+Rafael and acpi-ml On Fri, May 18, 2018 at 01:40:00PM +0300, Heikki Krogerus wrote: > Hi Paul, > > On Thu, May 17, 2018 at 03:34:49PM +0200, Paul Menzel wrote: > > Dear Heikki, > > > > Note, this problem also exists on the Dell XPS 13 9360 (Kaby Lake) device. > > OK. Good to know. Though, we don't have one of those either. > > > On 05/17/18 10:59, Heikki Krogerus wrote: > > > On Wed, May 16, 2018 at 04:13:31PM +0000, Mario.Limonciello@dell.com wrote: > > > > > > > > > > > > > -----Original Message----- > > > > > From: Heikki Krogerus [mailto:heikki.krogerus@linux.intel.com] > > > > > Sent: Wednesday, May 16, 2018 6:58 AM > > > > > To: Greg KH; Paul Menzel > > > > > Cc: linux-usb@vger.kernel.org; linux-kernel@vger.kernel.org; Limonciello, Mario > > > > > Subject: Re: `ucsi_acpi: probe of USBC000:00 failed with error -12` on Dell XPS 13 > > > > > 9370 > > > > > > > > > > Hi, > > > > > > > > > > On Wed, May 16, 2018 at 10:02:26AM +0200, Greg KH wrote: > > > > > > On Tue, May 15, 2018 at 06:47:37PM +0200, Paul Menzel wrote: > > > > > > > Dear Greg, > > > > > > > > > > > > > > > > > > > > > As always, thank you for the prompt response. > > > > > > > > > > > > > > > > > > > > > On 05/15/18 18:00, Greg KH wrote: > > > > > > > > On Tue, May 15, 2018 at 04:34:03PM +0200, Paul Menzel wrote: > > > > > > > > > > > > > > > > Linux 4.17-rc5 shows the error below on the Dell XPS 13 9370 with Debian > > > > > > > > > Sid/unstable. > > > > > > > > > > > > > > > > > > ``` > > > > > > > > > [???] > > > > > > > > > [ 0.440240] usb: port power management may be unreliable > > > > > > > > > [ 0.441358] usbcore: registered new interface driver usb-storage > > > > > > > > > [ 0.441367] usbcore: registered new interface driver usbserial_generic > > > > > > > > > [ 0.441369] usbserial: USB Serial support registered for generic > > > > > > > > > [ 0.441383] ioremap error for 0x3f799000-0x3f79a000, requested 0x2, got > > > > > > > > > 0x0 > > > > > > > > > [ 0.441518] ucsi_acpi: probe of USBC000:00 failed with error -12 > > > > > > > > > [???] > > > > > > > > > ``` > > > > > > > > > > > > > > > > > > 1. Are the ioremap and ucsi_acpi error related or is a separate report > > > > > > > > > needed? > > > > > > > > > > > > > > > > The ioremap error is what causes ucsi_acpi to fail the probe call (-12 > > > > > > > > is "out of memory".) > > > > > > > > > > > > > > > > > 2. Do you know the reason for the ucsi_acpi error? > > > > > > > > > > > > > > > > the call to ioremap failed. > > > > > > > > > > > > > > > > Does this device really have a working typec connector? > > > > > > > > > > > > > > Just to avoid misunderstandings, no device was connected to the laptop > > > > > > > during my test. > > > > > > > > > > > > > > But, from other boots, the Dell docking station TB16 kind of works with it, > > > > > > > so I???d say the USB Type-C connector is working. > > > > > > > > > > > > Ok, good, this might just be the acpi tables not set up properly for > > > > > > this type of connection. Odd that the tables show it should work, > > > > > > Heikki should know more about this. > > > > > > > > > > The firmware probable has not implemented UCSI on this board. I think > > > > > Dell always supplies the ACPI device node for UCSI in their acpi > > > > > tables. The _STA method in that device node is then used to inform the > > > > > OS if the interface exists or not. The return value for _STA comes > > > > > probable from BIOS, so this is most likely a BIOS problem. > > > > > > > > Heikki, > > > > > > > > I confirmed with internal team that UCSI is implemented on XPS 9370 > > > > and was confirmed to be working properly with Windows 10 RS2+. > > > > > > Just to double check: "UCSI was confirmed working properly", so not > > > "the Type-C ports were confirmed working properly"? > > > > > > > The reason that _STA is responding on this device node now but wasn't > > > > previously is it wasn't exposed in Linux until 4.16 when the Win 10 RS2 > > > > OSI string started to respond. > > > > > > OK. > > > > > > > Intel should internally have some XPS 9370 you can remotely access if > > > > you would like to poke around ACPI tables some. > > > > > > I will try get access to XPS 9370, but with the acpi tables, if > > > somebody could just send me acpidump, that would be enough: > > > > > > % acpidump -o xps9370_acpi.dump > > > > I uploaded them to another bug report [1]. Please find it attached. > > Thanks! At this point I need to ask help from ACPI experts. Rafael, the problem here with these Dell laptops is that a memory page that is used as a mailbox for special communication between EC FW and OS (called UCSI mailbox) should be in uncached memory, but the memory is marked as cached. I'm actually not even sure if the cache mode is something that is set based on information from ACPI tables, or is it something that the BIOS handles separately? If it's information coming from ACPI tables, can we help pin point the place the BIOS guys at Dell should fix? Thanks, -- heikki