dark0707
/
px4-firmware
forked from Archive/PX4-Autopilot
1
0
Fork 0
Code Pull Requests Activity
px4-firmware/nuttx-patches/00014-BACKPORT-i2c-hotfix.p...

326 lines
9.8 KiB
Diff
Raw Normal View History

HOTFIX:Backport of upstream NuttX i2c fix 5a6d95dd9f051be548a8d2378aaef75f0a1ba5e1 and ee5ae3a57dbbe835584f164c956e0374da1ed2eb Applies to STM32F4 and STM32F7 Save elapsed time before handling I2C in stm32_i2c_sem_waitdone() It is possible that a context switch occurs after stm32_i2c_isr() call but before elapsed time is saved in stm32_i2c_sem_waitdone(). It is then possible that the handling code was executed only once with "elapsed time" equal 0. When scheduler resumes this thread it is quite possible that now "elapsed time" will be well above timeout threshold. In that case the function returns and reports a timeout, even though the handling code was not executed "recently". Fix this by inverting the order of operations in the loop - save elapsed time before handling I2C. This way a context switch anywhere in the loop will not cause an erroneous "timeout" error.
2017-03-09 14:20:41 -04:00
diff --git NuttX/nuttx/arch/arm/src/stm32/stm32_i2c.c NuttX/nuttx/arch/arm/src/stm32/stm32_i2c.c
HOTFIX:Backport of second round upstream NuttX i2c fix (#6771) backport 3cd66af889b42b036d6c9d88e067fc3b8abbdb2a and pr 258 Applies to STM32F4 and STM32F7 STM32, STM32 F7, and STM32 L4: Clone Freddie Chopin's I2C change to similar STM32 I2C drivers. Save elapsed time before handling I2C in stm32_i2c_sem_waitstop() This patch follows the same logic as in previous fix to stm32_i2c_sem_waitdone(). It is possible that a context switch occurs after I2C registers are read but before elapsed time is saved in stm32_i2c_sem_waitstop(). It is then possible that the registers were read only once with "elapsed time" equal 0. When scheduler resumes this thread it is quite possible that now "elapsed time" will be well above timeout threshold. In that case the function returns and reports a timeout, even though the registers were not read "recently". Fix this by inverting the order of operations in the loop - save elapsed time before reading registers. This way a context switch anywhere in the loop will not cause an erroneous "timeout" error.
2017-03-10 13:26:51 -04:00
index 631ba66..669e0db 100644
HOTFIX:Backport of upstream NuttX i2c fix 5a6d95dd9f051be548a8d2378aaef75f0a1ba5e1 and ee5ae3a57dbbe835584f164c956e0374da1ed2eb Applies to STM32F4 and STM32F7 Save elapsed time before handling I2C in stm32_i2c_sem_waitdone() It is possible that a context switch occurs after stm32_i2c_isr() call but before elapsed time is saved in stm32_i2c_sem_waitdone(). It is then possible that the handling code was executed only once with "elapsed time" equal 0. When scheduler resumes this thread it is quite possible that now "elapsed time" will be well above timeout threshold. In that case the function returns and reports a timeout, even though the handling code was not executed "recently". Fix this by inverting the order of operations in the loop - save elapsed time before handling I2C. This way a context switch anywhere in the loop will not cause an erroneous "timeout" error.
2017-03-09 14:20:41 -04:00
--- NuttX/nuttx/arch/arm/src/stm32/stm32_i2c.c
+++ NuttX/nuttx/arch/arm/src/stm32/stm32_i2c.c
@@ -7,7 +7,7 @@
*
* With extensions, modifications by:
*
- * Copyright (C) 2011-2014, 2016 Gregory Nutt. All rights reserved.
+ * Copyright (C) 2011-2014, 2016-2017 Gregory Nutt. All rights reserved.
* Author: Gregory Nutt <gnutt@nuttx.org>
*
* Redistribution and use in source and binary forms, with or without
@@ -670,15 +670,15 @@ static inline int stm32_i2c_sem_waitdone(FAR struct stm32_i2c_priv_s *priv)
do
{
+ /* Calculate the elapsed time */
+
+ elapsed = clock_systimer() - start;
+
/* Poll by simply calling the timer interrupt handler until it
* reports that it is done.
*/
stm32_i2c_isr(priv);
-
- /* Calculate the elapsed time */
-
- elapsed = clock_systimer() - start;
}
/* Loop until the transfer is complete. */
HOTFIX:Backport of second round upstream NuttX i2c fix (#6771) backport 3cd66af889b42b036d6c9d88e067fc3b8abbdb2a and pr 258 Applies to STM32F4 and STM32F7 STM32, STM32 F7, and STM32 L4: Clone Freddie Chopin's I2C change to similar STM32 I2C drivers. Save elapsed time before handling I2C in stm32_i2c_sem_waitstop() This patch follows the same logic as in previous fix to stm32_i2c_sem_waitdone(). It is possible that a context switch occurs after I2C registers are read but before elapsed time is saved in stm32_i2c_sem_waitstop(). It is then possible that the registers were read only once with "elapsed time" equal 0. When scheduler resumes this thread it is quite possible that now "elapsed time" will be well above timeout threshold. In that case the function returns and reports a timeout, even though the registers were not read "recently". Fix this by inverting the order of operations in the loop - save elapsed time before reading registers. This way a context switch anywhere in the loop will not cause an erroneous "timeout" error.
2017-03-10 13:26:51 -04:00
@@ -729,6 +729,10 @@ static inline void stm32_i2c_sem_waitstop(FAR struct stm32_i2c_priv_s *priv)
start = clock_systimer();
do
{
+ /* Calculate the elapsed time */
+
+ elapsed = clock_systimer() - start;
+
/* Check for STOP condition */
cr1 = stm32_i2c_getreg(priv, STM32_I2C_CR1_OFFSET);
@@ -744,10 +748,6 @@ static inline void stm32_i2c_sem_waitstop(FAR struct stm32_i2c_priv_s *priv)
{
return;
}
-
- /* Calculate the elapsed time */
-
- elapsed = clock_systimer() - start;
}
/* Loop until the stop is complete or a timeout occurs. */
HOTFIX:Backport of upstream NuttX i2c fix 5a6d95dd9f051be548a8d2378aaef75f0a1ba5e1 and ee5ae3a57dbbe835584f164c956e0374da1ed2eb Applies to STM32F4 and STM32F7 Save elapsed time before handling I2C in stm32_i2c_sem_waitdone() It is possible that a context switch occurs after stm32_i2c_isr() call but before elapsed time is saved in stm32_i2c_sem_waitdone(). It is then possible that the handling code was executed only once with "elapsed time" equal 0. When scheduler resumes this thread it is quite possible that now "elapsed time" will be well above timeout threshold. In that case the function returns and reports a timeout, even though the handling code was not executed "recently". Fix this by inverting the order of operations in the loop - save elapsed time before handling I2C. This way a context switch anywhere in the loop will not cause an erroneous "timeout" error.
2017-03-09 14:20:41 -04:00
diff --git NuttX/nuttx/arch/arm/src/stm32/stm32_i2c_alt.c NuttX/nuttx/arch/arm/src/stm32/stm32_i2c_alt.c
HOTFIX:Backport of second round upstream NuttX i2c fix (#6771) backport 3cd66af889b42b036d6c9d88e067fc3b8abbdb2a and pr 258 Applies to STM32F4 and STM32F7 STM32, STM32 F7, and STM32 L4: Clone Freddie Chopin's I2C change to similar STM32 I2C drivers. Save elapsed time before handling I2C in stm32_i2c_sem_waitstop() This patch follows the same logic as in previous fix to stm32_i2c_sem_waitdone(). It is possible that a context switch occurs after I2C registers are read but before elapsed time is saved in stm32_i2c_sem_waitstop(). It is then possible that the registers were read only once with "elapsed time" equal 0. When scheduler resumes this thread it is quite possible that now "elapsed time" will be well above timeout threshold. In that case the function returns and reports a timeout, even though the registers were not read "recently". Fix this by inverting the order of operations in the loop - save elapsed time before reading registers. This way a context switch anywhere in the loop will not cause an erroneous "timeout" error.
2017-03-10 13:26:51 -04:00
index 545a647..981edb6 100644
HOTFIX:Backport of upstream NuttX i2c fix 5a6d95dd9f051be548a8d2378aaef75f0a1ba5e1 and ee5ae3a57dbbe835584f164c956e0374da1ed2eb Applies to STM32F4 and STM32F7 Save elapsed time before handling I2C in stm32_i2c_sem_waitdone() It is possible that a context switch occurs after stm32_i2c_isr() call but before elapsed time is saved in stm32_i2c_sem_waitdone(). It is then possible that the handling code was executed only once with "elapsed time" equal 0. When scheduler resumes this thread it is quite possible that now "elapsed time" will be well above timeout threshold. In that case the function returns and reports a timeout, even though the handling code was not executed "recently". Fix this by inverting the order of operations in the loop - save elapsed time before handling I2C. This way a context switch anywhere in the loop will not cause an erroneous "timeout" error.
2017-03-09 14:20:41 -04:00
--- NuttX/nuttx/arch/arm/src/stm32/stm32_i2c_alt.c
+++ NuttX/nuttx/arch/arm/src/stm32/stm32_i2c_alt.c
@@ -7,7 +7,7 @@
*
* With extensions, modifications by:
*
- * Copyright (C) 2011-2014, 2016 Gregory Nutt. All rights reserved.
+ * Copyright (C) 2011-2014, 2016-2017 Gregory Nutt. All rights reserved.
* Author: Gregory Nutt <gnutt@nuttx.org>
*
* Copyright( C) 2014 Patrizio Simona. All rights reserved.
@@ -678,15 +678,15 @@ static int stm32_i2c_sem_waitdone(FAR struct stm32_i2c_priv_s *priv)
do
{
+ /* Calculate the elapsed time */
+
+ elapsed = clock_systimer() - start;
+
/* Poll by simply calling the timer interrupt handler until it
* reports that it is done.
*/
stm32_i2c_isr(priv);
-
- /* Calculate the elapsed time */
-
- elapsed = clock_systimer() - start;
}
/* Loop until the transfer is complete. */
HOTFIX:Backport of second round upstream NuttX i2c fix (#6771) backport 3cd66af889b42b036d6c9d88e067fc3b8abbdb2a and pr 258 Applies to STM32F4 and STM32F7 STM32, STM32 F7, and STM32 L4: Clone Freddie Chopin's I2C change to similar STM32 I2C drivers. Save elapsed time before handling I2C in stm32_i2c_sem_waitstop() This patch follows the same logic as in previous fix to stm32_i2c_sem_waitdone(). It is possible that a context switch occurs after I2C registers are read but before elapsed time is saved in stm32_i2c_sem_waitstop(). It is then possible that the registers were read only once with "elapsed time" equal 0. When scheduler resumes this thread it is quite possible that now "elapsed time" will be well above timeout threshold. In that case the function returns and reports a timeout, even though the registers were not read "recently". Fix this by inverting the order of operations in the loop - save elapsed time before reading registers. This way a context switch anywhere in the loop will not cause an erroneous "timeout" error.
2017-03-10 13:26:51 -04:00
@@ -737,6 +737,10 @@ static inline void stm32_i2c_sem_waitstop(FAR struct stm32_i2c_priv_s *priv)
start = clock_systimer();
do
{
+ /* Calculate the elapsed time */
+
+ elapsed = clock_systimer() - start;
+
/* Check for STOP condition */
cr1 = stm32_i2c_getreg(priv, STM32_I2C_CR1_OFFSET);
@@ -753,9 +757,6 @@ static inline void stm32_i2c_sem_waitstop(FAR struct stm32_i2c_priv_s *priv)
return;
}
- /* Calculate the elapsed time */
-
- elapsed = clock_systimer() - start;
}
/* Loop until the stop is complete or a timeout occurs. */
HOTFIX:Backport of upstream NuttX i2c fix 5a6d95dd9f051be548a8d2378aaef75f0a1ba5e1 and ee5ae3a57dbbe835584f164c956e0374da1ed2eb Applies to STM32F4 and STM32F7 Save elapsed time before handling I2C in stm32_i2c_sem_waitdone() It is possible that a context switch occurs after stm32_i2c_isr() call but before elapsed time is saved in stm32_i2c_sem_waitdone(). It is then possible that the handling code was executed only once with "elapsed time" equal 0. When scheduler resumes this thread it is quite possible that now "elapsed time" will be well above timeout threshold. In that case the function returns and reports a timeout, even though the handling code was not executed "recently". Fix this by inverting the order of operations in the loop - save elapsed time before handling I2C. This way a context switch anywhere in the loop will not cause an erroneous "timeout" error.
2017-03-09 14:20:41 -04:00
diff --git NuttX/nuttx/arch/arm/src/stm32/stm32f30xxx_i2c.c NuttX/nuttx/arch/arm/src/stm32/stm32f30xxx_i2c.c
HOTFIX:Backport of second round upstream NuttX i2c fix (#6771) backport 3cd66af889b42b036d6c9d88e067fc3b8abbdb2a and pr 258 Applies to STM32F4 and STM32F7 STM32, STM32 F7, and STM32 L4: Clone Freddie Chopin's I2C change to similar STM32 I2C drivers. Save elapsed time before handling I2C in stm32_i2c_sem_waitstop() This patch follows the same logic as in previous fix to stm32_i2c_sem_waitdone(). It is possible that a context switch occurs after I2C registers are read but before elapsed time is saved in stm32_i2c_sem_waitstop(). It is then possible that the registers were read only once with "elapsed time" equal 0. When scheduler resumes this thread it is quite possible that now "elapsed time" will be well above timeout threshold. In that case the function returns and reports a timeout, even though the registers were not read "recently". Fix this by inverting the order of operations in the loop - save elapsed time before reading registers. This way a context switch anywhere in the loop will not cause an erroneous "timeout" error.
2017-03-10 13:26:51 -04:00
index 312e0b4..b2da861 100644
HOTFIX:Backport of upstream NuttX i2c fix 5a6d95dd9f051be548a8d2378aaef75f0a1ba5e1 and ee5ae3a57dbbe835584f164c956e0374da1ed2eb Applies to STM32F4 and STM32F7 Save elapsed time before handling I2C in stm32_i2c_sem_waitdone() It is possible that a context switch occurs after stm32_i2c_isr() call but before elapsed time is saved in stm32_i2c_sem_waitdone(). It is then possible that the handling code was executed only once with "elapsed time" equal 0. When scheduler resumes this thread it is quite possible that now "elapsed time" will be well above timeout threshold. In that case the function returns and reports a timeout, even though the handling code was not executed "recently". Fix this by inverting the order of operations in the loop - save elapsed time before handling I2C. This way a context switch anywhere in the loop will not cause an erroneous "timeout" error.
2017-03-09 14:20:41 -04:00
--- NuttX/nuttx/arch/arm/src/stm32/stm32f30xxx_i2c.c
+++ NuttX/nuttx/arch/arm/src/stm32/stm32f30xxx_i2c.c
@@ -7,7 +7,7 @@
*
* With extensions and modifications for the F1, F2, and F4 by:
*
- * Copyright (C) 2011-2013, 2016 Gregory Nutt. All rights reserved.
+ * Copyright (C) 2011-2013, 2016-2017 Gregory Nutt. All rights reserved.
* Author: Gregroy Nutt <gnutt@nuttx.org>
*
* And this version for the STM32 F3 by
@@ -704,15 +704,15 @@ static inline int stm32_i2c_sem_waitdone(FAR struct stm32_i2c_priv_s *priv)
do
{
+ /* Calculate the elapsed time */
+
+ elapsed = clock_systimer() - start;
+
/* Poll by simply calling the timer interrupt handler until it
* reports that it is done.
*/
stm32_i2c_isr(priv);
-
- /* Calculate the elapsed time */
-
- elapsed = clock_systimer() - start;
}
/* Loop until the transfer is complete. */
HOTFIX:Backport of second round upstream NuttX i2c fix (#6771) backport 3cd66af889b42b036d6c9d88e067fc3b8abbdb2a and pr 258 Applies to STM32F4 and STM32F7 STM32, STM32 F7, and STM32 L4: Clone Freddie Chopin's I2C change to similar STM32 I2C drivers. Save elapsed time before handling I2C in stm32_i2c_sem_waitstop() This patch follows the same logic as in previous fix to stm32_i2c_sem_waitdone(). It is possible that a context switch occurs after I2C registers are read but before elapsed time is saved in stm32_i2c_sem_waitstop(). It is then possible that the registers were read only once with "elapsed time" equal 0. When scheduler resumes this thread it is quite possible that now "elapsed time" will be well above timeout threshold. In that case the function returns and reports a timeout, even though the registers were not read "recently". Fix this by inverting the order of operations in the loop - save elapsed time before reading registers. This way a context switch anywhere in the loop will not cause an erroneous "timeout" error.
2017-03-10 13:26:51 -04:00
@@ -844,6 +844,10 @@ static inline void stm32_i2c_sem_waitstop(FAR struct stm32_i2c_priv_s *priv)
start = clock_systimer();
do
{
+ /* Calculate the elapsed time */
+
+ elapsed = clock_systimer() - start;
+
/* Check for STOP condition */
cr = stm32_i2c_getreg32(priv, STM32_I2C_CR2_OFFSET);
@@ -860,9 +864,6 @@ static inline void stm32_i2c_sem_waitstop(FAR struct stm32_i2c_priv_s *priv)
return;
}
- /* Calculate the elapsed time */
-
- elapsed = clock_systimer() - start;
}
/* Loop until the stop is complete or a timeout occurs. */
HOTFIX:Backport of upstream NuttX i2c fix 5a6d95dd9f051be548a8d2378aaef75f0a1ba5e1 and ee5ae3a57dbbe835584f164c956e0374da1ed2eb Applies to STM32F4 and STM32F7 Save elapsed time before handling I2C in stm32_i2c_sem_waitdone() It is possible that a context switch occurs after stm32_i2c_isr() call but before elapsed time is saved in stm32_i2c_sem_waitdone(). It is then possible that the handling code was executed only once with "elapsed time" equal 0. When scheduler resumes this thread it is quite possible that now "elapsed time" will be well above timeout threshold. In that case the function returns and reports a timeout, even though the handling code was not executed "recently". Fix this by inverting the order of operations in the loop - save elapsed time before handling I2C. This way a context switch anywhere in the loop will not cause an erroneous "timeout" error.
2017-03-09 14:20:41 -04:00
diff --git NuttX/nuttx/arch/arm/src/stm32/stm32f40xxx_i2c.c NuttX/nuttx/arch/arm/src/stm32/stm32f40xxx_i2c.c
HOTFIX:Backport of second round upstream NuttX i2c fix (#6771) backport 3cd66af889b42b036d6c9d88e067fc3b8abbdb2a and pr 258 Applies to STM32F4 and STM32F7 STM32, STM32 F7, and STM32 L4: Clone Freddie Chopin's I2C change to similar STM32 I2C drivers. Save elapsed time before handling I2C in stm32_i2c_sem_waitstop() This patch follows the same logic as in previous fix to stm32_i2c_sem_waitdone(). It is possible that a context switch occurs after I2C registers are read but before elapsed time is saved in stm32_i2c_sem_waitstop(). It is then possible that the registers were read only once with "elapsed time" equal 0. When scheduler resumes this thread it is quite possible that now "elapsed time" will be well above timeout threshold. In that case the function returns and reports a timeout, even though the registers were not read "recently". Fix this by inverting the order of operations in the loop - save elapsed time before reading registers. This way a context switch anywhere in the loop will not cause an erroneous "timeout" error.
2017-03-10 13:26:51 -04:00
index 2bb715c..e91058a 100644
HOTFIX:Backport of upstream NuttX i2c fix 5a6d95dd9f051be548a8d2378aaef75f0a1ba5e1 and ee5ae3a57dbbe835584f164c956e0374da1ed2eb Applies to STM32F4 and STM32F7 Save elapsed time before handling I2C in stm32_i2c_sem_waitdone() It is possible that a context switch occurs after stm32_i2c_isr() call but before elapsed time is saved in stm32_i2c_sem_waitdone(). It is then possible that the handling code was executed only once with "elapsed time" equal 0. When scheduler resumes this thread it is quite possible that now "elapsed time" will be well above timeout threshold. In that case the function returns and reports a timeout, even though the handling code was not executed "recently". Fix this by inverting the order of operations in the loop - save elapsed time before handling I2C. This way a context switch anywhere in the loop will not cause an erroneous "timeout" error.
2017-03-09 14:20:41 -04:00
--- NuttX/nuttx/arch/arm/src/stm32/stm32f40xxx_i2c.c
+++ NuttX/nuttx/arch/arm/src/stm32/stm32f40xxx_i2c.c
@@ -7,7 +7,7 @@
*
* With extensions, modifications by:
*
- * Copyright (C) 2011-2014, 2016 Gregory Nutt. All rights reserved.
+ * Copyright (C) 2011-2014, 2016-2017 Gregory Nutt. All rights reserved.
* Author: Gregory Nutt <gnutt@nuttx.org>
*
* Redistribution and use in source and binary forms, with or without
@@ -672,15 +672,15 @@ static inline int stm32_i2c_sem_waitdone(FAR struct stm32_i2c_priv_s *priv)
do
{
+ /* Calculate the elapsed time */
+
+ elapsed = clock_systimer() - start;
+
/* Poll by simply calling the timer interrupt handler until it
* reports that it is done.
*/
stm32_i2c_isr(priv);
-
- /* Calculate the elapsed time */
-
- elapsed = clock_systimer() - start;
}
/* Loop until the transfer is complete. */
HOTFIX:Backport of second round upstream NuttX i2c fix (#6771) backport 3cd66af889b42b036d6c9d88e067fc3b8abbdb2a and pr 258 Applies to STM32F4 and STM32F7 STM32, STM32 F7, and STM32 L4: Clone Freddie Chopin's I2C change to similar STM32 I2C drivers. Save elapsed time before handling I2C in stm32_i2c_sem_waitstop() This patch follows the same logic as in previous fix to stm32_i2c_sem_waitdone(). It is possible that a context switch occurs after I2C registers are read but before elapsed time is saved in stm32_i2c_sem_waitstop(). It is then possible that the registers were read only once with "elapsed time" equal 0. When scheduler resumes this thread it is quite possible that now "elapsed time" will be well above timeout threshold. In that case the function returns and reports a timeout, even though the registers were not read "recently". Fix this by inverting the order of operations in the loop - save elapsed time before reading registers. This way a context switch anywhere in the loop will not cause an erroneous "timeout" error.
2017-03-10 13:26:51 -04:00
@@ -731,6 +731,10 @@ static inline void stm32_i2c_sem_waitstop(FAR struct stm32_i2c_priv_s *priv)
start = clock_systimer();
do
{
+ /* Calculate the elapsed time */
+
+ elapsed = clock_systimer() - start;
+
/* Check for STOP condition */
cr1 = stm32_i2c_getreg(priv, STM32_I2C_CR1_OFFSET);
@@ -747,9 +751,6 @@ static inline void stm32_i2c_sem_waitstop(FAR struct stm32_i2c_priv_s *priv)
return;
}
- /* Calculate the elapsed time */
-
- elapsed = clock_systimer() - start;
}
/* Loop until the stop is complete or a timeout occurs. */
HOTFIX:Backport of upstream NuttX i2c fix 5a6d95dd9f051be548a8d2378aaef75f0a1ba5e1 and ee5ae3a57dbbe835584f164c956e0374da1ed2eb Applies to STM32F4 and STM32F7 Save elapsed time before handling I2C in stm32_i2c_sem_waitdone() It is possible that a context switch occurs after stm32_i2c_isr() call but before elapsed time is saved in stm32_i2c_sem_waitdone(). It is then possible that the handling code was executed only once with "elapsed time" equal 0. When scheduler resumes this thread it is quite possible that now "elapsed time" will be well above timeout threshold. In that case the function returns and reports a timeout, even though the handling code was not executed "recently". Fix this by inverting the order of operations in the loop - save elapsed time before handling I2C. This way a context switch anywhere in the loop will not cause an erroneous "timeout" error.
2017-03-09 14:20:41 -04:00
diff --git NuttX/nuttx/arch/arm/src/stm32f7/stm32_i2c.c NuttX/nuttx/arch/arm/src/stm32f7/stm32_i2c.c
HOTFIX:Backport of second round upstream NuttX i2c fix (#6771) backport 3cd66af889b42b036d6c9d88e067fc3b8abbdb2a and pr 258 Applies to STM32F4 and STM32F7 STM32, STM32 F7, and STM32 L4: Clone Freddie Chopin's I2C change to similar STM32 I2C drivers. Save elapsed time before handling I2C in stm32_i2c_sem_waitstop() This patch follows the same logic as in previous fix to stm32_i2c_sem_waitdone(). It is possible that a context switch occurs after I2C registers are read but before elapsed time is saved in stm32_i2c_sem_waitstop(). It is then possible that the registers were read only once with "elapsed time" equal 0. When scheduler resumes this thread it is quite possible that now "elapsed time" will be well above timeout threshold. In that case the function returns and reports a timeout, even though the registers were not read "recently". Fix this by inverting the order of operations in the loop - save elapsed time before reading registers. This way a context switch anywhere in the loop will not cause an erroneous "timeout" error.
2017-03-10 13:26:51 -04:00
index d089db9..fd73457 100644
HOTFIX:Backport of upstream NuttX i2c fix 5a6d95dd9f051be548a8d2378aaef75f0a1ba5e1 and ee5ae3a57dbbe835584f164c956e0374da1ed2eb Applies to STM32F4 and STM32F7 Save elapsed time before handling I2C in stm32_i2c_sem_waitdone() It is possible that a context switch occurs after stm32_i2c_isr() call but before elapsed time is saved in stm32_i2c_sem_waitdone(). It is then possible that the handling code was executed only once with "elapsed time" equal 0. When scheduler resumes this thread it is quite possible that now "elapsed time" will be well above timeout threshold. In that case the function returns and reports a timeout, even though the handling code was not executed "recently". Fix this by inverting the order of operations in the loop - save elapsed time before handling I2C. This way a context switch anywhere in the loop will not cause an erroneous "timeout" error.
2017-03-09 14:20:41 -04:00
--- NuttX/nuttx/arch/arm/src/stm32f7/stm32_i2c.c
+++ NuttX/nuttx/arch/arm/src/stm32f7/stm32_i2c.c
HOTFIX:Backport of second round upstream NuttX i2c fix (#6771) backport 3cd66af889b42b036d6c9d88e067fc3b8abbdb2a and pr 258 Applies to STM32F4 and STM32F7 STM32, STM32 F7, and STM32 L4: Clone Freddie Chopin's I2C change to similar STM32 I2C drivers. Save elapsed time before handling I2C in stm32_i2c_sem_waitstop() This patch follows the same logic as in previous fix to stm32_i2c_sem_waitdone(). It is possible that a context switch occurs after I2C registers are read but before elapsed time is saved in stm32_i2c_sem_waitstop(). It is then possible that the registers were read only once with "elapsed time" equal 0. When scheduler resumes this thread it is quite possible that now "elapsed time" will be well above timeout threshold. In that case the function returns and reports a timeout, even though the registers were not read "recently". Fix this by inverting the order of operations in the loop - save elapsed time before reading registers. This way a context switch anywhere in the loop will not cause an erroneous "timeout" error.
2017-03-10 13:26:51 -04:00
@@ -7,7 +7,7 @@
*
* With extensions and modifications for the F1, F2, and F4 by:
*
- * Copyright (C) 2016 Gregory Nutt. All rights reserved.
+ * Copyright (C) 2016-2017 Gregory Nutt. All rights reserved.
* Authors: Gregroy Nutt <gnutt@nuttx.org>
* John Wharington
* David Sidrane <david_s5@nscdg.com>
HOTFIX:Backport of upstream NuttX i2c fix 5a6d95dd9f051be548a8d2378aaef75f0a1ba5e1 and ee5ae3a57dbbe835584f164c956e0374da1ed2eb Applies to STM32F4 and STM32F7 Save elapsed time before handling I2C in stm32_i2c_sem_waitdone() It is possible that a context switch occurs after stm32_i2c_isr() call but before elapsed time is saved in stm32_i2c_sem_waitdone(). It is then possible that the handling code was executed only once with "elapsed time" equal 0. When scheduler resumes this thread it is quite possible that now "elapsed time" will be well above timeout threshold. In that case the function returns and reports a timeout, even though the handling code was not executed "recently". Fix this by inverting the order of operations in the loop - save elapsed time before handling I2C. This way a context switch anywhere in the loop will not cause an erroneous "timeout" error.
2017-03-09 14:20:41 -04:00
@@ -897,15 +897,15 @@ static inline int stm32_i2c_sem_waitdone(FAR struct stm32_i2c_priv_s *priv)
do
{
+ /* Calculate the elapsed time */
+
+ elapsed = clock_systimer() - start;
+
/* Poll by simply calling the timer interrupt handler until it
* reports that it is done.
*/
stm32_i2c_isr(priv);
-
- /* Calculate the elapsed time */
-
- elapsed = clock_systimer() - start;
}
/* Loop until the transfer is complete. */
HOTFIX:Backport of second round upstream NuttX i2c fix (#6771) backport 3cd66af889b42b036d6c9d88e067fc3b8abbdb2a and pr 258 Applies to STM32F4 and STM32F7 STM32, STM32 F7, and STM32 L4: Clone Freddie Chopin's I2C change to similar STM32 I2C drivers. Save elapsed time before handling I2C in stm32_i2c_sem_waitstop() This patch follows the same logic as in previous fix to stm32_i2c_sem_waitdone(). It is possible that a context switch occurs after I2C registers are read but before elapsed time is saved in stm32_i2c_sem_waitstop(). It is then possible that the registers were read only once with "elapsed time" equal 0. When scheduler resumes this thread it is quite possible that now "elapsed time" will be well above timeout threshold. In that case the function returns and reports a timeout, even though the registers were not read "recently". Fix this by inverting the order of operations in the loop - save elapsed time before reading registers. This way a context switch anywhere in the loop will not cause an erroneous "timeout" error.
2017-03-10 13:26:51 -04:00
@@ -1034,6 +1034,10 @@ static inline void stm32_i2c_sem_waitstop(FAR struct stm32_i2c_priv_s *priv)
start = clock_systimer();
do
{
+ /* Calculate the elapsed time */
+
+ elapsed = clock_systimer() - start;
+
/* Check for STOP condition */
cr = stm32_i2c_getreg32(priv, STM32_I2C_CR2_OFFSET);
@@ -1050,9 +1054,6 @@ static inline void stm32_i2c_sem_waitstop(FAR struct stm32_i2c_priv_s *priv)
return;
}
- /* Calculate the elapsed time */
-
- elapsed = clock_systimer() - start;
}
/* Loop until the stop is complete or a timeout occurs. */
HOTFIX:Backport of upstream NuttX i2c fix 5a6d95dd9f051be548a8d2378aaef75f0a1ba5e1 and ee5ae3a57dbbe835584f164c956e0374da1ed2eb Applies to STM32F4 and STM32F7 Save elapsed time before handling I2C in stm32_i2c_sem_waitdone() It is possible that a context switch occurs after stm32_i2c_isr() call but before elapsed time is saved in stm32_i2c_sem_waitdone(). It is then possible that the handling code was executed only once with "elapsed time" equal 0. When scheduler resumes this thread it is quite possible that now "elapsed time" will be well above timeout threshold. In that case the function returns and reports a timeout, even though the handling code was not executed "recently". Fix this by inverting the order of operations in the loop - save elapsed time before handling I2C. This way a context switch anywhere in the loop will not cause an erroneous "timeout" error.
