Re: [rtems commit] or1k: Implement context validate and context volatile clobber functions.

2014-09-02 Thread Chris Johns
On 3/09/2014 12:21 am, Joel Sherrill wrote: cpukit/score/cpu/or1k/preinstall.am|6 +- Did this patch happen before my fix to the preinstall went in ? I am seeing an issue with this one. Chris ___ devel mailing list devel@rtems.o

Re: compile rtems source code error

2014-09-02 Thread Chris Johns
On 3/09/2014 12:05 pm, 贾超 wrote: tools: sparc-rtems4.11. host: x86_64ubuntu. When I compile the source code in the sparc-rtems4.11 target, always appear the following error. ../../../sis/lib/include/rtems/rtems_bsdnet_internal.h:54:7: note: expected 'void *' but argument is of type 'volatile st

compile rtems source code error

2014-09-02 Thread 贾超
tools: sparc-rtems4.11. host: x86_64ubuntu. When Icompile thesource codein thesparc-rtems4.11target,always appearthe followingerror. ../../../sis/lib/include/rtems/rtems_bsdnet_internal.h:54:7: note: expected 'void *' but argument is of type 'volatile struct dwmac_desc_ext_erx *' ../../../../..

tms570 and networking

2014-09-02 Thread Joel Sherrill
Hi The tms570 has what appears to be a partial set of Makefile.am entries for a network driver but none is present so the BSP fails to build when networking is enabled. It appears to be missing the line which would specify a source file anyway. I have commented out the section locally. Should I d

BSPs which define NDEBUG in CFLAGS

2014-09-02 Thread Joel Sherrill
Hi The following BSPs define NDEBUG in their CFLAGS and thus fail to compile when RTEMS debug is enabled: mvme136 lpc1768_mbed_ahb_ram lpc1768_mbed rtl22xx I have tracked this down to newlib's assert.h (correctly) not defining __ASSERT_FUNC when NDEBUG is defined. This trips up the RTEMS assert

Warnings on Head

2014-09-02 Thread Joel Sherrill
Hi Doing some greps and mangling on the build logs from every BSP, I discovered we have ~2350 unique warnings output by gcc. - ~1250 of these only occur once. - ~1900 occur less than 5 times. - ~2270 occur less than 10 times. These are almost certainly in BSP specific code built for variants. My

Re: CMSIS?

2014-09-02 Thread Joel Sherrill
On 9/2/2014 9:52 AM, Daniel Gutson wrote: > On Tue, Sep 2, 2014 at 11:51 AM, Joel Sherrill > wrote: >> This is the license. >> >> The big problem I see first is the typical use only on our hardware >> and that in order to redistribute, we must ensure that it is only >> used on ARM hardware. >> >>

Re: CMSIS?

2014-09-02 Thread Daniel Gutson
On Tue, Sep 2, 2014 at 11:51 AM, Joel Sherrill wrote: > This is the license. > > The big problem I see first is the typical use only on our hardware > and that in order to redistribute, we must ensure that it is only > used on ARM hardware. > > This is a common restriction on vendor provided softw

Re: CMSIS?

2014-09-02 Thread Joel Sherrill
This is the license. The big problem I see first is the typical use only on our hardware and that in order to redistribute, we must ensure that it is only used on ARM hardware. This is a common restriction on vendor provided software and I honestly don't know how to address including anything whi

Re: [PATCH 2/3] or1ksim: Add tm27.h to include_HEADERS

2014-09-02 Thread Hesham Moustafa
On Tue, Sep 2, 2014 at 4:31 PM, Joel Sherrill wrote: > I think this patch installs tm27.h is redundant with one I already > committed. > > tm27 has "#include " and this patch would have installed > a second copy under the bsp/ subdirectory. That may in fact be the > correct location but it isn't w

Re: [PATCH 2/3] or1ksim: Add tm27.h to include_HEADERS

2014-09-02 Thread Joel Sherrill
I think this patch installs tm27.h is redundant with one I already committed. tm27 has "#include " and this patch would have installed a second copy under the bsp/ subdirectory. That may in fact be the correct location but it isn't where it is expected to be. It would have to be moved for all BSP

Re: [PATCH 1/3] Add missing r31 load instruction _ISR_Handler

2014-09-02 Thread Joel Sherrill
Committed. But no build. --joel On 8/31/2014 6:28 PM, Hesham ALMatary wrote: > _ISR_Handler: r31 was not loaded in restore function. This patch > adds this load. > --- > cpukit/score/cpu/or1k/or1k-exception-handler-low.S | 3 ++- > 1 file changed, 2 insertions(+), 1 deletion(-) > > diff --git a/c

Re: [PATCH 3/3] or1k: Implement context validate and context volatile clobber functions.

2014-09-02 Thread Joel Sherrill
Committed but no build. Thanks. On 8/31/2014 6:29 PM, Hesham ALMatary wrote: > score/cpu/or1k: Add two new assembly functions: _CPU_Context_validate > and _CPU_Context_volatile_clobber; their implementation follows > corresponding ARM functions. > --- > cpukit/score/cpu/or1k/Makefile.am

Re: [PATCH] Adding "eth" linker script section for arm bsp.

2014-09-02 Thread Federico Casares
Hi Sebastian, Are you suggesting to remove the "lpc176x/make/custom/lpc1768_mbed_ahb_ram_eth-testsuite.tcfg" and "lpc176x/make/custom/lpc1768_mbed_ahb_ram_eth.cfg" files? The tcfg includes in its list two tests that should not be compiled for this BSP configuration, due to BBS size required by t

Re: [PATCH] Adding "eth" linker script section for arm bsp.

2014-09-02 Thread Sebastian Huber
Hello Federico, since this just adds a linker command file variant, I would prefer to just add the linker command file without the BSP name. You can then use "-qnolinkcmds -T linkcmds.lpc1768_mbed_ahb_ram_eth" to select this memory map. On 01/09/14 15:21, Federico Casares wrote: Attached th