On 2 February 2016 at 10:53, Sebastian Huber <sebastian.hu...@embedded-brains.de> wrote: > On 01/02/16 14:39, Darshit Shah wrote: >>> >>> >Maybe we can implement >>> > >>> >Improve the SMP scheduler with arbitrary processor affinity support >>> >https://devel.rtems.org/ticket/2510 >>> > >>> >as part of this years GSoC. >>> > >> >> I've already mentioned that I'd be interested in picking this up as a >> potential project. However, the entire implementation may not be >> possible in a single summer. It is rather complex, and while a naive >> implementation may be possible to write, we would want something more >> efficient. I'm hoping for some pointers on where I should start >> looking for a better understanding of the codebase. > > > In order to get familiar with the scheduler, do the following. > > 0. Use a Linux system for development (use a Linux VM on Windows) > Check. Wouldn't survive without a Linux system
> 1. Install the RTEMS 4.12 ARM tool chain via the RSB > Check > 2. Build the realview_pbx_a9_qemu_smp BSP with > > configure --target=arm-rtems4.12 --prefix=/opt/rtems-4.12 > --enable-rtemsbsp=realview_pbx_a9_qemu_smp --enable-maintainer-mode > --enable-tests --enable-smp > Missed actually invoking "make" here. But that was rather obvious. As mentioned in a bug report I found on trac, `pax` despite being POSIX is not always available. On Arch Linux, it's not even in the official repositories. Maybe, we should add a note somewhere about this requirement? My make process failed the first time because of this missing dependency. > 3. Install a recent Qemu > > 4. Open two shell windows and go the the following subdirectory in the build > tree > > arm-rtems4.12/c/realview_pbx_a9_qemu_smp/testsuites/smptests/smp01 > > 5. In one start Qemu via > > qemu-system-arm -net none -nographic -M realview-pbx-a9 -m 256M -smp 2 -s -S > -kernel *.exe > > 6. In the other start GDB via > > arm-rtems4.12-gdb *.exe > Check, check, check and check Things do work! Thanks! > 7. Example GDB session, interesting functions are _Thread_Set_state() and > _Thread_Clear_state(). Just follow them in the debugger to figure out how > the scheduler works. > > GNU gdb (GDB) 7.9 > Copyright (C) 2015 Free Software Foundation, Inc. > License GPLv3+: GNU GPL version 3 or later > <http://gnu.org/licenses/gpl.html> > This is free software: you are free to change and redistribute it. > There is NO WARRANTY, to the extent permitted by law. Type "show copying" > and "show warranty" for details. > This GDB was configured as "--host=x86_64-linux-gnu --target=arm-rtems4.12". > Type "show configuration" for configuration details. > For bug reporting instructions, please see: > <http://www.gnu.org/software/gdb/bugs/>. > Find the GDB manual and other documentation resources online at: > <http://www.gnu.org/software/gdb/documentation/>. > For help, type "help". > Type "apropos word" to search for commands related to "word"... > Reading symbols from smpload01.exe...done. > (gdb) tar remote :1234 > Remote debugging using :1234 > bsp_start_vector_table_end () at > rtems-4.12/c/src/lib/libbsp/arm/realview-pbx-a9/../shared/start/start.S:142 > 142 mrc p15, 0, r0, c0, c0, 5 > (gdb) monitor system_reset > (gdb) load > Loading section .start, size 0x800 lma 0x100000 > Loading section .text, size 0x278e4 lma 0x100800 > Loading section .init, size 0xc lma 0x1280e4 > Loading section .fini, size 0xc lma 0x1280f0 > Loading section .rodata, size 0x1121 lma 0x1280fc > Loading section .ARM.exidx, size 0x8 lma 0x129220 > Loading section .eh_frame, size 0x4 lma 0x129228 > Loading section .init_array, size 0x4 lma 0x12922c > Loading section .fini_array, size 0x4 lma 0x129230 > Loading section .jcr, size 0x4 lma 0x129234 > Loading section .rtemsroset, size 0x20 lma 0x129238 > Loading section .data, size 0x97c lma 0x20a000 > Start address 0x100040, load size 170961 > Transfer rate: 18550 KB/sec, 1799 bytes/write. > (gdb) b _Thread_Set_state > Breakpoint 1 at 0x11bba6: file > rtems-4.12/c/src/../../cpukit/score/src/threadsetstate.c, line 39. > (gdb) b _Thread_Clear_state > Breakpoint 2 at 0x118092: file > rtems-4.12/c/src/../../cpukit/score/src/threadclearstate.c, line 35. > (gdb) c > Continuing. > > Breakpoint 2, _Thread_Clear_state (the_thread=0x21a3e0, state=4294967295) at > rtems-4.12/c/src/../../cpukit/score/src/threadclearstate.c:35 > 35 _Scheduler_Acquire( the_thread, &lock_context ); > (gdb) bt > #0 _Thread_Clear_state (the_thread=0x21a3e0, state=4294967295) at > rtems-4.12/c/src/../../cpukit/score/src/threadclearstate.c:35 > #1 0x0011bde0 in _Thread_Ready (the_thread=0x21a3e0) at > ../../cpukit/../../../realview_pbx_a9_qemu_smp/lib/include/rtems/score/threadimpl.h:277 > #2 0x0011be38 in _Thread_Start (the_thread=0x21a3e0, entry=0x200f64) at > rtems-4.12/c/src/../../cpukit/score/src/threadstart.c:34 > #3 0x0010f9d6 in rtems_task_start (id=167837697, entry_point=0x10f63f > <_RTEMS_Global_construction>, argument=2150284) at > rtems-4.12/c/src/../../cpukit/rtems/src/taskstart.c:67 > #4 0x0010f782 in _RTEMS_tasks_Initialize_user_tasks_body () at > rtems-4.12/c/src/../../cpukit/rtems/src/taskinitusers.c:110 > #5 0x00120252 in _RTEMS_tasks_Initialize_user_tasks () at > rtems-4.12/c/src/../../cpukit/rtems/src/tasks.c:229 > #6 0x00110728 in _API_extensions_Run_postdriver () at > rtems-4.12/c/src/../../cpukit/score/src/apiext.c:47 > #7 0x001104ee in rtems_initialize_device_drivers () at > rtems-4.12/c/src/../../cpukit/sapi/src/exinit.c:246 > #8 0x00110514 in rtems_initialize_executive () at > rtems-4.12/c/src/../../cpukit/sapi/src/exinit.c:276 > #9 0x00101cfa in boot_card (cmdline=0x0) at > rtems-4.12/c/src/lib/libbsp/arm/realview-pbx-a9/../../shared/bootcard.c:90 > #10 0x00100112 in bsp_vector_table_copy_done () at > rtems-4.12/c/src/lib/libbsp/arm/realview-pbx-a9/../shared/start/start.S:279 > Backtrace stopped: previous frame identical to this frame (corrupt stack?) > (gdb) del 2 > (gdb) c > Continuing. > [Switching to Thread 2] > > Breakpoint 1, _Thread_Set_state (the_thread=0x21a3e0, state=262144) at > rtems-4.12/c/src/../../cpukit/score/src/threadsetstate.c:39 > 39 _Scheduler_Acquire( the_thread, &lock_context ); > (gdb) bt > #0 _Thread_Set_state (the_thread=0x21a3e0, state=262144) at > rtems-4.12/c/src/../../cpukit/score/src/threadsetstate.c:39 > #1 0x0010c43e in _Event_Seize (event_in=2147483648, option_set=0, ticks=0, > event_out=0x24f96c, executing=0x21a3e0, event=0x21a5f8, wait_class=512, > block_state=262144, lock_context=0x24f944) > at rtems-4.12/c/src/../../cpukit/rtems/src/eventseize.c:101 > #2 0x0010f080 in rtems_event_system_receive (event_in=2147483648, > option_set=0, ticks=0, event_out=0x24f96c) at > rtems-4.12/c/src/../../cpukit/rtems/src/systemeventreceive.c:52 > #3 0x00100b58 in rtems_event_transient_receive (option_set=0, ticks=0) at > ../../../../../realview_pbx_a9_qemu_smp/lib/include/rtems/rtems/event.h:484 > #4 0x00101568 in test () at > rtems-4.12/c/src/../../testsuites/smptests/smpload01/init.c:333 > #5 0x0010177a in Init (arg=2150284) at > rtems-4.12/c/src/../../testsuites/smptests/smpload01/init.c:384 > #6 0x00118efa in _Thread_Entry_adaptor_numeric (executing=0x21a3e0) at > rtems-4.12/c/src/../../cpukit/score/src/threadentryadaptornumeric.c:25 > #7 0x00121524 in _Thread_Handler () at > rtems-4.12/c/src/../../cpukit/score/src/threadhandler.c:93 > #8 0x001214ce in _User_extensions_Thread_exitted (executing=0x1214cf > <_Thread_Handler>) at > ../../cpukit/../../../realview_pbx_a9_qemu_smp/lib/include/rtems/score/userextimpl.h:244 > #9 0x00001008 in ?? () > Backtrace stopped: previous frame identical to this frame (corrupt stack?) > (gdb) > > If you are not familiar with the command line GDB, then I recommend to do a > GDB tutorial first. > Now stepping through the scheduler. Will spend some time looking into this and the source directly to see what I am able to gather. And as Gedare mentioned, I should probably write up a bit about it. Though given the excellent state of GettingStarted pages already available, I'm not sure what new material I can contribute. > > -- > Sebastian Huber, embedded brains GmbH > > Address : Dornierstr. 4, D-82178 Puchheim, Germany > Phone : +49 89 189 47 41-16 > Fax : +49 89 189 47 41-09 > E-Mail : sebastian.hu...@embedded-brains.de > PGP : Public key available on request. > > Diese Nachricht ist keine geschäftliche Mitteilung im Sinne des EHUG. > -- Thanking You, Darshit Shah _______________________________________________ devel mailing list devel@rtems.org http://lists.rtems.org/mailman/listinfo/devel
