!= was used in place of <, while comparing expected and actual payload
size. The module may be running a higher version of the protocol and
might have some extra fields (towards the end) in the structure, and the
AP needs to ignore them.
This also updates the print (expected-payload-size <
actual-payload-size), when the size doesn't match for requests received
by the module. This gives more details required for debugging.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
Print (expected-payload-size actual-payload-size), when the size doesn't
match for requests received by the module. This gives more details
required for debugging the issue.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Alex Elder <elder@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
Print (expected-payload-size actual-payload-size), when the size doesn't
match for requests received by the module. This gives more details
required for debugging the issue.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Alex Elder <elder@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
These were not defined, and I just posted patches that use them.
Signed-off-by: Alex Elder <elder@linaro.org>
Tested-by: Mark Greer <mgreer@animalcreek.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
The error paths in gb_loopback_connection_init() are kind of screwed
up--not in proper order, and the label naming convention seems a
little inconsistent.
Fix this, ensuring each error cleans up the setup that's been done
up to that point. Use the convention that the label indicates the
first thing that needs to be cleaned up.
Reorder the statements in gb_loopback_connection_exit() to match
the order of cleanup in gb_loopback_connection_init().
Signed-off-by: Alex Elder <elder@linaro.org>
Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
Stop recording and updating the average every time a sample
is recorded. Instead, compute it from the sum and count only
when it's required.
Signed-off-by: Alex Elder <elder@linaro.org>
Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
Define a separate macro for displaying the average of the samples
collected. This will be used so we can calculate the average only
when requested, rather than every time a new value gets recorded.
Signed-off-by: Alex Elder <elder@linaro.org>
Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
The only values passed to the gb_loopback_ro_attr() macro are
unsigned 32-bit values. So there's no need to pass a "type"
format specifier.
Signed-off-by: Alex Elder <elder@linaro.org>
Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
The error count is unsigned, so fix the format specifier used in its
attribute definition.
Signed-off-by: Alex Elder <elder@linaro.org>
Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
The count of statistical samples recorded is currently a 64-bit
value. 32 bits is sufficient, and in fact anything more than
that won't work for the do_div() call it's pass to anyway. So make
the count field be 32 bits.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
The minimum and maximum values for stats values are always 32 bits.
Change the type for these fields to reflect this.
Signed-off-by: Alex Elder <elder@linaro.org>
Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
The only values supplied to gb_loopback_update_stats() are 32-bits,
so change the type of the second argument to reflect that.
Signed-off-by: Alex Elder <elder@linaro.org>
Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
Use the largest representable value when initializing the "min"
field when resetting loopback statistics.
Signed-off-by: Alex Elder <elder@linaro.org>
Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
There is no need to cast a void pointer to a particular type.
Drop the casts used in this way, mainly in the attribute definition
macros.
Signed-off-by: Alex Elder <elder@linaro.org>
Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
Fix two misspellings. And add spaces around a '%' operator.
Signed-off-by: Alex Elder <elder@linaro.org>
Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
Add a public declaration for gb_interface_destroy(), matching
gb_interface_create().
It's not yet used outside "interface.c" but I suppose it
could be, and its scope is currently public.
Signed-off-by: Alex Elder <elder@linaro.org>
Reviewed-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
For consistency reasons, use only type attributes for message packing.
Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Reviewed-by: Alex Elder <elder@linaro.org>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
Fix misspelled variable name in comment.
Reviewed-by: Alex Elder <elder@linaro.org>
Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
This should be exposed to external users (like gbsim). Move it to
greybus_protocols.h.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Johan Hovold <johan@hovoldconsulting.com>
Reviewed-by: Alex Elder <elder@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
The buffers allocated for message header is already 64 bit aligned and
we have explicit pad bytes in the header structure, to 64 bit align the
operation specific data.
And so there is no need to add the aligned attribute to the operation
message header. Drop it.
Suggested-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Alex Elder <elder@linaro.org>
Reviewed-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
The USB bridged-PHY protocol driver currently depends on changes to USB
core that are not yet upstream.
Disable for now.
Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
Fix allocation, deregistration and deallocation of USB HCD, and update
the hcd_priv helper functions.
