Mirrors_Framework/nuttx
1
0
Fork 0
mirror of https://github.com/apache/nuttx.git synced 2026-06-06 00:14:22 +08:00
Code Issues Actions 1 Packages Projects Releases Wiki Activity

Merged nuttx/nuttx into master

Browse Source
This commit is contained in:
Aleksandr Vyhovanec
2017-01-14 12:17:04 +03:00
parent 03fdaad3f4 fba247b119
commit 825979fdfe
31 changed files with 1082 additions and 1093 deletions
Download Patch File Download Diff File Expand all files Collapse all files
TODO
+14 -37
View File
@@ -1,4 +1,4 @@
NuttX TODO List (Last updated January 7, 2017)
NuttX TODO List (Last updated January 10, 2017)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
This file summarizes known NuttX bugs, limitations, inconsistencies with
@@ -31,8 +31,8 @@ nuttx/:
apps/ and other Add-Ons:
(3) Network Utilities (apps/netutils/)
(2) NuttShell (NSH) (apps/nshlib)
(2) Network Utilities (apps/netutils/)
(1) NuttShell (NSH) (apps/nshlib)
(1) System libraries apps/system (apps/system)
(1) Pascal add-on (pcode/)
(4) Other Applications & Tests (apps/examples/)
@@ -414,7 +414,7 @@ o Memory Management (mm/)
So in this case, NuttX work just link Linux or or *nix systems:
All memory allocated by processes or threads in processes will
be recovered when the process exists.
be recovered when the process exits.
But not for the flat memory build. In that case, the issues
above do apply. There is no safe way to recover the memory in
@@ -427,7 +427,9 @@ o Memory Management (mm/)
it is inherently unsafe, I would never incorporate anything
like that into NuttX.
Status: Open. No changes are planned.
Status: Open. No changes are planned. NOTE: This applies to the FLAT
and PROTECTED builds only. There is no such leaking of memory
in the KERNEL build mode.
Priority: Medium/Low, a good feature to prevent memory leaks but would
have negative impact on memory usage and code size.
@@ -1503,7 +1505,7 @@ o Libraries (libc/, libm/)
at the OS interface: I would like to see more primitive OS
system calls with more higher level logic in the C library.
One awkard thing is the incompatibility of KERNEL vs FLAT
One awkward thing is the incompatibility of KERNEL vs FLAT
builds: In the kernel build, it would be nice to move many
of the thread-specific data items out of the TCB and into
the process address environment where they belong. It is
@@ -1754,13 +1756,15 @@ o Graphics Subsystem (graphics/)
interactions are via a thin layer in libnx/. The OS
interface is only via messages sent and received using POSIX
message queues. So this is good code and respects all of the
POSIX interfacing rules.
POSIX interfacing rules. Hence, it works well in all build
modes (FLAT, PROTECTED, and KERNEL builds).
But without CONFIG_NX_MULTIUSER, the single user applications
violate all of the rules and call internal NX functions
violate all of the rules and calls internal NX functions
directly. This includes all calls to internal OSfunctions
with names like, nx_open, up_fbinitialize, board_lcd_*, and
others
others. This is a violation of interfacing standard in all
cases and can only be made to work in the FLAT build mode.
The single user mode does have some desirable properties: It
is lighter weight and so more suitable for very resource limited
@@ -1768,7 +1772,7 @@ o Graphics Subsystem (graphics/)
solution is to eliminate the single user mode and provide only
the multi-user mode with the message queue interface.
Status: Open
Priority: Low, not a serious issue but worth noting. Single user
Priority: Low-Medium, not a serious issue but worth noting. Single user
mode is a blemish on the OS and not compatible with the RTOS
roadmap. But neither is there any critical necessity to
remove the offending code immediately. Be aware: If you use
@@ -1804,21 +1808,6 @@ o Build system
Priority: Low, since I am not aware of anyone using the Windows Native build.
But, of course, very high if you want to use it.
Title: REMOVE SINGLE USER MODE
Description: The graphics sub-system can operate in either a single-user mode or
in a multi-user mode. In the multiple-user mode, a kernel thread
is used to support a graphics server. Multiple applications may then
communicate with the server using a message queue. This users only
standard POSIX interfaces and works in all build modes (FLAT,
PROTECTED, and KERNEL builds).
The single-user mode, on the hand, uses inappropriate calls directly
into the OS. This violates the POSIX interface and must, eventually,
be eliminated. These inappropriate calls can only be supported in
the FLAT build mode.
Status: Open
Priority: Medium-High
o Other drivers (drivers/)
^^^^^^^^^^^^^^^^^^^^^^^^
@@ -2011,13 +2000,6 @@ o Network Utilities (apps/netutils/)
Status: Open
Priority: Medium
Title: THTTPD WARNINGS
Description: If the network is enabled, but THTTPD is not configured, it spews out lots
of pointless warnings. This is kind of annoying and unprofessional; needs to
be fixed someday.
Status: Open. An annoyance, but not a real problem.
Priority: Low
Title: NETWORK MONITOR NOT GENERALLY AVAILABLE
Description: The NSH network management logic has general applicability
but is currently useful only because it is embedded in the NSH
@@ -2037,11 +2019,6 @@ o NuttShell (NSH) (apps/nshlib)
Status: Open
Priority: Low
Title: ARPPING COMMAND
Description: Add an arping command
Status: Open
Priority: Low (enhancement)
o System libraries apps/system (apps/system)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
arch/arm/include/kinetis/chip.h
+846 -805
View File
File diff suppressed because it is too large Load Diff
arch/arm/src/common/up_exit.c
+7 -9
View File
@@ -1,7 +1,7 @@
/****************************************************************************
* common/up_exit.c
*
* Copyright (C) 2007-2009, 201-2014 Gregory Nutt. All rights reserved.
