| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <86A0E76937111F4C92FABEC0A209885104426182@az33exm21>
Date: Thu, 4 Feb 2010 10:09:07 -0700
From: "Nguyen Dinh-R00091" <R00091@...escale.com>
To: "Amit Kucheria" <amit.kucheria@...onical.com>,
"Eric Miao" <eric.y.miao@...il.com>
Cc: "List Linux Kernel" <linux-kernel@...r.kernel.org>,
<linux@....linux.org.uk>, <s.hauer@...gutronix.de>,
<grant.likely@...retlab.ca>,
"Herring Robert-RA7055" <RA7055@...escale.com>,
<linux-arm-kernel@...ts.infradead.org>, <daniel@...aq.de>,
<bryan.wu@...onical.com>, <valentin.longchamp@...l.ch>
Subject: RE: [PATCHv2 01/11] arm: mxc: TrustZone interrupt controller (TZIC) for i.MX5 family
-----Original Message-----
From: Amit Kucheria [mailto:amit.kucheria@...onical.com]
Sent: Wednesday, February 03, 2010 7:24 AM
To: Eric Miao
Cc: List Linux Kernel; linux@....linux.org.uk; Nguyen Dinh-R00091; s.hauer@...gutronix.de; grant.likely@...retlab.ca; Herring Robert-RA7055; linux-arm-kernel@...ts.infradead.org; daniel@...aq.de; bryan.wu@...onical.com; valentin.longchamp@...l.ch
Subject: Re: [PATCHv2 01/11] arm: mxc: TrustZone interrupt controller (TZIC) for i.MX5 family
On 10 Feb 02, Eric Miao wrote:
> Hi Amit,
>
> Just some nit-picking review comments, see below:
>
> On Tue, Feb 2, 2010 at 9:16 PM, Amit Kucheria
> <amit.kucheria@...onical.com> wrote:
> > Freescale i.MX51 processor uses a new interrupt controller. Add
> > driver for TrustZone Interrupt Controller
> >
> > Signed-off-by: Amit Kucheria <amit.kucheria@...onical.com>
> > ---
> > arch/arm/plat-mxc/Kconfig | 8 ++
> > arch/arm/plat-mxc/Makefile | 3 +
> > arch/arm/plat-mxc/tzic.c | 182
> > ++++++++++++++++++++++++++++++++++++++++++++
> > 3 files changed, 193 insertions(+), 0 deletions(-)
> > create mode 100644 arch/arm/plat-mxc/tzic.c
> >
> > diff --git a/arch/arm/plat-mxc/Kconfig b/arch/arm/plat-mxc/Kconfig
> > index 8b0a1ee..59558c4 100644
> > --- a/arch/arm/plat-mxc/Kconfig
> > +++ b/arch/arm/plat-mxc/Kconfig
> > @@ -62,6 +62,14 @@ config MXC_IRQ_PRIOR
> > requirements for timing.
> > Say N here, unless you have a specialized requirement.
> >
> > +config MXC_TZIC
> > + bool "Enable TrustZone Interrupt Controller"
> > + depends on ARCH_MX51
>
> This is the first patch of the base port, yet I cannot find any
> reference to this ARCH_MX51, did you miss something?
ARCH_MX51 is only introduced in the later patches that add the core i.MX5 code. Since TZIC is not inherently dependent on i.MX5 (it's merely the first processor to use it), I thought of splitting it out as a separate patch.
Does this break the sanctity of one self-contained change?
> > + help
> > + This will be automatically selected for all processors
> > + containing this interrupt controller.
> > + Say N here only if you are really sure.
> > +
> > config MXC_PWM
> > tristate "Enable PWM driver"
> > depends on ARCH_MXC
> > diff --git a/arch/arm/plat-mxc/Makefile b/arch/arm/plat-mxc/Makefile
> > index 996cbac..0202ad9 100644
> > --- a/arch/arm/plat-mxc/Makefile
> > +++ b/arch/arm/plat-mxc/Makefile
> > @@ -5,6 +5,9 @@
> > # Common support
> > obj-y := irq.o clock.o gpio.o time.o devices.o cpu.o system.o
> >
> > +# MX51 uses the TZIC interrupt controller, older platforms use AVIC
> > +(irq.o)
> > +obj-$(CONFIG_MXC_TZIC) += tzic.o
> > +
> > obj-$(CONFIG_ARCH_MX1) += iomux-mx1-mx2.o dma-mx1-mx2.o
> > obj-$(CONFIG_ARCH_MX2) += iomux-mx1-mx2.o dma-mx1-mx2.o
> > obj-$(CONFIG_ARCH_MXC_IOMUX_V3) += iomux-v3.o diff --git
> > a/arch/arm/plat-mxc/tzic.c b/arch/arm/plat-mxc/tzic.c new file mode
> > 100644 index 0000000..00cb0ad
> > --- /dev/null
> > +++ b/arch/arm/plat-mxc/tzic.c
> > @@ -0,0 +1,182 @@
> > +/*
> > + * Copyright 2004-2009 Freescale Semiconductor, Inc. All Rights Reserved.