The HCD private data was not allocated correctly, something which would
lead to a crash when accessed in hcd_start. The HCD was neither
reregistered or deallocated on connection tear down.
Signed-off-by: Johan Hovold <johan@hovoldconsulting.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
This is required to get things working for now, after the latest revert
of svc protocol is done.
Currently svc's cport id is set to 2 and that hd cport id will be used
for the third connection we make. And that protocol (which is i2c in one
of the cases), may not work as the (dis)connected event isn't sent for
it.
Fix this by getting rid of svc protocol check from (dis)connected
events for now. This must be reverted later, once svc protocol is
included again.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Tested-by: Mark Greer <mgreer@animalcreek.com>
Reviewed-by: Alex Elder <elder@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
The firmware for the svc changes isn't quite ready, so revert the whole
set of patches in one hunk to get things back to a working state for the
other firmware developers. The svc patches will be added back in a
separate branch.
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
The depth of the kfifo used to log the latency data for user-space can be
moved upwards or downward by way of a module parameter. The user may still
specify a test set that's larger than the number of kfifo elements we have
available. If the user specifies more iterations than can be logged give a
warning as feedback and continue with the test.
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
The current loopback code provides the minimum, maximum and average latency
values for a given test set. It would be highly useful for user-space to
have access to each one of the latency metrics in order to graph outliers.
This patch adds a simple character device interface implmenting a read()
interface that allows user-space to read out the saved latency metrics
which have been stored in a kfifo for this purpose.
A module parameter is provided to allow varying the depth of the kfifo in
order to allow a user to capture potentially large data sets. This version
sets the default depth for the kfifo at 8192 dwords.
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
Current code allows a sysfs callback and a kernel worker thread to write
all over and act upon data that could be in the process of being updated by
the other. This patch adds a reasonably coarse mutex to enscure sync
between the two.
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
It is of more interest to graphing system performance to base our
timestamps on the time it takes a greybus_operation_sync() to complete.
Higher level timestamping code is less accurate and not relevant to
throughput and latency characterization.
Signed-off-by: Bryan O'Donoghue <bryan.odonoghue@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
Its really part of initializing the host device and is required for
every 'hd' that is created. Lets move the call to do basic
initialization of svc connection to greybus_create_hd().
Also add a comment to specify why we need to do it that early.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
We forgot to free any ida internal structures that were used by this
host controller structure when we free the memory for the controller.
So fix that up by doing so in the release function.
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
The svc connection needs to be ready before creating the URBs, otherwise
the svc version request might come in before the AP was ready to parse
them.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
The function is only called locally, so mark it static to make sparse
happy.
Tested-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
The callback is never used anymore, so remove it from struct
greybus_host_driver as well as from the es1 and es2 drivers.
Tested-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
The function is gone, remove it from the header file as well.
Tested-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
The AP needs to be able to specify L4 CPort flags and traffic class
parameters on a connection-by-connection basis. Extend the connection
create operation to accept these. Since there's no policy to decide
these, fix them at TC0 with end-to-end-flow control, controlled segment
dropping, and CPort safety valve enabled.
Tested-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Perry Hung <perry@leaflabs.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
Ask the SVC to do all the necessary bits for creating a new connection.
This is skipped for the initial SVC connection.
Tested-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Perry Hung <perry@leaflabs.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
Upon receiving a hotplug request, we need to prepare the routing table
to allow packets to flow between the AP interface and the newly detected
interface.
Tested-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Perry Hung <perry@leaflabs.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
Its not used anymore as we have more sophisticated svc protocol in
place, lets get rid of earlier code.
Signed-off-by: Viresh Kumar <viresh.kumar@linaro.org>
Tested-by: Perry Hung <perry@leaflabs.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
We have switched over to use the "new" svc messages, no more need to
have a special USB endpoint to handle them, they come through the normal
CPort messages.
Based on a patch from Viresh.
Tested-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
We have switched over to use the "new" svc messages, no more need to
have a special USB endpoint to handle them, they come through the normal
CPort messages.
Based on a patch from Viresh.
Tested-by: Viresh Kumar <viresh.kumar@linaro.org>
Reviewed-by: Viresh Kumar <viresh.kumar@linaro.org>
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>
