[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Date: Sat, 4 Apr 2020 12:00:18 +1100
From: Evan Benn <evanbenn@...omium.org>
To: Julius Werner <jwerner@...omium.org>
Cc: LKML <linux-kernel@...r.kernel.org>,
Xingyu Chen <xingyu.chen@...ogic.com>,
Andy Shevchenko <andriy.shevchenko@...ux.intel.com>,
Anson Huang <Anson.Huang@....com>,
Bjorn Andersson <bjorn.andersson@...aro.org>,
Catalin Marinas <catalin.marinas@....com>,
"David S. Miller" <davem@...emloft.net>,
Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
Guenter Roeck <linux@...ck-us.net>,
Jonathan Cameron <Jonathan.Cameron@...wei.com>,
Leonard Crestez <leonard.crestez@....com>,
Manivannan Sadhasivam <manivannan.sadhasivam@...aro.org>,
Marcin Juszkiewicz <marcin.juszkiewicz@...aro.org>,
Matthias Brugger <matthias.bgg@...il.com>,
Mauro Carvalho Chehab <mchehab+huawei@...nel.org>,
Olof Johansson <olof@...om.net>, Rob Herring <robh@...nel.org>,
Shawn Guo <shawnguo@...nel.org>,
Valentin Schneider <valentin.schneider@....com>,
Vinod Koul <vkoul@...nel.org>, Will Deacon <will@...nel.org>,
Wim Van Sebroeck <wim@...ux-watchdog.org>,
"moderated list:ARM/FREESCALE IMX / MXC ARM ARCHITECTURE"
<linux-arm-kernel@...ts.infradead.org>,
"moderated list:ARM/Mediatek SoC support"
<linux-mediatek@...ts.infradead.org>,
LINUX-WATCHDOG <linux-watchdog@...r.kernel.org>
Subject: Re: [PATCH v2 2/2] watchdog: Add new arm_smd_wdt watchdog driver
On Sat, Apr 4, 2020 at 9:56 AM Julius Werner <jwerner@...omium.org> wrote:
>
> > + wdd->info = &smcwd_info;
> > + /* get_timeleft is optional */
> > + if (smcwd_call(SMCWD_GET_TIMELEFT, 0, NULL))
>
> How is this supposed to work? A firmware that implements this call
> would return the time left here which may not be 0 (maybe the watchdog
> was already primed by the bootloader or whatever), so smcwd_call()
> would interpret it as an error.
>
> I think the cleanest solution would be to stick to the same return
> codes in a0 and use a1 to report the time left when a0 is
> PSCI_SUCCESS. This is more consistent with SMCWD_INIT too.
Yes you are right, I have the wrong return code in the get_timeleft
implementation. It should use ->a1 for the actual timeleft, a0 is for
the error code.
Here smcwd_call returns the error code, which is NOT_IMPLEMENTED if
the firmware does not implement timeleft. The timeleft itself cannot
return error codes else we would just return that there I guess.
Powered by blists - more mailing lists