Re: [PATCH v2 1/2] xen/arm: Add imx8q{m,x} platform glue

[John Ernberg <john.ernberg@xxxxxxxx>]

Hi Michal,

On 2/15/24 10:02, Michal Orzel wrote:
> Hi,
> 
> On 14/02/2024 17:06, John Ernberg wrote:
>>
>>
>> When using Linux for dom0 there are a bunch of drivers that need to do SMC
>> SIP calls into the firmware to enable certain hardware bits like the
>> watchdog.
>>
>> Provide a basic platform glue that implements the needed SMC forwarding.
>>
>> The format of these calls are as follows:
>>   - reg 0: service ID
>>   - reg 1: function ID
>>   remaining regs: args
>>
>> For now we only allow Dom0 to make these calls as they are all managing
>> hardware. There is no specification for these SIP calls, the IDs and names
>> have been extracted from the upstream linux kernel and the vendor kernel.
>>
>> Most of the SIP calls are only available for the iMX8M series of SoCs, so
>> they are easy to reject and they need to be revisited when iMX8M series
>> support is added.
> Given that you named your driver *qm and you search for dt compatible *qm, 
> *qxp, does it really
> make sense to add SIP calls for iMX8M?

The idea was to be explicit about why it makes no sense to support those 
SIPs on a qm/qxp implementation.

I can take it out if that's not argument enough to have them.
> 
>>
>>  From the other calls we can reject CPUFREQ because Dom0 cannot make an
>> informed decision regarding CPU frequency scaling, WAKEUP_SRC is to wake
>> up from suspend, which Xen doesn't support at this time.
>>
>> This leaves the TIME SIP and the OTP_WRITE SIP, which for now are allowed
>> to Dom0.
>>
>> NOTE: This code is based on code found in NXP Xen tree located here:
>> https://github.com/nxp-imx/imx-xen/blob/lf-5.10.y_4.13/xen/arch/arm/platforms/imx8qm.c
>>
>> Signed-off-by: Peng Fan <peng.fan@xxxxxxx>
>> [jernberg: Add SIP call filtering]
>> Signed-off-by: John Ernberg <john.ernberg@xxxxxxxx>
>>
>> ---
>>
>> v2:
>>   - Reword the commit message to be a bit clearer
>>   - Include the link previously added as a context note to the commit 
>> message (Julien Grall)
>>   - Add Pengs signed off (Julien Grall, Peng Fan)
>>   - Add basic SIP call filter (Julien Grall)
>>   - Expand the commit message a whole bunch because of the changes to the 
>> code
>> ---
>>   xen/arch/arm/platforms/Makefile |   1 +
>>   xen/arch/arm/platforms/imx8qm.c | 143 ++++++++++++++++++++++++++++++++
>>   2 files changed, 144 insertions(+)
>>   create mode 100644 xen/arch/arm/platforms/imx8qm.c
>>
>> diff --git a/xen/arch/arm/platforms/Makefile 
>> b/xen/arch/arm/platforms/Makefile
>> index 8632f4115f..bec6e55d1f 100644
>> --- a/xen/arch/arm/platforms/Makefile
>> +++ b/xen/arch/arm/platforms/Makefile
>> @@ -9,5 +9,6 @@ obj-$(CONFIG_ALL_PLAT)   += sunxi.o
>>   obj-$(CONFIG_ALL64_PLAT) += thunderx.o
>>   obj-$(CONFIG_ALL64_PLAT) += xgene-storm.o
>>   obj-$(CONFIG_ALL64_PLAT) += brcm-raspberry-pi.o
>> +obj-$(CONFIG_ALL64_PLAT) += imx8qm.o
>>   obj-$(CONFIG_MPSOC_PLATFORM)  += xilinx-zynqmp.o
>>   obj-$(CONFIG_MPSOC_PLATFORM)  += xilinx-zynqmp-eemi.o
>> diff --git a/xen/arch/arm/platforms/imx8qm.c 
>> b/xen/arch/arm/platforms/imx8qm.c
>> new file mode 100644
>> index 0000000000..4515c75935
>> --- /dev/null
>> +++ b/xen/arch/arm/platforms/imx8qm.c
>> @@ -0,0 +1,143 @@
>> +/* SPDX-License-Identifier: GPL-2.0-or-later */
>> +/*
>> + * xen/arch/arm/platforms/imx8qm.c
>> + *
>> + * i.MX 8QM setup
>> + *
>> + * Copyright (c) 2016 Freescale Inc.
>> + * Copyright 2018-2019 NXP
>> + *
>> + *
>> + * Peng Fan <peng.fan@xxxxxxx>
>> + */
>> +
>> +#include <xen/sched.h>
>> +#include <asm/platform.h>
>> +#include <asm/smccc.h>
>> +
>> +static const char * const imx8qm_dt_compat[] __initconst =
>> +{
>> +    "fsl,imx8qm",
>> +    "fsl,imx8qxp",
>> +    NULL
>> +};
>> +
>> +#define IMX_SIP_GPC          0xC2000000
> NIT: A matter of personal opinion but all the Arm64 SIP services starts with 
> 0xc2000000,
> so you could just define a function id i.e. 1, 2, ... and use a macro similar 
> to EEMI_FID(fid).
> This is just a suggestion.

