linux-kernel - RE: [PATCH v3] TCO watchdog pretimeout handler

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
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <B9C02DB17496AC4197F41A146D371B9B1A63E612@hasmsx107.ger.corp.intel.com>
Date:	Thu, 18 Jun 2015 12:41:23 +0000
From:	"Muller, Francois-nicolas" <francois-nicolas.muller@...el.com>
To:	Andy Shevchenko <andy.shevchenko@...il.com>
CC:	"linux@...ck-us.net" <linux@...ck-us.net>,
	"wim@...ana.be" <wim@...ana.be>,
	"linux-watchdog@...r.kernel.org" <linux-watchdog@...r.kernel.org>,
	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: RE: [PATCH v3] TCO watchdog pretimeout handler

On Wed, June 17, 2015 6:06 PM, Andy Shevchenko <andy.shevchenko@...il.com> wrote:

> On Wed, Jun 17, 2015 at 4:34 PM, Francois-Nicolas Muller <francois-nicolas.muller@...el.com> wrote:
>
>> +#define DEFAULT_PRETIMEOUT 0
>> +static bool pretimeout = DEFAULT_PRETIMEOUT;
>
> Static variables should not be assigned to zero explicitly (it doesn't make sense).
> Moreover you have integer -> boolean implicit conversion.

Agree, I will remove the DEFAULT_PRETIMEOUT define.

> The below should go under CONFIG_ACPI.
>
>> +static unsigned char *tco_hid = "8086229C"; /* Intel Cherrytrail LPC 
>> +*/

This codes also compiles with CONFIG_ACPI not defined, #if CONFIG_ACPI is not required here.

>> +static unsigned char *tco_hid = "8086229C"; /* Intel Cherrytrail LPC 
>> +*/
>
> Why not to use acpi_device_id ?

As long as there is only one id, I think this is not required.
Anyway I will use acpi_device_id instead to be prepared for future ids.

>> +       if (object.type != ACPI_TYPE_INTEGER)
>> +               return AE_BAD_DATA;
>
> Do we really need this right now? Existing users of _GPE are considering the result as integer w/o an additional check.

Right, as this is already done elsewhere without the check, I will remove it.
 
Thanks,
Francois-Nicolas

-----Original Message-----
From: Andy Shevchenko [mailto:andy.shevchenko@...il.com] 
Sent: Wednesday, June 17, 2015 6:06 PM
To: Muller, Francois-nicolas
Cc: linux@...ck-us.net; wim@...ana.be; linux-watchdog@...r.kernel.org; linux-kernel@...r.kernel.org
Subject: Re: [PATCH v3] TCO watchdog pretimeout handler

On Wed, Jun 17, 2015 at 4:34 PM, Francois-Nicolas Muller <francois-nicolas.muller@...el.com> wrote:

First of all it would be nice to start a new thread per iteration.

> Use TCO watchdog first timeout (pretimeout) to dump CPU backtraces and 
> ease debug of watchdog expiration causes.
> On Intel Cherrytrail, TCO logic generates a SMI, then SMI handler 
> triggers a SCI to the kernel, on a specific GPE.
> The GPE handler dumps all CPU backtraces and calls panic (in order to 
> execute registered panic callbacks).
> GPE number is configured from ACPI tables if LPC HID exists.
>
> Signed-off-by: Francois-Nicolas Muller 
> <francois-nicolas.muller@...el.com>
> ---
> SMI is not supported by the driver, only SCI.
>
> On Intel Cherrytrail, TCO watchdog raises an SMI, then the SMI handler 
> in Bios trigs a SCI to the kernel (in Android OS configuration).
>
> The patch has some effect only on Cherrytrail platform.
> On Braswell, the TCO HID exists in ACPI tables, so SCI is configured. 
> But the SMI handler does not trig the SCI.
> On other platforms, the HID does not exist in ACPI tables, and the SMI 
> handler does not trig the SCI.
>
> In ACPI tables, _GPE is associated to this HID, so no possible confusion.
>
> François-Nicolas
> ---
>  drivers/watchdog/iTCO_wdt.c | 50 
> +++++++++++++++++++++++++++++++++++++++++++++
>  1 file changed, 50 insertions(+)
>
> diff --git a/drivers/watchdog/iTCO_wdt.c b/drivers/watchdog/iTCO_wdt.c 
> index 3c3fd41..3e9ec8b 100644
> --- a/drivers/watchdog/iTCO_wdt.c
> +++ b/drivers/watchdog/iTCO_wdt.c
> @@ -68,6 +68,8 @@
>  #include <linux/io.h>                  /* For inb/outb/... */
>  #include <linux/mfd/core.h>
>  #include <linux/mfd/lpc_ich.h>
> +#include <linux/nmi.h>
> +#include <linux/acpi.h>
>
>  #include "iTCO_vendor.h"
>
> @@ -127,6 +129,12 @@ module_param(turn_SMI_watchdog_clear_off, int, 
> 0);  MODULE_PARM_DESC(turn_SMI_watchdog_clear_off,
>         "Turn off SMI clearing watchdog (depends on 
> TCO-version)(default=1)");
>
> +#define DEFAULT_PRETIMEOUT 0
> +static bool pretimeout = DEFAULT_PRETIMEOUT;

