| 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:
<TYZPR06MB52032C46C1AB2C18B923A8E6B21F2@TYZPR06MB5203.apcprd06.prod.outlook.com>
Date: Mon, 13 Jan 2025 03:41:09 +0000
From: Chin-Ting Kuo <chin-ting_kuo@...eedtech.com>
To: Andrew Jeffery <andrew@...econstruct.com.au>, "patrick@...cx.xyz"
<patrick@...cx.xyz>, "linux@...ck-us.net" <linux@...ck-us.net>,
"wim@...ux-watchdog.org" <wim@...ux-watchdog.org>, "joel@....id.au"
<joel@....id.au>, "linux-watchdog@...r.kernel.org"
<linux-watchdog@...r.kernel.org>, "linux-arm-kernel@...ts.infradead.org"
<linux-arm-kernel@...ts.infradead.org>, "linux-aspeed@...ts.ozlabs.org"
<linux-aspeed@...ts.ozlabs.org>, "linux-kernel@...r.kernel.org"
<linux-kernel@...r.kernel.org>
CC: "Peter.Yin@...ntatw.com" <Peter.Yin@...ntatw.com>,
"Patrick_NC_Lin@...ynn.com" <Patrick_NC_Lin@...ynn.com>, BMC-SW
<BMC-SW@...eedtech.com>, "chnguyen@...erecomputing.com"
<chnguyen@...erecomputing.com>, Aaron Lee <aaron_lee@...eedtech.com>
Subject: RE: [PATCH v6 1/1] watchdog: aspeed: Update bootstatus handling
Hi Andrew,
> -----Original Message-----
> From: Andrew Jeffery <andrew@...econstruct.com.au>
> Sent: Monday, January 13, 2025 11:29 AM
> Subject: Re: [PATCH v6 1/1] watchdog: aspeed: Update bootstatus handling
>
> On Sun, 2025-01-12 at 16:12 +0800, Chin-Ting Kuo wrote:
> > The boot status in the watchdog device struct is updated during
> > controller probe stage. Application layer can get the boot status
> > through the command, cat /sys/class/watchdog/watchdogX/bootstatus.
> > The bootstatus can be,
> > WDIOF_CARDRESET => the system is reset by WDT SoC reset.
> > Others => other reset events, e.g., power on reset.
> >
> > On ASPEED platform, the boot status is recorded in the SCU registers.
> > - AST2400: Only a bit represents for any WDT reset.
> > - AST2500/AST2600: The reset triggered by different WDT controllers
> > can be distinguished by different SCU bits.
> >
> > Besides, on AST2400 and AST2500, since alternating boot event is
> > triggered by WDT SoC reset, it is classified as WDIOF_CARDRESET.
> >
> > Signed-off-by: Chin-Ting Kuo <chin-ting_kuo@...eedtech.com>
> > ---
> > drivers/watchdog/aspeed_wdt.c | 81
> > ++++++++++++++++++++++++++++++++++-
> > 1 file changed, 79 insertions(+), 2 deletions(-)
> >
> > diff --git a/drivers/watchdog/aspeed_wdt.c
> > b/drivers/watchdog/aspeed_wdt.c index b4773a6aaf8c..369635b38ca0
> > 100644
> > --- a/drivers/watchdog/aspeed_wdt.c
> > +++ b/drivers/watchdog/aspeed_wdt.c
> > @@ -11,21 +11,30 @@
> > #include <linux/io.h>
> > #include <linux/kernel.h>
> > #include <linux/kstrtox.h>
> > +#include <linux/mfd/syscon.h>
> > #include <linux/module.h>
> > #include <linux/of.h>
> > #include <linux/of_irq.h>
> > #include <linux/platform_device.h>
> > +#include <linux/regmap.h>
> > #include <linux/watchdog.h>
> >
> > static bool nowayout = WATCHDOG_NOWAYOUT;
> > module_param(nowayout, bool, 0);
> > MODULE_PARM_DESC(nowayout, "Watchdog cannot be stopped once
> started
> > (default="
> > __MODULE_STRING(WATC
> HDOG_NOWAYOUT)
> > ")");
> > +struct aspeed_wdt_scu {
> > + const char *compatible;
> > + u32 reset_status_reg;
> > + u32 wdt_reset_mask;
> > + u32 wdt_reset_mask_shift;
> > +};
> >
> > struct aspeed_wdt_config {
> > u32 ext_pulse_width_mask;
> > u32 irq_shift;
> > u32 irq_mask;
> > + struct aspeed_wdt_scu scu;
> > };
> >
> > struct aspeed_wdt {
> > @@ -39,18 +48,36 @@ static const struct aspeed_wdt_config
> > ast2400_config = {
> > .ext_pulse_width_mask = 0xff,
> > .irq_shift = 0,
> > .irq_mask = 0,
> > + .scu = {
> > + .compatible = "aspeed,ast2400-scu",
> > + .reset_status_reg = 0x3c,
> > + .wdt_reset_mask = 0x1,
> > + .wdt_reset_mask_shift = 1,
> > + },
> > };
> >
> > static const struct aspeed_wdt_config ast2500_config = {
> > .ext_pulse_width_mask = 0xfffff,
> > .irq_shift = 12,
> > .irq_mask = GENMASK(31, 12),
> > + .scu = {
> > + .compatible = "aspeed,ast2500-scu",
> > + .reset_status_reg = 0x3c,
> > + .wdt_reset_mask = 0x1,
> > + .wdt_reset_mask_shift = 2,
> > + },
> > };
> >
> > static const struct aspeed_wdt_config ast2600_config = {
> > .ext_pulse_width_mask = 0xfffff,
> > .irq_shift = 0,
> > .irq_mask = GENMASK(31, 10),
> > + .scu = {
> > + .compatible = "aspeed,ast2600-scu",
> > + .reset_status_reg = 0x74,
> > + .wdt_reset_mask = 0xf,
> > + .wdt_reset_mask_shift = 16,
> > + },
> > };
> >
> > static const struct of_device_id aspeed_wdt_of_table[] = { @@ -213,6
> > +240,56 @@ static int aspeed_wdt_restart(struct watchdog_device *wdd,
> > return 0;
> > }
> >
> > +static void aspeed_wdt_update_bootstatus(struct platform_device
> > *pdev,
> > + struct
> aspeed_wdt *wdt) {
> > + const struct resource *res;
> > + struct aspeed_wdt_scu scu = wdt->cfg->scu;
> > + struct regmap *scu_base;
> > + u32 reset_mask_width;
> > + u32 reset_mask_shift;
> > + u32 idx = 0;
> > + u32 status;
> > + int ret;
> > +
> > + if (!of_device_is_compatible(pdev->dev.of_node,
> > "aspeed,ast2400-wdt")) {
> > + res = platform_get_resource(pdev,
> IORESOURCE_MEM, 0);
> > + idx = ((intptr_t)wdt->base & 0x00000fff) /
> > resource_size(res);
> > + }
> > +
> > + scu_base =
> > syscon_regmap_lookup_by_compatible(scu.compatible);
> > + if (IS_ERR(scu_base)) {
> > + wdt->wdd.bootstatus = WDIOS_UNKNOWN;
> > + return;
> > + }
> > +
> > + ret = regmap_read(scu_base, scu.reset_status_reg, &status);
> > + if (ret) {
> > + wdt->wdd.bootstatus = WDIOS_UNKNOWN;
> > + return;
> > + }
> > +
> > + reset_mask_width = hweight32(scu.wdt_reset_mask);
> > + reset_mask_shift = scu.wdt_reset_mask_shift +
> > + reset_mask_width * idx;
> > +
> > + if (status & (scu.wdt_reset_mask << reset_mask_shift))
> > + wdt->wdd.bootstatus = WDIOF_CARDRESET;
>
> How precise is the wording in your commit message? The AST2600, for
> instance, has 4 reset types:
>
> 1. "SoC"
> 2. "Full"
> 3. "ARM"
> 4. "Software"
>
> In the commit message, you said:
>
> > WDIOF_CARDRESET => the system is reset by WDT SoC reset.
>
The commit message should be changed for more precisely.
"WDIOF_CARDRESET => the system is reset due to WDT timeout occurs."
This change can be updated with a new patch version.
> However the logic here (with the way you've initialised the config struct for the
> AST2600) will signal WDIOF_CARDRESET even if what occurred was e.g. a
> (graceful?) software reset.
>
As the discussion in the previous patch series, there is no consensus for graceful
reset flag in WDT framework. Thus, this patch only distinguishes WDT reset from
other reset reasons.
> The only thing WDIOF_CARDRESET differentiates is a power-on reset from any
> other kind of watchdog-driven reset.
>
Yes.
> Is that what's intended? I'm just wary of confusion for what appears to be a
> generic use of "SoC" in the commit message vs the specific "SoC"
> mode of the watchdog controller (should some documentation be
> written/updated?)
>
The commit message can be updated.
Chin-Ting
Powered by blists - more mailing lists