]> Gitweb @ Texas Instruments - Open Source Git Repositories - git.TI.com/gitweb - ipc/ipcdev.git/log
10 years agoQNX: Fix multi_process.sh for < 10 processes
vwan@ti.com [Fri, 21 Jun 2013 22:28:13 +0000 (15:28 -0700)]
QNX: Fix multi_process.sh for < 10 processes

The slave side executables can only handle a maximum of 10 threads,
so the script did not work corrently when it inadvertently tried to
look for an 11th thread.

This addresses SDOCM00102067.

Signed-off-by: VW <vwan@ti.com>
10 years agoFix coding conventions in a few headers
Chris Ring [Mon, 24 Jun 2013 22:14:03 +0000 (15:14 -0700)]
Fix coding conventions in a few headers

Correct some IPC header issues recently identified by static

10 years agoVayu: Remove ti.platform.vayu.* pkgs
Chris Ring [Sat, 22 Jun 2013 00:07:49 +0000 (17:07 -0700)]
Vayu: Remove ti.platform.vayu.* pkgs

In the IPC 3.00 stream the tests/examples switched to
ti.platforms.evmDRA7XX platform instances, but we left
these ti.platform.vayu.* platform packages in the IPC product
to not disrupt existing users of them.

Users have had adequate time to move to the XDC-provided
platform instances, so we are removing them in the IPC 3.10

10 years agoQNX: Fix default build to not 'install'
Chris Ring [Fri, 21 Jun 2013 21:56:52 +0000 (14:56 -0700)]
QNX: Fix default build to not 'install'

The default build goal shouldn't run the 'install' goal.

This addresses SDOCM00102137.

10 years agoQNX IPC: MmRpc - Generate .so Libs also
Angela Stegmaier [Thu, 4 Apr 2013 17:42:46 +0000 (12:42 -0500)]
QNX IPC: MmRpc - Generate .so Libs also

Generate .so libraries in addition to .a so that the
MmRpc can be dynamically linked.

Signed-off-by: Angela Stegmaier <angelabaker@ti.com>
10 years agoQNX IPC: Export MmRpc Libs and Headers
Chris Ring [Thu, 20 Jun 2013 23:22:44 +0000 (16:22 -0700)]
QNX IPC: Export MmRpc Libs and Headers

Export the MmRpc generated libraries to usr/lib and
export the header file to user/include/ti/ipc/mm.

This is so that apps can easily include the header and
link with the library.

Signed-off-by: Angela Stegmaier <angelabaker@ti.com>
10 years agoReplaced usage of pathconf() with PATH_MAX
vwan@ti.com [Tue, 18 Jun 2013 19:18:02 +0000 (12:18 -0700)]
Replaced usage of pathconf() with PATH_MAX

pathconf() returns error when file is located in /tmp. We switch to using PATH_MAX to avoid this problem.

Signed-off-by: VW <vwan@ti.com>
10 years agoVayu: Increased processor table size in NotifySetup module
Ramsey Harris [Tue, 18 Jun 2013 22:11:28 +0000 (15:11 -0700)]
Vayu: Increased processor table size in NotifySetup module

When building for all cores on Vayu device, with non-SMP mode
on IPU1, the processor table was too small.

Signed-off-by: Ramsey Harris <ramsey@ti.com>
10 years agoMmRpc: Remove 'base' field from MmRpc_Xlt struct
Ramsey Harris [Tue, 18 Jun 2013 01:06:42 +0000 (18:06 -0700)]
MmRpc: Remove 'base' field from MmRpc_Xlt struct

This field was confusing and redundant. Given that the address
and offset to the embedded pointer are already available, the
base value can be derived in the MmRpc API layer.

Signed-off-by: Ramsey Harris <ramsey@ti.com>
10 years agoQNX: Remove tiler and update makefiles
vwan@ti.com [Mon, 17 Jun 2013 22:06:05 +0000 (15:06 -0700)]
QNX: Remove tiler and update makefiles

Tiler is now [correctly] managed independent from IPC.  This
commit removes tiler support from IPC.

10 years agoTests: Update tests to support concurrently loaded slaves
vwan@ti.com [Mon, 17 Jun 2013 05:00:59 +0000 (22:00 -0700)]
Tests: Update tests to support concurrently loaded slaves

Update the tests so that each core registers a different
endpoint name.  In this way multiple images can be loaded
at the same time and these tests can be run for each core
without having to unload and reload another image.

10 years agoTests: Update RPC Stress test
Angela Stegmaier [Mon, 17 Jun 2013 04:55:01 +0000 (21:55 -0700)]
Tests: Update RPC Stress test

Update the test so that each core registers a different
endpoint name.  In this way multiple images can be loaded
at the same time and this test can be run for each core
without having to unload and reaload another image.

Signed-off-by: Angela Stegmaier <angelabaker@ti.com>
10 years agoQNX: Don't Disable Host->rproc Interrupts During Register
vwan@ti.com [Mon, 17 Jun 2013 04:36:16 +0000 (21:36 -0700)]
QNX: Don't Disable Host->rproc Interrupts During Register

The register/unregister functions are clearing mailbox
messages and disabling the host->rproc interrupts.  When loading
multiple remote procs, this can cause an issue where the
remote proc's host->rproc interrupt remains disabled and the
remote proc is unable to receive interrupts.

This patch removes the code to avoid disabling the interrupt.

10 years agoBIOS: Remove M3 executable support for Vayu and OMAP5
Chris Ring [Sat, 15 Jun 2013 18:52:43 +0000 (11:52 -0700)]
BIOS: Remove M3 executable support for Vayu and OMAP5

No longer build the OMAP5 and Vayu tests when the M3 target
is set.  We want to migrate users of these devices to the M4

Also, an issue in BIOS 6.35.02 restricts Vayu to only M4
target support, so this ensures BIOS 6.35.02 users don't
hit this issue.