* Copyright (C) 2007-2009, 201-2014, 2017 Gregory Nutt. All rights reserved.
* Author: Gregory Nutt <gnutt@nuttx.org>
*
* Redistribution and use in source and binary forms, with or without
@@ -41,10 +41,11 @@
#include <sched.h>
#include <debug.h>
#include <nuttx/arch.h>
#include <nuttx/arch.h>
#include <nuttx/irq.h>
#ifdef CONFIG_DUMP_ON_EXIT
#include <nuttx/fs/fs.h>
# include <nuttx/fs/fs.h>
#endif
#include "task/task.h"
@@ -141,14 +142,11 @@ void _exit(int status)
{
struct tcb_s *tcb;
/* Disable interrupts. They will be restored when the next task is
* started.
/* Make sure that we are in a critical section with local interrupts.
* The IRQ state will be restored when the next task is started.
*/
(void)up_irq_save();
#ifdef CONFIG_SMP
(void)spin_trylock(&g_cpu_irqlock);
#endif
(void)enter_critical_section();
sinfo("TCB=%p exiting\n", this_task());
arch/arm/src/kinetis/Kconfig
+6
View File
@@ -120,6 +120,12 @@ config ARCH_CHIP_MK60X256VMD100
select KINETIS_HAVE_I2C1
select KINETIS_HAVE_I2C2
config ARCH_CHIP_MK60FN1M0VLQ12
bool "MK60FN1M0VLQ12"
select ARCH_FAMILY_K60
select KINETIS_HAVE_I2C1
select KINETIS_HAVE_I2C2
config ARCH_CHIP_MK64FN1M0VLL12
bool "MK64FN1M0VLL12"
select ARCH_FAMILY_K64
arch/arm/src/kinetis/chip/kinetis_k60pinmux.h
+2 -2
View File
@@ -55,10 +55,10 @@
* configuration (with no suffix) that maps to the correct alternative.
*/
#if defined(CONFIG_ARCH_CHIP_MK60N256VLQ100) || defined(CONFIG_ARCH_CHIP_MK60X256VLQ100) || \
#if defined(CONFIG_ARCH_CHIP_MK60N256VLQ100) || defined(CONFIG_ARCH_CHIP_MK60X256VLQ100) || \
defined(CONFIG_ARCH_CHIP_MK60N512VLQ100) || defined(CONFIG_ARCH_CHIP_MK60N256VMD100) || \
defined(CONFIG_ARCH_CHIP_MK60X256VMD100) || defined(CONFIG_ARCH_CHIP_MK60N512VMD100) || \
defined(CONFIG_ARCH_CHIP_MK60N512VLL100)
defined(CONFIG_ARCH_CHIP_MK60N512VLL100) || defined(CONFIG_ARCH_CHIP_MK60FN1M0VLQ12)
#define PIN_TSI0_CH1 (PIN_ANALOG | PIN_PORTA | PIN0)
#define PIN_UART0_CTS_1 (PIN_ALT2 | PIN_PORTA | PIN0)
arch/arm/src/kinetis/chip/kinetis_mcg.h
+4
View File
@@ -151,7 +151,11 @@
#define MCG_C6_VDIV_SHIFT (0) /* Bits 0-4: VCO Divider */
#define MCG_C6_VDIV_MASK (31 << MCG_C6_VDIV_SHIFT)
#ifdef KINETIS_NEW_MCG
# define MCG_C6_VDIV(n) ((uint32_t)((n)-16) << MCG_C6_VDIV_SHIFT) /* n=16..47 */
#else
# define MCG_C6_VDIV(n) ((uint32_t)((n)-24) << MCG_C6_VDIV_SHIFT) /* n=24..55 */
#endif
#define MCG_C6_CME (1 << 5) /* Bit 5: Clock Monitor Enable */
#define MCG_C6_PLLS (1 << 6) /* Bit 6: PLL Select */
#define MCG_C6_LOLIE (1 << 7) /* Bit 7: Loss of Lock Interrrupt Enable */
arch/arm/src/kinetis/kinetis_clockconfig.c
+1 -1
View File
@@ -152,7 +152,7 @@ void kinetis_pllconfig(void)
*/
#ifdef BOARD_FRDIV
putreg8(BOARD_FRDIV | MCG_C1_CLKS_EXTREF, KINETIS_MCG_C1);
putreg8((BOARD_FRDIV << MCG_C1_FRDIV_SHIFT) | MCG_C1_CLKS_EXTREF, KINETIS_MCG_C1);
#else
putreg8(MCG_C1_FRDIV_DIV256 | MCG_C1_CLKS_EXTREF, KINETIS_MCG_C1);
#endif
arch/arm/src/kinetis/kinetis_config.h
+1 -1
View File
@@ -80,7 +80,7 @@
#undef HAVE_UART_DEVICE
#if defined(CONFIG_KINETIS_UART0) || defined(CONFIG_KINETIS_UART1) || \
defined(CONFIG_KINETIS_UART2) || defined(CONFIG_KINETIS_UART3) || \
defined(CONFIG_KINETIS_UART5)
defined(CONFIG_KINETIS_UART4) || defined(CONFIG_KINETIS_UART5)
# define HAVE_UART_DEVICE 1
#endif
arch/arm/src/kinetis/kinetis_serial.c
+1 -1
View File
@@ -875,7 +875,7 @@ static int up_interrupts(int irq, void *context)
else
#endif
#ifdef CONFIG_KINETIS_UART5
if (g_uart5priv.irq == irqs)
if (g_uart5priv.irqs == irq)
{
dev = &g_uart5port;
}
arch/arm/src/stm32/stm32_flash.c
+8 -4
View File
@@ -231,12 +231,14 @@ ssize_t up_progmem_erasepage(size_t page)
return -EFAULT;
}
/* Get flash ready and begin erasing single page */
#if !defined(CONFIG_STM32_STM32F40XX)
if (!(getreg32(STM32_RCC_CR) & RCC_CR_HSION))
{
return -EPERM;
}
#endif
/* Get flash ready and begin erasing single page */
stm32_flash_unlock();
@@ -318,12 +320,14 @@ ssize_t up_progmem_write(size_t addr, const void *buf, size_t count)
return -EFAULT;
}
/* Get flash ready and begin flashing */
#if !defined(CONFIG_STM32_STM32F40XX)
if (!(getreg32(STM32_RCC_CR) & RCC_CR_HSION))
{
return -EPERM;
}
#endif
/* Get flash ready and begin flashing */
stm32_flash_unlock();
arch/arm/src/stm32/stm32_serial.c
+23 -133
View File
@@ -1,7 +1,7 @@
/****************************************************************************
* arch/arm/src/stm32/stm32_serial.c
*
* Copyright (C) 2009-2014, 2016 Gregory Nutt. All rights reserved.
