]>
rtime.felk.cvut.cz Git - frescor/frsh-forb.git/log
Michal Sojka [Fri, 18 Feb 2011 22:25:28 +0000 (23:25 +0100)]
forbrun: Simplify argument processing
Michal Sojka [Fri, 18 Feb 2011 22:03:58 +0000 (23:03 +0100)]
forb: Fix races between instances of forb_main_thread
The data passed to forb_main_thread must be allocated for each thread
separately. Otherwise, threads may access data intended for other threads.
Michal Sojka [Fri, 18 Feb 2011 21:40:50 +0000 (22:40 +0100)]
forb: Add tests for forbrun command line parameters
Michal Sojka [Fri, 18 Feb 2011 21:19:23 +0000 (22:19 +0100)]
forb: Update forbrun usage help
Michal Sojka [Thu, 17 Feb 2011 23:33:37 +0000 (00:33 +0100)]
Update Makefile.rules to the latest version
There is an improved support for wvtest framework.
Petr Benes [Thu, 17 Feb 2011 15:59:16 +0000 (16:59 +0100)]
forb: fixed passing parameters to shared libraries in forbrun
Michal Sojka [Thu, 17 Feb 2011 14:32:30 +0000 (15:32 +0100)]
forbrun: Exit whenever the first forb_main() returns
The exit code of forbrun is the returned value of forb_main().
Michal Sojka [Thu, 17 Feb 2011 14:29:37 +0000 (15:29 +0100)]
forb: Use new syntax for forbrun command line
We have added '--' before the name of the first library as it simplifies
command line parsing.
Michal Sojka [Thu, 17 Feb 2011 13:55:50 +0000 (14:55 +0100)]
forb: Add tests for forbrun
Michal Sojka [Thu, 17 Feb 2011 11:18:57 +0000 (12:18 +0100)]
forb: Move shared library tests under forbrun/ directory
Michal Sojka [Tue, 15 Feb 2011 16:12:15 +0000 (17:12 +0100)]
forb: First implementation of forbrun
Not tested! forbrun should executed FORB servers implemented as shared
libraries. Its aim is to allow execution of multiple independent servers
in a single process. This is done in preparation of porting FRSH/FORB to
RTEMS (and other single-address space executives).
Michal Sojka [Tue, 15 Feb 2011 15:24:32 +0000 (16:24 +0100)]
forb: Add support for synchronized initialization of multiple servers
This is needed in forbrun (future commit) to start multiple servers in a
single address space.
Michal Sojka [Tue, 15 Feb 2011 15:23:06 +0000 (16:23 +0100)]
forb: Add comments to forb_daemon_prepare()
Michal Sojka [Tue, 15 Feb 2011 13:40:05 +0000 (14:40 +0100)]
forb: Added shared library for testing of forbrun-based server execution
Petr Benes [Tue, 15 Feb 2011 12:24:14 +0000 (13:24 +0100)]
forb: object serialization working even for inter-executor invocation
in case the object being deserialized indicates being local,
it is not being created from scratch, however just copied
Petr Benes [Fri, 11 Feb 2011 15:22:43 +0000 (16:22 +0100)]
forb: End of header index removed from stub.
It is now embedded into the request structure. Does not make
so much mess in stubs.
Michal Sojka [Fri, 11 Feb 2011 15:07:40 +0000 (16:07 +0100)]
forb: Objects must be properly reference-counted
This fixes some nasty segfaults.
Petr Benes [Fri, 11 Feb 2011 14:35:55 +0000 (15:35 +0100)]
Inter-executor invocation bugs fixed.
Double free bug on request codec buffer fixed.
Stub index indicating a codec read position for inter-executor invocation fixed.
Test application->executor->executor does not fire properly. Needs a bug fix.
Petr Benes [Thu, 10 Feb 2011 14:27:26 +0000 (15:27 +0100)]
Struct request adjusted for inter-executor invocation.
Bug fix required.
Petr Benes [Tue, 8 Feb 2011 17:05:38 +0000 (18:05 +0100)]
Merge branch 'master' of rtime.felk.cvut.cz:/frescor/frsh-forb
Petr Benes [Tue, 8 Feb 2011 17:04:55 +0000 (18:04 +0100)]
Merge branch 'master' of rtime.felk.cvut.cz:/frescor/frsh-forb
Conflicts:
build/aquosa/wvtest
Michal Sojka [Tue, 8 Feb 2011 17:01:37 +0000 (18:01 +0100)]
forb: Add test for various invocation methods
This tests the following:
1) remote invocation (currently broken)
2) local invocation from application to the executor
3) local invocation from an executor to another one
4) direct invocation from an executor to the same one
Michal Sojka [Tue, 8 Feb 2011 16:45:13 +0000 (17:45 +0100)]
IDL compiler: Invoke methods directly only within the same executor
Michal Sojka [Tue, 8 Feb 2011 16:40:54 +0000 (17:40 +0100)]
Simplify frob_get_current_executor()
Return the executor as return value and not via a parameter.