10 years agoConfig and package updates for new XDCtools and SYS/BIOS builds
Ramsey Harris [Thu, 13 Jun 2013 23:14:08 +0000 (16:14 -0700)]
Config and package updates for new XDCtools and SYS/BIOS builds

This fixes build and configuration issues for the following imports:


Signed-off-by: Ramsey Harris <ramsey@ti.com>
10 years agoTests: Update tests to use DRA7XX platform instance
Chris Ring [Thu, 13 Jun 2013 22:34:17 +0000 (15:34 -0700)]
Tests: Update tests to use DRA7XX platform instance

Teach ti.ipc.tests to use instances of the
ti.platforms.evmDRA7XX platform rather than the IPC-specific
vayu platforms.  These IPC-specific platforms can now be removed,
although we'll save that for the 3.10 stream so we don't break
existing users of them.

Note the two subtle changes in ipc-bios.bld.  First, the
platform name should be 'ti.platforms.evmDRA7XX' (note the
plural 'platforms').  Second, the platform defined by XDC includes
an instance named "IPU2", so our locally defined instance name
cannot match that - so it's changed to lowercase 'ipu2'.  Going
forward, we should standardize on platform instances with
all lower-case names.

Signed-off-by: Chris Ring <cring@ti.com>
10 years agoVayu: More support for the DRA7XX part number
Chris Ring [Thu, 13 Jun 2013 22:26:14 +0000 (15:26 -0700)]
Vayu: More support for the DRA7XX part number

'DRA7XX' and 'Vayu' are the same device.  An earlier
commit added support for 'DRA7XX' but missed a few places,
which this commit addresses.

These missing changes were found when enabling platform instance
support to HLOS-based examples, so these changes are isolated to
packages used in HLOS environments.

Signed-off-by: Chris Ring <cring@ti.com>
10 years agoVayu: Remove unnecessary dependency in InterruptIpu
Chris Ring [Wed, 12 Jun 2013 22:13:23 +0000 (15:13 -0700)]
Vayu: Remove unnecessary dependency in InterruptIpu

There was an unnecessary dependency from the IPU interrupt handler
(ti.sdo.ipc.family.vayu.InterruptIpu) onto the core of IPC
(ti.sdo.ipc.Ipc).  For the MmServiceMgr use case, which needs the
interrupt handler but not the core of IPC, this dependency brought
in unnecessary code/data.

Signed-off-by: Chris Ring <cring@ti.com>
10 years agoMmServiceMgr: Hide internal MmServiceMgr_start() fxn
Chris Ring [Wed, 12 Jun 2013 22:12:35 +0000 (15:12 -0700)]
MmServiceMgr: Hide internal MmServiceMgr_start() fxn

Minor documentation change to hide the internal
MmServiceMgr_start() from doxygen-generated API reference guides.

Signed-off-by: Chris Ring <cring@ti.com>
10 years agoVayu: Enable configuration of IPU mailbox virt addrs
Chris Ring [Fri, 7 Jun 2013 21:59:17 +0000 (14:59 -0700)]
Vayu: Enable configuration of IPU mailbox virt addrs

IPC must access the physical mailbox addresses (physically
at 0x4XXX_XXXX) via virtual addresses (typically set up
using the IPU's AMMU).  System integrators need the
flexibility to place these mailboxes at different virtual

System integrators can now use the InteruptIpu.mailboxBaseAddr[]
config params to place these mailboxes at virtual
addresses other than the default (which is 0x6XXX_XXXX).

Also clarified the cdoc documentation.

Signed-off-by: Chris Ring <cring@ti.com>
10 years agoQNX IPC: Main - Use Absolute Path of Firmware
Angela Stegmaier [Thu, 30 May 2013 18:52:38 +0000 (13:52 -0500)]
QNX IPC: Main - Use Absolute Path of Firmware

Save the absolute path of the firmware file to use
when loading the firmware.

This is useful for the recovery situation, where the
relative path will no longer work.

Signed-off-by: Angela Stegmaier <angelabaker@ti.com>
10 years agoAdded support for new platform: ti.platforms.evmDRA7XX
Ramsey Harris [Sun, 9 Jun 2013 22:26:03 +0000 (15:26 -0700)]
Added support for new platform: ti.platforms.evmDRA7XX

Added evmDRA7XX alias for Vayu in family Settings.xs. Removed
references to simVayu platform.

Signed-off-by: Ramsey Harris <ramsey@ti.com>
10 years agoQNX IPC: RPMSG-RPC - Don't Wait for Response During Destroy
Angela Stegmaier [Wed, 29 May 2013 16:55:25 +0000 (11:55 -0500)]
QNX IPC: RPMSG-RPC - Don't Wait for Response During Destroy

When calling rpmsg_rpc_destroy, all existing connections
are disconnected so that the remote core may be reloaded
and stale connections are not left open.

When rpmsg_rpc_destroy is called, it is assumed that
the remote processor has already been stopped.  In this
case, there is no need to wait for a response from the
disconnect call.  Instead, the connection can immediately
be marked as disconnected and destroy can continue.

This patch modifies each remote core connection to have
a parameter that tells whether or not the destroy function
is currently executing.  Then, the disconnect call is
updated to check to see if destroy is being executed.  If
destroy is being executed, then it does not send the
message to the remote core and wait for a response.  Instead,
the connection is simply marked as disconnected.