* Copyright (C) 2009-2014, 2016, 2017 Gregory Nutt. All rights reserved.
* Author: Gregory Nutt <gnutt@nuttx.org>
*
* Redistribution and use in source and binary forms, with or without
@@ -70,6 +70,7 @@
/****************************************************************************
* Pre-processor Definitions
****************************************************************************/
/* Some sanity checks *******************************************************/
/* DMA configuration */
@@ -208,7 +209,7 @@
# error "Unknown STM32 DMA"
# endif
/* DMA control words */
/* DMA control word */
# if defined(CONFIG_STM32_STM32F20XX) || defined(CONFIG_STM32_STM32F40XX)
# define SERIAL_DMA_CONTROL_WORD \
@@ -220,16 +221,6 @@
CONFIG_USART_DMAPRIO | \
DMA_SCR_PBURST_SINGLE | \
DMA_SCR_MBURST_SINGLE)
# ifdef CONFIG_SERIAL_IFLOWCONTROL
# define SERIAL_DMA_IFLOW_CONTROL_WORD \
(DMA_SCR_DIR_P2M | \
DMA_SCR_MINC | \
DMA_SCR_PSIZE_8BITS | \
DMA_SCR_MSIZE_8BITS | \
CONFIG_USART_DMAPRIO | \
DMA_SCR_PBURST_SINGLE | \
DMA_SCR_MBURST_SINGLE)
# endif
# else
# define SERIAL_DMA_CONTROL_WORD \
(DMA_CCR_CIRC | \
@@ -237,13 +228,6 @@
DMA_CCR_PSIZE_8BITS | \
DMA_CCR_MSIZE_8BITS | \
CONFIG_USART_DMAPRIO)
# ifdef CONFIG_SERIAL_IFLOWCONTROL
# define SERIAL_DMA_IFLOW_CONTROL_WORD \
(DMA_CCR_MINC | \
DMA_CCR_PSIZE_8BITS | \
DMA_CCR_MSIZE_8BITS | \
CONFIG_USART_DMAPRIO)
# endif
# endif
#endif
@@ -1608,28 +1592,13 @@ static int up_dma_setup(struct uart_dev_s *dev)
priv->rxdma = stm32_dmachannel(priv->rxdma_channel);
#ifdef CONFIG_SERIAL_IFLOWCONTROL
if (priv->iflow)
{
/* Configure for non-circular DMA reception into the RX FIFO */
/* Configure for circular DMA reception into the RX fifo */
stm32_dmasetup(priv->rxdma,
priv->usartbase + STM32_USART_RDR_OFFSET,
(uint32_t)priv->rxfifo,
RXDMA_BUFFER_SIZE,
SERIAL_DMA_IFLOW_CONTROL_WORD);
}
else
#endif
{
/* Configure for circular DMA reception into the RX FIFO */
stm32_dmasetup(priv->rxdma,
priv->usartbase + STM32_USART_RDR_OFFSET,
(uint32_t)priv->rxfifo,
RXDMA_BUFFER_SIZE,
SERIAL_DMA_CONTROL_WORD);
}
stm32_dmasetup(priv->rxdma,
priv->usartbase + STM32_USART_RDR_OFFSET,
(uint32_t)priv->rxfifo,
RXDMA_BUFFER_SIZE,
SERIAL_DMA_CONTROL_WORD);
/* Reset our DMA shadow pointer to match the address just
* programmed above.
@@ -1643,26 +1612,12 @@ static int up_dma_setup(struct uart_dev_s *dev)
regval |= USART_CR3_DMAR;
up_serialout(priv, STM32_USART_CR3_OFFSET, regval);
#ifdef CONFIG_SERIAL_IFLOWCONTROL
if (priv->iflow)
{
/* Start the DMA channel, and arrange for callbacks at the full point
* in the FIFO. After buffer gets full, hardware flow-control kicks
* in and DMA transfer is stopped.
*/
/* Start the DMA channel, and arrange for callbacks at the half and
* full points in the FIFO. This ensures that we have half a FIFO
* worth of time to claim bytes before they are overwritten.
*/
stm32_dmastart(priv->rxdma, up_dma_rxcallback, (void *)priv, false);
}
else
#endif
{
/* Start the DMA channel, and arrange for callbacks at the half and
* full points in the FIFO. This ensures that we have half a FIFO
* worth of time to claim bytes before they are overwritten.
