| Message ID | 1444163325-3041-1-git-send-email-sylvain.rochet@finsecur.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
Hi Sylvain, On 06/10/2015 at 22:28:45 +0200, Sylvain Rochet wrote : > Since turning on idle-halt in commit 5161b31dc39a (watchdog: > at91sam9_wdt: better watchdog support"), SoCs compatible with > at91sam9260-wdt not using a device tree no longer reboot if the watchdog > times out while the CPU is in idle state. Removing the > AT91_WDT_WDIDLEHLT flag that was set by default fixes this. > > Signed-off-by: Sylvain Rochet <sylvain.rochet@finsecur.com> > Fixes: 5161b31dc39a ("watchdog: at91sam9_wdt: better watchdog support") > Cc: <stable@vger.kernel.org> # 3.14+ Acked-by: Alexandre Belloni <alexandre.belloni@free-electrons.com> However, we don't really care about that for kernels after 3.18 as no users are using pdata anymore. I think you could send a follow-up patch removing pdata support completely.
Hi, On Wed, Oct 07, 2015 at 01:01:12PM +0200, Alexandre Belloni wrote: > On 06/10/2015 at 22:28:45 +0200, Sylvain Rochet wrote : > > Since turning on idle-halt in commit 5161b31dc39a (watchdog: > > at91sam9_wdt: better watchdog support"), SoCs compatible with > > at91sam9260-wdt not using a device tree no longer reboot if the watchdog > > times out while the CPU is in idle state. Removing the > > AT91_WDT_WDIDLEHLT flag that was set by default fixes this. > > > > Signed-off-by: Sylvain Rochet <sylvain.rochet@finsecur.com> > > Fixes: 5161b31dc39a ("watchdog: at91sam9_wdt: better watchdog support") > > Cc: <stable@vger.kernel.org> # 3.14+ > > Acked-by: Alexandre Belloni <alexandre.belloni@free-electrons.com> > > However, we don't really care about that for kernels after 3.18 as no > users are using pdata anymore. > I think you could send a follow-up patch removing pdata support > completely. Looks like this one falls through the cracks, it didn't reach mainline and therefore wasn't applied to stable branches. I just checked, it still apply properly on today's linux-next branch. Cheers, Sylvain
On 10/16/2016 08:33 AM, Sylvain Rochet wrote: > Hi, > > On Wed, Oct 07, 2015 at 01:01:12PM +0200, Alexandre Belloni wrote: >> On 06/10/2015 at 22:28:45 +0200, Sylvain Rochet wrote : >>> Since turning on idle-halt in commit 5161b31dc39a (watchdog: >>> at91sam9_wdt: better watchdog support"), SoCs compatible with >>> at91sam9260-wdt not using a device tree no longer reboot if the watchdog >>> times out while the CPU is in idle state. Removing the >>> AT91_WDT_WDIDLEHLT flag that was set by default fixes this. >>> >>> Signed-off-by: Sylvain Rochet <sylvain.rochet@finsecur.com> >>> Fixes: 5161b31dc39a ("watchdog: at91sam9_wdt: better watchdog support") >>> Cc: <stable@vger.kernel.org> # 3.14+ >> >> Acked-by: Alexandre Belloni <alexandre.belloni@free-electrons.com> >> >> However, we don't really care about that for kernels after 3.18 as no >> users are using pdata anymore. >> I think you could send a follow-up patch removing pdata support >> completely. > > Looks like this one falls through the cracks, it didn't reach mainline > and therefore wasn't applied to stable branches. > Possibly that happened because you did not copy the watchdog mailing list. Guenter > I just checked, it still apply properly on today's linux-next branch. > > Cheers, > Sylvain >
On 10/16/2016 08:50 AM, Guenter Roeck wrote: > On 10/16/2016 08:33 AM, Sylvain Rochet wrote: >> Hi, >> >> On Wed, Oct 07, 2015 at 01:01:12PM +0200, Alexandre Belloni wrote: >>> On 06/10/2015 at 22:28:45 +0200, Sylvain Rochet wrote : >>>> Since turning on idle-halt in commit 5161b31dc39a (watchdog: >>>> at91sam9_wdt: better watchdog support"), SoCs compatible with >>>> at91sam9260-wdt not using a device tree no longer reboot if the watchdog >>>> times out while the CPU is in idle state. Removing the >>>> AT91_WDT_WDIDLEHLT flag that was set by default fixes this. >>>> >>>> Signed-off-by: Sylvain Rochet <sylvain.rochet@finsecur.com> >>>> Fixes: 5161b31dc39a ("watchdog: at91sam9_wdt: better watchdog support") >>>> Cc: <stable@vger.kernel.org> # 3.14+ >>> >>> Acked-by: Alexandre Belloni <alexandre.belloni@free-electrons.com> >>> >>> However, we don't really care about that for kernels after 3.18 as no >>> users are using pdata anymore. >>> I think you could send a follow-up patch removing pdata support >>> completely. >> >> Looks like this one falls through the cracks, it didn't reach mainline >> and therefore wasn't applied to stable branches. >> > Possibly that happened because you did not copy the watchdog mailing list. > ... and your other patches were not sent to the watchdog mailing list either, so you should not expect them to be picked up either. Seriously, how do you expect _any_ patch to be picked up if you neither copy the subsystem mailing list nor the subsystem maintainer ? Guenter
Hi Guenter, On Sun, Oct 16, 2016 at 08:55:41AM -0700, Guenter Roeck wrote: > On 10/16/2016 08:50 AM, Guenter Roeck wrote: > > On 10/16/2016 08:33 AM, Sylvain Rochet wrote: > > > > > > Looks like this one falls through the cracks, it didn't reach mainline > > > and therefore wasn't applied to stable branches. > > > > Possibly that happened because you did not copy the watchdog mailing list. > > ... and your other patches were not sent to the watchdog mailing list either, > so you should not expect them to be picked up either. > > Seriously, how do you expect _any_ patch to be picked up if you neither copy > the subsystem mailing list nor the subsystem maintainer ? Whoops, thanks for the heads up, I wonder how I managed to mess up that at that time, I usually take care of that. The other series need respinning anyway so I can fix it for v2 (if any). Cheers, Sylvain
diff --git a/drivers/watchdog/at91sam9_wdt.c b/drivers/watchdog/at91sam9_wdt.c index 7e6acaf..6e3a167 100644 --- a/drivers/watchdog/at91sam9_wdt.c +++ b/drivers/watchdog/at91sam9_wdt.c @@ -339,7 +339,7 @@ static int __init at91wdt_probe(struct platform_device *pdev) return -ENOMEM; wdt->mr = (WDT_HW_TIMEOUT * 256) | AT91_WDT_WDRSTEN | AT91_WDT_WDD | - AT91_WDT_WDDBGHLT | AT91_WDT_WDIDLEHLT; + AT91_WDT_WDDBGHLT; wdt->mr_mask = 0x3FFFFFFF; wdt->nowayout = nowayout; wdt->wdd.parent = &pdev->dev;
Since turning on idle-halt in commit 5161b31dc39a (watchdog: at91sam9_wdt: better watchdog support"), SoCs compatible with at91sam9260-wdt not using a device tree no longer reboot if the watchdog times out while the CPU is in idle state. Removing the AT91_WDT_WDIDLEHLT flag that was set by default fixes this. Signed-off-by: Sylvain Rochet <sylvain.rochet@finsecur.com> Fixes: 5161b31dc39a ("watchdog: at91sam9_wdt: better watchdog support") Cc: <stable@vger.kernel.org> # 3.14+ --- drivers/watchdog/at91sam9_wdt.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)