[PATCH v1 1/2] tpm: msleep() delays - replace with usleep_range() in i2c nuvoton driver

Commit 500462a9de65 "timers: Switch to a non-cascading wheel" replaced
the 'classic' timer wheel, which aimed for near 'exact' expiry of the
timers.  Their analysis was that the vast majority of timeout timers
are used as safeguards, not as real timers, and are cancelled or
rearmed before expiration.  The only exception noted to this were
networking timers with a small expiry time.

Not included in the analysis was the TPM polling timer, which resulted
in a longer normal delay and, every so often, a very long delay.  The
non-cascading wheel delay is based on CONFIG_HZ.  For a description of
the different rings and their delays, refer to the comments in
kernel/time/timer.c.

Below are the delays given for rings 0 - 2, which explains the longer
"normal" delays and the very, long delays as seen on systems with
CONFIG_HZ 250.

* HZ 1000 steps
 * Level Offset  Granularity            Range
 *  0      0         1 ms                0 ms - 63 ms
 *  1     64         8 ms               64 ms - 511 ms
 *  2    128        64 ms              512 ms - 4095 ms (512ms - ~4s)

* HZ  250
 * Level Offset  Granularity            Range
 *  0      0         4 ms                0 ms - 255 ms
 *  1     64        32 ms              256 ms - 2047 ms (256ms - ~2s)
 *  2    128       256 ms             2048 ms - 16383 ms (~2s - ~16s)

Below is a comparison of extending the TPM with 1000 measurements,
using msleep() vs. usleep_delay() when configured for 1000 hz vs. 250
hz, before and after commit 500462a9de65.

linux-4.7 | msleep() usleep_range()
1000 hz: 0m44.628s | 1m34.497s 29.243s
250 hz: 1m28.510s | 4m49.269s 32.386s

linux-4.7  | min-max (msleep)  min-max (usleep_range)
1000 hz: 0:017 - 2:760s | 0:015 - 3:967s    0:014 - 0:418s
250 hz: 0:028 - 1:954s | 0:040 - 4:096s    0:016 - 0:816s

This patch replaces the msleep() with usleep_range() calls in the
i2c nuvoton driver with a consistent max range value.

Signed-of-by: Mimi Zohar [off-list ref]
Cc: stable at vger.kernel.org (linux-4.8)
Signed-off-by: Nayna Jain <redacted>
---
Changelog v1:

- Included Jason's feedbacks related to #defines.

 drivers/char/tpm/tpm_i2c_nuvoton.c | 23 +++++++++++++++--------
 1 file changed, 15 insertions(+), 8 deletions(-)

diff --git a/drivers/char/tpm/tpm_i2c_nuvoton.c b/drivers/char/tpm/tpm_i2c_nuvoton.c
index e3a9155..0c98c42 100644
--- a/drivers/char/tpm/tpm_i2c_nuvoton.c
+++ b/drivers/char/tpm/tpm_i2c_nuvoton.c

@@ -49,9 +49,10 @@
  */
 #define TPM_I2C_MAX_BUF_SIZE           32
 #define TPM_I2C_RETRY_COUNT            32
-#define TPM_I2C_BUS_DELAY              1       /* msec */
-#define TPM_I2C_RETRY_DELAY_SHORT      2       /* msec */
-#define TPM_I2C_RETRY_DELAY_LONG       10      /* msec */
+#define TPM_I2C_BUS_DELAY              1000      	/* usec */
+#define TPM_I2C_RETRY_DELAY_SHORT      (2 * 1000)	/* usec */
+#define TPM_I2C_RETRY_DELAY_LONG       (10 * 1000) 	/* usec */
+#define TPM_I2C_DELAY_RANGE            300		/* usec */
 
 #define OF_IS_TPM2 ((void *)1)
 #define I2C_IS_TPM2 1

@@ -123,7 +124,8 @@ static s32 i2c_nuvoton_write_status(struct i2c_client *client, u8 data)
 	/* this causes the current command to be aborted */
 	for (i = 0, status = -1; i < TPM_I2C_RETRY_COUNT && status < 0; i++) {
 		status = i2c_nuvoton_write_buf(client, TPM_STS, 1, &data);
-		msleep(TPM_I2C_BUS_DELAY);
+		usleep_range(TPM_I2C_BUS_DELAY, TPM_I2C_BUS_DELAY
+			     + TPM_I2C_DELAY_RANGE);
 	}
 	return status;
 }