Signed-off-by: Angela Stegmaier <angelabaker@ti.com>
10 years agoQNX IPC: Main - Fix Compile Warnings
Angela Stegmaier [Wed, 29 May 2013 15:35:46 +0000 (10:35 -0500)]
QNX IPC: Main - Fix Compile Warnings

This patch fixes the following compilation warnings:

In function 'deinit_syslink_device':
warning: unused variable 'i'
In function 'deinit_syslink_trace_device':
warning: control reaches end of non-void function
In function 'init_syslink_trace_device':
warning: control reaches end of non-void function

Signed-off-by: Angela Stegmaier <angelabaker@ti.com>
10 years agoQNX IPC: Vayu - Fix Compile Warnings
Angela Stegmaier [Wed, 29 May 2013 15:31:10 +0000 (10:31 -0500)]
QNX IPC: Vayu - Fix Compile Warnings

This patch fixes the following compilation warnings:

In function 'VAYUIPUCORE1_halResetCtrl':
warning: unused variable 'counter'

In function 'VAYUIPUPWR_on':
warning: implicit declaration of function 'VAYUIPUMMU_enable'
In function 'VAYUIPUPWR_off':
warning: implicit declaration of function 'VAYUIPUMMU_disable'

Signed-off-by: Angela Stegmaier <angelabaker@ti.com>
10 years agoQNX IPC: Samples - Fix Compile Warnings
Angela Stegmaier [Wed, 29 May 2013 15:38:06 +0000 (10:38 -0500)]
QNX IPC: Samples - Fix Compile Warnings

This patch fixes the following compilation warnings:

In function 'test_exec_call':
warning: assignment makes integer from pointer without a cast
warning: format '%x' expects type 'unsigned int', but argument 2 has type
'int *'
warning: format '%x' expects type 'unsigned int', but argument 3 has type
'void *'

Signed-off-by: Angela Stegmaier <angelabaker@ti.com>
10 years agoResource: Introduce customTable to override resource table
Chris Ring [Tue, 28 May 2013 20:25:13 +0000 (13:25 -0700)]
Resource: Introduce customTable to override resource table

Users need to override the resource table.  Introduce the
customTable boolean config param which uses set to true
to indicate they will provide their own table (and IPC shouldn't
generate a table for them).

They then have the responsibility to provide their own table
(with a symbol named ti_ipc_remoteproc_ResourceTable) at link

10 years agoRemoteProc: rework resource_table definition/usage
Chris Ring [Fri, 31 May 2013 01:50:48 +0000 (18:50 -0700)]
RemoteProc: rework resource_table definition/usage

Introduce the resource_table struct (defined in Linux) and
rework the device/core-specific resource tables to derive from
that struct (rather than redefine the struct themselves.

No functional difference, but this provides a better
framework and example for end users when they (soon!) are enabled
to define their own resource_table-based structs.

10 years agoTest: Remove unnecessary trace
Chris Ring [Fri, 31 May 2013 17:48:54 +0000 (10:48 -0700)]
Test: Remove unnecessary trace

Minor test cleanup, removing unnecessary trace statements.

Removed trivial copy-n-paste comment at the end of the file.

10 years agoFix MessageQMulti's multi_process.sh script for QNX and install it
vwan@ti.com [Thu, 30 May 2013 23:05:12 +0000 (16:05 -0700)]
Fix MessageQMulti's multi_process.sh script for QNX and install it

10 years agoFixed build warning in MessageQMulti test
vwan@ti.com [Thu, 30 May 2013 20:36:47 +0000 (13:36 -0700)]
Fixed build warning in MessageQMulti test

Signed-off-by: Chris Ring <cring@ti.com>
10 years agoCleanup: remove directory qnx/src/ipc3x_dev/ti/syslink/rpmsg-resmgr
vwan@ti.com [Fri, 31 May 2013 17:02:55 +0000 (10:02 -0700)]
Cleanup: remove directory qnx/src/ipc3x_dev/ti/syslink/rpmsg-resmgr

OMAP5 power management code in this directory is moved to qnx/src/ipc3x_dev/ti/syslink/family/omap5430, and the power management code itself was modified to remove checks with rpmsg-resmgr.

Signed-off-by: VW <vwan@ti.com>
10 years agoCleanup: remove directory qnx/src/ipc3x_dev/ti/syslink/rpmsg-dce
vwan@ti.com [Wed, 29 May 2013 00:22:09 +0000 (17:22 -0700)]
Cleanup: remove directory qnx/src/ipc3x_dev/ti/syslink/rpmsg-dce

Signed-off-by: VW <vwan@ti.com>
10 years agoCleanup: remove directory qnx/src/ipc3x_dev/ti/syslink/rpmsg-omx
vwan@ti.com [Wed, 29 May 2013 00:17:25 +0000 (17:17 -0700)]
Cleanup: remove directory qnx/src/ipc3x_dev/ti/syslink/rpmsg-omx

Signed-off-by: VW <vwan@ti.com>
10 years agoCleanup: removed directory qnx/src/ipc3x_dev/perfutils
vwan@ti.com [Tue, 28 May 2013 23:44:57 +0000 (16:44 -0700)]
Cleanup: removed directory qnx/src/ipc3x_dev/perfutils

Signed-off-by: VW <vwan@ti.com>
10 years agoChanged Makefile.am to enable 'make dist' of files in hlos_common for Linux
vwan@ti.com [Tue, 28 May 2013 23:34:08 +0000 (16:34 -0700)]
Changed Makefile.am to enable 'make dist' of files in hlos_common for Linux

Signed-off-by: VW <vwan@ti.com>
10 years agoMoved utils/MultiProc.c into hlos_common directory for both Linux and QNX.
vwan@ti.com [Sat, 25 May 2013 00:20:21 +0000 (17:20 -0700)]
Moved utils/MultiProc.c into hlos_common directory for both Linux and QNX.

