2.6.7, amd64: PS/2 Mouse detection doesn't work

2.6.7, amd64: PS/2 Mouse detection doesn't work

[Harald Dunkel]

Hi folks,

Most of the times my mouse is detected as generic PS/2,
even though it is a Logitech. I have to reload the mousedev
and psmouse modules to make it work.

# grep -i mouse /var/log/kern.log
Jul 10 17:23:08 r101 kernel: input: PS/2 Generic Mouse on isa0060/serio1
Jul 10 17:23:08 r101 kernel: mice: PS/2 mouse device common for all mice
Jul 10 17:24:59 r101 kernel: input: PS/2 Generic Mouse on isa0060/serio1
Jul 10 17:24:59 r101 kernel: mice: PS/2 mouse device common for all mice
Jul 10 17:43:41 r101 kernel: input: PS2++ Logitech Mouse on isa0060/serio1
Jul 10 17:43:41 r101 kernel: mice: PS/2 mouse device common for all mice
Jul 10 17:46:59 r101 kernel: input: PS2++ Logitech Mouse on isa0060/serio1
Jul 10 17:47:00 r101 kernel: mice: PS/2 mouse device common for all mice
Jul 11 07:34:34 r101 kernel: input: PS/2 Generic Mouse on isa0060/serio1
Jul 11 07:34:34 r101 kernel: mice: PS/2 mouse device common for all mice
Jul 11 07:36:01 r101 kernel: input: PS/2 Generic Mouse on isa0060/serio1
Jul 11 07:36:01 r101 kernel: mice: PS/2 mouse device common for all mice
Jul 11 08:43:33 r101 kernel: mice: PS/2 mouse device common for all mice
Jul 11 08:43:33 r101 kernel: input: PS2++ Logitech Mouse on isa0060/serio1


Usually I wouldn't care, but I can go mad if the 4th mouse
button doesn't work :-).


Any idea?


Regards

Harri

Re: 2.6.7, amd64: PS/2 Mouse detection doesn't work

[Vojtech Pavlik]

[-- Attachment #1: Type: text/plain, Size: 1577 bytes --]

On Sun, Jul 11, 2004 at 09:00:22AM +0200, Harald Dunkel wrote:
> Hi folks,
> 
> Most of the times my mouse is detected as generic PS/2,
> even though it is a Logitech. I have to reload the mousedev
> and psmouse modules to make it work.
> 
> # grep -i mouse /var/log/kern.log
> Jul 10 17:23:08 r101 kernel: input: PS/2 Generic Mouse on isa0060/serio1
> Jul 10 17:23:08 r101 kernel: mice: PS/2 mouse device common for all mice
> Jul 10 17:24:59 r101 kernel: input: PS/2 Generic Mouse on isa0060/serio1
> Jul 10 17:24:59 r101 kernel: mice: PS/2 mouse device common for all mice
> Jul 10 17:43:41 r101 kernel: input: PS2++ Logitech Mouse on isa0060/serio1
> Jul 10 17:43:41 r101 kernel: mice: PS/2 mouse device common for all mice
> Jul 10 17:46:59 r101 kernel: input: PS2++ Logitech Mouse on isa0060/serio1
> Jul 10 17:47:00 r101 kernel: mice: PS/2 mouse device common for all mice
> Jul 11 07:34:34 r101 kernel: input: PS/2 Generic Mouse on isa0060/serio1
> Jul 11 07:34:34 r101 kernel: mice: PS/2 mouse device common for all mice
> Jul 11 07:36:01 r101 kernel: input: PS/2 Generic Mouse on isa0060/serio1
> Jul 11 07:36:01 r101 kernel: mice: PS/2 mouse device common for all mice
> Jul 11 08:43:33 r101 kernel: mice: PS/2 mouse device common for all mice
> Jul 11 08:43:33 r101 kernel: input: PS2++ Logitech Mouse on isa0060/serio1
> 
> 
> Usually I wouldn't care, but I can go mad if the 4th mouse
> button doesn't work :-).
> 
> Any idea?
 
