From: Bryan O'Donoghue Date: Fri, 26 May 2006 22:04:54 +0100 Subject: [U-Boot-Users] [PATCH 0/14] Various USB device patches Message-ID: <20060526220454.22355181.bodonoghue@codehermit.ie> List-Id: MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: u-boot@lists.denx.de Greetings. The next few posts contain patches to support a number of USB device modifications in Das U-Boot. As per netiquette I'm breaking the patches up into multiple emails. The diffs for the next 14 patches come from top of git today 26/05/06 3663e51eea05583c903e07dd90461fd879d27260 The following patches : - Add USB device support on the mpc8xx. - Add CDC ACM aka the USB standard serial emulation layer. - Add flow control to USB devices with a simple set_nak, unset_nak infrastructure. - Fix some spurious byte ordering assumptions in usbdcore* - Support Switching between gserial and cdc_acm in the usbtty console. - Extended the USB device logic to allow for CDC or Vendor specific Setup Requests. I've included my own setup for include/configs/Adder.h so that people with 8xx boards can use this code easily. The mpc8xx USB device code works perfectly with Linux 2.6.x using either gserial or cdc_acm. Windows includes a driver called usbser.sys which supports CDC ACM devices and I've used it in the past with other devices, however for some reason with the attached code the Windows host after *successfully* enumerating Das U-Boot, sends just *one* BULK IN token and then never polls again, this I believe is a bug with the Windows driver and is commented on here in my blog. More about this here http://codehermit.ie/blog/?postid=10 I have had to make some superficial changes to usbdcore_omap1510 which I haven't tested because I don't have omap hardware. I *did* actually investigate getting an omap development board but a) I couldn't find a 1510 and b) the one I could find on the TI website seemed a bit costly for the sake of implementing the two flow control functions, my thinking is that somebody with omap1510 hardware could possibly test this out and give feedback/implement the flow control functions. Let me know what you think. Best Regards, Bryan