> > + *
> > + * The code contained herein is licensed under the GNU General
> > +Public
> > + * License. You may obtain a copy of the GNU General Public License
> > + * Version 2 or later at the following locations:
> > + *
> > + * http://www.opensource.org/licenses/gpl-license.html
> > + * http://www.gnu.org/copyleft/gpl.html
> > + */
> > +
> > +#include <linux/module.h>
> > +#include <linux/moduleparam.h>
> > +#include <linux/init.h>
> > +#include <linux/device.h>
> > +#include <linux/errno.h>
> > +#include <linux/io.h>
> > +
> > +#include <asm/mach/irq.h>
> > +
> > +#include <mach/hardware.h>
> > +
> > +/*
> > + *****************************************
> > + * TZIC Registers *
> > + *****************************************
> > + */
> > +
> > +#define TZIC_INTCNTL 0x0000 /* Control register */
> > +#define TZIC_INTTYPE 0x0004 /* Controller Type register
> > +*/ #define TZIC_IMPID 0x0008 /* Distributor
> > +Implementer Identification */ #define TZIC_PRIOMASK
> > +0x000C /* Priority Mask Reg */ #define TZIC_SYNCCTRL
> > +0x0010 /* Synchronizer Control register */ #define TZIC_DSMINT
> > +0x0014 /* DSM interrupt Holdoffregister */ #define TZIC_INTSEC0
> > +0x0080 /* Interrupt Security register 0 */ #define TZIC_ENSET0
> > +0x0100 /* Enable Set Register 0 */ #define TZIC_ENCLEAR0
> > +0x0180 /* Enable Clear Register 0 */ #define TZIC_SRCSET0
> > +0x0200 /* Source Set Register 0 */ #define TZIC_SRCCLAR0
> > +0x0280 /* Source Clear Register 0 */ #define TZIC_PRIORITY0
> > +0x0400 /* Priority Register 0 */ #define TZIC_PND0
> > +0x0D00 /* Pending Register 0 */ #define TZIC_HIPND0
> > +0x0D80 /* High Priority Pending Register */ #define TZIC_WAKEUP0
> > +0x0E00 /* Wakeup Config Register */ #define TZIC_SWINT
> > +0x0F00 /* Software Interrupt Rigger Register */ #define TZIC_ID0
> > +0x0FD0 /* Indentification Register 0 */
> > +
> > +void __iomem *tzic_base;
>
> This can just be made to 'static' if it's not used elsewhere, and I'm
> wondering if it's neater to define them as:
>
> #define TZIC_INTCNTL (tzic_base + 0x0000)
>
> so to make the code below short and handy.
tzic_base is actually used in entry-macro.S in patch 0004. I've tried to follow AVIC's way of doing things.
> > +
> > +/*
> > + * Disable interrupt number "irq" in the TZIC
>
> I don't think this follows kernel API doc exactly, you may want to
> have a look into Documentation/kernel-doc-nano-HOWTO.txt.
OK.
> > + *
> > + * @param irq interrupt source number */ static void
> > +tzic_mask_irq(unsigned int irq) {
> > + int index, off;
> > +
> > + index = irq >> 5;
> > + off = irq & 0x1F;
> > + __raw_writel(1 << off, tzic_base + TZIC_ENCLEAR0 + (index <<
> > + 2));
>
> I'll normally define TZIC_ENCLEAR0 then as:
>
> #define TZIC_ENCLEAR(i) (0x0180 + ((i) << 2))
>
> so the above can be written as:
>
> __raw_writel(1 << off, tzic_base + TZIC_ENCLEAR(index));
>
> or by including tzic_base into TZIC_*, simply as:
>
> __raw_writel(1 << off, TZIC_ENCLEAR(index));
OK.
> > +}
> > +
> > +/*
> > + * Enable interrupt number "irq" in the TZIC
> > + *
> > + * @param irq interrupt source number */ static void
> > +tzic_unmask_irq(unsigned int irq) {
> > + int index, off;
> > +
> > + index = irq >> 5;
> > + off = irq & 0x1F;
> > + __raw_writel(1 << off, tzic_base + TZIC_ENSET0 + (index <<
> > +2)); }
> > +
> > +static unsigned int wakeup_intr[4];
> > +
> > +/*
> > + * Set interrupt number "irq" in the TZIC as a wake-up source.
> > + *
> > + * @param irq interrupt source number
> > + * @param enable enable as wake-up if equal to non-zero
> > + * disble as wake-up if equal to zero
> > + *
> > + * @return This function returns 0 on success.