Signed-off-by: VW <vwan@ti.com>
10 years agoMoved Std.h into ti/ipc namespace for QNX.
vwan@ti.com [Tue, 28 May 2013 22:31:56 +0000 (15:31 -0700)]
Moved Std.h into ti/ipc namespace for QNX.

Signed-off-by: VW <vwan@ti.com>
10 years agoMoved headers common to both Linux and QNX into hlos_common directory.
vwan@ti.com [Fri, 24 May 2013 22:37:23 +0000 (15:37 -0700)]
Moved headers common to both Linux and QNX into hlos_common directory.

This is done to avoid unnecessary duplication of code.

Signed-off-by: VW <vwan@ti.com>
10 years agoVayu: Fixed AMMU fault when running SYS/BIOS on IPU1
Ramsey Harris [Thu, 30 May 2013 02:33:15 +0000 (19:33 -0700)]
Vayu: Fixed AMMU fault when running SYS/BIOS on IPU1

Prevent invalid mailbox accesses for interrupts between core0 and core1
because those interrupts do not use mailboxes.

10 years agoQNX IPC: Vayu - Add DSP1 MMU1 Programming
Angela Stegmaier [Wed, 29 May 2013 16:12:25 +0000 (09:12 -0700)]
QNX IPC: Vayu - Add DSP1 MMU1 Programming

Currently, only MMU0 is being programmed for DSP1.
For EDMA traffic, there is a second MMU (MMU1), which
needs to be programmed.  DSP1 applications which
use EDMA would want to also have the MMU1 programmed
so that the address space is the same.

This patch adds programming of the DSP MMU1
with the same entries to match the programming
of the DSP MMU0.

Additionally, this patch attaches an interrupt
to the DSP1 MMU1 for notification of MMU events.
No default interrupt is programmed for DSP1 MMU1,
so an unused MPU interrupt is selected for use and
the interrupt crossbar is programmed accordingly.
The interrupt selected is MPU interrupt 143.

10 years agoOMAP4: Removed files specific to OMAP4 in qnx directory
vwan@ti.com [Fri, 24 May 2013 22:47:05 +0000 (15:47 -0700)]
OMAP4: Removed files specific to OMAP4 in qnx directory

This commit removes all source files that are specific to OMAP4,
which the product does not support. Note that there are still file
names, variable names, macros, etc. related to OMAP4 that will need
subsequent clean up.

Files in tiler and rpmsg-resmgr directories have been left in place
as we are discussing their removal in an on-going thread.

A minor change was needed in syslink_main.c as well in order to
eliminate its use of type 'u32' defined in rprcfmt.h, so that we can
remove the latter file, which is specific to the OMAP4 loader.

Signed-off-by: VW <vwan@ti.com>
10 years agoStyle: Change //-style comments to /* */ style
Chris Ring [Wed, 22 May 2013 22:49:45 +0000 (15:49 -0700)]
Style: Change //-style comments to /* */ style

Some users, especially those with MISRA-C requirements, don't
allow //-style comments.

Note that this addresses the files in packages/ti/ipc/*, there may
be more, but I didn't hunt further.

10 years agoBIOS: Fix minor typo in comment/docs
Chris Ring [Mon, 20 May 2013 20:07:02 +0000 (13:07 -0700)]
BIOS: Fix minor typo in comment/docs

10 years agoTCI6614: Improve error message during config
Chris Ring [Mon, 20 May 2013 20:00:02 +0000 (13:00 -0700)]
TCI6614: Improve error message during config

When this package is integrated into an unsupported environment,
it provides a better error message.

For example, before this change, when [errantly] integrating into
an OMAP5-based system, configuration would report:
    TypeError: Cannot read property "IPCGR0" from undefined

Now it reports:
    ti.ipc.family.tci6614.Interrupt : OMAP5430 unsupported

10 years agoRemoved duplicated MultiProc config info from Platform.c and simply use the info...
vwan@ti.com [Thu, 16 May 2013 18:11:28 +0000 (11:11 -0700)]
Removed duplicated MultiProc config info from Platform.c and simply use the info from _MultiProc_cfg.

10 years agoRemoved libcfg.a and moved _MultiProc_cfg into resource manager for QNX.
vwan@ti.com [Thu, 16 May 2013 17:34:02 +0000 (10:34 -0700)]
Removed libcfg.a and moved _MultiProc_cfg into resource manager for QNX.

This is done to match a similar change in Linux that moved _MultiProc_cfg into the daemon process,

A side-effect is that Ipc_start must now be called prior to any MultiProc API.

10 years agoRemove BUILD_FOR_VIRTIO from products.mak and added support for DEVICE simvayu
vwan@ti.com [Mon, 13 May 2013 18:56:32 +0000 (11:56 -0700)]
Remove BUILD_FOR_VIRTIO from products.mak and added support for DEVICE simvayu

DEVICE simvayu corresponds to a simulated vayu device

Signed-off-by: Chris Ring <cring@ti.com>
10 years agoIPC: Added conditional variable assignments to missing
Arnie Reynoso [Thu, 9 May 2013 20:24:18 +0000 (13:24 -0700)]
IPC: Added conditional variable assignments to missing

Signed-off-by: Chris Ring <cring@ti.com>
10 years agoNameServerRemoteRpmsg: Fixed strncpy operation to prevent overflow.
G Anthony [Wed, 15 May 2013 23:41:35 +0000 (16:41 -0700)]
NameServerRemoteRpmsg: Fixed strncpy operation to prevent overflow.