Build the USB drivers into the kernel, or use the attached patch.
If it helps, please tell me.

-- 
Vojtech Pavlik
SuSE Labs, SuSE CR

[-- Attachment #2: psmouse-usb-fix-3 --]
[-- Type: text/plain, Size: 2719 bytes --]

ChangeSet@1.1831, 2004-04-03 23:30:39+02:00, vojtech@suse.cz
  input: Disable USB Legacy emulation in PCI quirks.


 quirks.c |   62 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 files changed, 62 insertions(+)


diff -Nru a/drivers/pci/quirks.c b/drivers/pci/quirks.c
--- a/drivers/pci/quirks.c	Sat Apr  3 23:30:49 2004
+++ b/drivers/pci/quirks.c	Sat Apr  3 23:30:49 2004
@@ -868,6 +868,68 @@
 }
 #endif /* CONFIG_SCSI_SATA */
 
+
+#define UHCI_USBLEGSUP		0xc0		/* legacy support */
+#define UHCI_USBCMD		0		/* command register */
+#define UHCI_USBINTR		4		/* interrupt register */
+#define UHCI_USBLEGSUP_DEFAULT	0x2000		/* only PIRQ enable set */
+#define UHCI_USBCMD_GRESET	0x0004		/* Global reset */
+
+#define OHCI_CONTROL		0x04
+#define OHCI_CMDSTATUS		0x08
+#define OHCI_INTRENABLE		0x10
+#define OHCI_OCR		(1 << 3)	/* ownership change request */
+#define OHCI_CTRL_IR		(1 << 8)	/* interrupt routing */
+#define OHCI_INTR_OC		(1 << 30)	/* ownership change */
+
+#if defined(__i386__) || defined(__x86_64__)
+static void __init quirk_usb_disable_smm_bios(struct pci_dev *pdev)
+{
+
+	if (pdev->class == ((PCI_CLASS_SERIAL_USB << 8) | 0x00)) { /* UHCI */
+		int i;
+		unsigned long base = 0;;
+
+		for (i = 0; i < PCI_ROM_RESOURCE; i++) 
+			if ((pci_resource_flags(pdev, i) & IORESOURCE_IO)) {
+				base = pci_resource_start(pdev, i);
+				break;
+			}
+
+		if (!base)
+			return;
+
+		outw(0, base + UHCI_USBINTR);
+		outw(UHCI_USBCMD_GRESET, base + UHCI_USBCMD);
+		set_current_state(TASK_UNINTERRUPTIBLE);
+       		schedule_timeout((HZ*50+999) / 1000);
+		outw(0, base + UHCI_USBCMD);
+		set_current_state(TASK_UNINTERRUPTIBLE);
+		schedule_timeout((HZ*10+999) / 1000);
+
+		pci_write_config_word(pdev, UHCI_USBLEGSUP, UHCI_USBLEGSUP_DEFAULT);
+	}
+
+	if (pdev->class == ((PCI_CLASS_SERIAL_USB << 8) | 0x10)) { /* OHCI */
+		char *base = ioremap_nocache(pci_resource_start(pdev, 0),
+							pci_resource_len(pdev, 0));
+		if (base == NULL) return;
+
+		if (readl(base + OHCI_CONTROL) & OHCI_CTRL_IR) {
+			int temp = 500;     /* arbitrary: five seconds */
+			writel(OHCI_INTR_OC, base + OHCI_INTRENABLE);
+			writel(OHCI_OCR, base + OHCI_CMDSTATUS);
+			while (temp && readl(base + OHCI_CONTROL) & OHCI_CTRL_IR) {
+				temp--;
+				set_current_state(TASK_UNINTERRUPTIBLE);
+				schedule_timeout( HZ / 100);
+			}
+		}
+		iounmap(base);
+	}
+}
+#endif
+
 /*
  *  The main table of quirks.
  *
@@ -983,6 +1043,10 @@
 	{ PCI_FIXUP_FINAL,      PCI_VENDOR_ID_INTEL,    PCI_ANY_ID,
 	  quirk_intel_ide_combined },
 #endif /* CONFIG_SCSI_SATA */
+
+#if defined(__i386__) || defined(__x86_64__)
+	{ PCI_FIXUP_FINAL,	PCI_ANY_ID,		PCI_ANY_ID,			quirk_usb_disable_smm_bios },
+#endif
 
