| Message ID | 20200515140445.15218-1-ulf.hansson@linaro.org (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | [1/2] mmc: tmio: Further fixup runtime PM management at remove | expand |
On Fri, 15 May 2020 at 16:04, Ulf Hansson <ulf.hansson@linaro.org> wrote: > > Before calling tmio_mmc_host_probe(), the caller is required to enable > clocks for its device, as to make it accessible when reading/writing > registers during probe. > > Therefore, the responsibility to disable these clocks, in the error path of > ->probe() and during ->remove(), is better managed outside > tmio_mmc_host_remove(). As a matter of fact, callers of > tmio_mmc_host_remove() already expects this to be the behaviour. > > However, there's a problem with tmio_mmc_host_remove() when the Kconfig > option, CONFIG_PM, is set. More precisely, tmio_mmc_host_remove() may then > disable the clock via runtime PM, which leads to clock enable/disable > imbalance problems, when the caller of tmio_mmc_host_remove() also tries to > disable the same clocks. Just realized that part of the commit message is missing, explaining the change. I intend to add this: To solve the problem, let's make sure tmio_mmc_host_remove() leaves the device with clocks enabled, but also make sure to disable the IRQs, as we normally do at ->runtime_suspend(). Also, this is untested, so relying on you guys with the HW at hand. Kind regards Uffe > > Reported-by: Geert Uytterhoeven <geert+renesas@glider.be> > Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org> > --- > drivers/mmc/host/tmio_mmc_core.c | 6 ++++-- > 1 file changed, 4 insertions(+), 2 deletions(-) > > diff --git a/drivers/mmc/host/tmio_mmc_core.c b/drivers/mmc/host/tmio_mmc_core.c > index f31afd1c2671..ba301fb7656b 100644 > --- a/drivers/mmc/host/tmio_mmc_core.c > +++ b/drivers/mmc/host/tmio_mmc_core.c > @@ -1231,12 +1231,14 @@ void tmio_mmc_host_remove(struct tmio_mmc_host *host) > cancel_work_sync(&host->done); > cancel_delayed_work_sync(&host->delayed_reset_work); > tmio_mmc_release_dma(host); > + tmio_mmc_disable_mmc_irqs(host, TMIO_MASK_ALL); > > - pm_runtime_dont_use_autosuspend(&pdev->dev); > if (host->native_hotplug) > pm_runtime_put_noidle(&pdev->dev); > - pm_runtime_put_sync(&pdev->dev); > + > pm_runtime_disable(&pdev->dev); > + pm_runtime_dont_use_autosuspend(&pdev->dev); > + pm_runtime_put_noidle(&pdev->dev); > } > EXPORT_SYMBOL_GPL(tmio_mmc_host_remove); > > -- > 2.20.1 >
On Fri, May 15, 2020 at 05:02:08PM +0200, Ulf Hansson wrote: > On Fri, 15 May 2020 at 16:04, Ulf Hansson <ulf.hansson@linaro.org> wrote: > > > > Before calling tmio_mmc_host_probe(), the caller is required to enable > > clocks for its device, as to make it accessible when reading/writing > > registers during probe. > > > > Therefore, the responsibility to disable these clocks, in the error path of > > ->probe() and during ->remove(), is better managed outside > > tmio_mmc_host_remove(). As a matter of fact, callers of > > tmio_mmc_host_remove() already expects this to be the behaviour. > > > > However, there's a problem with tmio_mmc_host_remove() when the Kconfig > > option, CONFIG_PM, is set. More precisely, tmio_mmc_host_remove() may then > > disable the clock via runtime PM, which leads to clock enable/disable > > imbalance problems, when the caller of tmio_mmc_host_remove() also tries to > > disable the same clocks. > > Just realized that part of the commit message is missing, explaining > the change. I intend to add this: > > To solve the problem, let's make sure tmio_mmc_host_remove() leaves > the device with clocks enabled, but also make sure to disable the > IRQs, as we normally do at ->runtime_suspend(). Yeah, this paragraph is definately needed. > Also, this is untested, so relying on you guys with the HW at hand. I tested this on Gen3 M3-N (works fine!) and will next check your other patch. If this is okay, too, I will check them on Gen2 tomorrow before I give my Tested-by. Ulf, thanks for working on this, too!
On Fri, May 15, 2020 at 04:04:45PM +0200, Ulf Hansson wrote: > Before calling tmio_mmc_host_probe(), the caller is required to enable > clocks for its device, as to make it accessible when reading/writing > registers during probe. > > Therefore, the responsibility to disable these clocks, in the error path of > ->probe() and during ->remove(), is better managed outside > tmio_mmc_host_remove(). As a matter of fact, callers of > tmio_mmc_host_remove() already expects this to be the behaviour. > > However, there's a problem with tmio_mmc_host_remove() when the Kconfig > option, CONFIG_PM, is set. More precisely, tmio_mmc_host_remove() may then > disable the clock via runtime PM, which leads to clock enable/disable > imbalance problems, when the caller of tmio_mmc_host_remove() also tries to > disable the same clocks. > > Reported-by: Geert Uytterhoeven <geert+renesas@glider.be> > Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org> Works fine on my Lager board, too. Imbalance is gone when rebinding, and I can access large files on the SD card. Reviewed-by: Wolfram Sang <wsa+renesas@sang-engineering.com> Tested-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
diff --git a/drivers/mmc/host/tmio_mmc_core.c b/drivers/mmc/host/tmio_mmc_core.c index f31afd1c2671..ba301fb7656b 100644 --- a/drivers/mmc/host/tmio_mmc_core.c +++ b/drivers/mmc/host/tmio_mmc_core.c @@ -1231,12 +1231,14 @@ void tmio_mmc_host_remove(struct tmio_mmc_host *host) cancel_work_sync(&host->done); cancel_delayed_work_sync(&host->delayed_reset_work); tmio_mmc_release_dma(host); + tmio_mmc_disable_mmc_irqs(host, TMIO_MASK_ALL); - pm_runtime_dont_use_autosuspend(&pdev->dev); if (host->native_hotplug) pm_runtime_put_noidle(&pdev->dev); - pm_runtime_put_sync(&pdev->dev); + pm_runtime_disable(&pdev->dev); + pm_runtime_dont_use_autosuspend(&pdev->dev); + pm_runtime_put_noidle(&pdev->dev); } EXPORT_SYMBOL_GPL(tmio_mmc_host_remove);
Before calling tmio_mmc_host_probe(), the caller is required to enable clocks for its device, as to make it accessible when reading/writing registers during probe. Therefore, the responsibility to disable these clocks, in the error path of ->probe() and during ->remove(), is better managed outside tmio_mmc_host_remove(). As a matter of fact, callers of tmio_mmc_host_remove() already expects this to be the behaviour. However, there's a problem with tmio_mmc_host_remove() when the Kconfig option, CONFIG_PM, is set. More precisely, tmio_mmc_host_remove() may then disable the clock via runtime PM, which leads to clock enable/disable imbalance problems, when the caller of tmio_mmc_host_remove() also tries to disable the same clocks. Reported-by: Geert Uytterhoeven <geert+renesas@glider.be> Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org> --- drivers/mmc/host/tmio_mmc_core.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-)