*/
stm32_dmastart(priv->rxdma, up_dma_rxcallback, (void *)priv, true);
}
stm32_dmastart(priv->rxdma, up_dma_rxcallback, (void *)priv, true);
return OK;
}
@@ -2316,9 +2271,9 @@ static bool up_rxflowcontrol(struct uart_dev_s *dev,
unsigned int nbuffered, bool upper)
{
struct up_dev_s *priv = (struct up_dev_s *)dev->priv;
uint16_t ie;
#if defined(CONFIG_SERIAL_IFLOWCONTROL_WATERMARKS) && \
defined(CONFIG_STM32_FLOWCONTROL_BROKEN)
#if defined(CONFIG_SERIAL_IFLOWCONTROL_WATERMARKS) && defined(CONFIG_STM32_FLOWCONTROL_BROKEN)
if (priv->iflow && (priv->rts_gpio != 0))
{
/* Assert/de-assert nRTS set it high resume/stop sending */
@@ -2346,7 +2301,9 @@ static bool up_rxflowcontrol(struct uart_dev_s *dev,
* enable Rx interrupts.
*/
uart_disablerxint(dev);
ie = priv->ie;
ie &= ~USART_CR1_RXNEIE;
up_restoreusartint(priv, ie);
return true;
}
@@ -2359,7 +2316,7 @@ static bool up_rxflowcontrol(struct uart_dev_s *dev,
* received.
*/
uart_enablerxint(dev);
up_rxint(dev, true);
}
}
#endif
@@ -2391,18 +2348,7 @@ static int up_dma_receive(struct uart_dev_s *dev, unsigned int *status)
priv->rxdmanext++;
if (priv->rxdmanext == RXDMA_BUFFER_SIZE)
{
#ifdef CONFIG_SERIAL_IFLOWCONTROL
if (priv->iflow)
{
/* RX DMA buffer full. RX paused, RTS line pulled up to prevent
* more input data from other end.
*/
}
else
#endif
{
priv->rxdmanext = 0;
}
priv->rxdmanext = 0;
}
}
@@ -2410,40 +2356,6 @@ static int up_dma_receive(struct uart_dev_s *dev, unsigned int *status)
}
#endif
/****************************************************************************
* Name: up_dma_reenable
*
* Description:
* Call to re-enable RX DMA.
*
****************************************************************************/
#if defined(SERIAL_HAVE_DMA) && defined(CONFIG_SERIAL_IFLOWCONTROL)
static void up_dma_reenable(struct up_dev_s *priv)
{
/* Configure for non-circular DMA reception into the RX fifo */
stm32_dmasetup(priv->rxdma,
priv->usartbase + STM32_USART_RDR_OFFSET,
(uint32_t)priv->rxfifo,
RXDMA_BUFFER_SIZE,
SERIAL_DMA_IFLOW_CONTROL_WORD);
/* Reset our DMA shadow pointer to match the address just
* programmed above.
*/
priv->rxdmanext = 0;
/* Start the DMA channel, and arrange for callbacks at the full point in
* the FIFO. After buffer gets full, hardware flow-control kicks in and
* DMA transfer is stopped.
*/
stm32_dmastart(priv->rxdma, up_dma_rxcallback, (void *)priv, false);
}
#endif
/****************************************************************************
* Name: up_dma_rxint
*
@@ -2466,15 +2378,6 @@ static void up_dma_rxint(struct uart_dev_s *dev, bool enable)
*/
priv->rxenable = enable;
#ifdef CONFIG_SERIAL_IFLOWCONTROL
if (priv->iflow && priv->rxenable && (priv->rxdmanext == RXDMA_BUFFER_SIZE))
{
/* Re-enable RX DMA. */
up_dma_reenable(priv);
}
#endif
}
#endif
@@ -2510,14 +2413,10 @@ static bool up_dma_rxavailable(struct uart_dev_s *dev)
static void up_send(struct uart_dev_s *dev, int ch)
{
struct up_dev_s *priv = (struct up_dev_s *)dev->priv;
#ifdef HAVE_RS485
if (priv->rs485_dir_gpio != 0)
{
stm32_gpiowrite(priv->rs485_dir_gpio, priv->rs485_dir_polarity);
}
stm32_gpiowrite(priv->rs485_dir_gpio, priv->rs485_dir_polarity);
#endif
up_serialout(priv, STM32_USART_TDR_OFFSET, (uint32_t)ch);
}
@@ -2683,16 +2582,6 @@ static void up_dma_rxcallback(DMA_HANDLE handle, uint8_t status, void *arg)
if (priv->rxenable && up_dma_rxavailable(&priv->dev))
{
uart_recvchars(&priv->dev);
#ifdef CONFIG_SERIAL_IFLOWCONTROL
if (priv->iflow && priv->rxenable &&
(priv->rxdmanext == RXDMA_BUFFER_SIZE))
{
/* Re-enable RX DMA. */
up_dma_reenable(priv);
}
#endif
}
}
#endif
@@ -2791,6 +2680,7 @@ static void up_pm_notify(struct pm_callback_s *cb, int domain,
* return non-zero values when reverting back to higher power
* consumption modes!
*
*
****************************************************************************/
#ifdef CONFIG_PM
arch/arm/src/stm32/stm32f40xxx_rcc.c
+2 -7
View File
@@ -95,10 +95,10 @@ static inline void rcc_reset(void)
putreg32(0x00000000, STM32_RCC_CFGR);
/* Reset HSION, HSEON, CSSON and PLLON bits */
/* Reset HSEON, CSSON and PLLON bits */
regval = getreg32(STM32_RCC_CR);
regval &= ~(RCC_CR_HSION | RCC_CR_HSEON | RCC_CR_CSSON | RCC_CR_PLLON);
regval &= ~(RCC_CR_HSEON | RCC_CR_CSSON | RCC_CR_PLLON);
putreg32(regval, STM32_RCC_CR);
/* Reset PLLCFGR register to reset default */
@@ -619,11 +619,6 @@ static void stm32_stdclockconfig(void)
volatile int32_t timeout;
#ifdef STM32_BOARD_USEHSI
/* Enable Internal High-Speed Clock (HSI) */
regval = getreg32(STM32_RCC_CR);
regval |= RCC_CR_HSION; /* Enable HSI */
putreg32(regval, STM32_RCC_CR);
/* Wait until the HSI is ready (or until a timeout elapsed) */
arch/avr/src/common/up_exit.c
+7 -6
View File
@@ -1,7 +1,7 @@
/****************************************************************************
* arch/avr/src/common/up_exit.c
*
* Copyright (C) 2010, 2013-2014 Gregory Nutt. All rights reserved.
