Re: memstick: Alternative approach to proposed fixes

[Maxim Levitsky <maximlevitsky@gmail.com>]

On Wed, 2010-10-27 at 01:07 +0200, Maxim Levitsky wrote:
> On Wed, 2010-10-27 at 00:55 +0200, Maxim Levitsky wrote:
> > On Wed, 2010-10-27 at 04:41 +1100, Alex Dubov wrote:
> > > As I was not able to convince myself that profound changes proposed by Maxim
> > > are really necessary, I propose to follow a much milder path with this update
> > > round (as I outlined in my previous emails).
> > It is really sad to see that patchset.
> > 
> > 
> > > 
> > > In this small patchset, I fix a couple of small omissions as well as introduce
> > > support for "extended command" MSPro transfer method (currently enabled for
> > > JMicron, but I'll clear the TI host for this functionality soon). I expect the
> > > current solution to scale trivially for HG and XC transfer methods (I don't
> > > have the spec for the later yet).
> > Indeed small.
> > All the changes in this patchset are really minor and optional.
> > I could have dropped them.
> > 
> > 
> > I did much more that that.
> > I made the memstick code readable so not only you could understand it
> > I also wrote ms_block.c while adding common support.
> > I made your code use that support.
> > 
> > It is really sad that you are doing that.
> To futher add to that you are just resending my patch #7,
> which is just a minor bugfix, and add evem more complexity to mspro when
> you add CMDEX.
> Why you made it host dependant?, could I ask.
Abd besides CMDEX is really minor thing.
I even initially thought not to add it.


> 
> To be honest the ms subsystem is one of most obscure places I have ever
> seen.
> It took me much more time to write the driver that what it took me to
> write xD driver.
> 
> Alex, I really really hope you will accept my work.
> 
> 
> Best regards,
> 	Maxim Levitsky
>