Felipe Balbi wrote: > On Tue, Apr 06, 2010 at 01:46:29PM +0200, ext Gadiyar, Anand wrote: > >With g_zero, do you see a hang or do you have data corruption? > > data corruption. But with our internal mode1 changes, we can use double > buffering with rtl1.8 on omap36xx, so at least rtl1.8 can work. Yup - that's what I'm seeing too. I've given up on 3430, but 3630 seems to be much better behaved. (Which is why I'm surprised at the patch that restricts double buffering to RTL 2.0 and lesser) I'm also investigating without those internal changes - just to find out what's wrong in current code. I did catch one issue on the TX path, and still need to look at RX. Those internal patches are pretty much a rewrite from scratch of the DMA code and transfer handling code. I feel they need more testing and review at least. (Nothing obvious breaks on OMAP3, but none of the other architectures have reported anything. IIRC, only comment I had on OMAP3 was that if DMA programming failed, we were ignoring it). If we get those Mode1 changes you posted tested, reviewed, and queued up, that would be nice. No point debugging old code. - Anand -- To unsubscribe from this list: send the line "unsubscribe linux-usb" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