* Copyright (C) 2010, 2013-2014, 2017 Gregory Nutt. All rights reserved.
* Author: Gregory Nutt <gnutt@nuttx.org>
*
* Redistribution and use in source and binary forms, with or without
@@ -41,10 +41,11 @@
#include <sched.h>
#include <debug.h>
#include <nuttx/arch.h>
#include <nuttx/arch.h>
#include <nuttx/irq.h>
#ifdef CONFIG_DUMP_ON_EXIT
#include <nuttx/fs/fs.h>
# include <nuttx/fs/fs.h>
#endif
#include "task/task.h"
@@ -140,11 +141,11 @@ void _exit(int status)
{
struct tcb_s *tcb;
/* Disable interrupts. They will be restored when the next
* task is started.
/* Make sure that we are in a critical section with local interrupts.
* The IRQ state will be restored when the next task is started.
*/
(void)up_irq_save();
(void)enter_critical_section();
sinfo("TCB=%p exiting\n", this_task());
arch/hc/src/common/up_exit.c
+7 -6
View File
@@ -1,7 +1,7 @@
/****************************************************************************
* arch/hc/src/common/up_exit.c
*
* Copyright (C) 2011, 2013-2014 Gregory Nutt. All rights reserved.
* Copyright (C) 2011, 2013-2014, 2017 Gregory Nutt. All rights reserved.
* Author: Gregory Nutt <gnutt@nuttx.org>
*
* Redistribution and use in source and binary forms, with or without
@@ -41,10 +41,11 @@
#include <sched.h>
#include <debug.h>
#include <nuttx/arch.h>
#include <nuttx/arch.h>
#include <nuttx/irq.h>
#ifdef CONFIG_DUMP_ON_EXIT
#include <nuttx/fs/fs.h>
# include <nuttx/fs/fs.h>
#endif
#include "task/task.h"
@@ -140,11 +141,11 @@ void _exit(int status)
{
struct tcb_s* tcb;
/* Disable interrupts. They will be restored when the next
* task is started.
/* Make sure that we are in a critical section with local interrupts.
* The IRQ state will be restored when the next task is started.
*/
(void)up_irq_save();
(void)enter_critical_section();
sinfo("TCB=%p exiting\n", this_task());
arch/mips/src/common/up_exit.c
+7 -6
View File
@@ -1,7 +1,7 @@
/****************************************************************************
* arch/mips/src/common/up_exit.c
*
* Copyright (C) 2011, 2013-2014 Gregory Nutt. All rights reserved.
* Copyright (C) 2011, 2013-2014, 2017 Gregory Nutt. All rights reserved.
* Author: Gregory Nutt <gnutt@nuttx.org>
*
* Redistribution and use in source and binary forms, with or without
@@ -43,10 +43,11 @@
#include <syscall.h>
#include <assert.h>
#include <debug.h>
#include <nuttx/arch.h>
#include <nuttx/arch.h>
#include <nuttx/irq.h>
#ifdef CONFIG_DUMP_ON_EXIT
#include <nuttx/fs/fs.h>
# include <nuttx/fs/fs.h>
#endif
#include "task/task.h"
@@ -142,11 +143,11 @@ void _exit(int status)
{
struct tcb_s *tcb;
/* Disable interrupts. They will be restored when the next
* task is started.
/* Make sure that we are in a critical section with local interrupts.
* The IRQ state will be restored when the next task is started.
*/
(void)up_irq_save();
(void)enter_critical_section();
sinfo("TCB=%p exiting\n", this_task());
arch/misoc/src/lm32/lm32_exit.c
+7 -6
View File
@@ -1,9 +1,9 @@
/****************************************************************************
* arch/misoc/src/lm32/lm32_exit.c
*
* Copyright (C) 2010, 2013-2014 Gregory Nutt. All rights reserved.
* Copyright (C) 2010, 2013-2014, 2017 Gregory Nutt. All rights reserved.
* Author: Gregory Nutt <gnutt@nuttx.org>
* Ramtin Amin <keytwo@gmail.com>
* Ramtin Amin <keytwo@gmail.com>
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
@@ -42,8 +42,9 @@
#include <sched.h>
#include <debug.h>
#include <nuttx/arch.h>
#include <nuttx/arch.h>
#include <nuttx/irq.h>
#ifdef CONFIG_DUMP_ON_EXIT
# include <nuttx/fs/fs.h>
#endif
@@ -139,11 +140,11 @@ void _exit(int status)
{
struct tcb_s *tcb;
/* Disable interrupts. They will be restored when the next
* task is started.
/* Make sure that we are in a critical section with local interrupts.
* The IRQ state will be restored when the next task is started.
*/
(void)up_irq_save();
(void)enter_critical_section();
sinfo("TCB=%p exiting\n", this_task());
arch/renesas/src/common/up_exit.c
+6 -6
View File
@@ -1,7 +1,7 @@
/****************************************************************************
* arch/renesas/src/=common/up_exit.c
*
* Copyright (C) 2008-2009, 2013-2014 Gregory Nutt. All rights reserved.
* Copyright (C) 2008-2009, 2013-2014, 2017 Gregory Nutt. All rights reserved.
* Author: Gregory Nutt <gnutt@nuttx.org>
*
* Redistribution and use in source and binary forms, with or without
@@ -43,9 +43,9 @@
#include <debug.h>
#include <nuttx/arch.h>
#include <nuttx/irq.h>
#ifdef CONFIG_DUMP_ON_EXIT
#include <nuttx/fs/fs.h>
# include <nuttx/fs/fs.h>
#endif
#include "task/task.h"
@@ -141,11 +141,11 @@ void _exit(int status)
{
struct tcb_s* tcb;
/* Disable interrupts. They will be restored when the next
* task is started.
