arm shared code contains not shared code?

Hello, I am learning the rtems source, and when i look into this folder: rtems@book64:~/rtems-lite/rtems/c/src/lib/libcpu/arm/shared$ tree . ├── arm920 │   └── mmu.c └── include ├── am335x.h ├── arm-cp15.h ├── cache_.h ├── mmu.h ├── omap3.h └──

Re: [PATCH 00/13] Replace mongoose with civetweb.

Yes that's right. 05/13 just adds the unchanged sources from civetweb. Beneath civetweb.c and civetweb.h it also adds handle_form.inl and md5.inl. The last two files are included into civetweb.c. According to the documentation "The *INL* file extension represents code that is statically included

Re: the rtems kernel source list

On 22/04/2016 12:19, printk wrote: Hello, I am learning RTEMS, where is the kernel source, I find some place seems holding the kernel source: rtems/c/src/support/ - only version.c there. rtems/c/src/ada/ - is this kernel source? No. rtems/cpukit/score/src

the rtems kernel source list

Hello, I am learning RTEMS, where is the kernel source, I find some place seems holding the kernel source:

Re: [PATCH 00/13] Replace mongoose with civetweb.

On Thu, Apr 21, 2016 at 7:36 PM, Chris Johns wrote: > On 21/04/2016 18:49, Christian Mauderer wrote: > >> This patch series replaces the mongoose webserver by its still MIT >> licensed fork civetweb. >> >> Please note that I try to get some (currently two) of the patches >>

Re: [PATCH 00/13] Replace mongoose with civetweb.

On 21/04/2016 18:49, Christian Mauderer wrote: This patch series replaces the mongoose webserver by its still MIT licensed fork civetweb. Please note that I try to get some (currently two) of the patches directly into civetweb too. But I think that it might need some time and adaption till they

Managing RTEMS time in a paravirtualized environment

Hi Gedare and I are working on a paravirtualized RTEMS executing in an ARINC 653 partition. One challenge is ensuring the time passage in RTEMS is correct. The problem in a nutshell is that RTEMS is not executing the entire time. Each partition is an independent RTEMS instance. There will be no

Re: RKI Build Error - rtems_clock_get not available on 4.12

I should be able to update my RKI repository soon.. I’m building the latest tools and RTEMS repo now. Alan > On Apr 21, 2016, at 6:14 AM, Pavel Pisa wrote: > > Hello Habeeb, > > rtems_clock_get() has been obsoleted long time ago > and has been removed in > >

Re: RKI Build Error - rtems_clock_get not available on 4.12

Hello Habeeb, rtems_clock_get() has been obsoleted long time ago and has been removed in e65c45c4b6cf6dfb485bef48385e39969de8b361 Obsolete rtems_clock_get() directive. This service was marked as deprecated long prior to the 4.11 release series and is now being removed.

Provide POSIX network header for all normal Newlib targets?

Hello, with the recent type changes for etc. we are now able to build the FreeBSD 9.3 network stack on RTEMS. We directly use * , * , * , * , * , * , and * from Newlib with mostly unmodified FreeBSD header and source files. Cygwin and Linux provide currently their own set of

[PATCH 08/13] mghttpd: Use MD5 library.

From: Christian Mauderer This applies the changes from ea008e2d0d78b49e0f793949ab4a665b69283372. --- cpukit/mghttpd/civetweb.c | 5 + 1 file changed, 5 insertions(+) diff --git a/cpukit/mghttpd/civetweb.c b/cpukit/mghttpd/civetweb.c index

[PATCH 12/13] mghttpd: Use gethostbyname instead of getaddrinfo.

From: Christian Mauderer RTEMS does not have a getaddrinfo() so use gethostbyname() instead. This reintroduces a not thread save function into the code but it seems that we have no better replacement at the moment. --- cpukit/mghttpd/civetweb.c | 28

[PATCH 10/13] mghttpd: Don't use __sync_XXX_and_fetch functions.

From: Christian Mauderer This patch should be replaced by one that uses libatomic as soon as it is provided by our toolchain. --- cpukit/mghttpd/civetweb.c | 8 ++-- 1 file changed, 6 insertions(+), 2 deletions(-) diff --git

[PATCH 02/13] libnetworking: Add minimal getnameinfo().

From: Christian Mauderer This implementation just falls back to giving a string representation of the IP. It supports IPv4 only. --- cpukit/libnetworking/Makefile.am| 2 +- cpukit/libnetworking/libc/getnameinfo.c | 61

[PATCH 01/13] libnetworking: Add prototype for getnameinfo().

From: Christian Mauderer Copy the prototype and necessary defines for getnameinfo() from current FreeBSD. --- cpukit/libnetworking/netdb.h | 47 1 file changed, 47 insertions(+) diff --git

[PATCH 04/13] libcsupport: Add dummy for setgroups().

From: Christian Mauderer The dummy for setgroups() allows applications using it to build (for example civetweb webserver). --- cpukit/libcsupport/Makefile.am | 2 +- cpukit/libcsupport/src/setgroups.c | 26

[PATCH 07/13] mghttpd: Use HAVE_CONFIG_H.

From: Christian Mauderer This applies the changes from 4248b28296af7a56c3b1fb14c53debd8d79ed6d9. --- cpukit/mghttpd/civetweb.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/cpukit/mghttpd/civetweb.c b/cpukit/mghttpd/civetweb.c index

[PATCH 06/13] mghttpd: Use poll() replacement.

From: Christian Mauderer This patch provides a poll() replacement based on the poll() used for civetweb on windows. A similar patch based on the one on http://forums.bannister.org/ubbthreads.php?ubb=showflat=7600=1 had been integrated into the RTEMS

[PATCH 09/13] mghttpd: Add stack size and scheduling options.

From: Christian Mauderer This applies the changes from 54da7c3e55056efd93adece52076b720490258d6. --- cpukit/mghttpd/civetweb.c | 117 +- 1 file changed, 96 insertions(+), 21 deletions(-) diff --git

[PATCH 00/13] Replace mongoose with civetweb.

This patch series replaces the mongoose webserver by its still MIT licensed fork civetweb. Please note that I try to get some (currently two) of the patches directly into civetweb too. But I think that it might need some time and adaption till they are accepted. So I thought that adding them to

[PATCH 03/13] testsuite: Add test for getnameinfo().

From: Christian Mauderer --- testsuites/libtests/Makefile.am| 1 + testsuites/libtests/configure.ac | 1 + testsuites/libtests/libnetworking01/Makefile.am| 22 +++ testsuites/libtests/libnetworking01/init.c

[PATCH 11/13] mghttpd: Allow to have no timegm().

From: Christian Mauderer We currently don't have a timegm function. There seems to be no easy way to replace this function with POSIX functions in a thread save way. This patch just avoids the call. Note that with this solution, the caching feature is

[PATCH 01/10] posix: Avoid Giant lock in sem_getvalue()

Update #2555. --- cpukit/posix/src/semgetvalue.c | 13 +++-- 1 file changed, 11 insertions(+), 2 deletions(-) diff --git a/cpukit/posix/src/semgetvalue.c b/cpukit/posix/src/semgetvalue.c index ed49c09..f895266 100644 --- a/cpukit/posix/src/semgetvalue.c +++

[PATCH 08/10] score: Avoid Giant lock for barriers

Use _Thread_queue_Flush_critical() to atomically release the barrier. Update #2555. --- cpukit/posix/include/rtems/posix/barrierimpl.h | 24 ++--- cpukit/posix/src/pbarrierdestroy.c | 45 -- cpukit/posix/src/pbarrierwait.c| 50 +--

[PATCH 05/10] score: Use _Thread_queue_Flush_critical for futex

--- cpukit/score/src/futex.c | 98 ++-- 1 file changed, 28 insertions(+), 70 deletions(-) diff --git a/cpukit/score/src/futex.c b/cpukit/score/src/futex.c index 73d48fd..59a625c 100644 --- a/cpukit/score/src/futex.c +++ b/cpukit/score/src/futex.c @@

[PATCH 06/10] score: Use _Thread_queue_Flush_critical() for cond

--- cpukit/score/src/condition.c | 108 --- 1 file changed, 41 insertions(+), 67 deletions(-) diff --git a/cpukit/score/src/condition.c b/cpukit/score/src/condition.c index da3d133..ba255fc 100644 --- a/cpukit/score/src/condition.c +++

[PATCH 04/10] score: Add _Thread_queue_Is_empty()

--- cpukit/score/include/rtems/score/threadqimpl.h | 7 +++ testsuites/sptests/spthreadq01/init.c | 2 +- 2 files changed, 8 insertions(+), 1 deletion(-) diff --git a/cpukit/score/include/rtems/score/threadqimpl.h b/cpukit/score/include/rtems/score/threadqimpl.h index

[PATCH 10/10] rtems: Avoid Giant lock for dual ported memory

There is no need for an ISR lock since the Dual_ported_memory_Control is immutable after initialization. ISR disable is enough for deletion safety on uni-processor configurations. Update #2555. --- cpukit/rtems/include/rtems/rtems/dpmemimpl.h | 18 +++ cpukit/rtems/src/dpmemdelete.c

[PATCH 03/10] posix: Use _Objects_Get_local() for semaphores

This simplifies the code since the object location is no longer used. Remove superfluous header includes. --- cpukit/posix/include/rtems/posix/semaphoreimpl.h | 11 ++--- cpukit/posix/src/semaphorewaitsupp.c | 60 +--- cpukit/posix/src/semclose.c

[PATCH v3 2/3] Always provide register_t via

Always provide register_t via for glibc and BSD compatibility. Define __BIT_TYPES_DEFINED__ to 1 like glibc for legacy header files. Signed-off-by: Sebastian Huber --- newlib/libc/include/sys/types.h | 6 ++ winsup/cygwin/include/machine/types.h |

[PATCH v3 3/3] Provide FreeBSD types for on RTEMS

Provide the following types via on RTEMS for FreeBSD compatibility if __BSD_VISIBLE * accmode_t, * cap_rights_t, * c_caddr_t, * cpulevel_t, * fixpt_t, * lwpid_t, * uintfptr_t, * vm_offset_t, * vm_ooffset_t, * vm_paddr_t, * vm_pindex_t, and * vm_size_t. Signed-off-by: Sebastian Huber

[PATCH v3 1/3] Resurrect for

Resurrect for use in . Newlib targets may provide an own version of in their machine directory to add custom user types for . Check the _SYS_TYPES_H header guard to prevent a direct include of , since the file is a Newlib speciality. Signed-off-by: Sebastian Huber