Ack.
> 
>> +#define IMX_SIP_CPUFREQ      0xC2000001
>> +#define IMX_SIP_TIME         0xC2000002
>> +#define IMX_SIP_DDR_DVFS     0xC2000004
>> +#define IMX_SIP_SRC          0xC2000005
>> +#define IMX_SIP_GET_SOC_INFO 0xC2000006
>> +#define IMX_SIP_NOC          0xC2000008
>> +#define IMX_SIP_WAKEUP_SRC   0xC2000009
>> +#define IMX_SIP_OTP_WRITE    0xC200000B
> Looking at ATF, for QM,QXP there are also:
> BUILDINFO 0xC2000003
> OTP_READ  0xC200000A
> SET_TEMP  0xC200000C

I can't find them in the current Linux code, perhaps deprecated.
I can add them and deny them if it makes sense.
> 
>> +
>> +#define IMX_SIP_TIME_F_RTC_SET_TIME     0x00
>> +#define IMX_SIP_TIME_F_WDOG_START       0x01
>> +#define IMX_SIP_TIME_F_WDOG_STOP        0x02
>> +#define IMX_SIP_TIME_F_WDOG_SET_ACT     0x03
>> +#define IMX_SIP_TIME_F_WDOG_PING        0x04
>> +#define IMX_SIP_TIME_F_WDOG_SET_TIMEOUT 0x05
>> +#define IMX_SIP_TIME_F_WDOG_GET_STAT    0x06
>> +#define IMX_SIP_TIME_F_WDOG_SET_PRETIME 0x07
>> +
>> +static bool imx8qm_is_sip_time_call_ok(uint32_t function_id)
>> +{
>> +    switch ( function_id )
>> +    {
>> +    case IMX_SIP_TIME_F_RTC_SET_TIME:
>> +        return true;
>> +    case IMX_SIP_TIME_F_WDOG_START:
>> +    case IMX_SIP_TIME_F_WDOG_STOP:
>> +    case IMX_SIP_TIME_F_WDOG_SET_ACT:
>> +    case IMX_SIP_TIME_F_WDOG_PING:
>> +    case IMX_SIP_TIME_F_WDOG_SET_TIMEOUT:
>> +    case IMX_SIP_TIME_F_WDOG_GET_STAT:
>> +    case IMX_SIP_TIME_F_WDOG_SET_PRETIME:
>> +        return true;
>> +    default:
>> +        printk(XENLOG_WARNING "imx8qm: smc: time: Unknown function id 
>> %x\n", function_id);
> 80 chars limit, move argument list to the next line

Ack. I forgot to check the coding standard on line length.
> 
>> +        return false;
>> +    }
>> +}
>> +
>> +static bool imx8qm_smc(struct cpu_user_regs *regs)
>> +{
>> +    uint32_t service_id = get_user_reg(regs, 0);
> In SMCCC naming convention, W0 is called function identifier. Instead you 
> call X1 function_id
> which is a bit misleading.

Ack. Will use function_id and subfunction_id instead.
> 
>> +    uint32_t function_id = get_user_reg(regs, 1);
>> +    struct arm_smccc_res res;
>> +
>> +    if ( !cpus_have_const_cap(ARM_SMCCC_1_1) )
>> +    {
>> +        printk_once(XENLOG_WARNING "no SMCCC 1.1 support. Disabling 
>> firmware calls\n");
> 80 chars limit, move string to the next line
Ack.
> 
>> +
>> +        return false;
>> +    }
>> +
>> +    /* Only hardware domain may use the SIP calls */
>> +    if ( !is_hardware_domain(current->domain) )
>> +    {
>> +        gprintk(XENLOG_WARNING, "imx8qm: smc: No access\n");
>> +        return false;
>> +    }
>> +
>> +    switch ( service_id )
>> +    {
>> +    case IMX_SIP_GPC: /* Only available on imx8m series */
> If we decide to keep iMX8M FIDs, I think adding comments on top of a case 
> would result in a more readable code.
> 
>> +        return false;
>> +    case IMX_SIP_CPUFREQ: /* Dom0 can't take any informed descision here */
>> +        return false;
>> +    case IMX_SIP_TIME:
>> +        if ( imx8qm_is_sip_time_call_ok(function_id) )
>> +            goto allow_call;
>> +        return false;
>> +    case IMX_SIP_DDR_DVFS: /* Only available on imx8m series */
>> +        return false;
>> +    case IMX_SIP_SRC: /* Only available on imx8m series */
>> +        return false;
>> +    case IMX_SIP_GET_SOC_INFO: /* Only available on imx8m series */
>> +        return false;
>> +    case IMX_SIP_NOC: /* Only available on imx8m series */
>> +        return false;
>> +    case IMX_SIP_WAKEUP_SRC: /* Xen doesn't have suspend support */
>> +        return false;
>> +    case IMX_SIP_OTP_WRITE:
>> +        /* function_id is the fuse number, no sensible check possible */
>> +        goto allow_call;
>> +    default:
>> +        printk(XENLOG_WARNING "imx8qm: smc: Unknown service id %x\n", 
>> service_id);
> 80 chars limit, move argument list to the next line

Ack
> 
>> +        return false;
>> +    }
>> +
>> +allow_call:
> labels need to be indented with one space
Ack. Missed that clause in the indentation chapter.
> 
>> +    arm_smccc_1_1_smc(service_id,
>> +                      function_id,
>> +                      get_user_reg(regs, 2),
>> +                      get_user_reg(regs, 3),
>> +                      get_user_reg(regs, 4),
>> +                      get_user_reg(regs, 5),
>> +                      get_user_reg(regs, 6),
>> +                      get_user_reg(regs, 7),
>> +                      &res);
>> +
>> +    set_user_reg(regs, 0, res.a0);
>> +    set_user_reg(regs, 1, res.a1);
>> +    set_user_reg(regs, 2, res.a2);
>> +    set_user_reg(regs, 3, res.a3);
>> +
>> +    return true;
>> +}
>> +
>> +PLATFORM_START(imx8qm, "i.MX 8")
> Shouldn't it be i.MX 8Q{M,XP} ?
Ack.
> 
> ~Michal

Thanks! // John Ernberg