/* Make sure that we are in a critical section with local interrupts.
* The IRQ state will be restored when the next task is started.
*/
(void)up_irq_save();
(void)enter_critical_section();
sinfo("TCB=%p exiting\n", this_task());
arch/risc-v/src/common/up_exit.c
+7 -6
View File
@@ -1,7 +1,7 @@
/****************************************************************************
* arch/risc-v/src/common/up_exit.c
*
* Copyright (C) 2011, 2013-2014 Gregory Nutt. All rights reserved.
* Copyright (C) 2011, 2013-2014, 2017 Gregory Nutt. All rights reserved.
* Author: Gregory Nutt <gnutt@nuttx.org>
*
* Redistribution and use in source and binary forms, with or without
@@ -43,10 +43,11 @@
#include <syscall.h>
#include <assert.h>
#include <debug.h>
#include <nuttx/arch.h>
#include <nuttx/arch.h>
#include <nuttx/irq.h>
#ifdef CONFIG_DUMP_ON_EXIT
#include <nuttx/fs/fs.h>
# include <nuttx/fs/fs.h>
#endif
#include "task/task.h"
@@ -142,11 +143,11 @@ void _exit(int status)
{
struct tcb_s *tcb;
/* Disable interrupts. They will be restored when the next
* task is started.
/* Make sure that we are in a critical section with local interrupts.
* The IRQ state will be restored when the next task is started.
*/
(void)up_irq_save();
(void)enter_critical_section();
sinfo("TCB=%p exiting\n", this_task());
arch/x86/src/common/up_exit.c
+7 -6
View File
@@ -1,7 +1,7 @@
/****************************************************************************
* common/up_exit.c
*
* Copyright (C) 2011, 2013-2014, 2016 Gregory Nutt. All rights reserved.
* Copyright (C) 2011, 2013-2014, 2016-2017 Gregory Nutt. All rights reserved.
* Author: Gregory Nutt <gnutt@nuttx.org>
*
* Redistribution and use in source and binary forms, with or without
@@ -41,10 +41,11 @@
#include <sched.h>
#include <debug.h>
#include <nuttx/arch.h>
#include <nuttx/arch.h>
#include <nuttx/irq.h>
#ifdef CONFIG_DUMP_ON_EXIT
#include <nuttx/fs/fs.h>
# include <nuttx/fs/fs.h>
#endif
#include "task/task.h"
@@ -140,11 +141,11 @@ void _exit(int status)
{
struct tcb_s* tcb;
/* Disable interrupts. They will be restored when the next
* task is started.
/* Make sure that we are in a critical section with local interrupts.
* The IRQ state will be restored when the next task is started.
*/
(void)up_irq_save();
(void)enter_critical_section();
sinfo("TCB=%p exiting\n", this_task());
arch/xtensa/src/common/xtensa_exit.c
+7 -6
View File
@@ -1,7 +1,7 @@
/****************************************************************************
* arch/xtensa/src/common/xtensa_exit.c
*
* Copyright (C) 2016 Gregory Nutt. All rights reserved.
* Copyright (C) 2016-2017 Gregory Nutt. All rights reserved.
* Author: Gregory Nutt <gnutt@nuttx.org>
*
* Redistribution and use in source and binary forms, with or without
@@ -43,10 +43,11 @@
#include <syscall.h>
#include <assert.h>
#include <debug.h>
#include <nuttx/arch.h>
#include <nuttx/arch.h>
#include <nuttx/irq.h>
#ifdef CONFIG_DUMP_ON_EXIT
#include <nuttx/fs/fs.h>
# include <nuttx/fs/fs.h>
#endif
#include "task/task.h"
@@ -142,11 +143,11 @@ void _exit(int status)
{
struct tcb_s *tcb;
/* Disable interrupts. They will be restored when the next task is
* started.
/* Make sure that we are in a critical section with local interrupts.
* The IRQ state will be restored when the next task is started.
*/
(void)up_irq_save();
(void)enter_critical_section();
sinfo("TCB=%p exiting\n", this_task());
arch/z16/src/common/up_exit.c
+6 -6
View File
@@ -1,7 +1,7 @@
/****************************************************************************
* arch/z16/src/common/up_exit.c
*
* Copyright (C) 2008-2009, 2013 Gregory Nutt. All rights reserved.
* Copyright (C) 2008-2009, 2013, 2017 Gregory Nutt. All rights reserved.
* Author: Gregory Nutt <gnutt@nuttx.org>
*
* Redistribution and use in source and binary forms, with or without
@@ -43,9 +43,9 @@
#include <debug.h>
#include <nuttx/arch.h>
#include <nuttx/irq.h>
#ifdef CONFIG_DUMP_ON_EXIT
#include <nuttx/fs/fs.h>
# include <nuttx/fs/fs.h>
#endif
#include "chip/chip.h"
@@ -141,11 +141,11 @@ void _exit(int status)
{
FAR struct tcb_s* tcb;
/* Disable interrupts. Interrupts will remain disabled until
* the new task is resumed below.
/* Make sure that we are in a critical section with local interrupts.
* The IRQ state will be restored when the next task is started.
*/
(void)up_irq_save();
(void)enter_critical_section();
sinfo("TCB=%p exiting\n", tcb);
arch/z80/src/common/up_exit.c
+6 -6
View File
@@ -1,7 +1,7 @@
/****************************************************************************
* arch/z80/src/common/up_exit.c
*
* Copyright (C) 2007-2009, 2013-2014 Gregory Nutt. All rights reserved.
* Copyright (C) 2007-2009, 2013-2014, 2017 Gregory Nutt. All rights reserved.
