Age | Commit message (Collapse) | Author |
|
Without this patch time code interrupts was never cleared
in the STATUS register.
|
|
By introducing a spearate RX and TX DMA channel semaphore RX
operations and TX operations does not block each other. Stopping
a DMA channel needs now to take both RX and TX semaphores to
block out both RX and TX tasks.
|
|
|
|
The user has already the power to control which DMA buffer
will generate interrupt, but no clean way to enable RX/TX
interrupts on DMA channel. Without this patch the user had
to init DMA config rx/tx_irq_en_cnt to a very large value.
|
|
* Add support for reading the GRSPW TX/RX descriptor counters
directly.
* Add semaphore lock to counters to avoid couters being out
of sync
* This makes it possible for the user in polling mode to check
the amount of work before entering RX/TX descriptor table
processing.
|
|
The user is now responsible to stop and close the DMA channels
before closing the device. To prevent complicated situations and
blocking the caller of grspw_close and grspw_dma_close a return
code was added to indicate to the user that the DMA may not have
been stopped or that blocked tasks are still active within the
driver for the specified device.
|
|
|
|
|
|
|
|
Added functionality for manual handling of link status events,
configurable via grspw_link_ctrl.
Added statistics counter for disconnect error.
|
|
Improved the link error handling options. Its now possible to
disable the link on individual link errors/warnings instead of
always on all or none.
Changed name of LINKOPTS_IRQ to LINKOPTS_EIRQ to match Linux
and VxWorks SpW driver.
|
|
|
|
|
|
|
|
|
|
NOTE that this will change the association of /dev/occanN
with OCCAN[N] device to the same order as in hardware. This means
that if you have been useing /dev/occan0 before you have to use
/dev/occan1 on a dual OCCAN device system (this affects GR712RC).
|
|
The EN bit (enable separate node address for this channel) is preserved when
grspw_dma_start is called. This preserves any previous address configuration.
|
|
There was a mixup between the transmit descriptor Link error (LE) bit and
Interrupt enable (IE) bit. TXPKT_FLAG_LINKERR now correctly indicates the link
error condition.
|
|
|
|
|
|
The fix does not affect the driver since the bit definitions
were never used by the GR5133B drivers. However it could affect
and application using the definitions.
|
|
|
|
|
|
|
|
Added External datation registers to the GRSPWTDP register description.
|
|
|
|
|
|
|
|
|
|
added clock cycles after GRSPW reset to make sure CTRL.START bit
write actually have an effect. Wait until reset is completed.
|
|
|
|
Update: Daniel Hellstrom updated SpW-IRQ implementation accoring to
changes in hardware register layout and features.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
This driver supports mainly device discovery and
interrupt handling.
|
|
From this driver's point of view the major new thing is that the
GRPCI peripheral PCI bridge has been updated to GRPCI2, the second
version. This means that both Big and Little Endian systems are now
supported and autodetected on runtime.
|
|
From this driver's point of view the major new thing is that the
GRPCI peripheral PCI bridge has been updated to GRPCI2, the second
version. This means that both Big and Little Endian systems are now
supported and autodetected on runtime.
The PCI frequency is used as AMBA frequency of the GR-RASTA-TMTC.
|
|
|
|
|
|
GRTIMER is compatible with GPTIMER, however the GPTIMER driver
does not support the extended features of GRTIMER.
|
|
|
|
|
|
Hardware generated checksum UDP packets does not work for
IP fragmented UDP packets. It seems as if the BSD stack never
signals to the GRETH driver to generate or not to generate
TCP/UDP checksum for different cases. The GRETH driver never
tells the BSD stack about it capabilities either so there is
no point generating the cksums in HW when its done in SW any
way.
This patch disables hardware generated UDP and TCP checksums.
|
|
Hardware generated checksum UDP packets does not work for
IP fragmented UDP packets. It seems as if the BSD stack never
signals to the GRETH driver to generate or not to generate
TCP/UDP checksum for different cases. The GRETH driver never
tells the BSD stack about it capabilities either so there is
no point generating the cksums in HW when its done in SW any
way.
This patch disables hardware generated UDP and TCP checksums.
|
|
|