This is an improved solution to the patch:

195d2692fa95224d3c634daf8fe9bbd3860c7e16 BIOS: Fixed NameServerRemoteRpmsg MessageQ instanceName and name length

which now uses the actual destination length as the max characters for
strncpy() to copy, and ensures null termination of the copied string.

Improvement pointed out by Rob Tivy.

Signed-off-by: G Anthony <a0783926@ti.com>
10 years agoVirtQueue: VirtQueue_init() reads vring addr from resource table, checks MAR.
G Anthony [Tue, 7 May 2013 01:21:13 +0000 (18:21 -0700)]
VirtQueue: VirtQueue_init() reads vring addr from resource table, checks MAR.

Previously, VirtQueue used hardcoded vring addresses in the code.  Now, these
are read from the Resource Table, allowing the host to update them at load

However, it is still expected that the vring memory is non-cached, and that
the system configuration has set this up in the CFG file.

VirtQueue_init() asserts if the vring addresses are not in the resource table,
or if the MAR bits for the vring memory show it as being caceable.

This is done only for OMAPL138 and TCI6638 VirtQueue in this patch.

Signed-off-by: G Anthony <a0783926@ti.com>
10 years agoResource: Added accessor to read vring addresses from resource table.
G Anthony [Mon, 6 May 2013 23:56:06 +0000 (16:56 -0700)]
Resource: Added accessor to read vring addresses from resource table.

This will support the capability of allowing the host to allocate the vrings,
and specify the vring control structure address to the slave.

Also, renamed the existing getEntry() method to getMemEntry(), as it only
works for CARVEOUT and DEVMEM types.

Signed-off-by: G Anthony <a0783926@ti.com>
10 years agoTCI6638: Switched to ROM model (-c) from RAM model (-cr) link option.
G Anthony [Mon, 6 May 2013 22:56:16 +0000 (15:56 -0700)]
TCI6638: Switched to ROM model (-c) from RAM model (-cr) link option.

Kestone II DSP images were not able to be reloaded without rebooting the EVM.
This was due to the "-cr" linker options being thrown (RAM model), which
prevented processing of .cinit records at run time (done in ROM model).

So, "-cr" option was removed.

Keystone team mentioned that RLE compression was keeping the resource table
from being read properly without the "-cr" option, so added in the
"--cinit_compression=off" option just in case.  This also has the benefit
of improving DSP boot times.

Reloading the DSPs via MPM server was validated on Keystone II EVM
(over remote connection).

Signed-off-by: G Anthony <a0783926@ti.com>
10 years agoTCI6638: Linux target side scripts for loading, running lad, dumping trace.
G Anthony [Mon, 29 Apr 2013 23:39:33 +0000 (16:39 -0700)]
TCI6638: Linux target side scripts for loading, running lad, dumping trace.

Useful scripts to be used for the 8 core TCI6638 platformm.

Signed-off-by: G Anthony <a0783926@ti.com>
10 years agoLinux: Test - Cleanup to mmrpc build after sources moved to a common location
Arnie Reynoso [Fri, 10 May 2013 18:31:31 +0000 (11:31 -0700)]
Linux: Test - Cleanup to mmrpc build after sources moved to a common location

Signed-off-by: Chris Ring <cring@ti.com>
10 years agoLinux: Building test applications as platform independent where appropriate
Arnie Reynoso [Fri, 10 May 2013 18:12:15 +0000 (11:12 -0700)]
Linux: Building test applications as platform independent where appropriate

Platform dependent application's are no longer needed since a MultiProc
            LAD daemon has been added.

Signed-off-by: Chris Ring <cring@ti.com>
10 years agoIPC: Tests - Add an Error Test Case
Angela Stegmaier [Tue, 7 May 2013 09:41:45 +0000 (04:41 -0500)]
IPC: Tests - Add an Error Test Case

Add an error test case to test for MMU fault

Signed-off-by: Angela Stegmaier <angelabaker@ti.com>
10 years agoQNX IPC: DSP Fault Recovery Fix
Angela Stegmaier [Tue, 7 May 2013 09:39:31 +0000 (04:39 -0500)]
QNX IPC: DSP Fault Recovery Fix

Fix an issue where the procId and procHandle are not
being properly set.  If it is not properly set, then
error recovery doesn't happen properly.

This patch fixes that so that error recovery can happen

Signed-off-by: Angela Stegmaier <angelabaker@ti.com>
10 years agoQNX IPC: Add IPU2 MMU Fault Handling
Angela Stegmaier [Tue, 7 May 2013 09:02:45 +0000 (04:02 -0500)]
QNX IPC: Add IPU2 MMU Fault Handling

There is no default interrupt mapped to the IPU2 MMU
for Vayu. This patch programs the interrupt crossbar
for the IPU2 MMU interrupt to MPU interrupt 142.

Also, this patch enables receiving the interrupt in the
case of an MMU fault event in order to print the required
information, as well as schedule recovery.

Signed-off-by: Angela Stegmaier <angelabaker@ti.com>
10 years agoQNX IPC: RpmsgRpc - Fix Tiler Address Translations
Angela Stegmaier [Mon, 6 May 2013 20:38:01 +0000 (15:38 -0500)]
QNX IPC: RpmsgRpc - Fix Tiler Address Translations

Tiler address translations should be done thorugh the
ProcMgr_translate function instead of just assuming
the remote core VA is the same as the PA.

Signed-off-by: Angela Stegmaier <angelabaker@ti.com>
Signed-off-by: Chris Ring <cring@ti.com>
10 years agoIPC - Match Resource Table for Vayu Based on Device
Angela Stegmaier [Wed, 8 May 2013 22:56:50 +0000 (17:56 -0500)]
IPC - Match Resource Table for Vayu Based on Device

Match the resource table for Vayu based on the Device
name instead of the Platform name.

Currently, if the platform simVayu is used, it fails
to find the resource table.  If the match is changed
to match the device name, then the simVayu platform,
which uses the Vayu device name, can compile.

Signed-off-by: Chris Ring <cring@ti.com>
10 years agoChanged all references to IPU's Core 0 to 'IPU' when in SMP mode for OMAP5.
vwan@ti.com [Tue, 7 May 2013 22:32:46 +0000 (15:32 -0700)]
Changed all references to IPU's Core 0 to 'IPU' when in SMP mode for OMAP5.

10 years agoBIOS: Minor log statement formatting
Chris Ring [Wed, 8 May 2013 01:04:45 +0000 (18:04 -0700)]
BIOS: Minor log statement formatting

Remove some new lines and misc characters to better align with
other log statements throughout IPC.

10 years agoBIOS: Add missing dependency (rpmsg uses List)
Chris Ring [Wed, 8 May 2013 00:58:00 +0000 (17:58 -0700)]
BIOS: Add missing dependency (rpmsg uses List)

Add the rpmsg dependency on List.  This removes the need for
users to know this implementation detail of rpmsg and explicitly
bring in the List module themselves.

10 years agoBIOS: Remove unused include file
Chris Ring [Wed, 8 May 2013 00:42:07 +0000 (17:42 -0700)]
BIOS: Remove unused include file

10 years agoTest: Build script refactoring
Chris Ring [Tue, 7 May 2013 22:56:16 +0000 (15:56 -0700)]
Test: Build script refactoring

Consolidate some of the test package.bld script.  No functional

10 years agoOMAP5: Switch from custom DSP platforms to platform instances
Chris Ring [Tue, 7 May 2013 16:55:39 +0000 (09:55 -0700)]
OMAP5: Switch from custom DSP platforms to platform instances

Updated build scripts to use platform instances (currently of
ti.platforms.sdp5430) rather than the custom DSP platform
in ipcdev.

Remove ti.platform.omap54xx.dsp.

10 years agoLinux: Add MultiProc trace to log file
Chris Ring [Mon, 6 May 2013 23:54:25 +0000 (16:54 -0700)]
Linux: Add MultiProc trace to log file

To improve visibility and debug, log the MultiProc configuration in the
LAD log file (when enabled) when a client asks for it.

10 years agoOMAP5: Rework package validation check for SMP-BIOS
Chris Ring [Mon, 6 May 2013 23:45:16 +0000 (16:45 -0700)]
OMAP5: Rework package validation check for SMP-BIOS

10 years agoOMAP5: Switch from custom IPU platforms to platform instances
Chris Ring [Mon, 29 Apr 2013 20:10:52 +0000 (13:10 -0700)]
OMAP5: Switch from custom IPU platforms to platform instances

Updated build scripts to use platform instances (currently of
ti.platforms.sdp5430) rather than the custom IPU platform
in ipcdev.

Remove ti.platform.omap54xx.ipu.

Also, since OMAP5 requires SMP-BIOS - remove the
ti.platform.omap54xx.core* platforms as well.

10 years agoLinux: Fix OMAP5 IPU MultiProc name
Chris Ring [Tue, 7 May 2013 17:14:36 +0000 (10:14 -0700)]
Linux:  Fix OMAP5 IPU MultiProc name

The Linux-side MultiProc names must be aligned with the BIOS-side

10 years agoVayu: New crossbar mapping for ipu2 and dsp2 to host interrupt
Ramsey Harris [Sun, 5 May 2013 18:21:15 +0000 (11:21 -0700)]
Vayu:  New crossbar mapping for ipu2 and dsp2 to host interrupt

Original crossbar instance does not exist.

10 years agoRevert "QNX IPC: WA For DSP Not Receiving Messages"
Angela Stegmaier [Sat, 4 May 2013 09:27:07 +0000 (04:27 -0500)]
Revert "QNX IPC: WA For DSP Not Receiving Messages"

This reverts commit f83beccbebf69871fcd15d65525d0a8802850afa.


10 years agoQNX IPC: DSP1 - Need to Enable DSP_SYSTEM Config for MMU0
Angela Stegmaier [Fri, 3 May 2013 14:02:31 +0000 (09:02 -0500)]
QNX IPC: DSP1 - Need to Enable DSP_SYSTEM Config for MMU0

If using DSP1 with the MMU enabled, need to also enable
the global config for MMU0 through the DSP_SYS_MMU_CONFIG
MMU0_EN bit.

Signed-off-by: Angela Stegmaier <angelabaker@ti.com>
10 years agoIPC: DSP1 - Re-arrange Memory Map
Angela Stegmaier [Fri, 3 May 2013 13:46:37 +0000 (08:46 -0500)]
IPC: DSP1 - Re-arrange Memory Map

Move the text section out of the 0x20000000 virtual address space
and group together the sections at 0x95000000 virtual address.

Signed-off-by: Angela Stegmaier <angelabaker@ti.com>
10 years agoQNX IPC: Samples - Only Open SharedMemLib Once
Angela Stegmaier [Fri, 3 May 2013 14:09:33 +0000 (09:09 -0500)]
QNX IPC: Samples - Only Open SharedMemLib Once

A race condition in the rpmsg-rpc stress test could
result in one thread trying to access a NULL pointer
as there was no protection around the call to
initialize the sharedmemallocator library functions.

To avoid the race condition, move the initialization
of the sharedmemallocator functions to the main function,
to happen only once.