Static variables should not be assigned to zero explicitly (it doesn't make sense).
Moreover you have integer -> boolean implicit conversion.

> +module_param(pretimeout, bool, 0);
> +MODULE_PARM_DESC(pretimeout, "Enable watchdog pretimeout (default="
> +                               __MODULE_STRING(DEFAULT_PRETIMEOUT) 
> +")");

Since it's boolean, I suppose to use direct value here. It would be one line.

> +
>  /*
>   * Some TCO specific functions
>   */
> @@ -201,6 +209,45 @@ static int iTCO_wdt_unset_NO_REBOOT_bit(void)
>         return ret; /* returns: 0 = OK, -EIO = Error */  }
>

The below should go under CONFIG_ACPI.

> +static unsigned char *tco_hid = "8086229C"; /* Intel Cherrytrail LPC 
> +*/

Why not to use acpi_device_id ?

> +
> +static u32 iTCO_wdt_pretimeout_handler(acpi_handle gpe_device, u32 gpe,
> +                                      void *context) {
> +       /* dump backtraces for all available cores */
> +       trigger_all_cpu_backtrace();
> +
> +       /* call panic notifiers */
> +       panic("Kernel Watchdog");
> +
> +       return ACPI_INTERRUPT_HANDLED; }
> +
> +static acpi_status __init iTCO_wdt_register_gpe(acpi_handle handle,
> +                                       u32 lvl, void *context, void 
> +**rv) {
> +       unsigned long long gpe;
> +       acpi_status status;
> +       union acpi_object object = { 0 };
> +       struct acpi_buffer buffer = { sizeof(union acpi_object), 
> +&object };
> +
> +       status = acpi_evaluate_object(handle, "_GPE", NULL, &buffer);
> +       if (ACPI_FAILURE(status))
> +               return status;
> +
> +       if (object.type != ACPI_TYPE_INTEGER)
> +               return AE_BAD_DATA;

Do we really need this right now? Existing users of _GPE are considering the result as integer w/o an additional check.

> +
> +       gpe = object.integer.value;
> +       status = acpi_install_gpe_handler(NULL, gpe, ACPI_GPE_EDGE_TRIGGERED,
> +                                         iTCO_wdt_pretimeout_handler, 
> + NULL);

> +       if (ACPI_FAILURE(status))
> +               return status;
> +
> +       acpi_enable_gpe(NULL, gpe);
> +       return AE_OK;
> +}
> +
>  static int iTCO_wdt_start(struct watchdog_device *wd_dev)  {
>         unsigned int val;
> @@ -641,6 +688,9 @@ static int __init iTCO_wdt_init_module(void)
>         if (err)
>                 return err;
>
> +       if (pretimeout)
> +               acpi_get_devices(tco_hid, iTCO_wdt_register_gpe, NULL, 
> + NULL);
> +
>         return 0;
>  }
>
> --
> 1.9.1
>
> --
> 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/



--
With Best Regards,
Andy Shevchenko