Re: [PATCH v2 3/4] firmware: qcom_scm: Refactor code to support multiple download mode

[Mukesh Ojha <quic_mojha@quicinc.com>]

On 3/16/2023 9:57 PM, Dmitry Baryshkov wrote:
> On 03/03/2023 09:46, Mukesh Ojha wrote:
>> Thanks for your time in checking this..
>>
>> On 3/1/2023 4:29 PM, Dmitry Baryshkov wrote:
>>> On 01/03/2023 11:55, Mukesh Ojha wrote:
>>>> Currently on Qualcomm SoC, download_mode is enabled if
>>>> CONFIG_QCOM_SCM_DOWNLOAD_MODE_DEFAULT is selected.
>>>>
>>>> Refactor the code such that it supports multiple download
>>>> modes and drop CONFIG_QCOM_SCM_DOWNLOAD_MODE_DEFAULT config
>>>> instead, give interface to set the download mode from
>>>> module parameter.
>>>>
>>>> Signed-off-by: Mukesh Ojha <quic_mojha@quicinc.com>
>>>> ---
>>>> Changes in v2:
>>>>   - Passed download mode as parameter
>>>>   - Accept human accepatable download mode string.
>>>>   - enable = !!dload_mode
>>>>   - Shifted module param callback to somewhere down in
>>>>     the file so that it no longer need to know the
>>>>     prototype of qcom_scm_set_download_mode()
>>>>   - updated commit text.
>>>>
>>>>   drivers/firmware/Kconfig    | 11 --------
>>>>   drivers/firmware/qcom_scm.c | 65 
>>>> ++++++++++++++++++++++++++++++++++++++-------
>>>>   2 files changed, 56 insertions(+), 20 deletions(-)
>>>>
>>>> diff --git a/drivers/firmware/Kconfig b/drivers/firmware/Kconfig
>>>> index b59e304..ff7e9f3 100644
>>>> --- a/drivers/firmware/Kconfig
>>>> +++ b/drivers/firmware/Kconfig
>>>> @@ -215,17 +215,6 @@ config MTK_ADSP_IPC
>>>>   config QCOM_SCM
>>>>       tristate
>>>> -config QCOM_SCM_DOWNLOAD_MODE_DEFAULT
>>>> -    bool "Qualcomm download mode enabled by default"
>>>> -    depends on QCOM_SCM
>>>> -    help
>>>> -      A device with "download mode" enabled will upon an unexpected
>>>> -      warm-restart enter a special debug mode that allows the user to
>>>> -      "download" memory content over USB for offline postmortem 
>>>> analysis.
>>>> -      The feature can be enabled/disabled on the kernel command line.
>>>> -
>>>> -      Say Y here to enable "download mode" by default.
>>>> -
>>>>   config SYSFB
>>>>       bool
>>>>       select BOOT_VESA_SUPPORT
>>>> diff --git a/drivers/firmware/qcom_scm.c b/drivers/firmware/qcom_scm.c
>>>> index c9f1fad..ca07208 100644
>>>> --- a/drivers/firmware/qcom_scm.c
>>>> +++ b/drivers/firmware/qcom_scm.c
>>>> @@ -17,17 +17,22 @@
>>>>   #include <linux/clk.h>
>>>>   #include <linux/reset-controller.h>
>>>>   #include <linux/arm-smccc.h>
>>>> +#include <linux/kstrtox.h>
>>>>   #include "qcom_scm.h"
>>>> -static bool download_mode = 
>>>> IS_ENABLED(CONFIG_QCOM_SCM_DOWNLOAD_MODE_DEFAULT);
>>>> -module_param(download_mode, bool, 0);
>>>> -
>>>>   #define SCM_HAS_CORE_CLK    BIT(0)
>>>>   #define SCM_HAS_IFACE_CLK    BIT(1)
>>>>   #define SCM_HAS_BUS_CLK        BIT(2)
>>>>   #define QCOM_DOWNLOAD_MODE_MASK 0x30
>>>> +#define QCOM_DOWNLOAD_FULLDUMP    0x10
>>>> +#define QCOM_DOWNLOAD_NODUMP    0x0
>>>> +
>>>> +#define MAX_DLOAD_LEN    8
>>>> +
>>>> +static char download_mode[] = "off";
>>>> +static u32 dload_mode;
>>>>   struct qcom_scm {
>>>>       struct device *dev;
>>>> @@ -417,8 +422,9 @@ static int __qcom_scm_set_dload_mode(struct 
>>>> device *dev, bool enable)
>>>>       return qcom_scm_call_atomic(__scm->dev, &desc, NULL);
>>>>   }
>>>> -static void qcom_scm_set_download_mode(bool enable)
>>>> +static void qcom_scm_set_download_mode(u32 dload_mode)
>>>>   {
>>>> +    bool enable = !!dload_mode;
>>>>       bool avail;
>>>>       int ret = 0;
>>>>       u32 val;
>>>> @@ -438,7 +444,7 @@ static void qcom_scm_set_download_mode(bool enable)
>>>>           val &= ~QCOM_DOWNLOAD_MODE_MASK;
>>>>           if (enable)
>>>> -            val |= QCOM_SCM_BOOT_SET_DLOAD_MODE;
>>>> +            val |= dload_mode;
>>>>           ret = qcom_scm_io_writel(__scm->dload_mode_addr, val);
>>>>       } else {
>>>> @@ -1338,6 +1344,47 @@ bool qcom_scm_is_available(void)
>>>>   }
>>>>   EXPORT_SYMBOL(qcom_scm_is_available);
>>>> +static int set_dload_mode(const char *val, const struct 
>>>> kernel_param *kp)
>>>> +{
>>>> +    int ret;
>>>> +
>>>> +    if (!strncmp(val, "full", strlen("full"))) {
>>>> +        dload_mode = QCOM_DOWNLOAD_FULLDUMP;
>>>> +    } else if (!strncmp(val, "off", strlen("off"))) {
>>>> +        dload_mode = QCOM_DOWNLOAD_NODUMP;
>>>> +    } else {
>>>> +        if (kstrtouint(val, 0, &dload_mode) ||
>>>> +             !(dload_mode == 0 || dload_mode == 1)) {
>>>> +            pr_err("unknown download mode\n");
>>>> +            return -EINVAL;
>>>> +        }
>>>> +
>>>> +    }
>>>> +
>>>> +    ret = param_set_copystring(val, kp);
>>>> +    if (ret)
>>>> +        return ret;
>>>> +
>>>> +    if (__scm)
>>>> +        qcom_scm_set_download_mode(dload_mode);
>>>> +
>>>> +    return 0;
>>>> +}
>>>> +
>>>> +static const struct kernel_param_ops dload_mode_param_ops = {
>>>> +    .set = set_dload_mode,
>>>> +    .get = param_get_string,
>>>
>>> Please follow the param_get_bool approach and return sanitized data 
>>> here. In other words, "full" / "none" depending on the dload_mode 
>>> instead of storing the passed string and returning it later.
>>>
>>
>> This could be done.
>>
>>>> +};
>>>> +
>>>> +static struct kparam_string dload_mode_param = {
>>>> +    .string = download_mode,
>>>> +    .maxlen = MAX_DLOAD_LEN,
>>>
>>> I think writing "full" to this param might overwrite some kernel 
>>> data. "00000000" should be even worse.
>>
>> There is check in param_set_copystring() which would avoid that.
> 
> 
> No. The check will validate the value's length against MAX_DLOAD_LEN. 
> But it will not safeguard your code. download_mode's size is less than 
> MAX_DLOAD_LEN.

Oops !! you are right..,
Will fix it in v3.

-Mukesh

>