| Commit message (Collapse) | Author | Files | Lines |
|
Updates #3053.
|
|
|
|
This function was only used on some m68k variants. On these m68k
variants there is no need to use a global symbol. Use a local label
instead.
Remove _ISR_Dispatch() from the architecture-independent layer.
|
|
Statically initialize the interrupt stack area
(_Configuration_Interrupt_stack_area_begin,
_Configuration_Interrupt_stack_area_end, and
_Configuration_Interrupt_stack_size) via <rtems/confdefs.h>. Place the
interrupt stack area in a special section ".rtemsstack.interrupt". Let
BSPs define the optimal placement of this section in their linker
command files (e.g. in a fast on-chip memory).
This change makes makes the CPU_HAS_SOFTWARE_INTERRUPT_STACK and
CPU_HAS_HARDWARE_INTERRUPT_STACK CPU port defines superfluous, since the
low level initialization code has all information available via global
symbols.
This change makes the CPU_ALLOCATE_INTERRUPT_STACK CPU port define
superfluous, since the interrupt stacks are allocated by confdefs.h for
all architectures. There is no need for BSP-specific linker command
file magic (except the section placement), see previous ARM linker
command file as a bad example.
Remove _CPU_Install_interrupt_stack(). Initialize the hardware
interrupt stack in _CPU_Initialize() if necessary (e.g.
m68k_install_interrupt_stack()).
The optional _CPU_Interrupt_stack_setup() is still useful to customize
the registration of the interrupt stack area in the per-CPU information.
The initialization stack can reuse the interrupt stack, since
* interrupts are disabled during the sequential system initialization,
and
* the boot_card() function does not return.
This stack resuse saves memory.
Changes per architecture:
arm:
* Mostly replace the linker symbol based configuration of stacks with
the standard <rtems/confdefs.h> configuration via
CONFIGURE_INTERRUPT_STACK_SIZE. The size of the FIQ, ABT and UND
mode stack is still defined via linker symbols. These modes are
rarely used in applications and the default values provided by the
BSP should be sufficient in most cases.
* Remove the bsp_processor_count linker symbol hack used for the SMP
support. This is possible since the interrupt stack area is now
allocated by the linker and not allocated from the heap. This makes
some configure.ac stuff obsolete. Remove the now superfluous BSP
variants altcycv_devkit_smp and realview_pbx_a9_qemu_smp.
bfin:
* Remove unused magic linker command file allocation of initialization
stack. Maybe a previous linker command file copy and paste problem?
In the start.S the initialization stack is set to a hard coded value.
lm32, m32c, mips, nios2, riscv, sh, v850:
* Remove magic linker command file allocation of initialization stack.
Reuse interrupt stack for initialization stack.
m68k:
* Remove magic linker command file allocation of initialization stack.
Reuse interrupt stack for initialization stack.
powerpc:
* Remove magic linker command file allocation of initialization stack.
Reuse interrupt stack for initialization stack.
* Used dedicated memory region (REGION_RTEMSSTACK) for the interrupt
stack on BSPs using the shared linkcmds.base (replacement for
REGION_RWEXTRA).
sparc:
* Remove the hard coded initialization stack. Use the interrupt stack
for the initialization stack on the boot processor. This saves
16KiB of RAM.
Update #3459.
|
|
|
|
|
|
Use a per-CPU thread dispatch disable level. So instead of one global
thread dispatch disable level we have now one instance per processor.
This is a major performance improvement for SMP. On non-SMP
configurations this may simplifiy the interrupt entry/exit code.
The giant lock is still present, but it is now decoupled from the thread
dispatching in _Thread_Dispatch(), _Thread_Handler(),
_Thread_Restart_self() and the interrupt entry/exit. Access to the
giant lock is now available via _Giant_Acquire() and _Giant_Release().
The giant lock is still implicitly acquired via
_Thread_Dispatch_decrement_disable_level().
The giant lock is only acquired for high-level operations in interrupt
handlers (e.g. release of a semaphore, sending of an event).
As a side-effect this change fixes the lost thread dispatch necessary
indication bug in _Thread_Dispatch().
A per-CPU thread dispatch disable level greatly simplifies the SMP
support for the interrupt entry/exit code since no spin locks have to be
acquired in this area. It is only necessary to get the current
processor index and use this to calculate the address of the own per-CPU
control. This reduces the interrupt latency considerably.
All elements for the interrupt entry/exit code are now part of the
Per_CPU_Control structure: thread dispatch disable level, ISR nest level
and thread dispatch necessary. Nothing else is required (except CPU
port specific stuff like on SPARC).
|
|
Script does what is expected and tries to do it as
smartly as possible.
+ remove occurrences of two blank comment lines
next to each other after Id string line removed.
