This is the mail archive of the ecos-discuss@sources.redhat.com mailing list for the eCos project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

RE: HAL startup help.


Nick

	Thanks for your help.. That was exactly it.  I don't know how I got the memory map/alignment from.
	But after reviewing it against some of the other PPC targets, I have found some "issues".

	Once this was corrected... Life is Beautiful

	roman


-----Original Message-----
From: nickg@xl5.calivar.com [mailto:nickg@xl5.calivar.com] On Behalf Of Nick Garnett
Sent: Monday, August 02, 2004 5:07 AM
To: Rycerz Roman-G11565
Cc: Ecos-Discuss@Sources. Com (ecos-discuss@sources.redhat.com)
Subject: Re: [ECOS] HAL startup help.


Rycerz Roman-G11565 <roman.rycerz@motorola.com> writes:

>     I am hoping someone can help me.   I have a rather strange problem that comes and goes for reasons that so far escape me.
>     While I am not sure what makes this situation appear and 
> disappear, maybe this behavior will ring some bells.
>  
>  
>     When the ecos world is happy, the routine "cyg_hal_invoke_constructors" executes, behaves nicely and executes a list of constructors.
>     such as
>                 Global_I10100_diag_write_char
>                 Global_I11000_cyg_scheduler_sched_lock
>                 Global_I12000_ZB13cyg_Interrupt15....
>  
>  
>     When ecos is unhappy "cyg_hal_invoke_constructors" actually 
> executes a list of destructors
>  
>                 Global_D14000......
>                 Global_D15000.......
>  
>  
>    Anyone know what might cause this ????   I am sure its something I am doing, but so far I have no clue....

Cannot help you specifically, but in the past when I have seen things like this happening, it has often been an alignment issue. It appears and disappears as changing sizes of text and data change the alignment of later data items.

Check that the alignment requirements in your linker script are correct. If you try single stepping through
cyg_hal_invoke_constructors() when it is misbehaving you should be able to work out what is going wrong.

-- 
Nick Garnett                    eCos Kernel Architect
http://www.ecoscentric.com/     The eCos and RedBoot experts

-- 
Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos
and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]