| Message ID | 1438335599-3301-19-git-send-email-alexandre.belloni@free-electrons.com (mailing list archive) |
|---|---|
| State | Not Applicable, archived |
| Headers | show |
Hi, On Fri, Jul 31, 2015 at 11:39:54AM +0200, Alexandre Belloni wrote: > Commit dca1a4b5ff6e ("clk: at91: keep slow clk enabled to prevent system > hang") added a workaround for the slow clock as it is not properly handled > by its users. > > Get and use the slow clock as it is necessary for the at91 reset > controller. You call clk_prepare_enable during probe without a matching clk_disable_unprepare on module removal. -- Sebastian
On 05/08/2015 at 20:22:20 +0200, Sebastian Reichel wrote : > Hi, > > On Fri, Jul 31, 2015 at 11:39:54AM +0200, Alexandre Belloni wrote: > > Commit dca1a4b5ff6e ("clk: at91: keep slow clk enabled to prevent system > > hang") added a workaround for the slow clock as it is not properly handled > > by its users. > > > > Get and use the slow clock as it is necessary for the at91 reset > > controller. > > You call clk_prepare_enable during probe without a matching > clk_disable_unprepare on module removal. > They will never be compiled as modules as the Kconfig option is a bool so we don't have a .remove and the module will never be removed. Do you want me to change at91-reset and at91-poweroff to work as modules?
Hi, On Mon, Aug 10, 2015 at 05:50:05PM +0200, Alexandre Belloni wrote: > On 05/08/2015 at 20:22:20 +0200, Sebastian Reichel wrote : > > On Fri, Jul 31, 2015 at 11:39:54AM +0200, Alexandre Belloni wrote: > > > Commit dca1a4b5ff6e ("clk: at91: keep slow clk enabled to prevent system > > > hang") added a workaround for the slow clock as it is not properly handled > > > by its users. > > > > > > Get and use the slow clock as it is necessary for the at91 reset > > > controller. > > > > You call clk_prepare_enable during probe without a matching > > clk_disable_unprepare on module removal. > > They will never be compiled as modules as the Kconfig option is a > bool so we don't have a .remove and the module will never be removed. In that case the module_platform_driver() statement at the bottom should be replaced with builtin_platform_driver(). > Do you want me to change at91-reset and at91-poweroff to work as > modules? That would also be fine. -- Sebastian
diff --git a/drivers/power/reset/at91-reset.c b/drivers/power/reset/at91-reset.c index 3225c8590ed0..19fdd4c107dc 100644 --- a/drivers/power/reset/at91-reset.c +++ b/drivers/power/reset/at91-reset.c @@ -11,6 +11,7 @@ * warranty of any kind, whether express or implied. */ +#include <linux/clk.h> #include <linux/io.h> #include <linux/module.h> #include <linux/of_address.h> @@ -46,6 +47,7 @@ enum reset_type { }; static void __iomem *at91_ramc_base[2], *at91_rstc_base; +static struct clk *sclk; /* * unless the SDRAM is cleanly shutdown before we hit the @@ -193,9 +195,21 @@ static int at91_reset_probe(struct platform_device *pdev) match = of_match_node(at91_reset_of_match, pdev->dev.of_node); at91_restart_nb.notifier_call = match->data; + sclk = devm_clk_get(&pdev->dev, NULL); + if (IS_ERR(sclk)) + return PTR_ERR(sclk); + + ret = clk_prepare_enable(sclk); + if (ret) { + dev_err(&pdev->dev, "Could not enable slow clock\n"); + return ret; + } + ret = register_restart_handler(&at91_restart_nb); - if (ret) + if (ret) { + clk_disable_unprepare(sclk); return ret; + } at91_reset_status(pdev);
Commit dca1a4b5ff6e ("clk: at91: keep slow clk enabled to prevent system hang") added a workaround for the slow clock as it is not properly handled by its users. Get and use the slow clock as it is necessary for the at91 reset controller. Signed-off-by: Alexandre Belloni <alexandre.belloni@free-electrons.com> --- Cc: Sebastian Reichel <sre@kernel.org> Cc: Dmitry Eremin-Solenikov <dbaryshkov@gmail.com> Cc: linux-pm@vger.kernel.org drivers/power/reset/at91-reset.c | 16 +++++++++++++++- 1 file changed, 15 insertions(+), 1 deletion(-)