Hi Bjorn,

Hi Georgi,

Thanks for working on this, unfortunately it seems like I need to port
the rpm regulators to be able to test this.
But I did find that...

Yes, the regulators support is still missing upstream. Until then, a basic
testing can be done by bypassing the regulators init part in order to use
the power on defaults for regulators and maybe also clocks.

On Fri, Sep 27, 2013 at 9:57 AM, Georgi Djakov [off-list ref] wrote:

quoted

...
+static irqreturn_t sdhci_msm_pwr_irq(int irq, void *data)
+{
+       struct sdhci_msm_host *msm_host = (struct sdhci_msm_host *)data;
...
+static int sdhci_msm_probe(struct platform_device *pdev)
+{
+       struct sdhci_host *host;
+       struct sdhci_msm_host *msm_host;
...
+       ret = devm_request_threaded_irq(&pdev->dev, msm_host->pwr_irq,
NULL,
+                                       sdhci_msm_pwr_irq, IRQF_ONESHOT,
+                                       dev_name(&pdev->dev), host);
...

...sdhci_msm_pwr_irq() expects the data pointer to be msm_host, not
sdhci_host, so right now I only get a panic. So I suspect you mean
msm_host as the last param here.

Yes, you are right, I have also noticed this and it is fixed in the next
version which i will be posting in the next few days. Thanks!

BR,
Georgi