Hello all I have returned from vacations and managed actual RTEMS git (4.11-pre) to run a simple template
http://rtime.felk.cvut.cz/gitweb/rtems-devel.git/tree/:/rtems-omk-template and our full medical instrument application to run on PiMX1 i.MX1 board (compatible to RTEMS i.MXL csb336). There are changes bsp/csb336: Fix MMU _ttbl_base location to not overlap with exception vectors. This is a fix of MMU table overwrite by exception vectors. The patch is required to make any csb336 based board to run. bsp/csb336: Memory map update and jump to start at image start provided. More optimized cache setup and cover of i.MX1 eSRAM region. bsp/csb336: implement bsp_interrupt_vector_enable/disable. Real common implementation of AITC vector enable/disable. I am not 100% sure about this patch. I am sure that move of enable/disable to the common place is correct but there are spread some copies (replacement of the central support) of this functionality in other files rtems/c/src/lib/libbsp/arm/csb336/console/uart.c: imx_uart_isr_on() imx_uart_isr_off() rtems/c/src/lib/libbsp/arm/csb336/network/network.c enet_isr_on(void) Partially overlaps with rtems/c/src/lib/libcpu/arm/mc9328mxl/clock/clockdrv.c clock_isr_on() clock_isr_off() clock_isr_is_on() The local functions are used by rtems_irq_connect_data xxx_isr_data and BSP_install_rtems_irq_handler(&clock_isr_data) I think that these structures are not necessary for most cases and generic vectors enable/disable manipulation functions can be/should be used. But I am not 100% about future intention with connect_data use/necessity in future RTEMS versions. Even if this patch is not necessary it does not seem to cause troubles for our quite complex environment. c/src/lib/libbsp/arm/csb336/start/start.S | 7 +++++++ c/src/lib/libbsp/arm/csb336/startup/linkcmds | 7 ++++--- c/src/lib/libbsp/arm/csb336/startup/memmap.c | 10 +++++++--- c/src/lib/libcpu/arm/mc9328mxl/irq/irq.c | 10 ++++++++++ 4 files changed, 28 insertions(+), 6 deletions(-) -- 1.7.10.4 _______________________________________________ rtems-devel mailing list [email protected] http://www.rtems.org/mailman/listinfo/rtems-devel