 	{ 0 }
 };

Re: 2.6.7, amd64: PS/2 Mouse detection doesn't work

[Harald Dunkel]

Vojtech Pavlik wrote:
> On Sun, Jul 11, 2004 at 09:00:22AM +0200, Harald Dunkel wrote:
> 
>>
>>Usually I wouldn't care, but I can go mad if the 4th mouse
>>button doesn't work :-).
>>
>>Any idea?
> 
>  
> Build the USB drivers into the kernel, or use the attached patch.
> If it helps, please tell me.
> 
> 

I tried the patch: Seems to work. After a power cycle the
mouse was detected as

Jul 11 15:09:29 r101 kernel: input: PS2++ Logitech Mouse on isa0060/serio1
Jul 11 15:09:29 r101 kernel: mice: PS/2 mouse device common for all mice

as expected.


Many thanx

Harri

Re: 2.6.7, amd64: PS/2 Mouse detection doesn't work

[Harald Dunkel]

Vojtech Pavlik wrote:

>  
> Build the USB drivers into the kernel, or use the attached patch.
> If it helps, please tell me.
> 

The patch worked, as written before. Would it be possible
to permanently add it to the kernel for 2.6.8-rc3?


Many thanx

Harri

Re: 2.6.7, amd64: PS/2 Mouse detection doesn't work

[Harald Dunkel]

Hi folks,

Vojtech Pavlik wrote:
> On Sun, Jul 11, 2004 at 09:00:22AM +0200, Harald Dunkel wrote:
> 
>>Hi folks,
>>
>>Most of the times my mouse is detected as generic PS/2,
>>even though it is a Logitech. I have to reload the mousedev
>>and psmouse modules to make it work.
>>

Would it be possible to get Vojtech's patch in for 2.6.8?
Or are there some unpredictable side effects?


Many thanx

Harri

Re: 2.6.7, amd64: PS/2 Mouse detection doesn't work

[Harald Dunkel]

[-- Attachment #1: Type: text/plain, Size: 79 bytes --]

Hi folks,

Attached you can find an updated patch for 2.6.9.


Regards

Harri


[-- Attachment #2: psmouse-usb-fix-3.new --]
[-- Type: text/plain, Size: 2382 bytes --]

--- /tmp/linux-2.6.9/drivers/pci/quirks.c	2004-10-18 23:53:06.000000000 +0200
+++ linux-2.6.9/drivers/pci/quirks.c	2004-10-20 14:21:02.000000000 +0200
@@ -826,6 +826,69 @@
 	pci_read_config_byte(dev, 0x77, &val);
 }
 
+
+#define UHCI_USBLEGSUP		0xc0		/* legacy support */
+#define UHCI_USBCMD		0		/* command register */
+#define UHCI_USBINTR		4		/* interrupt register */
+#define UHCI_USBLEGSUP_DEFAULT	0x2000		/* only PIRQ enable set */
+#define UHCI_USBCMD_GRESET	0x0004		/* Global reset */
+
+#define OHCI_CONTROL		0x04
+#define OHCI_CMDSTATUS		0x08
+#define OHCI_INTRENABLE		0x10
+#define OHCI_OCR		(1 << 3)	/* ownership change request */
+#define OHCI_CTRL_IR		(1 << 8)	/* interrupt routing */
+#define OHCI_INTR_OC		(1 << 30)	/* ownership change */
+
+#if defined(__i386__) || defined(__x86_64__)
+static void __init quirk_usb_disable_smm_bios(struct pci_dev *pdev)
+{
+
+	if (pdev->class == ((PCI_CLASS_SERIAL_USB << 8) | 0x00)) { /* UHCI */
+		int i;
+		unsigned long base = 0;;
+
+		for (i = 0; i < PCI_ROM_RESOURCE; i++) 
+			if ((pci_resource_flags(pdev, i) & IORESOURCE_IO)) {
+				base = pci_resource_start(pdev, i);
+				break;
+			}
+
+		if (!base)
+			return;
+
+		outw(0, base + UHCI_USBINTR);
+		outw(UHCI_USBCMD_GRESET, base + UHCI_USBCMD);
+		set_current_state(TASK_UNINTERRUPTIBLE);
+       		schedule_timeout((HZ*50+999) / 1000);
+		outw(0, base + UHCI_USBCMD);
+		set_current_state(TASK_UNINTERRUPTIBLE);
+		schedule_timeout((HZ*10+999) / 1000);
+
+		pci_write_config_word(pdev, UHCI_USBLEGSUP, UHCI_USBLEGSUP_DEFAULT);
+	}
+
+	if (pdev->class == ((PCI_CLASS_SERIAL_USB << 8) | 0x10)) { /* OHCI */
+		char *base = ioremap_nocache(pci_resource_start(pdev, 0),
+							pci_resource_len(pdev, 0));
+		if (base == NULL) return;
+
+		if (readl(base + OHCI_CONTROL) & OHCI_CTRL_IR) {
+			int temp = 500;     /* arbitrary: five seconds */
+			writel(OHCI_INTR_OC, base + OHCI_INTRENABLE);
+			writel(OHCI_OCR, base + OHCI_CMDSTATUS);
+			while (temp && readl(base + OHCI_CONTROL) & OHCI_CTRL_IR) {
+				temp--;
+				set_current_state(TASK_UNINTERRUPTIBLE);
+				schedule_timeout( HZ / 100);
+			}
+		}
+		iounmap(base);
+	}
+}
+DECLARE_PCI_FIXUP_FINAL(PCI_ANY_ID,		PCI_ANY_ID,			quirk_usb_disable_smm_bios );
+#endif
+
 /*
  * ... This is further complicated by the fact that some SiS96x south
  * bridges pretend to be 85C503/5513 instead.  In that case see if we

Re: 2.6.7, amd64: PS/2 Mouse detection doesn't work

[Alan Cox]

On Mer, 2004-10-20 at 13:57, Harald Dunkel wrote:
> Hi folks,
> 
> Attached you can find an updated patch for 2.6.9.

You need to handle the USB reset quirks for OHCI and also catch any
escaped interrupts. It also seems you have to deal with UHCI (at least
my E750x fixes arent sufficient with a keyboard on an EHCI hub)

Re: 2.6.7, amd64: PS/2 Mouse detection doesn't work

[Harald Dunkel]

Alan Cox wrote:
> On Mer, 2004-10-20 at 13:57, Harald Dunkel wrote:
> 
>>Hi folks,
>>
>>Attached you can find an updated patch for 2.6.9.
> 
> 
> You need to handle the USB reset quirks for OHCI and also catch any
> escaped interrupts. It also seems you have to deal with UHCI (at least
> my E750x fixes arent sufficient with a keyboard on an EHCI hub)
> 

I am not a kernel developer. I just modified Vojtech's patch
for 2.6.9. His patch works for me since 2.6.7.

But it seems that UHCI _is_ handled, isn't it?


Regards

Harri

Re: 2.6.7, amd64: PS/2 Mouse detection doesn't work

[Alan Cox]

On Iau, 2004-10-21 at 09:11, Harald Dunkel wrote:
> I am not a kernel developer. I just modified Vojtech's patch
> for 2.6.9. His patch works for me since 2.6.7.
> 
> But it seems that UHCI _is_ handled, isn't it?

Braino .. I meant EHCI