- a/Documentation/devicetree/bindings/rng/mtk-rng.txt +++ b/Documentation/devicetree/bindings/rng/mtk-rng.txt @@ -3,9 +3,12 @@ found in MediaTek SoC family Required properties: - compatible : Should be - "mediatek,mt7622-rng", "mediatek,mt7623-rng" : for MT7622 - "mediatek,mt7629-rng", "mediatek,mt7623-rng" : for MT7629 - "mediatek,mt7623-rng" : for MT7623 + "mediatek,mt7622-rng", "mediatek,mt7623-rng" for MT7622 + "mediatek,mt7629-rng", "mediatek,mt7623-rng" for MT7629 + "mediatek,mt7623-rng" for MT7623 + "mediatek,mt67xx-rng" for MT67xx ARMv8 SoCs + +Optional properties: - clocks : list of clock specifiers, corresponding to entries in clock-names property; - clock-names : Should contain "rng" entries; @@ -19,3 +22,7 @@ rng: rng@1020f000 { clocks = <&infracfg CLK_INFRA_TRNG>; clock-names = "rng"; }; + +hwrng: hwrng { + compatible = "mediatek,mt67xx-rng"; +}; -- 1.7.9.5 ^ permalink raw reply related [flat|nested] 12+ messages in thread
* [PATCH 3/3] hwrng: add mt67xx-rng driver 2019-05-08 3:58 [PATCH 0/3] MT67XX random number generator support neal.liu 2019-05-08 3:58 ` [PATCH 1/3] soc: mediatek: add SMC fid table for SIP interface neal.liu 2019-05-08 3:58 ` [PATCH 2/3] dt-bindings: rng: update bindings for MT67xx SoCs neal.liu @ 2019-05-08 3:58 ` neal.liu [not found] ` <1557287937-2410-4-git-send-email-neal.liu-NuS5LvNUpcJWk0Htik3J/w@public.gmane.org> 2019-05-24 19:00 ` Sean Wang 2 siblings, 2 replies; 12+ messages in thread From: neal.liu @ 2019-05-08 3:58 UTC (permalink / raw) To: mpm, herbert, robh+dt, mark.rutland, matthias.bgg Cc: wsd_upstream, linux-crypto, linux-kernel, devicetree, linux-mediatek, Neal Liu From: Neal Liu <neal.liu@mediatek.com> For Mediatek SoCs on ARMv8 with TrustZone enabled, peripherals like entropy sources is not accessible from normal world (linux) and rather accessible from secure world (ATF/TEE) only. This driver aims to provide a generic interface to ATF rng service. Signed-off-by: Neal Liu <neal.liu@mediatek.com> --- drivers/char/hw_random/Kconfig | 16 ++++++ drivers/char/hw_random/Makefile | 1 + drivers/char/hw_random/mt67xx-rng.c | 104 +++++++++++++++++++++++++++++++++++ 3 files changed, 121 insertions(+) create mode 100644 drivers/char/hw_random/mt67xx-rng.c diff --git a/drivers/char/hw_random/Kconfig b/drivers/char/hw_random/Kconfig index 25a7d8f..98751d3 100644 --- a/drivers/char/hw_random/Kconfig +++ b/drivers/char/hw_random/Kconfig @@ -398,6 +398,22 @@ config HW_RANDOM_MTK If unsure, say Y. +config HW_RANDOM_MT67XX + tristate "Mediatek MT67XX Random Number Generator support" + depends on HW_RANDOM + depends on ARCH_MEDIATEK || COMPILE_TEST + default HW_RANDOM + help + This driver provides kernel-side support for the Random Number + Generator hardware found on Mediatek MT67xx SoCs. The difference + with mtk-rng is the Random Number Generator hardware is secure + access only. + + To compile this driver as a module, choose M here. the + module will be called mt67xx-rng. + + If unsure, say Y. + config HW_RANDOM_S390 tristate "S390 True Random Number Generator support" depends on S390 diff --git a/drivers/char/hw_random/Makefile b/drivers/char/hw_random/Makefile index 7c9ef4a..4be95ab 100644 --- a/drivers/char/hw_random/Makefile +++ b/drivers/char/hw_random/Makefile @@ -36,6 +36,7 @@ obj-$(CONFIG_HW_RANDOM_PIC32) += pic32-rng.o obj-$(CONFIG_HW_RANDOM_MESON) += meson-rng.o obj-$(CONFIG_HW_RANDOM_CAVIUM) += cavium-rng.o cavium-rng-vf.o obj-$(CONFIG_HW_RANDOM_MTK) += mtk-rng.o +obj-$(CONFIG_HW_RANDOM_MT67XX) += mt67xx-rng.o obj-$(CONFIG_HW_RANDOM_S390) += s390-trng.o obj-$(CONFIG_HW_RANDOM_KEYSTONE) += ks-sa-rng.o obj-$(CONFIG_HW_RANDOM_OPTEE) += optee-rng.o diff --git a/drivers/char/hw_random/mt67xx-rng.c b/drivers/char/hw_random/mt67xx-rng.c new file mode 100644 index 0000000..e70cbbe --- /dev/null +++ b/drivers/char/hw_random/mt67xx-rng.c @@ -0,0 +1,104 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Copyright (C) 2019 MediaTek Inc. + */ + +#include <linux/module.h> +#include <linux/platform_device.h> +#include <linux/hw_random.h> +#include <linux/of.h> +#include <linux/arm-smccc.h> +#include <linux/soc/mediatek/mtk_sip_svc.h> + +#define PFX KBUILD_MODNAME ": " +#define MT67XX_RNG_MAGIC 0x74726e67 +#define SMC_RET_NUM 4 + +struct mt67xx_rng_priv { + struct hwrng rng; +}; + + +static void __rng_sec_read(uint32_t *val) +{ + struct arm_smccc_res res; + + arm_smccc_smc(MTK_SIP_KERNEL_GET_RND, + MT67XX_RNG_MAGIC, 0, 0, 0, 0, 0, 0, &res); + + val[0] = res.a0; + val[1] = res.a1; + val[2] = res.a2; + val[3] = res.a3; +} + +static int mt67xx_rng_read(struct hwrng *rng, void *buf, size_t max, bool wait) +{ + int i, retval = 0; + uint32_t val[4] = {0}; + size_t get_rnd_size = sizeof(u32) * SMC_RET_NUM; + + if (!buf) { + pr_err("%s, buf is NULL\n", __func__); + return -EFAULT; + } + + while (max >= get_rnd_size) { + __rng_sec_read(val); + + for (i = 0; i < SMC_RET_NUM; i++) { + *(u32 *)buf = val[i]; + buf += sizeof(u32); + } + + retval += get_rnd_size; + max -= get_rnd_size; + } + + return retval; +} + +static int mt67xx_rng_probe(struct platform_device *pdev) +{ + int ret; + struct mt67xx_rng_priv *priv; + + pr_info(PFX "driver registered\n"); + priv = devm_kzalloc(&pdev->dev, sizeof(*priv), GFP_KERNEL); + if (!priv) + return -ENOMEM; + + priv->rng.name = KBUILD_MODNAME; + priv->rng.read = mt67xx_rng_read; + priv->rng.priv = (unsigned long)&pdev->dev; + priv->rng.quality = 900; + + ret = devm_hwrng_register(&pdev->dev, &priv->rng); + if (ret) { + dev_err(&pdev->dev, "failed to register rng device: %d\n", ret); + return ret; + } + + return 0; +} + +static const struct of_device_id mt67xx_rng_match[] = { + { .compatible = "mediatek,mt67xx-rng", }, + {} +}; +MODULE_DEVICE_TABLE(of, mt67xx_rng_match); + +static struct platform_driver mt67xx_rng_driver = { + .probe = mt67xx_rng_probe, + .driver = { + .name = KBUILD_MODNAME, + .owner = THIS_MODULE, + .of_match_table = mt67xx_rng_match, + }, +}; + +module_platform_driver(mt67xx_rng_driver); + +MODULE_DESCRIPTION("Mediatek MT67XX Random Number Generator Driver"); +MODULE_AUTHOR("Neal Liu <neal.liu@mediatek.com>"); +MODULE_LICENSE("GPL"); -- 1.7.9.5 ^ permalink raw reply related [flat|nested] 12+ messages in thread
[parent not found: <1557287937-2410-4-git-send-email-neal.liu-NuS5LvNUpcJWk0Htik3J/w@public.gmane.org>]
* Re: [PATCH 3/3] hwrng: add mt67xx-rng driver [not found] ` <1557287937-2410-4-git-send-email-neal.liu-NuS5LvNUpcJWk0Htik3J/w@public.gmane.org> @ 2019-05-08 6:34 ` Stephan Mueller 2019-05-08 10:35 ` Neal Liu 0 siblings, 1 reply; 12+ messages in thread From: Stephan Mueller @ 2019-05-08 6:34 UTC (permalink / raw) To: neal.liu-NuS5LvNUpcJWk0Htik3J/w Cc: mark.rutland-5wv7dgnIgG8, devicetree-u79uwXL29TY76Z2rM5mHXA, herbert-lOAM2aK0SrRLBo1qDEOMRrpzq4S04n8Q, wsd_upstream-NuS5LvNUpcJWk0Htik3J/w, linux-kernel-u79uwXL29TY76Z2rM5mHXA, robh+dt-DgEjT+Ai2ygdnm+yROfE0A, linux-mediatek-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r, linux-crypto-u79uwXL29TY76Z2rM5mHXA, mpm-VDJrAJ4Gl5ZBDgjK7y7TUQ, matthias.bgg-Re5JQEeQqe8AvxtiuMwx3w Am Mittwoch, 8. Mai 2019, 05:58:57 CEST schrieb neal.liu-NuS5LvNUpcJWk0Htik3J/w@public.gmane.org: Hi liu, > From: Neal Liu <neal.liu-NuS5LvNUpcJWk0Htik3J/w@public.gmane.org> > > For Mediatek SoCs on ARMv8 with TrustZone enabled, peripherals like > entropy sources is not accessible from normal world (linux) and > rather accessible from secure world (ATF/TEE) only. This driver aims > to provide a generic interface to ATF rng service. > > Signed-off-by: Neal Liu <neal.liu-NuS5LvNUpcJWk0Htik3J/w@public.gmane.org> > --- > drivers/char/hw_random/Kconfig | 16 ++++++ > drivers/char/hw_random/Makefile | 1 + > drivers/char/hw_random/mt67xx-rng.c | 104 > +++++++++++++++++++++++++++++++++++ 3 files changed, 121 insertions(+) > create mode 100644 drivers/char/hw_random/mt67xx-rng.c > > diff --git a/drivers/char/hw_random/Kconfig b/drivers/char/hw_random/Kconfig > index 25a7d8f..98751d3 100644 > --- a/drivers/char/hw_random/Kconfig > +++ b/drivers/char/hw_random/Kconfig > @@ -398,6 +398,22 @@ config HW_RANDOM_MTK > > If unsure, say Y. > > +config HW_RANDOM_MT67XX > + tristate "Mediatek MT67XX Random Number Generator support" > + depends on HW_RANDOM > + depends on ARCH_MEDIATEK || COMPILE_TEST > + default HW_RANDOM > + help > + This driver provides kernel-side support for the Random Number > + Generator hardware found on Mediatek MT67xx SoCs. The difference > + with mtk-rng is the Random Number Generator hardware is secure > + access only. > + > + To compile this driver as a module, choose M here. the > + module will be called mt67xx-rng. > + > + If unsure, say Y. > + > config HW_RANDOM_S390 > tristate "S390 True Random Number Generator support" > depends on S390 > diff --git a/drivers/char/hw_random/Makefile > b/drivers/char/hw_random/Makefile index 7c9ef4a..4be95ab 100644 > --- a/drivers/char/hw_random/Makefile > +++ b/drivers/char/hw_random/Makefile > @@ -36,6 +36,7 @@ obj-$(CONFIG_HW_RANDOM_PIC32) += pic32-rng.o > obj-$(CONFIG_HW_RANDOM_MESON) += meson-rng.o > obj-$(CONFIG_HW_RANDOM_CAVIUM) += cavium-rng.o cavium-rng-vf.o > obj-$(CONFIG_HW_RANDOM_MTK) += mtk-rng.o > +obj-$(CONFIG_HW_RANDOM_MT67XX) += mt67xx-rng.o > obj-$(CONFIG_HW_RANDOM_S390) += s390-trng.o > obj-$(CONFIG_HW_RANDOM_KEYSTONE) += ks-sa-rng.o > obj-$(CONFIG_HW_RANDOM_OPTEE) += optee-rng.o > diff --git a/drivers/char/hw_random/mt67xx-rng.c > b/drivers/char/hw_random/mt67xx-rng.c new file mode 100644 > index 0000000..e70cbbe > --- /dev/null > +++ b/drivers/char/hw_random/mt67xx-rng.c > @@ -0,0 +1,104 @@ > +// SPDX-License-Identifier: GPL-2.0 > +/* > + * Copyright (C) 2019 MediaTek Inc. > + */ > + > +#include <linux/module.h> > +#include <linux/platform_device.h> > +#include <linux/hw_random.h> > +#include <linux/of.h> > +#include <linux/arm-smccc.h> > +#include <linux/soc/mediatek/mtk_sip_svc.h> > + > +#define PFX KBUILD_MODNAME ": " > +#define MT67XX_RNG_MAGIC 0x74726e67 > +#define SMC_RET_NUM 4 > + > +struct mt67xx_rng_priv { > + struct hwrng rng; > +}; > + > + > +static void __rng_sec_read(uint32_t *val) > +{ > + struct arm_smccc_res res; > + > + arm_smccc_smc(MTK_SIP_KERNEL_GET_RND, > + MT67XX_RNG_MAGIC, 0, 0, 0, 0, 0, 0, &res); > + > + val[0] = res.a0; > + val[1] = res.a1; > + val[2] = res.a2; > + val[3] = res.a3; > +} > + > +static int mt67xx_rng_read(struct hwrng *rng, void *buf, size_t max, bool > wait) +{ > + int i, retval = 0; > + uint32_t val[4] = {0}; > + size_t get_rnd_size = sizeof(u32) * SMC_RET_NUM; > + > + if (!buf) { > + pr_err("%s, buf is NULL\n", __func__); > + return -EFAULT; > + } > + > + while (max >= get_rnd_size) { > + __rng_sec_read(val); > + > + for (i = 0; i < SMC_RET_NUM; i++) { > + *(u32 *)buf = val[i]; I am not sure this cast is right - or how is it guaranteed that buf is word- aligned? > + buf += sizeof(u32); > + } > + > + retval += get_rnd_size; > + max -= get_rnd_size; > + } > + > + return retval; > +} > + > +static int mt67xx_rng_probe(struct platform_device *pdev) > +{ > + int ret; > + struct mt67xx_rng_priv *priv; > + > + pr_info(PFX "driver registered\n"); > + priv = devm_kzalloc(&pdev->dev, sizeof(*priv), GFP_KERNEL); > + if (!priv) > + return -ENOMEM; > + > + priv->rng.name = KBUILD_MODNAME; > + priv->rng.read = mt67xx_rng_read; > + priv->rng.priv = (unsigned long)&pdev->dev; > + priv->rng.quality = 900; > + > + ret = devm_hwrng_register(&pdev->dev, &priv->rng); > + if (ret) { > + dev_err(&pdev->dev, "failed to register rng device: %d\n", ret); > + return ret; > + } > + > + return 0; > +} > + > +static const struct of_device_id mt67xx_rng_match[] = { > + { .compatible = "mediatek,mt67xx-rng", }, > + {} > +}; > +MODULE_DEVICE_TABLE(of, mt67xx_rng_match); > + > +static struct platform_driver mt67xx_rng_driver = { > + .probe = mt67xx_rng_probe, > + .driver = { > + .name = KBUILD_MODNAME, > + .owner = THIS_MODULE, > + .of_match_table = mt67xx_rng_match, > + }, > +}; > + > +module_platform_driver(mt67xx_rng_driver); > + > +MODULE_DESCRIPTION("Mediatek MT67XX Random Number Generator Driver"); > +MODULE_AUTHOR("Neal Liu <neal.liu-NuS5LvNUpcJWk0Htik3J/w@public.gmane.org>"); > +MODULE_LICENSE("GPL"); Ciao Stephan ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [PATCH 3/3] hwrng: add mt67xx-rng driver 2019-05-08 6:34 ` Stephan Mueller @ 2019-05-08 10:35 ` Neal Liu 2019-05-09 5:26 ` Herbert Xu 0 siblings, 1 reply; 12+ messages in thread From: Neal Liu @ 2019-05-08 10:35 UTC (permalink / raw) To: Stephan Mueller Cc: mpm, herbert, robh+dt, mark.rutland, matthias.bgg, wsd_upstream, linux-crypto, linux-kernel, devicetree, linux-mediatek, Crystal.Guo, neal.liu Hi Stephan, We think the cast is fine, and it cannot guarantee the buf is word-align. I reference multiple rng driver's implementation and found it's common usage for this. So it might be general usage for community. Is there any suggestion that is more appropriate? Thanks Best Regards, -Neal Liu On Wed, 2019-05-08 at 08:34 +0200, Stephan Mueller wrote: > Am Mittwoch, 8. Mai 2019, 05:58:57 CEST schrieb neal.liu@mediatek.com: > > Hi liu, > > > From: Neal Liu <neal.liu@mediatek.com> > > > > For Mediatek SoCs on ARMv8 with TrustZone enabled, peripherals like > > entropy sources is not accessible from normal world (linux) and > > rather accessible from secure world (ATF/TEE) only. This driver aims > > to provide a generic interface to ATF rng service. > > > > Signed-off-by: Neal Liu <neal.liu@mediatek.com> > > --- > > drivers/char/hw_random/Kconfig | 16 ++++++ > > drivers/char/hw_random/Makefile | 1 + > > drivers/char/hw_random/mt67xx-rng.c | 104 > > +++++++++++++++++++++++++++++++++++ 3 files changed, 121 insertions(+) > > create mode 100644 drivers/char/hw_random/mt67xx-rng.c > > > > diff --git a/drivers/char/hw_random/Kconfig b/drivers/char/hw_random/Kconfig > > index 25a7d8f..98751d3 100644 > > --- a/drivers/char/hw_random/Kconfig > > +++ b/drivers/char/hw_random/Kconfig > > @@ -398,6 +398,22 @@ config HW_RANDOM_MTK > > > > If unsure, say Y. > > > > +config HW_RANDOM_MT67XX > > + tristate "Mediatek MT67XX Random Number Generator support" > > + depends on HW_RANDOM > > + depends on ARCH_MEDIATEK || COMPILE_TEST > > + default HW_RANDOM > > + help > > + This driver provides kernel-side support for the Random Number > > + Generator hardware found on Mediatek MT67xx SoCs. The difference > > + with mtk-rng is the Random Number Generator hardware is secure > > + access only. > > + > > + To compile this driver as a module, choose M here. the > > + module will be called mt67xx-rng. > > + > > + If unsure, say Y. > > + > > config HW_RANDOM_S390 > > tristate "S390 True Random Number Generator support" > > depends on S390 > > diff --git a/drivers/char/hw_random/Makefile > > b/drivers/char/hw_random/Makefile index 7c9ef4a..4be95ab 100644 > > --- a/drivers/char/hw_random/Makefile > > +++ b/drivers/char/hw_random/Makefile > > @@ -36,6 +36,7 @@ obj-$(CONFIG_HW_RANDOM_PIC32) += pic32-rng.o > > obj-$(CONFIG_HW_RANDOM_MESON) += meson-rng.o > > obj-$(CONFIG_HW_RANDOM_CAVIUM) += cavium-rng.o cavium-rng-vf.o > > obj-$(CONFIG_HW_RANDOM_MTK) += mtk-rng.o > > +obj-$(CONFIG_HW_RANDOM_MT67XX) += mt67xx-rng.o > > obj-$(CONFIG_HW_RANDOM_S390) += s390-trng.o > > obj-$(CONFIG_HW_RANDOM_KEYSTONE) += ks-sa-rng.o > > obj-$(CONFIG_HW_RANDOM_OPTEE) += optee-rng.o > > diff --git a/drivers/char/hw_random/mt67xx-rng.c > > b/drivers/char/hw_random/mt67xx-rng.c new file mode 100644 > > index 0000000..e70cbbe > > --- /dev/null > > +++ b/drivers/char/hw_random/mt67xx-rng.c > > @@ -0,0 +1,104 @@ > > +// SPDX-License-Identifier: GPL-2.0 > > +/* > > + * Copyright (C) 2019 MediaTek Inc. > > + */ > > + > > +#include <linux/module.h> > > +#include <linux/platform_device.h> > > +#include <linux/hw_random.h> > > +#include <linux/of.h> > > +#include <linux/arm-smccc.h> > > +#include <linux/soc/mediatek/mtk_sip_svc.h> > > + > > +#define PFX KBUILD_MODNAME ": " > > +#define MT67XX_RNG_MAGIC 0x74726e67 > > +#define SMC_RET_NUM 4 > > + > > +struct mt67xx_rng_priv { > > + struct hwrng rng; > > +}; > > + > > + > > +static void __rng_sec_read(uint32_t *val) > > +{ > > + struct arm_smccc_res res; > > + > > + arm_smccc_smc(MTK_SIP_KERNEL_GET_RND, > > + MT67XX_RNG_MAGIC, 0, 0, 0, 0, 0, 0, &res); > > + > > + val[0] = res.a0; > > + val[1] = res.a1; > > + val[2] = res.a2; > > + val[3] = res.a3; > > +} > > + > > +static int mt67xx_rng_read(struct hwrng *rng, void *buf, size_t max, bool > > wait) +{ > > + int i, retval = 0; > > + uint32_t val[4] = {0}; > > + size_t get_rnd_size = sizeof(u32) * SMC_RET_NUM; > > + > > + if (!buf) { > > + pr_err("%s, buf is NULL\n", __func__); > > + return -EFAULT; > > + } > > + > > + while (max >= get_rnd_size) { > > + __rng_sec_read(val); > > + > > + for (i = 0; i < SMC_RET_NUM; i++) { > > + *(u32 *)buf = val[i]; > > I am not sure this cast is right - or how is it guaranteed that buf is word- > aligned? > > > + buf += sizeof(u32); > > + } > > + > > + retval += get_rnd_size; > > + max -= get_rnd_size; > > + } > > + > > + return retval; > > +} > > + > > +static int mt67xx_rng_probe(struct platform_device *pdev) > > +{ > > + int ret; > > + struct mt67xx_rng_priv *priv; > > + > > + pr_info(PFX "driver registered\n"); > > + priv = devm_kzalloc(&pdev->dev, sizeof(*priv), GFP_KERNEL); > > + if (!priv) > > + return -ENOMEM; > > + > > + priv->rng.name = KBUILD_MODNAME; > > + priv->rng.read = mt67xx_rng_read; > > + priv->rng.priv = (unsigned long)&pdev->dev; > > + priv->rng.quality = 900; > > + > > + ret = devm_hwrng_register(&pdev->dev, &priv->rng); > > + if (ret) { > > + dev_err(&pdev->dev, "failed to register rng device: %d\n", > ret); > > + return ret; > > + } > > + > > + return 0; > > +} > > + > > +static const struct of_device_id mt67xx_rng_match[] = { > > + { .compatible = "mediatek,mt67xx-rng", }, > > + {} > > +}; > > +MODULE_DEVICE_TABLE(of, mt67xx_rng_match); > > + > > +static struct platform_driver mt67xx_rng_driver = { > > + .probe = mt67xx_rng_probe, > > + .driver = { > > + .name = KBUILD_MODNAME, > > + .owner = THIS_MODULE, > > + .of_match_table = mt67xx_rng_match, > > + }, > > +}; > > + > > +module_platform_driver(mt67xx_rng_driver); > > + > > +MODULE_DESCRIPTION("Mediatek MT67XX Random Number Generator Driver"); > > +MODULE_AUTHOR("Neal Liu <neal.liu@mediatek.com>"); > > +MODULE_LICENSE("GPL"); > > > > Ciao > Stephan > > ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [PATCH 3/3] hwrng: add mt67xx-rng driver 2019-05-08 10:35 ` Neal Liu @ 2019-05-09 5:26 ` Herbert Xu 2019-05-09 14:54 ` Neal Liu 0 siblings, 1 reply; 12+ messages in thread From: Herbert Xu @ 2019-05-09 5:26 UTC (permalink / raw) To: Neal Liu Cc: Stephan Mueller, mpm, robh+dt, mark.rutland, matthias.bgg, wsd_upstream, linux-crypto, linux-kernel, devicetree, linux-mediatek, Crystal.Guo On Wed, May 08, 2019 at 06:35:37PM +0800, Neal Liu wrote: > Hi Stephan, > We think the cast is fine, and it cannot guarantee the buf is > word-align. > I reference multiple rng driver's implementation and found it's common > usage for this. So it might be general usage for community. Is there any > suggestion that is more appropriate? If you don't know whether it's unaligned or not then you should do an unaligned operation. Cheers, -- Email: Herbert Xu <herbert@gondor.apana.org.au> Home Page: http://gondor.apana.org.au/~herbert/ PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [PATCH 3/3] hwrng: add mt67xx-rng driver 2019-05-09 5:26 ` Herbert Xu @ 2019-05-09 14:54 ` Neal Liu 2019-05-10 6:39 ` Herbert Xu 0 siblings, 1 reply; 12+ messages in thread From: Neal Liu @ 2019-05-09 14:54 UTC (permalink / raw) To: Herbert Xu Cc: Stephan