Re: [PATCH] bus: mhi: pci_generic: Apply no-op for wake using inband wake support flag

Bhaumik Bhatt <bbhatt@xxxxxxxxxxxxxx> · Fri, 18 Jun 2021 10:41:34 -0700

On 2021-06-18 10:27 AM, Manivannan Sadhasivam wrote:
On Fri, Jun 18, 2021 at 10:06:28AM -0700, Bhaumik Bhatt wrote:
Hi Mani,
On 2021-06-17 11:52 PM, Manivannan Sadhasivam wrote:
> On Thu, Jun 17, 2021 at 11:23:53AM -0700, Bhaumik Bhatt wrote:
> > Devices such as SDX24 do not have the provision for inband wake
> > doorbell in the form of channel 127. Newer devices such as SDX55
> > or SDX65 have it by default. Ensure the functionality is used
> > based on this such that device wake stays held when a client
> > driver uses mhi_device_get() API or the equivalent debugfs entry.
> >
> > Fixes: e3e5e6508fc1 ("bus: mhi: pci_generic: No-Op for device_wake
> > operations")
> > Signed-off-by: Bhaumik Bhatt <bbhatt@xxxxxxxxxxxxxx>
> > ---
> >  drivers/bus/mhi/pci_generic.c | 26 ++++++++++++++++++--------
> >  1 file changed, 18 insertions(+), 8 deletions(-)
> >
> > diff --git a/drivers/bus/mhi/pci_generic.c
> > b/drivers/bus/mhi/pci_generic.c
> > index d84b743..31360a2 100644
> > --- a/drivers/bus/mhi/pci_generic.c
> > +++ b/drivers/bus/mhi/pci_generic.c
> > @@ -32,6 +32,7 @@
> >   * @edl: emergency download mode firmware path (if any)
> >   * @bar_num: PCI base address register to use for MHI MMIO register
> > space
> >   * @dma_data_width: DMA transfer word size (32 or 64 bits)
> > + * @no_inband_wake: Devices without inband wake support (such as
> > sdx24)
>
> I'd rather like this field to be "inband_wake" and set to false/true
> based on the capability of the devices. Rest looks good.
>
> Thanks,
> Mani
>
I should have known this was coming :)

Can I use sideband_wake instead of no_inband_wake and leave the 
booleans as
is?

Do you mean, the older devices uses a dedicated sideband GPIO for
controlling the wakeup? If so, what is it?

From what I know, this used to be the DEVICE_WAKE GPIO for SDX24. We 
moved to use
the channel 127 doorbell as WAKE doorbell for this purpose SDX55 onwards 
for ease
of deployment.

By default, inband_wake will have to be true for any and all devices 
moving
forward.

Right but still the functionality is not present in older devices. So I
think it makes sense to stay with "inband_wake".

Thanks,
Mani

So, by default we would not have to add explicit "true" for inband_wake 
as
newer devices will have the feature and controller drivers do not need 
to
populate the entry for a static const struct.

