This fixes a logic inversion that was preventing creation of a Bad Block
Table (BBT) from scratch on devices that lack one. This was discovered
during upstream integration testing. The BBT management layer in this
driver is not designed to be easily testable other than on real hardware.
This adds a test to very that remounting a completely used, but empty
NOR JFFS2 filesystem is possible. Previously, this triggered an edge
condition in the JFFS2 scan code that prevented remount of a correctly
formed and uncorrupted filesystem.
On JFFS2 file systems on NOR flash or dataflash that does not have spare
area for metadata and thus does not invoke delayed writes, it is
possible to put the file system into a state where all blocks have been
written to and all files have been deleted from the filesystem. There is
a bug in the JFFS2 file system scan routine that detects this situation
as a corrupted file system since the scan routine relies on "used" space
to discriminate a valid file system when there are blocks that need to
be erased. The correct fix would require a partial rewrite of the scan
routine, so instead this patch tracks the space marked as obsolete along
with space at the end of each block that is otherwise too small to
contain a JFFS2 node so that it can me compared with the dirty space.
Corrupted data (or otherwise non-JFFS2 blocks) will still cause this
check to fail as corrupted data isn't recognized as obsoleted (deleted)
nodes.
The RAM origin had an unnecessary alignment requirement and the RAM
lenght had a reference that couldn't possibly be used. This removes
those obsolete option specifiers.
Make sure there is no spurious empty line between the gcov info and the
*** END OF GCOV INFO BASE64 ***
line. This helps to recalculate the hash correctly at the consumer
side.
This adds a pthread that runs the test as well to increase test
coverage. The original test would have passed if all threads returned
the address of the Init task's errno since no additional threads or
tasks were checked.
Stacked mode doubles the number of sectors and device size while
parallel mode doubles the sector size and the device size. Make sure
that this is accounted for in the device size accessor.
When the switch to waf occurred, the SMP support in i386 was left out
causing it to accumulate a minor amount of cruft. This enables SMP for
the i386 BSPs that support it and updates them for the API drift that
has occurred since the change.
`tms570_sci_interrupt_handler()` is called when an RX interrupt fires. It checks
in the register `FLR`, the `RXRDY` bit (Receiver ready flag - indicate that the
SCIRD contains new data). If it is set, it calls
`tms570_sci_read_received_chars()`.
`tms570_sci_read_received_chars()` checks the register `RD` against 0. If it is
non zero, it returns 1 to indicate that one byte was read.
In the old behavior, if it is zero, the function returns 0 to indicate that no
data was read.
The new behavior is to not silently drop 0x00 bytes. Ignoring 0x00 bytes is fine
when working with printable text (which, I assume, is how this driver was
tested), but as soon as the UART is used in non canonical (raw) mode, with
potentially 0x00 bytes, these bytes will be silently dropped, causing issues in
the data/protocol layer above.
Update #4982.
This enables the VME support for the MVME2500. Note that the PCIe
support from libbsd is used. So you need the related libbsd patches for
this to work.
If the drivers in libbsd are not enabled, the linker should not pick up
anything from this patch.
Add a function that allows to set the polarity (active-low / negative
edge triggered or active-high / positive edge triggered) and sense
(level or edge sensitive) of the external interrupts.
This alters the API for rtems_cache_coherent_add_area to allow reporting
of failures that can occur during the process of adding a new area to
the coherent cache heap.