@@ -160,7 +162,8 @@ static int i2c_nuvoton_get_burstcount(struct i2c_client *client,
 			burst_count = min_t(u8, TPM_I2C_MAX_BUF_SIZE, data);
 			break;
 		}
-		msleep(TPM_I2C_BUS_DELAY);
+		usleep_range(TPM_I2C_BUS_DELAY, TPM_I2C_BUS_DELAY
+			     + TPM_I2C_DELAY_RANGE);
 	} while (time_before(jiffies, stop));
 
 	return burst_count;

@@ -203,13 +206,17 @@ static int i2c_nuvoton_wait_for_stat(struct tpm_chip *chip, u8 mask, u8 value,
 			return 0;
 
 		/* use polling to wait for the event */
-		ten_msec = jiffies + msecs_to_jiffies(TPM_I2C_RETRY_DELAY_LONG);
+		ten_msec = jiffies + usecs_to_jiffies(TPM_I2C_RETRY_DELAY_LONG);
 		stop = jiffies + timeout;
 		do {
 			if (time_before(jiffies, ten_msec))
-				msleep(TPM_I2C_RETRY_DELAY_SHORT);
+				usleep_range(TPM_I2C_RETRY_DELAY_SHORT,
+					     TPM_I2C_RETRY_DELAY_SHORT
+					     + TPM_I2C_DELAY_RANGE);
 			else
-				msleep(TPM_I2C_RETRY_DELAY_LONG);
+				usleep_range(TPM_I2C_RETRY_DELAY_LONG,
+					     TPM_I2C_RETRY_DELAY_LONG
+					     + TPM_I2C_DELAY_RANGE);
 			status_valid = i2c_nuvoton_check_status(chip, mask,
 								value);
 			if (status_valid)

-- 
2.9.3

--
To unsubscribe from this list: send the line "unsubscribe linux-security-module" in
the body of a message to majordomo at vger.kernel.org
More majordomo info@ http://vger.kernel.org/majordomo-info.html

Currently, there is an unnecessary 1 msec delay added in
i2c_nuvoton_write_status() for the successful case. This 
function is called multiple times during send() and recv(),
which implies adding multiple extra delays for every TPM
operation.

This patch calls usleep_range() only if retry is to be done.

Signed-off-by: Nayna Jain <redacted>
Cc: stable at vger.kernel.org (linux-4.8)
Reviewed-by: Mimi Zohar <redacted>
---
 drivers/char/tpm/tpm_i2c_nuvoton.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/drivers/char/tpm/tpm_i2c_nuvoton.c b/drivers/char/tpm/tpm_i2c_nuvoton.c
index 0c98c42..c642877 100644
--- a/drivers/char/tpm/tpm_i2c_nuvoton.c
+++ b/drivers/char/tpm/tpm_i2c_nuvoton.c

@@ -124,8 +124,9 @@ static s32 i2c_nuvoton_write_status(struct i2c_client *client, u8 data)
 	/* this causes the current command to be aborted */
 	for (i = 0, status = -1; i < TPM_I2C_RETRY_COUNT && status < 0; i++) {
 		status = i2c_nuvoton_write_buf(client, TPM_STS, 1, &data);
-		usleep_range(TPM_I2C_BUS_DELAY, TPM_I2C_BUS_DELAY
-			     + TPM_I2C_DELAY_RANGE);
+		if (status < 0)
+			usleep_range(TPM_I2C_BUS_DELAY, TPM_I2C_BUS_DELAY
+				     + TPM_I2C_DELAY_RANGE);
 	}
 	return status;
 }

-- 
2.9.3

--
To unsubscribe from this list: send the line "unsubscribe linux-security-module" in
the body of a message to majordomo at vger.kernel.org
More majordomo info@ http://vger.kernel.org/majordomo-info.html

Hi Jarkko,

On Fri, 2017-03-10 at 13:45 -0500, Nayna Jain wrote:

Currently, there is an unnecessary 1 msec delay added in
i2c_nuvoton_write_status() for the successful case. This 
function is called multiple times during send() and recv(),
which implies adding multiple extra delays for every TPM
operation.

