Under unknown conditions the receive path ended up in a frozen state.
In this state, the DMA and driver descriptor head were equal and all
receive descriptors had the used bit set. So, the DMA was unable to
store received frames. However, the receive daemon was never woken up
to refill the receive buffers. It seems that the RXUBR interrupt can be
used to recover from this state.
Update #4652.
Do not use the interface mutex in the receive loop. Avoid multiple
reads of DMA descriptor words. Use a compile-time constant for the
receive DMA descriptor count to simplify calculations.
Update #4652.
Use the transmit interface handler to avoid a transmit task/interrupt.
Use a compile-time constant for the transmit DMA descriptor count to
simplify calculations.
Update #4652.
The pppstart expected that a driver write would somehow magically
process all data passed to the write function. Because ppp disables all
buffering that originally has been in termios, that assumption is not
true for all but polled drivers.
With this patch, the pppstart now gets and processes the feedback that
is returned from the driver via rtems_termios_dequeue_characters.
Update #4493
Since the epoch call callbacks may lock/unlock a mutex the callbacks must be
called from within thread context with thread dispatching enabled. Use thread
pinning to ensure that the calling thread stays on the right processor. Use
the interrupt server to ensure the thread context for the epoch drain.
Changed nfs_utime() to nfs_utimens(), changed the arguments to use
a timespec array instead of individual variables for access and
modified time.
Updates #4400
Possible data sources for SD driver:
- mmc_sd_switch():
- length: 64 byte;
- buffer on stack
- mmc_test_bus_width():
- length: 4 or 8 byte
- buffer in program memory or on stack
- mmc_app_send_scr():
- length: 8 byte
- buffer from device ivar structure
- mmc_app_sd_status():
- length: 64 byte
- buffer from device ivar structure
- mmc_send_ext_csd():
- length: MMC_EXTCSD_SIZE = 512
- buffer from device ivar structure
- rtems_bsd_mmcsd_disk_read_write():
- length: depends on read
- buffer from rtems_blkdev buffer -> already aligned
- mmcsd_ioctl_cmd():
- length: depends on call
- buffer malloced, not aligned -> patched in RTEMS
So the problematic buffers are only the ones up to 512 bytes. Copy these data
into a buffer to avoid that problem.
The TI driver has been moved to RTEMS so the driver can be removed
from libBSD.
The following files have been removed from libBSD and moved to
RTEMS.
1) ti/am335x/am335x_scm_padconf.c
2) ti/am335x/am335x_scm_padconf.h
3) ti/ti_pinmux.c
4) ti/ti_pinmux.h
Update #3784
- The files in the i386 directory have been moved to common x86 directories by FreeBSD:
- freebsd/sys/i386/include/machine/bus.h
- freebsd/sys/x86/include/machine/legacyvar.h
- freebsd/sys/x86/include/machine/specialreg.h
- Add header files in rtemsbsd directory to direct compiler to new files
At least the powerpc C++ compiler defines currently _GNU_SOURCE due to a GCC
configuration error. This leads to conflicting qsort_r() definitions.
As a workaround undefine _GNU_SOURCE for this C++ compatibility test file.
Update #4078.