* Author: Gregory Nutt <gnutt@nuttx.org>
*
* Redistribution and use in source and binary forms, with or without
@@ -44,9 +44,9 @@
#include <debug.h>
#include <nuttx/arch.h>
#include <nuttx/irq.h>
#ifdef CONFIG_DUMP_ON_EXIT
#include <nuttx/fs/fs.h>
# include <nuttx/fs/fs.h>
#endif
#include "chip/chip.h"
@@ -143,11 +143,11 @@ void _exit(int status)
{
FAR struct tcb_s* tcb;
/* Disable interrupts. Interrupts will remain disabled until
* the new task is resumed below.
/* Make sure that we are in a critical section with local interrupts.
* The IRQ state will be restored when the next task is started.
*/
(void)up_irq_save();
(void)enter_critical_section();
sinfo("TCB=%p exiting\n", tcb);
drivers/mmcsd/mmcsd_sdio.c
+1
View File
@@ -1476,6 +1476,7 @@ static ssize_t mmcsd_readmultiple(FAR struct mmcsd_state_s *priv,
{
offset = startblock << priv->blockshift;
}
finfo("nbytes=%d byte offset=%d\n", nbytes, offset);
/* Select the block size for the card */
fs/mqueue/mq_open.c
+1 -1
View File
@@ -100,7 +100,7 @@ mqd_t mq_open(FAR const char *mq_name, int oflags, ...)
/* Make sure that a non-NULL name is supplied */
if (mq_name == NULL || *mq_name == '/0')
if (mq_name == NULL || *mq_name == '\0')
{
errcode = EINVAL;
goto errout;
sched/Kconfig
+1 -1
View File
@@ -732,7 +732,7 @@ config SCHED_INSTRUMENTATION_CSECTION
void sched_note_csection(FAR struct tcb_s *tcb, bool state);
config SCHED_INSTRUMENTATION_SPINLOCK
config SCHED_INSTRUMENTATION_SPINLOCKS
bool "Spinlock monitor hooks"
default n
---help---
sched/irq/irq.h
+5 -1
View File
@@ -1,7 +1,7 @@
/****************************************************************************
* sched/irq/irq.h
*
* Copyright (C) 2007, 2008, 2013-2014 Gregory Nutt. All rights reserved.
* Copyright (C) 2007, 2008, 2013-2014, 2017 Gregory Nutt. All rights reserved.
* Author: Gregory Nutt <gnutt@nuttx.org>
*
* Redistribution and use in source and binary forms, with or without
@@ -72,6 +72,10 @@ extern volatile spinlock_t g_cpu_irqlock SP_SECTION;
extern volatile spinlock_t g_cpu_irqsetlock SP_SECTION;
extern volatile cpu_set_t g_cpu_irqset SP_SECTION;
/* Handles nested calls to enter_critical section from interrupt handlers */
extern volatile uint8_t g_cpu_nestcount[CONFIG_SMP_NCPUS];
#endif
/****************************************************************************
sched/irq/irq_csection.c
+18 -13
View File
@@ -1,7 +1,7 @@
/****************************************************************************
* sched/irq/irq_csection.c
*
* Copyright (C) 2016 Gregory Nutt. All rights reserved.
* Copyright (C) 2016-2017 Gregory Nutt. All rights reserved.
* Author: Gregory Nutt <gnutt@nuttx.org>
*
* Redistribution and use in source and binary forms, with or without
@@ -66,16 +66,10 @@ volatile spinlock_t g_cpu_irqlock SP_SECTION = SP_UNLOCKED;
volatile spinlock_t g_cpu_irqsetlock SP_SECTION;
volatile cpu_set_t g_cpu_irqset SP_SECTION;
#endif
/****************************************************************************
* Private Data
****************************************************************************/
#ifdef CONFIG_SMP
/* Handles nested calls to enter_critical section from interrupt handlers */
static uint8_t g_cpu_nestcount[CONFIG_SMP_NCPUS];
volatile uint8_t g_cpu_nestcount[CONFIG_SMP_NCPUS];
#endif
/****************************************************************************
@@ -303,6 +297,13 @@ try_again:
/* In any event, the nesting count is now one */
g_cpu_nestcount[cpu] = 1;
/* Also set the CPU bit so that other CPUs will be aware that this
* CPU holds the critical section.
*/
spin_setbit(&g_cpu_irqset, cpu, &g_cpu_irqsetlock,
&g_cpu_irqlock);
}
}
else
@@ -455,18 +456,22 @@ void leave_critical_section(irqstate_t flags)
}
else
{
/* No, not nested. Release the spinlock. */
/* No, not nested. Restore the g_cpu_irqset for this CPU
* and release the spinlock (if necessary).
*/
DEBUGASSERT(spin_islocked(&g_cpu_irqlock) &&
g_cpu_nestcount[cpu] == 1);
spin_lock(&g_cpu_irqsetlock); /* Protects g_cpu_irqset */
if (g_cpu_irqset == 0)
FAR struct tcb_s *rtcb = this_task();
DEBUGASSERT(rtcb != NULL);
if (rtcb->irqcount <= 0)
{
spin_unlock(&g_cpu_irqlock);
spin_clrbit(&g_cpu_irqset, cpu, &g_cpu_irqsetlock,
&g_cpu_irqlock);
}
spin_unlock(&g_cpu_irqsetlock);
g_cpu_nestcount[cpu] = 0;
}
}
sched/sched/sched_addreadytorun.c
+30 -3
View File
@@ -1,7 +1,7 @@
/****************************************************************************
* sched/sched/sched_addreadytorun.c
*
* Copyright (C) 2007-2009, 2014, 2016 Gregory Nutt. All rights reserved.
* Copyright (C) 2007-2009, 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
@@ -322,13 +322,40 @@ bool sched_addreadytorun(FAR struct tcb_s *btcb)
if (btcb->irqcount > 0)
{
/* Yes... make sure that scheduling logic on other CPUs knows
* that we hold the IRQ lock.
