Am Dienstag, 23. Februar 2010 22:19:08 schrieb Alan Stern: > to the new sg-list processing, which involved changes to message.c, > ehci-q.c, and usbmon. Maybe the data is getting transmitted correctly > but usbmon is looking in the wrong place. Hi Pete, it seems like usbmon has been broken. This is the trace in 2.6.31.11. The returned sense makes sense ffff8801084c3600 1834878914 S Bo:2:004:1 -115 31 = 55534243 84000000 00020000 80000a28 00000000 ea000001 00000000 000000 ffff8801084c3600 1834879043 C Bo:2:004:1 0 31 > ffff88010e0092c0 1834879213 S Bi:2:004:2 -115 512 < ffff88010e0092c0 1835094446 C Bi:2:004:2 -32 0 ffff8801084c3600 1835094744 S Co:2:004:0 s 02 01 0000 0082 0000 0 ffff8801084c3600 1835094973 C Co:2:004:0 0 0 ffff8801084c3600 1835095124 S Bi:2:004:2 -115 13 < ffff8801084c3600 1835095319 C Bi:2:004:2 0 13 = 55534253 84000000 00020000 01 ffff8801084c3600 1835095448 S Bo:2:004:1 -115 31 = 55534243 85000000 12000000 80000603 00000012 00000000 00000000 000000 ffff8801084c3600 1835095543 C Bo:2:004:1 0 31 > ffff88010e0092c0 1835095620 S Bi:2:004:2 -115 18 < ffff88010e0092c0 1835095667 C Bi:2:004:2 0 18 = f0010600 0000000a 00aa5501 28000100 0000 ffff8801084c3600 1835095731 S Bi:2:004:2 -115 13 < ffff8801084c3600 1835095793 C Bi:2:004:2 0 13 = 55534253 85000000 00000000 00 ffff88010e0092c0 1837914420 S Co:2:001:0 s 23 03 0002 0004 0000 0 ffff88010e0092c0 1837914488 C Co:2:001:0 0 0 ffff88013619b0c0 1840914542 C Ii:2:001:1 -2:2048 0 2.6.33-rc8, same hardware, same test: ffff880132c9c0c0 2502851442 S Bo:2:003:1 -115 31 = 55534243 63000000 00020000 80000a28 00000000 ea000001 00000000 000000 ffff880132c9c0c0 2502851712 C Bo:2:003:1 0 31 > ffff880136636840 2502851826 S Bi:2:003:2 -115 512 < ffff880136636840 2503066840 C Bi:2:003:2 -32 0 ffff880132c9c0c0 2503066896 S Co:2:003:0 s 02 01 0000 0082 0000 0 ffff880132c9c0c0 2503067089 C Co:2:003:0 0 0 ffff880132c9c0c0 2503067130 S Bi:2:003:2 -115 13 < ffff880132c9c0c0 2503067209 C Bi:2:003:2 0 13 = 55534253 63000000 00020000 01 ffff880132c9c0c0 2503067245 S Bo:2:003:1 -115 31 = 55534243 64000000 12000000 80000603 00000012 00000000 00000000 000000 ffff880132c9c0c0 2503067334 C Bo:2:003:1 0 31 > ffff880136636840 2503067383 S Bi:2:003:2 -115 18 < ffff880136636840 2503067585 C Bi:2:003:2 0 18 = 00000000 00000000 00000000 00000000 0000 ffff880132c9c0c0 2503067623 S Bi:2:003:2 -115 13 < ffff880132c9c0c0 2503067709 C Bi:2:003:2 0 13 = 55534253 64000000 00000000 00 ffff880136636840 2505110643 S Co:2:001:0 s 23 03 0002 0003 0000 0 As you can see the sense buffer is zeroed. The SCSI layer still reports that it got ASC 28 ASCQ 00. That makes it very likely that the malfunction is in usbmon and that the sense buffer that actually goes over the wire is identical in both cases. Do you want me to bisect, or do you have any others tests to rather run? Regards Oliver -- 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
