Re: Release DLL not working
- From: "Dean Ramsier" <ramsiernospam@xxxxxxxxxx>
- Date: Wed, 28 Dec 2005 08:59:39 -0500
I've not yet used this driver, so I can't give you specifics. However,
there are a few differences between debug and release mode drivers due to
differences in compiler optimizations that could give you a starting point.
- Debug versions are not optimized, and will therefore run a little slower.
In addition, some code sequences will be optimized away completely in a
retail build, so a missing volatile qualifier show up with these symptoms.
- I believe local variables on the stack get zeroed in a debug build, but
not in a retail build. This could hide a problem with using a variable
before it is initialized. There may be other differences, but these are the
big ones
The documentation for this controller mentions some minimum timing
requirements between certain types of register access. It's possible you're
on the edge here, although you would have hoped the driver from SMSC would
have handled the problem. You could try slowing down the chip select
speeds, see if that makes a difference.
--
Dean Ramsier - eMVP
BSQUARE Corporation
"gd.shukla" <gdshukla@xxxxxxxxxxxxxxxxxxxxxxxxx> wrote in message
news:C07BA1BE-9AFB-4BCD-BD07-42BF8D06F44D@xxxxxxxxxxxxxxxx
> Hi
> I am building SMSC9118 ethernet driver. The strange thing that is
> happening
> is that if I build my platform in debug mode, the LAN driver functions
> properly and answers to the PING done from another system. But if I build
> the
> same platform in release mode, the driver fails at the initial LOOPBACK
> test
> which is done to check the basixc functionality of the hardware. And if I
> check the status registers I get the correct values but somehow the chip
> fails to transmit even the loopback test data but the FIFO counter shows
> that
> the no. of bytes remaining are decrementing.
> The same driver works if I copy the debug mode DLL into the release folder
> and do the makeinage.
> Please suggest.
> --
> GD Shukla
.
- Prev by Date: Re: 20min system freeze
- Next by Date: Re: 20min system freeze
- Previous by thread: Re: byteswapping eboot.nbo
- Next by thread: Cloning Windows CE 3.0
- Index(es):
Relevant Pages
|