This patch calls usleep_range() only if retry is to be done.

Signed-off-by: Nayna Jain <redacted>
Cc: stable at vger.kernel.org (linux-4.8)
Reviewed-by: Mimi Zohar <redacted>

Either Reviewed-by/Acked-by is fine. 

Can you pick up this patch and replace the original version of "tpm:
msleep() delays - replace with usleep_range()" with the one Nayna
posted?

Thanks!

Mimi

--
To unsubscribe from this list: send the line "unsubscribe linux-security-module" in
the body of a message to majordomo at vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

On Mon, Mar 13, 2017 at 10:54:22AM -0400, Mimi Zohar wrote:

Hi Jarkko,

On Fri, 2017-03-10 at 13:45 -0500, Nayna Jain wrote:

quoted

Currently, there is an unnecessary 1 msec delay added in
i2c_nuvoton_write_status() for the successful case. This 
function is called multiple times during send() and recv(),
which implies adding multiple extra delays for every TPM
operation.

This patch calls usleep_range() only if retry is to be done.

Signed-off-by: Nayna Jain <redacted>
Cc: stable at vger.kernel.org (linux-4.8)
Reviewed-by: Mimi Zohar <redacted>

Either Reviewed-by/Acked-by is fine. 

Can you pick up this patch and replace the original version of "tpm:
msleep() delays - replace with usleep_range()" with the one Nayna
posted?

Thanks!

Mimi

Yes, sure. I'll try to get that done as soon as possible..

/Jarkko

quoted

---
 drivers/char/tpm/tpm_i2c_nuvoton.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/drivers/char/tpm/tpm_i2c_nuvoton.c b/drivers/char/tpm/tpm_i2c_nuvoton.c
index 0c98c42..c642877 100644
--- a/drivers/char/tpm/tpm_i2c_nuvoton.c
+++ b/drivers/char/tpm/tpm_i2c_nuvoton.c

@@ -124,8 +124,9 @@ static s32 i2c_nuvoton_write_status(struct i2c_client *client, u8 data)
 	/* this causes the current command to be aborted */
 	for (i = 0, status = -1; i < TPM_I2C_RETRY_COUNT && status < 0; i++) {
 		status = i2c_nuvoton_write_buf(client, TPM_STS, 1, &data);
-		usleep_range(TPM_I2C_BUS_DELAY, TPM_I2C_BUS_DELAY
-			     + TPM_I2C_DELAY_RANGE);
+		if (status < 0)
+			usleep_range(TPM_I2C_BUS_DELAY, TPM_I2C_BUS_DELAY
+				     + TPM_I2C_DELAY_RANGE);
 	}
 	return status;
 }

--
To unsubscribe from this list: send the line "unsubscribe linux-security-module" in
the body of a message to majordomo at vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

On Fri, Mar 10, 2017 at 01:45:53PM -0500, Nayna Jain wrote:

Commit 500462a9de65 "timers: Switch to a non-cascading wheel" replaced
the 'classic' timer wheel, which aimed for near 'exact' expiry of the
timers.  Their analysis was that the vast majority of timeout timers
are used as safeguards, not as real timers, and are cancelled or
rearmed before expiration.  The only exception noted to this were
networking timers with a small expiry time.

Not included in the analysis was the TPM polling timer, which resulted
in a longer normal delay and, every so often, a very long delay.  The
non-cascading wheel delay is based on CONFIG_HZ.  For a description of
the different rings and their delays, refer to the comments in
kernel/time/timer.c.

Below are the delays given for rings 0 - 2, which explains the longer
"normal" delays and the very, long delays as seen on systems with
CONFIG_HZ 250.

* HZ 1000 steps
 * Level Offset  Granularity            Range
 *  0      0         1 ms                0 ms - 63 ms
 *  1     64         8 ms               64 ms - 511 ms
 *  2    128        64 ms              512 ms - 4095 ms (512ms - ~4s)

* HZ  250
 * Level Offset  Granularity            Range
 *  0      0         4 ms                0 ms - 255 ms
 *  1     64        32 ms              256 ms - 2047 ms (256ms - ~2s)
 *  2    128       256 ms             2048 ms - 16383 ms (~2s - ~16s)

Below is a comparison of extending the TPM with 1000 measurements,
using msleep() vs. usleep_delay() when configured for 1000 hz vs. 250
hz, before and after commit 500462a9de65.

linux-4.7 | msleep() usleep_range()
1000 hz: 0m44.628s | 1m34.497s 29.243s
250 hz: 1m28.510s | 4m49.269s 32.386s

linux-4.7  | min-max (msleep)  min-max (usleep_range)
1000 hz: 0:017 - 2:760s | 0:015 - 3:967s    0:014 - 0:418s
250 hz: 0:028 - 1:954s | 0:040 - 4:096s    0:016 - 0:816s

This patch replaces the msleep() with usleep_range() calls in the
i2c nuvoton driver with a consistent max range value.

Signed-of-by: Mimi Zohar [off-list ref]
Cc: stable at vger.kernel.org (linux-4.8)
Signed-off-by: Nayna Jain <redacted>
---
Changelog v1:

- Included Jason's feedbacks related to #defines.

What was changed?

/Jarkko
--
To unsubscribe from this list: send the line "unsubscribe linux-security-module" in
the body of a message to majordomo at vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

On 03/15/2017 09:22 PM, Jarkko Sakkinen wrote:

On Fri, Mar 10, 2017 at 01:45:53PM -0500, Nayna Jain wrote:

quoted

Commit 500462a9de65 "timers: Switch to a non-cascading wheel" replaced
the 'classic' timer wheel, which aimed for near 'exact' expiry of the
timers.  Their analysis was that the vast majority of timeout timers
are used as safeguards, not as real timers, and are cancelled or
rearmed before expiration.  The only exception noted to this were
networking timers with a small expiry time.

Not included in the analysis was the TPM polling timer, which resulted
in a longer normal delay and, every so often, a very long delay.  The
non-cascading wheel delay is based on CONFIG_HZ.  For a description of
the different rings and their delays, refer to the comments in
kernel/time/timer.c.

Below are the delays given for rings 0 - 2, which explains the longer
"normal" delays and the very, long delays as seen on systems with
CONFIG_HZ 250.

* HZ 1000 steps
  * Level Offset  Granularity            Range
  *  0      0         1 ms                0 ms - 63 ms
  *  1     64         8 ms               64 ms - 511 ms
  *  2    128        64 ms              512 ms - 4095 ms (512ms - ~4s)

* HZ  250
  * Level Offset  Granularity            Range
  *  0      0         4 ms                0 ms - 255 ms
  *  1     64        32 ms              256 ms - 2047 ms (256ms - ~2s)
  *  2    128       256 ms             2048 ms - 16383 ms (~2s - ~16s)

Below is a comparison of extending the TPM with 1000 measurements,
using msleep() vs. usleep_delay() when configured for 1000 hz vs. 250
hz, before and after commit 500462a9de65.

linux-4.7 | msleep() usleep_range()
1000 hz: 0m44.628s | 1m34.497s 29.243s
250 hz: 1m28.510s | 4m49.269s 32.386s

linux-4.7  | min-max (msleep)  min-max (usleep_range)
1000 hz: 0:017 - 2:760s | 0:015 - 3:967s    0:014 - 0:418s
250 hz: 0:028 - 1:954s | 0:040 - 4:096s    0:016 - 0:816s

This patch replaces the msleep() with usleep_range() calls in the
i2c nuvoton driver with a consistent max range value.

Signed-of-by: Mimi Zohar [off-list ref]
Cc: stable at vger.kernel.org (linux-4.8)
Signed-off-by: Nayna Jain <redacted>
---
Changelog v1:

- Included Jason's feedbacks related to #defines.

What was changed?

Changelog v1:
 >>
 >> - Included Jason's feedbacks related to #defines.

Based on Jason's review:
- Added () in #define
- Replaced hardcoded maximum range value with defined name.

Hmm.. could have included exact details.

Thanks & Regards,
- Nayna

/Jarkko

--
To unsubscribe from this list: send the line "unsubscribe linux-security-module" in
the body of a message to majordomo at vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

On Wed, Mar 15, 2017 at 09:51:47PM +0530, Nayna wrote:

On 03/15/2017 09:22 PM, Jarkko Sakkinen wrote:

quoted

On Fri, Mar 10, 2017 at 01:45:53PM -0500, Nayna Jain wrote:

quoted

Commit 500462a9de65 "timers: Switch to a non-cascading wheel" replaced
the 'classic' timer wheel, which aimed for near 'exact' expiry of the
timers.  Their analysis was that the vast majority of timeout timers
are used as safeguards, not as real timers, and are cancelled or
rearmed before expiration.  The only exception noted to this were
networking timers with a small expiry time.

Not included in the analysis was the TPM polling timer, which resulted
in a longer normal delay and, every so often, a very long delay.  The
non-cascading wheel delay is based on CONFIG_HZ.  For a description of
the different rings and their delays, refer to the comments in
kernel/time/timer.c.

Below are the delays given for rings 0 - 2, which explains the longer
"normal" delays and the very, long delays as seen on systems with
CONFIG_HZ 250.

* HZ 1000 steps
  * Level Offset  Granularity            Range
  *  0      0         1 ms                0 ms - 63 ms
  *  1     64         8 ms               64 ms - 511 ms
  *  2    128        64 ms              512 ms - 4095 ms (512ms - ~4s)

* HZ  250
  * Level Offset  Granularity            Range
  *  0      0         4 ms                0 ms - 255 ms
  *  1     64        32 ms              256 ms - 2047 ms (256ms - ~2s)
  *  2    128       256 ms             2048 ms - 16383 ms (~2s - ~16s)

Below is a comparison of extending the TPM with 1000 measurements,
using msleep() vs. usleep_delay() when configured for 1000 hz vs. 250
hz, before and after commit 500462a9de65.

linux-4.7 | msleep() usleep_range()
1000 hz: 0m44.628s | 1m34.497s 29.243s
250 hz: 1m28.510s | 4m49.269s 32.386s

linux-4.7  | min-max (msleep)  min-max (usleep_range)
1000 hz: 0:017 - 2:760s | 0:015 - 3:967s    0:014 - 0:418s
250 hz: 0:028 - 1:954s | 0:040 - 4:096s    0:016 - 0:816s

This patch replaces the msleep() with usleep_range() calls in the
i2c nuvoton driver with a consistent max range value.

Signed-of-by: Mimi Zohar [off-list ref]
Cc: stable at vger.kernel.org (linux-4.8)
Signed-off-by: Nayna Jain <redacted>
---
Changelog v1:

- Included Jason's feedbacks related to #defines.

What was changed?

Changelog v1:

quoted

quoted

- Included Jason's feedbacks related to #defines.

Based on Jason's review:
- Added () in #define
- Replaced hardcoded maximum range value with defined name.

Hmm.. could have included exact details.

Thanks & Regards,
- Nayna

quoted

/Jarkko

OK, I'll replace the original patch with this. Thanks.

/Jarkko
--
To unsubscribe from this list: send the line "unsubscribe linux-security-module" in
the body of a message to majordomo at vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

On Fri, Mar 10, 2017 at 01:45:54PM -0500, Nayna Jain wrote:

Currently, there is an unnecessary 1 msec delay added in
i2c_nuvoton_write_status() for the successful case. This 
function is called multiple times during send() and recv(),
which implies adding multiple extra delays for every TPM
operation.

This patch calls usleep_range() only if retry is to be done.

Signed-off-by: Nayna Jain <redacted>
Cc: stable at vger.kernel.org (linux-4.8)
Reviewed-by: Mimi Zohar <redacted>

I applied this too.

/Jarkko

--
To unsubscribe from this list: send the line "unsubscribe linux-security-module" in
the body of a message to majordomo at vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

I replaced the patch.

/Jarkko

On Fri, Mar 10, 2017 at 01:45:53PM -0500, Nayna Jain wrote:

--
To unsubscribe from this list: send the line "unsubscribe linux-security-module" in
the body of a message to majordomo at vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html