2017-03-09 14:20:41 -04:00
diff --git NuttX/nuttx/arch/arm/src/stm32l4/stm32l4_i2c.c NuttX/nuttx/arch/arm/src/stm32l4/stm32l4_i2c.c
HOTFIX:Backport of second round upstream NuttX i2c fix (#6771) backport 3cd66af889b42b036d6c9d88e067fc3b8abbdb2a and pr 258 Applies to STM32F4 and STM32F7 STM32, STM32 F7, and STM32 L4: Clone Freddie Chopin's I2C change to similar STM32 I2C drivers. Save elapsed time before handling I2C in stm32_i2c_sem_waitstop() This patch follows the same logic as in previous fix to stm32_i2c_sem_waitdone(). It is possible that a context switch occurs after I2C registers are read but before elapsed time is saved in stm32_i2c_sem_waitstop(). It is then possible that the registers were read only once with "elapsed time" equal 0. When scheduler resumes this thread it is quite possible that now "elapsed time" will be well above timeout threshold. In that case the function returns and reports a timeout, even though the registers were not read "recently". Fix this by inverting the order of operations in the loop - save elapsed time before reading registers. This way a context switch anywhere in the loop will not cause an erroneous "timeout" error.
2017-03-10 13:26:51 -04:00
index eed199f..be46524 100644
HOTFIX:Backport of upstream NuttX i2c fix 5a6d95dd9f051be548a8d2378aaef75f0a1ba5e1 and ee5ae3a57dbbe835584f164c956e0374da1ed2eb Applies to STM32F4 and STM32F7 Save elapsed time before handling I2C in stm32_i2c_sem_waitdone() It is possible that a context switch occurs after stm32_i2c_isr() call but before elapsed time is saved in stm32_i2c_sem_waitdone(). It is then possible that the handling code was executed only once with "elapsed time" equal 0. When scheduler resumes this thread it is quite possible that now "elapsed time" will be well above timeout threshold. In that case the function returns and reports a timeout, even though the handling code was not executed "recently". Fix this by inverting the order of operations in the loop - save elapsed time before handling I2C. This way a context switch anywhere in the loop will not cause an erroneous "timeout" error.
2017-03-09 14:20:41 -04:00
--- NuttX/nuttx/arch/arm/src/stm32l4/stm32l4_i2c.c
+++ NuttX/nuttx/arch/arm/src/stm32l4/stm32l4_i2c.c
@@ -4,7 +4,7 @@
*
* Copyright (C) 2011 Uros Platise. All rights reserved.
* Author: Uros Platise <uros.platise@isotel.eu>
- * Copyright (C) 2011-2013, 2016 Gregory Nutt. All rights reserved.
+ * Copyright (C) 2011-2013, 2016-2017 Gregory Nutt. All rights reserved.
* Author: Gregroy Nutt <gnutt@nuttx.org>
* Author: John Wharington
* Author: Sebastien Lorquet
@@ -648,15 +648,15 @@ static inline int stm32l4_i2c_sem_waitdone(FAR struct stm32l4_i2c_priv_s *priv)
do
{
+ /* Calculate the elapsed time */
+
+ elapsed = clock_systimer() - start;
+
/* Poll by simply calling the timer interrupt handler until it
* reports that it is done.
*/
stm32l4_i2c_isr(priv);
-
- /* Calculate the elapsed time */
-
- elapsed = clock_systimer() - start;
}
/* Loop until the transfer is complete. */
HOTFIX:Backport of second round upstream NuttX i2c fix (#6771) backport 3cd66af889b42b036d6c9d88e067fc3b8abbdb2a and pr 258 Applies to STM32F4 and STM32F7 STM32, STM32 F7, and STM32 L4: Clone Freddie Chopin's I2C change to similar STM32 I2C drivers. Save elapsed time before handling I2C in stm32_i2c_sem_waitstop() This patch follows the same logic as in previous fix to stm32_i2c_sem_waitdone(). It is possible that a context switch occurs after I2C registers are read but before elapsed time is saved in stm32_i2c_sem_waitstop(). It is then possible that the registers were read only once with "elapsed time" equal 0. When scheduler resumes this thread it is quite possible that now "elapsed time" will be well above timeout threshold. In that case the function returns and reports a timeout, even though the registers were not read "recently". Fix this by inverting the order of operations in the loop - save elapsed time before reading registers. This way a context switch anywhere in the loop will not cause an erroneous "timeout" error.
2017-03-10 13:26:51 -04:00
@@ -788,6 +788,10 @@ static inline void stm32l4_i2c_sem_waitstop(FAR struct stm32l4_i2c_priv_s *priv)
start = clock_systimer();
do
{
+ /* Calculate the elapsed time */
+
+ elapsed = clock_systimer() - start;
+
/* Check for STOP condition */
cr = stm32l4_i2c_getreg32(priv, STM32L4_I2C_CR2_OFFSET);
@@ -804,9 +808,6 @@ static inline void stm32l4_i2c_sem_waitstop(FAR struct stm32l4_i2c_priv_s *priv)
return;
}
- /* Calculate the elapsed time */
-
- elapsed = clock_systimer() - start;
}
/* Loop until the stop is complete or a timeout occurs. */