+ remove entire comment blocks which only exited to
contain CVS Ids
+ If the processing left a blank line at the top of
a file, it was removed.
|
|
* cpu_asm.S: _Thread_Executing was not used.
|
|
PR 1599/cpukit
* cpu_asm.S: Rename _Context_Switch_necessary to
_Thread_Dispatch_necessary to more properly reflect the intent.
|
|
PR 1573/cpukit
* cpu_asm.S, rtems/score/cpu.h: Add a per cpu data structure which
contains the information required by RTEMS for each CPU core. This
encapsulates information such as thread executing, heir, idle and
dispatch needed.
|
|
* cpu_asm.S: Remove trailing tabs.
|
|
* cpu.c, cpu_asm.S: Add include of config.h
|
|
|
|
PR 1385/cpukit
* cpu_asm.S: When the type rtems_boolean was switched to the C99 bool,
the size changed from 4 bytes to 1 byte. The interrupt dispatching
code accesses two boolean variables for scheduling purposes and the
assembly implementations of this code did not get updated.
|
|
|
|
|
|
* cpu_asm.S: Collected PR 601 changes for commit to cvshead
for rtems-4.7
|
|
* cpu_asm.S: Remove use of C++ style comments and make this compile
again.
|
|
|
|
PR 598/bsps
* cpu_asm.S, rtems/score/cpu.h: Add save of floating point
status/control register on context switches. Missing this register
was causing intermittent floating point errors.
|
|
* Makefile.am: Install iregdefs.h and idtcpu.h to
$(includedir)/rtems/mips.
* cpu_asm.S: Include <rtems/mips/iregdef.h> instead of <iregdef.h>.
* rtems/score/mips.h, cpu_asm.S: Include <rtems/mips/idtcpu.h>
instead of <idtcpu.h>.
|
|
* cpu_asm.S: Include <rtems/asm.h> instead of <asm.h>.
|
|
* cpu.c, cpu_asm.S, rtems/score/cpu.h, rtems/score/mips.h,
rtems/score/types.h: URL for license changed.
|
|
* cpu_asm.S: Clarified some comments, removed code that forced
SR_IEP on when returning from an interrupt.
|
|
* cpu_asm.S: Added SR_IEO to context restore to fix isr disabled
deadlock caused by interrupt arriving while dispatching.
|
|
* cpu_asm.S: Now compiles on 4600 and 4650.
|
|
* cpu_asm.S: Fixed a sneaky return from int w/ ints disabled bug.
* rtems/score/cpu.h: Fixed register numbering in comments and made
interrupt enable/disable more robust.
|
|
* cpu_asm.S: Added support for the debug exception vector, cleaned
up the exception processing & exception return stuff. Re-added
EPC in the task context structure so the gdb stub will know where
a thread is executing. Should've left it there in the first place...
* idtcpu.h: Added support for the debug exception vector.
* cpu.c: Added ___exceptionTaskStack to hold a pointer to the
stack frame in an interrupt so context switch code can get the
userspace EPC when scheduling.
* rtems/score/cpu.h: Re-added EPC to the task context.
|
|
* cpu_asm.S: Fixed exception return address, modified FP context
switch so FPU is properly enabled and also doesn't screw up the
exception FP handling.
* idtcpu.h: Added C0_TAR, the MIPS target address register used for
returning from exceptions.
* iregdef.h: Added R_TAR to the stack frame so the target address
can be saved on a per-exception basis. The new entry is past the
end of the frame gdb cares about, so doesn't affect gdb or cpu.h
stuff.
* rtems/score/cpu.h: added an #ifdef so cpu_asm.S can include it
to obtain FPU defines without systax errors generated by the C
defintions.
* cpu.c: Improved interrupt level saves & restores.
|
|
* cpu_asm.S: Enhanced to save/restore more registers on
exceptions.
* rtems/score/cpu.h (CPU_Interrupt_frame): Enhanced to list every
register individually and document when it is saved.
* idtcpu.h: Added constants for the coprocessor 1 registers
revision and status.
|
|
* cpu.c: Enhancements and fixes for modifying the SR when changing
the interrupt level.
* cpu_asm.S: Fixed handling of FP enable bit so it is properly
managed on a per-task basis, improved handling of interrupt levels,
and made deferred FP contexts work on the MIPS.
* rtems/score/cpu.h: Modified to support above changes.
|
|
* cpu_asm.S: _CPU_Context_save_fp in was incorrectly in conditional
compilation block with (CPU_HARDWARE_FP == FALSE). Reported by
Wayne Bullaughey <wayne@wmi.com>.
|
|
* Assisted in design and debug by Joel Sherrill <joel@OARcorp.com>.
* cpu_asm.S: Now works on Mongoose-V. Missed in previous patch.
|
|
* cpu_asm.S: Merged patches from Gregory Menke
<Gregory.D.Menke.1@gsfc.nasa.gov> that clean up
stack usage and include nops in the delay slots.
|
|
* cpu_asm.S: Added code to save and restore SR and EPC to
properly support nested interrupts. Note that the ISR
(not RTEMS) enables interrupts allowing the nesting to occur.
|
|
* cpu.c, cpu_asm.S, iregdef.h, rtems/score/cpu.h, rtems/score/mips.h:
Merged MIPS1 and MIPS3 code reducing the number of lines of assembly.
