Hi Arvid, On Thu, 2012-06-07 at 13:52 +0000, Arvid Brodin wrote: > Is it correct that the device "sort of" works, just with frequent resets? "Sort of", yes. HID works, but mass storage is pretty flaky with a lot of resets and errors. > If it was an > endian problem, it seems to me that things should not work at all? It seems to me logical too. > Some of the changes made in this patch also affect memory timings. Do you use a > custom-made hardware? If so, you could try to modify the ndelay(90) calls (increase the > value). I already tried that with no luck, but there are more timings involved. The chip is connected to a "WEIM"-Interface (freescale arm). And this interface needs a lots of timing-settings. I suppose these have been wrong since the early days of this custom embedded arm board, but still good enough for the old driver. The isp1760 is connected by PIO not DMA (Pin-DREQ not connected, Pin-DACK to 3V). Do you know of any limitations regarding no DMA? > Also, if your isp176x is not connected as a PCI device, make sure the signal > timings are set up correctly (i.e. if you used a custom avr32 board you would call > smc_set_timing() in your platform device init routine; I'm not sure about the procedure > for this on arm). Thanks a lot for you input, -- Christoph -- 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
