In addition to the problems that were previously identified there were a few other bits and pieces outstanding;
* Timing was dependent on CPU speed rather than absolute time
* End of transfer handling was a bit mixed up
* It's possible for data to still be in the FIFO (i.e. not have reached
the card) when a next write is requested, so we need to wait for that to
complete
* Interrupt Status could be carried over from one transfer episode to the
next, corrupting progress
* Multi-descriptor DMA writing simply wasn't implemented, but there were no
indications ... it just failed silently
(1) Now the Ethernet is completely re-initialized when an error occurs by means of taking the interface down and back up but the PHY is _not_ renegotiated for that case because that is very time consuming and an error in the Ethernet is no reflection on the state of the PHY anyway.
(2) Explicitly sets the expected PHY address to zero (this could be moved into the config) rather than searching for it which takes ages, and it's zero anyway for this board (that's the broadcast address, and anything that cannot respond on that has multiple PHYs, so that would be a new board).
(3) Allows for the renegotiation of the PHY to be optional when a reset is needed. If a non-renegotiated reset doesn't result in good comms to the PHY then it'll automatically be escalated to a renegotiated one.
(4) Only performs a reset for errors that need it (the CRITICAL_ERROR define). The list of errors that need reset are somewhat arbitrarily chosen based on my prejudices and might need to be revisited, but certainly the jabber errors don't need reset, the partial packet is thrown away by the layer above anyway.
(5) Re-loads the multicast table on reset.
(6) Adds a bit more logging into the imxrt Ethernet module.
Pull request for mtd/progmem refactoring.
* mtd/progmem: change up_progmem_npages to up_progmem_neraseblocks
page is a unit for read/write operation.
eraseblock is a unit for erase operation.
up_progmem_npages is a little bit confusing because it returns number of
erase blocks in flash memory. This patch changes up_progmem_npages to
up_progmem_neraseblocks. There is no logical change.
Signed-off-by: EunBong Song <eunb.song@samsung.com>
* mtd/progmem: up_progmem_erasesize.
Change argument name to be more readable.
Signed-off-by: EunBong Song <eunb.song@samsung.com>
* mtd/progmem: up_progmem_eraseblock
Change up_progmem_erasepage to up_progmem_eraseblock.
eraseblock is more readable name than erasepage.
Signed-off-by: EunBong Song <eunb.song@samsung.com>
* mtd/progmem: change up_progmem_eraseblock's return value.
up_progmem_eraseblock erase a block. so it's better to return the erase block
size than page size.
Signed-off-by: EunBong Song <eunb.song@samsung.com>
* arm/cortex-r : fix wrong cp15_flash_dcache.
change mcrne to mcr for unconditional dcache.
Signed-off-by: sungduk.cho <sungduk.cho@samsung.com>
Approved-by: GregoryN <gnutt@nuttx.org>
drivers/usbdev: Fix buffer overrun check in rndis.c
The rndis driver has been working since 13 Nov 2017.
However, I finally found that it depends on network and buffer
configurations. If a receiving TCP packet is devided into smaller
ones based on USB max packet size, this condition check works
correctly.
Signed-off-by: Masayuki Ishikawa <Masayuki.Ishikawa@jp.sony.com>
Approved-by: GregoryN <gnutt@nuttx.org>
many locations: Change occurences of open() followed by file_detach() to file_open(). Change most non-controversion calls to open() to nx_open().
fs/inode/fs_fileopen.c: Flesh out file_open() with some interim, placeholder logic.
fs/inode/fs_fileopen.c: Add a framework for a file_open() implementation (no real logic in place yet).
fs/vfs/fs_open.c: Add nx_open() which is the same as open() except that it does not create a cancellation point nor does it modify the errno variable.
Squashed commit of the following:
Add procfs support to show stopped tasks. Add nxsig_action() to solve a chicken and egg problem: We needed to use sigaction to set default actions, but sigaction() would refuse to set actions if the default actions could not be caught or ignored.
sched/signal: Add configuration option to selectively enabled/disable default signal actions for SIGSTOP/SIGSTP/SIGCONT and SIGKILL/SIGINT. Fix some compilation issues.
sched/sched: Okay.. I figured out a way to handle state changes that may occur while they were stopped. If a task/thread was already blocked when SIGSTOP/SIGSTP was received, it will restart in the running state. I will appear that to the task/thread that the blocked condition was interrupt by a signal and returns the EINTR error.
sched/group and sched/sched: Finish framework for continue/resume logic.
sched/signal: Roughing out basic structure to support task suspend/resume
Pull request for cortex-r4 codes
* arm/armv7-r: Add general interrupt controller.
This is based on armv7-a gic controller code.
Signed-off-by: EunBong Song <eunb.song@samsung.com>
* arm/armv7-r: add invalidate dcache in arm_head.S
Adding invalidate dcache as a comment in arm_head.S.
Signed-off-by: EunBong Song <eunb.song@samsung.com>
* arm/armv7-r: Fix some wrong configuration of program status register.
PSR_E_BIT bit should be set for big endian system.
PSR_A_BIT bis is set automatically as arm cortex-r4 reference manual 3.7.4.
So we don't need to set this bit.
Signed-off-by: EunBong Song <eunb.song@samsung.com>
* arm/armv7-r: Fix some wrong MPU register definition.
Change MPU_RBAR_ADDR_MASK and MPU_RACR_TEX_SHIFT mask as
arm cortex-r4 reference manual.
Region Base Address Register 0-4 bits are reserved.
MPU Region Access control register type 3-5 bits.
Signed-off-by: EunBong Song <eunb.song@samsung.com>
* driver/mtd: fix compilation error.
This commit fixes below compilation errors.
CC: mtd/smart.c
mtd/smart.c:182:22: error: 'gWearBitToLevelMap4' defined but not used [-Werror=unused-const-variable=]
static const uint8_t gWearBitToLevelMap4[] =
^~~~~~~~~~~~~~~~~~~
mtd/smart.c:170:22: error: 'gWearLevelToBitMap4' defined but not used [-Werror=unused-const-variable=]
static const uint8_t gWearLevelToBitMap4[] =
^~~~~~~~~~~~~~~~~~~
cc1: all warnings being treated as errors
make[1]: *** [smart.o] Error 1
Signed-off-by: Junyeon LEE <junyeon2.lee@samsung.com>
Approved-by: GregoryN <gnutt@nuttx.org>