Also reimplemented some assembly routines in C further reducing
the amount of assembly and increasing maintainability.
|
|
* cpu_asm.S: Use SR_INTERRUPT_ENABLE_BITS instead of SR_XXX constants
to make it easier to conditionalize the code for various ISA levels.
|
|
* rtems/score/cpu.h: Added _CPU_Initialize_vectors().
* cpu_asm.S: Eliminated warning for duplicate definition of EXTERN.
|
|
* cpu_asm.S (_ISR_Handler): Return to the address in the EPC register.
Previous code resulting in the interrupted immediately returning
to the caller of the routine it was inside.
|
|
* cpu.c: Removed duplicate declaration for _ISR_Vector_table.
* cpu_asm.S: Removed assembly language to vector ISR handler
on MIPS ISA I. Now call mips_vector_isr_handlers() in libcpu or BSP.
* rtems/score/cpu.h (CPU_INTERRUPT_NUMBER_OF_VECTORS): No
longer a constant -- get the real value from libcpu.
|
|
* cpu_asm.h: Removed.
* Makefile.am: Remove cpu_asm.h.
* rtems/score/mips64orion.h: Renamed mips.h.
* rtems/score/mips.h: New file, formerly mips64orion.h.
Header rewritten.
(mips_get_sr, mips_set_sr, mips_enable_in_interrupt_mask,
mips_disable_in_interrupt_mask): New macros.
* rtems/score/Makefile.am: Reflect renaming mips64orion.h.
* asm.h: Include <mips.h> not <mips64orion.h>. Now includes the
few defines that were in <cpu_asm.h>.
* cpu.c (_CPU_ISR_Get_level): Added MIPS ISA I version of this routine.
MIPS ISA 3 is still in assembly for now.
(_CPU_Thread_Idle_body): Rewrote in C.
* cpu_asm.S: Rewrote file header.
(FRAME,ENDFRAME) now in asm.h.
(_CPU_ISR_Get_level): Removed ISA I version and rewrote in C.
(_CPU_ISR_Set_level): Removed ISA I version and rewrote in C.
(_CPU_Context_switch): MIPS ISA I now manages preserves SR_IEC and
leaves other bits in SR alone on task switch.
(mips_enable_interrupts,mips_disable_interrupts,
mips_enable_global_interrupts,mips_disable_global_interrupts,
disable_int, enable_int): Removed.
(mips_get_sr): Rewritten as C macro.
(_CPU_Thread_Idle_body): Rewritten in C.
(init_exc_vecs): Rewritten in C as mips_install_isr_entries() and
placed in libcpu.
(exc_tlb_code, exc_xtlb_code, exc_cache_code, exc_norm_code): Moved
to libcpu/mips/shared/interrupts.
(general): Cleaned up comment blocks and #if 0 areas.
* idtcpu.h: Made ifdef report an error.
* iregdef.h: Removed warning.
* rtems/score/cpu.h (CPU_INTERRUPT_NUMBER_OF_VECTORS): Now a variable
number defined by libcpu.
(_CPU_ISR_Disable, _CPU_ISR_Enable): Rewritten to use new routines
to access SR.
(_CPU_ISR_Set_level): Rewritten as macro for ISA I.
(_CPU_Context_Initialize): Honor ISR level in task initialization.
(_CPU_Fatal_halt): Use new _CPU_ISR_Disable() macro.
|
|
* cpu_asm.S: Changed "_CPU_Ccontext_switch_restore: typo to
correct name of _CPU_Context_switch_restore. Added dummy
version of exc_utlb_code() so applications would link.
|
|
Joel Sherrill <joel@OARcorp.com>
* This is a major reworking of the mips64orion port to use
gcc predefines as much as possible and a big push to multilib
the mips port. The mips64orion port was copied/renamed to mips
to be more like other GNU tools. Alan did most of the technical
work of determining how to map old macro names used by the mips64orion
port to standard compiler macro definitions. Joel did the merge
with CVS magic to keep individual file history and did the BSP
modifications. Details follow:
* Makefile.am: idtmon.h in mips64orion port not present.
* asm.h: MIPS64ORION replaced with MIPS. Frame setup macros added.
* cpu.c: Comments added.
* cpu_asm.S: Conditionals changed. MIPS ISA level 1 support added.
First attempt at exception/interrupt processing for ISA level 1
and minus any use of IDT/MON added.
* idtcpu.h: Conditionals changed to use gcc predefines.
* iregdef.h: Ditto.
* cpu_asm.h: No real change. Merger required commit.
* rtems/Makefile.am: Ditto.
* rtems/score/Makefile.am: Ditto.
* rtems/score/cpu.h: Change MIPS64ORION to MIPS.
* rtems/score/mips64orion.h: Change MIPS64ORION to MIPS. Convert
from using RTEMS_CPU_MODEL to gcc predefines to figre things out.
|
|
|
|
|
|
I found a small buglet in the mips64orion _CPU_ISR_Set_level; the
original was wiping out the level argument, and then comparing the
current interrupt level with some random value of v0. See patch below.
|
|
|
|
|
|
of switching to the modified GNU GPL.
|