This is very interesting indeed.

I won't claim to be an expert on the rt_mem_mgr stuff, I guess Pierre would
have to take care of this one. Maybe the RTL guys could help you out too
for the verification and provide ftp space for this. But I can't speak
for anyone. Just hope the pointers help out.

If nothing else, this gives a brief outline of the changes necessary to
port something between RTAI and RTL and vice-versa, which could be useful.

Best regards

Karim

Joel Vallier wrote:
> 
> I've done the porting to RTL 3.0 of the RTAI Dynamic Memory Manager
> (rt_mem_mgr) distributed with RTAI-1.5.
> 
> It works for me, but I'd like that some RTL experts verify that all is Ok.
> 
> Here is a list of main changes for the porting:
> ---------------------------------------------------------------------------------
> flags = rt_spin_lock_irqsave(&rt_mem_lock);                             // RTAI
> rtl_spin_lock_irqsave(&rtl_dmm_lock, flags);                            // RTL3.0
> ---------------------------------------------------------------------------------
> rt_spin_unlock_irqrestore(flags, &rt_mem_lock);                         // RTAI
> rtl_spin_unlock_irqrestore(&rtl_dmm_lock, flags);                       // RTL3.0
> ---------------------------------------------------------------------------------
> alloc_sysrq.srq = rt_request_srq(0, rt_alloc_sysrq_handler, 0)          // RTAI
> alloc_sysrq.srq = rtl_get_soft_irq (rtl_alloc_sysrq_handler, "rtl_dmm") // RTL3.0
> ---------------------------------------------------------------------------------
> rt_pend_linux_srq(alloc_sysrq.srq);                                     // RTAI
> rtl_global_pend_irq(alloc_sysrq.srq);                                   // RTL3.0
> ---------------------------------------------------------------------------------
> void rt_alloc_sysrq_handler(void)                                       // RTAI
> void rtl_alloc_sysrq_handler(int irq, void *dev_id, struct pt_regs *p)  // RTL3.0
> ---------------------------------------------------------------------------------
> kmalloc(granularity, GFP_KERNEL)                                        // RTAI
> kmalloc(granularity, GFP_ATOMIC)                                        // RTL3.0
> ---------------------------------------------------------------------------------
> 
> I can make source public if you give me a FTP server.
> 
> Regards,
> 
> Joel
> 
> Karim Yaghmour wrote:
> >
> > You can dynamically allocate memory in real-time using the rt_mem_mgr
> > stuff in RTAI.
> >
> > Cheers
> >
> > Karim
> >
> > [EMAIL PROTECTED] wrote:
> > >
> > > I remember seeing many discussions on the list about memory allocation,
> > > but i am failing to find them now.
> > >  I remember that you can only do memory allocation in initmodule. Is
> > > this right?  I would like to allocate a few
> > > kbytes either in a fifo handler or a real time thread.  Is this ok?
> > > eric
> > >
> >
> > ===================================================
> >                  Karim Yaghmour
> >                [EMAIL PROTECTED]
> >           Operating System Consultant
> >  (Linux kernel, real-time and distributed systems)
> > ===================================================

-- 
===================================================
                 Karim Yaghmour
               [EMAIL PROTECTED]
          Operating System Consultant
 (Linux kernel, real-time and distributed systems)
===================================================
-- [rtl] ---
To unsubscribe:
echo "unsubscribe rtl" | mail [EMAIL PROTECTED] OR
echo "unsubscribe rtl <Your_email>" | mail [EMAIL PROTECTED]
---
For more information on Real-Time Linux see:
http://www.rtlinux.org/rtlinux/

Reply via email to