Mueller, mpm, robh+dt, mark.rutland, matthias.bgg, wsd_upstream, linux-crypto, linux-kernel, devicetree, linux-mediatek, Crystal.Guo On Thu, 2019-05-09 at 13:26 +0800, Herbert Xu wrote: > On Wed, May 08, 2019 at 06:35:37PM +0800, Neal Liu wrote: > > Hi Stephan, > > We think the cast is fine, and it cannot guarantee the buf is > > word-align. > > I reference multiple rng driver's implementation and found it's common > > usage for this. So it might be general usage for community. Is there any > > suggestion that is more appropriate? > > If you don't know whether it's unaligned or not then you should > do an unaligned operation. Hi Stephan/Herbert, My mistake. This buffer is allocated by kmalloc with larger than 32 bytes. So yes, it's word-align for sure. reference: https://elixir.bootlin.com/linux/latest/source/drivers/char/hw_random/core.c#L590 Thanks Best Regards, -Neal Liu ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [PATCH 3/3] hwrng: add mt67xx-rng driver 2019-05-09 14:54 ` Neal Liu @ 2019-05-10 6:39 ` Herbert Xu 2019-05-24 7:42 ` Neal Liu 0 siblings, 1 reply; 12+ messages in thread From: Herbert Xu @ 2019-05-10 6:39 UTC (permalink / raw) To: Neal Liu Cc: Stephan Mueller, mpm, robh+dt, mark.rutland, matthias.bgg, wsd_upstream, linux-crypto, linux-kernel, devicetree, linux-mediatek, Crystal.Guo On Thu, May 09, 2019 at 10:54:46PM +0800, Neal Liu wrote: > > Hi Stephan/Herbert, > My mistake. This buffer is allocated by kmalloc with larger than 32 > bytes. So yes, it's word-align for sure. > reference: > https://elixir.bootlin.com/linux/latest/source/drivers/char/hw_random/core.c#L590 Yes you're right. Cheers, -- Email: Herbert Xu <herbert@gondor.apana.org.au> Home Page: http://gondor.apana.org.au/~herbert/ PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [PATCH 3/3] hwrng: add mt67xx-rng driver 2019-05-10 6:39 ` Herbert Xu @ 2019-05-24 7:42 ` Neal Liu 2019-05-24 12:55 ` Herbert Xu 0 siblings, 1 reply; 12+ messages in thread From: Neal Liu @ 2019-05-24 7:42 UTC (permalink / raw) To: Herbert Xu Cc: mark.rutland@arm.com, devicetree@vger.kernel.org, wsd_upstream, Stephan Mueller, linux-kernel@vger.kernel.org, robh+dt@kernel.org, linux-mediatek@lists.infradead.org, linux-crypto@vger.kernel.org, mpm@selenic.com, matthias.bgg@gmail.com, Crystal Guo (郭晶), Neal Liu Hi Herbert, Could you kindly help to review our patches? Thanks From Neal Liu <> Subject [PATCH 0/3] MT67XX random number generator support Date Wed, 8 May 2019 11:58:54 +0800 share These patch series introduce a generic rng driver for Trustzone based kernel driver which would like to communicate with ATF SIP services. Patch #1 initials SMC fid table for Mediatek SIP interfaces and adds HWRNG related SMC call. Patch #2..3 adds mt67xx-rng kernel driver for Trustzone based SoCs. For Mediatek SoCs on ARMv8 with TrustZone enabled, peripherals like entropy sources is not accessible from normal world (linux) and rather accessible from secure world (ATF/TEE) only. This driver aims to provide a generic interface to ATF rng service. Neal Liu (3): soc: mediatek: add SMC fid table for SIP interface dt-bindings: rng: update bindings for MT67xx SoCs hwrng: add mt67xx-rng driver Documentation/devicetree/bindings/rng/mtk-rng.txt | 13 ++- drivers/char/hw_random/Kconfig | 16 ++++ drivers/char/hw_random/Makefile | 1 + drivers/char/hw_random/mt67xx-rng.c | 104 +++++++++++++++++++++ include/linux/soc/mediatek/mtk_sip_svc.h | 55 +++++++++++ 5 files changed, 186 insertions(+), 3 deletions(-) create mode 100644 drivers/char/hw_random/mt67xx-rng.c create mode 100644 include/linux/soc/mediatek/mtk_sip_svc.h -- 1.7.9.5 Best Regards, -Neal Liu ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [PATCH 3/3] hwrng: add mt67xx-rng driver 2019-05-24 7:42 ` Neal Liu @ 2019-05-24 12:55 ` Herbert Xu 0 siblings, 0 replies; 12+ messages in thread From: Herbert Xu @ 2019-05-24 12:55 UTC (permalink / raw) To: Neal Liu