Re: DMA problem in net2280?

On Thu, 16 Aug 2007, David Brownell wrote:

> On Thursday 16 August 2007, Alan Stern wrote:
> > Dave:
> > 
> > While doing some testing today, I found that net2280 wasn't working
> > right.  When DMA is enabled, the bulk-OUT data buffers presented to
> > g_file_storage contain nothing but 0x6b bytes.  With DMA disabled
> > everything works okay, so this definitely has to be something wrong in
> > the controller driver.
> 
> You surely know that POISON_FREE is 0x6b...
> 
> Doesn't need to be the controller driver.  Something else could
> be making trouble.

Like g_file_storage.  The fact that it works with DMA disabled makes 
this less likely, however.

> > I haven't noticed any recent changes that would cause this, and in fact 
> > I tested net2280 a few weeks ago (can't remember exactly when) with no 
> > problems.  So maybe this was caused by a change elsewhere, like in the 
> > DMA API.  Regardless, it needs to be fixed.
> > 
> > Have you seen this, or can you confirm it?
> 
> A quick check with usbtest/g_zero using the default pattern
> (all zeroes) doesn't fail ... same results with DMA and PIO.
> 
> That's a slightly hacked g_zero, which I'll post in a while,
> but I can't see how that would have affected that.

I didn't see any ill effects with g_zero either.  Can you try 
g_file_storage?

BTW, the same thing happened with both an EHCI and an OHCI host.  This 
was on 2.6.23-rc3 with the gregkh-all patch.

Alan Stern


-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems?  Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >>  http://get.splunk.com/
_______________________________________________
[email protected]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel