Skip to contentSkip to conversation

tpm_tis: Re-enable interrupts upon resume

PATCH Superseded Stefan Berger 2 participants 4 messages Feb 22, 2011 – Mar 1, 2011 TPM device driver

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 tpm_tis: Re-enable interrupts upon resume

Below patch applies to the tip of the git tree.

This patch makes sure that if the TPM TIS interface is run in interrupt mode (rather than polling mode) that the interrupts are enabled in the TPM's interrupt enable register which may either have been cleared by the TPM's TIS loosing its state during device sleep in ACPI S3 (suspend) or by the BIOS, which upon resume sends a TPM_Startup() command to the TPM, and may run the TPM in polling mode and leave the TIS interrupts disabled once it transfers control to the OS again.

Problem is, I don't currently have a machine running the TPM in interrupt mode. I found this through a self-built TPM device model for Qemu and SeaBIOS patches, where this does resolve a problem upon resume.

You may want to check if your TPM runs with interrupts by doing

cat /proc/interrupts | grep -i tpm

and see whether there is an entry.

drivers/char/tpm/tpm_tis.c +17 -0
Index: linux-2.6/drivers/char/tpm/tpm_tis.c
===================================================================
@@ -651,6 +651,23 @@ static int tpm_tis_pnp_resume(struct pnp
 {
 	struct tpm_chip *chip = pnp_get_drvdata(dev);
 	int ret;
+	u32 intmask;
+
+	if (chip->vendor.irq) {
+		/* reenable interrupts that device may have lost or
+		   BIOS/firmware may have disabled */
+		intmask =
+		    ioread32(chip->vendor.iobase +
+			     TPM_INT_ENABLE(chip->vendor.locality));
+
+		intmask |= TPM_INTF_CMD_READY_INT
+		    | TPM_INTF_LOCALITY_CHANGE_INT | TPM_INTF_DATA_AVAIL_INT
+		    | TPM_INTF_STS_VALID_INT | TPM_GLOBAL_INT_ENABLE;
+
+		iowrite32(intmask,
+			  chip->vendor.iobase +
+			  TPM_INT_ENABLE(chip->vendor.locality));
+	}
 
 	ret = tpm_pm_resume(&dev->dev);
 	if (!ret)
#1 lore raw
Quoted 10 lines · Stefan Berger

Below patch applies to the tip of the git tree.

This patch makes sure that if the TPM TIS interface is run in interrupt mode (rather than polling mode) that the interrupts are enabled in the TPM's interrupt enable register which may either have been cleared by the TPM's TIS loosing its state during device sleep in ACPI S3 (suspend) or by the BIOS, which upon resume sends a TPM_Startup() command to the TPM, and may run the TPM in polling mode and leave the TIS interrupts disabled once it transfers control to the OS again.

Quoted 7 lines

Problem is, I don't currently have a machine running the TPM in interrupt mode. I found this through a self-built TPM device model for Qemu and SeaBIOS patches, where this does resolve a problem upon resume.

You may want to check if your TPM runs with interrupts by doing

cat /proc/interrupts | grep -i tpm

No, this is empty output.

regards,

Signature · 1 line
js
#2 lore raw
Quoted 11 lines · Jiri Slaby
Quoted 10 lines · Stefan Berger

Below patch applies to the tip of the git tree.

This patch makes sure that if the TPM TIS interface is run in interrupt mode (rather than polling mode) that the interrupts are enabled in the TPM's interrupt enable register which may either have been cleared by the TPM's TIS loosing its state during device sleep in ACPI S3 (suspend) or by the BIOS, which upon resume sends a TPM_Startup() command to the TPM, and may run the TPM in polling mode and leave the TIS interrupts disabled once it transfers control to the OS again.

Quoted 8 lines
Quoted 7 lines

Problem is, I don't currently have a machine running the TPM in interrupt mode. I found this through a self-built TPM device model for Qemu and SeaBIOS patches, where this does resolve a problem upon resume.

You may want to check if your TPM runs with interrupts by doing

cat /proc/interrupts | grep -i tpm

No, this is empty output.

Ok. That's what most TPM seem to do now -- they run in polling mode.

Stefan

#3 lore raw

PATCH v2 1/1 tpm_tis: Re-enable interrupts upon resume

In reply to Stefan Berger

v2:
- the patch was adapted to also work with a machine with a Intel TPM

Below patch applies to the tip of the git tree.

This patch makes sure that if the TPM TIS interface is run in interrupt mode (rather than polling mode) that the interrupts are enabled in the TPM's interrupt enable register which may either have been cleared by the TPM's TIS loosing its state during device sleep in ACPI S3 (suspend) or by the BIOS, which upon resume sends a TPM_Startup() command to the TPM, and may run the TPM in polling mode and leave the TIS interrupts disabled once it transfers control to the OS again.

You may want to check if your TPM runs with interrupts by doing

cat /proc/interrupts | grep -i tpm

and see whether there is an entry or otherwise for it to use interrupts:

modprobe tpm_tis interrupts=1 [add 'itpm=1' for Intel TPM ]

---
 drivers/char/tpm/tpm_tis.c |   31 +++++++++++++++++++++++++++++++
 1 file changed, 31 insertions(+)
drivers/char/tpm/tpm_tis.c +31 -0
Index: linux-2.6/drivers/char/tpm/tpm_tis.c
===================================================================
@@ -647,11 +647,37 @@ static int tpm_tis_pnp_suspend(struct pn
 	return tpm_pm_suspend(&dev->dev, msg);
 }
 
+static void tpm_tis_reenable_interrupts(struct tpm_chip *chip)
+{
+	u32 intmask;
+
+	/* reenable interrupts that device may have lost or
+	   BIOS/firmware may have disabled */
+	iowrite8(chip->vendor.irq, chip->vendor.iobase +
+		 TPM_INT_VECTOR(chip->vendor.locality));
+
+	intmask =
+	    ioread32(chip->vendor.iobase +
+		     TPM_INT_ENABLE(chip->vendor.locality));
+
+	intmask |= TPM_INTF_CMD_READY_INT
+	    | TPM_INTF_LOCALITY_CHANGE_INT | TPM_INTF_DATA_AVAIL_INT
+	    | TPM_INTF_STS_VALID_INT | TPM_GLOBAL_INT_ENABLE;
+
+	iowrite32(intmask,
+		  chip->vendor.iobase +
+		  TPM_INT_ENABLE(chip->vendor.locality));
+}
+
+
 static int tpm_tis_pnp_resume(struct pnp_dev *dev)
 {
 	struct tpm_chip *chip = pnp_get_drvdata(dev);
 	int ret;
 
+	if (chip->vendor.irq)
+		tpm_tis_reenable_interrupts(chip);
+
 	ret = tpm_pm_resume(&dev->dev);
 	if (!ret)
 		tpm_continue_selftest(chip);
@@ -704,6 +730,11 @@ static int tpm_tis_suspend(struct platfo
 
 static int tpm_tis_resume(struct platform_device *dev)
 {
+	struct tpm_chip *chip = dev_get_drvdata(&dev->dev);
+
+	if (chip->vendor.irq)
+		tpm_tis_reenable_interrupts(chip);
+
 	return tpm_pm_resume(&dev->dev);
 }
 static struct platform_driver tis_drv = {