Signed-off-by: Angela Stegmaier <angelabaker@ti.com>
10 years agoQNX IPC: Update Int Xbar Programming to Match BIOS-side IPC
Angela Stegmaier [Fri, 3 May 2013 17:13:44 +0000 (12:13 -0500)]
QNX IPC: Update Int Xbar Programming to Match BIOS-side IPC

There was a conflict with the MPU Int being used by QNX IPC
for the DSP Mailbox.

The BIOS-side IPC is using MPU INT 141 for HOST receiving
interrupts from IPU2, and MPU INT 136 for HOST receiving
interrupts from DSP1.

QNX IPC is updated to use the same as we know these interrupts
are available.

Signed-off-by: Angela Stegmaier <angelabaker@ti.com>
10 years agoQNX IPC: Remove EVE from MultiProc as it is Not Used Currently
Angela Stegmaier [Fri, 3 May 2013 11:33:51 +0000 (06:33 -0500)]
QNX IPC: Remove EVE from MultiProc as it is Not Used Currently

If EVE is defined in the MultiProc, then when the IPU is
starting up, the interrupt module will try to access the EVE
mailboxes, which leads to a crash as we are not mapping
the EVE mailboxes.

This patch updates the MultiProc config to remove the EVEs
so that the EVE mailboxes are not accessed.

Signed-off-by: Angela Stegmaier <angelabaker@ti.com>
10 years agoIPC: Vayu - Fix GPTimer Values and Control Module AMMU Config
Angela Stegmaier [Thu, 2 May 2013 19:10:41 +0000 (14:10 -0500)]
IPC: Vayu - Fix GPTimer Values and Control Module AMMU Config

The GPTimer values in IpcCommon.cfg.xs were using the
old virtual addresses that started with 0xA8.  A
recent patch changed the virtual address mapping for
0x48000000 from 0xA8000000 to 0x68000000.  The
GPTimer base addresses missed being updated.

This patch updates the GPTimer values to the correct
virtual addresses.

Also, the control module entry in IpuAmmu.cfg is
updated to do a translation from 0x4a000000 to
0x6a000000, to take advantage of the existing
mapping of the 0x6a000000 to the 0x4a000000 registers.

Signed-off-by: Angela Stegmaier <angelabaker@ti.com>
10 years agoBuild: Don't build big endian tests for Vayu or TCI6638
Chris Ring [Wed, 1 May 2013 03:39:05 +0000 (20:39 -0700)]
Build: Don't build big endian tests for Vayu or TCI6638

Limit the platform array for the C66 targets by only adding
TCI6638 and Vayu to the C66 little endian target, not the
big endian one.

10 years agoBuild: Remove errant warning when building for Vayu
Chris Ring [Wed, 1 May 2013 03:36:28 +0000 (20:36 -0700)]
Build: Remove errant warning when building for Vayu

The resource table generation during configuration (by the
Resource.xdt template) errantly warns about not being able
to create the resource table.

Restructure the code to remove the warning for this successful
and valid use case.

10 years agoDocs: Minor doc typo fix
Chris Ring [Tue, 30 Apr 2013 23:17:45 +0000 (16:17 -0700)]
Docs: Minor doc typo fix

10 years agoLinux: Moved IPC specific types file (Std.h) into the ti/ipc namespace
Arnie Reynoso [Tue, 30 Apr 2013 21:07:49 +0000 (14:07 -0700)]
Linux: Moved IPC specific types file (Std.h) into the ti/ipc namespace

10 years agoBIOS: Fix ti.ipc.transport build script for rpmsg2-based libs
Angela Stegmaier [Fri, 26 Apr 2013 22:48:50 +0000 (17:48 -0500)]
BIOS: Fix ti.ipc.transport build script for rpmsg2-based libs

Also rework ti.ipc.transport build scripts to remove unnecessary
platform-specific lib names.

Also update BIOS-side example to indicate the correct rpmsg version.

10 years agoQNX: Return E_NOTFOUND instead of E_FAIL if MQCopy_send fails
Angela Stegmaier [Mon, 29 Apr 2013 17:42:20 +0000 (12:42 -0500)]
QNX: Return E_NOTFOUND instead of E_FAIL if MQCopy_send fails

This allows MessageQ NameServer lookups to slaves that aren't
loaded to gracefully fail.

10 years agoIn VirtQueue_create, clear the VRING_USED_F_NO_NOTIFY flag to ensure all VirtQueue_ki...
VW [Mon, 29 Apr 2013 18:38:46 +0000 (11:38 -0700)]
In VirtQueue_create, clear the VRING_USED_F_NO_NOTIFY flag to ensure all VirtQueue_kick calls from the host results in an interrupt on slave.

10 years agoChanged GateAll to GateHwi since GateAll does not work in SMP BIOS according to SDOCM...
vwan@ti.com [Thu, 25 Apr 2013 17:42:22 +0000 (10:42 -0700)]
Changed GateAll to GateHwi since GateAll does not work in SMP BIOS according to SDOCM00100833

10 years agoFixed MultiProc config so that device names match the slave side on OMAP5
vwan@ti.com [Thu, 25 Apr 2013 17:40:05 +0000 (10:40 -0700)]
Fixed MultiProc config so that device names match the slave side on OMAP5

10 years agoQNX IPC: Vayu - Update Xbar Config for 16_eng
Angela Stegmaier [Fri, 26 Apr 2013 14:44:09 +0000 (09:44 -0500)]
QNX IPC: Vayu - Update Xbar Config for 16_eng

In the 16_eng release of IPC 3x, the mailbox
usages is updated, so the interrupt crossbar
configuration must be updated.  The new configuration
has the following mailbox usage:

Mailbox for DSP1->HOST is Mbox 5 User 2
Mailbox for IPU2->HOST is Mbox 6 User 2

Signed-off-by: Angela Stegmaier <angelabaker@ti.com>
10 years agoBIOS IPC: Vayu - Update AMMU and MMU Configuration
Angela Stegmaier [Wed, 24 Apr 2013 18:44:21 +0000 (13:44 -0500)]
BIOS IPC: Vayu - Update AMMU and MMU Configuration

In the 16_eng release, the mailbox virtual address
changed to 0x60000000 block.  This was previously where
the TILER addresses were mapped.  As a consequence, the
TILER mappings needed to be moved, and the AMMU config
as well as the resource table needed to be updated
to reflect the changes.  Also, as a result of the move,
the VRING virtual addresses for the IPU have changed.