Cc: mark.rutland@arm.com, devicetree@vger.kernel.org, wsd_upstream, Stephan Mueller, linux-kernel@vger.kernel.org, robh+dt@kernel.org, linux-mediatek@lists.infradead.org, linux-crypto@vger.kernel.org, mpm@selenic.com, matthias.bgg@gmail.com, Crystal Guo (郭晶) On Fri, May 24, 2019 at 03:42:34PM +0800, Neal Liu wrote: > Hi Herbert, > Could you kindly help to review our patches? You need acks for patches 1 and 2. Cheers, -- Email: Herbert Xu <herbert@gondor.apana.org.au> Home Page: http://gondor.apana.org.au/~herbert/ PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt ^ permalink raw reply [flat|nested] 12+ messages in thread
* Re: [PATCH 3/3] hwrng: add mt67xx-rng driver 2019-05-08 3:58 ` [PATCH 3/3] hwrng: add mt67xx-rng driver neal.liu [not found] ` <1557287937-2410-4-git-send-email-neal.liu-NuS5LvNUpcJWk0Htik3J/w@public.gmane.org> @ 2019-05-24 19:00 ` Sean Wang 1 sibling, 0 replies; 12+ messages in thread From: Sean Wang @ 2019-05-24 19:00 UTC (permalink / raw) To: neal.liu Cc: mpm, herbert, Rob Herring, Mark Rutland, Matthias Brugger, devicetree, wsd_upstream, lkml, linux-crypto, moderated list:ARM/Mediatek SoC support Hi Neal, On Tue, May 7, 2019 at 9:00 PM <neal.liu@mediatek.com> wrote: > > From: Neal Liu <neal.liu@mediatek.com> > > For Mediatek SoCs on ARMv8 with TrustZone enabled, peripherals like MediaTek > entropy sources is not accessible from normal world (linux) and > rather accessible from secure world (ATF/TEE) only. This driver aims > to provide a generic interface to ATF rng service. Could we rename the mt67xx-rng.c to mtk-sec-rng.c? It seems all MediaTek SoCs with ATF/TEE can reuse this driver, not be limited in MT67xx. > > Signed-off-by: Neal Liu <neal.liu@mediatek.com> > --- > drivers/char/hw_random/Kconfig | 16 ++++++ > drivers/char/hw_random/Makefile | 1 + > drivers/char/hw_random/mt67xx-rng.c | 104 +++++++++++++++++++++++++++++++++++ > 3 files changed, 121 insertions(+) > create mode 100644 drivers/char/hw_random/mt67xx-rng.c > > diff --git a/drivers/char/hw_random/Kconfig b/drivers/char/hw_random/Kconfig > index 25a7d8f..98751d3 100644 > --- a/drivers/char/hw_random/Kconfig > +++ b/drivers/char/hw_random/Kconfig > @@ -398,6 +398,22 @@ config HW_RANDOM_MTK > > If unsure, say Y. > > +config HW_RANDOM_MT67XX > + tristate "Mediatek MT67XX Random Number Generator support" MediaTek Security > + depends on HW_RANDOM > + depends on ARCH_MEDIATEK || COMPILE_TEST > + default HW_RANDOM > + help > + This driver provides kernel-side support for the Random Number > + Generator hardware found on Mediatek MT67xx SoCs. The difference MediaTek > + with mtk-rng is the Random Number Generator hardware is secure > + access only. > + > + To compile this driver as a module, choose M here. the > + module will be called mt67xx-rng. > + > + If unsure, say Y. > + > config HW_RANDOM_S390 > tristate "S390 True Random Number Generator support" > depends on S390 > diff --git a/drivers/char/hw_random/Makefile b/drivers/char/hw_random/Makefile > index 7c9ef4a..4be95ab 100644 > --- a/drivers/char/hw_random/Makefile > +++ b/drivers/char/hw_random/Makefile > @@ -36,6 +36,7 @@ obj-$(CONFIG_HW_RANDOM_PIC32) += pic32-rng.o > obj-$(CONFIG_HW_RANDOM_MESON) += meson-rng.o > obj-$(CONFIG_HW_RANDOM_CAVIUM) += cavium-rng.o cavium-rng-vf.o > obj-$(CONFIG_HW_RANDOM_MTK) += mtk-rng.o > +obj-$(CONFIG_HW_RANDOM_MT67XX) += mt67xx-rng.o > obj-$(CONFIG_HW_RANDOM_S390) += s390-trng.o > obj-$(CONFIG_HW_RANDOM_KEYSTONE) += ks-sa-rng.o > obj-$(CONFIG_HW_RANDOM_OPTEE) += optee-rng.o > diff --git a/drivers/char/hw_random/mt67xx-rng.c b/drivers/char/hw_random/mt67xx-rng.c > new file mode 100644 > index 0000000..e70cbbe > --- /dev/null > +++ b/drivers/char/hw_random/mt67xx-rng.c > @@ -0,0 +1,104 @@ > +// SPDX-License-Identifier: GPL-2.0 > +/* > + * Copyright (C) 2019 MediaTek Inc. > + */ > + > +#include <linux/module.h> > +#include <linux/platform_device.h> > +#include <linux/hw_random.h> > +#include <linux/of.h> > +#include <linux/arm-smccc.h> > +#include <linux/soc/mediatek/mtk_sip_svc.h> > + sort these headers in alphabetical order > +#define PFX KBUILD_MODNAME ": " PFX can be dropped that is less useful > +#define MT67XX_RNG_MAGIC 0x74726e67 > +#define SMC_RET_NUM 4 > + > +struct mt67xx_rng_priv { > + struct hwrng rng; > +}; > + > + > +static void __rng_sec_read(uint32_t *val) > +{ add a prefix like mt67xx, the function call is specific to mtk > + struct arm_smccc_res res; > + > + arm_smccc_smc(MTK_SIP_KERNEL_GET_RND, > + MT67XX_RNG_MAGIC, 0, 0, 0, 0, 0, 0, &res); > + > + val[0] = res.a0; > + val[1] = res.a1; > + val[2] = res.a2; > + val[3] = res.a3; > +} > + > +static int mt67xx_rng_read(struct hwrng *rng, void *buf, size_t max, bool wait) > +{ > + int i, retval = 0; > + uint32_t val[4] = {0}; > + size_t get_rnd_size = sizeof(u32) * SMC_RET_NUM; sort declarations in the reverse-Xmas tree. and get_rnd_size always be unchanged so it can be a macro > + > + if (!buf) { the sanity check is unnecessary, the parameter max would ensure the size of the whole buffer > + pr_err("%s, buf is NULL\n", __func__); > + return -EFAULT; > + } > + > + while (max >= get_rnd_size) { > + __rng_sec_read(val); > + > + for (i = 0; i < SMC_RET_NUM; i++) { > + *(u32 *)buf = val[i]; > + buf += sizeof(u32); > + } > + > + retval += get_rnd_size; > + max -= get_rnd_size; > + } > + > + return retval; > +} > + > +static int mt67xx_rng_probe(struct platform_device *pdev) > +{ > + int ret; > + struct mt67xx_rng_priv *priv; sort declarations in the reverse-Xmas tree. > + > + pr_info(PFX "driver registered\n"); drop the message > + priv = devm_kzalloc(&pdev->dev, sizeof(*priv), GFP_KERNEL); > + if (!priv) > + return -ENOMEM; > + > + priv->rng.name = KBUILD_MODNAME; suggest using the device name priv->rng.name = pdev->name; > + priv->rng.read = mt67xx_rng_read; > + priv->rng.priv = (unsigned long)&pdev->dev; > + priv->rng.quality = 900; > + > + ret = devm_hwrng_register(&pdev->dev, &priv->rng); > + if (ret) { > + dev_err(&pdev->dev, "failed to register rng device: %d\n", ret); > + return ret; > + } > + > + return 0; > +} > + > +static const struct of_device_id mt67xx_rng_match[] = { > + { .compatible = "mediatek,mt67xx-rng", }, > + {} > +}; > +MODULE_DEVICE_TABLE(of, mt67xx_rng_match); > + > +static struct platform_driver mt67xx_rng_driver = { > + .probe = mt67xx_rng_probe, > + .driver = { > + .name = KBUILD_MODNAME, > + .owner = THIS_MODULE, > + .of_match_table = mt67xx_rng_match, > + }, > +}; > + > +module_platform_driver(mt67xx_rng_driver); > + > +MODULE_DESCRIPTION("Mediatek MT67XX Random Number Generator Driver"); MediaTek > +MODULE_AUTHOR("Neal Liu <neal.liu@mediatek.com>"); > +MODULE_LICENSE("GPL"); > -- > 1.7.9.5 > > > _______________________________________________ > Linux-mediatek mailing list > Linux-mediatek@lists.infradead.org > http://lists.infradead.org/mailman/listinfo/linux-mediatek ^ permalink raw reply [flat|nested] 12+ messages in thread
end of thread, other threads:[~2019-05-24 19:00 UTC | newest] Thread overview: 12+ messages (download: mbox.gz follow: Atom feed -- links below jump to the message on this page -- 2019-05-08 3:58 [PATCH 0/3] MT67XX random number generator support neal.liu 2019-05-08 3:58 ` [PATCH 1/3] soc: mediatek: add SMC fid table for SIP interface neal.liu 2019-05-08 3:58 ` [PATCH 2/3] dt-bindings: rng: update bindings for MT67xx SoCs neal.liu 2019-05-08 3:58 ` [PATCH 3/3] hwrng: add mt67xx-rng driver neal.liu [not found] ` <1557287937-2410-4-git-send-email-neal.liu-NuS5LvNUpcJWk0Htik3J/w@public.gmane.org> 2019-05-08 6:34 ` Stephan Mueller 2019-05-08 10:35 ` Neal Liu 2019-05-09 5:26 ` Herbert Xu 2019-05-09 14:54 ` Neal Liu 2019-05-10 6:39 ` Herbert Xu 2019-05-24 7:42 ` Neal Liu 2019-05-24 12:55 ` Herbert Xu 2019-05-24 19:00 ` Sean Wang
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for NNTP newsgroup(s).