NEC ELECTRONICS GLOBAL
nec electronics global
HOME
APPLICATIONS
PRODUCTS
TECHNOLOGY
SUPPORT
BUY ONLINE
NEWS & EVENTS
ABOUT US
header
GO
AdvancedParametric
SITE MAP CONTACT US

RX4000

Contents

    
FAQ-ID = VRrx4000-nnnn
0001: System calls
0002: UserPool
0003: FPU instruction
VRrx4000
-0001
System calls
Q1
When the RX4000 outputs the get_blk system call, and then outputs rel_blk after a memory write operation from the start of the acquisition memory block, the error message "Memory block is used as a message" is output.
Must the first four bytes be cleared to zero before use, such as when snd_msg is output?
(Although this is not specifically mentioned in the User's Manual, normal status is restored by executing rel_blk after zero-clearing the first four bytes.)
A1
Yes, the first four bytes must be cleared to zero.

This is because this system is designed so that rel_blk cannot be executed when snd_msg has been executed and messages are queued in the mailbox.
At this time, the RX4000 performs judgments based on whether or not the first four bytes = 0.
When they equal zero, queuing is not recognized. Since this judgment is always performed, the first four bytes of that memory block must be cleared to zero when rel_blk is executed.
Is this information useful for you ?
back to top  
(2006/03)

VRrx4000
-0002
UserPool
Q1
I have a question about the system memory's UserPool.
When dividing the UserPool itself for each execution task, can I generate multiple UserPools?
Specifically, I would like to create an independent UserPool for the cacheable area and uncacheable area, and wonder if an operation can determine which UserPool to use for each task.
A1
Unfortunately, you cannot create multiple SystemPools, UserPools, or StackPools.
Consequently, use of the UserPool cannot be divided between the cacheable area and uncacheable area.
For now, the best method would be to place the UserPool in the cacheable area.
Is this information useful for you ?
back to top  
(2006/03)

VRrx4000
-0003
FPU instruction
Q1
When initializing the VR4310, use of co-processors CP0 and CP1 is enabled and the kernel is initialized.
However, in the main routine following kernel initialization, CP1 use becomes disabled and a CP1 use-disabled exception occurs when the FPU instruction is executed.
I tried to enable CP0 and CP1 in other locations, but it seems these co-processors become use-disabled when a system call is executed.
As a temporary workaround, I disable the interrupts immediately before the FPU, and then I enable CP0 and CP1.
Is there a better method (such as setting information to the kernel)?
  Environment
        CPU:  V4310
        Kernel:  RX4000
        Compiler:  Green Hills
        Debugger:  Partner ET2
A1
If your RX4000 version is Ver. 3.10, the FPU instruction cannot be used.
To upgrade to RX4000 Ver. 3.11, contact the NEC Electronics sales representative where you purchased your RX4000.
Is this information useful for you ?
back to top  
(2006/03)









































 LEGAL  RSS Feeds       © 1995-2008  NEC Electronics Corporation