TILER VA: Moved from 0x60000000 to 0xA0000000
Peripherals and VRING: Moved from 0xA0000000 to 0x60000000

Signed-off-by: Angela Stegmaier <angelabaker@ti.com>
10 years agoQNX IPC - Update to New Mailbox Usage
Angela Stegmaier [Tue, 23 Apr 2013 19:07:39 +0000 (14:07 -0500)]
QNX IPC - Update to New Mailbox Usage

With the latest 16_eng IPC 3.x release, the mailbox
usage has changed.

The QNX IPC is updated to be in sync with these
changes for communication with IPU2 and DSP1.

Signed-off-by: Angela Stegmaier <angelabaker@ti.com>
10 years agoQNX IPC: WA For DSP Not Receiving Messages
Angela Stegmaier [Wed, 24 Apr 2013 21:25:47 +0000 (16:25 -0500)]
QNX IPC: WA For DSP Not Receiving Messages

There is an issue observed when testing HOST->DSP1
communication using Vayu Virtio where if more than one
message is in the mailbox FIFO, then the DSP only
picks up one message, and the interrupt callback
stops running.

To avoid this issue, the workaround is to make sure
that we only send one message at a time from HOST->DSP1.

A side affect of this is that messages will be dropped,
but since (for the most part), the contents of the mailbox
message are the same, we can get away with this for now.

This may be just a Vayu Virtio issue, and we will
attempt to revert the WA in the future.

Signed-off-by: Angela Stegmaier <angelabaker@ti.com>
10 years agoQNX IPC: Build - Add Vayu Build to the QNX IPC
Angela Stegmaier [Mon, 8 Apr 2013 21:18:38 +0000 (16:18 -0500)]
QNX IPC: Build - Add Vayu Build to the QNX IPC

Add the Vayu Build to the QNX IPC and update the build
files to allow specification of the device in the
products.mak file.

Signed-off-by: Angela Stegmaier <angelabaker@ti.com>
10 years agoAdd Vayu Family Support for IPU2 (SMP) and DSP1
Angela Stegmaier [Mon, 1 Apr 2013 22:40:33 +0000 (17:40 -0500)]
Add Vayu Family Support for IPU2 (SMP) and DSP1

Add support for HOST->IPU2 and HOST->DSP1 support
for Vayu.

10 years agoQNX IPC: Vayu - Add Vayu IPU2 and DSP1 Support
Angela Stegmaier [Thu, 18 Apr 2013 18:34:39 +0000 (13:34 -0500)]
QNX IPC: Vayu - Add Vayu IPU2 and DSP1 Support

Add the Vayu family files to support IPC communication
between the HOST running QNX and the IPU2 and/or DSP1
running SYS/BIOS IPC3.x.

Signed-off-by: Angela Stegmaier <angelabaker@ti.com>
10 years agoAdd OMAP5 DSP Support for HLOS->SYSBIOS IPC
Angela Stegmaier [Mon, 11 Mar 2013 22:42:26 +0000 (17:42 -0500)]

Signed-off-by: Angela Stegmaier <angelabaker@ti.com>
10 years agoTCI6638: Linux target side scripts for loading, running lad, dumping trace.
G Anthony [Mon, 29 Apr 2013 23:39:33 +0000 (16:39 -0700)]
TCI6638: Linux target side scripts for loading, running lad, dumping trace.

Useful scripts to be used for the 8 core TCI6638 platformm.

Signed-off-by: G Anthony <a0783926@ti.com>
Signed-off-by: Chris Ring <cring@ti.com>
10 years agoResource: Added config option to build images with SysMin trace resource only.
G Anthony [Mon, 29 Apr 2013 21:05:00 +0000 (14:05 -0700)]
Resource: Added config option to build images with SysMin trace resource only.

Set Resource.traceOnly to true in order to get an image with trace resources,
and without VDEV and VRING resources.

This feature was requested by the Keystone II (TCI6638) team for testing,
and so for now is only implemented for that platform's resource table.

Signed-off-by: Chris Ring <cring@ti.com>
10 years agoBuild: Fix SMP support for platforms without 'ipu' in the name
Chris Ring [Mon, 29 Apr 2013 19:30:53 +0000 (12:30 -0700)]
Build: Fix SMP support for platforms without 'ipu' in the name

A few SMP-aware packages only support platforms with the name
'ipu' in them.  Remove this explicit platform name check and replace
it with a more correct check (e.g. by detecting BIOS.smpEnabled, or
using the platform's .deviceName field - which doesn't vary across
platform names).

10 years agoLinux: Moved MultiProc config into the platform specific LAD daemon.
Arnie Reynoso [Thu, 25 Apr 2013 21:57:06 +0000 (14:57 -0700)]
Linux: Moved MultiProc config into the platform specific LAD daemon.

Also updated example apps to correctly call Ipc_start() before any
MultiProc functions.