*/
spin_setbit(&g_cpu_irqset, cpu, &g_cpu_irqsetlock,
&g_cpu_irqlock);
}
/* No.. This CPU will be relinquishing the lock. But this works
* differently if we are performing a context switch from an
* interrupt handler and the interrupt handler has established
* a critical section. We can detect this case when
* g_cpu_nestcount[me] > 0.
*/
else if (g_cpu_nestcount[me] <= 0)
{
/* Release our hold on the IRQ lock. */
spin_clrbit(&g_cpu_irqset, cpu, &g_cpu_irqsetlock,
&g_cpu_irqlock);
}
/* Sanity check. g_cpu_netcount should be greater than zero
* only while we are within the critical section and within
* an interrupt handler. If we are not in an interrupt handler
* then there is a problem; perhaps some logic previously
* called enter_critical_section() with no matching call to
* leave_critical_section(), leaving the non-zero count.
*/
else
{
spin_clrbit(&g_cpu_irqset, cpu, &g_cpu_irqsetlock,
&g_cpu_irqlock);
DEBUGASSERT(up_interrupt_context());
}
/* If the following task is not locked to this CPU, then it must
sched/sched/sched_note.c
+13 -7
View File
@@ -82,6 +82,12 @@ struct note_startalloc_s
# define SIZEOF_NOTE_START(n) (sizeof(struct note_start_s))
#endif
/****************************************************************************
* Private Function Prototypes
****************************************************************************/
static void note_add(FAR const uint8_t *note, uint8_t notelen);
/****************************************************************************
* Private Data
****************************************************************************/
@@ -558,21 +564,21 @@ void sched_note_csection(FAR struct tcb_s *tcb, bool enter)
#ifdef CONFIG_SCHED_INSTRUMENTATION_SPINLOCKS
void sched_note_spinlock(FAR struct tcb_s *tcb, FAR volatile void *spinlock)
{
note_spincommon(tcb, spinlock, NOTE_SPINLOCK_LOCK)
note_spincommon(tcb, spinlock, NOTE_SPINLOCK_LOCK);
}
void sched_note_spinlocked(FAR struct tcb_s *tcb, FAR volatile void *spinlock);
void sched_note_spinlocked(FAR struct tcb_s *tcb, FAR volatile void *spinlock)
{
note_spincommon(tcb, spinlock, NOTE_SPINLOCK_LOCKED)
note_spincommon(tcb, spinlock, NOTE_SPINLOCK_LOCKED);
}
void sched_note_spinunlock(FAR struct tcb_s *tcb, FAR volatile void *spinlock);
void sched_note_spinunlock(FAR struct tcb_s *tcb, FAR volatile void *spinlock)
{
note_spincommon(tcb, spinlock, NOTE_SPINLOCK_UNLOCK)
note_spincommon(tcb, spinlock, NOTE_SPINLOCK_UNLOCK);
}
void sched_note_spinabort(FAR struct tcb_s *tcb, FAR volatile void *spinlock);
void sched_note_spinabort(FAR struct tcb_s *tcb, FAR volatile void *spinlock)
{
note_spincommon(tcb, spinlock, NOTE_SPINLOCK_ABORT)
note_spincommon(tcb, spinlock, NOTE_SPINLOCK_ABORT);
}
#endif
sched/sched/sched_removereadytorun.c
+30 -7
View File
@@ -1,7 +1,7 @@
/****************************************************************************
* sched/sched_removereadytorun.c
*
* Copyright (C) 2007-2009, 2012, 2016 Gregory Nutt. All rights reserved.
* Copyright (C) 2007-2009, 2012, 2016-2017 Gregory Nutt. All rights reserved.
* Author: Gregory Nutt <gnutt@nuttx.org>
*
* Redistribution and use in source and binary forms, with or without
@@ -260,23 +260,46 @@ bool sched_removereadytorun(FAR struct tcb_s *rtcb)
&g_cpu_schedlock);
}
/* Interrupts may be disabled after the switch. If irqcount is greater
* than zero, then this task/this CPU holds the IRQ lock
/* Adjust global IRQ controls. If irqcount is greater than zero,
* then this task/this CPU holds the IRQ lock
*/
if (nxttcb->irqcount > 0)
{
/* Yes... make sure that scheduling logic knows about this */
/* Yes... make sure that scheduling logic on other CPUs knows
* that we hold the IRQ lock.
*/
spin_setbit(&g_cpu_irqset, cpu, &g_cpu_irqsetlock,
&g_cpu_irqlock);
}
else
/* No.. This CPU will be relinquishing the lock. But this works
* differently if we are performing a context switch from an
* interrupt handler and the interrupt handler has established
* a critical section. We can detect this case when
* g_cpu_nestcount[me] > 0.
*/
else if (g_cpu_nestcount[me] <= 0)
{
/* No.. we may need to release our hold on the irq state. */
/* Release our hold on the IRQ lock. */
spin_clrbit(&g_cpu_irqset, cpu, &g_cpu_irqsetlock,
&g_cpu_irqlock);
&g_cpu_irqlock);
}
/* Sanity check. g_cpu_netcount should be greater than zero
* only while we are within the critical section and within
* an interrupt handler. If we are not in an interrupt handler
* then there is a problem; perhaps some logic previously
* called enter_critical_section() with no matching call to
* leave_critical_section(), leaving the non-zero count.
*/
else
{
DEBUGASSERT(up_interrupt_context());
}
nxttcb->task_state = TSTATE_TASK_RUNNING;
sched/sched/sched_setpriority.c
+1
View File
@@ -46,6 +46,7 @@
#include <nuttx/irq.h>
#include <nuttx/arch.h>
#include "irq/irq.h"
#include "sched/sched.h"
/****************************************************************************