> > + */
> > +static int tzic_set_wake_irq(unsigned int irq, unsigned int enable)
> > +{
> > + unsigned int index, off;
> > +
> > + index = irq >> 5;
> > + off = irq & 0x1F;
> > +
> > + if (index > 3)
> > + return -EINVAL;
> > +
> > + if (enable)
> > + wakeup_intr[index] |= (1 << off);
> > + else
> > + wakeup_intr[index] &= ~(1 << off);
> > +
> > + return 0;
> > +}
> > +
> > +static struct irq_chip mxc_tzic_chip = {
> > + .name = "MXC_TZIC",
> > + .ack = tzic_mask_irq,
> > + .mask = tzic_mask_irq,
> > + .unmask = tzic_unmask_irq,
> > + .set_wake = tzic_set_wake_irq, };
> > +
> > +/*
> > + * This function initializes the TZIC hardware and disables all the
> > + * interrupts. It registers the interrupt enable and disable
> > +functions
> > + * to the kernel for each interrupt source.
> > + */
> > +void __init tzic_init_irq(void __iomem *irqbase) {
> > + int i;
> > +
> > + tzic_base = irqbase;
> > + /* put the TZIC into the reset value with
> > + * all interrupts disabled
> > + */
> > + i = __raw_readl(tzic_base + TZIC_INTCNTL);
>
> Mixing the use of 'i' as both a signed counter and register value
> might not be a good idea, provided it's not guaranteed from theory
> that 'i' as an integer could not be sufficient to hold the value
> returned from
> __raw_readl()
Fair enough.
> > +
> > + __raw_writel(0x80010001, tzic_base + TZIC_INTCNTL);
> > + i = __raw_readl(tzic_base + TZIC_INTCNTL);
> > + __raw_writel(0x1f, tzic_base + TZIC_PRIOMASK);
> > + i = __raw_readl(tzic_base + TZIC_PRIOMASK);
> > + __raw_writel(0x02, tzic_base + TZIC_SYNCCTRL);
> > + i = __raw_readl(tzic_base + TZIC_SYNCCTRL);
>
> Are these read-back really necessary? We can start without them and
> add them later if they do cause issues.
Can anybody from Freescale comment whether the read-back is necessary?
I'll remove it for now to see what happens in my testing.
[Dinh] - This piece of code was taken from our silicon validation code where it was check the values of the read-backs. The read-backs can be removed.
> > +
> > + for (i = 0; i < 4; i++)
> > + __raw_writel(0xFFFFFFFF, tzic_base + TZIC_INTSEC0 +
> > + i * 4);
> > +
> > + /* disable all interrupts */
> > + for (i = 0; i < 4; i++)
> > + __raw_writel(0xFFFFFFFF, tzic_base + TZIC_ENCLEAR0 +
> > + i * 4);
> > +
> > + /* all IRQ no FIQ Warning :: No selection */
> > +
> > + for (i = 0; i < MXC_INTERNAL_IRQS; i++) {
> > + set_irq_chip(i, &mxc_tzic_chip);
> > + set_irq_handler(i, handle_level_irq);
> > + set_irq_flags(i, IRQF_VALID);
> > + }
> > +
> > + printk(KERN_INFO "TrustZone Interrupt Controller (TZIC)
> > + initialized\n");
>
> You may want to use pr_info() for short.
OK
> > +}
> > +
> > +/*
> > + * enable wakeup interrupt
> > + *
> > + * @param is_idle 1 if called in idle loop (ENSET
> > +register);
> > + * 0 to be used when called from low
> > +power entry
> > + * @return 0 if successful; non-zero otherwise
> > + *
> > + */
> > +int tzic_enable_wake(int is_idle)
> > +{
> > + unsigned int i, v;
> > +
> > + __raw_writel(1, tzic_base + TZIC_DSMINT);
> > + if (unlikely(__raw_readl(tzic_base + TZIC_DSMINT) == 0))
> > + return -EAGAIN;
>
> Looks like an unnecessary read-back provided the silicon is sane enough.
Again, Dinh/Rob can you comment?
[Dinh] - Can be removed.
> > +
> > + if (likely(is_idle)) {
> > + for (i = 0; i < 4; i++) {
> > + v = __raw_readl(tzic_base + TZIC_ENSET0 + i
> > + * 4);
> > + __raw_writel(v, tzic_base + TZIC_WAKEUP0 + i
> > + * 4);
> > + }
> > + } else {
> > + for (i = 0; i < 4; i++) {
> > + v = wakeup_intr[i];
> > + __raw_writel(v, tzic_base + TZIC_WAKEUP0 + i
> > + * 4);
> > + }
> > + }
>
> Or could be simplified to:
>
> for (i = 0; i < 4; i++) {
> v = is_idle ? __raw_readl(TZIC_ENSET(i)) : wakeup_intr[i];
> __raw_writel(v, TZIC_WAKEUP(i));
> }
OK
> but just nit-picking comments, so it's up to you.
>
> > + return 0;
> > +}
>
> Mmmm.... this being called elsewhere, I'm thinking about making this a
> sys_device and having this called within sysdev_class.suspend() to
> make this file rather self-contained.
That is the idea once the base port is upstream.
Thanks for the review.
/Amit
--
----------------------------------------------------------------------
Amit Kucheria, Kernel Engineer || amit.kucheria@...onical.com
----------------------------------------------------------------------
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
Powered by blists - more mailing lists