Please let me know your preference.
> >   */
> >  struct mhi_pci_dev_info {
> >  	const struct mhi_controller_config *config;
> > @@ -40,6 +41,7 @@ struct mhi_pci_dev_info {
> >  	const char *edl;
> >  	unsigned int bar_num;
> >  	unsigned int dma_data_width;
> > +	bool no_inband_wake;
> >  };
> >
> >  #define MHI_CHANNEL_CONFIG_UL(ch_num, ch_name, el_count, ev_ring) \
> > @@ -242,7 +244,8 @@ static const struct mhi_pci_dev_info
> > mhi_qcom_sdx65_info = {
> >  	.edl = "qcom/sdx65m/edl.mbn",
> >  	.config = &modem_qcom_v1_mhiv_config,
> >  	.bar_num = MHI_PCI_DEFAULT_BAR_NUM,
> > -	.dma_data_width = 32
> > +	.dma_data_width = 32,
> > +	.no_inband_wake = false
> >  };
> >
> >  static const struct mhi_pci_dev_info mhi_qcom_sdx55_info = {
> > @@ -251,7 +254,8 @@ static const struct mhi_pci_dev_info
> > mhi_qcom_sdx55_info = {
> >  	.edl = "qcom/sdx55m/edl.mbn",
> >  	.config = &modem_qcom_v1_mhiv_config,
> >  	.bar_num = MHI_PCI_DEFAULT_BAR_NUM,
> > -	.dma_data_width = 32
> > +	.dma_data_width = 32,
> > +	.no_inband_wake = false
> >  };
> >
> >  static const struct mhi_pci_dev_info mhi_qcom_sdx24_info = {
> > @@ -259,7 +263,8 @@ static const struct mhi_pci_dev_info
> > mhi_qcom_sdx24_info = {
> >  	.edl = "qcom/prog_firehose_sdx24.mbn",
> >  	.config = &modem_qcom_v1_mhiv_config,
> >  	.bar_num = MHI_PCI_DEFAULT_BAR_NUM,
> > -	.dma_data_width = 32
> > +	.dma_data_width = 32,
> > +	.no_inband_wake = true
> >  };
> >
> >  static const struct mhi_channel_config mhi_quectel_em1xx_channels[]
> > = {
> > @@ -301,7 +306,8 @@ static const struct mhi_pci_dev_info
> > mhi_quectel_em1xx_info = {
> >  	.edl = "qcom/prog_firehose_sdx24.mbn",
> >  	.config = &modem_quectel_em1xx_config,
> >  	.bar_num = MHI_PCI_DEFAULT_BAR_NUM,
> > -	.dma_data_width = 32
> > +	.dma_data_width = 32,
> > +	.no_inband_wake = true
> >  };
> >
> >  static const struct mhi_channel_config mhi_foxconn_sdx55_channels[]
> > = {
> > @@ -339,7 +345,8 @@ static const struct mhi_pci_dev_info
> > mhi_foxconn_sdx55_info = {
> >  	.edl = "qcom/sdx55m/edl.mbn",
> >  	.config = &modem_foxconn_sdx55_config,
> >  	.bar_num = MHI_PCI_DEFAULT_BAR_NUM,
> > -	.dma_data_width = 32
> > +	.dma_data_width = 32,
> > +	.no_inband_wake = false
> >  };
> >
> >  static const struct pci_device_id mhi_pci_id_table[] = {
> > @@ -640,9 +647,12 @@ static int mhi_pci_probe(struct pci_dev *pdev,
> > const struct pci_device_id *id)
> >  	mhi_cntrl->status_cb = mhi_pci_status_cb;
> >  	mhi_cntrl->runtime_get = mhi_pci_runtime_get;
> >  	mhi_cntrl->runtime_put = mhi_pci_runtime_put;
> > -	mhi_cntrl->wake_get = mhi_pci_wake_get_nop;
> > -	mhi_cntrl->wake_put = mhi_pci_wake_put_nop;
> > -	mhi_cntrl->wake_toggle = mhi_pci_wake_toggle_nop;
> > +
> > +	if (info->no_inband_wake) {
> > +		mhi_cntrl->wake_get = mhi_pci_wake_get_nop;
> > +		mhi_cntrl->wake_put = mhi_pci_wake_put_nop;
> > +		mhi_cntrl->wake_toggle = mhi_pci_wake_toggle_nop;
> > +	}
> >
> >  	err = mhi_pci_claim(mhi_cntrl, info->bar_num,
> > DMA_BIT_MASK(info->dma_data_width));
> >  	if (err)
> > --
> > The Qualcomm Innovation Center, Inc. is a member of the Code Aurora
> > Forum,
> > a Linux Foundation Collaborative Project
> >

Thanks,
Bhaumik
---
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora 
Forum,
a Linux Foundation Collaborative Project

Thanks,
Bhaumik
---
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora 
Forum,
a Linux Foundation Collaborative Project