| Message ID | 1708023665-1441674-1-git-send-email-radhey.shyam.pandey@amd.com (mailing list archive) |
|---|---|
| State | New |
| Headers | show |
| Series | dt-bindings: usb: dwc3: Add snps,enable_guctl1_ipd_quirk | expand |
On Fri, Feb 16, 2024 at 12:31:05AM +0530, Radhey Shyam Pandey wrote: > From: Piyush Mehta <piyush.mehta@amd.com> > > SNPS controller when configured in HOST mode maintains Inter Packet > Delay (IPD) of ~380ns which works with most of the super-speed hubs > except VIA-LAB hubs. When IPD is ~380ns HOST controller fails to > enumerate FS/LS devices when connected behind VIA-LAB hubs. > > To address the above issue, add 'snps,enable_guctl1_ipd_quirk' quirk, > This quirk set the bit 9 of GUCTL1 that enables the workaround in HW to > reduce the ULPI clock latency by 1 cycle, thus reducing the IPD (~360ns). > > Signed-off-by: Piyush Mehta <piyush.mehta@amd.com> > Signed-off-by: Radhey Shyam Pandey <radhey.shyam.pandey@amd.com> > --- > In the zynqmp public database GUCTL1 bit 9 is reserved but it is used to > enable a fix related to Inter Packet Delay in HW. The documentation team > is working to update GUCTL1 bit 9 description. Does this just affect the zynqmp? If it does, then you don't need a property - do this based on compatible. If it does affect other devices, what prevents the workaround being performed for all dwc3 controllers? Cheers, Conor. > --- > Documentation/devicetree/bindings/usb/snps,dwc3.yaml | 7 +++++++ > 1 file changed, 7 insertions(+) > > diff --git a/Documentation/devicetree/bindings/usb/snps,dwc3.yaml b/Documentation/devicetree/bindings/usb/snps,dwc3.yaml > index 8f5d250070c7..b226457a6e50 100644 > --- a/Documentation/devicetree/bindings/usb/snps,dwc3.yaml > +++ b/Documentation/devicetree/bindings/usb/snps,dwc3.yaml > @@ -280,6 +280,13 @@ properties: > xhci reset. And the vbus will back to 5V automatically when reset done. > type: boolean > > + snps,enable_guctl1_ipd_quirk: No underscores in properties please. > + description: > + When set, HW reduce the ULPI clock latency by 1 cycle, thus reducing > + the IPD (~360ns) and making controller enumerate FS/LS devices > + connected behind via-hubs. > + type: boolean > + > snps,is-utmi-l1-suspend: > description: > True when DWC3 asserts output signal utmi_l1_suspend_n, false when > -- > 2.34.1 >
> -----Original Message----- > From: Conor Dooley <conor@kernel.org> > Sent: Friday, February 16, 2024 1:09 AM > To: Pandey, Radhey Shyam <radhey.shyam.pandey@amd.com> > Cc: gregkh@linuxfoundation.org; robh@kernel.org; > krzysztof.kozlowski+dt@linaro.org; conor+dt@kernel.org; balbi@kernel.org; > Thinh.Nguyen@synopsys.com; linux-usb@vger.kernel.org; > devicetree@vger.kernel.org; linux-kernel@vger.kernel.org; git (AMD-Xilinx) > <git@amd.com>; Piyush Mehta <piyush.mehta@amd.com> > Subject: Re: [PATCH] dt-bindings: usb: dwc3: Add > snps,enable_guctl1_ipd_quirk > > On Fri, Feb 16, 2024 at 12:31:05AM +0530, Radhey Shyam Pandey wrote: > > From: Piyush Mehta <piyush.mehta@amd.com> > > > > SNPS controller when configured in HOST mode maintains Inter Packet > > Delay (IPD) of ~380ns which works with most of the super-speed hubs > > except VIA-LAB hubs. When IPD is ~380ns HOST controller fails to > > enumerate FS/LS devices when connected behind VIA-LAB hubs. > > > > To address the above issue, add 'snps,enable_guctl1_ipd_quirk' quirk, > > This quirk set the bit 9 of GUCTL1 that enables the workaround in HW > > to reduce the ULPI clock latency by 1 cycle, thus reducing the IPD (~360ns). > > > > Signed-off-by: Piyush Mehta <piyush.mehta@amd.com> > > Signed-off-by: Radhey Shyam Pandey <radhey.shyam.pandey@amd.com> > > --- > > In the zynqmp public database GUCTL1 bit 9 is reserved but it is used > > to enable a fix related to Inter Packet Delay in HW. The documentation > > team is working to update GUCTL1 bit 9 description. > > Does this just affect the zynqmp? > If it does, then you don't need a property - do this based on compatible. Yes this only affect zynqmp. I will apply IPD quirk based on zynqmp compatible. There is related discussion ongoing[1] on how to do it ? i.e in core or platform glue driver. Thinh suggested have it in platform code and I am working on it. So will drop this binding patch. [1]: https://lore.kernel.org/all/20240223230758.s7rodlxbsfa44frw@synopsys.com/ Thanks, Radhey > If it does affect other devices, what prevents the workaround being > performed for all dwc3 controllers? > > Cheers, > Conor. > > > --- > > Documentation/devicetree/bindings/usb/snps,dwc3.yaml | 7 +++++++ > > 1 file changed, 7 insertions(+) > > > > diff --git a/Documentation/devicetree/bindings/usb/snps,dwc3.yaml > > b/Documentation/devicetree/bindings/usb/snps,dwc3.yaml > > index 8f5d250070c7..b226457a6e50 100644 > > --- a/Documentation/devicetree/bindings/usb/snps,dwc3.yaml > > +++ b/Documentation/devicetree/bindings/usb/snps,dwc3.yaml > > @@ -280,6 +280,13 @@ properties: > > xhci reset. And the vbus will back to 5V automatically when reset done. > > type: boolean > > > > + snps,enable_guctl1_ipd_quirk: > > No underscores in properties please. > > > + description: > > + When set, HW reduce the ULPI clock latency by 1 cycle, thus reducing > > + the IPD (~360ns) and making controller enumerate FS/LS devices > > + connected behind via-hubs. > > + type: boolean > > + > > snps,is-utmi-l1-suspend: > > description: > > True when DWC3 asserts output signal utmi_l1_suspend_n, false > > when > > -- > > 2.34.1 > >
diff --git a/Documentation/devicetree/bindings/usb/snps,dwc3.yaml b/Documentation/devicetree/bindings/usb/snps,dwc3.yaml index 8f5d250070c7..b226457a6e50 100644 --- a/Documentation/devicetree/bindings/usb/snps,dwc3.yaml +++ b/Documentation/devicetree/bindings/usb/snps,dwc3.yaml @@ -280,6 +280,13 @@ properties: xhci reset. And the vbus will back to 5V automatically when reset done. type: boolean + snps,enable_guctl1_ipd_quirk: + description: + When set, HW reduce the ULPI clock latency by 1 cycle, thus reducing + the IPD (~360ns) and making controller enumerate FS/LS devices + connected behind via-hubs. + type: boolean + snps,is-utmi-l1-suspend: description: True when DWC3 asserts output signal utmi_l1_suspend_n, false when