Michal Sojka [Tue, 8 Feb 2011 15:34:33 +0000 (16:34 +0100)]
wvtest: Show stringified expressions in WVPASSEQ/NE
Previously, WVPASSEQ(func(), 0) printed only "0 == 0 ok", whereas now,
it prints "func() == 0 == 0 ok". This is useful when the test fails so
that the real value is seen in the output e.g. "func() == 1 == 0 FAILED".
Petr Benes [Tue, 8 Feb 2011 15:30:23 +0000 (16:30 +0100)]
Fixed synchronization for codec buffer.
Michal Sojka [Tue, 8 Feb 2011 15:29:56 +0000 (16:29 +0100)]
Makefile: Add missing header file
Petr Benes [Tue, 8 Feb 2011 14:38:21 +0000 (15:38 +0100)]
Fixed typography.
Petr Benes [Tue, 8 Feb 2011 13:53:32 +0000 (14:53 +0100)]
Merge branch 'master' of rtime.felk.cvut.cz:/frescor/frsh-forb
Petr Benes [Tue, 8 Feb 2011 13:52:56 +0000 (14:52 +0100)]
Prepared for multiple executors in a single address space
Object implementation invocation prepared for a use of multiple
executors running in a single address space.
One can call the iop methods in the same way as for the case of
a remote invocation.
Michal Sojka [Tue, 8 Feb 2011 08:42:03 +0000 (09:42 +0100)]
Align wvtest results
Now, the results of wvtest output is aligned to the 10 right characters
of the terminal.
Michal Sojka [Tue, 8 Feb 2011 08:41:49 +0000 (09:41 +0100)]
Remove a copy of wvtest
Use the version from src/
Michal Sojka [Tue, 8 Feb 2011 08:18:46 +0000 (09:18 +0100)]
Add some patterns to gitignore
Michal Sojka [Mon, 7 Feb 2011 21:41:09 +0000 (22:41 +0100)]
Merge branch 'master' of rtime.felk.cvut.cz:frescor/frsh-forb
Michal Sojka [Mon, 7 Feb 2011 20:49:25 +0000 (21:49 +0100)]
Document better forb_init()
Michal Sojka [Mon, 31 Jan 2011 09:00:49 +0000 (10:00 +0100)]
Split FOSA opaque types to multiple files based on platform name
The aim of this is to have more readable code. As we are now focused mainly
on Linux platform, we do not want to mistakenly find declarations for
other platforms.
Petr Benes [Fri, 28 Jan 2011 17:47:01 +0000 (18:47 +0100)]
Using plain POSIX instead of FOSA for thread specific data.
Michal Sojka [Fri, 28 Jan 2011 16:04:37 +0000 (17:04 +0100)]
Initialize thread specific data in forb_init()
This was done to avoid race conditions which causes errors in code.
Unfortunately this change does not help since FOSA is broken and we will
better use plain POSIX here.
Michal Sojka [Fri, 28 Jan 2011 15:03:07 +0000 (16:03 +0100)]
Michal Sojka [Thu, 27 Jan 2011 14:37:56 +0000 (15:37 +0100)]
Add script for splitting subtrees to the original repositories
Michal Sojka [Thu, 27 Jan 2011 14:22:41 +0000 (15:22 +0100)]
Update README and remove .gitmodules
Now, when all sources are in the single repository, our users do not need
to know about submodules at all.
Michal Sojka [Fri, 28 Jan 2011 14:58:51 +0000 (15:58 +0100)]
Michal Sojka [Fri, 28 Jan 2011 14:58:51 +0000 (15:58 +0100)]
Squashed 'src/wvtest/' content from commit
e54dada
git-subtree-dir: src/wvtest
git-subtree-split:
e54dada3feae72c9420d94c42840b423fad5bc9c
Michal Sojka [Thu, 27 Jan 2011 13:43:49 +0000 (14:43 +0100)]
Michal Sojka [Thu, 27 Jan 2011 13:43:14 +0000 (14:43 +0100)]
Michal Sojka [Thu, 27 Jan 2011 13:43:14 +0000 (14:43 +0100)]
Michal Sojka [Thu, 27 Jan 2011 13:43:13 +0000 (14:43 +0100)]
Michal Sojka [Thu, 27 Jan 2011 13:43:13 +0000 (14:43 +0100)]
Michal Sojka [Thu, 27 Jan 2011 13:43:12 +0000 (14:43 +0100)]
Michal Sojka [Thu, 27 Jan 2011 13:43:12 +0000 (14:43 +0100)]
Michal Sojka [Thu, 27 Jan 2011 13:21:03 +0000 (14:21 +0100)]
Remove submodules in order to be included as subtrees
See https://github.com/apenwarr/git-subtree
Michal Sojka [Thu, 27 Jan 2011 12:22:18 +0000 (13:22 +0100)]
Update submodules
Petr Benes [Thu, 27 Jan 2011 10:13:34 +0000 (11:13 +0100)]
Bug fixed in the function fosa_thread_get_specific_data()
Michal Sojka [Thu, 27 Jan 2011 10:27:38 +0000 (11:27 +0100)]
Merge branch 'master' of rtime.felk.cvut.cz:frescor/fwp
Conflicts:
fwp/mngr/fwp_mngr.c
Petr Beneš [Tue, 25 Jan 2011 16:13:36 +0000 (17:13 +0100)]
Prepared test cases for executors (thread specific data)
tests/executor_id.c as a simple test
tests/executor_calls.c as a manually created stub
Petr Beneš [Tue, 25 Jan 2011 16:10:14 +0000 (17:10 +0100)]
Added a function forb_object_get_executor() in object.h
Petr Beneš [Tue, 25 Jan 2011 16:08:54 +0000 (17:08 +0100)]
Added a function forb_get_current_executor()
Pavel Pisa [Fri, 21 Jan 2011 23:41:28 +0000 (00:41 +0100)]
uLUt library header files can be moved to include/ulut subdirectory.
If the option
CONFIG_OC_ULUT_INCDIR=y
is selected, the copy of uLUt library header files
is added into "ulut" subdirectory of include location
but copy in commonly used top level location is
provided as well.
If the
CONFIG_OC_ULUT_INCDIRONLY=y
option is enabled the the uLUt header files are placed
only in "ulut" subdirectory location.
Signed-off-by: Pavel Pisa <pisa@cmp.felk.cvut.cz>
Michal Sojka [Fri, 21 Jan 2011 20:25:01 +0000 (21:25 +0100)]
Merge branch 'master' of frsh-forb.git.sourceforge.net:/gitroot/frsh-forb/frsh-forb
Michal Sojka [Fri, 21 Jan 2011 20:24:52 +0000 (21:24 +0100)]
Merge branch 'master' of git://frsh-forb.git.sourceforge.net/gitroot/frsh-forb/frsh
Michal Sojka [Tue, 11 Jan 2011 09:19:57 +0000 (10:19 +0100)]
Add help option
Michal Sojka [Tue, 14 Sep 2010 13:20:56 +0000 (15:20 +0200)]
README: Show how to disable AQUOSA
Pavel Pisa [Fri, 5 Nov 2010 13:13:18 +0000 (14:13 +0100)]
SDCC 3.0 can build uLUt but its GCSE optimization goes mad for gavl_balance_one.
The common expression elimination optimization
leads to allocation of whole internal memory
when enabled for gavl_balance_one function.
Signed-off-by: Pavel Pisa <pisa@cmp.felk.cvut.cz>
Petr Beneš [Sun, 17 Oct 2010 23:48:07 +0000 (01:48 +0200)]
Added thread specific data to executor
Edited executor.h and executor.c
Requires a testcase
Michal Sojka [Fri, 15 Oct 2010 06:15:02 +0000 (08:15 +0200)]
Allow daemonizing FWP manager
Michal Sojka [Fri, 15 Oct 2010 06:14:46 +0000 (08:14 +0200)]
Allow daemonizing CPU resource managers
Michal Sojka [Fri, 15 Oct 2010 05:42:33 +0000 (07:42 +0200)]
fcb: Enable -h option (help)
Michal Sojka [Tue, 14 Sep 2010 13:20:56 +0000 (15:20 +0200)]
README: Show how to disable AQUOSA
Michal Sojka [Tue, 14 Sep 2010 13:17:55 +0000 (15:17 +0200)]
make: Disable kernel passes
We do not need to compile anything for the kernel, so disable the passes
to not confuse our users.
Michal Sojka [Tue, 14 Sep 2010 13:15:14 +0000 (15:15 +0200)]
Add prerequisites to README
Michal Sojka [Mon, 30 Aug 2010 14:28:02 +0000 (16:28 +0200)]
README: Add how to test the framework
Michal Sojka [Thu, 8 Jul 2010 14:17:13 +0000 (16:17 +0200)]
Figure: Remove DTM and change "Data exchange" to "Communication"
Michal Sojka [Sun, 27 Jun 2010 12:48:44 +0000 (14:48 +0200)]
Add elisp helper function for easier editing
Michal Sojka [Sat, 26 Jun 2010 21:13:02 +0000 (23:13 +0200)]
Add a comment with elisp helper function
I have this function defined in my .emacs and it automatically updates
frsh_error.c with the current list of errors. Whenever I add a new error,
I place the point after ")" and press C-x C-e to execute the function.
Michal Sojka [Fri, 25 Jun 2010 23:16:40 +0000 (01:16 +0200)]
Add basic transaction support
Michal Sojka [Fri, 25 Jun 2010 23:16:14 +0000 (01:16 +0200)]
Update transaction API
Michal Sojka [Fri, 25 Jun 2010 22:54:47 +0000 (00:54 +0200)]
Log time
Michal Sojka [Fri, 25 Jun 2010 22:54:17 +0000 (00:54 +0200)]
Allow specifying loglevels in environment
Michal Sojka [Fri, 25 Jun 2010 22:51:35 +0000 (00:51 +0200)]
Add ul_log to executor
Michal Sojka [Fri, 25 Jun 2010 21:46:56 +0000 (23:46 +0200)]
fcb: Do not allow multiple allocations of the same VRES (within a transaction)
Michal Sojka [Fri, 25 Jun 2010 21:29:23 +0000 (23:29 +0200)]
fcb: Reject direct contract manipulation in transactions
Michal Sojka [Fri, 25 Jun 2010 21:05:10 +0000 (23:05 +0200)]
frsh_transaction_alloc_vres works
Michal Sojka [Fri, 25 Jun 2010 20:12:09 +0000 (22:12 +0200)]
gui: Fix compile error caused by change in fres_contract.h
Michal Sojka [Fri, 25 Jun 2010 14:41:17 +0000 (16:41 +0200)]
Started implementation of frsh_transaction_alloc_vres()
Michal Sojka [Fri, 25 Jun 2010 14:18:34 +0000 (16:18 +0200)]
fcb: Handle transaction cancelation - not tested
Michal Sojka [Fri, 25 Jun 2010 12:38:08 +0000 (14:38 +0200)]
negotiate_transaction() hopefully completed
Michal Sojka [Fri, 25 Jun 2010 08:47:01 +0000 (10:47 +0200)]
Add fcb_remember_contracts() to simplify the code
Michal Sojka [Fri, 25 Jun 2010 07:45:33 +0000 (09:45 +0200)]
Suppress warning in fres_contract_id_to_string() on 64 bit system
Michal Sojka [Fri, 25 Jun 2010 07:44:34 +0000 (09:44 +0200)]
fcb: Deallocate res_array properly
Michal Sojka [Thu, 24 Jun 2010 22:53:17 +0000 (00:53 +0200)]
Progress on transaction negotiation
Michal Sojka [Thu, 24 Jun 2010 22:41:01 +0000 (00:41 +0200)]
Do not use random contract IDs
Instead use a counter in FCB which is used as a value of a contracts ID.
Currently, we have a centralized FCB, so this is no harm for us and we get
a more readable log output as well as repeatable ID for testing.
Michal Sojka [Thu, 24 Jun 2010 21:04:34 +0000 (23:04 +0200)]
fcb: Correct bug to pass the test suite
FCB didn't detect rejected contracts.
Michal Sojka [Thu, 24 Jun 2010 22:38:26 +0000 (00:38 +0200)]
Get compilable output for sequences of a typedeffed type
This change is a workaround rather than a real fix. This way be generate
unnecessarily more [de]serialization functions even if they do the same
as the function of the original type.
Michal Sojka [Thu, 24 Jun 2010 14:56:46 +0000 (16:56 +0200)]
fcb: Add reserve_resource helper function
TODO: Fix a bug in the implementation - test suite fails!
Michal Sojka [Thu, 24 Jun 2010 13:38:20 +0000 (15:38 +0200)]
fcb: Refactor resource reserve/commit
This allows sharing some code with transaction negotiation.
Michal Sojka [Wed, 23 Jun 2010 23:23:00 +0000 (01:23 +0200)]
fcb: Work on transaction support continues
Michal Sojka [Wed, 23 Jun 2010 21:56:37 +0000 (23:56 +0200)]
Small fixes
Michal Sojka [Wed, 23 Jun 2010 21:56:10 +0000 (23:56 +0200)]
fcb: Refactor prepare_fcb_contracts()
Michal Sojka [Wed, 23 Jun 2010 20:30:22 +0000 (22:30 +0200)]
Fix contract test
Critical sections were uninitialized and as such caused segfault.
Remove them, since they are not used in any resource.
Michal Sojka [Wed, 23 Jun 2010 15:39:26 +0000 (17:39 +0200)]
Transaction support - part 1
* src/forb
1e0ff4c ...
0f6c74a (1):
> test: Sequence is initilized to zero
* src/frsh
45d3688 ...
cd50efb (4):
> Start of transaction implementation and tests
> Rename fres test binaries to have more unique names
> Update fres_transaction API
> fcb: Add skeleton of transaction handling
* src/frsh-include
a2be805 ...
4057d7b (1):
> Add FRSH API for transactions